IDWeek 2012: Antimicrobial Stewardship in Special Populations and Circumstances

By Stan Deresinski MD is Clinical Professor of Medicine, Stanford University, Hospital Epidemiologist, Sequoia Hospital, Redwood City, CA, is Editor for Infectious Disease Alert.

Dr.Elizabeth Robilotti MD is an Infectious Disease Fellow at Stanford University and a member of the Stanford Hospital Antimicrobial Stewardship Program.

Emily Mui PharmD, is the Stanford program's Infectious Disease Pharmacist.

Dr. Robilotti and Emily Mui report no financial relationships in this field of study.

A remarkably large number of presentations at IDWeek held in San Diego October 17-21, 2012, dealt directly with issues of antimicrobial stewardship. These included aspects such as documenting the frequency of inappropriate antibiotic use in various settings together with descriptions of successful interventions. Here we have extracted information of interest from published abstracts from the meeting that deal with special populations and clinical circumstances.

Critical Care Patients

A prospective analysis of 67 consecutive patients in a Canadian community hospital Medical-Surgical ICU who were prescribed antibiotics found that recommendations made by the stewardship team were accepted in 72 of 78 (92.3%) of instances (#709). The two most frequent suggestions were further testing (31.9% of recommendations) and antibiotic discontinuation (25%). This stewardship program was associated with only a relatively modest decrease in antimicrobial costs relative to that seen in an earlier patient cohort, but an approximately 1.5 day reduction in estimated median time to discharge.

A retrospective review of records of 123 medical ICU patients found evidence of inappropriate antimicrobial therapy in 5 of 12 (42%) with community acquired pneumonia, 26 of 62 (45%) with healthcare associated pneumonia, 8 of 49 (42%) with bacteremia, 11 of 16 (69%) with intra-abdominal infection, 5 of 5 (100%) with "supposed CNS infections," 5 of 10 (50%) with skin and skin structure infections, 7 of 18 (38%) with urinary tract infection, and 2 of 2 (100%) with sepsis of unknown etiology (#702). Prolonging the duration of antimicrobial administration beyond national guideline recommendations accounted for approximately 80% of inappropriate use.

An adjudication committee evaluated the appropriateness of diagnosis and of antibiotic therapy of 232 episodes in 203 patients with a diagnosis of ventilator-associated pneumonia (VAP) on the day of diagnosis (day 1) and on days 3 and 7 in six adult ICUs (#766). On day 1, 94 (40.5%) cases were adjudicated to truly be VAP while by day 3 that number had declined to only 73 (31.5%). Antibiotics were continued on day 3 in 82 (51.9%) despite no evidence of other infection. The apparent misdiagnoses resulted in 1193 excess antibiotic days. In addition, antibiotics were continued for >8 days in 36 (49.3%) with a legitimate diagnosis of VAP, adding 374 excess antibiotic days. Factors associated with inappropriate antibiotic prolongation included pressors on day 1(OR 2.26 95% CI 1.27, 9.77) and sputum culture on day 3 (OR 3.53 95% CI 1.27, 9.77).

A retrospective review of 426 episodes of VAP in 273 patients in whom a BAL specimen had been obtained examined the value of the Gram stain of BAL specimens in directing the need for MRSA coverage (#751). No Gram positive cocci were seen in 302 of the 426 samples. The negative predictive value for recovery in culture of Gram positive coccal organisms was 96%. MRSA was recovered from only 4 of the 302 (1.3%), for a negative predictive value of 98.6%. Linezolid had been initiated in 142 of the 302 episodes and was subsequently discontinued in 76 of these. In comparing the linezolid discontinuation group with those in whom it was continued, the authors found a significantly lower mortality rate (10.8% vs. 36.4%, p=0.037), fewer adverse events (9.2% vs 24.2%, p=0.022) and a mean cost savings of $528.82 per episode. Length of stay did not differ significantly between the two groups.

Cancer Patients

Tverdek and colleagues described their experience with antimicrobial stewardship in cancer patients at the MD Anderson Cancer Center with targeting meropenem, vancomycin, linezolid, daptomycin and tigecycline (#707). The responsible clinician was notified via email on day 5 of therapy to either discontinue the target drug(s) or to submit an online form providing a rationale for continuation. The intervention resulted in significant decrease in Defined Daily Dose/1000 patient days for all targeted antibiotics with the exception of linezolid. There were significant acquisition cost savings for meropenem, vancomycin and tigecycline, but an estimated increase for linezolid and daptomycin. The increase of linezolid use was the result of a shift from vancomycin in patients with leukemia. The incidence of nosocomial infections with multi-drug resistant Pseudomonas aeruginosa, MRSA, VRE and Clostridium difficile decreased post-intervention.

