TY - JOUR
AU1 - Traugott, Kristi, A.
AU2 - Maxwell, Pamela, R.
AU3 - Green,, Kay
AU4 - Frei,, Christopher
AU5 - Lewis, James, S.
AB - Abstract Purpose The effects of therapeutic drug monitoring (TDM) criteria in a computerized prescriber-order-entry (CPOE) system on the appropriateness of orders for vancomycin levels were evaluated. Methods Vancomycin TDM criteria were developed and implemented in a CPOE system. These criteria were displayed via a pop-up alert message when vancomycin levels were ordered and included directions for appropriate timing and justification for routine monitoring. Data for two groups of adult inpatients who had vancomycin levels ordered before and after criteria implementation were compared. Medical records were retrospectively reviewed for these patients to collect information regarding patient demographics, vancomycin dosage and indication, concurrent antibiotics and nephrotoxic agents during vancomycin therapy, length of stay, duration of vancomycin therapy, and number of vancomycin levels drawn. The primary outcome was the percent change in appropriate vancomycin levels ordered after criteria implementation. Results A total of 200 patients were analyzed, 100 in each group. The percentage of appropriate orders for vancomycin levels significantly increased after criteria implementation (from 58% to 68%, p = 0.02). The greatest effect on appropriateness occurred with the first level ordered (52% versus 70% in the preimplementation and postimplementation groups, respectively; p = 0.01). The majority of inappropriate levels were due to improper timing of sample collections, accounting for 55% of the inappropriate levels evaluated. Conclusion A significant increase in the number of appropriately ordered and drawn serum vancomycin levels occurred after implementation of TDM criteria in the hospital’s CPOE system. The majority of orders that were deemed inappropriate were due to improper timing of laboratory collection. Antibiotics, Blood levels, Computers, Dosage, Medication orders, Prescribing, Protocols, Rational therapy, Vancomycin Confusion regarding the appropriate use of therapeutic drug monitoring (TDM) for van-comycin has plagued hospitals for years.1–5 In early 2009, the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists published long-awaited consensus guidelines on this controversial subject.5 While these guidelines address many critical issues surrounding the appropriate use of TDM for vancomycin, confusion remains common among many prescribers attempting to monitor vancomycin therapy. We have observed practitioners at University Hospital, a 604-bed acute care academic teaching institution in San Antonio, Texas, become frustrated with the large number of serum vancomycin levels that are incorrectly ordered and obtained. Examples of incorrect levels include ordering vancomycin levels before reaching steady state, not indicating that they are time-critical levels (resulting in levels being obtained at random times), and nurses obtaining samples more than one hour before the next dose. These errors often lead to difficulty in interpreting serum vancomycin levels, which may result in inappropriate modifications of vancomycin regimens, such as the unnecessary withholding of doses due to perceived high trough serum vancomycin levels. Unnecessary costs are incurred when levels are ordered or obtained incorrectly and additional serum drug levels are needed. A 2007 study revealed that many U.S. hospitals reported the cost of obtaining a serum vancomycin level as less than $25; however, this does not include the time for physicians and nurses to reorder and redraw levels.6 When these same hospitals were asked to take all steps of obtaining a vancomycin level into consideration (including cost of materials, time for assay, quality control, obtaining samples, and reporting results), the majority reported a