American Journal of Health-System Pharmacy

doi: 10.1093/ajhp/71.19.1606pmid: N/A

Amino acids, electrolytes, dextrose, and lipid injectable emulsion (Kabiven, Fresenius Kabi USA): The hypertonic emulsion, for infusion only into a central vein, is indicated as a source of calories, protein, electrolytes, and essential fatty acids for adults requiring parenteral nutrition when oral intake or enteral nutrition is insufficient, contraindicated, or not possible. Once the components of the three-chamber container are mixed, the resultant liquid has an osmolarity of 1060 mOsm/L. Amino acids, electrolytes, dextrose, and lipid injectable emulsion (Perikabiven, Fresenius Kabi USA): The hypertonic emulsion, for infusion into a peripheral or central vein, is indicated as a source of calories, protein, electrolytes, and essential fatty acids for adults requiring parenteral nutrition when oral intake or enteral nutrition is insufficient, contraindicated, or not possible. Once the components of the three-chamber container are mixed, the resultant liquid has an osmolarity of 750 mOsm/L. Abacavir, dolutegravir, and lamivudine tablets (Triumeq, ViiV Healthcare): The combination product is indicated for the treatment of HIV-1 infection. The product is subject to FDA Medication Guide requirements. Epinephrine injection (no brand name, Belcher Pharmaceuticals): The α- and β-adrenergic agonist is indicated to increase mean arterial blood pressure in adults with hypotension associated with septic shock. Fluticasone furoate inhalation powder (Arnuity Ellipta, GlaxoSmithKline): The corticosteroid is indicated as a prophylactic for the once-daily maintenance treatment of asthma in patients 12 years of age or older. Peginterferon beta-1a injection (Plegridy, Biogen Idec): The interferon beta product, intended for use every 14 days, is indicated for the treatment of patients with relapsing forms of multiple sclerosis. Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Thompson, Cheryl, A.

doi: 10.2146/news140068pmid: 25225441

FDA and Sanofi on August 19 announced the approval of eliglustat, or Cerdelga, for the long-term treatment of adults with type 1 Gaucher disease who have a certain drug-metabolizing-enzyme genotype. Eliglustat, a glucosylceramide analog, was designed to partially inhibit the enzyme that synthesizes glucocerebroside, the company said. Patients with Gaucher disease do not adequately catabolize glucocerebroside, which is a lipid molecule in the cell membrane of the brain, nervous tissue, and other tissues. Insufficient activity by the enzyme responsible for breaking down glucocerebroside, FDA said, causes fatty materials to accumulate in the spleen, liver, and bone marrow. Patients can have spleen and liver enlargement, anemia, low blood platelet counts, and bone problems. The labeling for eliglustat instructs clinicians to select patients with type 1 Gaucher disease for treatment with the drug on the basis of their genotype for cytochrome P-450 (CYP) isoenzyme 2D6. Eliglustat is a substrate of CYP2D6. The drug’s labeling recommends dosages for patients who are extensive, intermediate, or poor metabolizers of CYP2D6 substrates but not dosages for ultrarapid metabolizers or patients whose CYP2D6 genotype cannot be determined. Patients who are identified as extensive or intermediate metabolizers should take 84 mg of eliglustat by mouth twice daily. Poor metabolizers should take 84 mg of eliglustat by mouth once daily. This dosage is based on a pharmacokinetic prediction rather than clinical study results. In addition to being a substrate of CYP2D6, eliglustat is a substrate of CYP3A. The labeling states that eliglustat must not be taken by patients who are extensive or intermediate CYP2D6 metabolizers and using a strong or moderate CYP2D6 inhibitor along with a strong or moderate CYP3A inhibitor. Also, eliglustat must not be taken by patients who are intermediate or poor CYP2D6 metabolizers and using a strong CYP3A inhibitor. The Drug Interactions section of the labeling explains the potential for drugs to affect eliglustat, recommends dosage adjustments, and offers clinical recommendations. The FDA-required Medication Guide for eliglustat tells patients that their use of other medications or herbal supplements may increase the risk of adverse events. Patients are advised to report their use of St. John’s wort and medications that treat fungal infections, tuberculosis, seizures, heart conditions, hypertension, depression, or other mental health problems. According to the labeling, the most common adverse reactions among patients who took eliglustat in either of the two major clinical studies were fatigue, headache, nausea, diarrhea, back pain, pain in an extremity, and upper abdominal pain. Cerdelga will be available in capsules containing 84 mg of eliglustat—actually 100 mg of eliglustat hemitartrate. The capsules are packaged 14 to a blister card, with one or four blister cards to a carton. These should be stored at 20–25 °C. Sanofi said it expects the product “to be available to patients within a month.” Cerdelga is a product of Genzyme, a Sanofi company. Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Traynor,, Kate

doi: 10.2146/news140069pmid: 25225443

FDA is pondering the appropriateness of releasing interim clinical trial data on cardiovascular risks for diabetes medications and, potentially, reconsidering how critical the trials are in the first place. “We did hear calls for whether we need to be reevaluating whether we need to be doing these large outcomes studies for all diabetes agents, going forward,” said John Jenkins, director of FDA’s Office of New Drugs, during an August 11 public hearing. The issue of early data disclosure arose from a 2008 FDA decision to require cardiovascular outcomes studies for investigational diabetes drugs. For statistical validity, sponsors of new drug and biologics license applications for diabetes drugs must design trials of sufficient size and duration to accrue at least 620 adverse cardiovascular events among comparator groups. This FDA guidance was a response to public outcry after a meta-analysis published in 2007 suggested that rosiglitazone increased patients’ risks of major adverse cardiovascular events and death. “Rosiglitazone really set the stage for these cardiovascular outcomes trials,” said Robert Ratner, chief scientific and medical officer of the American Diabetes Association. Notably, FDA recently concluded, on the basis of available data, that rosiglitazone does not pose undue cardiovascular safety concerns. Interim data The intent of the August hearing, according to FDA, was to obtain feedback about the “appropriate handling of interim analysis results” of ongoing clinical trials that assess the cardiovascular risks of diabetes drugs. Of particular concern is whether the release of such data imperils the ability to complete the studies, which may involve thousands of patients and are designed to answer important questions about drug safety. Lisa LaVange, director of FDA’s Office of Biostatistics and the chairwoman of the hearing, said that quickly bringing safe and effective drugs to the market without adversely affecting important clinical research constitutes a “balancing act” for the agency. FDA had laid the groundwork for the meeting with the January 2013 approval, on the basis of interim data from an ongoing clinical trial, of three alogliptin-containing products. The medications are indicated for the improvement of glycemic control in adults with type 2 diabetes, and the relevant data were from the Examination of Cardiovascular Outcomes with Alogliptin versus Standard of Care (EXAMINE) trial. The 5380-patient trial began in 2009, and final data on major adverse cardiac events were collected in April 2013. According to FDA, data from EXAMINE that were unblinded for a planned review before the study’s conclusion supported the preliminary finding that alogliptin improves glycemic control without any “offsetting” cardiovascular or other risks. FDA approved the drug products on the basis of that data, in part, but did not release the customary summary of that data or include the information in the products’ labeling. Instead, the agency released general background information about the trial, a summary indicating that the interim results were in line with FDA guidance on evaluating cardiovascular risks, and additional data unrelated to those risks. A March 2013 FDA memo emphasized that approval decisions are normally based on data from completed clinical trials, which was not the case for alogliptin. The memo stated that the approval of alogliptin thus raised “novel issues” about the release of data from unfinished studies but acknowledged that similar situations, though rare, may arise in the future. Big numbers Cleveland Clinic cardiologist Steven Nissen, the lead author of the 2007 rosiglitazone meta-analysis, told hearing participants that 16 cardiovascular outcomes studies for diabetes drugs were initiated during the first three years after FDA’s 2008 decision. He said the planned recruitment for the studies exceeds 110,000 patients, and data from the studies, once they are completed, will provide major benefits to society. “This is a terribly important issue,” Nissen said. But Ratner noted that improvements in the care of patients with diabetes have caused a decline in cardiovascular event rates over the past 20 years, which makes it difficult to complete the type of study FDA requires. Ratner, like most presenters at the hearing, was leery of public disclosure of interim data because of the potential to further complicate the ability to complete clinical trials. University of Washington biostatistician Tom Fleming recalled that a confidential analysis of interim data from a study involving HIV-infected patients revealed 39 cases of death or progression to AIDS in patients treated with zalcitabine, compared with 19 such events in didanosine recipients. But by the end of the study a year later, adverse events had equalized between the groups. Although the interim adverse event data were within the study’s prespecified threshold, the data could well have alarmed clinicians, Fleming said. “If the early results were released, it would have been very misleading,” Fleming said. “There was essentially no chance that [the study] would have been completed.” FDA’s 2008 guidance on diabetes drugs allows for the possibility of early data disclosure to occur after accrual of 122 adverse cardiovascular events among clinical trial participants. This figure represents about 20% of the total events needed for statistical validity. FDA allows drug sponsors to apply for product approval at this point during the study if the hazard ratio for the investigational diabetes drug does not exceed a predetermined threshold. If a drug sponsor files for approval using the interim safety data, FDA requires that the trial be continued to completion—with accrual of 620 adverse events—into the postmarketing period. Alternatively, drug sponsors may conduct a new cardiovascular outcomes study to obtain the required data. But several speakers at the hearing indicated that the latter option is not practical. Nissen said it’s critically important to avoid the disclosure of interim hazard ratios and other data that may prompt clinicians and patients to alter their medication decisions on the basis of statistical interpretations that may not hold up over the long term. “I believe that the agency is acting wisely,” Nissen said regarding the data-disclosure model FDA used for approving alogliptin. “It’s a tremendous benefit to society to have clarity about what these drugs do and what they don’t do,” he said. Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Traynor,, Kate

doi: 10.2146/news140070pmid: 25225444

The Centers for Medicare and Medicaid Services (CMS) on July 18 eased a weeks-old policy that had sent hospice programs scrambling to secure Medicare Part D coverage for medically necessary medications not included in the hospice benefit. “We are all kind of breathing a sigh of relief” after the July policy revision, said Pamela S. Moore, clinical lead pharmacist in pain and palliative care at Summa Health System in Akron, Ohio. The original policy, which went into effect May 1, 2014, had instructed Part D plan sponsors to subject all drug claims for hospice patients to prior-authorization requirements. “We have over 200 patients on our home hospice service, so it had a pretty significant impact,” Moore said. “It was very difficult to maintain continuity of those patients’ regimens.” Moore said each day at the hospice program began with “multiple physicians, multiple nurses, myself, hospice administrators, all in a room just going through the new admissions” to determine how to bill for each patient’s medications. She said patients with Part D coverage needed prior authorization to continue receiving medications for diabetes, dementia, and other common conditions. “It would take hours of the office staff time to get the forms filled out,” Moore said. “Each Part D company wanted something different, and they didn’t have a uniform way of accepting the information.” The original CMS policy was a response to instances in which Medicare Part D plans might have billed the agency for medications that should be covered under the Medicare Part A hospice benefit. This problem was described in a June 2012 report from the Health and Human Services department’s Office of Inspector General. The report had concluded that Part D plans “paid for prescription analgesic, antinausea, laxative, and antianxiety drugs that likely should have been covered under the per diem payments made to hospice organizations.” The July policy revision asked Part D plans to limit prior authorizations for hospice patients to those four drug categories. CMS acknowledged the “operational challenges” in the original policy and said it “created difficulties for Part D sponsors and hospice providers, and in some cases, barriers to access for beneficiaries.” The revised policy has an October 1 effective date, but CMS asked Part D plans to implement it as soon as possible. CMS requires hospice programs to cover the cost of drugs for conditions related to each patient’s terminal condition. Medications unrelated to that diagnosis may be covered by Medicare Part D or another third-party payer. But CMS has not precisely defined what the agency means by a “terminal condition” or “related conditions,” said Jason Kimbrel. Kimbrel is director of the HospiScript–Ohio State University College of Pharmacy postgraduate year 2 residency in hospice and palliative care and vice president of clinical services for HospiScript, a hospice-focused unit of the pharmacy benefits management company Catamaran. In the absence of clear guidance from CMS, Kimbrel said, decisions about a medication’s relation to the illness for which the patient is receiving hospice services are made by each hospice organization’s medical team, and there is no standardized approach to the process. And, Kimbrel explained, “there are legitimate medications that are unrelated to the terminal prognosis but are still medically necessary,” like eye drops for patients with glaucoma. “A patient could have glaucoma and they could have six months to live, and you don’t want them to have increased intraocular pressure and have vision issues,” Kimbrel said. “But there’s nothing about their glaucoma that’s going to cause them to pass away in the next six months. So it’s not a life-limiting illness at that point, but it’s medically necessary to still be managed.” Kimbrel said such medications should be eligible for Part D coverage. But when the prior-authorization policy first went into effect, he said, “those patients were unable to obtain their meds” from their Part D plans. For diabetes medications, identifying the correct payer can be difficult, Kimbrel said. If a patient is in hospice care because of Parkinson’s disease but also has well-controlled diabetes, the diabetes may not be contributing to the dying process. In this case, diabetes drugs could appropriately be covered under Medicare Part D, Kimbrel said. But Parkinson’s disease can make it difficult for patients to eat or swallow, which may affect glucose control in patients with diabetes. “If you say hospice’s definitions of ‘related conditions’ are ‘any symptom that the patient needs to manage at the last six months of their life,’ well, then I would say hospice should cover the drug, because you would have to manage that person’s diabetes,” Kimbrel said. A rare but potentially difficult coverage decision could involve a patient who takes an analgesic for the treatment of a chronic condition, such as pain from a long-ago traffic accident. If that patient requires hospice care for a condition that causes little or no pain, Part D could conceivably cover the medication under a prior authorization, Kimbrel said. CMS in May solicited comments on the definitions of “terminal illness” and “related conditions” for future rule-making. The agency stated that it will consider those comments and the effects the definitions may have on hospice services. Kimbrel said the initial policy on prior authorization “made sense, to some degree, on paper, but it didn’t necessarily actually come full circle on process.” He called the subsequent limiting of prior authorization “an interim solution” pending additional guidance from the agency about medication coverage determinations. “The intent is to limit patient harm and to ensure point-of-sale rejections don’t get in the way of patient care,” Kimbrel said. Moore said Part D plans’ policies on prior authorization have been “hit or miss” since CMS revised its policy. “Some of these Part D companies have reverted back. Some of them are still requiring prior authorizations for things; they haven’t quite gotten on board,” she said in mid-August. “So we’ve got a number of scenarios happening now, and we’re never sure which one we’re going to encounter” or when coverage will be denied. Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Thompson, Cheryl, A.

doi: 10.2146/news140071pmid: 25225445

Janet L. Teeters, M.S., director of the ASHP Accreditation Services Division, died on August 15 after an acute illness. She was 58 years old. Open in new tabDownload slide Janet L. Teeters Open in new tabDownload slide Janet L. Teeters To hundreds of residency program directors and preceptors, Teeters was the person who answered questions during the annual Town Hall on Sunday afternoon at the ASHP Midyear Clinical Meeting. To thousands of pharmacy students, she was the greeter at the Residency Showcase, answering questions and helping to manage the eager, perhaps anxiety-ridden, crowd. Throughout the year, Teeters was a facilitator for people who understood the value of residency training for pharmacists and accreditation of those training programs. “Janet’s commitment to helping pharmacists prepare for their patient care roles has had a significant impact on the lives and careers of thousands of practitioners,” said ASHP Chief Executive Officer Paul W. Abramowitz. “Her contributions were instrumental to the tremendous growth of ASHP’s accreditation programs and services, and her influence, both professionally and personally, will be felt for many years to come.” Open in new tabDownload slide Janet Lee Hamm Teeters, an avid horsewoman, after a first-level dressage competition. Photo printed with permission of Diane Katt. Open in new tabDownload slide Janet Lee Hamm Teeters, an avid horsewoman, after a first-level dressage competition. Photo printed with permission of Diane Katt. Vice President Janet A. Silvester, of the Accreditation Services Office, said Teeters “had a sincere dedication to excellence in her professional endeavors—from residency and technician training to her role on the Council on Credentialing in Pharmacy. Her significant impact will long be remembered.” The number of residency training positions in the United States “more than tripled in her 12 years at ASHP,” said Vice President Douglas J. Scheckelhoff, Teeters’ supervisor for most of that time. Scheckelhoff credited Teeters’ approach to the accreditation process for helping to fuel that growth. Her approach, he said, reflected her philosophy regarding accreditation visits: “We’re there to help people be successful in training residents.” Visits by the division’s lead surveyors combined evaluation against ASHP’s standard with consultative advice on how to improve a site’s residency program and expand its capacity. Teeters, ever mindful of others’ needs, also worked toward fulfilling participants’ desire for ASHP to streamline its accreditation standards and the processes for evaluating residents and applying for positions, Scheckelhoff said. Thoughtfulness, generosity of time, and optimism were traits of Teeters for years before she worked at ASHP, recalled Debby Gwozdz Bryniarski. “She also had great empathy…. She went into every situation with an open mind and the idea that she could somehow learn from it,” said Bryniarski, who was a pharmacy resident when Teeters joined the pharmacy department at Lutheran General Hospital in suburban Chicago in 1989. Bryniarski, now director of pharmacy and respiratory services at Advocate Lutheran General Hospital, said she tries to instill those same qualities in her own residents. She described the years with Teeters at the helm of pharmacy services for the large community hospital as “very exciting.” Teeters saw the benefit of deploying automation and pharmacy technicians in drug distribution to expand clinical pharmacy services. She also wanted pharmacists to be in close proximity to patients. The hospital started using an automated dispensing system in Teeters’ first year as the pharmacy director, Bryniarski said. In 1993, the hospital opened a warfarin clinic that, at the time, was the largest one in the area run by pharmacists who met face-to-face with patients. Last year while in the Chicago area, Teeters saw what the pharmacy had accomplished since her departure. The patient tower that had been built in 2009 has office spaces for the decentralized pharmacists, providing structural support for them to perform all of their functions in the vicinity of patients and other health professionals. “This is really like a very long-term dream” come true for her, Bryniarski said. Looking back on their medical mission to Guatemala, Bryniarski called Teeters “sort of fearless,” particularly in travel. “She was willing to just jump into anything. And I think that was true for her professional life as well.” Teeters was someone who, in offering career guidance, truly looked out for the best interest of the person seeking counsel, recalled Sarah C. Erush, now pharmacy clinical manager and residency director at Children’s Hospital of Philadelphia. In the late 1980s, Erush faced a choice in pursuing a Pharm.D. degree. She spoke with Teeters, who held a leadership role at Erush’s first-choice college. Teeters set aside the college’s interest and advised Erush to pursue the path that would better fuel her academic growth and development. For Erush, this exposure to Teeters’ selfless nature proved valuable in the short term and established a principle to which she has adhered for some 25 years: Always have the resident’s best interest at heart. Teeters served as the 1995–96 president of the Illinois Council of Health-System Pharmacists and won its Pharmacist of the Year Award in 1998. She was a member of the organization’s Technician Certification Commission in the years leading up to creation of the Pharmacy Technician Certification Board. In 1993–99, she represented Illinois on the executive committee for the Great Lakes Pharmacy Resident Conference. Teeters received her B.S. degree in pharmacy from the University of Wisconsin in Madison. On graduation, she worked in Eau Claire and then Chippewa Falls. Four years after graduation, she left Wisconsin to enter graduate school at the University of Minnesota in Minneapolis and the ASHP-accredited general pharmacy residency program at University of Minnesota Hospital and Clinics. Teeters then worked at the Veterans Administration Medical Center in Philadelphia, New England Medical Center in Boston, and Lutheran General Hospital in Park Ridge, Illinois. In 1990, the year after arriving at Lutheran General, she was promoted to director of pharmacy. She rose to director of operations for pharmacy and rehabilitation services in 1997. By then, Lutheran General was part of Advocate Health Care, an integrated health system. In 1999, she became director of pharmacy for Lutheran General and all of Advocate Health Care. She served as the director for the pharmacy practice residency program at Lutheran General for the entire 13 years she worked there. Teeters is survived by her husband of 30 years, Joseph, of Laurel, Maryland, and sister Diane Katt. Contributions in Teeters’ memory may be made to the ASHP Foundation Pharmacy Residency Expansion Grant Fund. Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

doi: 10.1093/ajhp/71.19.1611pmid: N/A

Mallinckrodt plc acquired Questcor Pharmaceuticals Inc. in August. Questcor is now known as the autoimmune and rare diseases business in Mallinckrodt’s specialty pharmaceuticals segment. ASHP Chief Executive Officer Paul W. Abramowitz, Pharm.D., Sc.D. (Hon), FASHP, visited Indianapolis on August 17–19. He visited four Indianapolis health systems: Indiana University Health (IU Health), Eskenazi Health, St. Vincent Health System, and Community Health Network. He met with pharmacy leadership, staff, and residents. Dr. Abramowitz also gave a presentation at IU Health, titled “Transforming Patient Care: Paramount Issues and Opportunities in Pharmacy Practice.” Open in new tabDownload slide Open in new tabDownload slide Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Christensen,, Traci

