Risk adjustment is needed to fairly compare central-line–associated bloodstream infection (CLABSI) rates between hospitals. Until 2017, the Centers for Disease Control and Prevention (CDC) methodology adjusted CLABSI rates only by type of intensive care unit (ICU). The 2017 CDC models also adjust for hospital size and medical school affiliation. We hypothesized that risk adjustment would be improved by including patient demographics and comorbidities from electronically available hospital discharge codes.

Using a cohort design across 22 hospitals, we analyzed data from ICU patients admitted between January 2012 and December 2013. Demographics and International Classification of Diseases, Ninth Edition, Clinical Modification (ICD-9-CM) discharge codes were obtained for each patient, and CLABSIs were identified by trained infection preventionists. Models adjusting only for ICU type and for ICU type plus patient case mix were built and compared using discrimination and standardized infection ratio (SIR). Hospitals were ranked by SIR for each model to examine and compare the changes in rank.

Overall, 85,849 ICU patients were analyzed and 162 (0.2%) developed CLABSI. The significant variables added to the ICU model were coagulopathy, paralysis, renal failure, malnutrition, and age. The C statistics were 0.55 (95% CI, 0.51–0.59) for the ICU-type model and 0.64 (95% CI, 0.60–0.69) for the ICU-type plus patient case-mix model. When the hospitals were ranked by adjusted SIRs, 10 hospitals (45%) changed rank when comorbidity was added to the ICU-type model.

Our risk-adjustment model for CLABSI using electronically available comorbidities demonstrated better discrimination than did the CDC model. The CDC should strongly consider comorbidity-based risk adjustment to more accurately compare CLABSI rates across hospitals.

Infect Control Hosp Epidemiol 2017;38:1019–1024

To describe the investigation and control of a cluster of Serratia marcescens bacteremia in a 505-bed tertiary-care center.

Cluster cases were defined as all patients with S. marcescens bacteremia between March 2 and April 7, 2014, who were found to have identical or related blood isolates determined by molecular typing with pulsed-field gel electrophoresis. Cases were compared using bivariate analysis with controls admitted at the same time and to the same service as the cases, in a 4:1 ratio.

In total, 6 patients developed S. marcescens bacteremia within 48 hours after admission within the above period. Of these, 5 patients had identical Serratia isolates determined by molecular typing, and were included in a case-control study. Exposure to the post-anesthesia care unit was a risk factor identified in bivariate analysis. Evidence of tampered opioid-containing syringes on several hospital units was discovered soon after the initial cluster case presented, and a full narcotic diversion investigation was conducted. A nurse working in the post-anesthesia care unit was identified as the employee responsible for the drug diversion and was epidemiologically linked to all 5 patients in the cluster. No further cases were identified once the implicated employee’s job was terminated.

Illicit drug use by healthcare workers remains an important mechanism for the development of bloodstream infections in hospitalized patients. Active mechanisms and systems should remain in place to prevent, detect, and control narcotic drug diversions and associated patient harm in the healthcare setting.

Infect Control Hosp Epidemiol 2017;38:1027–1031

To evaluate the impact of early infectious diseases (ID) antimicrobial stewardship (AMS) intervention on inpatient sepsis antibiotic management.

Interventional, nonrandomized, controlled study.

Tertiary-care referral hospital, Sydney, Australia.

Consecutive, adult, non–intensive care unit (non-ICU) inpatients triggering an institutional clinical sepsis pathway from May to August 2015.

All patients reviewed by an ID Fellow within 24 hours of sepsis pathway trigger underwent case review and clinic file documentation of recommendations. Those not reviewed by an ID Fellow were considered controls and received standard sepsis pathway care. The primary outcome was antibiotic appropriateness 48 hours after sepsis trigger.