doubling of the cost per serum level ordered.6 One contributing factor to inappropriate ordering of serum vancomycin levels at our institution is inexperience with the hospital’s computerized prescriber-order-entry (CPOE) system among the interns and residents who frequently order these tests. Although CPOE systems have been shown to decrease medication errors, they can be confusing for an inexperienced user.7 For example, before implementation of the interventions in this study, prescribers had the option to order a peak, trough, or random vancomycin level with no formal guidance regarding appropriate TDM. Depending on patient location, orders for vancomycin levels defaulted to different collection times unless the prescriber fully displayed the order and selected a specific collection time. As a result, many serum vancomycin levels were obtained at inappropriate times. Despite potentially creating confusion among new users, an advantage of a CPOE system is the ability to use the system to continually educate a large number of prescribers by using pop-up alert messages. Previous investigators have utilized a CPOE system to implement vancomycin nomogram guidelines at their institution, demonstrating a significant increase in initial vancomycin orders that met nomogram recommendations after implementation.8 The purpose of this study was to examine the percent change in appropriate orders for vancomycin levels before and after an alert message providing instruction regarding the appropriate TDM of vancomycin was implemented in the CPOE system. Methods Study design This project was deemed to be a quality-improvement project by the institutional review board at the University of Texas Health Science Center at San Antonio and did not require its approval. A literature search was performed, and criteria for appropriate vancomycin TDM ( appendix)2,3,9–17 were developed and approved by our institution’s pharmacy and therapeutics (P&T) committee. Approved criteria were very similar to previously mentioned consensus guidelines for monitoring vancomycin.5 Before criteria implementation in the CPOE system, a comprehensive list of vancomycin levels ordered between January 1 and August 31, 2008, was generated from the CPOE system. Drug levels were then randomly chosen using Excel (Microsoft Corporation, Redmond, WA), excluding duplicate patients, outpatients, and patients younger than 18 years, until 100 patients were selected for the control group (pre-implementation group). The medical records for these patients were retrospectively reviewed to collect the following information: patient characteristics (age, sex, weight, height, patient care unit and service), vancomycin dosage and indication, concurrent antibiotics and nephrotoxic agents during vancomycin therapy (aminoglycosides, long-term high-dose nonsteroidal antiinflammatory drugs, tacrolimus, cyclosporine, amphotericin B, acyclovir, ganciclovir, valganciclovir, furosemide, contrast media, angiotensin-converting-enzyme [ACE] inhibitors, angiotensin II receptor blockers, chemotherapeutic agents, lithium, vasopressors), length of stay, duration of vancomycin therapy, number of levels drawn, and the first four vancomycin levels ordered (to determine appropriateness based on TDM criteria [ appendix]). Additional serum vancomycin levels were not evaluated for appropriateness but were included in the total number of levels drawn for each patient. Due to clinical feasibility, trough vancomycin levels were considered appropriate if drawn within one hour of the patient’s next dose. Also, due to the limitations of the study design, the criterion of not monitoring vancomycin levels if expected duration of therapy was less than one week was not assessed, because this information is difficult to interpret in chart reviews. After completion of the retrospective chart review of the control group, the TDM criteria