doi: 10.2146/ajhp130630pmid: 25225446

Assessing and documenting the competence of clinical pharmacists is required by the Joint Commission in order for a hospital to receive accreditation.1 Currently there is no standard for best practices for competency program development and assessment. The purpose of this article is to describe a process for designing a clinical pharmacist competency program that meets the requirements and recommendations of various agencies and organizations; this process may be used as a model and customized for implementation at other facilities. The Joint Commission has standards to ensure that all hospital staff members have the skills and knowledge necessary to perform their job duties. Human resources standard HR.01.05.03 stipulates that “staff participate in ongoing education and training”1; element of performance (EP) 1 further states that staff are required to participate in education for the purpose of maintaining or increasing their competency. EP 5 states that education should be specific to the needs of the population of patients served by the hospital. EP 6 states that education should include skills such as “team communication, collaboration and coordination of care.” The Joint Commission further clarifies, in HR.01.06.01, that competency should be assessed and documented, at minimum, once every three years.1 An American College of Clinical Pharmacy (ACCP) white paper on clinical pharmacist competencies contains a few more specific recommendations.2 According to ACCP, competency programs should include clinical problem solving, judgment, and decision making; communication and education; medical information evaluation and management; management of patient populations; and therapeutic knowledge. The American Society of Health-System Pharmacists (ASHP) publishes a book containing competency modules that may be used and customized by hospitals as a competency program.3 The book addresses many topics, such as fire safety, infection control, renal dosing, patient counseling, and medication safety, in addition to competencies directed toward specific patient population age groups. However, this book is not updated frequently; the last edition was published in 2008. One of the challenges of developing any competency program is ensuring that educational materials are up-to-date and applicable. Methods of assessment There are three basic methods for assessing competence: simulations, cognitive tests, and direct observation.3 Cognitive tests are the most commonly used method of assessing competency and usually include multiple-choice tests and quizzes. This type of competency assessment is useful for assuring competency when tasks or patient cases similar to those encountered in actual practice are presented to the practitioner, but it is not a good predictor of real-life practice because patient-specific factors may influence a practitioner’s decision making.3 Simulations can provide a good indication of how a clinician might handle similar patient care situations, but, as in clinical practice, there may be more than one correct course of action, and it is difficult to simulate complex situations.3 Direct observation is an excellent way to assess how a pharmacist practices in real-life situations but is difficult due to limitations on the number of observations that may be performed. There is a considerable time commitment involved in observing each pharmacist, and the data may be subjective unless a good objective assessment tool is used. ASHP recommends direct observation as the preferred method for assessing competence.3 Due to the strengths and weaknesses of the different methods of competency assessment, it is best to use a combination of methods to assess overall competency. This article describes the competency assessment methods in use at a specific facility and how each satisfies the requirements of the various organizations involved in ensuring competency. Background and program development Utah Valley Regional Medical Center (UVRMC) is a 400-bed community hospital in Provo, Utah. UVRMC is the second largest hospital in the Intermountain Healthcare system of 22 hospitals. The pharmacy staff consists of 32 decentralized pharmacists practicing within a mixed clinical model in a variety of settings, including ambulatory care, acute and critical care (adult and pediatric), infectious diseases, behavioral medicine, and emergency medicine. In addition, there are 10 pharmacists working mostly in the central pharmacy in the areas of i.v. compounding, automation, and distribution. In terms of experience levels, the pharmacists at UVRMC are diverse, ranging from recent graduates with no postgraduate training to postgraduate year 1 and 2 residency–trained pharmacists to individuals with more than 30 years of hospital pharmacy experience. The responsibility of ensuring competence currently lies with the clinical pharmacy manager. In addition to the competency program, UVRMC has separate quarterly basic education requirements and a weekly professional development program. The material covered through basic education includes subjects necessary for clinical practitioners in the hospital setting, such as Health Insurance Portability and Accountability Act compliance, fire safety, quarterly policy updates, and signs of abuse in patients, which satisfy all the requirements set by the Joint Commission. The weekly professional development program includes accredited continuing-education topics presented by the hospital’s clinical pharmacists, topics to help prepare for board certification exams, and preceptor development topics. The pharmacists are encouraged to attend these weekly meetings, but they are not required to do so. Additionally, all pharmacists are required to maintain certification in basic life support. On development of the current competency program, it was decided that the basic education requirements were sufficient for ensuring compliance with Joint Commission standards but were not sufficient for assessing or developing the clinical abilities or skills of the pharmacists. The weekly professional and preceptor development meetings help to develop clinical knowledge, but since these meetings are optional, it was decided that they were not a reliable means to ensure competence. The decision was made to develop required monthly competencies for all pharmacists employed by the hospital. Competency program assignments are distributed to the pharmacists on the first day of each month and are to be completed by the end of each month. The pharmacy administrative assistant checks answers to tests and quizzes and records the completion of each pharmacist’s assignments on a spreadsheet. A schedule outlining the curriculum for the competency program was developed based on the needs of the hospital and recommendations from ACCP and the Joint Commission. The competencies covered were of six broad types: clinical area–specific, operational and regulatory, direct observation, drug information (new drugs), collaborative practice agreements, and remediation. Each competency program component is described below. Clinical area–specific competencies Assignments in this area are made during the first month of each new quarter, and they are designed to allow pharmacists to choose an activity to enhance their knowledge in their current area of clinical practice. Since clinical pharmacists practice in a wide variety of settings, providing content that is applicable to each of them requires flexibility. The pharmacists can choose to read a case chosen from a bank of disease-specific case reports and then answer questions, or review Board of Pharmaceutical Specialties preparatory material and answer questions based on the module, or prepare a written review of a current journal article. Other options for demonstration of competency may be approved at the discretion of the clinical manager; examples of activities that have been approved in this category include advanced cardiac life support (ACLS) or pediatric advanced life support (PALS) training, attendance at a conference related to the individual’s area of practice, and media-fill testing for i.v. room pharmacists. These practice area–specific clinical competencies satisfy Joint Commission requirements to provide patient age-specific competencies to pharmacists. They also satisfy ACCP competency requirements pertaining to clinical problem solving, judgment, and decision making as well as knowledge of therapeutics. Cognitive tests are used to evaluate these competencies; in addition, simulation methods are used to ensure ACLS, PALS, and media-fill testing competencies. Operational and regulatory competencies This portion of the competency program was designed to convey information important to pharmacists from a regulatory standpoint, to convey information about a process change at the facility, or to help improve efficiency. Over the past three years, programs in this area have focused on many different topics, such as the Joint Commission’s Surgical Care Improvement Project, value-based purchasing initiatives and what a clinical pharmacist can do to ensure compliance, the medication reconciliation process at the facility, risk evaluation and mitigation strategy processes, Joint Commission survey preparation, pharmacy computer system reports and shortcuts, and policy changes with an impact on pharmacy. These programs, usually consisting of a short PowerPoint (Microsoft Corporation, Redmond, WA) presentation or reading assignments followed by a multiple-choice quiz, are often developed by the clinical pharmacy manager, pharmacists on staff, or students or residents on rotation at UVRMC. These competencies ensure that all pharmacists are knowledgeable of laws, regulations, and hospital policies related to pharmacy practice; ensure that pharmacists are prepared for Joint Commission surveys; and help to promote efficiency in using the pharmacy software system. Direct observation of competency ASHP recommends using direct observation to assess competence.3 The direct-observation portion of the UVRMC competency program was designed to evaluate the pharmacists using a predetermined list of criteria while allowing for customization based on practice area. Annual competency assessment in this area is accomplished by having our pharmacist team leads spend two to three hours observing the pharmacists in their respective areas of practice and evaluating them on several criteria. The team leads are clinical pharmacists who also supervise 7–10 clinical pharmacists in a given service line, such as critical care, acute care, and ambulatory care. The team leads assist with annual goal setting and performance evaluations for the members of their teams. The pharmacist team leads are, in turn, observed and evaluated by the pharmacy clinical manager. The criteria for direct observation are created based on the pharmacist’s practice area and the actual workflow for the day on which observations are to be conducted. For instance, if a clinical pharmacist and a student are rounding with physicians and entering medication orders during the evaluation period, they might be evaluated on items such as precepting skills, the appropriateness and thoroughness of prerounding diagnostic workups, their interaction with other members of the healthcare team (including physicians and nurses), the accuracy of order entry, and their ability to triage priority orders. If the clinical pharmacist is working in an ambulatory care clinic accompanied by a student, the pair might be evaluated on interaction with the team at the clinic, recommendations made to the patient or prescribers, and skill in communicating with patients. In this manner, the evaluation is customized to the practice area of the clinical pharmacist. Observed activities and actions are scored on a four-point scale (0 = no, or inadequate, performance, 1 = minimum performance, 2 = proficient in all aspects, 3 = exceeds expectations). These scores have not been used to compare the pharmacists with one another; rather, the intent is to assess each pharmacist’s progression relative to scores from previous years. This method of competency assessment has been used for the past two years; when a larger body of data has been compiled, it will be evaluated to assess the degree of individual pharmacist improvement. Direct observation of pharmacist competency satisfies Joint Commission requirements for team communication and collaboration and coordination of care, in addition to satisfying all the pertinent requirements of ACCP. New drug competencies Competency assessment activities in this category were developed to ensure that the pharmacists are aware of newly approved medications through the distribution of monographs and educational modules conveying information on the key properties and characteristics, formulary status, and potential place in therapy of select new medications approved by the Food and Drug Administration in the previous six months. After reading the materials provided, the pharmacists are required to complete a short quiz assessing their knowledge of the new medications. Information on new drug approvals, new indications, and complete drug monographs are supplied by Thomson Reuters through its Micromedex P&T QUIK Report system (Micromedex products are now the property of Truven Health Analytics, Greenwood Village, CO). The data from these reports are reviewed biannually, and a learning module on each new drug selected is developed; the list of new medications selected for inclusion is determined according to predicted levels of use in the facility’s patient population. This portion of the competency assessment program is usually prepared by the pharmacy clinical manager, a clinical pharmacist, or a pharmacy student on rotation at the facility. The drug information component of the competency program satisfies the ACCP recommendations regarding the assessment of pharmacist competence in the areas of medical information evaluation and management and knowledge of therapeutics. Collaborative practice agreement competencies Competency assessment activities in this area include live presentations, distribution of reading materials based on current clinical practice guidelines, and a written assessment involving simulated patient care scenarios accompanied by multiple-choice questions. UVRMC pharmacists currently participate in collaborative practice agreements for anticoagulation therapy and vancomycin dosing. In order to meet the requirements set forth in the collaborative practice agreements, the pharmacy department educates its clinical pharmacists about the latest guidelines regarding anticoagulation and vancomycin dosing and monitoring on an annual basis. This education consists of attending a live presentation, where updates to the guidelines and general therapeutic knowledge are presented. The presentations are recorded so that staff unable to attend a live session may watch them at a later date. In addition, the pharmacists are required to complete reading assignments on the subjects of anticoagulation and vancomycin dosing and monitoring before attending the presentations. After the presentations, the pharmacists are required to pass a written assessment on the reading material and the presentation material in order to be eligible to participate in a collaborative practice agreement. This portion of the competency program satisfies the collaborative practice agreement requirements to provide ongoing education, in addition to satisfying the ACCP requirements on management of patient populations and therapeutic knowledge, as well as the ACCP requirements pertaining to clinical problem solving, judgment, and decision making. Remediation Since it is imperative that all pharmacists complete all scheduled competency assessments throughout the year, and it is also imperative that the pharmacists master all the educational material provided in all applicable competency assignments, it was decided that one month a year be set aside for remediation purposes. The month set aside for remediation is the month just prior to annual evaluations, so that the pharmacists can address any missing assignments and, for already completed assignments, review the material and correct any incorrect answers on previous competency quizzes. Pharmacists at UVRMC receive annual evaluations based on their job performance and adherence to company values. In order for pharmacists to receive an evaluation indicating that the expectations of his or her job have been met, all competency program assignments for the year must be completed; to help ensure an optimal score on the evaluation, all assignments must be completed on time. Conclusion Designing and implementing a competency program for clinical pharmacists can be a daunting process, with several organizations making only broad recommendations and very little specific guidance available in the literature. A successful program should incorporate various methods for assessing competence on a regular basis, be consistently applied, allow for customization by area of practice, be timely, be flexible, and allow for changes as different needs arise. Footnotes The Management Consultation column gives readers an opportunity to obtain advice on common management problems from pharmacists practicing in health systems. The assistance of ASHP’s Section of Pharmacy Practice Managers and its Advisory Group on Manager Development in soliciting Management Consultation submissions is acknowledged. Unsolicited submissions are also welcome. Readers are invited to submit topics for this column to [email protected] or ASHP c/o David Chen, Director, Pharmacy Practice Sections, 7272 Wisconsin Avenue, Bethesda, MD 20814 ([email protected]). The author has declared no potential conflicts of interest. References 1 Joint Commission E-dition . Performance improvement . https://e-dition.jcrinc.com/MainContent.aspx (accessed 2013 Aug 13). 2 Burke JM Miller WA Spencer AP et al. . ACCP white paper: clinical pharmacist competencies . Pharmacotherapy . 2008 ; 28 : 806 – 15 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Murdaugh LB . Designing and managing a competence assessment program . In: Competence assessment tools for health-system pharmacies . 4th ed . Bethesda, MD : Oxford University Press ; 2008 : 13 – 6 . Google Preview WorldCat COPAC Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Awad, Nadia, I.;Bridgeman, Mary, Barna

doi: 10.2146/ajhp130446pmid: 25225447

Continuing education (CE) is an essential component of professional development in addition to being a necessary requirement for pharmacy licensure renewal. This professional development education serves to strengthen competencies and skills and mandates that practitioners keep abreast of evolving trends and practices within the profession. Regardless of their practice setting, new practitioners are often tasked with the planning and presentation of CE programming for pharmacy staff development. In addition, pharmacy residents completing their postgraduate training are often expected to attend and provide CE programming as part of residency program requirements or activities. New practitioners can provide several types of CE programs through various venues. Moreover, the role of the new practitioner may include not only the delivery but also the development of the CE program. Live CE programs may be presented to pharmacy staff members employed at the institution at which the new practitioner is employed. In addition, a new practitioner may be invited to participate in presenting CE programs through the office of CE at a pharmacy school or at a local chapter meeting of his or her state affiliate society of health-system pharmacists. CE programs may also be presented in the form of a webinar regarding a common disease state or controversial topic in the practice of pharmacy. Written CE programs and monographs can reach clinicians throughout the country, depending on the platform selected for presentation. Regardless of the form of delivery of the CE program and the audience, new practitioners provided with an opportunity to present such a program may be unsure of where to begin. When tasked with planning a program, an essential first step is to conduct a needs assessment. A needs assessment can broadly identify the educational needs of the audience; it can also be useful after a CE program is implemented in gauging the fundamental knowledge of the audience on a specific topic to assess improvement and, thus, the effectiveness of the program. The purposes of conducting a needs assessment are to determine the educational needs of the learner (in the scope of pharmacy CE, the learner may be a pharmacist or a pharmacy technician) and to assure that program content is appropriate for the learner’s educational level and applicable to his or her daily practice. According to the Accreditation Council for Pharmacy Education (ACPE), a needs assessment should be systematic and multifaceted and incorporate formal, objective procedures to assure relevance, balance, and use of the best available evidence.1 The following step-by-step approach can be used by new practitioners in order to conduct a needs assessment. Step 1: Audience identification Is the CE program intended for pharmacists, pharmacy technicians, administrative personnel, students, or a mixed group? It is important to note that learning needs will vary based on the intended audience for a program. If the audience consists of pharmacists, pharmacy technicians, or both, the individual planning the program may need to ascertain the status of the organization or institution with regard to CE provider (i.e., the agency from which CE accreditation is desired) to determine what is needed to provide CE credit for the program. If the program is to be accredited through ACPE, the individual tasked with coordinating the program should work with the CE provider to obtain the necessary information to acquire ACPE approval and a universal activity number so that credit can be distributed to participants. According to ACPE requirements for programs seeking accreditation, if the audience consists of both pharmacists and pharmacy technicians, two sets of learning objectives based on the two groups of learners’ respective professional competencies (as they relate to the selected topic) must be created.1 Step 2: Evaluating the needs of the audience Thinking about the best way to connect with the intended CE program audience to identify its specific needs represents the second step in program planning. Audience feedback on professional education needs can be gathered through the use of multiple mechanisms, including the evaluation of proven needs (as established by the professional literature, evidence-based guidelines, or epidemiologic data, for example), inferred staff needs (e.g., needs that arise pursuant to the incorporation of new protocols, technology, facilities, and regulatory or legal changes), and needs elicited through staff feedback mechanisms such as verbalized requests for specific education and requests conveyed via informal staff surveys.2,3 Some questions to consider include the following: Has there been a problem related to insufficient education or training? Is there a new service area, a new drug product, or a new trend in pharmacy practice that needs to be addressed? Conducting the needs assessment and evaluating specific areas of audience need will get the new practitioner off to a good start in developing a quality program. Step 3: What does the literature say? A literature review may be one aspect of conducting a needs assessment and ensuring that the selected educational topic is relevant and in congruence with the latest information available. Identifying the most current guideline recommendations and expert consensus opinions on disease management topics or other subjects of interest is a critical step in ensuring timely topic selection and accurate information. Step 4: Narrowing the scope of the program All aspects of a broad topic (e.g., congestive heart failure) cannot be covered in a single-session, one- or two-hour CE program; a multiple-session program may be required to address such a topic. In thinking about time constraints and what is feasible for a particular audience or practice setting, the CE planner should consider what information can specifically be conveyed to the audience in the allotted time. For instance, if the topic chosen is congestive heart failure, the program can be narrowed to focus on newly released guidelines for the treatment of acute decompensated heart failure, with a discussion regarding how these new recommendations are being incorporated into the pertinent institutional protocols. Enumerating the specific goals and objectives for learning about a selected topic can help to prioritize CE activities within the allotted time. One institution’s experience At Robert Wood Johnson University Hospital, a tertiary academic medical center located in New Brunswick, New Jersey, staff pharmacists, postgraduate trainees in the pharmacy practice and specialty residency programs, and faculty members affiliated with the Ernest Mario School of Pharmacy provide CE programs that are accredited through the New Jersey Board of Pharmacy to the pharmacy department staff on a routine basis. Although not a nationally accredited provider of continuing pharmacy education, the department conducted a formal educational needs assessment survey in 2012 to help with CE program planning, development of content, and communication among potential presenters to ensure that all CE programming is planned around key areas of educational interest to the staff. A survey was developed and distributed electronically to 152 pharmacists and pharmacy technicians employed in the pharmacy department. To assess professional and scholarly needs, a five-point Likert scale was used to determine the level of interest for approximately 7–10 topics in each of the following areas: critical care, internal medicine, infectious diseases, and emergency medicine. Survey respondents were also asked to identify additional topics not addressed in the survey for consideration as potential CE programs. Out of the 152 invitations distributed, a total of 70 surveys (46.1%) were completed; 44 (62.9%) were completed by pharmacists, and 26 (37.1%) were completed by pharmacy technicians. The top five areas of interest for CE programs, as determined by survey responses from both pharmacists and pharmacy technicians, along with the respective percentages of each group indicating they were “very interested,” were as follows: sepsis and targeted antimicrobial therapy (60.5% and 52.0%), common infections in the critically ill (55.8% and 68.0%), new drugs of abuse (53.5% and 54.2%), institutional multidrug resistance (51.2% and 45.8%), and fluid and electrolyte replacement therapy (44.2% and 48.0%). From these topics, CE programs were developed in the areas identified by staff as being of high interest that were within the expertise and interest of the residents and faculty involved in program planning. The information gathered in the needs assessment survey is discussed annually at the first meeting of the institution’s residency advisory council, when CE program planning is introduced, and is also used in planning upcoming staff education programs. Reassessment of staff educational needs on a biennial basis is planned. We recommend that new practitioners tasked with planning CE activities consider using survey methodology to evaluate audience educational preferences and topics for review. Providing limited topics of current-event and institution-specific interest helps eliminate personal educational preferences and favors identification of group educational needs. Closing notes A formal needs assessment can be conducted by new practitioners to improve the process of planning pharmacy-related CE activities and to ensure that the educational needs of the audience are met in terms of topic selection and covered information. The methodology used to conduct a needs assessment can be revisited periodically and refined to determine emerging topics of CE program interest among pharmacists and pharmacy technicians in accordance with changing and current practice trends. Footnotes The New Practitioners Forum column features articles that address the special professional needs of pharmacists early in their careers as they transition from students to practitioners. Authors include new practitioners or others with expertise in a topic of interest to new practitioners. AJHP readers are invited to submit topics or articles for this column to the New Practitioners Forum, c/o Jill Haug, 7272 Wisconsin Avenue, Bethesda, MD 20814 (301-664-8821 or [email protected]). At the time the needs assessment described in this article was developed and conducted, Dr. Awad was Postgraduate Year 2 Emergency Medicine Pharmacy Resident, Ernest Mario School of Pharmacy, Rutgers, The State University of New Jersey, and Robert Wood Johnson University Hospital. The authors have declared no potential conflicts of interest. References 1 Accreditation Council for Pharmacy Education . Accreditation standards for continuing pharmacy education. Version 2 ( March 2014 ). www.acpe-accredit.org/pdf/CPE_Standards_Final.pdf (accessed 2014 Jul 15). 2 Ratnapalan S Hilliard RI . Needs assessment in postgraduate medical education: a review . http://med-ed-online.net/index.php/meo/article/view/4542/4722 (accessed 2013 Jul 15). 3 Grant J . Learning needs assessment: assessing the need . BMJ . 2002 ; 324 : 156 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Halilovic,, Jenana;Heintz, Brett, H.