Rehabilitation Inpatients

Lewis and colleagues evaluated the effect of implementation of guidelines aimed at reducing antibiotic administration for asymptomatic bacteriuria by a team of physicians, nurses and pharmacists (#710). Relative to a baseline period with no intervention and a period with a pharmacist-only-centered intervention, implementation by the team that included all 3 professional types was associated with a significant 36.8% decrease in days of antibiotic therapy per 1000 patient-days and a significant 49% decrease in the number of patients prescribed antibiotics.

Long-Term Care Patients

During a pre-intervention study period at a 77-bed long-term acute care facility, the costs attributable to unnecessary antimicrobial use constituted 41% of the total expenditure on antimicrobials (#705). Implementation of stewardship policies and guidelines was followed by a 42% decrease in daptomycin use and a 58% decrease in tigecycline use with resultant significant cost savings.

Only 8 of 100 long-term care facility residents prescribed antibiotics for urinary tract infection met criteria for that diagnosis (#765). The stewardship team, consisting of a pharmacist and a physician, who performed weekly prospective audit and feedback, were able to make real-time recommendations in 40 patients. The recommendations were accepted in 9 (23%) of these and none experienced infection-related adverse events.

Ambulatory Care Patients

Analysis of data from a national representative sample of ambulatory visits by adults >18 years of age in 2007-9 in the National Ambulatory and National Hospital Ambulatory Medical Care Surveys found that antibiotics were prescribed during 101 million visits per year — amounting to 10% of all ambulatory visits (#714). Quinolones (25%), macrolides (20%) and aminopenicillins (12%) were most frequently prescribed, most often for respiratory (40%), skin (12%), and urinary tract (8%) infections. Antibiotics were prescribed for respiratory conditions for which antibiotics are rarely indicated, such as acute bronchitis, 24 million times and 80% the antibiotics were broad spectrum.

Jenkins and colleagues performed a cluster-randomized trial in 8 primary care clinics targeting both adults and children with upper respiratory infection, acute bronchitis, acute rhinosinusitis, pharyngitis, acute otitis media, urinary tract infection, skin and soft tissue infection, and community-acquired pneumonia (#731). The experimental clinics received clinical pathways and patient education materials while the control group received no intervention. The intervention was associated with a significant decrease in antibiotic prescriptions for acute respiratory infections and overall use of broad spectrum antibiotics.

Brown and colleagues searched PubMed and EMBASE to assess the association between antibiotic class and the risk of Clostridium difficile infection (CDI) in the community setting by meta-analysis (#723). Pooled odds ratios indicated that clindamycin (OR=16.80, 95%CI: 7.48-37.76), fluoroquinolones (OR=5.40, 95%CI: 4.02-7.26) and cephalosporins (OR=5.55, 95%CI: 1.85-16.66) had the largest effects, while macrolides (OR=2.61, 95%CI: 1.84-3.72), sulfonamides and trimethoprim (OR=1.76, 95%CI: 1.31-2.38) and penicillins (OR=2.52, 95%CI: 1.61-3.95) had lesser, but nevertheless statistically significant associations with CDI. There was, in contrast, no evidence of an effect of tetracyclines on CDI risk (OR=0.92, 95%CI: 0.61-1.40).

OPAT (CoPAT) Patients

The University of Utah Infectious Disease Division developed a protocol defining the OPAT process and performed a Gap analysis. Interventions centered on achieving swift communication among the primary team, discharge planners, home infusion companies, and outpatient ID providers. Important shared information included antibiotic dose, duration and monitoring; timely follow up; and a mechanism to catch near misses (#804). The program accomplished the following:

  • an OPAT Standard Operating Procedure documenting anticipated stop dates, appropriate monitoring, and follow-up;
  • an OPAT Note - a concise and easily identifiable electronic note documenting the diagnosis, antibiotic, dose, anticipated stop date, lab monitoring and frequency, follow up provider's name, and follow up date/ time;
  • an OPAT Masterlist designed to track all ID patients slated for OPAT patients being followed by ID, changing patient tracking from 0% (0/21) to 100% (20/20).

Each of the above components was demonstrated to result in substantial improvement in OPAT process, quality, and patient care.

Analysis of the results of mandatory Infectious Disease consultation for OPAT at Henry Ford Hospital found that it led to a change in diagnosis in 12 (21%) patients and a change in antibiotic therapy in 32 (55%) (#703). Therapy was de-escalated in 12 (21%), with switch from IV to PO routes of administration in 11 of these and antibiotic discontinuation in 10. The antimicrobial spectrum was broadened in 3 (5%) patients. Readmission within 30 days was frequent, occurring in 15 (26%) patients with 10 (17%) being readmitted for non-infectious processes, 3 (5%) for new infection, and 2(4%) for same infection.