were incorporated into the CPOE system. Implementation of the criteria resulted in the removal of the option to order peak serum vancomycin levels, as they were no longer considered necessary for vancomycin TDM.5 The remaining monitoring options consisted of random or trough levels. When prescribers ordered random or trough serum vancomycin levels, a pop-up alert message provided the criteria for appropriate monitoring. In addition, trough levels were defaulted to “predose” specifications, and random levels were defaulted to routine morning laboratory sample collections. All changes in the CPOE system were made in collaboration with the hospital’s clinical informatics specialist, the clinical informatics pharmacist, and the primary investigator with the prior knowledge and approval of the hospital’s P&T committee. After the changes were implemented on January 22, 2009, a consecutive comparator group (post-implementation group) was selected using the same exclusion criteria. This group consisted of the first 100 patients for whom vancomycin levels were ordered after criteria implementation. Both groups were selected and reviewed by the unblinded primary investigator. The data collected for the preimplementation group, as described above, were also collected for the postimplementation group. Outcomes and data analysis The primary outcome was the percent change in appropriate vancomycin levels ordered after criteria implementation. Secondary outcomes included differences in appropriate orders for therapeutic vancomycin levels among medical services and hospital units and explanations for orders for vancomycin levels that were deemed inappropriate. Continuous data were tested for normality by the Shapiro-Wilk W test.18 Normally distributed continuous data were summarized by means ± S.D. and compared by Student’s t test. Nonnormally distributed data were summarized by medians (interquartile range) and compared by the Wilcoxon rank sum test. Nominal data were analyzed by chi-square or Fisher’s exact test. All tests were two tailed, and a p of <0.05 was considered statistically significant. Data were analyzed using JMP 7.0 (SAS Institute, Cary, NC). Results Data were collected for 200 patients: 100 patients in the preimplementation group, analyzing 310 serum vancomycin concentrations, and 100 patients in the postimplementation group, analyzing 235 concentrations. Patients’ baseline characteristics are described in Table 1. Table 1 Baseline Characteristics of Patients Variable Preimplementation (n = 100) Postimplementation (n = 100) p Mean ± S.D. age, yr 51 ± 15 48 ± 15 0.1 Male, no. pts 59 69 0.1 Race, no. pts  White 61 65 0.4  Hispanic 32 23  African American 5 7  Asian 1 1  Other 1 4 Median (IQRa) body mass index, kg/m2 27.0 (24–34) 26.4 (22–30) 0.2 Median (IQR) serum creatinine conc., mg/dL 0.8 (0.6–1.5) 0.8 (0.6–1.5) 0.9 Median (IQR) length of stay, days 18 (8–39) 8 (5–15) <0.0001 Median (IQR) duration of vancomycin, days 8 (5–15) 5 (3–8) <0.0001 Median (IQR) no. vancomycin levels ordered 5 (2–10) 2 (1–4) <0.0001 Nephrotoxic agent use overall, no. pts 71 63 0.2 Specific nephrotoxic agents used, no. pts  Contrast media 30 21 0.1  Furosemide 21 15 0.3  Vasopressor 14 12 0.7  Calcineurin inhibitor 12 10 0.7  Other 35 35 . . . Site/type of infection, no. pts  Empirical 40 54 0.05  Bacteremia 34 14 0.0008  Skin or soft tissue infection 14 20 0.3  Pneumonia 13 10 0.5  Other 18 8 . . . Organism, no. pts  Coagulase-negative Staphylococcus 20 7 0.006  Methicillin-resistant Staphylococcus aureus 19 14 0.3  Methicillin-susceptible S. aureus 15 18 0.6  Enterococcus species 11 7 0.3  Other 6 9 . . . Variable Preimplementation (n = 100) Postimplementation (n = 100) p Mean ± S.D. age, yr 51 ± 15 48 ± 15 0.1 Male, no. pts 59 69 0.1 Race, no. pts  White 61 65 0.4  Hispanic 32 23  African American 5 7  Asian 1 1  Other 1 4 Median (IQRa) body mass index, kg/m2 