In total, 164 patients triggered the sepsis pathway: 6 patients were excluded (previous sepsis trigger); 158 patients were eligible; 106 had ID intervention; and 52 were control cases. Of these 158 patients, 91 (58%) had sepsis, and 15 of these 158 (9.5%) had severe sepsis. Initial antibiotic appropriateness, assessable in 152 of 158 patients, was appropriate in 80 (53%) of these 152 patients and inappropriate in 72 (47%) of these patients. In the intervention arm, 93% of ID Fellow recommendations were followed or partially followed, including 53% of cases in which antibiotics were de-escalated. ID Fellow intervention improved antibiotic appropriateness at 48 hours by 24% (adjusted risk ratio, 1.24; 95% confidence interval, 1.04–1.47; P=.035). The appropriateness agreement among 3 blinded ID staff opinions was 95%. Differences in intervention and control group mortality (13% vs 17%) and median length of stay (13 vs 17.5 days) were not statistically significant.

Sepsis overdiagnosis and delayed antibiotic optimization may reduce sepsis pathway effectiveness. Early ID AMS improved antibiotic management of non-ICU inpatients with suspected sepsis, predominantly by de-escalation. Further studies are needed to evaluate clinical outcomes.

Infect Control Hosp Epidemiol 2017;38:1032–1038

To evaluate interventions to reduce avoidable antibiotic use on pediatric oncology and hematopoietic stem cell transplantation (HSCT) services.

Interrupted time series.

Academic pediatric hospital with separate oncology and HSCT services.

Children admitted to the services during baseline (October 2011–August 2013) and 2 intervention periods, September 2013–June 2015 and July 2015–June 2016, including 1,525 oncology hospitalizations and 301 HSCT hospitalizations.

In phase 1, we completed an update of the institutional febrile neutropenia (FN) guideline for the pediatric oncology service, recommending first-line β-lactam monotherapy rather than routine use of 2 gram-negative agents. Phase 2 included updating the HSCT service FN guideline and engagement with a new pediatric antimicrobial stewardship program. The use of target antibiotics (tobramycin and ciprofloxacin) was measured in days of therapy per 1,000 patient days collected from administrative data. Intervention effects were evaluated using interrupted time series with segmented regression.

Phase 1 had mixed effects–long-term reduction in tobramycin use (97% below projected at 18 months) but rebound with increasing slope in ciprofloxacin use (+18% per month). Following phase 2, tobramycin and ciprofloxacin use on the oncology service were both 99% below projected levels at 12 months. On the HSCT service, tobramycin use was 99% below the projected level and ciprofloxacin use was 96% below the projected level at 12 months.

Locally adapted guidelines can facilitate practice changes in oncology and HSCT settings. More comprehensive and ongoing interventions, including follow-up education, feedback, and engagement of companion services may be needed to sustain changes.

Infect Control Hosp Epidemiol 2017;38:1039–1047

Survey results suggest that prolonged administration of prophylactic antibiotics is common after mastectomy with reconstruction. We determined utilization, predictors, and outcomes of postdischarge prophylactic antibiotics after mastectomy with or without immediate breast reconstruction.

Retrospective cohort.

Commercially insured women aged 18–64 years coded for mastectomy from January 2004 to December 2011 were included in the study. Women with a preexisting wound complication or septicemia were excluded.

Predictors of prophylactic antibiotics within 5 days after discharge were identified in women with 1 year of prior insurance enrollment; relative risks (RR) were calculated using generalized estimating equations.

Overall, 12,501 mastectomy procedures were identified; immediate reconstruction was performed in 7,912 of these procedures (63.3%). Postdischarge prophylactic antibiotics were used in 4,439 procedures (56.1%) with immediate reconstruction and 1,053 procedures (22.9%) without immediate reconstruction (P<.001). The antibiotics most commonly prescribed were cephalosporins (75.1%) and fluoroquinolones (11.1%). Independent predictors of postdischarge antibiotics were implant reconstruction (RR, 2.41; 95% confidence interval [CI], 2.23–2.60), autologous reconstruction (RR, 2.17; 95% CI, 1.93–2.45), autologous reconstruction plus implant (RR, 2.11; 95% CI, 1.92–2.31), hypertension (RR, 1.05; 95% CI, 1.00–1.10), tobacco use (RR, 1.07; 95% CI, 1.01–1.14), surgery at an academic hospital (RR, 1.14; 95% CI, 1.07–1.21), and receipt of home health care (RR, 1.11; 95% CI, 1.04–1.18). Postdischarge prophylactic antibiotics were not associated with SSI after mastectomy with or without immediate reconstruction (both P>.05).

