Implementing a hospital-wide protocol for Staphylococcus aureus bacteremia

Implementing a hospital-wide protocol for Staphylococcus aureus bacteremia Staphylococcus aureus bacteraemia (SAB) is associated with high-mortality and complication rates. A multidisciplinary ap- proach is needed to predict, detect and treat complications. In this pre- and post-intervention study, we investigated the effects of a hospital-wide protocol for diagnosis, classification and treatment of SAB. It was hypothesized that complications and endocar- ditis would be better identified and treated. Medical records of SAB patients admitted in 2011 and 2012 (pre) were analysed. In 2013, a protocol, describing risk factors, diagnostic classification and recommended treatment, was implemented. In 2014 and 2015 (post), SAB patients were followed prospectively. Transthoracic (TTE) or transoesophageal cardiac ultrasound (TEE) was chosen following a decision tree. A resident internal medicine acted as contact person. Pre-intervention, 98 patients were eligible for analysis compared to 85 patients post-intervention. Age and number of risk factors were slightly higher post-intervention; other baseline characteristics were similar. Most SAB-patients were classified as complicated (89 and 82% pre- and post- intervention, respectively). Follow-up blood cultures drawn within 2 days after initiating treatment increased from 51 to 85%. Cardiac ultrasounds increased from 44 to 83% for TTE and 13 to 24% for TEE. Endocarditis was more frequently diagnosed (4 vs. 12%). Additionally, duration of antibiotic therapy increased. The 3-month mortality did not change significantly (33% pre- intervention vs. 35% post-intervention; p > 0.05). Introduction of a hospital-wide protocol for SAB management increased standard of care, created awareness among clinicians to properly classify SAB, search for endocarditis and adapt duration of antibiotic treatment. Mortality did not decrease. . . . . Keywords Staphylococcus aureus Bacteraemia Treatment Echocardiography Infective endocarditis Introduction acutely life-threatening complications such as endocarditis, intracardiac abscesses and valve perforation [3]. Important in Staphylococcus aureus bacteraemia (SAB) is associated with the management of SAB is to correctly classify SAB as either a long hospital stay, a high morbidity and high 30-day mor- uncomplicated or complicated and to detect the complications tality numbers ranging from 10 to 30% [1, 2]. The mortality is as quickly as possible. Risk factors for a complicated course associated with the development of serious complications include community acquisition, persistent positive blood cul- such as metastatic abscesses, vertebral osteomyelitis and tures after 48–72 h, persistent fever after 72 h, diabetes mellitus, purulent thrombophlebitis, presence of prosthetic de- vices and immunosuppressive therapy [4–6]. In addition, re- * K. Bolhuis petitive follow-up blood cultures and echocardiography are k.bolhuis@amc.nl recommended for patients with risk factors to detect (imminent) complications. Department of Internal Medicine, Academic Medical Center, The antibiotic treatment of SAB aims to prevent deteriora- Meijbergdreef 9, 1105 AZ Amsterdam, The Netherlands tion of disease and the development of complications. The Department of Medical Microbiology, Tergooi Hospital, Van duration of treatment remains somewhat controversial, but Riebeeckweg 212, 1213 XZ Hilversum, The Netherlands current guidelines regarding management of SAB patients Department of Cardiology, Tergooi Hospital, Van Riebeeckweg 212, recommend at least 14 days of antimicrobial therapy for a 1213 XZ Hilversum, The Netherlands simple bacteraemia with low risk of complications. For Department of Internal Medicine, Tergooi Hospital, Van bacteraemia with risk factors associated with a complicated Riebeeckweg 212, 1213 XZ Hilversum, The Netherlands Eur J Clin Microbiol Infect Dis course due to metastatic infections or endocarditis, at least All events of S. aureus bacteraemia in patients > 16 years 28 days of antimicrobial therapy is recommended [4, 7–10]. were analysed during the study period. A SAB > 12 weeks In contrast to many other blood stream infections, the instal- after cessation of antimicrobial therapy for a previous SAB ment of adequate antibiotic treatment is not enough to warrant was considered to be a distinct event and was included into the a favourable outcome. analysis. Because the study was designed to evaluate the effect Although SAB is among the deadliest bloodstream infec- of implementation of a protocol, exclusion criteria included tions, previous studies showed that potentially complicated contaminated blood culture specimen or any of the following infections are not well recognized by physicians resulting in within 3 days after collection of blood culture sample: death, non-adherence to standard of care therapy in 10 to 41% of withdrawing of therapy or transfer to another facility. SAB cases [11–13]. Detection of complications remains a Pregnancy was an exclusion criterion as well. challenge because the clinical symptoms and findings are non-specific [14, 15]. Despite the identification of several risk Pre-intervention group factors for developing these complications, mortality rates have only improved marginally [16]. The pre-intervention group cohort was studied in retrospect. Non-adherence to standard of care is most often the result of All records of hospitalized patients with blood cultures posi- inadequate duration of antibiotic treatment, lack of follow-up tive for S. aureus between 1 January 2011 and 31 December blood cultures, or omitting echocardiography [12–14]. In par- 2012 were retrieved. The date of the first sample taken was ticular, transoesophageal echocardiography (TEE) is often not considered the date of diagnosis. This group was called the performed due to comorbidity, limited acceptance by patients or pre-intervention group. treating physicians [12], even though other authors showed that performing routinely TEE results in a high incidence of the Intervention protocol most feared complication: infective endocarditis [17]. Multiple studies have shown an increase in adherence to In 2013, accompanying the introduction of a hospital-wide standard of care by routine infectious disease (ID) consulta- protocol, awareness was created by oral and visual presenta- tion and some found that this adherence seemed to preface a tions to all medical staff. The potential complications of SAB significant lower short-term mortality rate (or variable effect and the added value of TEE were emphasized in these educa- on mortality rates) [14, 18–21]. Almost all those studies were tional sessions. All cardiologists were made aware of the pro- performed in large academic, tertiary hospitals, and it remains tocol. The protocol was implemented in all medical and sur- to be confirmed whether this can be generalized to non-aca- gical wards and was designed to function as a simple guideline demic, regional or smaller hospitals. for physicians by classifying patients to groups with uncom- To improve quality of care, we conducted an intervention plicated SAB or with potentially complicated SAB. Within study, a multidisciplinary approach to predict, detect and treat complicated SAB, possible endocarditis was identified as a complications of SAB and adapt the treatment to the risk profiles. separate group. The primary goal of this study was to implement a clear, struc- Patients with SAB were reported by the laboratory of mi- tured, safe and univocal hospital-wide protocol regarding the crobiology as soon as a blood culture became positive for S. management and treatment of patients with SAB and to evaluate aureus. The attending physician was informed about this SAB the effects of this protocol in a regional teaching hospital. and referred to the protocol. According to the protocol SAB was considered complicated when at least one of the following risk factors was present: community acquisition, diabetes Materials and methods mellitus, persistent positive blood cultures after 48/72 h of antibiotic therapy, prosthetic material in situ, persistent fever Study design after 72 h of therapy and metastatic infections. These risk factors are known to be associated with a high risk for a com- This pre- and post-intervention retrospective study was per- plicated course [4–6]. formed in a 550-bed regional teaching hospital in the centre of Before and during the intervention, an ID specialist (PdV) the Netherlands. In cooperation with a cardiologist, medical was working in the hospital. The introduction of the protocol microbiologist and ID specialist, a protocol was designed that introduced a closer involvement of the ID specialist and the describes the risk factors of complicated SAB and includes a dedicated resident internal medicine (KB) to the patient care flow chart for diagnostic classification and recommendations but did not include bedside visits by them. The flow chart for therapy according to the current guidelines. Pre- guided the treating physicians to score these risk factors and intervention data were collected retrospectively from the hos- classify the patients in complicated versus uncomplicated pital records. The post-intervention data were collected (Fig. 1). Flucloxacillin is the drug of first choice in the prospectively. Netherlands for patients with SAB. According to the flowchart, Eur J Clin Microbiol Infect Dis Box 1: Criteria for complicated SAB SAB DM Immune deficiency or suppressive therapy Artificial material in situ Purulent thromboflebitis Persistent fever 72 hrs after start of therapy See Box 1 Positive blood cultures 2-3 days after start of therapy Signs of septic embolisms Community acquired Uncomplicated Complicated SAB SAB Intracardiac device Hemodialysis Spinal infection Non-vertebral osteomyelitis Yes No Diagnosis endocarditis rejected Yes No TTE Positive TEE Negative Inconclusive unless TTE conclusive Complicated SAB Complicated SAB without with endocarditis endocarditis 14 days of > 4 weeks flucloxacillin 6 dd 2 gr i.v. 6 weeks of flucloxacillin 6 dd 2 gr i.v. flucloxacillin 6 dd 1 gr Re-evaluation Fig. 1 Protocol for diagnosis, classification and treatment of SAB. SAB, Staphylococcus aureus bacteraemia; TTE, transthoracic