27.0 (24–34) 26.4 (22–30) 0.2 Median (IQR) serum creatinine conc., mg/dL 0.8 (0.6–1.5) 0.8 (0.6–1.5) 0.9 Median (IQR) length of stay, days 18 (8–39) 8 (5–15) <0.0001 Median (IQR) duration of vancomycin, days 8 (5–15) 5 (3–8) <0.0001 Median (IQR) no. vancomycin levels ordered 5 (2–10) 2 (1–4) <0.0001 Nephrotoxic agent use overall, no. pts 71 63 0.2 Specific nephrotoxic agents used, no. pts  Contrast media 30 21 0.1  Furosemide 21 15 0.3  Vasopressor 14 12 0.7  Calcineurin inhibitor 12 10 0.7  Other 35 35 . . . Site/type of infection, no. pts  Empirical 40 54 0.05  Bacteremia 34 14 0.0008  Skin or soft tissue infection 14 20 0.3  Pneumonia 13 10 0.5  Other 18 8 . . . Organism, no. pts  Coagulase-negative Staphylococcus 20 7 0.006  Methicillin-resistant Staphylococcus aureus 19 14 0.3  Methicillin-susceptible S. aureus 15 18 0.6  Enterococcus species 11 7 0.3  Other 6 9 . . . a IQR = interquartile range. Open in new tab Table 1 Baseline Characteristics of Patients Variable Preimplementation (n = 100) Postimplementation (n = 100) p Mean ± S.D. age, yr 51 ± 15 48 ± 15 0.1 Male, no. pts 59 69 0.1 Race, no. pts  White 61 65 0.4  Hispanic 32 23  African American 5 7  Asian 1 1  Other 1 4 Median (IQRa) body mass index, kg/m2 27.0 (24–34) 26.4 (22–30) 0.2 Median (IQR) serum creatinine conc., mg/dL 0.8 (0.6–1.5) 0.8 (0.6–1.5) 0.9 Median (IQR) length of stay, days 18 (8–39) 8 (5–15) <0.0001 Median (IQR) duration of vancomycin, days 8 (5–15) 5 (3–8) <0.0001 Median (IQR) no. vancomycin levels ordered 5 (2–10) 2 (1–4) <0.0001 Nephrotoxic agent use overall, no. pts 71 63 0.2 Specific nephrotoxic agents used, no. pts  Contrast media 30 21 0.1  Furosemide 21 15 0.3  Vasopressor 14 12 0.7  Calcineurin inhibitor 12 10 0.7  Other 35 35 . . . Site/type of infection, no. pts  Empirical 40 54 0.05  Bacteremia 34 14 0.0008  Skin or soft tissue infection 14 20 0.3  Pneumonia 13 10 0.5  Other 18 8 . . . Organism, no. pts  Coagulase-negative Staphylococcus 20 7 0.006  Methicillin-resistant Staphylococcus aureus 19 14 0.3  Methicillin-susceptible S. aureus 15 18 0.6  Enterococcus species 11 7 0.3  Other 6 9 . . . Variable Preimplementation (n = 100) Postimplementation (n = 100) p Mean ± S.D. age, yr 51 ± 15 48 ± 15 0.1 Male, no. pts 59 69 0.1 Race, no. pts  White 61 65 0.4  Hispanic 32 23  African American 5 7  Asian 1 1  Other 1 4 Median (IQRa) body mass index, kg/m2 27.0 (24–34) 26.4 (22–30) 0.2 Median (IQR) serum creatinine conc., mg/dL 0.8 (0.6–1.5) 0.8 (0.6–1.5) 0.9 Median (IQR) length of stay, days 18 (8–39) 8 (5–15) <0.0001 Median (IQR) duration of vancomycin, days 8 (5–15) 5 (3–8) <0.0001 Median (IQR) no. vancomycin levels ordered 5 (2–10) 2 (1–4) <0.0001 Nephrotoxic agent use overall, no. pts 71 63 0.2 Specific nephrotoxic agents used, no. pts  Contrast media 30 21 0.1  Furosemide 21 15 0.3  Vasopressor 14 12 0.7  Calcineurin inhibitor 12 10 0.7  Other 35 35 . . . Site/type of infection, no. pts  Empirical 40 54 0.05  Bacteremia 34 14 0.0008  Skin or soft tissue infection 14 20 0.3  Pneumonia 13 10 0.5  Other 18 8 . . . Organism, no. pts  Coagulase-negative Staphylococcus 20 7 0.006  Methicillin-resistant Staphylococcus aureus 19 14 0.3  Methicillin-susceptible S. aureus 15 18 0.6  Enterococcus species 11 7 0.3  Other 6 9 . . . a IQR = interquartile range. Open in new tab No significant differences were found in age, race, sex, body mass index, serum creatinine concentration, or nephrotoxic agent use between the two groups. However, the median length of stay and duration of vancomycin therapy were longer and the number of serum vancomycin levels obtained were higher in the preimplementation group (p < 0.001, Table 1). Significant differences also existed when analyzing the site or type of infection and organism identified. The preimplementation group had a higher rate of bacteremia and coagulase-negative Staphylococcus infection (the majority of isolates were