Christine B. Teng

Randomized Trial of Rapid Multiplex Polymerase Chain Reaction–Based Blood Culture Identification and Susceptibility Testing

BACKGROUND The value of rapid, panel-based molecular diagnostics for positive blood culture bottles (BCBs) has not been rigorously assessed. We performed a prospective randomized controlled trial evaluating outcomes associated with rapid multiplex PCR (rmPCR) detection of bacteria, fungi, and resistance genes directly from positive BCBs. METHODS A total of 617 patients with positive BCBs underwent stratified randomization into 3 arms: standard BCB processing (control, n = 207), rmPCR reported with templated comments (rmPCR, n = 198), or rmPCR reported with templated comments and real-time audit and feedback of antimicrobial orders by an antimicrobial stewardship team (rmPCR/AS, n = 212). The primary outcome was antimicrobial therapy duration. Secondary outcomes were time to antimicrobial de-escalation or escalation, length of stay (LOS), mortality, and cost. RESULTS Time from BCB Gram stain to microorganism identification was shorter in the intervention group (1.3 hours) vs control (22.3 hours) (P < .001). Compared to the control group, both intervention groups had decreased broad-spectrum piperacillin-tazobactam (control 56 hours, rmPCR 44 hours, rmPCR/AS 45 hours; P = .01) and increased narrow-spectrum β-lactam (control 42 hours, rmPCR 71 hours, rmPCR/AS 85 hours; P = .04) use, and less treatment of contaminants (control 25%, rmPCR 11%, rmPCR/AS 8%; P = .015). Time from Gram stain to appropriate antimicrobial de-escalation or escalation was shortest in the rmPCR/AS group (de-escalation: rmPCR/AS 21 hours, control 34 hours, rmPCR 38 hours, P < .001; escalation: rmPCR/AS 5 hours, control 24 hours, rmPCR 6 hours, P = .04). Groups did not differ in mortality, LOS, or cost. CONCLUSIONS rmPCR reported with templated comments reduced treatment of contaminants and use of broad-spectrum antimicrobials. Addition of antimicrobial stewardship enhanced antimicrobial de-escalation. CLINICAL TRIALS REGISTRATION NCT01898208.

Timing of susceptibility-based antifungal drug administration in patients with Candida bloodstream infection: correlation with outcomes

OBJECTIVES We sought to determine the impact of timing of appropriate antifungal therapy, as assessed by susceptibility results, on patient survival. METHODS Patients ≥16 years of age with first episodes of candidaemia during 2001-09 were included. Clinical data were collected retrospectively, including time to appropriate antifungal therapy and patient survival. RESULTS The study population included 446 patients [243 (54%) female, mean age 53 years] with candidaemia, 380 (85%) of whom had antifungal susceptibility data. Candida albicans was the most common pathogen (221, 50%) followed by Candida glabrata (99, 22%), Candida parapsilosis (59, 13%), Candida tropicalis (48, 11%) and Candida krusei (6, 1%). Appropriate antifungal therapy consisted of fluconazole (177, 40%), an echinocandin (125, 28%), amphotericin B (41, 9%) and voriconazole (6, 1%); 97 (22%) failed to receive appropriate antifungal therapy. The 30 day mortality was 34% (151/446) and there was no clear relationship between time from positive culture to receipt of appropriate antifungal therapy and 30 day survival. On multivariable Cox regression, increased APACHE II score [hazard ratio (HR) 1.11, 95% CI 1.09-1.13, P<0.001], cirrhosis (HR 2.15, 95% CI 1.48-3.13, P<0.001) and HIV infection (HR 2.03, 95% CI 1.11-3.72, P=0.02) were independent predictors of mortality. A secondary analysis requiring patients in the early treatment group to have received ≥24 h of effective antifungal therapy did show a significant mortality benefit to receiving antifungal treatment within 72 h of a positive blood culture being drawn (30 day mortality for early treatment: 27% versus 40%, P=0.004; HR for mortality with delayed treatment on multivariable analysis: 1.41, 95% CI 1.01-1.98, P=0.045). CONCLUSIONS Candida bloodstream infection is associated with high mortality, despite timely receipt of appropriate antifungal therapy.