echocardiography; TEE, transoesophageal echocardiography uncomplicated SAB was treated with flucloxacillin intrave- these complicated SAB-patients. If infective endocarditis was nously for 14 days. For complicated SAB, the flowchart advises confirmed, the patient was treated according to national guide- to assess the modified Duke criteria for endocarditis and to lines which includes 6 weeks of IV flucloxacillin and 3 to 5 days perform a transthoracic echocardiography (TTE) for all pa- of low-dose gentamicin, with or without rifampicin (Fig. 2). tients. When the echocardiographic results were uncertain or inconclusive, a transoesophageal echocardiography (TEE) Post-intervention group was suggested. When indicated, echocardiography was con- ducted by the attending cardiologist, usually the same day. The post-intervention group consisted of all SAB episodes of Depending on the outcome of the echocardiography, at least admitted patients between 1 January 2014 and 31 December 4 weeks of IV flucloxacillin therapy was recommended for 2015. These patients were followed prospectively with respect Eur J Clin Microbiol Infect Dis Fig. 2 Flow diagram of events 219 events of SAB included in the study. SAB, Staphylococcus aureus bacteraemia 98 events SAB 121 events SAB 23 events excluded 13 events excluded 4 age 1 age 3 contamination 3 contamination 10 missing follow up 9 missing follow up 4 relapse SAB < 12 wks 2 other 98 events SAB analysed 85 events SAB analysed to the routinely collected data. Consequently, in this study, When the source of infection was, for example, an indwell- prospectively collected data were compared to data of recent ing peripheral intravenous catheter, a pacemaker or a central historic cohort. In the post-intervention period, a dedicated venous catheter, this was defined as a health care-related in- resident internal medicine, advising microbiologist and ID fection, and removal of these intravenous devices was scored. specialist actively recommended the attending physician to Adequate antibiotic treatment was defined as the initiation consult and follow the protocol, which is available on the of parenteral administration of antibiotics with documented hospital’s intranet pages. The management of individual pa- activity against the isolated S. aureus strain, with adequate tients including actually performing echocardiography dosage and duration of therapy (14 days for uncomplicated remained the responsibility of the treating physician. and at least 28 days for complicated SAB). The study was approved by the hospital’s scientific review Infective endocarditis was scored based on echocardiogra- committee responsible for improving the quality of care. This phy, modified Duke criteria [22], strongly suggestive clinical study did not collect experimental data, and since the data findings, PET-CT or endocarditis found at autopsy. collection itself is part of the routine quality of care assessment Relapse of infection was defined as recurrent isolation of S. system, no additional ethical clearance was required. aureus in blood cultures within 12 weeks after cessation of antimicrobial therapy. Patient follow-up The most important post-discharge outcomes were relapse of SAB and mortality. In case of relapse, it was expected that Patients were followed until death or 6 months after diagnosis, most if not all patients would be referred and admitted to the whichever of the events occurred first. The medical records hospital due to the complicated natural course of SAB and were analysed, and data were collected by a single reasoning that SAB is not self-limiting in case no intravenous investigator. antimicrobial treatment is offered to the patient. Following this theory, possible relapses were scored by analysing the medical Data collection records. Mortality outside the hospital was collected by medical Medical records were analysed for patient characteristics, records as well. Tergooi hospital receives and records the in- medical history, underlying diseases, source of infection, pres- formation of deaths of all patients, even if they do not die in ence of prosthetic devices and other known risk factors. the hospital. Information about choice, dosage and duration of antimicro- bial therapy was recorded; number of follow-up blood cultures and number and results of echocardiograms were also includ- Primary endpoint output ed. All SAB-related complications, including the occurrence of endocarditis, relapse of infection and mortality, were Adherence to standards of care was scored by three indicators: recorded. (1) obtaining follow-up blood cultures 2 to 4 days after initi- Hospital-acquired bacteraemia was defined by isolation of ation of (adequate) therapy; (2) administration of adequate S. aureus from a blood culture obtained > 72 h after hospital therapy scored by the use of appropriate antibiotics, optimal admission. Purulent thrombophlebitis was diagnosed, based parenteral treatment dosage and duration; and (3) performing on clinical symptoms and signs. echocardiography. Eur J Clin Microbiol Infect Dis Data analysis Results Data were initially entered in Microsoft Excel and analysed in A number of 219 patients was diagnosed with at least one SPSS (v. 23, SPSS Statistics for Windows, Version 23.0. blood culture specimen positive for S. aureus.Thirty-sixpa- Released 2015. IBM Corp. Armonk, NY). tients were excluded, six due to contaminated blood cultures, Parametric tests were used for comparison of normally dis- five patients were < 16 years old, four because of a relapse tributed variables. Non-parametric tests were used otherwise. within 12 weeks and 19 were lost to follow-up within 3 days Categorical variables were analysed with chi-square tests. because of death, withdrawing of therapy or translocation to Statistically significant difference was accepted when p < another facility. A total of 98 patients remained eligible for 0.05. analysis in the pre-intervention group and 85 patients in the post-intervention group (Fig. 1). Data availability The datasets generated during and/or Demographic characteristics are shown in Table 1.Two analysed during the current study are available from the cor- patients, both in the pre-intervention group, suffered from responding author on reasonable request. more than one episode of S. aureus bacteraemia (defined as Table 1 Patient demographic Variable Pre-intervention Post-intervention p value characteristics. Staphylococcus (n =98) group (n =85) aureus bacteraemia before (2011– 2012) and after a protocol Age, median years (min, max) 69, 65 (19, 94) 76.49 (35, 98) 0.016 adherence intervention (2014– 2015) Female sex 47 (48%) 35 (41%) 0.44 Classification SAB Uncomplicated SAB 11 (11%) 15 (18%) 0.303 Complicated SAB 87 (89%) 70 (82%) Risk factors Community acquisition 63 (64%) 55 (65%) Diabetes mellitus 28 (29%) 26 (31%) Positive follow-up blood cultures 48/72 h 13 (13%) 21 (25%) Prosthetic material 31 (32%) 29 (34%) Persistent fever after 72 h 4 (4%) 9 (11%) Catheter-related and in situ 3 (3%) 2 (2%) Immunocompromised 12 (12%) 9 (11%) Metastatic infections 23 (23%) 23 (27%) Purulent thrombophlebitis 11 (11%) 0 (0%) Malignancy 18 (18%) 10 (12%) Alcohol 10 (10%) 7 (8%) Haemodialysis 3 (3%) 1 (1%) No. of risk factors 0 11 (11%) 15 (18%) 0.590 1 27 (28%) 16 (19%) 2 31 (32%) 28 (33%) 3 22 (22%) 16 (19%) 45(5%) 6(7%) 53(3%) 4(5%) Source of infection Health care-related 50 (51%) 37 (44%) Surgery 11 (22%) 6 (16%) Urinary tract/CAD 6 (12%) 8 (22%) CVC/PAC 11 (22%) 5 (14%) Peripheral catheter 12 (24%) 11 (30%) Unknown source 26 (52%) 21 (57%) Patient demographics. SAB, Staphylococcus aureus bacteraemia; CAD, urine catheter; CVC, central venous catheter; PAC, port a cath; NS, not significant. A p value < 0.05 is considered statistically significant Eur J Clin Microbiol Infect Dis a new episode of SAB > 12 weeks post-therapy of the previ- classified as complicated (89 and 82% pre- and post-interven- ous episode of SAB). In the first patient, the episodes were tion, respectively). separated by 9 months, and in the second patient, the episodes The adequacy of care for SAB increased during the inter- were separated by 3 years. vention (Table 2). Firstly, the frequency of drawing follow-up The median age was higher in the post-intervention group blood cultures 48–72 h after initiating treatment significantly compared to the pre-intervention group, and malignancy was increased, resulting in the identification of more patients with more frequent in the pre-intervention group, but other baseline positive blood cultures during follow-up (15 vs. 25%). In both characteristics were similar. Most SAB patients were groups, the proportion of positive blood cultures remained the Table 2 Quality of care Diagnostic and therapeutic management of Staphylococcus aureus bacteraemia, indicators before (2011–2012) and after a protocol adherence intervention (2014–2015) Variable Pre-intervention (n = 98) Post-intervention p value (n =85) Diagnostic workup Echocardiography performed in complicated SAB patients TTE 38 (44%) 58 (83%) <0.001 TEE 11 (13%) 17 (24%) 0.09 2nd blood cultures obtained after 48–72 h 50 (51%) 72 (85%) <0.001 Adequate antibiotic therapy Yes 38 (39%) 60 (71%) <0.001 No 59 (58%) 24 (28%) Not started 4 0 Inadequate length 53 24 Inadequate dose 3 0 Unknown 1 (1%) 1 (1%) Intravenous antibiotic treatment < 14 days 32 (33%) 8 (9%) ≥ 14 days 44 (45%) 58 (68%) Lost to FU < 14 days 18 (18%) 19 (22%) Death/withdrawing therapy 15 16 Transfer 2 2 Unknown 1 1 Adequate AB not started 4 (4%) 0 (0%) Antibiotics stopped due to death or withdrawing therapy Adequate AB not started 4 0 Yes 23 21 No 65 57 Transfer 5 5 Unknown 1 2 Treatment duration (mean) 17.3 (n = 65) 23.0 (n =57) 0.014 Uncomplicated SAB (mean/days) 10.3 (n =7) 13.8 (n =15) Complicated SAB (days, mean) 18.2 (n = 58) 26.3 (n =42) Length of hospital stay (mean, days) 27.5 22.4 NS SAB, Staphylococcus aureus bacteraemia; TTE, transthoracic echocardiography; TEE, transoesophageal echocar- diography; FU, follow-up; AB antibiotic treatment; NS, not significant. A p value < 0.05 is considered statistically significant. Mean treatment duration. Patients were excluded when adequate antibiotics were not started or if antibiotic therapy was stopped due to death, withdrawing therapy or transfer to a different health care clinic