obtained from blood), while the postimplementation group was more likely to be started on vancomycin for empirical treatment. A greater number of appropriate vancomycin levels were obtained after implementation of the criteria in the CPOE system (180 of 310 [58%] versus 160 of 235 [68%] in the pre-implementation group, p = 0.02). Analysis of the first four individual levels for each patient demonstrated that the greatest effect on appropriateness occurred with the first level ordered (52% of the preimplementation levels ordered versus 70% of the postimplementation levels ordered, p = 0.01) (Table 2). Table 2 Appropriateness of Vancomycin Levels Ordered Vancomycin Level Ordered No. Appropriate/No. Levels Ordered (%) p Preimplementation Postimplementation First 52/100 (52) 70/100 (70) 0.01 Second 50/84 (60) 44/67 (66) 0.4 Third 43/67 (64) 25/39 (64) 1.0 Fourth 35/59 (59) 21/29 (72) 0.2  Total 180/310 (58) 160/235 (68) 0.02 Vancomycin Level Ordered No. Appropriate/No. Levels Ordered (%) p Preimplementation Postimplementation First 52/100 (52) 70/100 (70) 0.01 Second 50/84 (60) 44/67 (66) 0.4 Third 43/67 (64) 25/39 (64) 1.0 Fourth 35/59 (59) 21/29 (72) 0.2  Total 180/310 (58) 160/235 (68) 0.02 Open in new tab Table 2 Appropriateness of Vancomycin Levels Ordered Vancomycin Level Ordered No. Appropriate/No. Levels Ordered (%) p Preimplementation Postimplementation First 52/100 (52) 70/100 (70) 0.01 Second 50/84 (60) 44/67 (66) 0.4 Third 43/67 (64) 25/39 (64) 1.0 Fourth 35/59 (59) 21/29 (72) 0.2  Total 180/310 (58) 160/235 (68) 0.02 Vancomycin Level Ordered No. Appropriate/No. Levels Ordered (%) p Preimplementation Postimplementation First 52/100 (52) 70/100 (70) 0.01 Second 50/84 (60) 44/67 (66) 0.4 Third 43/67 (64) 25/39 (64) 1.0 Fourth 35/59 (59) 21/29 (72) 0.2  Total 180/310 (58) 160/235 (68) 0.02 Open in new tab Of the services that most often ordered vancomycin levels, the adult medicine service had the highest percentage of appropriately obtained serum vancomycin levels (79%), and those obtained by the solid organ transplant service were the least likely to be fully consistent with the aforementioned criteria (39%). When examining all the inappropriate serum vancomycin levels before and after implementation, the most common cause of inappropriately obtained serum vancomycin levels was sample collection more than one hour before the next dose, which accounted for 55% of the inappropriate levels evaluated (Table 3). Other reasons for serum vancomycin levels that did not meet our criteria included ordering the level before steady state had been achieved in patients receiving scheduled vancomycin (i.e., before the third to fifth dose), inappropriately selecting random levels in place of trough concentrations and vice versa, ordering peak levels, and unnecessarily repeating trough levels after an appropriate trough level was ordered with the previous dose. On several occasions, serum vancomycin levels were ordered and reported for patients not receiving vancomycin therapy (7% of levels). Table 3 Reasons for Inappropriateness of Order for Vancomycin Levels Reason No. (%) Inappropriate Vancomycin Levels Ordered p Overall Preimplementation Postimplementation Level drawn >1 hr before next dose 112 (55) 65 (50) 47 (63) 0.08 Should have been ordered as random level 30 (15) 19 (15) 11 (15) 1.0 Level was drawn before steady state achieved 25 (12) 15 (12) 10 (13) 0.7 Patient not treated with vancomycin 14 (7) 11 (8) 3 (4) 0.2 Should have been ordered as trough level 12 (6) 8 (6) 4 (5) 1.0 Peak level ordered 4 (2) 4 (3) 0 0.3 Trough level drawn with previous dosea 4 (2) 4 (3) 0 0.3 Previous dose missedb 2 (1) 2 (2) 0 0.5 Other 2 (1) 2 (2) 0 . . .  