doi: 10.2146/ajhp140031pmid: 25225448

Abstract Purpose Published evidence regarding the influence of cirrhosis on the clinical pharmacokinetics of antibacterial agents is reviewed; dosing recommendations and a decision algorithm are provided. Summary A systematic PubMed search (1960–2013) was conducted to identify literature pertaining to the use of antibacterials with hepatobiliary clearance in adult patients with cirrhosis. Clinical drug databases, conference abstracts, and package inserts were also reviewed for pertinent information. Twenty-two antibiotics that undergo hepatic or mixed renal–hepatobiliary clearance were identified. Overall, published pharmacokinetic data to guide antibiotic dosing in adults with cirrhosis are sparse, and many relevant studies were conducted before wide adoption of the Child–Pugh method for classifying the severity of cirrhosis. Dose adjustments should be considered in the setting of decompensated liver disease, particularly with antibiotics that undergo phase I metabolism, have high protein binding, or are associated with a high frequency of hepatotoxicity or other concentration-dependent toxicities. Individualization of dosing regimens should take into account a number of variables: the intent of therapy (treatment versus prophylaxis); the duration of therapy; the site and severity of infection; the degree of organ dysfunction, as indicated by Child–Pugh class; the patient’s immune status, weight, and fluid status; and the pharmacokinetic and pharmacodynamic properties of the antibacterial agents under consideration. Conclusion Cirrhosis has multiple effects on the disposition of a wide range of antibacterial agents. Appropriate antibiotic therapy selection and individualized dosing can contribute to optimal clinical outcomes while decreasing the risk of hepatotoxicity. Bacterial infection is one of the most frequent complications in patients with cirrhosis, occurring in an estimated 34–38% of those requiring hospitalization.1 The high risk of infection in cirrhosis is associated with multiple immune defects as well as increased bacterial translocation. Further, infection increases morbidity, mortality, and hospital length of stay and may precipitate other complications such as hepatic encephalopathy.2,3 It is estimated that 30% of patients with cirrhosis who have a bacterial infection will die within one month, while another 30% will die within one year. The most common types of bacterial infection in patients with cirrhosis are spontaneous bacterial peritonitis, urinary tract infections, pneumonia, and bacteremia.1 Appropriate antibacterial therapy is essential to ensure favorable clinical outcomes in patients with cirrhosis. However, hepatic impairment, especially chronic impairment, may directly or indirectly decrease protein binding, metabolism, and renal elimination of antibiotics.4,–8 Several pathophysiologic processes contribute to the resultant changes in drug disposition, including reduced hepatic blood flow, impaired biliary clearance, ascites, hypoalbuminemia, and loss of cytochrome metabolic activity. Collectively, the altered pharmacokinetic parameters may increase serum drug concentrations and increase the risk of drug toxicity.4,–8 The clinician must carefully assess the implication of the alterations to determine whether or not a dose adjustment is needed. The scarcity of published literature on the pharmacokinetics of antibiotics in hepatic impairment makes it particularly challenging for the clinician to determine the most appropriate dose for a patient with cirrhosis. In fact, among 251 new molecular entities approved by the Food and Drug Administration over the 10-year period 1998–2007, dosing recommendations applicable in the setting of hepatic impairment were available for only 4.9 Further, many pharmacokinetic studies that guide dosing of antibiotics in patients with hepatic impairment were published before wide adoption of the Child–Pugh scoring system, included a small number of patients, employed limited sampling strategies, and used nonrigorous analytical methods that did not control for confounding variables.10 Given the limited information available, caution should be exercised in the prescribing of hepatically cleared agents in the setting of cirrhosis. This review explores alterations in the pharmacokinetics of commonly used antibiotics that undergo hepatobiliary clearance in the context of hepatic disease. In addition, a decision algorithm for dosing antibiotics in the context of cirrhosis is provided to assist clinicians when no pertinent pharmacokinetic data are available. Data sources A systematic PubMed search was conducted to identify all English-language literature published from January 1960 to October 2013 that would be considered useful in formulating dosing recommendations for antibacterials that undergo hepatobiliary clearance and are used in adult patients with cirrhosis. All human studies (including case series and case reports) that evaluated the clinical pharmacokinetics of systemically available (i.e., i.v. or oral) antibacterials and antibacterial drug classes that undergo either hepatic or mixed renal–hepatic clearance in adults with cirrhosis were reviewed. Search terms included antibacterials, antibacterial agents, antibiotics, antimicrobials, anti-infective agents, hepatitis, hepatotoxicity, hepatic or liver impairment, hepatic or liver failure, hepatic or liver dysfunction, cirrhosis, pharmacokinetics, and dose adjustments. Clinical drug databases, conference abstracts, and package inserts were also reviewed. Studies evaluating acute hepatitis, pediatric populations, or critically ill patients were excluded from review. Further, antiparasitic, antifungal, antiretroviral, and antimycobacterial agents (except rifamycins, isoniazid, and pyrazinamide) were considered to be beyond the scope of this review and were excluded. The following drug classes were found to include antibacterials with hepatic or mixed renal–hepatobiliary clearance: β-lactams, fluoroquinolones, macrolides, oxazolidinones, rifamycins, and tetracyclines and glycylcyclines; in addition, a number of specific individual agents (clindamycin, metronidazole, nitrofurantoin, pyrazinamide, quinupristin–dalfopristin, and sulfamethoxazole–trimethoprim) exhibit these clearance characteristics. We identified 26 original research articles and abstracts describing pharmacokinetic data on specific antibacterials with hepatic or mixed renal–hepatobiliary clearance in patients with cirrhosis. The referenced antibacterials included (in alphabetical order): azithromycin, cefotaxime, ceftriaxone, ciprofloxacin, clarithromycin, clindamycin, erythromycin, isoniazid, linezolid, metronidazole, moxifloxacin, nafcillin, nitrofurantoin, pyrazinamide, quinupristin–dalfopristin, rifabutin, rifampin, and tigecycline. Data synthesis All drug classes containing antibacterials that undergo either primarily hepatobiliary or mixed renal–hepatobiliary clearance were included in the data synthesis. If pharmacokinetic data on the use of an individual agent in patients with cirrhosis were available, the data are discussed below and summarized in Tables 1 and 2. If pharmacokinetic data on an individual agent specifically pertaining to patients with cirrhosis were not available but the general pharmacokinetics clearly indicated that the agent undergoes either hepatobiliary or mixed renal–hepatobiliary clearance, guidance on the need for dosing adjustment is provided. Table 2 Summary of Dosing and Pharmacokinetic Data From Reviewed Clinical Studiesa,b Antibiotic Patients Dose and Frequency Mean ± S.D.b Cmax (mg/L) V (L) t½ (hr) CL (mL/min) AUC (mg · hr/L) Azithromycin34 Child–Pugh class A (n = 10) 500 mg × 1 dose 0.39 ± 0.25 1787.4 ± 270.9 60.6 ± 19.2 608.0 ± 191.1 4.8 ± 2.0 Child–Pugh class B (n = 6) 500 mg × 1 dose 0.5 ± 0.6 1731.6 ± 362.1 68.1 ± 13.2 643.7 ± 157.0 4.0 ± 2.0 Healthy controls (n = 6) 500 mg × 1 dose 0.29 ± 0.1 1783.0 ± 329.1 53.5 ± 7.1 773.1 ± 203.7 4.9 ± 2.4 Cefotaxime16,17  Reference 16 Child–Pugh class B (n = 4) or C (n = 4) 1 g i.v. × 1 dose …c 0.87 ± 0.4 4.8 ± 3.0 185 ± 88 105.9 ± 53.7 Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.49 ± 0.2 1.6 ± 0.6 291 ± 90 70.4 ± 28.8  Reference 17 Cirrhosis (n = 11) 2 g i.v. × 1 dose … … … 1.4 ± 0.7 … Cirrhosis (n = 7) 2 g i.v. × 1 dose … … … 1.3 ± 0.6 … Ceftriaxone16,19  Reference 19 Cirrhosis with ascites (n = 6) 1 g i.v. × 1 dose … 0.23 ± 0.077d 9.7 ± 1.8 0.39 ± 0.16e … Cirrhosis without ascites (n = 4) 1 g i.v. × 1 dose … 0.12 ± 0.035d 8.0 ± 2.0 0.23 ± 0.11e … Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.142 ± 0.017d 8.4 ± 1.8 0.23 ± 0.064e …  Reference 16 Child–Pugh class B (n = 5) or C (n = 3) 1 g i.v. × 1 dose 1 g i.v. × 1 … 0.23 ± 0.06 10.7 ± 4.3 20.7 ± 9.3 Healthy controls (n = 8) 1 g i.v. × 1 dose … … 0.13 ± 0.06 9.0 ± 1.7 13.1 ± 4.8 Ciprofloxacin24,25  Reference 24 Cirrhosis (n = 7) 750 mg p.o. q 12 hr × 5 days 3.71 ± 0.81 2.83 ± 0.59f 3.47 ± 1.23 765.1 ± 233.4 18.4 ± 7.1 Healthy controls (n = 7) 750 mg p.o. q 12 hr × 5 days 3.49 ± 0.68 3.48 ± 1.0f 3.71 ± 0.41 840.2 ± 240.0 16.2 ± 4.6  Reference 25 Child–Pugh class B (n = 20) 500 mg p.o. × 1 dose 2.6 ± 1.3 … 3.2 ± 1.8 … 21.9 ± 4.5 Healthy controls (n = 10) 500 mg p.o. × 1 dose 2.6 ± 0.6 … 3.6 ± 1.2 … 19.3 ± 3.8 Clarithromycin33 Child–Pugh class B or C (n = 7) 250 mg p.o. q 12 hr × 5 doses 1.22 ± 0.82 305 ± 211f 5.0g 713 ± 593 9.29 ± 6.2h Healthy controls (n = 6) 250 mg p.o. q 12 hr × 5 doses 1.52 ± 0.35 138 ± 19f 3.3g 486 ± 166 9.29 ± 2.6h Clindamycin49,50  Reference 49 Chronic hepatitis (n = 6) 300 mg i.v. q 12 hr × 4 doses … … 2.1g … … Alcoholic cirrhosis (n = 9) 300 mg i.v. q 12 hr × 4 doses … … 2.5g … … Healthy controls (n = 8) 300 mg i.v. q 12 hr × 4 doses … … 1.8g … …  Reference 50 Alcoholic cirrhosis (n = 7) 300 mg i.v. × 1 dose … 0.693 ± 0.258d 4.5 ± 0.9 121.7 ± 23.6 … Healthy controls (n = 7) 300 mg i.v. × 1 dose … 0.583 ± 0.18d 3.4 ± 0.4 159.2 ± 39.6 … Erythromycin30,–32  Reference 31 Alcoholic cirrhosis (n = 8) 500 mg p.o. × 1 dose 2.04 ± 1.13 93.5 ± 88.9f 4.5g … 11.6 ± 8.9i Healthy controls (n = 6) 500 mg p.o. × 1 dose 1.5 ± 0.91 98.9 ± 34.2f 6.6g … 9.0 ± 5.9i  Reference 32 Alcoholic cirrhosis (n = 6) 500 mg p.o. × 1 dose … … 3.2 ± 0.4 … … Healthy controls (n = 6) 500 mg p.o. × 1 dose … … 2.0 ± 0.7 … …  Reference 30 Alcoholic cirrhosis (n = 5) 500 mg i.v. × 1 dose … 85.5 ± 23.8 2.24 ± 0.89 403.3 ± 138.3j … Healthy controls (n = 6) 500 mg i.v. × 1 dose … 57.6 ± 14.8 1.36 ± 0.43 570.0 ± 201.7j … Isoniazid41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 6.7g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 3.2g … … Linezolid36 Child–Pugh class A or B (n = 7) 600 mg p.o. × 1 dose 11.5 ± 2.0 0.563 ± 0.058d 6.8 ± 3.1 1.12 ± 0.46e 128 ± 60 Healthy controls (n = 8) 600 mg p.o. × 1 dose 11.9 ± 9.8 0.618 ± 0.132d 5.4 ± 1.6 1.39 ± 0.35e 97 ± 31 Metronidazole54,–56  Reference 55 Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3  Reference 56 Child–Pugh class A (n = 14) 500 mg i.v. × 1 dose … 0.74 ± 0.11d 10.7 ± 2.3 0.85 ± 0.26e 124.9 ± 42.3i Child–Pugh class B (n = 9) 500 mg i.v. × 1 dose … 0.79 ± 0.12d 13.5 ± 5.1 0.79 ± 0.36e 124.4 ± 25.8i Child–Pugh class C (n = 12) 500 mg i.v. × 1 dose … 0.81 ± 0.14d 21.5 ± 12.7 0.56 ± 0.28e 174.1 ± 52.0i Healthy controls (n = 7) 500 mg i.v. × 1 dose … 0.80 ± 0.32d 7.4 ± 2.2 1.53 ± 0.37e 81.4 ± 27.0i Cirrhosis with coma (n = 8) 500 mg i.v. × 1 dose … 44 ± 9 20 ± 9 29 ± 10 …  Reference 54 Healthy controls (n = 8) 500 mg i.v. × 1 dose … 48 ± 7 7.3 ± 0.9 83 ± 14 … Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3 Moxifloxacin27 Child–Pugh class C (n = 9) 400 mg i.v. q 24 hr × 3 doses … … 10.4 (8.5–16.0)k 146.7 (106.7–175.0)k 45.5 (38.1–62.2)k Nafcillin13 Alcoholic cirrhosis (n = 12) 500 mg i.v. × 1 dose … 19.9 ± 5.8 1.2 ± 0.3 291.5 ± 147.6 … Healthy controls (n = 12) 500 mg i.v. × 1 dose … 27.1 ± 7.3 1.0 ± 0.2 583.7 ± 144.2 … Pyrazinamide60 Cirrhosis (n = 10) 19.3 ± 0.6 mg/kg × 1 dose 30.1 ± 5.4 0.6 ± 0.1d 15.1 ± 3.3 0.48 ± 0.15e 280 Healthy controls (n = 9) 19.3 ± 0.6 mg/kg × 1 dose 35.6 ± 5.4 0.66 ± 0.05d 9.2 ± 1.8 0.84 ± 0.2e 725 ± 219 Quinupristin (Q)–dalfopristin (D)61,62 Child–Pugh class A or B (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 3.16g (Q) 7.34g (D) … … 0.91g (Q) 0.61g (D) … … 3.50g,i (Q) 7.36g,i (D) Healthy controls (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 2.72g (Q) 7.24g (D) … … 0.91g (Q) 0.45g (D) … … 2.9g,i (Q) 7.36 g,i (D) Rifabutin43 Child–Pugh class A (n = 18) 300 mg p.o. × 1 dose … … 16 ± 11 … 4.39 ± 1.16 Child–Pugh class B (n = 6) 300 mg p.o. × 1 dose … … 11 ± 10 … 3.45 ± 1.12 Child–Pugh class C (n = 4) 300 mg p.o. × 1 dose … … 42 ± 18 … 9.43 ± 2.73 Healthy controls (n = 12) 300 mg p.o. × 1 dose … … 67 ± 45 … 8.85 ± 4.40 Rifampin41,42  Reference 41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 5.4g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 2.8g … …  Reference 42 Cirrhosis (n = 26) 4–10 mg/kg twice weekly × 4 doses … … 3.0–6.8g … … Healthy controls 8–16 mg/kg twice weekly × 4 doses … … 1.9–2.7g … … Tigecycline47 Child–Pugh class A (n = 10) 100 mg i.v. × 1 dose 0.86 ± 0.38 617 ± 234 19.1 ± 5.4 520.0 ± 231.7 3.8 ± 1.8 Child–Pugh class B (n = 10) 100 mg i.v. × 1 dose 0.91 ± 0.55 542 ± 246 23.0 ± 5.0 368.3 ± 155.0 5.6 ± 3.4 Child–Pugh class C (n = 5) 100 mg i.v. × 1 dose 1.21 ± 0.41 378 ± 107 26.8 ± 6.1 225.0 ± 45.0 7.6 ± 1.5 Healthy controls (n = 23) 100 mg i.v. × 1 dose 0.98 ± 0.54 524 ± 157 18.7 ± 7.2 496.7 ± 188.3 3.7 ± 1.3 Antibiotic Patients Dose and Frequency Mean ± S.D.b Cmax (mg/L) V (L) t½ (hr) CL (mL/min) AUC (mg · hr/L) Azithromycin34 Child–Pugh class A (n = 10) 500 mg × 1 dose 0.39 ± 0.25 1787.4 ± 270.9 60.6 ± 19.2 608.0 ± 191.1 4.8 ± 2.0 Child–Pugh class B (n = 6) 500 mg × 1 dose 0.5 ± 0.6 1731.6 ± 362.1 68.1 ± 13.2 643.7 ± 157.0 4.0 ± 2.0 Healthy controls (n = 6) 500 mg × 1 dose 0.29 ± 0.1 1783.0 ± 329.1 53.5 ± 7.1 773.1 ± 203.7 4.9 ± 2.4 Cefotaxime16,17  Reference 16 Child–Pugh class B (n = 4) or C (n = 4) 1 g i.v. × 1 dose …c 0.87 ± 0.4 4.8 ± 3.0 185 ± 88 105.9 ± 53.7 Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.49 ± 0.2 1.6 ± 0.6 291 ± 90 70.4 ± 28.8  Reference 17 Cirrhosis (n = 11) 2 g i.v. × 1 dose … … … 1.4 ± 0.7 … Cirrhosis (n = 7) 2 g i.v. × 1 dose … … … 1.3 ± 0.6 … Ceftriaxone16,19  Reference 19 Cirrhosis with ascites (n = 6) 1 g i.v. × 1 dose … 0.23 ± 0.077d 9.7 ± 1.8 0.39 ± 0.16e … Cirrhosis without ascites (n = 4) 1 g i.v. × 1 dose … 0.12 ± 0.035d 8.0 ± 2.0 0.23 ± 0.11e … Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.142 ± 0.017d 8.4 ± 1.8 0.23 ± 0.064e …  Reference 16 Child–Pugh class B (n = 5) or C (n = 3) 1 g i.v. × 1 dose 1 g i.v. × 1 … 0.23 ± 0.06 10.7 ± 4.3 20.7 ± 9.3 Healthy controls (n = 8) 1 g i.v. × 1 dose … … 0.13 ± 0.06 9.0 ± 1.7 13.1 ± 4.8 Ciprofloxacin24,25  Reference 24 Cirrhosis (n = 7) 750 mg p.o. q 12 hr × 5 days 3.71 ± 0.81 2.83 ± 0.59f 3.47 ± 1.23 765.1 ± 233.4 18.4 ± 7.1 Healthy controls (n = 7) 750 mg p.o. q 12 hr × 5 days 3.49 ± 0.68 3.48 ± 1.0f 3.71 ± 0.41 840.2 ± 240.0 16.2 ± 4.6  Reference 25 Child–Pugh class B (n = 20) 500 mg p.o. × 1 dose 2.6 ± 1.3 … 3.2 ± 1.8 … 21.9 ± 4.5 Healthy controls (n = 10) 500 mg p.o. × 1 dose 2.6 ± 0.6 … 3.6 ± 1.2 … 19.3 ± 3.8 Clarithromycin33 Child–Pugh class B or C (n = 7) 250 mg p.o. q 12 hr × 5 doses 1.22 ± 0.82 305 ± 211f 5.0g 713 ± 593 9.29 ± 6.2h Healthy controls (n = 6) 250 mg p.o. q 12 hr × 5 doses 1.52 ± 0.35 138 ± 19f 3.3g 486 ± 166 9.29 ± 2.6h Clindamycin49,50  Reference 49 Chronic hepatitis (n = 6) 300 mg i.v. q 12 hr × 4 doses … … 2.1g … … Alcoholic cirrhosis (n = 9) 300 mg i.v. q 12 hr × 4 doses … … 2.5g … … Healthy controls (n = 8) 300 mg i.v. q 12 hr × 4 doses … … 1.8g … …  Reference 50 Alcoholic cirrhosis (n = 7) 300 mg i.v. × 1 dose … 0.693 ± 0.258d 4.5 ± 0.9 121.7 ± 23.6 … Healthy controls (n = 7) 300 mg i.v. × 1 dose … 0.583 ± 0.18d 3.4 ± 0.4 159.2 ± 39.6 … Erythromycin30,–32  Reference 31 Alcoholic cirrhosis (n = 8) 500 mg p.o. × 1 dose 2.04 ± 1.13 93.5 ± 88.9f 4.5g … 11.6 ± 8.9i Healthy controls (n = 6) 500 mg p.o. × 1 dose 1.5 ± 0.91 98.9 ± 34.2f 6.6g … 9.0 ± 5.9i  Reference 32 Alcoholic cirrhosis (n = 6) 500 mg p.o. × 1 dose … … 3.2 ± 0.4 … … Healthy controls (n = 6) 500 mg p.o. × 1 dose … … 2.0 ± 0.7 … …  Reference 30 Alcoholic cirrhosis (n = 5) 500 mg i.v. × 1 dose … 85.5 ± 23.8 2.24 ± 0.89 403.3 ± 138.3j … Healthy controls (n = 6) 500 mg i.v. × 1 dose … 57.6 ± 14.8 1.36 ± 0.43 570.0 ± 201.7j … Isoniazid41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 6.7g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 3.2g … … Linezolid36 Child–Pugh class A or B (n = 7) 600 mg p.o. × 1 dose 11.5 ± 2.0 0.563 ± 0.058d 6.8 ± 3.1 1.12 ± 0.46e 128 ± 60 Healthy controls (n = 8) 600 mg p.o. × 1 dose 11.9 ± 9.8 0.618 ± 0.132d 5.4 ± 1.6 1.39 ± 0.35e 97 ± 31 Metronidazole54,–56  Reference 55 Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3  Reference 56 Child–Pugh class A (n = 14) 500 mg i.v. × 1 dose … 0.74 ± 0.11d 10.7 ± 2.3 0.85 ± 0.26e 124.9 ± 42.3i Child–Pugh class B (n = 9) 500 mg i.v. × 1 dose … 0.79 ± 0.12d 13.5 ± 5.1 0.79 ± 0.36e 124.4 ± 25.8i Child–Pugh class C (n = 12) 500 mg i.v. × 1 dose … 0.81 ± 0.14d 21.5 ± 12.7 0.56 ± 0.28e 174.1 ± 52.0i Healthy controls (n = 7) 500 mg i.v. × 1 dose … 0.80 ± 0.32d 7.4 ± 2.2 1.53 ± 0.37e 81.4 ± 27.0i Cirrhosis with coma (n = 8) 500 mg i.v. × 1 dose … 44 ± 9 20 ± 9 29 ± 10 …  Reference 54 Healthy controls (n = 8) 500 mg i.v. × 1 dose … 48 ± 7 7.3 ± 0.9 83 ± 14 … Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3 Moxifloxacin27 Child–Pugh class C (n = 9) 400 mg i.v. q 24 hr × 3 doses … … 10.4 (8.5–16.0)k 146.7 (106.7–175.0)k 45.5 (38.1–62.2)k Nafcillin13 Alcoholic cirrhosis (n = 12) 500 mg i.v. × 1 dose … 19.9 ± 5.8 1.2 ± 0.3 291.5 ± 147.6 … Healthy controls (n = 12) 500 mg i.v. × 1 dose … 27.1 ± 7.3 1.0 ± 0.2 583.7 ± 144.2 … Pyrazinamide60 Cirrhosis (n = 10) 19.3 ± 0.6 mg/kg × 1 dose 30.1 ± 5.4 0.6 ± 0.1d 15.1 ± 3.3 0.48 ± 0.15e 280 Healthy controls (n = 9) 19.3 ± 0.6 mg/kg × 1 dose 35.6 ± 5.4 0.66 ± 0.05d 9.2 ± 1.8 0.84 ± 0.2e 725 ± 219 Quinupristin (Q)–dalfopristin (D)61,62 Child–Pugh class A or B (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 3.16g (Q) 7.34g (D) … … 0.91g (Q) 0.61g (D) … … 3.50g,i (Q) 7.36g,i (D) Healthy controls (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 2.72g (Q) 7.24g (D) … … 0.91g (Q) 0.45g (D) … … 2.9g,i (Q) 7.36 g,i (D) Rifabutin43 Child–Pugh class A (n = 18) 300 mg p.o. × 1 dose … … 16 ± 11 … 4.39 ± 1.16 Child–Pugh class B (n = 6) 300 mg p.o. × 1 dose … … 11 ± 10 … 3.45 ± 1.12 Child–Pugh class C (n = 4) 300 mg p.o. × 1 dose … … 42 ± 18 … 9.43 ± 2.73 Healthy controls (n = 12) 300 mg p.o. × 1 dose … … 67 ± 45 … 8.85 ± 4.40 Rifampin41,42  Reference 41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 5.4g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 2.8g … …  Reference 42 Cirrhosis (n = 26) 4–10 mg/kg twice weekly × 4 doses … … 3.0–6.8g … … Healthy controls 8–16 mg/kg twice weekly × 4 doses … … 1.9–2.7g … … Tigecycline47 Child–Pugh class A (n = 10) 100 mg i.v. × 1 dose 0.86 ± 0.38 617 ± 234 19.1 ± 5.4 520.0 ± 231.7 3.8 ± 1.8 Child–Pugh class B (n = 10) 100 mg i.v. × 1 dose 0.91 ± 0.55 542 ± 246 23.0 ± 5.0 368.3 ± 155.0 5.6 ± 3.4 Child–Pugh class C (n = 5) 100 mg i.v. × 1 dose 1.21 ± 0.41 378 ± 107 26.8 ± 6.1 225.0 ± 45.0 7.6 ± 1.5 Healthy controls (n = 23) 100 mg i.v. × 1 dose 0.98 ± 0.54 524 ± 157 18.7 ± 7.2 496.7 ± 188.3 3.7 ± 1.3 a Cmax= maximum concentration (plasma or serum), V = volume of distribution, t½ = half-life (terminal phase), CL = total body clearance, AUC = area under the concentration–time curve (from time zero to infinity unless otherwise noted). b Some values are expressed here in units of measure different from those originally reported. c Not reported. d Expressed as L/kg. e Expressed as mL/min/kg. f Expressed as V/F (ratio of volume of distribution to bioavailability). g S.D. not reported. h AUC from time zero to 12 hours. i AUC from time zero to 24 hours. j A more significant difference in clearance of unbound erythromycin was observed in patients with alcoholic cirrhosis versus healthy controls (mean ± S.D., 42.2 ± 10.11 L/hr versus 113.2 ± 44.2 L/hr). k Reported as median (range). Open in new tab Table 2 Summary of Dosing and Pharmacokinetic Data From Reviewed Clinical Studiesa,b Antibiotic Patients Dose and Frequency Mean ± S.D.b Cmax (mg/L) V (L) t½ (hr) CL (mL/min) AUC (mg · hr/L) Azithromycin34 Child–Pugh class A (n = 10) 500 mg × 1 dose 0.39 ± 0.25 1787.4 ± 270.9 60.6 ± 19.2 608.0 ± 191.1 4.8 ± 2.0 Child–Pugh class B (n = 6) 500 mg × 1 dose 0.5 ± 0.6 1731.6 ± 362.1 68.1 ± 13.2 643.7 ± 157.0 4.0 ± 2.0 Healthy controls (n = 6) 500 mg × 1 dose 0.29 ± 0.1 1783.0 ± 329.1 53.5 ± 7.1 773.1 ± 203.7 4.9 ± 2.4 Cefotaxime16,17  Reference 16 Child–Pugh class B (n = 4) or C (n = 4) 1 g i.v. × 1 dose …c 0.87 ± 0.4 4.8 ± 3.0 185 ± 88 105.9 ± 53.7 Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.49 ± 0.2 1.6 ± 0.6 291 ± 90 70.4 ± 28.8  Reference 17 Cirrhosis (n = 11) 2 g i.v. × 1 dose … … … 1.4 ± 0.7 … Cirrhosis (n = 7) 2 g i.v. × 1 dose … … … 1.3 ± 0.6 … Ceftriaxone16,19  Reference 19 Cirrhosis with ascites (n = 6) 1 g i.v. × 1 dose … 0.23 ± 0.077d 9.7 ± 1.8 0.39 ± 0.16e … Cirrhosis without ascites (n = 4) 1 g i.v. × 1 dose … 0.12 ± 0.035d 8.0 ± 2.0 0.23 ± 0.11e … Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.142 ± 0.017d 8.4 ± 1.8 0.23 ± 0.064e …  Reference 16 Child–Pugh class B (n = 5) or C (n = 3) 1 g i.v. × 1 dose 1 g i.v. × 1 … 0.23 ± 0.06 10.7 ± 4.3 20.7 ± 9.3 Healthy controls (n = 8) 1 g i.v. × 1 dose … … 0.13 ± 0.06 9.0 ± 1.7 13.1 ± 4.8 Ciprofloxacin24,25  Reference 24 Cirrhosis (n = 7) 750 mg p.o. q 12 hr × 5 days 3.71 ± 0.81 2.83 ± 0.59f 3.47 ± 1.23 765.1 ± 233.4 18.4 ± 7.1 Healthy controls (n = 7) 750 mg p.o. q 12 hr × 5 days 3.49 ± 0.68 3.48 ± 1.0f 3.71 ± 0.41 840.2 ± 240.0 16.2 ± 4.6  Reference 25 Child–Pugh class B (n = 20) 500 mg p.o. × 1 dose 2.6 ± 1.3 … 