Nonconcordance with Surgical Site Infection Prevention Guidelines and Rates of Surgical Site Infections for General Surgical, Neurological, and Orthopedic Procedures

ABSTRACT Surgical site infection (SSI) is a common and preventable complication of surgery, but the relative importance of individual measures recommended by guidelines has not been determined. Elective general surgical, neurological, and orthopedic procedures requiring antibiotic prophylaxis from a 3-month period were retrospectively studied to determine concordance with SSI prevention guidelines and to identify factors which predicted the development of SSIs. A total of 216 surgeries were reviewed, with 18 SSIs (8.3%). A mean of 1.4 antibiotic prophylaxis errors per surgery were identified, with correct antibiotic type identified for 64%, antibiotic timing for 83%, supplemental antibiotic dosing for 34%, and antibiotic duration of less than 24 h for 44%. Normothermia was present in 79% of surgeries, and normoglycemia was present in 17%. Univariate analysis of the SSI rate identified four significant factors. Antibiotic prophylaxis for less than 24 h postoperatively (odds ratio [OR], 0.213; 95% confidence interval [95% CI]0.060 to 0.757) and neurosurgery (OR, 0.118; 95% CI, 0.15 to 0.903) correlated with a reduced risk of SSI. The mean number of prophylaxis errors (OR, 1.6; 95% CI, 1.02 to 2.4) and a duration of surgical drainage for more than 3 days (OR, 2.679; 95% CI, 1.009 to 7.113) predicted SSI. By multivariate analysis, errors in individual antibiotic prophylaxis measures were not significantly associated with SSI; however, the presence of more than two errors was significant (OR, 4.030; 95% CI, 1.018 to 15.96). A strong correlation was identified between the degree of concordance to SSI prevention guidelines and the SSI rate (P = 0.001, Mantel-Haenszel linear-by-linear association chi-square test).

Safety and clinical outcomes of carbapenem de-escalation as part of an antimicrobial stewardship programme in an ESBL-endemic setting

OBJECTIVES To evaluate the safety and clinical outcomes of patients who received carbapenem de-escalation as guided by an antimicrobial stewardship programme (ASP) in a setting where ESBL-producing Enterobacteriaceae are endemic. METHODS Patients receiving meropenem or imipenem underwent a prospective ASP review for eligibility for de-escalation according to defined institutional guidelines. Patients in whom carbapenem was de-escalated or not de-escalated, representing the acceptance and rejection of the ASP recommendation, respectively, were compared. The primary outcome was the clinical success rate; secondary outcomes included the 30 day readmission and mortality rates, the duration of carbapenem therapy, the incidence of adverse drug reactions due to antimicrobials, the acquisition of carbapenem-resistant Gram-negative bacteria and the occurrence of Clostridium difficile-associated diarrhoea (CDAD). RESULTS The de-escalation recommendations for 300 patients were evaluated; 204 (68.0%) were accepted. The patient demographics and disease severity were similar. The clinical success rates were similar [de-escalated versus not de-escalated, 183/204 (89.7%) versus 85/96 (88.5%), P=0.84], as was the survival at hospital discharge [173/204 (84.8%) versus 79/96 (82.3%), P=0.58]. In the de-escalated group, the duration of carbapenem therapy was shorter (6 versus 8 days, P<0.001), the rate of adverse drug reactions was lower [11/204 (5.4%) versus 12/96 (12.5%), P=0.037], there was less diarrhoea [9/204 (4.4%) versus 12/96 (12.5%), P=0.015], there was a lower incidence of carbapenem-resistant Acinetobacter baumannii acquisition [4/204 (2.0%) versus 7/96 (7.3%), P=0.042] and there was a lower incidence of CDAD [2/204 (1.0%) versus 4/96 (4.2%), P=0.081]. CONCLUSIONS This study suggests that the ASP-guided de-escalation of carbapenems led to comparable clinical success, fewer adverse effects and a lower incidence of the development of resistance. This approach is safe and practicable, and should be a key component of an ASP.

A multicenter case-case control study for risk factors and outcomes of extensively drug-resistant Acinetobacter baumannii bacteremia.