Keller and colleagues analyzed the relationship of process errors and patient outcomes before (238 patients) and after (44 patients) introduction of a post-discharge Infectious Diseases Transition Service (IDTS) for OPAT patients at an academic medical center (#1365). The presence of the IDTS was associated with an increased receipt of laboratory results by the infectious diseases clinic (91% vs. 41%, p<0.0001), more follow-up appointments kept (77% vs. 61%, p=0.045), and fewer medication errors at discharge (2.1% vs. 18%, p=0.008). There was no difference after introduction of the IDTS in the incidence of adverse drug reactions or infection relapse, but there were non-significant trends in the incidence of catheter complications (11% vs. 4.5%, p=0.38) and readmissions or emergency department visits within 60 days of discharge (46% vs. 36%, p=0.25).

Bittner and colleagues evaluated the success of self-administered OPAT in 73 veterans with bone and joint infections and found that treatment failure occurred in 47 (64.4%) (#752). A significantly increased risk of failure was associated with increased travel distance, osteomyelitis, and isolates other than S. aureus from surface culture, while a reduced risk of failure was associated with prosthetic joint infection and MSSA alone from bone/fluid culture. Nineteen (26%) patients were non-adherent and 17 of these experience treatment failure.

Donnelley and colleagues prospectively examined the effect of referral for ASP evaluation of 98 OPAT patients at the time of hospital discharge (#738). They made 199 interventions in 90 of the 98 (92%) patients; 7 of the remaining 8 had Infectious Disease service consultations. Complications in OPAT patients involved IV access (4.1%), adverse drug events (8.2%), and re-hospitalization or emergency care (8%). Overall, 90% of patients completed their OPAT course as planned on discharge. Regimen complexity was associated with reported patient dissatisfaction with OPAT.

A matched control study with propensity matching examined antimicrobial-related adverse events (AE) and healthcare utilization in patients receiving community-based parenteral anti-infective therapy (CoPAT) with either daptomycin or vancomycin. The mean duration of OPAT was 19 days (#794). The incidences of antimicrobial AEs were 36.9 and 42.4 per 1000 CoPAT days, that of vascular access complications were 4.8 and 4.2 per 1000 CoPAT days, vascular access interventions were 4.4 and 3.3 and the incidences of emergency department visits were 7.9 and 8.6 per 1000 CoPAT days in daptomycin and vancomycin recipients, respectively — none of these comparisons were statistically significant. In contrast, daptomycin use was associated with significantly fewer antimicrobial interventions (5.6 vs. 27.1 per 1000 CoPAT days; RR 0.21, 95% CI 0.12-0.35; P <0.001) and telephone encounters (242.7 vs. 273.7 per 1000 CoPAT days; RR 0.88, 95% CI 0.81-0.97; P = 0.01).

Investigators examined the effects of non-adherence with laboratory monitoring recommendations in 408 patients receiving CoPAT - 37% in skilled nursing facilities or long term acute care centers, and 4% in other settings (#801). There were 88 readmissions/deaths (22%) while on CoPAT. There was complete or partial adherence to test recommendations in 294 (72%) patients. Lack of adherence to lab monitoring recommendations was independently associated with readmission/death while on CoPAT (adjusted odds ratio (OR) 2.24; 95% CI 1.24 to 4.06).

Review of 92 OPAT patients with either a midline (26%) or PICC (74%) found that that the overall incidences of catheter placement-related complications, 30-day readmission for any cause, and 30-day readmission for catheter-related reasons were 8%, 30.4% and 6.5%, respectively (#803). Complications related to line placement were seen in 4.2% of PICC line patients and 8.8% of midline patients (p=0.46) while line-related readmissions were observed in 5.9% vs 8.3% (PICC vs Midline; p=0.82), respectively. In multivariate analysis, the rate of complications or readmission was similar among patients with PICC lines and Midlines.

An analysis of complications in 337 patients received 365 OPAT courses found that the most frequently encountered complication was an adverse reaction to antibiotics, which occurred in 94 (26%) OPAT courses (#808). This was followed by a problem with vascular access in 52 (14%) and a failed OPAT plan in 20 (6%); there were 2 (1%) deaths. Experiencing a first OPAT complication was significantly associated with having no primary care provider (p=0.008), a comorbid psychiatric disease (p=0.05), and receipt of IV vancomycin (p=<0.001).