3.2 ± 1.8 … 21.9 ± 4.5 Healthy controls (n = 10) 500 mg p.o. × 1 dose 2.6 ± 0.6 … 3.6 ± 1.2 … 19.3 ± 3.8 Clarithromycin33 Child–Pugh class B or C (n = 7) 250 mg p.o. q 12 hr × 5 doses 1.22 ± 0.82 305 ± 211f 5.0g 713 ± 593 9.29 ± 6.2h Healthy controls (n = 6) 250 mg p.o. q 12 hr × 5 doses 1.52 ± 0.35 138 ± 19f 3.3g 486 ± 166 9.29 ± 2.6h Clindamycin49,50  Reference 49 Chronic hepatitis (n = 6) 300 mg i.v. q 12 hr × 4 doses … … 2.1g … … Alcoholic cirrhosis (n = 9) 300 mg i.v. q 12 hr × 4 doses … … 2.5g … … Healthy controls (n = 8) 300 mg i.v. q 12 hr × 4 doses … … 1.8g … …  Reference 50 Alcoholic cirrhosis (n = 7) 300 mg i.v. × 1 dose … 0.693 ± 0.258d 4.5 ± 0.9 121.7 ± 23.6 … Healthy controls (n = 7) 300 mg i.v. × 1 dose … 0.583 ± 0.18d 3.4 ± 0.4 159.2 ± 39.6 … Erythromycin30,–32  Reference 31 Alcoholic cirrhosis (n = 8) 500 mg p.o. × 1 dose 2.04 ± 1.13 93.5 ± 88.9f 4.5g … 11.6 ± 8.9i Healthy controls (n = 6) 500 mg p.o. × 1 dose 1.5 ± 0.91 98.9 ± 34.2f 6.6g … 9.0 ± 5.9i  Reference 32 Alcoholic cirrhosis (n = 6) 500 mg p.o. × 1 dose … … 3.2 ± 0.4 … … Healthy controls (n = 6) 500 mg p.o. × 1 dose … … 2.0 ± 0.7 … …  Reference 30 Alcoholic cirrhosis (n = 5) 500 mg i.v. × 1 dose … 85.5 ± 23.8 2.24 ± 0.89 403.3 ± 138.3j … Healthy controls (n = 6) 500 mg i.v. × 1 dose … 57.6 ± 14.8 1.36 ± 0.43 570.0 ± 201.7j … Isoniazid41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 6.7g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 3.2g … … Linezolid36 Child–Pugh class A or B (n = 7) 600 mg p.o. × 1 dose 11.5 ± 2.0 0.563 ± 0.058d 6.8 ± 3.1 1.12 ± 0.46e 128 ± 60 Healthy controls (n = 8) 600 mg p.o. × 1 dose 11.9 ± 9.8 0.618 ± 0.132d 5.4 ± 1.6 1.39 ± 0.35e 97 ± 31 Metronidazole54,–56  Reference 55 Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3  Reference 56 Child–Pugh class A (n = 14) 500 mg i.v. × 1 dose … 0.74 ± 0.11d 10.7 ± 2.3 0.85 ± 0.26e 124.9 ± 42.3i Child–Pugh class B (n = 9) 500 mg i.v. × 1 dose … 0.79 ± 0.12d 13.5 ± 5.1 0.79 ± 0.36e 124.4 ± 25.8i Child–Pugh class C (n = 12) 500 mg i.v. × 1 dose … 0.81 ± 0.14d 21.5 ± 12.7 0.56 ± 0.28e 174.1 ± 52.0i Healthy controls (n = 7) 500 mg i.v. × 1 dose … 0.80 ± 0.32d 7.4 ± 2.2 1.53 ± 0.37e 81.4 ± 27.0i Cirrhosis with coma (n = 8) 500 mg i.v. × 1 dose … 44 ± 9 20 ± 9 29 ± 10 …  Reference 54 Healthy controls (n = 8) 500 mg i.v. × 1 dose … 48 ± 7 7.3 ± 0.9 83 ± 14 … Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3 Moxifloxacin27 Child–Pugh class C (n = 9) 400 mg i.v. q 24 hr × 3 doses … … 10.4 (8.5–16.0)k 146.7 (106.7–175.0)k 45.5 (38.1–62.2)k Nafcillin13 Alcoholic cirrhosis (n = 12) 500 mg i.v. × 1 dose … 19.9 ± 5.8 1.2 ± 0.3 291.5 ± 147.6 … Healthy controls (n = 12) 500 mg i.v. × 1 dose … 27.1 ± 7.3 1.0 ± 0.2 583.7 ± 144.2 … Pyrazinamide60 Cirrhosis (n = 10) 19.3 ± 0.6 mg/kg × 1 dose 30.1 ± 5.4 0.6 ± 0.1d 15.1 ± 3.3 0.48 ± 0.15e 280 Healthy controls (n = 9) 19.3 ± 0.6 mg/kg × 1 dose 35.6 ± 5.4 0.66 ± 0.05d 9.2 ± 1.8 0.84 ± 0.2e 725 ± 219 Quinupristin (Q)–dalfopristin (D)61,62 Child–Pugh class A or B (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 3.16g (Q) 7.34g (D) … … 0.91g (Q) 0.61g (D) … … 3.50g,i (Q) 7.36g,i (D) Healthy controls (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 2.72g (Q) 7.24g (D) … … 0.91g (Q) 0.45g (D) … … 2.9g,i (Q) 7.36 g,i (D) Rifabutin43 Child–Pugh class A (n = 18) 300 mg p.o. × 1 dose … … 16 ± 11 … 4.39 ± 1.16 Child–Pugh class B (n = 6) 300 mg p.o. × 1 dose … … 11 ± 10 … 3.45 ± 1.12 Child–Pugh class C (n = 4) 300 mg p.o. × 1 dose … … 42 ± 18 … 9.43 ± 2.73 Healthy controls (n = 12) 300 mg p.o. × 1 dose … … 67 ± 45 … 8.85 ± 4.40 Rifampin41,42  Reference 41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 5.4g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 2.8g … …  Reference 42 Cirrhosis (n = 26) 4–10 mg/kg twice weekly × 4 doses … … 3.0–6.8g … … Healthy controls 8–16 mg/kg twice weekly × 4 doses … … 1.9–2.7g … … Tigecycline47 Child–Pugh class A (n = 10) 100 mg i.v. × 1 dose 0.86 ± 0.38 617 ± 234 19.1 ± 5.4 520.0 ± 231.7 3.8 ± 1.8 Child–Pugh class B (n = 10) 100 mg i.v. × 1 dose 0.91 ± 0.55 542 ± 246 23.0 ± 5.0 368.3 ± 155.0 5.6 ± 3.4 Child–Pugh class C (n = 5) 100 mg i.v. × 1 dose 1.21 ± 0.41 378 ± 107 26.8 ± 6.1 225.0 ± 45.0 7.6 ± 1.5 Healthy controls (n = 23) 100 mg i.v. × 1 dose 0.98 ± 0.54 524 ± 157 18.7 ± 7.2 496.7 ± 188.3 3.7 ± 1.3 Antibiotic Patients Dose and Frequency Mean ± S.D.b Cmax (mg/L) V (L) t½ (hr) CL (mL/min) AUC (mg · hr/L) Azithromycin34 Child–Pugh class A (n = 10) 500 mg × 1 dose 0.39 ± 0.25 1787.4 ± 270.9 60.6 ± 19.2 608.0 ± 191.1 4.8 ± 2.0 Child–Pugh class B (n = 6) 500 mg × 1 dose 0.5 ± 0.6 1731.6 ± 362.1 68.1 ± 13.2 643.7 ± 157.0 4.0 ± 2.0 Healthy controls (n = 6) 500 mg × 1 dose 0.29 ± 0.1 1783.0 ± 329.1 53.5 ± 7.1 773.1 ± 203.7 4.9 ± 2.4 Cefotaxime16,17  Reference 16 Child–Pugh class B (n = 4) or C (n = 4) 1 g i.v. × 1 dose …c 0.87 ± 0.4 4.8 ± 3.0 185 ± 88 105.9 ± 53.7 Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.49 ± 0.2 1.6 ± 0.6 291 ± 90 70.4 ± 28.8  Reference 17 Cirrhosis (n = 11) 2 g i.v. × 1 dose … … … 1.4 ± 0.7 … Cirrhosis (n = 7) 2 g i.v. × 1 dose … … … 1.3 ± 0.6 … Ceftriaxone16,19  Reference 19 Cirrhosis with ascites (n = 6) 1 g i.v. × 1 dose … 0.23 ± 0.077d 9.7 ± 1.8 0.39 ± 0.16e … Cirrhosis without ascites (n = 4) 1 g i.v. × 1 dose … 0.12 ± 0.035d 8.0 ± 2.0 0.23 ± 0.11e … Healthy controls (n = 8) 1 g i.v. × 1 dose … 0.142 ± 0.017d 8.4 ± 1.8 0.23 ± 0.064e …  Reference 16 Child–Pugh class B (n = 5) or C (n = 3) 1 g i.v. × 1 dose 1 g i.v. × 1 … 0.23 ± 0.06 10.7 ± 4.3 20.7 ± 9.3 Healthy controls (n = 8) 1 g i.v. × 1 dose … … 0.13 ± 0.06 9.0 ± 1.7 13.1 ± 4.8 Ciprofloxacin24,25  Reference 24 Cirrhosis (n = 7) 750 mg p.o. q 12 hr × 5 days 3.71 ± 0.81 2.83 ± 0.59f 3.47 ± 1.23 765.1 ± 233.4 18.4 ± 7.1 Healthy controls (n = 7) 750 mg p.o. q 12 hr × 5 days 3.49 ± 0.68 3.48 ± 1.0f 3.71 ± 0.41 840.2 ± 240.0 16.2 ± 4.6  Reference 25 Child–Pugh class B (n = 20) 500 mg p.o. × 1 dose 2.6 ± 1.3 … 3.2 ± 1.8 … 21.9 ± 4.5 Healthy controls (n = 10) 500 mg p.o. × 1 dose 2.6 ± 0.6 … 3.6 ± 1.2 … 19.3 ± 3.8 Clarithromycin33 Child–Pugh class B or C (n = 7) 250 mg p.o. q 12 hr × 5 doses 1.22 ± 0.82 305 ± 211f 5.0g 713 ± 593 9.29 ± 6.2h Healthy controls (n = 6) 250 mg p.o. q 12 hr × 5 doses 1.52 ± 0.35 138 ± 19f 3.3g 486 ± 166 9.29 ± 2.6h Clindamycin49,50  Reference 49 Chronic hepatitis (n = 6) 300 mg i.v. q 12 hr × 4 doses … … 2.1g … … Alcoholic cirrhosis (n = 9) 300 mg i.v. q 12 hr × 4 doses … … 2.5g … … Healthy controls (n = 8) 300 mg i.v. q 12 hr × 4 doses … … 1.8g … …  Reference 50 Alcoholic cirrhosis (n = 7) 300 mg i.v. × 1 dose … 0.693 ± 0.258d 4.5 ± 0.9 121.7 ± 23.6 … Healthy controls (n = 7) 300 mg i.v. × 1 dose … 0.583 ± 0.18d 3.4 ± 0.4 159.2 ± 39.6 … Erythromycin30,–32  Reference 31 Alcoholic cirrhosis (n = 8) 500 mg p.o. × 1 dose 2.04 ± 1.13 93.5 ± 88.9f 4.5g … 11.6 ± 8.9i Healthy controls (n = 6) 500 mg p.o. × 1 dose 1.5 ± 0.91 98.9 ± 34.2f 6.6g … 9.0 ± 5.9i  Reference 32 Alcoholic cirrhosis (n = 6) 500 mg p.o. × 1 dose … … 3.2 ± 0.4 … … Healthy controls (n = 6) 500 mg p.o. × 1 dose … … 2.0 ± 0.7 … …  Reference 30 Alcoholic cirrhosis (n = 5) 500 mg i.v. × 1 dose … 85.5 ± 23.8 2.24 ± 0.89 403.3 ± 138.3j … Healthy controls (n = 6) 500 mg i.v. × 1 dose … 57.6 ± 14.8 1.36 ± 0.43 570.0 ± 201.7j … Isoniazid41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 6.7g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 3.2g … … Linezolid36 Child–Pugh class A or B (n = 7) 600 mg p.o. × 1 dose 11.5 ± 2.0 0.563 ± 0.058d 6.8 ± 3.1 1.12 ± 0.46e 128 ± 60 Healthy controls (n = 8) 600 mg p.o. × 1 dose 11.9 ± 9.8 0.618 ± 0.132d 5.4 ± 1.6 1.39 ± 0.35e 97 ± 31 Metronidazole54,–56  Reference 55 Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3  Reference 56 Child–Pugh class A (n = 14) 500 mg i.v. × 1 dose … 0.74 ± 0.11d 10.7 ± 2.3 0.85 ± 0.26e 124.9 ± 42.3i Child–Pugh class B (n = 9) 500 mg i.v. × 1 dose … 0.79 ± 0.12d 13.5 ± 5.1 0.79 ± 0.36e 124.4 ± 25.8i Child–Pugh class C (n = 12) 500 mg i.v. × 1 dose … 0.81 ± 0.14d 21.5 ± 12.7 0.56 ± 0.28e 174.1 ± 52.0i Healthy controls (n = 7) 500 mg i.v. × 1 dose … 0.80 ± 0.32d 7.4 ± 2.2 1.53 ± 0.37e 81.4 ± 27.0i Cirrhosis with coma (n = 8) 500 mg i.v. × 1 dose … 44 ± 9 20 ± 9 29 ± 10 …  Reference 54 Healthy controls (n = 8) 500 mg i.v. × 1 dose … 48 ± 7 7.3 ± 0.9 83 ± 14 … Alcoholic cirrhosis (n = 8) 7.5 mg/kg i.v. × 1 dose … 0.77 ± 0.16d 18.3 ± 6.1 0.51 ± 0.11e 256.8 ± 56.3 Moxifloxacin27 Child–Pugh class C (n = 9) 400 mg i.v. q 24 hr × 3 doses … … 10.4 (8.5–16.0)k 146.7 (106.7–175.0)k 45.5 (38.1–62.2)k Nafcillin13 Alcoholic cirrhosis (n = 12) 500 mg i.v. × 1 dose … 19.9 ± 5.8 1.2 ± 0.3 291.5 ± 147.6 … Healthy controls (n = 12) 500 mg i.v. × 1 dose … 27.1 ± 7.3 1.0 ± 0.2 583.7 ± 144.2 … Pyrazinamide60 Cirrhosis (n = 10) 19.3 ± 0.6 mg/kg × 1 dose 30.1 ± 5.4 0.6 ± 0.1d 15.1 ± 3.3 0.48 ± 0.15e 280 Healthy controls (n = 9) 19.3 ± 0.6 mg/kg × 1 dose 35.6 ± 5.4 0.66 ± 0.05d 9.2 ± 1.8 0.84 ± 0.2e 725 ± 219 Quinupristin (Q)–dalfopristin (D)61,62 Child–Pugh class A or B (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 3.16g (Q) 7.34g (D) … … 0.91g (Q) 0.61g (D) … … 3.50g,i (Q) 7.36g,i (D) Healthy controls (n = 16) 7.5 mg/kg i.v. × 1 dose 7.5 mg/kg i.v. × 1 dose 2.72g (Q) 7.24g (D) … … 0.91g (Q) 0.45g (D) … … 2.9g,i (Q) 7.36 g,i (D) Rifabutin43 Child–Pugh class A (n = 18) 300 mg p.o. × 1 dose … … 16 ± 11 … 4.39 ± 1.16 Child–Pugh class B (n = 6) 300 mg p.o. × 1 dose … … 11 ± 10 … 3.45 ± 1.12 Child–Pugh class C (n = 4) 300 mg p.o. × 1 dose … … 42 ± 18 … 9.43 ± 2.73 Healthy controls (n = 12) 300 mg p.o. × 1 dose … … 67 ± 45 … 8.85 ± 4.40 Rifampin41,42  Reference 41 Cirrhosis (n = 7) 600 mg p.o. × 1 dose … … 5.4g … … Healthy controls (n = 6) 600 mg p.o. × 1 dose … … 2.8g … …  Reference 42 Cirrhosis (n = 26) 4–10 mg/kg twice weekly × 4 doses … … 3.0–6.8g … … Healthy controls 8–16 mg/kg twice weekly × 4 doses … … 1.9–2.7g … … Tigecycline47 Child–Pugh class A (n = 10) 100 mg i.v. × 1 dose 0.86 ± 0.38 617 ± 234 19.1 ± 5.4 520.0 ± 231.7 3.8 ± 1.8 Child–Pugh class B (n = 10) 100 mg i.v. × 1 dose 0.91 ± 0.55 542 ± 246 23.0 ± 5.0 368.3 ± 155.0 5.6 ± 3.4 Child–Pugh class C (n = 5) 100 mg i.v. × 1 dose 1.21 ± 0.41 378 ± 107 26.8 ± 6.1 225.0 ± 45.0 7.6 ± 1.5 Healthy controls (n = 23) 100 mg i.v. × 1 dose 0.98 ± 0.54 524 ± 157 18.7 ± 7.2 496.7 ± 188.3 3.7 ± 1.3 a Cmax= maximum concentration (plasma or serum), V = volume of distribution, t½ = half-life (terminal phase), CL = total body clearance, AUC = area under the concentration–time curve (from time zero to infinity unless otherwise noted). b Some values are expressed here in units of measure different from those originally reported. c Not reported. d Expressed as L/kg. e Expressed as mL/min/kg. f Expressed as V/F (ratio of volume of distribution to bioavailability). g S.D. not reported. h AUC from time zero to 12 hours. i AUC from time zero to 24 hours. j A more significant difference in clearance of unbound erythromycin was observed in patients with alcoholic cirrhosis versus healthy controls (mean ± S.D., 42.2 ± 10.11 L/hr versus 113.2 ± 44.2 L/hr). k Reported as median (range). Open in new tab Table 1 Summary of Key Pharmacologic and Pharmacokinetic Variables and Dosing Recommendations for Select Antibiotic Agents That Undergo Hepatobiliary Elimination10–75,a Drug Hepatic Metabolism (Phase I or II) Protein Binding (%) CYP Interaction Primary Type(s) of Hepatotoxicity and Frequency (%)b Dose Adjustment Recommendations, by Child–Pugh Classc A B C Azithromycin Demethylation (I) 12–50 …d Cholestatic (<1) None None ND; adjustment likely not needede Cefotaxime Desacetylation (II) 30–50 … …f (<1) None None None Ceftriaxone Nonenzymatic (biliary clearance) 83–96 … Cholestatic (<1) None None Noneg Ciprofloxacin Oxidation (I) 20–40 1A2 inhibitor Hepatocellular, cholestatic (<1) None None None Clarithromycin Hydroxylation (I), demethylation (I) 42–50 3A4 substrate and inhibitor Cholestatic (<1) None None None Clindamycin Sulfoxide metabolite (I), demethylation (I) 60–95 … Hepatocellular, mixed (<1) None None 50% reduction Doxycycline Nonenzymatic 80–90 … Hepatocellular, cholestatic (<1) ND ND ND Erythromycin Demethylation (I) 73–96 3A4 substrate and inhibitor Cholestatic (1–5) Adjustment likely needede,h Adjustment likely needede,h Adjustment likely needede,h Isoniazid Acetylation (II) 4–30 2E1 inducer; 2C19, 1A2, 3A4 inhibitor Hepatocellular (1–5) Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Linezolid Oxidation (I) 31 …i Other (steatosis) (<1) None None ND (use caution) Metronidazole Hydroxylation (I), oxidation (I), glucuronidation (II) <20 2C9 inhibitor Hepatocellular (<1) 500 mg q 12–24 hre 500 mg q 12–24 hre 500 mg q 12–24 hre Minocycline Hydroxylation (I), demethylation (I) 76 … Hepatocellular (<1) ND (caution advised with chronic use)c ND (caution advised with chronic use)c ND (caution advised with chronic use)c Moxifloxacin Glucuronidation (II), sulfation (II) 30–50 … Hepatocellular, cholestatic (<1) None None None Nafcillin Oxidation (I), other unknown methods 90 1A2 inducer Cholestatic (<1) Adjustment may be needede,h Adjustment may be needede,h Adjustment may be needede,h Nitrofurantoin Unknown 90 … Hepatocellular (<1) ND (caution advised with chronic use)e ND (caution advised with chronic use)e ND (caution advised with chronic use)e Norfloxacin Oxidation (I) 14 1A2 inhibitor Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Pyrazinamide Hydrolysis (I) 5–10 … Hepatocellular (>5) 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e Quinupristin–dalfopristin Quinupristin: cysteine–glutathione conjugation (II) Dalfopristin: hydrolysis (I) 55–78, 11–26 3A4 inhibitor …f (<1) 5 mg/kg i.v. q 12 hr 5 mg/kg i.v. q 12 hr ND (further adjustment may be needed)e Rifabutin Deacetylation (II), hydroxylation (I) 72–85 3A4 inducer Hepatocellular (1–5) None (use caution) None (use caution) Adjustment may be needed (use caution) Rifampin Deacetylation (II), oxidation (I) 60–90 3A, 1A2, 2C inducer Hepatocellular (1–5) Consider 50% reductione,h,j Consider 50% reductione,h,j Consider 50% reductione,h,j Sulfamethoxazole–trimethoprim Sulfamethoxazole: acetylation (II) Trimethoprim: hydroxylation (I), oxidation (I) 70, 44 2C9 inhibitor and substrate Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Tigecycline Glucuronidation (II), N-acetylation (II) 71–89 … Cholestatic (<1) None None 50% reduction Drug Hepatic Metabolism (Phase I or II) Protein Binding (%) CYP Interaction Primary Type(s) of Hepatotoxicity and Frequency (%)b Dose Adjustment Recommendations, by Child–Pugh Classc A B C Azithromycin Demethylation (I) 12–50 …d Cholestatic (<1) None None ND; adjustment likely not needede Cefotaxime Desacetylation (II) 30–50 … …f (<1) None None None Ceftriaxone Nonenzymatic (biliary clearance) 83–96 … Cholestatic (<1) None None Noneg Ciprofloxacin Oxidation (I) 20–40 1A2 inhibitor Hepatocellular, cholestatic (<1) None None None Clarithromycin Hydroxylation (I), demethylation (I) 42–50 3A4 substrate and inhibitor Cholestatic (<1) None None None Clindamycin Sulfoxide metabolite (I), demethylation (I) 60–95 … Hepatocellular, mixed (<1) None None 50% reduction Doxycycline Nonenzymatic 80–90 … Hepatocellular, cholestatic (<1) ND ND ND Erythromycin Demethylation (I) 73–96 3A4 substrate and inhibitor Cholestatic (1–5) Adjustment likely needede,h Adjustment likely needede,h Adjustment likely needede,h Isoniazid Acetylation (II) 4–30 2E1 inducer; 2C19, 1A2, 3A4 inhibitor Hepatocellular (1–5) Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Linezolid Oxidation (I) 31 …i Other (steatosis) (<1) None None ND (use caution) Metronidazole Hydroxylation (I), oxidation (I), glucuronidation (II) <20 2C9 inhibitor Hepatocellular (<1) 500 mg q 12–24 hre 500 mg q 12–24 hre 500 mg q 12–24 hre Minocycline Hydroxylation (I), demethylation (I) 76 … Hepatocellular (<1) ND (caution advised with chronic use)c ND (caution advised with chronic use)c ND (caution advised with chronic use)c Moxifloxacin Glucuronidation (II), sulfation (II) 30–50 … Hepatocellular, cholestatic (<1) None None None Nafcillin Oxidation (I), other unknown methods 90 1A2 inducer Cholestatic (<1) Adjustment may be needede,h Adjustment may be needede,h Adjustment may be needede,h Nitrofurantoin Unknown 90 … Hepatocellular (<1) ND (caution advised with chronic use)e ND (caution advised with chronic use)e ND (caution advised with chronic use)e Norfloxacin Oxidation (I) 14 1A2 inhibitor Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Pyrazinamide Hydrolysis (I) 5–10 … Hepatocellular (>5) 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e Quinupristin–dalfopristin Quinupristin: cysteine–glutathione conjugation (II) Dalfopristin: hydrolysis (I) 55–78, 11–26 3A4 inhibitor …f (<1) 5 mg/kg i.v. q 12 hr 5 mg/kg i.v. q 12 hr ND (further adjustment may be needed)e Rifabutin Deacetylation (II), hydroxylation (I) 72–85 3A4 inducer Hepatocellular (1–5) None (use caution) None (use caution) Adjustment may be needed (use caution) Rifampin Deacetylation (II), oxidation (I) 60–90 3A, 1A2, 2C inducer Hepatocellular (1–5) Consider 50% reductione,h,j Consider 50% reductione,h,j Consider 50% reductione,h,j Sulfamethoxazole–trimethoprim Sulfamethoxazole: acetylation (II) Trimethoprim: hydroxylation (I), oxidation (I) 70, 44 2C9 inhibitor and substrate Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Tigecycline Glucuronidation (II), N-acetylation (II) 71–89 … Cholestatic (<1) None None 50% reduction a CYP = cytochrome P-450, ND = no data available. b Rates of asymptomatic liver function test abnormalities may be higher. c Dosing recommendations are direct recommendations from the reviewed studies unless otherwise noted. d None reported. e Author dosing recommendations. f Hepatotoxicity not reported but possible. g Data from one study suggest that a 50% dose decrease be considered due to an increased free fraction of drug in plasma. h Recommendations are based on research using various definitions of cirrhosis, with no Child–Pugh classification of disease severity. i Linezolid metabolism has not been fully elucidated; some data suggest linezolid is a CYP substrate. j A 50% dose reduction is suggested on the basis of a study finding that 6–8 mg/kg twice weekly in cirrhotic patients is equivalent in effect to a twice-weekly dose of 12 mg/kg in healthy controls. Open in new tab Table 1 Summary of Key Pharmacologic and Pharmacokinetic Variables and Dosing Recommendations for Select Antibiotic Agents That Undergo Hepatobiliary Elimination10–75,a Drug Hepatic Metabolism (Phase I or II) Protein Binding (%) CYP Interaction Primary Type(s) of Hepatotoxicity and Frequency (%)b Dose Adjustment Recommendations, by Child–Pugh Classc A B C Azithromycin Demethylation (I) 12–50 …d Cholestatic (<1) None None ND; adjustment likely not needede Cefotaxime Desacetylation (II) 30–50 … …f (<1) None None None Ceftriaxone Nonenzymatic (biliary clearance) 83–96 … Cholestatic (<1) None None Noneg Ciprofloxacin Oxidation (I) 20–40 1A2 inhibitor Hepatocellular, cholestatic (<1) None None None Clarithromycin Hydroxylation (I), demethylation (I) 42–50 3A4 substrate and inhibitor Cholestatic (<1) None None None Clindamycin Sulfoxide metabolite (I), demethylation (I) 60–95 … Hepatocellular, mixed (<1) None None 50% reduction Doxycycline Nonenzymatic 80–90 … Hepatocellular, cholestatic (<1) ND ND ND Erythromycin Demethylation (I) 73–96 3A4 substrate and inhibitor Cholestatic (1–5) Adjustment likely needede,h Adjustment likely needede,h Adjustment likely needede,h Isoniazid Acetylation (II) 4–30 2E1 inducer; 2C19, 1A2, 3A4 inhibitor Hepatocellular (1–5) Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Linezolid Oxidation (I) 31 …i Other (steatosis) (<1) None None ND (use caution) Metronidazole Hydroxylation (I), oxidation (I), glucuronidation (II) <20 2C9 inhibitor Hepatocellular (<1) 500 mg q 12–24 hre 500 mg q 12–24 hre 500 mg q 12–24 hre Minocycline Hydroxylation (I), demethylation (I) 76 … Hepatocellular (<1) ND (caution advised with chronic use)c ND (caution advised with chronic use)c ND (caution advised with chronic use)c Moxifloxacin Glucuronidation (II), sulfation (II) 30–50 … Hepatocellular, cholestatic (<1) None None None Nafcillin Oxidation (I), other unknown methods 90 1A2 inducer Cholestatic (<1) Adjustment may be needede,h Adjustment may be needede,h Adjustment may be needede,h Nitrofurantoin Unknown 90 … Hepatocellular (<1) ND (caution advised with chronic use)e ND (caution advised with chronic use)e ND (caution advised with chronic use)e Norfloxacin Oxidation (I) 14 1A2 inhibitor Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Pyrazinamide Hydrolysis (I) 5–10 … Hepatocellular (>5) 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e Quinupristin–dalfopristin Quinupristin: cysteine–glutathione conjugation (II) Dalfopristin: hydrolysis (I) 55–78, 11–26 3A4 inhibitor …f (<1) 5 mg/kg i.v. q 12 hr 5 mg/kg i.v. q 12 hr ND (further adjustment may be needed)e Rifabutin Deacetylation (II), hydroxylation (I) 72–85 3A4 inducer Hepatocellular (1–5) None (use caution) None (use caution) Adjustment may be needed (use caution) Rifampin Deacetylation (II), oxidation (I) 60–90 3A, 1A2, 2C inducer Hepatocellular (1–5) Consider 50% reductione,h,j Consider 50% reductione,h,j Consider 50% reductione,h,j Sulfamethoxazole–trimethoprim Sulfamethoxazole: acetylation (II) Trimethoprim: hydroxylation (I), oxidation (I) 70, 44 2C9 inhibitor and substrate Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Tigecycline Glucuronidation (II), N-acetylation (II) 71–89 … Cholestatic (<1) None None 50% reduction Drug Hepatic Metabolism (Phase I or II) Protein Binding (%) CYP Interaction Primary Type(s) of Hepatotoxicity and Frequency (%)b Dose Adjustment Recommendations, by Child–Pugh Classc A B C Azithromycin Demethylation (I) 12–50 …d Cholestatic (<1) None None ND; adjustment likely not needede Cefotaxime Desacetylation (II) 30–50 … …f (<1) None None None Ceftriaxone Nonenzymatic (biliary clearance) 83–96 … Cholestatic (<1) None None Noneg Ciprofloxacin Oxidation (I) 20–40 1A2 inhibitor Hepatocellular, cholestatic (<1) None None None Clarithromycin Hydroxylation (I), demethylation (I) 42–50 3A4 substrate and inhibitor Cholestatic (<1) None None None Clindamycin Sulfoxide metabolite (I), demethylation (I) 60–95 … Hepatocellular, mixed (<1) None None 50% reduction Doxycycline Nonenzymatic 80–90 … Hepatocellular, cholestatic (<1) ND ND ND Erythromycin Demethylation (I) 73–96 3A4 substrate and inhibitor Cholestatic (1–5) Adjustment likely needede,h Adjustment likely needede,h Adjustment likely needede,h Isoniazid Acetylation (II) 4–30 2E1 inducer; 2C19, 1A2, 3A4 inhibitor Hepatocellular (1–5) Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Adjustment likely needed (use caution)e,h Linezolid Oxidation (I) 31 …i Other (steatosis) (<1) None None ND (use caution) Metronidazole Hydroxylation (I), oxidation (I), glucuronidation (II) <20 2C9 inhibitor Hepatocellular (<1) 500 mg q 12–24 hre 500 mg q 12–24 hre 500 mg q 12–24 hre Minocycline Hydroxylation (I), demethylation (I) 76 … Hepatocellular (<1) ND (caution advised with chronic use)c ND (caution advised with chronic use)c ND (caution advised with chronic use)c Moxifloxacin Glucuronidation (II), sulfation (II) 30–50 … Hepatocellular, cholestatic (<1) None None None Nafcillin Oxidation (I), other unknown methods 90 1A2 inducer Cholestatic (<1) Adjustment may be needede,h Adjustment may be needede,h Adjustment may be needede,h Nitrofurantoin Unknown 90 … Hepatocellular (<1) ND (caution advised with chronic use)e ND (caution advised with chronic use)e ND (caution advised with chronic use)e Norfloxacin Oxidation (I) 14 1A2 inhibitor Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Pyrazinamide Hydrolysis (I) 5–10 … Hepatocellular (>5) 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e 50% reduction (avoid use if possible)e Quinupristin–dalfopristin Quinupristin: cysteine–glutathione conjugation (II) Dalfopristin: hydrolysis (I) 55–78, 11–26 3A4 inhibitor …f (<1) 5 mg/kg i.v. q 12 hr 5 mg/kg i.v. q 12 hr ND (further adjustment may be needed)e Rifabutin Deacetylation (II), hydroxylation (I) 72–85 3A4 inducer Hepatocellular (1–5) None (use caution) None (use caution) Adjustment may be needed (use caution) Rifampin Deacetylation (II), oxidation (I) 60–90 3A, 1A2, 2C inducer Hepatocellular (1–5) Consider 50% reductione,h,j Consider 50% reductione,h,j Consider 50% reductione,h,j Sulfamethoxazole–trimethoprim Sulfamethoxazole: acetylation (II) Trimethoprim: hydroxylation (I), oxidation (I) 70, 44 2C9 inhibitor and substrate Hepatocellular, cholestatic (<1) ND; likely no adjustment needede ND; likely no adjustment needede ND; likely no adjustment needede Tigecycline Glucuronidation (II), N-acetylation (II) 71–89 … Cholestatic (<1) None None 50% reduction a CYP = cytochrome P-450, ND = no data available. b Rates of asymptomatic liver function test abnormalities may be higher. c Dosing recommendations are direct recommendations from the reviewed studies unless otherwise noted. d None reported. e Author dosing recommendations. f Hepatotoxicity not reported but possible. g Data from one study suggest that a 50% dose decrease be considered due to an increased free fraction of drug in plasma. h Recommendations are based on research using various definitions of cirrhosis, with no Child–Pugh classification of disease severity. i Linezolid metabolism has not been fully elucidated; some data suggest linezolid is a CYP substrate. j A 50% dose reduction is suggested on the basis of a study finding that 6–8 mg/kg twice weekly in cirrhotic patients is equivalent in effect to a twice-weekly dose of 12 mg/kg in healthy controls. Open in new tab Beta-lactams Penicillins The majority of penicillins are eliminated via renal tubular secretion. They are also eliminated in bile to varying degrees, but, with the exception of nafcillin, the amount of drug eliminated in this manner is generally small. The extent of biotransformation by the liver varies among these agents, with penicillin V and oxacillin being most extensively inactivated.11,12 Nafcillin pharmacokinetics was studied in patients with cirrhosis.13 The investigators reported a roughly twofold decrease in nafcillin plasma clearance as well as an almost twofold increase in nafcillin urinary excretion among cirrhotic patients; this seems to indicate that the kidneys are able to compensate for some of the reduced nafcillin elimination. The article did not make any specific dose recommendations for nafcillin but advised decreasing the dose in patients with combined hepatic and renal impairment, as even lower plasma clearance of nafcillin is to be expected in such patients. Cephalosporins The pharmacokinetics of first- and second-generation cephalosporins is poorly studied. However, since these drugs are largely eliminated by the kidneys, no need for dose adjustment in the context of hepatic impairment is expected. Among third-generation cephalosporins, cefotaxime and ceftriaxone are noteworthy, as they both undergo substantial nonrenal clearance. About 40–50% of cefotaxime is metabolized by the liver to an active deacetylated metabolite.14 Two pharmacokinetic studies in cirrhotic patients demonstrated a threefold increase in cefotaxime’s half-life (t½) relative to values in healthy subjects, while one study did not find any significant difference15,–17; total cefotaxime clearance was also decreased, which was likely due to impaired biotransformation in the liver.15,16 Despite alterations in some pharmacokinetic parameters, dose adjustment was not found to be necessary in hepatic impairment due to the wide therapeutic index of the drug.15 Pharmacokinetic features of ceftriaxone include concentration-dependent protein binding as well as extensive secretion in the bile (30–60% of total clearance).18 The pharmacokinetics of single-dose ceftriaxone was studied in patients with Child–Pugh class B or C cirrhosis; compared with values for healthy subjects, there was no significant difference in either t½ or clearance.16 Another study, conducted in subjects with cirrhosis with or without ascites, also found no significant difference in t½ among cirrhotic and noncirrhotic patients and healthy controls; however, the fraction of the unbound drug in plasma was increased by 84% in cirrhotic patients without ascites and by 222% in those with ascites compared with healthy controls.19 The volume of distribution and the total drug clearance were increased by 35% and 60%, respectively, in patients with ascites. The investigators pointed out one outlier—a patient with cirrhosis and concomitant renal insufficiency—who had unbound drug clearance of only 0.752 mL/min/kg. As a result, the authors concluded that while a decrease in the ceftriaxone dose is likely not needed due to the agent’s wide therapeutic range, a 50% dose decrease can be considered for patients with severe hepatic disease. In addition, dose adjustment was recommended for patients with concomitant hepatic and renal failure.19 Cefepime and the newer cephalosporin ceftaroline are both primarily eliminated by the kidneys, and dose adjustment in hepatic impairment is likely not needed.20,21 Fluoroquinolones While all fluoroquinolones demonstrate low protein binding in plasma, the extent of biotransformation in the liver varies by specific agent. Levofloxacin is largely excreted by the kidneys and undergoes limited hepatic metabolism, resulting in two metabolites that account for less than 5% of the amount excreted in the urine.22 Norfloxacin and ciprofloxacin are both excreted by a combination of renal and hepatic mechanisms. Both are hepatically metabolized by the liver to several metabolites, some of which have minor antimicrobial activity.23 While we could not identify any pharmacokinetic studies of norfloxacin in the setting of chronic hepatic impairment, studies of ciprofloxacin indicate no need for dose adjustment in patients with cirrhosis.24,25 Of all the fluoroquinolones, moxifloxacin undergoes the most extensive biotransfomation in the liver: About half of the dose is converted to inactive metabolites.22 Two studies have evaluated the pharmacokinetics of moxifloxacin in patients with Child–Pugh class A, B, or C disease and found no significant differences in pharmacokinetics parameters relative to values in healthy controls.26,27 Collectively, these findings indicate that patients with mild, moderate, or severe hepatic disease will likely not require dosage adjustment while receiving fluoroquinolones. While there are no pertinent pharmacokinetic studies of norfloxacin, this agent has been widely used in patients with cirrhosis and demonstrated to have good tolerability.28,29 Macrolides Among the macrolides, erythromycin and azithromycin are primarily metabolized by the liver, while clarithromycin undergoes both hepatic and renal clearance. Erythromycin is used relatively infrequently, largely due to its adverse-effect profile and potential for drug interactions with cytochrome P-450 (CYP) 3A4 enzymes. Studies conducted in the 1980s demonstrated prolonged t½, decreased clearance, and increased concentration of unbound erythromycin in patients with alcoholic cirrhosis.30,–32 No specific dose adjustment recommendations were provided, and the Child–Pugh system was not utilized to classify patients’ severity of disease. However, based on the pharmacokinetic properties of the drug, dose adjustment and cautious use are prudent. Clarithromycin is metabolized by the liver in part to an active metabolite, 14-hydroxyclarithromycin. It also has the potential for drug interactions, though to a lesser extent than erythromycin. One pharmacokinetic study conducted in patients with Child–Pugh class B or C disease found no significant differences in pharmacokinetic parameters between cirrhotic and healthy subjects33; that suggests that no dose adjustment of clarithromycin is needed in patients with cirrhosis. Azithromycin has the least effect on CYP enzymes of all the macrolide agents included in this review and is mainly excreted in bile. The pharmacokinetics of azithromycin was studied in patients with Child–Pugh class A or B liver cirrhosis, and no changes in pharmacokinetic parameters were found between those patients and healthy volunteers.34 As a result, azithromycin does not require dose adjustment in patients with mild or moderate cirrhosis. Studies of azithromycin in the setting of severe cirrhosis are lacking. Oxazolidinones Linezolid undergoes both renal and hepatic clearance, with about 35% being renally excreted and 65% hepatically oxidized to inactive metabolites.35 Linezolid neither induces nor inhibits the CYP enzymes and is not significantly bound to plasma proteins. One single-dose study in patients with Child–Pugh class A or B cirrhosis did not show significant differences in the area under the concentration–time curve (AUC), t½, or total clearance relative to values in healthy controls.36 Those findings suggest that dose adjustment of linezolid in patients with mild-to-moderate liver disease is not needed. Studies in patients with severe liver disease are lacking, so no conclusive recommendations can be provided. Some data suggest that dose adjustments may be needed in the setting of combined renal and hepatic dysfunction.37 Rifamycins Rifamycins are mainly metabolized by the liver to both active and inactive metabolites and subsequently eliminated via bile and feces. Rifampin is known as a potent inducer of CYP3A, but it can induce other hepatic enzymes such as 1A2 and 2C. Rifabutin is both a substrate and an inducer of CYP3A enzymes, but relative to rifampin it has lower inductive potency toward CYP3A. Both drugs autoinduce their own metabolism, and steady-state conditions are usually achieved after about one week. Rifamycin monotherapy is associated with a modest risk of hepatotoxicity, estimated to be around 1–2%. The risk of hepatotoxicity is notably increased when rifampin is given in combination with other antitubercular medications such as isoniazid and pyrazinamide; frequent monitoring of liver function tests is recommended in such circumstances.38,–40 Two studies conducted in patients with cirrhosis demonstrated that hepatic impairment may require rifampin dose adjustments in order to prevent drug levels above the therapeutic range.41,42 While individuals with normal liver function experienced a drop in rifampin serum levels as a result of autoinduction of rifampin metabolism, the investigators found that cirrhotic patients did not demonstrate a similar trend but instead exhibited relatively higher serum concentrations initially and throughout the study period. Further, the t½ of rifampin was about twofold longer in patients with cirrhosis than in healthy controls. The investigators concluded that a dosage reduction in the context of severe hepatic disease is warranted; they reported a dose of 6–8 mg/kg in patients with severe hepatic disease as being equivalent to a 12-mg/kg dose in healthy subjects.42 One pharmacokinetic study of rifabutin demonstrated increases in AUC and t½ that were directly proportional to the severity of liver dysfunction.43 However, once the investigators included data from a group of healthy volunteers (who were studied separately), the aforementioned pharmacokinetic parameters were not notably different in cirrhotic and noncirrhotic individuals. The authors still advised using caution and considering a dose reduction of rifabutin in patients with severe cirrhosis. Tetracyclines and glycylcyclines All tetracyclines undergo biliary excretion to varying degrees, but all biliary concentrations reach levels severalfold higher than those found in serum. Minocycline and, to a much lesser extent, tigecycline undergo biotransformation in the liver, while doxycycline does not seem to be metabolized by the hepatic CYP enzyme system.44 No pharmacokinetic data or recommendations on the dosing of doxycycline or minocycline in hepatic impairment were found among the articles identified in our literature review.44,45 While the risk of hepatotoxicity is relatively low with these agents, cases of autoimmune hepatitis have been reported, largely in patients receiving long-term minocycline therapy.46 Tigecycline pharmacokinetics was studied in patients with varying degrees of cirrhosis.47 There was no significant difference in the pharmacokinetic parameters between patients of Child–Pugh class A or B and healthy controls. However, class C patients had more than twofold higher AUC values and about half the clearance of controls. As a result, patients with mild-to-moderate hepatic insufficiency do not require dose adjustments, while patients with severe disease should have tigecycline maintenance doses reduced by 50%. Other antibiotics Clindamycin Clindamycin is hepatically cleared to two active metabolites and has plasma protein binding ranging between 60% and 94%.48 Clindamycin is also found in the bile at concentrations twofold to threefold higher than those in serum. The pharmacokinetics of clindamycin was studied back in the 1970s and 1980s before the development and adoption of the Child–Pugh scoring system. As a result, the study findings are difficult to compare, as they were based on varying definitions of hepatic impairment and patient populations. Studies in cirrhotic patients have reported increased t½ and decreased total clearance of clindamycin.49,50 In another study, serum and urine concentrations of clindamycin in patients with varying degrees of hepatic disease (definitions were based on both total bilirubin and aspartate transaminase levels) were measured.51 The authors found that patients with moderate-to-severe hepatic disease not only had higher clindamycin serum concentrations but also had twofold greater urinary recovery of clindamycin compared with subjects with normal liver function. As a result, it is possible that patients with severe hepatic impairment have increased renal clearance of clindamycin as a compensatory response. Our findings indicate that clindamycin dose adjustment in patients with hepatic impairment may be warranted. Expert opinion published elsewhere suggests decreasing the dose by 50% in patients with severe hepatic disease.52 Isoniazid Isoniazid is metabolized by the liver into acetylhydrazine and isonicotinic acid.38,39 Acetylhydrazine is subsequently either hydrolyzed to hydrazine or acetylated into diacetylhydrazine. Isoniazid may be hepatotoxic, and research has suggested that hydrazine may be the cause. In fact, patients who are slow acetylators may be at higher risk for developing hepatotoxicity than fast acetylators. The combination of rifampin with isoniazid has an additive effect on hepatotoxicity risk, as rifampin induces isoniazid hydrolase, resulting in increased hydrazine production.38,39 One study has described the pharmacokinetics of isoniazid with or without concomitant rifampin use in patients with cirrhosis.41 Isoniazid’s t½ was twofold higher in cirrhotic patients. After seven days of therapy, patients with liver disease were more likely to accumulate isoniazid and had 30% higher serum levels. There were no significant differences in any evaluated isoniazid pharmacokinetic values between the patients who received rifampin and those who did not. Also, the authors did not comment on the need for dosing adjustments in patients with cirrhosis.41 Based on the limited pharmacokinetic information available, no isoniazid dosing recommendations can be made for patients with cirrhosis. Due to the risk of hepatotoxicity and the potential for increased serum concentrations and a prolonged t½, close monitoring is recommended for cirrhotic patients who may require isoniazid therapy. Metronidazole Metronidazole is metabolized by the liver to five metabolites, the most notable being the hydroxy metabolite, which carries antimicrobial activity that is 30–65% that of metronidazole.53 Patients with alcoholic cirrhosis have altered metronidazole pharmacokinetics, including an increased t½, a decreased total clearance, and a decreased production of the hydroxyl metabolite.54,55 Dose adjustment recommendations vary somewhat among the reviewed studies and generally suggest increasing the dosing frequency to 12–24 hours.54,55 However, the investigators did not use the Child–Pugh classification, so it remains unclear how the dose adjustments would be applied based on the severity of cirrhosis. Another trial conducted in patients with Child–Pugh class A, B, or C cirrhosis also reported a prolonged metronidazole t½, an increased AUC, and a lower total clearance in all three patient groups.56 However, no specific dosing recommendations were provided by the investigators. In summary, these findings indicate that dose adjustment of metronidazole is needed in patients with hepatic impairment. It is not fully clear at what Child–Pugh class the dose adjustment is to be implemented, but, at a minimum, patients with moderate-to-severe hepatic disease should have their daily dose decreased.55 Nitrofurantoin Most research on the pharmacokinetics of nitrofurantoin was performed in the 1960s and indicated both renal and hepatobiliary clearance as primary routes of elimination.57 In vitro and animal studies demonstrated that nitrofurantoin is eliminated via bile and that the liver plays a major role in its inactivation, but the specific enzymes involved in its biotransformation have not been fully elucidated.58,59 No studies on the pharmacokinetics of nitrofurantoin in the setting of hepatic impairment have been published, so caution should be exercised, particularly if long-term therapy with nitrofurantoin is required. Pyrazinamide Pyrazinamide is metabolized by the liver, and its metabolites are subsequently excreted in urine. Pyrazinamide is among the most hepatotoxic antituberculous medications, but the exact mechanism of hepatotoxicity has not been elucidated.38,39 The pharmacokinetic properties of pyrazinamide were studied in patients with cirrhosis and found to be altered compared with values in healthy individuals; cirrhotic subjects had a 34% higher AUC, a 64% longer t½, and a 43% lower total clearance.60 The researchers recommended decreasing the dose of pyrazinamide by 50% in patients with cirrhosis, but since the Child–Pugh classification was not used in the study, it is unclear if this decrease is needed at all stages of cirrhosis or not. Due to the high risk of hepatotoxicity and decreased metabolism, pyrazinamide is best avoided in patients with cirrhosis if at all possible. If no alternatives are available, frequent monitoring of liver function test results and close followup with a healthcare provider are recommended.38 Quinupristin–dalfopristin Quinupristin and dalfopristin are converted via nonenzymatic reactions to three main active metabolites. Both parent drugs and their metabolites are primarily eliminated via the biliary route.6 Quinupristin–dalfopristin pharmacokinetics was studied in patients with Child–Pugh class A or B cirrhosis.61,62 While the pharmacokinetic parameters of the parent compounds alone were not notably different from those of healthy controls, differences were observed when active metabolites were included in the bioassay; specifically, cirrhotic patients had AUC values that were almost 3-fold higher for quinupristin and 1.5-fold higher for dalfopristin. The investigators suggested a dosage reduction to 5 mg/kg in patients with Child–Pugh class A or B cirrhosis.61,62 Further dosage reduction in patients in Child–Pugh class C may be needed, but pharmacokinetic studies in this patient population have not been published. Sulfamethoxazole–trimethoprim Urinary excretion is the main route of elimination of sulfamethoxazole–trimethoprim. While trimethoprim is largely excreted as unchanged drug, sulfamethoxazole is hepatically transformed to several metabolites.63 We could not identify any