Total 205 130 75 . . . Reason No. (%) Inappropriate Vancomycin Levels Ordered p Overall Preimplementation Postimplementation Level drawn >1 hr before next dose 112 (55) 65 (50) 47 (63) 0.08 Should have been ordered as random level 30 (15) 19 (15) 11 (15) 1.0 Level was drawn before steady state achieved 25 (12) 15 (12) 10 (13) 0.7 Patient not treated with vancomycin 14 (7) 11 (8) 3 (4) 0.2 Should have been ordered as trough level 12 (6) 8 (6) 4 (5) 1.0 Peak level ordered 4 (2) 4 (3) 0 0.3 Trough level drawn with previous dosea 4 (2) 4 (3) 0 0.3 Previous dose missedb 2 (1) 2 (2) 0 0.5 Other 2 (1) 2 (2) 0 . . .  Total 205 130 75 . . . a Patient had an appropriately drawn trough level with previous dose, so order for second trough level was not necessary. b Patient did not receive previous dose of vancomycin, so result is not reflective of actual vancomycin level. Open in new tab Table 3 Reasons for Inappropriateness of Order for Vancomycin Levels Reason No. (%) Inappropriate Vancomycin Levels Ordered p Overall Preimplementation Postimplementation Level drawn >1 hr before next dose 112 (55) 65 (50) 47 (63) 0.08 Should have been ordered as random level 30 (15) 19 (15) 11 (15) 1.0 Level was drawn before steady state achieved 25 (12) 15 (12) 10 (13) 0.7 Patient not treated with vancomycin 14 (7) 11 (8) 3 (4) 0.2 Should have been ordered as trough level 12 (6) 8 (6) 4 (5) 1.0 Peak level ordered 4 (2) 4 (3) 0 0.3 Trough level drawn with previous dosea 4 (2) 4 (3) 0 0.3 Previous dose missedb 2 (1) 2 (2) 0 0.5 Other 2 (1) 2 (2) 0 . . .  Total 205 130 75 . . . Reason No. (%) Inappropriate Vancomycin Levels Ordered p Overall Preimplementation Postimplementation Level drawn >1 hr before next dose 112 (55) 65 (50) 47 (63) 0.08 Should have been ordered as random level 30 (15) 19 (15) 11 (15) 1.0 Level was drawn before steady state achieved 25 (12) 15 (12) 10 (13) 0.7 Patient not treated with vancomycin 14 (7) 11 (8) 3 (4) 0.2 Should have been ordered as trough level 12 (6) 8 (6) 4 (5) 1.0 Peak level ordered 4 (2) 4 (3) 0 0.3 Trough level drawn with previous dosea 4 (2) 4 (3) 0 0.3 Previous dose missedb 2 (1) 2 (2) 0 0.5 Other 2 (1) 2 (2) 0 . . .  Total 205 130 75 . . . a Patient had an appropriately drawn trough level with previous dose, so order for second trough level was not necessary. b Patient did not receive previous dose of vancomycin, so result is not reflective of actual vancomycin level. Open in new tab Discussion Overall, implementation of TDM criteria into our CPOE system significantly improved the ordering of appropriate vancomycin levels. However, even with this improvement, vancomycin levels were appropriate only 68% of the time, leaving considerable room for improvement. The reasons for inappropriate ordering of serum vancomycin levels suggest that incorrect timing of collection for serum vancomycin levels—not patient selection—is the major challenge at our center. The appropriate ordering of vancomycin levels did improve, but inappropriate collection times remain an issue, which will be addressed in the future through physician and nurse education. This study had several limitations. First, as a single-center, retrospective study, it had limited external validity. Second, selection bias existed between the two study groups. The preimplementation group appeared sicker, with longer hospital stays and vancomycin courses. The manner in which the preimplementation population was chosen may have contributed to this finding. This population was randomly selected from a list of every serum vancomycin level ordered during the designated time period. Therefore, a higher probability existed to select patients for whom multiple levels were ordered versus a patient for whom only one or two levels were ordered. The preimplementation group also had a significantly higher rate of bacteremia and coagulase-negative Staphylococcus infection. Because the preimplementation group had longer hospitalizations, they were more likely to have an associated nosocomial infection, such as an i.v. line infection. The significantly higher rate of empirical treatment with vancomycin