Echocardiography in complicated SAB patients Length of hospital stay was calculated from date positive blood culture was taken till day of discharge. Patients transferred to different hospital were excluded Eur J Clin Microbiol Infect Dis same (29 vs. 29%). Secondly, the mean duration of antibiotic It created awareness among clinicians to properly classify therapy significantly increased (17.3 vs. 23.0 days),resulting SAB, search for endocarditis, choose the right antibiotic reg- in more patients receiving adequate antimicrobial therapy (39 imen and adapt duration of antibiotic treatment. However, it vs. 71%). And, thirdly, among patients with complicated did not improve survival. SAB, more TTEs were performed (44 to 83%), and the num- One of the strong points of this study is its real-life setting. A ber of TEEs increased from 13 to 24%. Consequently, endo- small intervention such as the introduction of this protocol cre- carditis was detected more frequently (12%) during the inter- ates knowledge about the treatment approach involving patients vention period than before (4%). The increase of TEE did not with SAB among a wide range of medical specialists and there- decrease the proportion of positive TEEs (31 vs. 42%). This by directly improves standard of care. The introduction of pro- underlines the urgency to perform a TEE. tocols and local guidelines probably belong to the most com- More infectious complications were detected in the post- mon and feasible interventions in hospitals. The situation in our intervention group (39 vs. 45%), and the mean duration of hospital reflects the situation of many hospitals, and this inter- hospital stay, counting from the date of the first positive blood vention can easily be implemented in many other hospitals. In culture, decreased from 27 days before intervention to 23 days our knowledge, this study is among the largest studies per- post-intervention. In addition, lower number of relapses with- formed in the setting of a general hospital [23]. in 12 weeks were reported (5 vs. 2%) (Table 3). In this study, the motivations to change practices were not The 4- and 12-week mortality did not change significantly studied in depth. However, we encountered some expressions (26 vs. 26% and 33 vs. 35% in the pre-intervention vs. post- of inertia towards change. These were mainly the lack of intervention group; p >0.05). knowledge and sense of urgency with respect to the poor prognosis of SAB and its treacherous presentation of compli- cations. Another explanation for the relatively persistent low number of TEE might be the fact that, in exceptional cases, the Discussion protocol permitted clinicians to forego TEE in patients with- out any clinical signs of endocarditis, with excellent acoustic This study shows that introducing a hospital-wide protocol for window on TTE and no vegetations or valvular regurgitation management of SAB in a general teaching hospital improved (Fig. 1). We did not encounter any expressions of mistrust by the standard of care: The number of follow-up blood cultures the attending physicians towards the resident or ID specialist. significantly increased by two thirds, antimicrobial therapy There was often reluctance to conduct transoesophageal became significantly more adequate and the number of TTE echocardiography because it was considered too invasive for and TEE almost doubled post-intervention. ill and elderly patients and falsely relying on the acuity of the Table 3 Secondary outcomes Variable Pre-intervention (n = 98) Post-intervention group (n =85) Infectious complications 38 (39%) 38 (45%) Endocarditis 4 (4%) 10 (12%) Spondylodiscitis 8 (8%) 5 (6%) Abscesses 10 (10%) 13 (15%) Septic arthritis 3 (3%) 6 (7%) Prosthetic joint infection 5 (5%) 7 (8%) Endovascular infection 2 (2%) 2 (2%) Relapse within 3 months 5 (5%) 2 (2%) Cerebral septic embolism 4 (4%) 4 (5%) Mortality 30-day mortality 25 (26%) 22 (26%) Uncomplicated 3 1 Complicated 22 21 Unknown/lost to FU 2 1 90-day mortality 32 (33%) 30 (35%) Uncomplicated 4 3 Complicated 28 27 Unknown/lost to FU 3 1 Secondary outcome: complications. FU, follow-up Eur J Clin Microbiol Infect Dis cardiac valve images during transthoracic echocardiography. Endocarditis is among the most feared complications of PET-CT was available already before the intervention but locat- SAB with a reported incidence of 25–29% and a 1-year mor- ed in another hospital with which it was shared. After the study, tality rate up to 43% [3, 15, 24]. This complication is difficult a PET-CT facility was installed at the location of the study. It is to diagnose since the initial clinical symptoms and signs are nowadays increasingly used to detect complications of SAB. insensitive and non-specific [14, 15]. Echocardiography is The contribution of purulent phlebitis with bacteraemia recommended as primary tool to diagnose endocarditis but decreased (Table 1) substantially. This reduced the total num- clinicians should be aware about its limitations. TTE is insen- ber of complicated SAB while the proportional detection of sitive (sensitivity of 55%) compared with TEE for the detec- persistent bacteraemia increased. tion of native valve vegetations [25]. Although others suggest Previous studies showed the beneficial effects of imple- that TTE has a high negative predictive value of 97%, suffi- mentation of hospital-wide diagnostic and therapeutic guide- cient data are lacking, and further research is needed to better lines and of the standard involvement of an infectious disease identify those patients with SAB for whom TEE is required to consultant on adherence to standard of care [13, 18]. Borde et detect endocarditis. [26] The sensitivity of a TEE is 90–100% al. implemented a standardized bundle approach to the treat- and is the preferred screening tool in patients with a compli- ment of S. aureus bacteraemia which resulted in an increase of cated SAB and suspected infectious endocarditis. [27] conducting TEE by 42% and collecting follow-up blood cul- A recent Canadian multicentre study including 847 patients tures by 49% [18]. Appropriate therapy within 72 h of detec- compared the management of SAB in patients receiving in- tion of SAB increased by 81%. fectious disease (ID) consultation and patients without ID In our study, we also found significant effects on quality of consultation of an ID specialist. The number of patients re- care indicators; the adequacy of dosing and duration of anti- ceiving echocardiography increased, but despite bedside con- microbial treatment and on documentation of infectious com- sultation of an ID specialist, TEE was only performed in 17% plications. An interesting phenomenon was found by of patients [21]. Others suggested a simple set of criteria for analysing the duration of hospital stay before and after inter- nosocomial SAB to identify patients with low risk of infec- vention. Despite a longer mean antimicrobial treatment, the tious endocarditis and who might not routinely require TEE mean duration of hospital stay decreased after protocol imple- [28]. mentation. We hypothesize that, greater awareness among cli- A recently published large retrospective 12-year cohort nicians, more patients receiving optimal treatment and the use study showed that the use of three evidence-based care pro- of more diagnostic tools to find infectious complications pre- cesses, appropriate antibiotic therapy, echocardiography and vents serious complications during hospital stay and leads to consultation of an ID specialist, was associated with a reduc- faster hospital discharge. tion of 30-day mortality by 67% [29]. The risk-adjusted mor- The year 2013 was not analysed but used to implement the tality decreased from 23.5 to 18.2%, lower than in our hospi- SAB protocol. After proper implementation, we started to collect tal. In our study, we also found significant effects on adher- data prospectively from 2014 onwards. The long study period, ence to guidelines and the use of echocardiography after the 5 years, has the advantage of excluding season effects. Although introduction of a protocol. Although execution of our protocol this may also introduce effects of changing health care and out- was monitored by a dedicated person (KB), this is not equal to comes over time, we are not aware of important structural chang- a bedside consultation of an ID specialist. es regarding the medical care in this hospital during this period. Despite the improved adherence to guidelines in treating The only effect was a reduction of purulent phlebitis. The extra patients with SAB in the post-intervention group, overall mor- attention and instructions about SAB and its causes might have tality did not decrease in our study. Besides the fact that the improved awareness about the risks and better measures to pre- study itself was not designed to see changes in mortality, the vent infections due to peripheral catheters. slightly older age and higher incidence of risk factors for com- Retrospective studies may be subject to different forms of plicated SAB in the post-intervention group could be an addi- bias, but in this study, we captured all patients with document- tional explanation for the absent decrease in mortality. In ad- ed SAB in a similar manner before (the retrospective part) and dition, there was no adjustment for comorbidities and severity after intervention. This makes bias less likely. This study was of disease performed on admission in both groups. Another not blinded but since data collection was similar before inter- possible explanation could be the fact that in this study, the vention as after, we do not consider this a probable source of management strategies, including ordering for echocardiogra- bias. Follow-up was incomplete for several patients who were phy, remained the responsibility of the treating physician. We discharged from the hospital for palliative/terminal home care did not routinely offer bedside ID specialist consultation as or who were transferred to other institutions. was done in other studies of which some reported impressive We saw a threefold increase of the detection of endocarditis reductions of mortality [14, 18, 19]. A policy of routine ID after intervention. Still, most patients with complicated SAB consultations and additional thorough search for infectious did