pharmacokinetic studies of sulfamethoxazole or trimethoprim in patients with hepatic impairment. However, clinical studies in cirrhotic patients with spontaneous bacterial peritonitis have demonstrated that it is well tolerated and safe in this patient population.1,64 Dosing suggestions and monitoring General principles The liver is the primary site for drug metabolism (biotransformation), which occurs via two key phases to enhance solubility for renal elimination.6,65 Phase I metabolism includes oxidation (e.g., via CYP isoenzymes), reduction, hydroxylation, hydrolysis, and demethylation, which are capacity-limited processes that are compromised in decompensated hepatic failure.4–8,66 Phase II metabolism (conjugation) includes glucuronidation, acetylation, sulfation, methylation, and glycine–glutathione conjugation and is neither capacity limited nor meaningfully compromised in the context of decompensated hepatic failure.65, 67, 68 Various host factors may alter the metabolism of drugs, including genetics (as is the case with poor or extensive metabolizers), the environment (e.g., nutrition, diseases, infection), and physiological factors (e.g., age, sex, metabolic capacity of the liver, hepatic blood flow, first-pass dynamics). In general, impairment of drug metabolism is proportional to hepatic dysfunction. The liver’s metabolic capacity needs to be reduced by greater than 90% before drug metabolism is appreciably reduced and dose adjustments should be considered.4–7,65,69 In addition to altered drug metabolism, considerable hepatic impairment may alter other pharmacokinetic parameters of drugs, including protein binding, volume of distribution, and renal elimination.4–8,69 Reduced hepatic blood flow results in lower hepatic first-pass metabolism, leading to higher bioavailability and serum drug concentrations.5, 6, 69 Hypoalbuminemia resulting from cirrhosis decreases protein binding and increases unbound concentrations of highly protein-bound drugs (e.g., nafcillin). In addition, the volume of distribution of lipophilic agents with relatively high plasma protein binding (e.g., clindamycin) is increased. Ascites secondary to cirrhosis can increase the volume of distribution of hydrophilic agents (e.g., β-lactams). Decompensated hepatic failure may lead to renal vasoconstriction and subsequent renal failure, which reduces renal elimination of agents with primary renal clearance, leading to increased serum drug concentrations.5, 6, 69 Collectively, these altered pharmacokinetic parameters may increase serum drug concentrations. Idiosyncratic reactions resulting in drug-induced liver injury have been reported with many antibiotics. Drug-induced liver injury tends to be unpredictable and varies in its pathological and clinical features based on the offending agent. At the same time, one single agent can be responsible for different types of liver injury. The latency period also can vary from days to months, and signs and symptoms of liver injury may not be apparent for weeks after the offending drug has been discontinued. In general, the type of liver injury can be classified pathologically as hepatocellular, cholestatic, or mixed (Table 1).70,–72 While it is not fully clear whether preexisting hepatic impairment itself increases the risk of drug-induced liver injury, most experts agree that overall, patients with cirrhosis are at higher risk for a worse prognosis and complications.71 Specific recommendations A summary of dosing recommendations for commonly used antibiotics in patients with cirrhosis can be found in Table 1. To help clinicians in situations where pertinent data and recommendations are not available, we have developed Figure 1 as a guide to determining appropriate dosing regimens. When individualizing dosing regimens, clinicians need to consider the indication, the site and severity of infection, and the duration of therapy. For example, shorter courses of therapy (seven days or less), including treatment of mild infections or surgical prophylaxis, may not require dose adjustments. Clinicians should assess if hepatic disease is acute or chronic in nature as well as the degree of hepatic dysfunction. It is also important to evaluate the pharmacokinetic profile of the antibiotic being considered and whether the drug undergoes primarily renal, hepatic (phase I or II metabolism), or mixed clearance. In general, dose adjustments should be considered in the setting of decompensated liver disease, in particular for antibiotics that undergo phase I metabolism, have high protein binding (>90%), or are associated with high rates of hepatotoxicity and other concentration-dependent toxicities (Table 1).5–7,65,69 Figure 1 Open in new tabDownload slide Algorithm for antibiotic dosing and monitoring in adult patients with cirrhosis. Figure 1 Open in new tabDownload slide Algorithm for antibiotic dosing and monitoring in adult patients with cirrhosis. Agents associated with a high frequency of hepatotoxicity should generally be avoided unless other, safer alternatives are not available.5 Patients with hepatorenal syndrome are at considerable risk for accumulation of both hepatically and renally cleared antibiotics and may need further dose reductions. In addition, the risk of drug–drug interactions should be assessed. To minimize the potential for such interactions, caution is advised when using antibiotics that are CYP substrates or notable CYP isoenzyme inducers or inhibitors in the setting of hepatic disease. Further, the clinician should critically evaluate the quality of data available when dosing suggestions are recommended based on pharmacokinetic data and, in clinical practice, critically assess the extent to which participants in a given study were comparable to the patient at hand (Table 2). With select antibiotic agents, therapeutic drug monitoring may be warranted for patients who have severe infections, are not responding to current therapy, are at high risk for drug toxicity, or have developed signs and symptoms consistent with drug toxicity, especially for drugs with a narrow therapeutic range and when such monitoring is not cost or time prohibitive. However, a well-defined therapeutic range is not established for most of the agents discussed in this review, and monitoring may be time prohibitive in many cases, as therapy may be completed prior to the availability of laboratory test results. Conclusion Cirrhosis has multiple effects on the disposition of a wide range of antibacterial agents. Appropriate antibiotic therapy selection and individualized dosing can contribute to optimal clinical outcomes while decreasing the risk of hepatotoxicity. Footnotes The authors have declared no potential conflicts of interest. The Clinical Consultation section features articles that provide brief advice on how to handle specific drug therapy problems. All articles are based on a systematic review of the literature. The assistance of ASHP’s Section of Clinical Specialists and Scientists in soliciting Clinical Consultation submissions is acknowledged. Unsolicited submissions are also welcome. References 1 Fagiuoli S Colli A Bruno R et al. . Management of infections in cirrhotic patients: report of a consensus conference . Dig Liver Dis . 2014 ; 46 : 204 – 12 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Strauss E da Costa MF . The importance of bacterial infections as precipitating factors of chronic hepatic encephalopathy in cirrhosis . Hepatogastroenterology . 1998 ; 45 : 900 – 4 . Google Scholar PubMed WorldCat 3 Borzio M Salerno F Piantoni L et al. . Bacterial infection in patients with advanced cirrhosis: a multicentre prospective study . Dig Liver Dis . 2001 ; 33 : 41 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Delco F Tchambaz L Schlienger R et al. . Dose adjustment in patients with liver disease . Drug Saf . 2005 ; 28 : 529 – 45 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Lewis JH Stine JG . Review article: prescribing medications in patients with cirrhosis—a practical guide . Aliment Pharmacol Ther . 2013 ; 37 : 1132 – 56 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Verbeeck RK . Pharmacokinetics and dosage adjustment in patients with hepatic dysfunction . Eur J Clin Pharmacol . 2008 ; 64 : 1147 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Westphal JF Brogard JM . Clinical pharmacokinetics of newer antibacterial agents in liver disease . Clin Pharmacokinet . 1993 ; 24 : 46 – 58 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Morgan DJ McLean AJ . Clinical pharmacokinetic and pharmacodynamic considerations in patients with liver disease. An update . Clin Pharmacokinet . 1995 ; 29 : 370 – 91 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Dowling TC Matzke GR Murphy JE Burckart GJ . Evaluation of renal drug dosing: prescribing information and clinical pharmacist approaches . Pharmacotherapy . 2010 ; 30 : 776 – 86 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Pugh RN Murray-Lyon IM Dawson JL et al. . Transection of the oesophagus for bleeding oesophageal varices . Br J Surg . 1973 ; 60 : 646 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Barza M Weinstein L . Pharmacokinetics of the penicillins in man . Clin Pharmacokinet . 1976 ; 1 : 297 – 308 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Plaut ME O’Connell CJ Pabico RC Davidson D . Penicillin handling in normal and azotemic patients . J Lab Clin Med . 1969 ; 74 : 12 – 8 . Google Scholar PubMed WorldCat 13 Marshall JP 2nd Salt WB Elam RO et al. . Disposition of nafcillin in patients with cirrhosis and extrahepatic biliary obstruction . Gastroenterology . 1977 ; 73 : 1388 – 92 . Google Scholar PubMed WorldCat 14 LeFrock JL Prince RA Leff RD . Mechanism of action, antimicrobial activity, pharmacology, adverse effects, and clinical efficacy of cefotaxime . Pharmacotherapy . 1982 ; 2 : 174 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Ko RJ Sattler FR Nichols S et al. . Pharmacokinetics of cefotaxime and desacetyl-cefotaxime in patients with liver disease . Antimicrob Agents Chemother . 1991 ; 35 : 1376 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Hary L Andrejak M Leleu S et al. . The pharmacokinetics of ceftriaxone and cefotaxime in cirrhotic patients with ascites . Eur J Clin Pharmacol . 1989 ; 36 : 613 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Graninger W Uihlein M Ferenci P et al. . Cefotaxime and desacetyl-cefotaxime blood levels in hepatic dysfunction . J Antimicrob Chemother . 1984 ; 14 ( suppl B ): 143 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Balant L Dayer P Auckenthaler R . Clinical pharmacokinetics of the third generation cephalosporins . Clin Pharmacokinet . 1985 ; 10 : 101 – 43 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Stoeckel K Tuerk H Trueb V McNamara PJ . Single-dose ceftriaxone kinetics in liver insufficiency . Clin Pharmacol Ther . 1984 ; 36 : 500 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Cefepime injection package insert . Schaumburg, IL : Sagent Pharmaceuticals ; 2013 Jan . WorldCat COPAC 21 Teflaro (ceftaroline fosamil) injection package insert . St. Louis : Forest Pharmaceuticals ; 2013 . WorldCat COPAC 22 Rodvold KA Neuhauser M . Pharmacokinetics and pharmacodynamics of fluoroquinolones . Pharmacotherapy . 2001 ; 21 : 233S – 252S . Google Scholar Crossref Search ADS PubMed WorldCat 23 Lode H Hoffken G Boeckk M et al. . Quinolone pharmacokinetics and metabolism . J Antimicrob Chemother . 1990 ; 26 ( suppl B ): 41 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Frost RW Lettieri JT Krol G et al. . The effect of cirrhosis on the steady-state pharmacokinetics of oral ciprofloxacin . Clin Pharmacol Ther . 1989 ; 45 : 608 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Dixit RK Satapathy SK Kumar R et al. . Pharmacokinetics of ciprofloxacin in patients with liver cirrhosis . Ind J Gastroenterol . 2002 ; 21 : 62 – 3 . WorldCat 26 Avelox (moxifloxacin) package insert . West Haven, CT : Bayer Corp. ; 1999 . WorldCat COPAC 27 Barth J Jager D Mundkowski R et al. . Single- and multiple-dose pharmacokinetics of intravenous moxifloxacin in patients with severe hepatic impairment . J Antimicrob Chemother . 2008 ; 62 : 575 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Esposito S Galante D Barba D et al. . Efficacy and safety of ciprofloxacin in the treatment of UTIs and RTIs in patients affected by liver diseases . Infection . 1988 ; 16 ( suppl 1 ): S57 – 61 . Google Scholar Crossref Search ADS PubMed WorldCat 29 Gines P Rimola A Planas R et al. . Norfloxacin prevents spontaneous bacterial peritonitis recurrence in cirrhosis: results of a double-blind, placebo-controlled trial . Hepatology . 1990 ; 12 : 716 – 24 . Google Scholar Crossref Search ADS PubMed WorldCat 30 Barre J Mallat A Rosenbaum J et al. . Pharmacokinetics of erythromycin in patients with severe cirrhosis. Respective influence of decreased serum binding and impaired liver metabolic capacity . Br J Clin Pharmacol . 1987 ; 23 : 753 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 31 Kroboth PD Brown A Lyon JA et al. . Pharmacokinetics of single-dose erythromycin in normal and alcoholic liver disease subjects . Antimicrob Agents Chemother . 1982 ; 21 : 135 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat 32 Hall KW Nightingale CH Gibaldi M et al. . Pharmacokinetics of erythromycin in normal and alcoholic liver disease subjects . J Clin Pharmacol . 1982 ; 22 : 321 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 33 Chu SY Granneman GR Pichotta PJ et al. . Effect of moderate or severe hepatic impairment on clarithromycin pharmacokinetics . J Clin Pharmacol . 1993 ; 33 : 480 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 34 Mazzei T Surrenti C Novelli A et al. . Pharmacokinetics of azithromycin in patients with impaired hepatic function . J Antimicrob Chemother . 1993 ; 31 ( suppl E ): 57 – 63 . Google Scholar Crossref Search ADS PubMed WorldCat 35 Stalker DJ Jungbluth GL . Clinical pharmacokinetics of linezolid, a novel oxazolidinone antibacterial . Clin Pharmacokinet . 2003 ; 42 : 1129 – 40 . Google Scholar Crossref Search ADS PubMed WorldCat 36 Hendershot P Jungbluth G Cammarata S . Pharmacokinetics of linezolid in patients with liver disease . J Antimicrob Chemother . 1999 ; 44 ( suppl A ): 55 . WorldCat 37 Tsuji Y Hiraki Y Mizoguchi A et al. . Pharmacokinetics of repeated dosing of linezolid in a hemodialysis patient with chronic renal failure . J Infect Chemother . 2008 ; 14 : 156 – 60 . Google Scholar Crossref Search ADS PubMed WorldCat 38 Yew WW Leung CC . Antituberculosis drugs and hepatotoxicity . Respirology . 2006 ; 11 : 699 – 707 . Google Scholar Crossref Search ADS PubMed WorldCat 39 Tostmann A Boeree MJ Aarnoutse RE et al. . Antituberculosis drug-induced hepatotoxicity: concise up-to-date review . J Gastroenterol Hepatol . 2008 ; 23 : 192 – 202 . Google Scholar Crossref Search ADS PubMed WorldCat 40 Burman WJ Gallicano K Peloquin C . Comparative pharmacokinetics and pharmacodynamics of the rifamycin antibacterials . Clin Pharmacokinet . 2001 ; 40 : 327 – 41 . Google Scholar Crossref Search ADS PubMed WorldCat 41 Acocella G Bonollo L Garimoldi M et al. . Kinetics of rifampicin and isoniazid administered alone and in combination to normal subjects and patients with liver disease . Gut . 1972 ; 13 : 47 – 53 . Google Scholar Crossref Search ADS PubMed WorldCat 42 Curci G Claar E Bergamini N et al. . Studies on blood serum levels of rifampicin in patients with normal and impaired liver function . Chemotherapy . 1973 ; 19 : 197 – 205 . Google Scholar Crossref Search ADS PubMed WorldCat 43 Benedetti MS Edwards DM Poggesi I et al. . Pharmacokinetics of rifabutin in hepatic insufficiency . Abstract presented at 9th International Conference on AIDS . Berlin, Germany ; 1993 Jun 6 . 44 Agwuh KN MacGowan A . Pharmacokinetics and pharmacodynamics of the tetracyclines including glycylcyclines . J Antimicrob Chemother . 2006 ; 58 : 256 – 65 . Google Scholar Crossref Search ADS PubMed WorldCat 45 Saivin S Houin G . Clinical pharmacokinetics of doxycycline and minocycline . Clin Pharmacokinet . 1988 ; 15 : 355 – 66 . Google Scholar Crossref Search ADS PubMed WorldCat 46 Ford TJ Dillon JF . Minocycline hepatitis . Eur J Gastroenterol Hepatol . 2008 ; 20 : 796 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 47 Korth-Bradley JM Baird-Bellaire SJ Patat AA et al. . Pharmacokinetics and safety of a single intravenous dose of the antibiotic tigecycline in patients with cirrhosis . J Clin Pharmacol . 2011 ; 51 : 93 – 101 . Google Scholar Crossref Search ADS PubMed WorldCat 48 Cleocin phosphate (clindamycin) injection package insert . Deerfield, IL : Pfizer Inc. ; 2013 Oct . WorldCat COPAC 49 Hinthorn DR Baker LH Romig DA et al. . Use of clindamycin in patients with liver disease . Antimicrob Agents Chemother . 1976 ; 9 : 498 – 501 . Google Scholar Crossref Search ADS PubMed WorldCat 50 Avant GR Schenker S Alford RH . The effect of cirrhosis on the disposition and elimination of clindamycin . Am J Dig Dis . 1975 ; 20 : 223 – 30 . Google Scholar Crossref Search ADS PubMed WorldCat 51 Williams DN Crossley K Hoffman C Sabath LD . Parenteral clindamycin phosphate: pharmacology with normal and abnormal liver function and effect on nasal staphylococci . Antimicrob Agents Chemother . 1975 ; 7 : 153 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 52 Kluytmans J Murk J . Lincomycin and clindamycin . In: Grayson M , ed. Kucers’ the use of antibiotics . 6th ed . Boca Raton, FL : CRC Press Taylor and Francis Group ; 2010 : 987 – 1003 . Google Preview WorldCat COPAC 53 Lamp KC Freeman CD Klutman NE Lacy MK . Pharmacokinetics and pharmacodynamics of the nitroimidazole antimicrobials . Clin Pharmacokinet . 1999 ; 36 : 353 – 73 . Google Scholar Crossref Search ADS PubMed WorldCat 54 Loft S Sonne J Dossing M Andreasen PB . Metronidazole pharmacokinetics in patients with hepatic encephalopathy . Scand J Gastroenterol . 1987 ; 22 : 117 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat 55 Lau AH Evans R Chang CW Seligsohn R . Pharmacokinetics of metronidazole in patients with alcoholic liver disease . Antimicrob Agents Chemother . 1987 ; 31 : 1662 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat 56 Muscara MN Pedrazzoli J Jr Miranda EL et al. . Plasma hydroxy-metronidazole/metronidazole ratio in patients with liver disease and in healthy volunteers . Br J Clin Pharmacol . 1995 ; 40 : 477 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 57 Gleckman R Alvarez S Joubert DW . Drug therapy reviews: nitrofurantoin . Am J Hosp Pharm . 1979 ; 36 : 342 – 51 . Google Scholar PubMed WorldCat 58 Conklin JD Sobers RJ Wagner DL . Further studies on nitrofurantoin excretion in dog hepatic bile . Br J Clin Pharmacol . 1973 ; 48 : 273 – 7 . Google Scholar Crossref Search ADS WorldCat 59 Paul HE Ells VR Kopko F Bender RC . Metabolic degradation of the nitrofurans . J Med Pharm Chem . 1960 ; 2 : 563 – 84 . Google Scholar Crossref Search ADS PubMed WorldCat 60 Lacroix C Tranvouez JL Phan Hoang T et al. . Pharmacokinetics of pyrazinamide and its metabolites in patients with hepatic cirrhotic insufficiency . Arzneimittelforschung . 1990 ; 40 : 76 – 9 . Google Scholar PubMed WorldCat 61 Delgado G Jr Neuhauser MM Bearden DT Danziger LH . Quinupristin-dalfopristin: an overview . Pharmacotherapy . 2000 ; 20 : 1469 – 85 . Google Scholar Crossref Search ADS PubMed WorldCat 62 Chevalier P Rey J Boucher E et al. . Open, comparative, study of the pharmacokinetics and safety of a single dose of quinupristin/dalfopristin (Q/D, RP 59500, Synercid) in subjects with hepatic cirrhosis and healthy volunteers . Abstract presented at 20th International Congress of Chemotherapy . Sydney, Australia ; 1997 . 63 Bactrim (sulfamethoxazole and trimethoprim) package insert . Philadelphia : AR Scientific ; 2005 Mar . WorldCat COPAC 64 Lontos S Gow PJ Vaughan RB Angus PW . Norfloxacin and trimethoprim-sulfamethoxazole therapy have similar efficacy in prevention of spontaneous bacterial peritonitis . J Gastroenterol Hepatol . 2008 ; 23 : 252 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 65 Senior JR . Monitoring for hepatotoxicity: what is the predictive value of liver “function” tests ? Clin Pharmacol Ther . 2009 ; 85 : 331 – 4 . Google Scholar Crossref Search ADS PubMed WorldCat 66 Amarapurkar DN . Prescribing medications in patients with decompensated liver cirrhosis . Int J Hepatol . 2011 ; 2011 : 519526 . Google Scholar PubMed WorldCat 67 Abboud G Kaplowitz N . Drug-induced liver injury . Drug Saf . 2007 ; 30 : 277 – 94 . Google Scholar Crossref Search ADS PubMed WorldCat 68 Verma S Kaplowitz N . Diagnosis, management and prevention of drug-induced liver injury . Gut . 2009 ; 58 : 1555 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat 69 Rodighiero V . Effects of liver disease on pharmacokinetics. An update . Clin Pharmacokinet . 1999 ; 37 : 399 – 431 . Google Scholar Crossref Search ADS PubMed WorldCat 70 Kaplowitz N . Drug-induced liver injury . Clin Infect Dis . 2004 ; 38 ( suppl 2 ): S44 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 71 Chalasani N Bjornsson E . Risk factors for idiosyncratic drug-induced liver injury . Gastroenterology . 2010 ; 138 : 2246 – 59 . Google Scholar Crossref Search ADS PubMed WorldCat 72 Thiim M Friedman LS . Hepatotoxicity of antibiotics and antifungals . Clin Liver Dis . 2003 ; 7 : 381 – 99 . Google Scholar Crossref Search ADS PubMed WorldCat 73 Lexicomp Online [online database] . Hudson, OH : Lexi-Comp (accessed 2014 Apr 1). WorldCat COPAC 74 Micromedex 2.0 [online database] . Greenwood Village, CO : Truven Health Analytics (accessed 2014 Apr 1). WorldCat COPAC 75 National Institutes of Health . Livertox: clinical and research information on drug-induced liver injury [online database] . http://livertox.nih.gov/ (accessed 2014 Apr 1). Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.