Prophylactic postdischarge antibiotics are commonly prescribed after mastectomy; immediate reconstruction is the strongest predictor. Stewardship efforts in this population to limit continuation of prophylactic antibiotics after discharge are needed to limit antimicrobial resistance.

Infect Control Hosp Epidemiol 2017;38:1048–1054

While a direct relation between hospital construction and concomitant infection rates has been clearly established, few data are available regarding the environmental decontamination effects of renovation in which surfaces are replaced and regarding subsequent infection incidence.

Retrospective clinical study with vancomycin-resistant Enterococcus (VRE) molecular strain typing and environmental cultures.

A regional referral center for acute leukemia and hematopoietic stem-cell transplantation.

Overall, 536 consecutive hospital admissions for newly diagnosed acute leukemia or a first autologous or allogeneic stem-cell transplantation were reviewed.

During 2009–2010, our unit underwent complete remodeling including replacement of all surfaces. We assessed the effects of this construction on the incidence of hospital-acquired VRE colonization before, during, and after the renovation.

We observed a sharp decrease in VRE colonization rates (hazard ratio, <0.23; 95% confidence interval, 0.18–0.44; P<.0001) during the first year after the renovation, with a return to near baseline rates thereafter. The known risk factors for VRE colonization appeared to be stable over the study interval. Environmental cultures outside of patient rooms revealed several contaminated areas that are commonly touched by unit personnel. Multilocus sequence typing of VRE isolates that were cryopreserved over the study interval showed that dominant strains prior to construction disappeared and were replaced by other strains after the renovation.

Unit reconstruction interrupted endemic transmission of VRE, which resumed with novel strains upon reopening. Contamination of environmental surfaces and shared equipment may play an important role in endemic transmission of VRE.

Infect Control Hosp Epidemiol 2017;38:1055–1061

To compare different techniques of endoscope sampling to assess residual bacterial contamination.

Diagnostic study.

The endoscopy unit of an 1,100-bed university hospital performing ~13,000 endoscopic procedures annually.

In total, 4 sampling techniques, combining flushing fluid with or without a commercial endoscope brush, were compared in an endoscope model. Based on these results, sterile physiological saline flushing with or without PULL THRU brush was selected for evaluation on 40 flexible endoscopes by adenosine triphosphate (ATP) measurement and bacterial culture. Acceptance criteria from the French National guideline (<25 colony-forming units [CFU] per endoscope and absence of indicator microorganisms) were used as part of the evaluation.

On biofilm-coated PTFE tubes, physiological saline in combination with a PULL THRU brush generated higher mean ATP values (2,579 relative light units [RLU]) compared with saline alone (1,436 RLU; P=.047). In the endoscope samples, culture yield using saline plus the PULL THRU (mean, 43 CFU; range, 1–400 CFU) was significantly higher than that of saline alone (mean, 17 CFU; range, 0–500 CFU; P<.001). In samples obtained using the saline+PULL THRU brush method, ATP values of samples classified as unacceptable were significantly higher than those of samples classified as acceptable (P=.001).

Physiological saline flushing combined with PULL THRU brush to sample endoscopes generated higher ATP values and increased the yield of microbial surveillance culture. Consequently, the acceptance rate of endoscopes based on a defined CFU limit was significantly lower when the saline+PULL THRU method was used instead of saline alone.