OBJECTIVE Extensively drug resistant (XDR) Acinetobacter baumannii infections are increasing. Knowledge of risk factors can help to prevent these infections. METHODS We designed a 1∶1∶1 case-case-control study to identify risk factors for XDR A. baumannii bacteremia in Singapore and Thailand. Case group 1 was defined as having infection due to XDR A. baumannii, and case group 2 was defined as having infection due to non-XDR A. baumannii. The control group comprised patients with blood cultures obtained to determine possible infection. RESULTS There were 93 patients in each group. Pitt bacteremia score (adjusted odds ratio [aOR], 2.570 [95% confidence interval (CI), 1.528-4.322]), central venous catheters (CVCs; aOR, 12.644 [95% CI, 2.143-74.620]), use of carbapenems (aOR, 54.391 [95% CI, 3.869-764.674]), and piperacillin-tazobactam (aOR, 55.035 [95% CI, 4.803-630.613]) were independently associated with XDR A. baumannii bacteremia. In case group 2, Pitt bacteremia score (aOR, 1.667 [95% CI, 1.265-2.196]) and third-generation cephalosporins (aOR, 2.965 [95% CI, 1.224-7.182]) were independently associated with non-XDR A. baumannii bacteremia. Concurrent infections (aOR, 3.527 [95% CI, 1.479-8.411]), cancer (aOR, 3.172 [95% CI, 1.135-8.865]), and respiratory source (aOR, 2.690 [95% CI, 1.160-6.239]) were associated with an increased risk of 30-day mortality. Survivors received more active empirical therapy (16.7% vs 9.6%; P = .157), had fewer cases of XDR bacteremia (45.8% vs 52.6%; P = .452), and received higher median definitive polymyxin B doses (840,000 units vs 700,000 units; P = .339) CONCLUSIONS: Use of CVC and broad spectrum antibiotics were unique risk factors of XDR A. baumannii bacteremia. Effective antimicrobial stewardship together with use of a CVC bundle may reduce the incidence of these infections. Risk factors of acquisition and mortality may help identify patients for early initiation of polymyxin B therapy.

Implementation hurdles of an interactive, integrated, point-of-care computerised decision support system for hospital antibiotic prescription

Antimicrobial stewardship is used to combat antimicrobial resistance. In Singapore, a tertiary hospital has integrated a computerised decision support system, called Antibiotic Resistance Utilisation and Surveillance-Control (ARUSC), into the electronic inpatient prescribing system. ARUSC is launched either by the physician to seek guidance for an infectious disease condition or via auto-trigger when restricted antibiotics are prescribed. This paper describes the implementation of ARUSC over three phases from 1 May 2011 to 30 April 2013, compared factors between ARUSC launches via auto-trigger and for guidance, examined factors associated with acceptance of ARUSC recommendations, and assessed user acceptability. During the study period, a monthly average of 9072 antibiotic prescriptions was made, of which 2370 (26.1%) involved ARUSC launches. Launches via auto-trigger comprised 48.1% of ARUSC launches. In phase 1, 23% of ARUSC launches were completed. This rose to 38% in phase 2, then 87% in phase 3, as escapes from the ARUSC programme were sequentially disabled. Amongst completed launches for guidance, 89% of ARUSC recommendations were accepted versus 40% amongst completed launches via auto-trigger. Amongst ARUSC launches for guidance, being from a medical department [adjusted odds ratio (aOR)=1.20, 95% confidence interval (CI) 1.04-1.37] and ARUSC launch during on-call (aOR=1.81, 95% CI 1.61-2.05) were independently associated with acceptance of ARUSC recommendations. Junior physicians found ARUSC useful. Senior physicians found ARUSC reliable but admitted to having preferences for antibiotics that may conflict with ARUSC. Hospital-wide implementation of ARUSC encountered hurdles from physicians. With modifications, the completion rate improved.

Continuation of High-Dose Vancomycin despite Nephrotoxicity

In agreement with several other studies ([3][1], [4][2]), Bosso et al. reported that in treating serious methicillin-resistant Staphylococcus aureus (MRSA) infections, vancomycin trough concentrations of >15 mg/liter were independently associated with nephrotoxicity ([1][3]). However, how best to

An interactive, point-of-care, computerised antibiotic prescription decision support system and quality of antibiotic prescription in the management of complicated urinary tract infection ☆