not receive a TEE. complications may lead to additional benefits and could be Eur J Clin Microbiol Infect Dis used for future research. However, the lack of improved out- References come despite increased quality of care underscores once more that SAB is very difficult to manage, even in the absence of 1. Kaasch AJ, Barlow G, Edgeworth JD, Fowler VG, Hellmich M, Hopkins S et al (2014) Staphylococcus aureus bloodstream infec- methicillin resistance. tion: a pooled analysis of five prospective, observational studies. J The high mortality related to SAB will probably continue Inf Secur 68(3):242–251 to rise with the increasing use of invasive techniques such as 2. van Hal SJ, Jensen SO, Vaska VL, Espedido BA, Paterson DL, implementation of cardiac devices, prosthetic joints and Gosbell IB (2012) Predictors of mortality in Staphylococcus aureus bacteremia. Clin Microbiol Rev 25(2):362–386 endovascular prosthesis. In our study, health-care related in- 3. Lauridsen TK, Park L, Tong SYC, Selton-Suty C, Peterson G, fections contributed to 50% of all SABs. Tergooi hospital Cecchi E et al (2015) Echocardiographic findings predict in- serves a relatively old population, which is reflected by the hospital and 1-year mortalityinleft-sidednativevalve high age of in-hospital patients with multiple underlying dis- Staphylococcus aureus endocarditis: analysis from the international collaboration on endocarditis-prospective Echo Cohort Study. Circ eases and nosocomial complications. Cardiovasc Imaging 8(7):e003397 In conclusion, the management of SAB demands the 4. Fowler VG, Olsen MK, Corey GR, Woods CW, Cabell CH, Reller highest level of clinical commitment. The continuous rise of LB et al (2003 Sep 22) Clinical identifiers of complicated the use of invasive techniques and devices emphasizes the Staphylococcus aureus bacteremia. Arch Intern Med 163(17):2066 5. Kanafani ZA, Kourany WM, Fowler VG, Levine DP, Vigliani GA, need for preventive measures and optimal diagnostic and Campion M et al (2009) Clinical characteristics and outcomes of treatment strategies for this high-risk infection. The timely diabetic patients with Staphylococcus aureus bacteremia and endo- detection of complications of SAB requires more than just carditis. Eur J Clin Microbiol Infect Dis Off Publ Eur Soc Clin clinical expertise. Additional diagnostic techniques are need- Microbiol 28(12):1477–1482 6. Palraj BR, Baddour LM, Hess EP, Steckelberg JM, Wilson WR, ed. It was recently shown that imaging with F-FDG PET/CT is Lahr BD et al (2015) Predicting Risk of Endocarditis Using a a valuable technique for early detection of metastatic infec- Clinical Tool (PREDICT): scoring system to guide use of echocar- tious foci that often leads to modification of treatment and thus diography in the management of Staphylococcus aureus bacter- reduces mortality. [30] emia. Clin Infect Dis Off Publ Infect Dis Soc Am 61(1):18–28 7. Mitchell DH, Howden BP (2005) Diagnosis and management of Staphylococcus aureus bacteraemia. Intern Med J 35(Suppl 2): S17–S24 Conclusion 8. Holland TL, Arnold C, Fowler VG (2014) Clinical Management of Staphylococcus aureus bacteremia: a review. JAMA 312(13):1330 9. Corey GR (2009) Staphylococcus aureus bloodstream infections: Implementation of a hospital-wide protocol for the manage- definitions and treatment. Clin Infect Dis 48(s4):S254–S259 ment of uncomplicated and complicated SAB patients resulted 10. Chong YP, Moon SM, Bang K-M, Park HJ, Park S-Y, Kim M-N et in increased adherence to standard of care. More follow-up al (2013) Treatment duration for uncomplicated Staphylococcus aureus bacteremia to prevent relapse: analysis of a prospective ob- blood cultures were obtained, and more echocardiograms servational cohort study. Antimicrob Agents Chemother 57(3): were performed. More patients received adequate antibiotic 1150–1156 therapy—mostly in the form of an increased duration of ther- 11. Johnson LB, Almoujahed MO, Ilg K, Maolood L, Khatib R (2003) apy. More complicated SAB-patients with infective endocar- Staphylococcus aureus bacteremia: compliance with standard treat- ditis were identified but the overall mortality did not decrease. ment, long-term outcome and predictors of relapse. Scand J Infect Dis 35(11–12):782–789 12. Blyth CC, Darragh H, Whelan A, O’Shea JP, Beaman MH, Compliance with ethical standards McCarthy JS (2002) Evaluation of clinical guidelines for the man- agement of Staphylococcus aureus bacteraemia. Intern Med J 32(5– Conflict of interest Dr. de Vries reports personal fees from ACE 6):224–232 Pharma, outside the submitted work. The rest of the authors declare that 13. Verhagen DWM, Van Der Meer JTM, Hamming T, De Jong MD, they have no conflict of interest. Speelman P (2003) Management of patients with Staphylococcus aureus acteraemia in a university hospital: a retrospective study. Scand J Infect Dis 35(8):459–463 Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institu- 14. Fowler VG, Sanders LL, Sexton DJ, Kong L, Marr KA, Gopal AK tional and/or national research committee and with the 1964 Helsinki et al (1998) Outcome of Staphylococcus aureus bacteremia accord- declaration and its later amendments or comparable ethical standards. ing to compliance with recommendations of infectious diseases For this type of study, formal consent is not required. specialists: experience with 244 patients. Clin Infect Dis Off Publ Infect Dis Soc Am. 27(3):478–486 15. Incani A, Hair C, Purnell P, O’Brien DP, Cheng AC, Appelbe A, et al. Staphylococcus aureus bacteraemia: evaluation of the role of transoesophageal echocardiography in identifying clinically unsus- Open Access This article is distributed under the terms of the Creative pected endocarditis. Eur J Clin Microbiol Infect Dis Off Publ Eur Commons Attribution 4.0 International License (http:// Soc Clin Microbiol 2013; 32(8):1003–1008 creativecommons.org/licenses/by/4.0/), which permits unrestricted use, 16. Kern WV (2010) Management of Staphylococcus aureus bacter- distribution, and reproduction in any medium, provided you give appro- emia and endocarditis: progresses and challenges. Curr Opin priate credit to the original author(s) and the source, provide a link to the Infect Dis 23(4):346–358 Creative Commons license, and indicate if changes were made. Eur J Clin Microbiol Infect Dis 17. Holden E, Bashir A, Das I, Morton H, Steadman CD, Nightingale P Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 30(4):1072–1078 et al (2014) Staphylococcus aureus bacteraemia in a UK tertiary referral centre: a Btransoesophageal echocardiogram for all^ policy. 25. Reynolds HR, Jagen MA, Tunick PA, Kronzon I (2003) Sensitivity J Antimicrob Chemother 69(7):1960–1965 of transthoracic versus transesophageal echocardiography for the 18. Borde JP, Batin N, Rieg S, Feik R, Reimling C, Kern WV et al detection of native valve vegetations in the modern era. J Am Soc (2014) Adherence to an antibiotic stewardship bundle targeting Echocardiogr Off Publ Am Soc Echocardiogr 16(1):67–70 Staphylococcus aureus blood stream infections at a 200-bed com- 26. Barton TL, Mottram PM, Stuart RL, Cameron JD, Moir S (2014) munity hospital. Infection 42(4):713–719 Transthoracic echocardiography is still useful in the initial evalua- 19. Jenkins TC, Price CS, Sabel AL, Mehler PS, Burman WJ (2008) tion of patients with suspected infective endocarditis: evaluation of Impact of routine infectious diseases service consultation on the a large cohort at a tertiary referral center. Mayo Clin Proc 89(6): evaluation, management, and outcomes of Staphylococcus aureus 799–805 bacteremia. Clin Infect Dis 46(7):1000–1008 27. Baddour LM, Wilson WR, Bayer AS, Fowler VG, Tleyjeh IM, 20. Honda H, Krauss MJ, Jones JC, Olsen MA, Warren DK (2010) The Rybak MJ et al (2015) Infective endocarditis in adults: diagnosis, value of infectious diseases consultation in Staphylococcus aureus antimicrobial therapy, and management of complications: a scien- bacteremia. Am J Med 123(7):631–637 tific statement for healthcare professionals from the American Heart 21. Bai AD, Showler A, Burry L, Steinberg M, Ricciuto DR, Fernandes Association. Circulation 132(15):1435–1486 T, et al. Impact of infectious disease consultation on quality of care, 28. Kaasch AJ, Fowler VG, Rieg S, Peyerl-Hoffmann G, Birkholz H, mortality, and length of stay in Staphylococcus aureus bacteremia: Hellmich M et al (2011) Use of a simple criteria set for guiding results from a large multicenter cohort study. Clin Infect Dis echocardiography in nosocomial Staphylococcus aureus bacter- [Internet]. 2015 Feb 20 [cited 2016 Mar 16]; Available from: emia. Clin Infect Dis Off Publ Infect Dis Soc Am. 53(1):1):1–1):9 https://doi.org/10.1093/cid/civ120 29. Goto M, Schweizer ML, Vaughan-Sarrazin MS, Perencevich EN, 22. Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG, Ryan T et al Livorsi DJ, Diekema DJ, et al. Association of evidence-based care (2000) Proposed modifications to the Duke criteria for the diagnosis processes with mortality in Staphylococcus aureus bacteremia at of infective endocarditis. Clin Infect Dis Off Publ Infect Dis Soc veterans health administration. Hospitals, 2003-2014. JAMA Am 30(4):633–638 Intern Med 2017; 177 (10):1489 23. Vogel M, Schmitz RPH, Hagel S, Pletz MW, Gagelmann N, 30. Berrevoets MAH, Kouijzer IJE, Aarntzen EHJG, Janssen MJR, De Scherag A et al (2016) Infectious disease consultation for Geus-Oei L-F, Wertheim HFL et al (2017) F-FDG PET/CT opti- Staphylococcus aureus bacteremia—a systematic review and me- mizes treatment in Staphylococcus aureus bacteremia and is asso- ta-analysis. J Inf Secur 72(1):19–28 ciated with reduced mortality. J Nucl Med 58(9):1504–1510 24. Fowler VG, Li J, Corey GR, Boley J, Marr KA, Gopal AK et al (1997) Role of echocardiography in evaluation of patients with http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png European Journal of Clinical Microbiology Infectious Diseases Springer Journals

Implementing a hospital-wide protocol for Staphylococcus aureus bacteremia

Free
10 pages
Loading next page...
 
/lp/springer_journal/implementing-a-hospital-wide-protocol-for-staphylococcus-aureus-clOlZRGo3B
Publisher
Springer Berlin Heidelberg
Copyright
Copyright © 2018 by The Author(s)
Subject
Biomedicine; Medical Microbiology; Internal Medicine
ISSN
0934-9723
eISSN
1435-4373
D.O.I.