in the postimplementation group most likely resulted from the manner in which this population was selected. After implementation of criteria into the CPOE system, the postimplementation group comprised the first 100 patients with vancomycin levels ordered who met inclusion criteria. Therefore, patients receiving empirical therapy for whom only one or two levels were ordered were more likely to be selected compared with the preimplementation group. These inconsistencies may have influenced the outcomes of the study; however, this is unlikely, as indication and disease severity should not influence the appropriateness of orders for vancomycin levels. Third, the criteria were implemented into the CPOE system in the form of a pop-up alert message. Since the alert message did not require acknowledgement from the prescriber, the number of prescribers who read the criteria cannot be determined. Lastly, this intervention included strict criteria for appropriateness. For example, certain medications considered nephrotoxic in this study, such as loop diuretics and ACE inhibitors, may not always be considered nephrotoxic. Also, trough levels were considered appropriate only for those patients receiving scheduled vancomycin, and random levels were considered appropriate only for patients not receiving scheduled vancomycin. A strength of this study is that multiple serum vancomycin levels were analyzed from each patient, providing a large sample size. In addition, the multiple patient variables collected provided insight into specific areas for improvement. The results of the study indicate that making changes in CPOE systems to assist physicians in correctly ordering serum vancomycin concentrations improves the number of correctly ordered levels. The absence of physician acknowledgment of appropriate TDM of vancomycin severely limited our study. Therefore, mandatory acknowledgment from physicians may also assist in appropriate TDM of vancomycin. Lastly, nurse education regarding appropriate collection of vancomycin serum levels appears to be the key to improving vancomycin TDM. Conclusion A significant increase in the number of appropriately ordered and drawn serum vancomycin levels occurred after implementation of TDM criteria in the hospital’s CPOE system. The majority of orders that were deemed inappropriate were due to improper timing of laboratory collection. Appendix—Criteria for vancomycin therapeutic drug monitoring2,3,9–17 Criteria for routine vancomycin therapeutic drug monitoring Unstable renal function (serum creatinine concentration fluctuations of >0.3 mg/dL during therapy) Acute/chronic renal insufficiency Morbid obesity (body mass index of >40 kg/m2) Serious infections (osteomyelitis, bacteremia, pneumonia, infections in immunocompromised patients) Concurrent use of nephrotoxic agents Rapid drug clearance (patients with cystic fibrosis, burn patients, patients age <30 years) Poor response Routine vancomycin monitoring not recommended Expected therapy duration of less than one week Oral vancomycin used Monitor random levels only Severe renal dysfunction Dialysis Sample collection information Peak vancomycin levels should not be ordered Trough vancomycin levels should be ordered before third to fifth dose Subsequent levels should be ordered weekly Daily levels should be ordered for patients with renal dysfunction/dialysis and patients receiving concurrent therapy with nephrotoxic agents Footnotes The authors have declared no potential conflicts of interest. References 1 Cantu TG Yamanaka-Yuen NA Lietman PS . Serum vancomycin concentrations: reappraisal of their clinical value . Clin Infect Dis. 1994 ; 18 : 533 – 43 . Google Scholar Crossref Search ADS PubMed WorldCat 2 Moellering RC Jr . Monitoring serum vancomycin levels: climbing the mountain because it is there? Clin Infect Dis. 1994 ; 18 : 544 – 6 . Google Scholar Crossref Search ADS PubMed WorldCat 3 Rybak MJ . The pharmacokinetic and pharmacodynamic properties of vancomycin . Clin Infect Dis. 2006 ; 42 (suppl 1) : S35 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 4 Darko W Medicis JJ Smith A et al. . Mississippi mud no more: cost-effectiveness of pharmacokinetic dosage adjustment of vancomycin to prevent nephrotoxicity . Pharmacotherapy. 