Antimicrobial resistance is a major public health problem. Antimicrobial stewardship programmes (ASPs) have been shown to contain or reduce resistance without adversely impacting the quality of care [1]. Computerised antibiotic prescription systems have been used and can circumvent the limitations of lack of manpower [2]. At Tan Tock Seng Hospital, a 1400-bed university teaching hospital in Singapore, there was an average of 200 new antibiotic orders a day, and 37.8% of carbapenem prescriptions were inappropriate and represented an opportunity for antimicrobial stewardship [3]. In this study, the safety and efficacy of an in-house computerised decision support system in the treatment of complicated urinary tract infections (cUTI) was evaluated. A computer program based on the hospital’s guidelines was designed, which is integrated into a doctor’s antibiotic ordering and interfaced with laboratory and pharmacy information system called ARUS-C (Antibiotic Resistance Utilization and Surveillance–Control). ARUS-C is accessed through the electronic inpatient medication record. It provides recommendations for empirical, definitive and prophylactic antibiotic use. When positive microbiology results are unavailable, ARUS-C provides antibiotic recommendations depending on available clinical laboratory information and other data manually entered by the doctor. Specific infectious diseases and use of active empirical antibiotics are considered when determining the duration of antibiotics. All antibiotic recommendations are adjusted for renal function, and drug allergies are accounted for. ARUS-C also provides alerts and clues that will help decide on diagnosis and management. The doctor preserves autonomy and the whole process takes 20 s. ARUS-C can be a passive guidance tool that is initiated by the doctor or can provide active guidance by mandating compulsory guidance when ‘restricted antibiotics’ are ordered. Any override of ARUS-C recommendations can be routed to the ASP team to allow timely audit and feedback. Prior to implementation, briefing and orientation were conducted and a user manual was made available. A retrospective study was performed from 31 August 2009 to 3 January 2010 to evaluate the effect of passive guidance of

Clinical impact of non-antibiotic recommendations by a multi-disciplinary antimicrobial stewardship team

INTRODUCTION The multi-disciplinary antimicrobial stewardship team at the study hospital conducts prospective review and feedback on all inpatient orders of piperacillin-tazobactam and carbapenems. In addition, the team provides non-antibiotic recommendations (i.e. additional investigations and infectious disease reviews). This study aimed to describe the impact of these recommendations on patient outcomes. METHODS Patients on carbapenem and piperacillin-tazobactam who received at least one non-antibiotic recommendation between January 2012 and August 2014 were included in this study. Acceptance and rejection of non-antibiotic recommendations by the managing physician were compared. The primary outcome was 30-d mortality. RESULTS Non-antibiotic recommendations were made in 166 patients. There were no differences in baseline characteristics between patients for whom recommendations were accepted and patients for whom recommendations were rejected. Thirty-day mortality (18.0% vs. 34.5%, P = 0.02) was significantly lower in patients who had at least one non-antibiotic recommendation accepted. Multi-variate analysis found that Charlsons comorbidity score [odds ratio (OR) 1.20, 95% confidence interval (CI) 1.03-1.42, P = 0.03], APACHE II score (OR 1.10, 95% CI 1.01-1.19, P < 0.01), hepatobiliary source of infection (OR 10.19, 95% CI 1.44-72.13, P = 0.02) and acceptance of at least one non-antibiotic recommendation (OR 0.39, 95% CI 0.17-0.88, P = 0.02) were independently associated with 30-d mortality. CONCLUSIONS During prospective review and feedback of piperacillin-tazobactam and carbapenems, acceptance of non-antibiotic recommendations was found to be associated with a reduction in 30-d mortality.

Optimization of vancomycin dosing in very low-birth-weight preterm neonates.

OBJECTIVE To compare vancomycin serum trough concentrations and 24-hour area under the serum concentration-versus-time curve (AUC24) among very low-birth-weight (VLBW) premature infants before and after implementation of an institution-wide increase in neonatal vancomycin dosing. STUDY DESIGN We performed a retrospective analysis of vancomycin concentrations among preterm VLBW neonates before (2007-2010) and after (2010-2013) implementation of a new vancomycin dosing protocol consisting of increased vancomycin daily dose and frequency of administration. RESULTS Neonates weighing < 1,500 g and receiving the new vancomycin dosing regimen had lower rates of undetectable trough concentrations (24 vs. 50%, p = 0.04), higher median trough concentrations (10.8 vs. 5.9 µg/mL, p = 0.003), a higher proportion of goal trough concentrations of 10 to 20 µg/mL (35 vs. 4%, p = 0.005), and a significantly higher vancomycin AUC24 (438 vs. 320 mg·h/L, p = 0.004) compared with historical controls. CONCLUSION Increasing the vancomycin daily dose and dosing frequency led to an increase in vancomycin trough concentrations and AUC24, and a decrease in the proportion of undetectable (< 5.0 µg/mL) troughs, without an increase in toxicity among VLBW premature neonates.