10.1007/s10096-018-3284-9
Publisher site
See Article on Publisher Site

Abstract

Staphylococcus aureus bacteraemia (SAB) is associated with high-mortality and complication rates. A multidisciplinary ap- proach is needed to predict, detect and treat complications. In this pre- and post-intervention study, we investigated the effects of a hospital-wide protocol for diagnosis, classification and treatment of SAB. It was hypothesized that complications and endocar- ditis would be better identified and treated. Medical records of SAB patients admitted in 2011 and 2012 (pre) were analysed. In 2013, a protocol, describing risk factors, diagnostic classification and recommended treatment, was implemented. In 2014 and 2015 (post), SAB patients were followed prospectively. Transthoracic (TTE) or transoesophageal cardiac ultrasound (TEE) was chosen following a decision tree. A resident internal medicine acted as contact person. Pre-intervention, 98 patients were eligible for analysis compared to 85 patients post-intervention. Age and number of risk factors were slightly higher post-intervention; other baseline characteristics were similar. Most SAB-patients were classified as complicated (89 and 82% pre- and post- intervention, respectively). Follow-up blood cultures drawn within 2 days after initiating treatment increased from 51 to 85%. Cardiac ultrasounds increased from 44 to 83% for TTE and 13 to 24% for TEE. Endocarditis was more frequently diagnosed (4 vs. 12%). Additionally, duration of antibiotic therapy increased. The 3-month mortality did not change significantly (33% pre- intervention vs. 35% post-intervention; p > 0.05). Introduction of a hospital-wide protocol for SAB management increased standard of care, created awareness among clinicians to properly classify SAB, search for endocarditis and adapt duration of antibiotic treatment. Mortality did not decrease. . . . . Keywords Staphylococcus aureus Bacteraemia Treatment Echocardiography Infective endocarditis Introduction acutely life-threatening complications such as endocarditis, intracardiac abscesses and valve perforation [3]. Important in Staphylococcus aureus bacteraemia (SAB) is associated with the management of SAB is to correctly classify SAB as either a long hospital stay, a high morbidity and high 30-day mor- uncomplicated or complicated and to detect the complications tality numbers ranging from 10 to 30% [1, 2]. The mortality is as quickly as possible. Risk factors for a complicated course associated with the development of serious complications include community acquisition, persistent positive blood cul- such as metastatic abscesses, vertebral osteomyelitis and tures after 48–72 h, persistent fever after 72 h, diabetes mellitus, purulent thrombophlebitis, presence of prosthetic de- vices and immunosuppressive therapy [4–6]. In addition, re- * K. Bolhuis petitive follow-up blood cultures and echocardiography are k.bolhuis@amc.nl recommended for patients with risk factors to detect (imminent) complications. Department of Internal Medicine, Academic Medical Center, The antibiotic treatment of SAB aims to prevent deteriora- Meijbergdreef 9, 1105 AZ Amsterdam, The Netherlands tion of disease and the development of complications. The Department of Medical Microbiology, Tergooi Hospital, Van duration of treatment remains somewhat controversial, but Riebeeckweg 212, 1213 XZ Hilversum, The Netherlands current guidelines regarding management of SAB patients Department of Cardiology, Tergooi Hospital, Van Riebeeckweg 212, recommend at least 14 days of antimicrobial therapy for a 1213 XZ Hilversum, The Netherlands simple bacteraemia with low risk of complications. For Department of Internal Medicine, Tergooi Hospital, Van bacteraemia with risk factors associated with a complicated Riebeeckweg 212, 1213 XZ Hilversum, The Netherlands Eur J Clin Microbiol Infect Dis course due to metastatic infections or endocarditis, at least All events of S. aureus bacteraemia in patients > 16 years 28 days of antimicrobial therapy is recommended [4, 7–10]. were analysed during the study period. A SAB > 12 weeks In contrast to many other blood stream infections, the instal- after cessation of antimicrobial therapy for a previous SAB ment of adequate antibiotic treatment is not enough to warrant was considered to be a distinct event and was included into the a favourable outcome. analysis. Because the study was designed to evaluate the effect Although SAB is among the deadliest bloodstream infec- of implementation of a protocol, exclusion criteria included tions, previous studies showed that potentially complicated contaminated blood culture specimen or any of the following infections are not well recognized by physicians resulting in within 3 days after collection of blood culture sample: death, non-adherence to standard of care therapy in 10 to 41% of withdrawing of therapy or transfer to another facility. SAB cases [11–13]. Detection of complications remains a Pregnancy was an exclusion criterion as well. challenge because the clinical symptoms and findings are non-specific [14, 15]. Despite the identification of several risk Pre-intervention group factors for developing these complications, mortality rates have only improved marginally [16]. The pre-intervention group cohort was studied in retrospect. Non-adherence to standard of care is most often the result of All records of hospitalized patients with blood cultures posi- inadequate duration of antibiotic treatment, lack of follow-up tive for S. aureus between 1 January 2011 and 31 December blood cultures, or omitting echocardiography [12–14]. In par- 2012 were retrieved. The date of the first sample taken was ticular, transoesophageal echocardiography (TEE) is often not considered the date of diagnosis. This group was called the performed due to comorbidity, limited acceptance by patients or pre-intervention group. treating physicians [12], even though other authors showed that performing routinely TEE results in a high incidence of the Intervention protocol most feared complication: infective endocarditis [17]. Multiple studies have shown an increase in adherence to In 2013, accompanying the introduction of a hospital-wide standard of care by routine infectious disease (ID) consulta- protocol, awareness was created by oral and visual presenta- tion and some found that this adherence seemed to preface a tions to all medical staff. The potential complications of SAB significant lower short-term mortality rate (or variable effect and the added value of TEE were emphasized in these educa- on mortality rates) [14, 18–21]. Almost all those studies were tional sessions. All cardiologists were made aware of the pro- performed in large academic, tertiary hospitals, and it remains tocol. The protocol was implemented in all medical and sur- to be confirmed whether this can be generalized to non-aca- gical wards and was designed to function as a simple guideline demic, regional or smaller hospitals. for physicians by classifying patients to groups with uncom- To improve quality of care, we conducted an intervention plicated SAB or with potentially complicated SAB. Within study, a multidisciplinary approach to predict, detect and treat complicated SAB, possible endocarditis was identified as a complications of SAB and adapt the treatment to the risk profiles. separate group. The primary goal of this study was to implement a clear, struc- Patients with SAB were reported by the laboratory of mi- tured, safe and univocal hospital-wide protocol regarding the crobiology as soon as a blood culture became positive for S. management and treatment of patients with SAB and to evaluate aureus. The attending physician was informed about this SAB the effects of this protocol in a regional teaching hospital. and referred to the protocol. According to the protocol SAB was considered complicated when at least one of the following risk factors was present: community acquisition, diabetes Materials and methods mellitus, persistent positive blood cultures after 48/72 h of antibiotic therapy, prosthetic material in situ, persistent fever Study design after 72 h of therapy and metastatic infections. These risk factors are known to be associated with a high risk for a com- This pre- and post-intervention retrospective study was per- plicated course [4–6]. formed in a 550-bed regional teaching hospital in the centre of Before and during the intervention, an ID specialist (PdV) the Netherlands. In cooperation with a cardiologist, medical was working in the hospital. The introduction of the protocol microbiologist and ID specialist, a protocol was designed that introduced a closer involvement of the ID specialist and the describes the risk factors of complicated SAB and includes a dedicated resident internal medicine (KB) to the patient care flow chart for diagnostic classification and recommendations but did not include bedside visits by them. The flow chart for therapy according to the current guidelines. Pre- guided the treating physicians to score these risk factors and intervention data were collected retrospectively from the hos- classify the patients in complicated versus uncomplicated pital records. The post-intervention data were collected (Fig. 1). Flucloxacillin is the drug of first choice in the prospectively. Netherlands for patients with SAB. According to the flowchart, Eur J Clin Microbiol Infect Dis Box 1: Criteria for complicated SAB SAB DM Immune deficiency or suppressive therapy Artificial material in situ Purulent thromboflebitis Persistent fever 72 hrs after start of therapy See Box 1 Positive blood cultures 2-3 days after start of therapy Signs of septic embolisms Community acquired Uncomplicated Complicated SAB SAB Intracardiac device Hemodialysis Spinal infection Non-vertebral osteomyelitis Yes No Diagnosis endocarditis rejected Yes No TTE Positive TEE Negative Inconclusive unless TTE conclusive Complicated SAB Complicated SAB without with endocarditis endocarditis 14 days of > 4 weeks flucloxacillin 6 dd 2 gr i.v. 6 weeks of flucloxacillin 6 dd 2 gr i.v. flucloxacillin 6 dd 1 gr Re-evaluation Fig. 1 Protocol for diagnosis, classification and treatment of SAB. SAB, Staphylococcus aureus bacteraemia; TTE, transthoracic