Menzin,, Joseph;Sussman,, Matthew;Nichols,, Christine;Friedman,, Mark;Zbrozek,, Arthur

doi: 10.2146/ajhp130729pmid: 25225449

Abstract Purpose The impact of correcting elevated International Normalized Ratio (INR) values on inhospital mortality in patients with warfarin-associated major bleeding is presented. Methods Using patient information from the database of a large U.S. health system, a retrospective analysis was conducted to (1) evaluate inpatient practice patterns in correcting INR elevations among patients hospitalized with warfarin-related intracranial hemorrhage (ICH) or non-ICH bleeding and (2) test the hypothesis that achieving INR correction, defined as an INR of ≤1.5, at any point during the hospital stay is correlated with lower inhospital mortality. Cox proportional hazards models were constructed to assess predictors of inhospital death. Results Among the 354 patients who met the study selection criteria, INR correction was achieved in 87.9% overall (92.5% and 85.5% of patients with ICH and non-ICH bleeds, respectively). Patients whose elevated INR values were corrected had significantly lower inhospital death rates than those with uncorrected elevations: 15.3% versus 55.6% (p = 0.010) among patients with ICH and 2.0% versus 11.8% (p = 0.017) among those with non-ICH bleeds. After adjusting for baseline demographics and comorbidities, the correlation between failure to correct INR elevations and increased mortality risk was significant only for patients with ICH (hazard ratio, 8.04; 95% confidence interval, 2.07–31.18; p = 0.003). Conclusion Results of this study indicated that correction of elevated INR values was associated with a lower likelihood of inhospital death among warfarin-treated patients hospitalized for ICH or non-ICH major bleeding. The anticoagulant warfarin has been a standard-of-care treatment in the United States for over 50 years.1 Despite its declining use with the addition of novel oral anticoagulant treatment alternatives to the market, warfarin remains the most prevalent anticoagulant prescribed in the United States today.2,3 However, warfarin is associated with a high frequency of adverse events (AEs). During the period 2003–06, warfarin ranked ninth on the Food and Drug Administration (FDA) Adverse Event Reporting System list of prescription drugs associated with serious adverse outcomes.4 More recent data from an unpublished study of AEs, submitted to FDA by the Institute for Safe Medication Practices, showed that warfarin was the second highest-listed drug in terms of the total number of direct AE reports to FDA.5 Furthermore, actual rates of warfarin-related AEs in clinical practice may be even higher than FDA data indicate, since serious outcomes associated with anticoagulant use reported to FDA are often underestimates.6 Results from an independent review of published AE incidence studies suggested that warfarin-related AEs were severely underreported to FDA, with reporting rates as low as 1% for warfarin-associated hospitalizations among Medicare patients.6 Disparities between actual and reported AE rates may be due to various patient-, physician-, and organization-level forces; however, health professional knowledge and attitudes have been cited as key factors.7 The most common warfarin-associated AE is bleeding, with yearly rates of major bleeding of approximately 0.5–6.5% among warfarin-treated patients (including those with various indications for anticoagulation such as atrial fibrillation, venous thromboembolism, pulmonary embolism, and stroke), with fatal bleeding occurring in approximately 0.1–1.0% of treated patients each year.8,–13 This high variability in reported bleeding rates is due to differing definitions of major bleeding, which may be defined by evidence of overt gastrointestinal (GI) bleeding, the loss of a specific number of units of blood (e.g., >2 units within seven days, ≥3 units over a longer period), or evidence of an emergency department (ED) or inpatient visit for a hemorrhage during warfarin therapy.8,–12 Specific treatment for warfarin-associated bleeding is dependent on various factors, including International Normalized Ratio (INR) levels at the time of presentation with bleeding, the location and severity of bleeding, and underlying comorbidities.14 Products commonly administered to control major bleeding include fresh frozen plasma (FFP), prothrombin complex concentrates (PCCs), and phytonadione (vitamin K) preparations, which are commonly used in addition to FFP or PCCs. Prior studies have evaluated the associations of INR levels, hematoma size, and mortality. In a small retrospective analysis of patients admitted to the intensive care unit with oral anticoagulant–associated intracranial hemorrhage (ICH), an increased INR that persisted for 2 hours after hospital admission was found to be a significant predictor of increased hematoma size.15 In a separate study, greater hematoma size was associated with an increased risk of death within three months of hospital admission for ICH.16 Additionally, an analysis of electronic health record (EHR) data for patients receiving hospital care for warfarin-associated ICH indicated that failure to correct the INR (correction was defined as achieving an INR value of ≤1.3) within 48 hours of FFP administration was associated with a significantly higher risk of death within 30 days of admission.17 Taken together, the results of these retrospective analyses suggest that INR correction might reduce the risk of death after warfarin-associated ICH.15,–17 However, there is little information available regarding the association between INR correction and deaths during bleeding-related hospitalizations, taking into consideration both ICH and non-ICH major bleeding. As FFP alternatives offering more rapid INR reversal enter the U.S. market for the correction of anticoagulant-associated major bleeding, there is a need for greater understanding of the benefits of rapid reversal with regard to clinical outcomes. Therefore, the objectives of the study described here were to (1) describe inpatient clinical practice patterns for INR correction, as well as trends in the time to INR correction, among patients presenting with warfarin-related major bleeding, (2) test the hypothesis that achieving INR correction at any point during the hospital stay would be correlated with lower rates of inhospital mortality, and (3) carry out the above analyses in patients with either ICH or non-ICH major bleeding. Methods Design and setting This retrospective database analysis used EHR data from a large integrated health system for the period January 2004–January 2010. Study patients were identified based on EHR encounter records indicating receipt of FFP; diagnosis of major hemorrhage, as indicated by specified International Classification of Diseases, 9th Revision, Clinical Modification (ICD-9-CM) diagnosis codes (appendix); an elevated INR value at the time of arrival to the hospital; and a prescription for warfarin before and/or during the hospitalization episode. Appendix Major bleeding diagnosis codesa Condition ICD-9-CM Diagnosis Code Description Intracranial hemorrhage (primary or secondary diagnosis) 430.xx Subarachnoid hemorrhage 431.xx Intracerebral hemorrhage 432.0x, 432.1x, 432.9x Other and unspecified intracranial hemorrhage, including subdural 852.0x, 852.2x, 852.4x Subarachnoid, subdural, and extradural hemorrhage, following injury 853.0x Other and unspecified intracranial hemorrhage following injury Gastrointestinal hemorrhage (primary diagnosis only) 456.0x, 456.20 Esophageal varices with bleeding 530.7x, 530.82 Esophageal hemorrhage 531.00, 531.01, 531.20, 531.21, 531.40, 531.41, 531.60, 531.61 Gastric ulcer with hemorrhage 532.00, 532.01, 532.20, 532.21, 532.40, 532.41, 532.60, 532.61 Duodenal ulcer with hemorrhage 533.00, 533.01, 533.20, 533.21, 533.40, 533.41, 533.60, 533.61 Peptic ulcer with hemorrhage 534.00, 534.01, 534.20, 534.21, 534.40, 534.41, 534.60, 534.61 Gastrojejunal ulcer with hemorrhage 535.01, 535.11, 535.21, 535.31, 535.41, 535.51, 535.61 Gastritis and duodenitis with hemorrhage 537.83 Angiodysplasia of stomach and duodenum with hemorrhage 562.02, 562.03, 562.12, 562.13 Diverticulosis of intestine with hemorrhage 568.81 Hemoperitoneum (nontraumatic) 569.3x, 569.85 Other disorders of intestine with hemorrhage 578.0x, 578.1x, 578.9x Gastrointestinal hemorrhage Other major hemorrhages 459.0x Hemorrhage, unspecified 593.81 Vascular disorder of kidneys with hemorrhage 599.7x Hematuria 719.1x Hemarthrosis 784.7x, 784.8x Hemorrhage from nose or throat 958.2x Secondary and recurrent hemorrhage as a complication from trauma Condition ICD-9-CM Diagnosis Code Description Intracranial hemorrhage (primary or secondary diagnosis) 430.xx Subarachnoid hemorrhage 431.xx Intracerebral hemorrhage 432.0x, 432.1x, 432.9x Other and unspecified intracranial hemorrhage, including subdural 852.0x, 852.2x, 852.4x Subarachnoid, subdural, and extradural hemorrhage, following injury 853.0x Other and unspecified intracranial hemorrhage following injury Gastrointestinal hemorrhage (primary diagnosis only) 456.0x, 456.20 Esophageal varices with bleeding 530.7x, 530.82 Esophageal hemorrhage 531.00, 531.01, 531.20, 531.21, 531.40, 531.41, 531.60, 531.61 Gastric ulcer with hemorrhage 532.00, 532.01, 532.20, 532.21, 532.40, 532.41, 532.60, 532.61 Duodenal ulcer with hemorrhage 533.00, 533.01, 533.20, 533.21, 533.40, 533.41, 533.60, 533.61 Peptic ulcer with hemorrhage 534.00, 534.01, 534.20, 534.21, 534.40, 534.41, 534.60, 534.61 Gastrojejunal ulcer with hemorrhage 535.01, 535.11, 535.21, 535.31, 535.41, 535.51, 535.61 Gastritis and duodenitis with hemorrhage 537.83 Angiodysplasia of stomach and duodenum with hemorrhage 562.02, 562.03, 562.12, 562.13 Diverticulosis of intestine with hemorrhage 568.81 Hemoperitoneum (nontraumatic) 569.3x, 569.85 Other disorders of intestine with hemorrhage 578.0x, 578.1x, 578.9x Gastrointestinal hemorrhage Other major hemorrhages 459.0x Hemorrhage, unspecified 593.81 Vascular disorder of kidneys with hemorrhage 599.7x Hematuria 719.1x Hemarthrosis 784.7x, 784.8x Hemorrhage from nose or throat 958.2x Secondary and recurrent hemorrhage as a complication from trauma a ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification. Open in new tab Appendix Major bleeding diagnosis codesa Condition ICD-9-CM Diagnosis Code Description Intracranial hemorrhage (primary or secondary diagnosis) 430.xx Subarachnoid hemorrhage 431.xx Intracerebral hemorrhage 432.0x, 432.1x, 432.9x Other and unspecified intracranial hemorrhage, including subdural 852.0x, 852.2x, 852.4x Subarachnoid, subdural, and extradural hemorrhage, following injury 853.0x Other and unspecified intracranial hemorrhage following injury Gastrointestinal hemorrhage (primary diagnosis only) 456.0x, 456.20 Esophageal varices with bleeding 530.7x, 530.82 Esophageal hemorrhage 531.00, 531.01, 531.20, 531.21, 531.40, 531.41, 531.60, 531.61 Gastric ulcer with hemorrhage 532.00, 532.01, 532.20, 532.21, 532.40, 532.41, 532.60, 532.61 Duodenal ulcer with hemorrhage 533.00, 533.01, 533.20, 533.21, 533.40, 533.41, 533.60, 533.61 Peptic ulcer with hemorrhage 534.00, 534.01, 534.20, 534.21, 534.40, 534.41, 534.60, 534.61 Gastrojejunal ulcer with hemorrhage 535.01, 535.11, 535.21, 535.31, 535.41, 535.51, 535.61 Gastritis and duodenitis with hemorrhage 537.83 Angiodysplasia of stomach and duodenum with hemorrhage 562.02, 562.03, 562.12, 562.13 Diverticulosis of intestine with hemorrhage 568.81 Hemoperitoneum (nontraumatic) 569.3x, 569.85 Other disorders of intestine with hemorrhage 578.0x, 578.1x, 578.9x Gastrointestinal hemorrhage Other major hemorrhages 459.0x Hemorrhage, unspecified 593.81 Vascular disorder of kidneys with hemorrhage 599.7x Hematuria 719.1x Hemarthrosis 784.7x, 784.8x Hemorrhage from nose or throat 958.2x Secondary and recurrent hemorrhage as a complication from trauma Condition ICD-9-CM Diagnosis Code Description Intracranial hemorrhage (primary or secondary diagnosis) 430.xx Subarachnoid hemorrhage 431.xx Intracerebral hemorrhage 432.0x, 432.1x, 432.9x Other and unspecified intracranial hemorrhage, including subdural 852.0x, 852.2x, 852.4x Subarachnoid, subdural, and extradural hemorrhage, following injury 853.0x Other and unspecified intracranial hemorrhage following injury Gastrointestinal hemorrhage (primary diagnosis only) 456.0x, 456.20 Esophageal varices with bleeding 530.7x, 530.82 Esophageal hemorrhage 531.00, 531.01, 531.20, 531.21, 531.40, 531.41, 531.60, 531.61 Gastric ulcer with hemorrhage 532.00, 532.01, 532.20, 532.21, 532.40, 532.41, 532.60, 532.61 Duodenal ulcer with hemorrhage 533.00, 533.01, 533.20, 533.21, 533.40, 533.41, 533.60, 533.61 Peptic ulcer with hemorrhage 534.00, 534.01, 534.20, 534.21, 534.40, 534.41, 534.60, 534.61 Gastrojejunal ulcer with hemorrhage 535.01, 535.11, 535.21, 535.31, 535.41, 535.51, 535.61 Gastritis and duodenitis with hemorrhage 537.83 Angiodysplasia of stomach and duodenum with hemorrhage 562.02, 562.03, 562.12, 562.13 Diverticulosis of intestine with hemorrhage 568.81 Hemoperitoneum (nontraumatic) 569.3x, 569.85 Other disorders of intestine with hemorrhage 578.0x, 578.1x, 578.9x Gastrointestinal hemorrhage Other major hemorrhages 459.0x Hemorrhage, unspecified 593.81 Vascular disorder of kidneys with hemorrhage 599.7x Hematuria 719.1x Hemarthrosis 784.7x, 784.8x Hemorrhage from nose or throat 958.2x Secondary and recurrent hemorrhage as a complication from trauma a ICD-9-CM = International Classification of Diseases, 9th Revision, Clinical Modification. Open in new tab Data source The EHR database used for the analysis included data from healthcare service providers within a U.S. integrated health system; the data covered all in-network care information from primary care physicians, specialists, hospitals, and laboratory services, as well as prescription orders. The study was approved by the health system’s internal review board. The database included demographics, encounters, procedures, medication orders and administrations, laboratory readings, and date of death, among other elements. Diagnosis codes in the EHR system were entered by physicians and billing coders. In some cases, the evaluated data on inpatient visits covered services relating to multiple primary diagnoses (one from each hospital department, as applicable) and multiple secondary diagnoses. For all patients, the dates of laboratory test results and medication administrations were available for the entire six-year study period. Hourly laboratory time-stamp data and medication administration data were available for portions of the study period (January 2006–January 2010 and January 2008–January 2010, respectively). Laboratory and medication administration information was available for both ED and inpatient settings, allowing for the evaluation of a full set of information for patients admitted through the ED. Patient selection Patients were selected for inclusion in the study if they met all of the criteria outlined below. Patients were first selected on the basis of evidence of FFP administration during a hospitalization episode during the study period; receipt of FFP was defined as an EHR notation of at least one of seven Healthcare Common Procedure Coding System codes (P9017, P9020, P9023, P9043, P9044, P9059, P9060) or at least one medication order with “plasma” in the medication name. Second, patients were selected for inclusion in the study if there was a documented diagnosis of a major hemorrhage on the day before, the day of, or the day after receipt of FFP; major hemorrhages, which included both ICH and non-ICH bleeds (including GI or other major bleed types), were identified by the ICD-9-CM diagnosis codes listed on the EHR. Third, the inclusion criteria stipulated the notation of at least one elevated INR value (defined as an INR of ≥2) on either the day before or the day of FFP administration, as well as the documentation of at least one additional INR value during the hospitalization episode. Finally, only patients who had at least one medication order for warfarin with an order date in the 90 days before and/or during the hospitalization episode were included in the analysis. Patients were excluded from the analysis if FFP was administered in an ED setting and there was no inpatient hospitalization after the ED encounter. Study measures Baseline study measures included demographics (age, sex) and clinical characteristics (Deyo-modified Charlson Comorbidity Index [DM-CCI] score and rates of selected other diagnoses), which were evaluated in the 12-month period prior to the first day of the hospitalization episode (with the first day of the hospitalization designated as the index date for all analyses). The DM-CCI score is a measure of physical health status commonly used in studies of medical claims and chronic disease.18,–20 The score is calculated by assigning points for the presence of the following 17 conditions (with higher scores indicating poorer overall health status): myocardial infarction, congestive heart failure, peripheral vascular disease, cerebrovascular disease, dementia, chronic obstructive pulmonary disease, rheumatologic disease, peptic ulcer disease, mild liver disease, diabetes mellitus (uncomplicated), diabetes with chronic complications, hemiplegia or paraplegia, renal disease, any malignancy, liver disease, metastatic solid tumor, and acquired immunodeficiency syndrome. Other baseline diagnoses included in the analysis were atrial fibrillation, trauma, stroke, venous thromboembolism, coagulation defects (including an abnormal coagulation profile or symptomatic hemophilia A carrier status), GI (i.e., gastric, duodenal, peptic, gastrojejunal) ulcers, and inflammatory bowel disease. Descriptive analyses of blood product and medication use during the hospitalization included the number of units of FFP received and the percentages of patients receiving red blood cells, cryoprecipitate, platelets, diuretics, phytonadione, and warfarin. INR values were evaluated from the day before FFP administration through the day of the last available INR test result during the hospital stay, with specific measures including the INR value at the time of the first elevated result prior to FFP administration and the latest documented INR value during the hospitalization. INR correction was defined as any INR value during the hospitalization of 1.5 or less; this definition was based on various published guidelines for the correction of anticoagulant-associated major bleeding.21,–23 Among the patients in whom INR correction was achieved, the number of INR tests before correction and the number of days from FFP administration to INR correction were assessed. Additionally, the mean number of hours to INR correction was analyzed among the subset of patients for whom hourly laboratory and medication administration data (specifically, INR values and FFP administration times) were available. Finally, rates of inhospital mortality, stratified by bleed type and INR correction status, were examined. Analysis Descriptive analyses of all study measures were performed and reported as stratified by ICH and non-ICH bleed types, with most measures further stratified by INR correction status. Binary variables were summarized using percentages of patients, and continuous variables were summarized using mean ± S.D. and median values. Nonparametric rank-order Wilcoxon tests were used for statistical comparison of mean hospital length of stay, and Fisher’s exact test was used for comparison of inhospital mortality rates according to INR correction status. Kaplan–Meier curves were generated to compare time-to-correction values (i.e., days from FFP administration to correction), with data censorship for patient death, by bleeding type. In the subset of patients for whom hourly time-stamp data were available, Kaplan–Meier curves were generated to compare time-to-correction values (i.e., hours from FFP administration to correction) by bleeding type. A Cox proportional hazards model was constructed to assess factors associated with the time to INR correction. Specific covariates included patient age range (65–74 or ≥75 years versus the reference value of <65 years), sex, baseline DM-CCI score (>4 versus reference of ≤4), baseline diagnosis of trauma, warfarin use during the hospital admission, first INR value upon presentation with elevated INR (>3.5 versus reference of ≤3.5), mean number of FFP units administered (>4 versus ≤4), phytonadione use during the hospitalization episode, and bleeding type. Two additional Cox proportional hazards models (for ICH versus non-ICH bleeds) were constructed to assess factors associated with the time to inhospital mortality; specific covariates for this analysis were the same as those listed above, with the addition of INR correction status. Differences in baseline demographic and clinical characteristics between patients with ICH and patients with non-ICH bleeding were tested for statistical significance, with rank-order Wilcoxon tests used for continuous variables and Fisher’s exact test used for comparisons of proportions. Additionally, within each bleeding-type category, significance testing was conducted for differences in blood products administered and laboratory testing between patients who experienced and patients who did not experience INR correction. All analyses were carried out with INR correction defined as an INR of ≤1.5; however, sensitivity analyses were performed using an alternative definition (i.e., a value of ≤1.3). This alternative definition of INR correction was explored in recognition of variability in practice patterns dependent on physician preference and patient comorbidities. All analyses were completed with SAS, version 9.3 (SAS Institute, Cary, NC). The a priori level of significance was set at 0.05. Results From an initial sample of 6215 patients who received FFP during the period January 2004–January 2010, 354 patients met all inclusion criteria (Figure 1). Of those 354 patients, 120 (33.9%) had an ICH major bleed and 234 (66.1%) had a non-ICH major bleed. The overall mean age was 74.7 years, and nearly all patients (over 99%) were white. The baseline demographics and clinical characteristics of the study patients, by bleed type, are provided in Table 1. Table 1 Baseline Demographic and Clinical Characteristics of Study Population, by Bleeding Typea Characteristic Overall (n = 354) ICH Bleed (n = 120) Non-ICH Bleed (n = 234) pb Age range, yr (%)  <65 16.4 12.5 18.4  65–74 22.0 20.8 22.6  ≥75 61.6 66.7 59.0 Mean ± S.D. age, yr 74.7 ± 11.1 76.1 ± 10.4 74.0 ± 11.4 0.099 Median age, yr 77 78 77 Female (%) 46.9 50.0 45.3 0.416 Race/ethnicity (%)  White/Caucasian 99.2 98.3 99.6 0.023  Black/African-American 0.6 0.8 0.4  Hispanic 0.3 0.8 0.0 DM-CCI score  Mean ± S.D. 2.52 ± 2.65 1.65 ± 2.30 2.97 ± 2.71 <0.001  Median 2.00 1.00 2.00 Other diagnoses (%)  Atrial fibrillation 46.9 40.0 50.4 0.072  Trauma 18.6 15.8 20.1 0.388  Stroke 11.6 6.7 14.1 0.052  Venous thromboembolism 11.3 7.5 13.2 0.114  Coagulation defectsc 6.5 2.5 8.5 0.038  GI ulcerd 4.0 1.7 5.1 0.153  Inflammatory bowel disease 1.7 0.0 2.6 0.100 Characteristic Overall (n = 354) ICH Bleed (n = 120) Non-ICH Bleed (n = 234) pb Age range, yr (%)  <65 16.4 12.5 18.4  65–74 22.0 20.8 22.6  ≥75 61.6 66.7 59.0 Mean ± S.D. age, yr 74.7 ± 11.1 76.1 ± 10.4 74.0 ± 11.4 0.099 Median age, yr 77 78 77 Female (%) 46.9 50.0 45.3 0.416 Race/ethnicity (%)  White/Caucasian 99.2 98.3 99.6 0.023  Black/African-American 0.6 0.8 0.4  Hispanic 0.3 0.8 0.0 DM-CCI score  Mean ± S.D. 2.52 ± 2.65 1.65 ± 2.30 2.97 ± 2.71 <0.001  Median 2.00 1.00 2.00 Other diagnoses (%)  Atrial fibrillation 46.9 40.0 50.4 0.072  Trauma 18.6 15.8 20.1 0.388  Stroke 11.6 6.7 14.1 0.052  Venous thromboembolism 11.3 7.5 13.2 0.114  Coagulation defectsc 6.5 2.5 8.5 0.038  GI ulcerd 4.0 1.7 5.1 0.153  Inflammatory bowel disease 1.7 0.0 2.6 0.100 a ICH = intracranial hemorrhage, DM-CCI = Deyo-modified Charlson Comorbidity Index, GI = gastrointestinal. b Means tested with the nonparametric rank-order Wilcoxon test; proportions tested with Fisher’s exact test. c Includes history of International Classification of Diseases, 9th Revision, Clinical Modification diagnosis for coagulation defects (code 286.xx), abnormal coagulation profile (790.92), or symptomatic hemophilia A carrier (V83.02). d Includes history of gastric, duodenal, peptic, or gastrojejunal ulcers. Open in new tab Table 1 Baseline Demographic and Clinical Characteristics of Study Population, by Bleeding Typea Characteristic Overall (n = 354) ICH Bleed (n = 120) Non-ICH Bleed (n = 234) pb Age range, yr (%)  <65 16.4 12.5 18.4  65–74 22.0 20.8 22.6  ≥75 61.6 66.7 59.0 Mean ± S.D. age, yr 74.7 ± 11.1 76.1 ± 10.4 74.0 ± 11.4 0.099 Median age, yr 77 78 77 Female (%) 46.9 50.0 45.3 0.416 Race/ethnicity (%)  White/Caucasian 99.2 98.3 99.6 0.023  Black/African-American 0.6 0.8 0.4  Hispanic 0.3 0.8 0.0 DM-CCI score  Mean ± S.D. 2.52 ± 2.65 1.65 ± 2.30 2.97 ± 2.71 <0.001  Median 2.00 1.00 2.00 Other diagnoses (%)  Atrial fibrillation 46.9 40.0 50.4 0.072  Trauma 18.6 15.8 20.1 0.388  Stroke 11.6 6.7 14.1 0.052  Venous thromboembolism 11.3 7.5 13.2 0.114  Coagulation defectsc 6.5 2.5 8.5 0.038  GI ulcerd 4.0 1.7 5.1 0.153  Inflammatory bowel disease 1.7 0.0 2.6 0.100 Characteristic Overall (n = 354) ICH Bleed (n = 120) Non-ICH Bleed (n = 234) pb Age range, yr (%)  <65 16.4 12.5 18.4  65–74 22.0 20.8 22.6  ≥75 61.6 66.7 59.0 Mean ± S.D. age, yr 74.7 ± 11.1 76.1 ± 10.4 74.0 ± 11.4 0.099 Median age, yr 77 78 77 Female (%) 46.9 50.0 45.3 0.416 Race/ethnicity (%)  White/Caucasian 99.2 98.3 99.6 0.023  Black/African-American 0.6 0.8 0.4  Hispanic 0.3 0.8 0.0 DM-CCI score  Mean ± S.D. 2.52 ± 2.65 1.65 ± 2.30 2.97 ± 2.71 <0.001  Median 2.00 1.00 2.00 Other diagnoses (%)  Atrial fibrillation 46.9 40.0 50.4 0.072  Trauma 18.6 15.8 20.1 0.388  Stroke 11.6 6.7 14.1 0.052  Venous thromboembolism 11.3 7.5 13.2 0.114  Coagulation defectsc 6.5 2.5 8.5 0.038  GI ulcerd 4.0 1.7 5.1 0.153  Inflammatory bowel disease 1.7 0.0 2.6 0.100 a ICH = intracranial hemorrhage, DM-CCI = Deyo-modified Charlson Comorbidity Index, GI = gastrointestinal. b Means tested with the nonparametric rank-order Wilcoxon test; proportions tested with Fisher’s exact test. c Includes history of International Classification of Diseases, 9th Revision, Clinical Modification diagnosis for coagulation defects (code 286.xx), abnormal coagulation profile (790.92), or symptomatic hemophilia A carrier (V83.02). d Includes history of gastric, duodenal, peptic, or gastrojejunal ulcers. Open in new tab Figure 1 Open in new tabDownload slide Patient selection process. FFP = fresh frozen plasma, INR = International Normalized Ratio. Figure 1 Open in new tabDownload slide Patient selection process. FFP = fresh frozen plasma, INR = International Normalized Ratio. Overall, atrial fibrillation was the most prevalent baseline diagnosis (46.9%), followed by trauma (18.6%) and stroke (11.6%). The mean DM-CCI score was 2.52 among the overall study cohort but significantly lower for patients presenting with ICH versus non-ICH bleeding (1.65 versus 2.97, p < 0.001). The prevalence of other baseline diagnoses was similar among patients with ICH versus non-ICH bleeding, excluding the diagnosis of coagulation defects, which was significantly less common among patients with ICH versus non-ICH bleeding (2.5% versus 8.5%, p = 0.038). INR correction was achieved in the majority of patients (87.9% overall) at some point during hospitalization, with rates of 92.5% and 85.5% in patients with ICH bleeds and non-ICH bleeds, respectively (p = 0.060). In the sensitivity analysis using a narrower definition of INR correction (i.e., a cutoff value of ≤1.3), INR correction was achieved in 68.6% of patients overall (79.2% of those with ICH bleeds versus 63.7% of those with non-ICH bleeds, p = 0.004). On average, patients who experienced INR correction received more units of FFP than those whose INR remained uncorrected (5.2 units versus 3.7 units for those with ICH bleeds, p = 0.024; 4.2 units versus 3.8 units for those with non-ICH bleeds, p = 0.944), as shown in Table 2. Similarly, for both bleed types, a higher proportion of patients who experienced INR correction received red blood cells relative to patients whose INR remained uncorrected (30% versus 11% for ICH bleeds and 84% versus 71% for non-ICH bleeds). Compared with patients whose INR remained uncorrected, a higher proportion of patients who experienced INR correction received diuretics (74% versus 67% for ICH bleeds and 70% versus 68% for non-ICH bleeds), and a lower proportion had evidence of warfarin use during the hospitalization episode (50% versus 56% for ICH bleeds and 41% versus 47% for non-ICH bleeds). In general, none of these differences were significant (Table 2). Very few patients (<6%) received cryoprecipitate or platelets, regardless of INR correction status. No patients received PCCs or factor VIIa. Table 2 Use of Blood Products and Other