2003 ; 23 : 643 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 5 Rybak M Lomaestro B Rotschafer JC et al. . Therapeutic monitoring of vancomycin in adult patients: a consensus review of the American Society of Health-System Pharmacists, the Infectious Diseases Society of America, and the Society of Infectious Diseases Pharmacists . Am J Health-Syst Pharm. 2009 ; 66 : 82 – 98 . Google Scholar Crossref Search ADS PubMed WorldCat 6 Paladino JA Sunderlin JL Adelman MH et al. . Observations on vancomycin use in U.S. hospitals . Am J Health-Syst Pharm. 2007 ; 64 : 1633 – 41 . Google Scholar Crossref Search ADS PubMed WorldCat 7 Kaushal R Shojania KG Bates DW . Effects of computerized physician order entry and clinical decision support systems on medication safety: a systematic review . Arch Intern Med. 2003 ; 163 : 1409 – 16 . Google Scholar Crossref Search ADS PubMed WorldCat 8 McCluggage L Lee K Potter T et al. . Implementation and evaluation of vancomycin nomogram guidelines in a computerized prescriber-order-entry system . Am J Health-Syst Pharm. 2010 : 67 : 70 – 5 . Google Scholar Crossref Search ADS PubMed WorldCat 9 Freeman CD Quintiliani R Nightingale CH . Vancomycin therapeutic drug monitoring: is it necessary? Ann Pharmacother. 1993 ; 27 : 594 – 8 . Google Scholar Crossref Search ADS PubMed WorldCat 10 Pryka RD . Vancomycin serum concentration monitoring: a continued debate . Ann Pharmacother. 1994 ; 28 : 1397 – 9 . Google Scholar Crossref Search ADS PubMed WorldCat 11 Leader WG Chandler MH Castiglia M . Pharmacokinetic optimisation of vancomycin therapy . Clin Pharmacokinet. 1995 ; 28 : 327 – 42 . Google Scholar Crossref Search ADS PubMed WorldCat 12 James CW Gurk-Turner C . Recommendations for monitoring serum vancomycin concentrations . Proc (Bayl Univ Med Cent). 2001 ; 14 : 189 – 90 . Google Scholar Crossref Search ADS PubMed WorldCat 13 Begg EJ Barclay ML Kirkpatrick CM . The therapeutic monitoring of antimicrobial agents . Br J Clin Pharmacol. 2001 ; 52 (suppl 1) : 35S – 43S . Google Scholar Crossref Search ADS PubMed WorldCat 14 Tobin CM Darville JM Thomson AH et al. . Vancomycin therapeutic drug monitoring: is there a consensus view? The results of a UK National External Quality Assessment Scheme (UK NEQAS) for Antibiotic Assays questionnaire . J Antimi-crob Chemother. 2002 ; 50 : 713 – 8 . Google Scholar Crossref Search ADS WorldCat 15 Saunders NJ . Why monitor peak vancomycin concentrations? Lancet. 1994 ; 344 : 1748 – 50 . Google Scholar Crossref Search ADS PubMed WorldCat 16 Bates DW Soldin SJ Rainey PM et al. . Strategies for physician education in therapeutic drug monitoring . Clin Chem. 1998 ; 44 : 401 – 7 . Google Scholar PubMed WorldCat 17 Pannu N Nadim MK . An overview of drug-induced acute kidney injury . Crit Care Med. 2008 ; 36 (suppl) : S216 – 23 . Google Scholar Crossref Search ADS PubMed WorldCat 18 Shapiro SS Wilk MB . An analysis of variance test for normality . Biometrika. 1965 ; 52 : 591 – 611 . Google Scholar Crossref Search ADS WorldCat Copyright © 2011, American Society of Health-System Pharmacists, Inc. All rights reserved.
TI - Effects of therapeutic drug monitoring criteria in a computerized prescriber-order-entry system on the appropriateness of vancomycin level orders
JF - American Journal of Health-System Pharmacy
DO - 10.2146/ajhp090661
DA - 2011-02-15
UR - https://www.deepdyve.com/lp/oxford-university-press/effects-of-therapeutic-drug-monitoring-criteria-in-a-computerized-m1TdRawLsu
SP - 347
VL - 68
IS - 4
DP - DeepDyve
ER -