echocardiography; TEE, transoesophageal echocardiography uncomplicated SAB was treated with flucloxacillin intrave- these complicated SAB-patients. If infective endocarditis was nously for 14 days. For complicated SAB, the flowchart advises confirmed, the patient was treated according to national guide- to assess the modified Duke criteria for endocarditis and to lines which includes 6 weeks of IV flucloxacillin and 3 to 5 days perform a transthoracic echocardiography (TTE) for all pa- of low-dose gentamicin, with or without rifampicin (Fig. 2). tients. When the echocardiographic results were uncertain or inconclusive, a transoesophageal echocardiography (TEE) Post-intervention group was suggested. When indicated, echocardiography was con- ducted by the attending cardiologist, usually the same day. The post-intervention group consisted of all SAB episodes of Depending on the outcome of the echocardiography, at least admitted patients between 1 January 2014 and 31 December 4 weeks of IV flucloxacillin therapy was recommended for 2015. These patients were followed prospectively with respect Eur J Clin Microbiol Infect Dis Fig. 2 Flow diagram of events 219 events of SAB included in the study. SAB, Staphylococcus aureus bacteraemia 98 events SAB 121 events SAB 23 events excluded 13 events excluded 4 age 1 age 3 contamination 3 contamination 10 missing follow up 9 missing follow up 4 relapse SAB < 12 wks 2 other 98 events SAB analysed 85 events SAB analysed to the routinely collected data. Consequently, in this study, When the source of infection was, for example, an indwell- prospectively collected data were compared to data of recent ing peripheral intravenous catheter, a pacemaker or a central historic cohort. In the post-intervention period, a dedicated venous catheter, this was defined as a health care-related in- resident internal medicine, advising microbiologist and ID fection, and removal of these intravenous devices was scored. specialist actively recommended the attending physician to Adequate antibiotic treatment was defined as the initiation consult and follow the protocol, which is available on the of parenteral administration of antibiotics with documented hospital’s intranet pages. The management of individual pa- activity against the isolated S. aureus strain, with adequate tients including actually performing echocardiography dosage and duration of therapy (14 days for uncomplicated remained the responsibility of the treating physician. and at least 28 days for complicated SAB). The study was approved by the hospital’s scientific review Infective endocarditis was scored based on echocardiogra- committee responsible for improving the quality of care. This phy, modified Duke criteria [22], strongly suggestive clinical study did not collect experimental data, and since the data findings, PET-CT or endocarditis found at autopsy. collection itself is part of the routine quality of care assessment Relapse of infection was defined as recurrent isolation of S. system, no additional ethical clearance was required. aureus in blood cultures within 12 weeks after cessation of antimicrobial therapy. Patient follow-up The most important post-discharge outcomes were relapse of SAB and mortality. In case of relapse, it was expected that Patients were followed until death or 6 months after diagnosis, most if not all patients would be referred and admitted to the whichever of the events occurred first. The medical records hospital due to the complicated natural course of SAB and were analysed, and data were collected by a single reasoning that SAB is not self-limiting in case no intravenous investigator. antimicrobial treatment is offered to the patient. Following this theory, possible relapses were scored by analysing the medical Data collection records. Mortality outside the hospital was collected by medical Medical records were analysed for patient characteristics, records as well. Tergooi hospital receives and records the in- medical history, underlying diseases, source of infection, pres- formation of deaths of all patients, even if they do not die in ence of prosthetic devices and other known risk factors. the hospital. Information about choice, dosage and duration of antimicro- bial therapy was recorded; number of follow-up blood cultures and number and results of echocardiograms were also includ- Primary endpoint output ed. All SAB-related complications, including the occurrence of endocarditis, relapse of infection and mortality, were Adherence to standards of care was scored by three indicators: recorded. (1) obtaining follow-up blood cultures 2 to 4 days after initi- Hospital-acquired bacteraemia was defined by isolation of ation of (adequate) therapy; (2) administration of adequate S. aureus from a blood culture obtained > 72 h after hospital therapy scored by the use of appropriate antibiotics, optimal admission. Purulent thrombophlebitis was diagnosed, based parenteral treatment dosage and duration; and (3) performing on clinical symptoms and signs. echocardiography. Eur J Clin Microbiol Infect Dis Data analysis Results Data were initially entered in Microsoft Excel and analysed in A number of 219 patients was diagnosed with at least one SPSS (v. 23, SPSS Statistics for Windows, Version 23.0. blood culture specimen positive for S. aureus.Thirty-sixpa- Released 2015. IBM Corp. Armonk, NY). tients were excluded, six due to contaminated blood cultures, Parametric tests were used for comparison of normally dis- five patients were < 16 years old, four because of a relapse tributed variables. Non-parametric tests were used otherwise. within 12 weeks and 19 were lost to follow-up within 3 days Categorical variables were analysed with chi-square tests. because of death, withdrawing of therapy or translocation to Statistically significant difference was accepted when p < another facility. A total of 98 patients remained eligible for 0.05. analysis in the pre-intervention group and 85 patients in the post-intervention group (Fig. 1). Data availability The datasets generated during and/or Demographic characteristics are shown in Table 1.Two analysed during the current study are available from the cor- patients, both in the pre-intervention group, suffered from responding author on reasonable request. more than one episode of S. aureus bacteraemia (defined as Table 1 Patient demographic Variable Pre-intervention Post-intervention p value characteristics. Staphylococcus (n =98) group (n =85) aureus bacteraemia before (2011– 2012) and after a protocol Age, median years (min, max) 69, 65 (19, 94) 76.49 (35, 98) 0.016 adherence intervention (2014– 2015) Female sex 47 (48%) 35 (41%) 0.44 Classification SAB Uncomplicated SAB 11 (11%) 15 (18%) 0.303 Complicated SAB 87 (89%) 70 (82%) Risk factors Community acquisition 63 (64%) 55 (65%) Diabetes mellitus 28 (29%) 26 (31%) Positive follow-up blood cultures 48/72 h 13 (13%) 21 (25%) Prosthetic material 31 (32%) 29 (34%) Persistent fever after 72 h 4 (4%) 9 (11%) Catheter-related and in situ 3 (3%) 2 (2%) Immunocompromised 12 (12%) 9 (11%) Metastatic infections 23 (23%) 23 (27%) Purulent thrombophlebitis 11 (11%) 0 (0%) Malignancy 18 (18%) 10 (12%) Alcohol 10 (10%) 7 (8%) Haemodialysis 3 (3%) 1 (1%) No. of risk factors 0 11 (11%) 15 (18%) 0.590 1 27 (28%) 16 (19%) 2 31 (32%) 28 (33%) 3 22 (22%) 16 (19%) 45(5%) 6(7%) 53(3%) 4(5%) Source of infection Health care-related 50 (51%) 37 (44%) Surgery 11 (22%) 6 (16%) Urinary tract/CAD 6 (12%) 8 (22%) CVC/PAC 11 (22%) 5 (14%) Peripheral catheter 12 (24%) 11 (30%) Unknown source 26 (52%) 21 (57%) Patient demographics. SAB, Staphylococcus aureus bacteraemia; CAD, urine catheter; CVC, central venous catheter; PAC, port a cath; NS, not significant. A p value < 0.05 is considered statistically significant Eur J Clin Microbiol Infect Dis a new episode of SAB > 12 weeks post-therapy of the previ- classified as complicated (89 and 82% pre- and post-interven- ous episode of SAB). In the first patient, the episodes were tion, respectively). separated by 9 months, and in the second patient, the episodes The adequacy of care for SAB increased during the inter- were separated by 3 years. vention (Table 2). Firstly, the frequency of drawing follow-up The median age was higher in the post-intervention group blood cultures 48–72 h after initiating treatment significantly compared to the pre-intervention group, and malignancy was increased, resulting in the identification of more patients with more frequent in the pre-intervention group, but other baseline positive blood cultures during follow-up (15 vs. 25%). In both characteristics were similar. Most SAB patients were groups, the proportion of positive blood cultures remained the Table 2 Quality of care Diagnostic and therapeutic management of Staphylococcus aureus bacteraemia, indicators before (2011–2012) and after a protocol adherence intervention (2014–2015) Variable Pre-intervention (n = 98) Post-intervention p value (n =85) Diagnostic workup Echocardiography performed in complicated SAB patients TTE 38 (44%) 58 (83%) <0.001 TEE 11 (13%) 17 (24%) 0.09 2nd blood cultures obtained after 48–72 h 50 (51%) 72 (85%) <0.001 Adequate antibiotic therapy Yes 38 (39%) 60 (71%) <0.001 No 59 (58%) 24 (28%) Not started 4 0 Inadequate length 53 24 Inadequate dose 3 0 Unknown 1 (1%) 1 (1%) Intravenous antibiotic treatment < 14 days 32 (33%) 8 (9%) ≥ 14 days 44 (45%) 58 (68%) Lost to FU < 14 days 18 (18%) 19 (22%) Death/withdrawing therapy 15 16 Transfer 2 2 Unknown 1 1 Adequate AB not started 4 (4%) 0 (0%) Antibiotics stopped due to death or withdrawing therapy Adequate AB not started 4 0 Yes 23 21 No 65 57 Transfer 5 5 Unknown 1 2 Treatment duration (mean) 17.3 (n = 65) 23.0 (n =57) 0.014 Uncomplicated SAB (mean/days) 10.3 (n =7) 13.8 (n =15) Complicated SAB (days, mean) 18.2 (n = 58) 26.3 (n =42) Length of hospital stay (mean, days) 27.5 22.4 NS SAB, Staphylococcus aureus bacteraemia; TTE, transthoracic echocardiography; TEE, transoesophageal echocar- diography; FU, follow-up; AB antibiotic treatment; NS, not significant. A p value < 0.05 is considered statistically significant. Mean treatment duration. Patients were excluded when adequate antibiotics were not started or if antibiotic therapy was