Infect Control Hosp Epidemiol 2017;38:1062–1069

Clostridium difficile infection (CDI) and asymptomatic carriage of toxigenic C. difficile are common in long-term care facilities (LTCFs). However, whether C. difficile is frequently acquired in the LTCF versus during acute-care admissions remains unknown.

To test the hypothesis that LTCF residents often acquire C. difficile colonization and infection in the LTCF

This 5-month cohort study was conducted to determine the incidence of acquisition of C. difficile colonization and infection in asymptomatic patients transferred from a Veterans Affairs hospital to an affiliated LTCF.

Rectal swabs were cultured for toxigenic C. difficile at the time of transfer to the LTCF and weekly for up to 6 weeks. We calculated the proportion of LTCF-onset CDI cases within 1 month of transfer that occurred in residents colonized on admission versus those with new acquisition in the LTCF.

Of 110 patients transferred to the LTCF, 12 (11%) were asymptomatically colonized with toxigenic C. difficile upon admission, and 4 of these 12 patients (33%) developed CDI within 1 month, including 3 recurrent and 1 initial CDI episode. Of 82 patients with negative cultures on transfer and at least 1 follow-up culture, 22 (27%) acquired toxigenic C. difficile colonization, and 4 developed CDI within 1 month, including 1 recurrent and 3 initial CDI episodes.

LTCF residents frequently acquired colonization with toxigenic C. difficile after transfer from the hospital, and 3 of 4 initial CDI cases with onset within 1 month of transfer occurred in residents who acquired colonization in the LTCF.

Infect Control Hosp Epidemiol 2017;38:1070–1076

To evaluate healthcare worker (HCW) risk of self-contamination when donning and doffing personal protective equipment (PPE) using fluorescence and MS2 bacteriophage.

Prospective pilot study.

Tertiary-care hospital.

A total of 36 HCWs were included in this study: 18 donned/doffed contact precaution (CP) PPE and 18 donned/doffed Ebola virus disease (EVD) PPE.

HCWs donned PPE according to standard protocols. Fluorescent liquid and MS2 bacteriophage were applied to HCWs. HCWs then doffed their PPE. After doffing, HCWs were scanned for fluorescence and swabbed for MS2. MS2 detection was performed using reverse transcriptase PCR. The donning and doffing processes were videotaped, and protocol deviations were recorded.

Overall, 27% of EVD PPE HCWs and 50% of CP PPE HCWs made ≥1 protocol deviation while donning, and 100% of EVD PPE HCWs and 67% of CP PPE HCWs made ≥1 protocol deviation while doffing (P=.02). The median number of doffing protocol deviations among EVD PPE HCWs was 4, versus 1 among CP PPE HCWs. Also, 15 EVD PPE protocol deviations were committed by doffing assistants and/or trained observers. Fluorescence was detected on 8 EVD PPE HCWs (44%) and 5 CP PPE HCWs (28%), most commonly on hands. MS2 was recovered from 2 EVD PPE HCWs (11%) and 3 CP PPE HCWs (17%).

Protocol deviations were common during both EVD and CP PPE doffing, and some deviations during EVD PPE doffing were committed by the HCW doffing assistant and/or the trained observer. Self-contamination was common. PPE donning/doffing are complex and deserve additional study.

Infect Control Hosp Epidemiol 2017;38:1077–1083

To evaluate the incidence of surgical-site infections (SSIs) in a cohort of liver transplant recipients and to assess risk factors predisposing patients to these infections.

Prospective observational cohort study.

Single transplant center in Canada.

Patients who underwent liver transplantation between February 2011 and August 2014.

Multivariate logistic regression was used to identify independent risk factors for SSIs in liver transplant patients.