Medications in Study Population, by Bleeding Type and INR Correction Status Blood Product/Medication Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb FFP units per patient  Mean ± S.D. 4.55 ±3.77 3.79 ± 2.63 0.089 5.16 ±3.67 3.67 ± 4.03 0.024 4.21 ±3.78 3.82 ± 2.21 0.944  Median 4 3 4 2 4 3 Red blood cells (%) 64.60 58.10 0.403 29.70 11.10 0.443 84.00 70.60 0.087 Cryoprecipitate (%) 0.60 0 1.000 0 0 …c 1.00 0 1.000 Platelets (%) 5.80 4.70 1.000 6.30 0 1.000 5.50 5.90 1.000 Diuretics (%) 71.10 67.40 0.598 73.90 66.70 0.699 69.50 67.60 0.842 Phytonadione (%) 60.50 51.20 0.251 61.30 66.70 1.000 60.00 47.10 0.189 Warfarin (%) 43.70 48.80 0.622 49.50 55.60 0.731 40.50 47.10 0.851 Blood Product/Medication Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb FFP units per patient  Mean ± S.D. 4.55 ±3.77 3.79 ± 2.63 0.089 5.16 ±3.67 3.67 ± 4.03 0.024 4.21 ±3.78 3.82 ± 2.21 0.944  Median 4 3 4 2 4 3 Red blood cells (%) 64.60 58.10 0.403 29.70 11.10 0.443 84.00 70.60 0.087 Cryoprecipitate (%) 0.60 0 1.000 0 0 …c 1.00 0 1.000 Platelets (%) 5.80 4.70 1.000 6.30 0 1.000 5.50 5.90 1.000 Diuretics (%) 71.10 67.40 0.598 73.90 66.70 0.699 69.50 67.60 0.842 Phytonadione (%) 60.50 51.20 0.251 61.30 66.70 1.000 60.00 47.10 0.189 Warfarin (%) 43.70 48.80 0.622 49.50 55.60 0.731 40.50 47.10 0.851 a ICH = intracranial hemorrhage, INR= International Normalized Ratio, FFP = fresh frozen plasma. b Means tested with the non parametric rank-order Wilcoxon test; proportions tested with Fisher’s exact test. c Not applicable. Open in new tab Table 2 Use of Blood Products and Other Medications in Study Population, by Bleeding Type and INR Correction Status Blood Product/Medication Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb FFP units per patient  Mean ± S.D. 4.55 ±3.77 3.79 ± 2.63 0.089 5.16 ±3.67 3.67 ± 4.03 0.024 4.21 ±3.78 3.82 ± 2.21 0.944  Median 4 3 4 2 4 3 Red blood cells (%) 64.60 58.10 0.403 29.70 11.10 0.443 84.00 70.60 0.087 Cryoprecipitate (%) 0.60 0 1.000 0 0 …c 1.00 0 1.000 Platelets (%) 5.80 4.70 1.000 6.30 0 1.000 5.50 5.90 1.000 Diuretics (%) 71.10 67.40 0.598 73.90 66.70 0.699 69.50 67.60 0.842 Phytonadione (%) 60.50 51.20 0.251 61.30 66.70 1.000 60.00 47.10 0.189 Warfarin (%) 43.70 48.80 0.622 49.50 55.60 0.731 40.50 47.10 0.851 Blood Product/Medication Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb FFP units per patient  Mean ± S.D. 4.55 ±3.77 3.79 ± 2.63 0.089 5.16 ±3.67 3.67 ± 4.03 0.024 4.21 ±3.78 3.82 ± 2.21 0.944  Median 4 3 4 2 4 3 Red blood cells (%) 64.60 58.10 0.403 29.70 11.10 0.443 84.00 70.60 0.087 Cryoprecipitate (%) 0.60 0 1.000 0 0 …c 1.00 0 1.000 Platelets (%) 5.80 4.70 1.000 6.30 0 1.000 5.50 5.90 1.000 Diuretics (%) 71.10 67.40 0.598 73.90 66.70 0.699 69.50 67.60 0.842 Phytonadione (%) 60.50 51.20 0.251 61.30 66.70 1.000 60.00 47.10 0.189 Warfarin (%) 43.70 48.80 0.622 49.50 55.60 0.731 40.50 47.10 0.851 a ICH = intracranial hemorrhage, INR= International Normalized Ratio, FFP = fresh frozen plasma. b Means tested with the non parametric rank-order Wilcoxon test; proportions tested with Fisher’s exact test. c Not applicable. Open in new tab On average, the first elevated INR test result before FFP administration was similar among patients who achieved INR correction at any point during the hospitalization and among those whose INR remained uncorrected, with mean ± S.D. INR values of 3.4 ± 2.5 versus 3.2 ± 1.1 for ICH bleeds and 4.1 ± 2.0 versus 4.0 ± 1.5 for non-ICH bleeds, respectively (Table 3). Among patients in whom INR correction was achieved, the mean ± S.D. number of days from FFP administration to INR correction was slightly shorter for those with ICH bleeds (0.9 ± 0.9 day) versus non-ICH bleeds (1.8 ± 2.7 days). The mean ± S.D. number of INR tests between FFP administration and INR correction was slightly lower for patients with ICH bleeds (3.3 ± 1.9) versus non-ICH bleeds (4.6 ± 4.4). In the subset of patients with available hourly laboratory time-stamp data (n = 65), the mean time from FFP administration to INR correction (unadjusted for deaths) was shorter for those with ICH versus non-ICH bleeding (16.4 hours versus 55.2 hours). Table 3 INR Testing and Results in Study Population, by ICH Status and INR Correction Status3 Variable Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb INR value at first elevated result (INR of >2) prior to FFP use  Mean ± S.D. 3.84 ±2.18 3.85 ± 1.46 0.271 3.42 ± 2.48 3.19 ±1.07 0.964 4.07 ± 1.96 4.03 ±1.52 0.456  Median 3.13 3.56 2.81 2.86 3.38 3.9 INR value at last available result during hospitalization  Mean ± S.D. 1.37 ±0.44 1.84 ± 0.34 <0.001 1.28 ±0.33 1.92 ±0.35 <0.001 1.41 ± 0.48 1.82 ±0.34 <0.001  Median 1.24 1.76 1.2 1.91 1.26 1.7 No. tests until INR in target range  Mean ± S.D. 4.12 ±3.74 …c 3.32 ±1.87 … 4.57 ±4.4 …  Median 3 … 3 … 4 … Hours to INR in target range after FFP use 65 23 42  No. patients evaluated   Mean ± S.D. 41.45 ± 97.06 … 16.36 ±15.78 … 55.19 ± 118.43 …   Median 18.70 … 12.60 … 22.75 … Days to INR in target range from date of first FFP use  Mean ± S.D. 1.44 ± 2.29 … 0.88 ± 0.85 … 1.76 ± 2.74 …  Median 1 … 1 … 1 … Days to INR in target range from date of first FFP use (% of patients)  0 25.7 … 33.3 … 21.5 …  1 43.1 … 50.5 … 39.0 …  2 18.0 … 10.8 … 22.0 …  3 6.8 … 4.5 … 8.0 …  4 2.6 … 0.9 … 3.5 …  5 1.9 … 0.0 … 3.0 …  6 0.3 … 0.0 … 0.5 …  >7 1.6 … 0.0 … 2.5 … Variable Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb INR value at first elevated result (INR of >2) prior to FFP use  Mean ± S.D. 3.84 ±2.18 3.85 ± 1.46 0.271 3.42 ± 2.48 3.19 ±1.07 0.964 4.07 ± 1.96 4.03 ±1.52 0.456  Median 3.13 3.56 2.81 2.86 3.38 3.9 INR value at last available result during hospitalization  Mean ± S.D. 1.37 ±0.44 1.84 ± 0.34 <0.001 1.28 ±0.33 1.92 ±0.35 <0.001 1.41 ± 0.48 1.82 ±0.34 <0.001  Median 1.24 1.76 1.2 1.91 1.26 1.7 No. tests until INR in target range  Mean ± S.D. 4.12 ±3.74 …c 3.32 ±1.87 … 4.57 ±4.4 …  Median 3 … 3 … 4 … Hours to INR in target range after FFP use 65 23 42  No. patients evaluated   Mean ± S.D. 41.45 ± 97.06 … 16.36 ±15.78 … 55.19 ± 118.43 …   Median 18.70 … 12.60 … 22.75 … Days to INR in target range from date of first FFP use  Mean ± S.D. 1.44 ± 2.29 … 0.88 ± 0.85 … 1.76 ± 2.74 …  Median 1 … 1 … 1 … Days to INR in target range from date of first FFP use (% of patients)  0 25.7 … 33.3 … 21.5 …  1 43.1 … 50.5 … 39.0 …  2 18.0 … 10.8 … 22.0 …  3 6.8 … 4.5 … 8.0 …  4 2.6 … 0.9 … 3.5 …  5 1.9 … 0.0 … 3.0 …  6 0.3 … 0.0 … 0.5 …  >7 1.6 … 0.0 … 2.5 … a INR = International Normalized Ratio, ICH = intracranial hemorrhage, FFP = fresh frozen plasma. b Means tested with the nonparametric rank-order Wilcoxon test. c Not applicable. Open in new tab Table 3 INR Testing and Results in Study Population, by ICH Status and INR Correction Status3 Variable Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb INR value at first elevated result (INR of >2) prior to FFP use  Mean ± S.D. 3.84 ±2.18 3.85 ± 1.46 0.271 3.42 ± 2.48 3.19 ±1.07 0.964 4.07 ± 1.96 4.03 ±1.52 0.456  Median 3.13 3.56 2.81 2.86 3.38 3.9 INR value at last available result during hospitalization  Mean ± S.D. 1.37 ±0.44 1.84 ± 0.34 <0.001 1.28 ±0.33 1.92 ±0.35 <0.001 1.41 ± 0.48 1.82 ±0.34 <0.001  Median 1.24 1.76 1.2 1.91 1.26 1.7 No. tests until INR in target range  Mean ± S.D. 4.12 ±3.74 …c 3.32 ±1.87 … 4.57 ±4.4 …  Median 3 … 3 … 4 … Hours to INR in target range after FFP use 65 23 42  No. patients evaluated   Mean ± S.D. 41.45 ± 97.06 … 16.36 ±15.78 … 55.19 ± 118.43 …   Median 18.70 … 12.60 … 22.75 … Days to INR in target range from date of first FFP use  Mean ± S.D. 1.44 ± 2.29 … 0.88 ± 0.85 … 1.76 ± 2.74 …  Median 1 … 1 … 1 … Days to INR in target range from date of first FFP use (% of patients)  0 25.7 … 33.3 … 21.5 …  1 43.1 … 50.5 … 39.0 …  2 18.0 … 10.8 … 22.0 …  3 6.8 … 4.5 … 8.0 …  4 2.6 … 0.9 … 3.5 …  5 1.9 … 0.0 … 3.0 …  6 0.3 … 0.0 … 0.5 …  >7 1.6 … 0.0 … 2.5 … Variable Overall ICH Bleed Non-ICH Bleed INR Corrected (n = 311) INR Not Corrected (n = 43) pb INR Corrected (n = 111) INR Not Corrected (n = 9) pb INR Corrected (n = 200) INR Not Corrected (n = 34) pb INR value at first elevated result (INR of >2) prior to FFP use  Mean ± S.D. 3.84 ±2.18 3.85 ± 1.46 0.271 3.42 ± 2.48 3.19 ±1.07 0.964 4.07 ± 1.96 4.03 ±1.52 0.456  Median 3.13 3.56 2.81 2.86 3.38 3.9 INR value at last available result during hospitalization  Mean ± S.D. 1.37 ±0.44 1.84 ± 0.34 <0.001 1.28 ±0.33 1.92 ±0.35 <0.001 1.41 ± 0.48 1.82 ±0.34 <0.001  Median 1.24 1.76 1.2 1.91 1.26 1.7 No. tests until INR in target range  Mean ± S.D. 4.12 ±3.74 …c 3.32 ±1.87 … 4.57 ±4.4 …  Median 3 … 3 … 4 … Hours to INR in target range after FFP use 65 23 42  No. patients evaluated   Mean ± S.D. 41.45 ± 97.06 … 16.36 ±15.78 … 55.19 ± 118.43 …   Median 18.70 … 12.60 … 22.75 … Days to INR in target range from date of first FFP use  Mean ± S.D. 1.44 ± 2.29 … 0.88 ± 0.85 … 1.76 ± 2.74 …  Median 1 … 1 … 1 … Days to INR in target range from date of first FFP use (% of patients)  0 25.7 … 33.3 … 21.5 …  1 43.1 … 50.5 … 39.0 …  2 18.0 … 10.8 … 22.0 …  3 6.8 … 4.5 … 8.0 …  4 2.6 … 0.9 … 3.5 …  5 1.9 … 0.0 … 3.0 …  6 0.3 … 0.0 … 0.5 …  >7 1.6 … 0.0 … 2.5 … a INR = International Normalized Ratio, ICH = intracranial hemorrhage, FFP = fresh frozen plasma. b Means tested with the nonparametric rank-order Wilcoxon test. c Not applicable. Open in new tab With data censorship for patient deaths, the time to INR correction did not differ by bleeding type (a median of 1 day); however, in the sensitivity analysis using the narrow definition of INR correction (i.e., a value of ≤1.3), the median time to correction was significantly shorter for patients with ICH versus non-ICH bleeds (approximately 2 days versus approximately 3 days, p < 0.001). In the 65 patients for whom hourly laboratory time-stamp data were available, the median time to achieve an INR of ≤1.5, with censoring of data for patient death, was shorter for those with ICH versus non-ICH bleeds (16 hours versus 23 hours, p = 0.039), as shown in Figure 2. In the sensitivity analysis using the stricter definition of INR correction, a similar trend was noted, with median times to INR correction of 26 and 35 hours for patients with ICH and non-ICH bleeds, respectively (p = 0.387). Figure 2 Open in new tabDownload slide Kaplan–Meier curves depicting the likelihood of elevated International Normalized Ratio (INR) values remaining uncorrected (i.e., >1.5) at various time points after administration of fresh frozen plasma (FFP) in subsets of patients with intracranial hemorrhage (ICH) (n = 23) and non-ICH major bleeding (n = 42), represented by the yellow and red lines, respectively. The median time to INR correction (i.e., a value of ≤1.5) was significantly shorter in patients with ICH versus non-ICH bleeds (Wilcoxon p value, 0.039). Figure 2 Open in new tabDownload slide Kaplan–Meier curves depicting the likelihood of elevated International Normalized Ratio (INR) values remaining uncorrected (i.e., >1.5) at various time points after administration of fresh frozen plasma (FFP) in subsets of patients with intracranial hemorrhage (ICH) (n = 23) and non-ICH major bleeding (n = 42), represented by the yellow and red lines, respectively. The median time to INR correction (i.e., a value of ≤1.5) was significantly shorter in patients with ICH versus non-ICH bleeds (Wilcoxon p value, 0.039). Results from a Cox proportional hazards model evaluating predictors of the time to achieve INR correction (controlling for patient age, sex, DM-CCI score, baseline diagnosis of trauma, warfarin or phytonadione use during the episode, presenting INR value, and number of FFP units received) showed that presentation with an ICH bleed was associated with a higher probability of INR correction; however, results were not significant at the 95% confidence level (hazard ratio [HR], 1.29; 95% confidence interval [CI], 0.99–1.65; p = 0.051). In a sensitivity analysis using the same Cox model but with INR correction defined as the attainment of a value of ≤1.3, presentation with an ICH bleed was a significant predictor of achieving INR correction (HR, 1.41; 95% CI, 1.06–1.86; p = 0.018). Compared with patients whose INR remained uncorrected, fewer patients with corrected INR values died during their hospital stay. Overall, 6.8% of patients experiencing INR correction died during the hospitalization, compared with 20.9% of patients whose INR was not corrected (p < 0.001); among patients with an ICH bleed, 15.3% of patients in whom INR correction was achieved died, compared with 55.6% of patients with uncorrected INR levels (p = 0.010). Similarly, among patients with a non-ICH bleed, 2.0% of patients in the INR-corrected group died during the hospital stay, compared with 11.8% of patients whose INR was uncorrected (p = 0.017). In a sensitivity analysis using the narrower definition of INR correction, inhospital mortality was improved relative to mortality determined using a cutoff value of ≤1.5. Overall, 6.6% of patients in whom an INR of ≤1.3 was achieved died during the hospitalization, compared with 12.7% of patients whose INR remained uncorrected per that stricter definition (p = 0.001). Two Cox proportional hazards models (one for ICH and one for non-ICH bleeds) were conducted to evaluate potential predictors of time to inhospital mortality (controlling for patient age, sex, DM-CCI score, baseline diagnosis of trauma, warfarin or phytonadione use, presenting INR level, and number of FFP units received). The results of Cox modeling indicated that failure to achieve INR correction during the hospital stay (per the less strict cutoff value of ≤1.5) was associated with a significantly higher risk of death for patients presenting with an ICH bleed (HR, 8.04; 95% CI, 2.07–31.18; p = 0.003). When the Cox model for predicting time to inhospital mortality among patients presenting with a non-ICH bleed was conducted, the likelihood ratio test exceeded 0.05, indicating that the model was not an adequate fit (i.e., not sufficiently predictive of the dependent variable of inhospital mortality). Discussion After applying the patient selection criteria, we identified a total of 354 patients who received FFP for correction of warfarin-related major bleeding. Overall and among both ICH and non-ICH subgroups, most patients experienced INR correction at some point during their hospital stay. Among patients presenting with an ICH bleed, those whose INR value was corrected (to ≤1.5) had a significantly lower risk of inhospital death relative to patients whose INR remained uncorrected, even with controlling for baseline patient demographics and clinical characteristics; the adjusted association between INR correction and inhospital mortality was not significant among patients presenting with non-ICH bleeds. Inhospital mortality rates among those experiencing INR correction, as defined using the stricter cutoff value (i.e., an INR of ≤1.3) were numerically lower than mortality rates among patients whose INR was lowered to values of ≤1.5. These findings suggest that there is clinical utility in the rapid correction of anticoagulant-associated major bleeding, specifically in relation to short-term outcomes such as inhospital mortality, among patients presenting with ICH. Results from a review of a prospective, multicenter registry of patients hospitalized with warfarin-associated ICH showed mortality rates similar to those in the ICH group in our study. Mortality rates by specific ICH diagnosis, as documented in the registry, ranged from 12.5% for patients with subarachnoid hemorrhage to 42.3% among those with intraparenchymal hemorrhage, compared with a mortality rate of 18.3% among patients with an ICH bleed in our study.24 Additionally, the mortality rate among patients with non-ICH bleeding in our study (3.4%, including patients with GI bleeding) was similar to the inhospital mortality rate observed in a study of 111 patients receiving anticoagulant therapy who were hospitalized for upper GI bleeding (3.6%).25 The notion that uncorrected INR values are associated with greater mortality risk was suggested in previous studies of patients with anticoagulant-related major bleeding. Results from a small observational study of seven patients presenting to an intensive care unit with warfarin-related ICH showed that the median time to INR correction was shorter for patients surviving the hospitalization than for those who died (85 minutes versus 10 hours)26; this finding, although based on a small sample size, was in agreement with our findings that uncorrected INR values may be associated with increased mortality risk (specifically inhospital mortality). In an analysis of stroke registry data, inhospital mortality was compared among patients transferred to a comprehensive stroke center from community EDs with warfarin-associated ICH who received “early reversal” versus “nonearly reversal” therapy for INR elevations (early reversal was defined as receipt of at least 3 units of FFP or any factor VIIa use in the ED, and nonearly reversal was defined as receiving less than 3 units of FFP or no factor VIIa use). The results showed a numerically lower inhospital mortality rate among patients in the early-reversal group relative to the other patients (30% versus 39%, p = 0.72).27 The observed trend of reduced mortality risk with earlier INR reversal parallels results from our study (i.e., lower inhospital mortality among patients with ICH whose INR was corrected to a value of ≤1.5); however, our findings were significant at the 95% confidence level. A previously published study based on analyses of the same EHR data set used in the study described here evaluated the risk of 30-day mortality after hospital admission for warfarin-associated major bleeding using a more strict definition of INR correction (i.e., achieving an INR value of ≤1.3 within 48 hours of FFP administration).17 Results from that study indicated that INR correction was associated with a lower mortality risk, although the correlation was significant only for patients with ICH bleeds; that result was congruous with findings in the analyses presented here, which indicated that achieving INR correction was significantly associated with a lower risk of inhospital mortality among patients with ICH bleeds when INR correction was defined as attainment of an INR of ≤1.5. Our study entailed limitations common to all studies that rely on the retrospective review of medical records data, including potential coding errors.28 In our study, INR correction was defined a priori as a dichotomous variable using a threshold value (INR of ≤1.5) based on various guidelines for reversal of anticoagulant-related major bleeding21,–23; that definition limits the generalizability of the findings, as we recognize that the target value for correction of an elevated INR may not be the same for all patients due to underlying risk factors and concerns over volume overload. In order to account for variability in the definition of INR correction dependent on patient and provider characteristics and preferences, we performed a sensitivity analysis in which INR correction was defined by an INR of ≤1.3. In our analysis of time from FFP administration to INR correction, only 18% of patients (n = 65) had hourly time-stamp data available for both FFP administration and INR laboratory result measures. It is possible that patients with hourly time-stamp data may have differed from those without such data. Direct clinical markers of bleeding severity such as hematoma size and Glasgow Coma Scale scores were not available in the evaluated EHR data set; therefore, baseline diagnosis of trauma, number of FFP units received, and initial elevated INR level upon presentation were used in multivariate models to control for the severity of the bleed at the time of presentation. In another study limitation, the analysis of the mean number of INR tests ordered prior to INR correction (if correction was achieved) did not take into account variability in physician and hospital practice patterns regarding the frequency of laboratory testing. Therefore, that study measure should be considered purely descriptive and evaluated with this limitation in mind. Also, we did not evaluate the specific surgical interventions performed during the sampled hospitalization episodes, some of which (e.g., craniotomy, partial gastrectomy) could have affected the evaluated mortality outcomes. Moreover, any care that was received outside of the integrated care system was not captured in the EHR data set, potentially resulting in the exclusion of patients who may have been hospitalized in another network. Additionally, the EHR data used in this analysis may not be representative of care provided in all regions of the United States. Finally, no specific cause-of-death information was available in this data set; thus, any inhospital deaths may have been due to causes unrelated to the major bleeding episode. This study provided background on current clinical practice patterns regarding FFP administration, resulting INR correction or lack thereof, and the association between achieving corrected INR levels and inhospital mortality. With the introduction of alternatives to FFP to reverse anticoagulant-related major bleeding, with some potentially providing faster INR reversal, more research is needed on the association between time to correction and patient outcomes such as mortality and morbidity outcomes. Follow-on research should be conducted using larger overall sample sizes and methodology providing for greater availability of hourly laboratory data on the times from hospital admission to FFP administration and INR correction, as well as information on postdischarge treatment. Conclusion Results of this study indicated that INR value correction was associated with a lower likelihood of inhospital death among warfarin-treated patients hospitalized for ICH or non-ICH major bleeding. Footnotes Jordan Menzin, B.A., is acknowledged for assistance with computer programming. This research was performed by Boston Health Economics, Inc., and was funded by CSL Behring, Inc. Both Boston Health Economics, Inc. and CSL Behing, Inc. participated in the study design, interpretation of results, and final preparation of this manuscript. The study described herein was sponsored by CSL Behring. Drs. Menzin and Friedman, Mr. Sussman, and Ms. Nichols received funding from the study sponsor. The other author has declared no potential conflicts of interest. References 1 Pollock BE . Clinical experience with warfarin (Coumadin) sodium, a new anticoagulant . J Am Med Assoc . 1955 ; 159 : 1094 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Johnson JA . Warfarin pharmacogenetics: a rising tide for its clinical value . Circulation . 2012 ; 125 : 1964 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Kirley K Qato DM Kornfield R et al. . National trends in oral anticoagulant use in the United States, 2007 to 2011 . Circ Cardiovasc Qual Outcomes . 2012 ; 5 : 615 – 21 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Wysowski DK Nourjah P Swartz L . Bleeding complications with warfarin use: a prevalent adverse effect resulting in regulatory action . Arch Intern Med . 2007 ; 167 : 1414 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Institute for Safe Medication Practices . QuarterWatch: Anticoagulants the leading reported drug risk in 2011 ( May 31 , 2012 ). www.ismp.org/quarterwatch/pdfs/2011Q4.pdf (accessed 2014 Jun 18). 6 Moore TJ Bennett CL . Underreporting of hemorrhagic and thrombotic complications of pharmaceuticals to the U.S. Food and Drug Administration: empirical findings for warfarin, clopidogrel, ticlopidine, and thalidomide from the Southern Network on Adverse Reactions (SONAR) . Semin Thromb Hemost . 2012 ; 38 : 905 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Sakaeda T Tamon A Kadoyama K et al. . Data mining of the public version of the FDA Adverse Event Reporting System . Int J Med Sci . 2013 ; 10 : 796 – 803 . Google Scholar Crossref Search ADS PubMed WorldCat 8 Atrial Fibrillation Investigators . Risk factors for stroke and efficacy of antithrombotic therapy in atrial fibrillation: analysis of pooled data from five randomized controlled trials . Arch Intern Med . 1994 ; 154 : 1449 – 57 . Crossref Search ADS PubMed WorldCat 9 Beyth RJ Quinn LM Landefeld CS . Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin . Am J Med . 1998 ; 105 : 91 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Fihn SD Callahan CM Martin DC et al. . The risk for and severity of bleeding complications in elderly patients treated with warfarin. The National Consortium of Anticoagulation Clinics . Ann Intern Med . 1996 ; 124 : 970 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Gomes T Mamdani MM Holbrook AM et al. . Rates of hemorrhage during warfarin therapy for atrial fibrillation . CMAJ . 2013 ; 185 : E121 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 12 Landefeld CS Beyth RJ . Anticoagulant-related bleeding: clinical epidemiology, prediction, and prevention . Am J Med . 1993 ; 95 : 315 – 28 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Palareti G Leali N Coccheri S et al. . Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy . Lancet . 1996 ; 348 : 423 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 14 Wiedermann CJ Stockner I . Warfarin-induced bleeding complications—clinical presentation and therapeutic options . Thromb Res . 2008 ; 122 ( suppl 2 ): S13 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 15 Huttner HB Schellinger PD Hartmann M et al. . Hematoma growth and outcome in treated neurocritical care patients with intracerebral hemorrhage related to oral anticoagulant therapy: comparison of acute treatment strategies using vitamin K, fresh frozen plasma, and prothrombin complex concentrates . Stroke . 2006 ; 37 : 1465 – 70 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Flibotte JJ Hagan N O’Donnell J et al. . Warfarin, hematoma expansion, and outcome of intracerebral hemorrhage . Neurology . 2004 ; 63 : 1059 – 64 . Google Scholar Crossref Search ADS PubMed WorldCat 17 Menzin J Hoesche J Friedman M et al. . Failure to correct International Normalized Ratio and mortality among patients with warfarin-related major bleeding: an analysis of electronic health records . J Thromb Haemost . 2012 ; 10 : 596 – 605 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Romano PS Roos LL Jollis JG . Adapting a clinical comorbidity index for use with ICD-9-CM administrative data: differing perspectives . J Clin Epidemiol . 1993 ; 46 : 1075 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 19 Charlson ME Pompei P Ales KL et al. . A new method of classifying prognostic comorbidity in longitudinal studies: development and validation . J Chronic Dis . 1987 ; 40 : 373 – 83 . Google Scholar Crossref Search ADS PubMed WorldCat 20 Deyo RA Cherkin DC Ciol MA . Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases . J Clin Epidemiol . 1992 ; 45 : 613 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 21 AGS Clinical Practices Committee, American Geriatric Society . The use of oral anticoagulants (warfarin) in older people . J Am Geriatr Soc . 2000 ; 48 : 224 – 7 . Crossref Search ADS PubMed WorldCat 22 Pernod G Godier A Gozalo C et al. . French clinical practice guidelines on the management of patients on vitamin K antagonists in at-risk situations (overdose, risk of bleeding, and active bleeding) . Thromb Res . 2010 ; 126 : e167 – 74 . Google Scholar Crossref Search ADS PubMed WorldCat 23 Tran HA Chunilal SD Harper PL et al. for the Australasian Society of Thrombosis and Haemostasis . An update of consensus guidelines for warfarin reversal . Med J Aust . 2013 ; 198 : 198 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 24 Dowlatshahi D Butcher KS Asdaghi N et al. . Poor prognosis in warfarin-associated intracranial hemorrhage despite anticoagulation reversal . Stroke . 2012 ; 43 : 1812 – 7 . Google Scholar Crossref Search ADS PubMed WorldCat 25 Thomopoulos KC Mimidis KP Theocharis GJ et al. . Acute upper gastrointestinal bleeding in patients on long-term oral anticoagulation therapy: endoscopic findings, clinical management and outcome . World J Gastroenterol . 2005 ; 11 : 1365 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 26 Chong CT Lew TW Kuperan P et al. . Rapid reversal of coagulopathy in warfarin-related intracranial haemorrhages with prothrombin complex concentrates . Anaesth Intensive Care . 2010 ; 38 : 474 – 80 . Google Scholar Crossref Search ADS PubMed WorldCat 27 Liotta EM Garg RK Temes RE et al. . Warfarin-associated intracerebral hemorrhage is inadequately treated at community emergency departments . Stroke . 2012 ; 43 : 2503 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 28 Schneeweiss S Avorn J . A review of uses of health care utilization databases for epidemiologic research on therapeutics . J Clin Epidemiol . 2005 ; 58 : 323 – 37 . Google Scholar Crossref Search ADS PubMed WorldCat Copyright © 2014 by the American Society of Health-System Pharmacists, Inc. All rights reserved.