stopped due to death, withdrawing therapy or transfer to a different health care clinic Echocardiography in complicated SAB patients Length of hospital stay was calculated from date positive blood culture was taken till day of discharge. Patients transferred to different hospital were excluded Eur J Clin Microbiol Infect Dis same (29 vs. 29%). Secondly, the mean duration of antibiotic It created awareness among clinicians to properly classify therapy significantly increased (17.3 vs. 23.0 days),resulting SAB, search for endocarditis, choose the right antibiotic reg- in more patients receiving adequate antimicrobial therapy (39 imen and adapt duration of antibiotic treatment. However, it vs. 71%). And, thirdly, among patients with complicated did not improve survival. SAB, more TTEs were performed (44 to 83%), and the num- One of the strong points of this study is its real-life setting. A ber of TEEs increased from 13 to 24%. Consequently, endo- small intervention such as the introduction of this protocol cre- carditis was detected more frequently (12%) during the inter- ates knowledge about the treatment approach involving patients vention period than before (4%). The increase of TEE did not with SAB among a wide range of medical specialists and there- decrease the proportion of positive TEEs (31 vs. 42%). This by directly improves standard of care. The introduction of pro- underlines the urgency to perform a TEE. tocols and local guidelines probably belong to the most com- More infectious complications were detected in the post- mon and feasible interventions in hospitals. The situation in our intervention group (39 vs. 45%), and the mean duration of hospital reflects the situation of many hospitals, and this inter- hospital stay, counting from the date of the first positive blood vention can easily be implemented in many other hospitals. In culture, decreased from 27 days before intervention to 23 days our knowledge, this study is among the largest studies per- post-intervention. In addition, lower number of relapses with- formed in the setting of a general hospital [23]. in 12 weeks were reported (5 vs. 2%) (Table 3). In this study, the motivations to change practices were not The 4- and 12-week mortality did not change significantly studied in depth. However, we encountered some expressions (26 vs. 26% and 33 vs. 35% in the pre-intervention vs. post- of inertia towards change. These were mainly the lack of intervention group; p >0.05). knowledge and sense of urgency with respect to the poor prognosis of SAB and its treacherous presentation of compli- cations. Another explanation for the relatively persistent low number of TEE might be the fact that, in exceptional cases, the Discussion protocol permitted clinicians to forego TEE in patients with- out any clinical signs of endocarditis, with excellent acoustic This study shows that introducing a hospital-wide protocol for window on TTE and no vegetations or valvular regurgitation management of SAB in a general teaching hospital improved (Fig. 1). We did not encounter any expressions of mistrust by the standard of care: The number of follow-up blood cultures the attending physicians towards the resident or ID specialist. significantly increased by two thirds, antimicrobial therapy There was often reluctance to conduct transoesophageal became significantly more adequate and the number of TTE echocardiography because it was considered too invasive for and TEE almost doubled post-intervention. ill and elderly patients and falsely relying on the acuity of the Table 3 Secondary outcomes Variable Pre-intervention (n = 98) Post-intervention group (n =85) Infectious complications 38 (39%) 38 (45%) Endocarditis 4 (4%) 10 (12%) Spondylodiscitis 8 (8%) 5 (6%) Abscesses 10 (10%) 13 (15%) Septic arthritis 3 (3%) 6 (7%) Prosthetic joint infection 5 (5%) 7 (8%) Endovascular infection 2 (2%) 2 (2%) Relapse within 3 months 5 (5%) 2 (2%) Cerebral septic embolism 4 (4%) 4 (5%) Mortality 30-day mortality 25 (26%) 22 (26%) Uncomplicated 3 1 Complicated 22 21 Unknown/lost to FU 2 1 90-day mortality 32 (33%) 30 (35%) Uncomplicated 4 3 Complicated 28 27 Unknown/lost to FU 3 1 Secondary outcome: complications. FU, follow-up Eur J Clin Microbiol Infect Dis cardiac valve images during transthoracic echocardiography. Endocarditis is among the most feared complications of PET-CT was available already before the intervention but locat- SAB with a reported incidence of 25–29% and a 1-year mor- ed in another hospital with which it was shared. After the study, tality rate up to 43% [3, 15, 24]. This complication is difficult a PET-CT facility was installed at the location of the study. It is to diagnose since the initial clinical symptoms and signs are nowadays increasingly used to detect complications of SAB. insensitive and non-specific [14, 15]. Echocardiography is The contribution of purulent phlebitis with bacteraemia recommended as primary tool to diagnose endocarditis but decreased (Table 1) substantially. This reduced the total num- clinicians should be aware about its limitations. TTE is insen- ber of complicated SAB while the proportional detection of sitive (sensitivity of 55%) compared with TEE for the detec- persistent bacteraemia increased. tion of native valve vegetations [25]. Although others suggest Previous studies showed the beneficial effects of imple- that TTE has a high negative predictive value of 97%, suffi- mentation of hospital-wide diagnostic and therapeutic guide- cient data are lacking, and further research is needed to better lines and of the standard involvement of an infectious disease identify those patients with SAB for whom TEE is required to consultant on adherence to standard of care [13, 18]. Borde et detect endocarditis. [26] The sensitivity of a TEE is 90–100% al. implemented a standardized bundle approach to the treat- and is the preferred screening tool in patients with a compli- ment of S. aureus bacteraemia which resulted in an increase of cated SAB and suspected infectious endocarditis. [27] conducting TEE by 42% and collecting follow-up blood cul- A recent Canadian multicentre study including 847 patients tures by 49% [18]. Appropriate therapy within 72 h of detec- compared the management of SAB in patients receiving in- tion of SAB increased by 81%. fectious disease (ID) consultation and patients without ID In our study, we also found significant effects on quality of consultation of an ID specialist. The number of patients re- care indicators; the adequacy of dosing and duration of anti- ceiving echocardiography increased, but despite bedside con- microbial treatment and on documentation of infectious com- sultation of an ID specialist, TEE was only performed in 17% plications. An interesting phenomenon was found by of patients [21]. Others suggested a simple set of criteria for analysing the duration of hospital stay before and after inter- nosocomial SAB to identify patients with low risk of infec- vention. Despite a longer mean antimicrobial treatment, the tious endocarditis and who might not routinely require TEE mean duration of hospital stay decreased after protocol imple- [28]. mentation. We hypothesize that, greater awareness among cli- A recently published large retrospective 12-year cohort nicians, more patients receiving optimal treatment and the use study showed that the use of three evidence-based care pro- of more diagnostic tools to find infectious complications pre- cesses, appropriate antibiotic therapy, echocardiography and vents serious complications during hospital stay and leads to consultation of an ID specialist, was associated with a reduc- faster hospital discharge. tion of 30-day mortality by 67% [29]. The risk-adjusted mor- The year 2013 was not analysed but used to implement the tality decreased from 23.5 to 18.2%, lower than in our hospi- SAB protocol. After proper implementation, we started to collect tal. In our study, we also found significant effects on adher- data prospectively from 2014 onwards. The long study period, ence to guidelines and the use of echocardiography after the 5 years, has the advantage of excluding season effects. Although introduction of a protocol. Although execution of our protocol this may also introduce effects of changing health care and out- was monitored by a dedicated person (KB), this is not equal to comes over time, we are not aware of important structural chang- a bedside consultation of an ID specialist. es regarding the medical care in this hospital during this period. Despite the improved adherence to guidelines in treating The only effect was a reduction of purulent phlebitis. The extra patients with SAB in the post-intervention group, overall mor- attention and instructions about SAB and its causes might have tality did not decrease in our study. Besides the fact that the improved awareness about the risks and better measures to pre- study itself was not designed to see changes in mortality, the vent infections due to peripheral catheters. slightly older age and higher incidence of risk factors for com- Retrospective studies may be subject to different forms of plicated SAB in the post-intervention group could be an addi- bias, but in this study, we captured all patients with document- tional explanation for the absent decrease in mortality. In ad- ed SAB in a similar manner before (the retrospective part) and dition, there was no adjustment for comorbidities and severity after intervention. This makes bias less likely. This study was of disease performed on admission in both groups. Another not blinded but since data collection was similar before inter- possible explanation could be the fact that in this study, the vention as after, we do not consider this a probable source of management strategies, including ordering for echocardiogra- bias. Follow-up was incomplete for several patients who were phy, remained the responsibility of the treating physician. We discharged from the hospital for palliative/terminal home care did not routinely offer bedside ID specialist consultation as or who were transferred to other institutions. was done in other studies of which some reported impressive We saw a threefold increase of the detection of endocarditis reductions of mortality [14, 18, 19]. A policy of routine ID after intervention. Still, most patients with complicated SAB consultations and additional thorough search for infectious did not receive a TEE. complications may lead to additional benefits and could be Eur J Clin Microbiol Infect Dis used for future research. However, the lack of improved out- References come despite increased quality of care underscores once more that SAB is very difficult to manage, even in the absence of 1. Kaasch AJ, Barlow G, Edgeworth JD, Fowler VG, Hellmich M, Hopkins S et al (2014) Staphylococcus aureus bloodstream infec- methicillin resistance. tion: a pooled analysis of five prospective, observational studies. J The high mortality related to SAB will probably continue Inf Secur 68(3):242–251 to rise with the increasing use of invasive techniques such as 2. van Hal SJ, Jensen SO, Vaska VL, Espedido BA, Paterson DL, implementation of cardiac devices, prosthetic joints and Gosbell IB (2012) Predictors of mortality in Staphylococcus aureus bacteremia. Clin Microbiol Rev 25(2):362–386 endovascular prosthesis. In our study, health-care related in- 3. Lauridsen TK, Park L, Tong SYC, Selton-Suty C, Peterson G, fections contributed to 50% of all SABs. Tergooi hospital Cecchi E et al (2015) Echocardiographic findings predict in- serves a relatively old population, which is reflected by the hospital and 1-year mortalityinleft-sidednativevalve high age of in-hospital patients with multiple underlying dis- Staphylococcus aureus endocarditis: analysis from the international collaboration on endocarditis-prospective Echo Cohort Study. Circ eases and nosocomial complications. Cardiovasc Imaging 8(7):e003397 In conclusion, the management of SAB demands the 4. Fowler VG, Olsen MK, Corey GR, Woods CW, Cabell CH, Reller highest level of clinical commitment. The continuous rise of LB et al (2003 Sep 22) Clinical identifiers of complicated the use of invasive techniques and devices emphasizes the Staphylococcus aureus bacteremia. Arch Intern Med 163(17):2066 5. Kanafani ZA, Kourany WM, Fowler VG, Levine DP, Vigliani GA, need for preventive measures and optimal diagnostic and Campion M et al (2009) Clinical characteristics and outcomes of treatment strategies for this high-risk infection. The timely diabetic patients with Staphylococcus aureus bacteremia and endo- detection of complications of SAB requires more than just carditis. Eur J Clin Microbiol Infect Dis Off Publ Eur Soc Clin clinical expertise. Additional diagnostic techniques are need- Microbiol 28(12):1477–1482 6. Palraj BR, Baddour LM, Hess EP, Steckelberg JM, Wilson WR, ed. It was recently shown that imaging with F-FDG PET/CT is Lahr BD et al (2015) Predicting Risk of Endocarditis Using a a valuable technique for early detection of metastatic infec- Clinical Tool (PREDICT): scoring system to guide use of echocar- tious foci that often leads to modification of treatment and thus diography in the management of Staphylococcus aureus bacter- reduces mortality. [30] emia. Clin Infect Dis Off Publ Infect Dis Soc Am 61(1):18–28 7. Mitchell DH, Howden BP (2005) Diagnosis and management of Staphylococcus aureus bacteraemia. Intern Med J 35(Suppl 2): S17–S24 Conclusion 8. Holland TL, Arnold C, Fowler VG (2014) Clinical Management of Staphylococcus aureus bacteremia: a review. JAMA 312(13):1330 9. Corey GR (2009) Staphylococcus aureus bloodstream infections: Implementation of a hospital-wide protocol for the manage- definitions and treatment. Clin Infect Dis 48(s4):S254–S259 ment of uncomplicated and complicated SAB patients resulted 10. Chong YP, Moon SM, Bang K-M, Park HJ, Park S-Y, Kim M-N et in increased adherence to standard of care. More follow-up al (2013) Treatment duration for uncomplicated Staphylococcus aureus bacteremia to prevent relapse: analysis of a prospective ob- blood cultures were obtained, and more echocardiograms servational cohort study. Antimicrob Agents Chemother 57(3): were performed. More patients received adequate antibiotic 1150–1156 therapy—mostly in the form of an increased duration of ther- 11. Johnson LB, Almoujahed MO, Ilg K, Maolood L, Khatib R (2003) apy. More complicated SAB-patients with infective endocar- Staphylococcus aureus bacteremia: compliance with standard treat- ditis were identified but the overall mortality did not decrease. ment, long-term outcome and predictors of relapse. Scand J Infect Dis 35(11–12):782–789 12. Blyth CC, Darragh H, Whelan A, O’Shea JP, Beaman MH, Compliance with ethical standards McCarthy JS (2002) Evaluation of clinical guidelines for the man- agement of Staphylococcus aureus bacteraemia. Intern Med J 32(5– Conflict of interest Dr. de Vries reports personal fees from ACE 6):224–232 Pharma, outside the submitted work. The rest of the authors declare that 13. Verhagen DWM, Van Der Meer JTM, Hamming T, De Jong MD, they have no conflict of interest. Speelman P (2003) Management of patients with Staphylococcus aureus acteraemia in a university hospital: a retrospective study. Scand J Infect Dis 35(8):459–463 Ethical approval All procedures performed in studies involving human participants were in accordance with the ethical standards of the institu- 14. Fowler VG, Sanders LL, Sexton DJ, Kong L, Marr KA, Gopal AK tional and/or national research committee and with the 1964 Helsinki et al (1998) Outcome of Staphylococcus aureus bacteremia accord- declaration and its later amendments or comparable ethical standards. ing to compliance with recommendations of infectious diseases For this type of study, formal consent is not required. specialists: experience with 244 patients. Clin Infect Dis Off Publ Infect Dis Soc Am. 27(3):478–486 15. Incani A, Hair C, Purnell P, O’Brien DP, Cheng AC, Appelbe A, et al. Staphylococcus aureus bacteraemia: evaluation of the role of transoesophageal echocardiography in identifying clinically unsus- Open Access This article is distributed under the terms of the Creative pected endocarditis. Eur J Clin Microbiol Infect Dis Off Publ Eur Commons Attribution 4.0 International License (http:// Soc Clin Microbiol 2013; 32(8):1003–1008 creativecommons.org/licenses/by/4.0/), which permits unrestricted use, 16. Kern WV (2010) Management of Staphylococcus aureus bacter- distribution, and reproduction in any medium, provided you give appro- emia and endocarditis: progresses and challenges. Curr Opin priate credit to the original author(s) and the source, provide a link to the Infect Dis 23(4):346–358 Creative Commons license, and indicate if changes were made. Eur J Clin Microbiol Infect Dis 17. Holden E, Bashir A, Das I, Morton H, Steadman CD, Nightingale P Staphylococcus aureus bacteremia: experience in 103 patients. J Am Coll Cardiol 30(4):1072–1078 et al (2014) Staphylococcus aureus bacteraemia in a UK tertiary referral centre: a Btransoesophageal echocardiogram for all^ policy. 25. Reynolds HR, Jagen MA, Tunick PA, Kronzon I (2003) Sensitivity J Antimicrob Chemother 69(7):1960–1965 of transthoracic versus transesophageal echocardiography for the 18. Borde JP, Batin N, Rieg S, Feik R, Reimling C, Kern WV et al detection of native valve vegetations in the modern era. J Am Soc (2014) Adherence to an antibiotic stewardship bundle targeting Echocardiogr Off Publ Am Soc Echocardiogr 16(1):67–70 Staphylococcus aureus blood stream infections at a 200-bed com- 26. Barton TL, Mottram PM, Stuart RL, Cameron JD, Moir S (2014) munity hospital. Infection 42(4):713–719 Transthoracic echocardiography is still useful in the initial evalua- 19. Jenkins TC, Price CS, Sabel AL, Mehler PS, Burman WJ (2008) tion of patients with suspected infective endocarditis: evaluation of Impact of routine infectious diseases service consultation on the a large cohort at a tertiary referral center. Mayo Clin Proc 89(6): evaluation, management, and outcomes of Staphylococcus aureus 799–805 bacteremia. Clin Infect Dis 46(7):1000–1008 27. Baddour LM, Wilson WR, Bayer AS, Fowler VG, Tleyjeh IM, 20. Honda H, Krauss MJ, Jones JC, Olsen MA, Warren DK (2010) The Rybak MJ et al (2015) Infective endocarditis in adults: diagnosis, value of infectious diseases consultation in Staphylococcus aureus antimicrobial therapy, and management of complications: a scien- bacteremia. Am J Med 123(7):631–637 tific statement for healthcare professionals from the American Heart 21. Bai AD, Showler A, Burry L, Steinberg M, Ricciuto DR, Fernandes Association. Circulation 132(15):1435–1486 T, et al. Impact of infectious disease consultation on quality of care, 28. Kaasch AJ, Fowler VG, Rieg S, Peyerl-Hoffmann G, Birkholz H, mortality, and length of stay in Staphylococcus aureus bacteremia: Hellmich M et al (2011) Use of a simple criteria set for guiding results from a large multicenter cohort study. Clin Infect Dis echocardiography in nosocomial Staphylococcus aureus bacter- [Internet]. 2015 Feb 20 [cited 2016 Mar 16]; Available from: emia. Clin Infect Dis Off Publ Infect Dis Soc Am. 53(1):1):1–1):9 https://doi.org/10.1093/cid/civ120 29. Goto M, Schweizer ML, Vaughan-Sarrazin MS, Perencevich EN, 22. Li JS, Sexton DJ, Mick N, Nettles R, Fowler VG, Ryan T et al Livorsi DJ, Diekema DJ, et al. Association of evidence-based care (2000) Proposed modifications to the Duke criteria for the diagnosis processes with mortality in Staphylococcus aureus bacteremia at of infective endocarditis. Clin Infect Dis Off Publ Infect Dis Soc veterans health administration. Hospitals, 2003-2014. JAMA Am 30(4):633–638 Intern Med 2017; 177 (10):1489 23. Vogel M, Schmitz RPH, Hagel S, Pletz MW, Gagelmann N, 30. Berrevoets MAH, Kouijzer IJE, Aarntzen EHJG, Janssen MJR, De Scherag A et al (2016) Infectious disease consultation for Geus-Oei L-F, Wertheim HFL et al (2017) F-FDG PET/CT opti- Staphylococcus aureus bacteremia—a systematic review and me- mizes treatment in Staphylococcus aureus bacteremia and is asso- ta-analysis. J Inf Secur 72(1):19–28 ciated with reduced mortality. J Nucl Med 58(9):1504–1510 24. Fowler VG, Li J, Corey GR, Boley J, Marr KA, Gopal AK et al (1997) Role of echocardiography in evaluation of patients with

Journal

European Journal of Clinical Microbiology Infectious DiseasesSpringer Journals

Published: May 31, 2018

References

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off