We enrolled 250 liver transplant recipients. The recipients’ median age at the time of transplantation was 56 years (range, 19–70 years), and 166 patients (66.4%) were male. Moreover, 47 SSIs were documented in 43 patients (17.2%). Organ-space, superficial, and deep SSIs were noted in 29, 7, and 3 patients, respectively. In addition, 2 patients developed superficial and organ-space SSIs, and another 2 patients were found to have deep as well as organ-space infections. In total, we identified 33 organ-space SSIs (70.2%), 9 superficial SSIs (19.1%), and 5 deep SSIs (10.6%). Factors predictive of SSIs by multivariate analysis were duct-to-duct anastomosis (odds ratio [OR], 3.88; 95% CI, 1.85–8.13; P<.001) and dialysis (OR, 3.57; 95% CI, 1.02–12.50; P=.046). Of the 66 organisms isolated in both deep and organ-space SSIs, 55 (83%) were resistant to cefazolin.

Organ-space SSIs are a common complication after liver transplantation. Duct-to-duct anastomosis and dialysis were independent risk factors associated with SSIs. Appropriate perioperative prophylaxis targeting patients with duct-to-duct anastomosis and dialysis while simultaneously providing optimum coverage for the potential pathogens causing SSIs is warranted.

Infect Control Hosp Epidemiol 2017;38:1084–1090

To assess hospital surgical-site infection (SSI) identification and reporting following colon surgery and abdominal hysterectomy via a statewide external validation

Infection preventionists (IPs) from the California Department of Public Health (CDPH) performed on-site SSI validation for surgical procedures performed in hospitals that voluntarily participated. Validation involved chart review of SSI cases previously reported by hospitals plus review of patient records flagged for review by claims codes suggestive of SSI. We assessed the sensitivity of traditional surveillance and the added benefit of claims-based surveillance. We also evaluated the positive predictive value of claims-based surveillance (ie, workload efficiency).

Upon validation review, CDPH IPs identified 239 SSIs following colon surgery at 42 hospitals and 76 SSIs following abdominal hysterectomy at 34 hospitals. For colon surgery, traditional surveillance had a sensitivity of 50% (47% for deep incisional or organ/space [DI/OS] SSI), compared to 84% (88% for DI/OS SSI) for claims-based surveillance. For abdominal hysterectomy, traditional surveillance had a sensitivity of 68% (67% for DI/OS SSI) compared to 74% (78% for DI/OS SSI) for claims-based surveillance. Claims-based surveillance was also efficient, with 1 SSI identified for every 2 patients flagged for review who had undergone abdominal hysterectomy and for every 2.6 patients flagged for review who had undergone colon surgery. Overall, CDPH identified previously unreported SSIs in 74% of validation hospitals performing colon surgery and 35% of validation hospitals performing abdominal hysterectomy.

Claims-based surveillance is a standardized approach that hospitals can use to augment traditional surveillance methods and health departments can use for external validation.

Infect Control Hosp Epidemiol 2017;38:1091–1097

To explore the trends in infection control peer-reviewed journals, mainstream media, and blogs written by infection control professionals

Narrative and scoping reviews

Narrative and scoping reviews were performed to identify trending infection prevention and control topics from international journals, national news websites, newspapers, and so-called grey literature throughout 2015. Data were analyzed using word frequencies, and results are displayed in word clouds.

For 2015, our search identified 6 news websites with a total of published 116 articles, 71 articles from selected newspapers, and 214 publications from infection control websites. In total, 1,059 journal articles were initially identified; 98 articles were anonymous and thus were excluded, leaving 961 articles in the reviews. The terms ‘superbug’ and ‘antibiotics’ were most commonly used in titles of news websites and newspapers, whereas the terms ‘infection’ and ‘prevention’ were most commonly used in infection control websites or blogs. The word frequency differences among the 4 selected journals reflected their respective specialties.

In infection prevention and control, the integration of a range of mediums is necessary to best serve public interests. Whether the aim is advocacy, general health information dissemination, or warnings of imminent risk, health researchers have access to multiple forums with different strengths through which to influence public risk perceptions and responses.

Infect Control Hosp Epidemiol 2017;38:1098–1102