Surgical site infection (SSI) surveillance represents a key method of nosocomial infection control programs worldwide. However, most SSI surveillance systems are considered to be poorly cost effective regarding human and economic resources required for data collection and patient follow up. This study aims to assess the efficacy of using hospital discharge databases (HDDs) as a routine surveillance system for detecting hip or knee arthroplasty–related infections (HKAIs).

A case-control study was conducted among patients hospitalized in the Centre region of France between 2008 and 2010. HKAI cases were extracted from the HDD with various algorithms based on the International Classification of Diseases, Tenth Revision, and procedure codes. The control subjects were patients with hip or knee arthroplasty (HKA) without infection selected at random from the HDD during the study period. The gold standard was medical chart review. Sensitivity (Se), specificity (Spe), positive predictive value (PPV), and negative predictive value (NPV) were calculated to evaluate the efficacy of the surveillance system.

Among 18,265 hospital stays for HKA, corresponding to 17,388 patients, medical reports were checked for 1,010 hospital stays (989 patients). We identified 530 cases in total (incidence rate, 1% [95% confidence interval (CI), 0.4%–1.6%), and 333 cases were detected by routine surveillance. As compared with 480 controls, Se was 98%, Spe was 71%, PPV was 63%, and NPV was 99%. Using a more specific case definition, based on a sample of 681 hospital stays, Se was 97%, Spe was 95%, PPV was 87%, and NPV was 98%.

This study demonstrates the potential of HDD as a tool for routine SSI surveillance after low-risk surgery, under conditions of having an appropriate algorithm for selecting infections.

Infect Control Hosp Epidemiol 2014;35(6):646–651

To explore the feasibility of identifying anterior cruciate ligament (ACL) allograft implantations and infections using claims.

Retrospective cohort study.

We identified ACL reconstructions using procedure codes at 6 health plans from 2000 to 2008. We then identified potential infections using claims-based indicators of infection, including diagnoses, procedures, antibiotic dispensings, specialty consultations, emergency department visits, and hospitalizations. Patients’ medical records were reviewed to determine graft type, validate infection status, and calculate sensitivity and positive predictive value (PPV) for indicators of ACL allografts and infections.

A total of 11,778 patients with codes for ACL reconstruction were identified. After chart review, PPV for ACL reconstruction was 96% (95% confidence interval [CI], 94%–97%). Of the confirmed ACL reconstructions, 39% (95% CI, 35%–42%) used allograft tissues. The deep infection rate after ACL reconstruction was 1.0% (95% CI, 0.7%–1.4%). The odds ratio of infection for allografts versus autografts was 0.41 (95% CI, 0.19–0.78). Sensitivity of individual claims-based indicators for deep infection after ACL reconstruction ranged from 0% to 75% and PPV from 0% to 100%. Claims-based infection indicators could be combined to enhance sensitivity or PPV but not both.

While claims data accurately identify ACL reconstructions, they poorly distinguish between allografts and autografts and identify infections with variable accuracy. Claims data could be useful to monitor infection trends after ACL reconstruction, with different algorithms optimized for different surveillance goals.

Infect Control Hosp Epidemiol 2014;35(6):652–659

Surgical site infection (SSI) is one of the most common healthcare-associated infections (HAIs). This study aims to assess factors associated with SSI after colorectal surgery in Japan, using a Japanese national database for HAIs.

A retrospective nationwide surveillance-based study.

Japanese healthcare facilities.

Data on colon and rectal surgeries performed from 2008 through 2010 were extracted from a national monitoring system for healthcare-associated infections, the Japan Nosocomial Infections Surveillance (JANIS). Factors associated with SSI after colon and rectal surgery were assessed using multivariate logistic regression.

The cumulative incidence of SSI for colon and rectal surgery was 15.0% (6,691 of 44,751) and 17.8% (3,230 of 18,187), respectively. Traditional risk factors included in the National Nosocomial Infections Surveillance (NNIS) modified risk index were significant in predicting SSI in the final model for both colon and rectal surgery. Among the additional variables routinely collected in JANIS were factors independently associated with the development of SSI, such as male sex (adjusted odds ratio [aOR], 1.20 [95% confidence interval (CI), 1.14–1.27]), ileostomy or colostomy placement (aOR, 1.13 [95% CI, 1.04–1.21]), emergency operation (aOR, 1.40 [95% CI, 1.29–1.52]), and multiple procedures (aOR, 1.22 [95% CI, 1.13–1.33]) for colon surgery as well as male sex (aOR, 1.43 [95% CI, 1.31–1.55]), ileostomy or colostomy placement (aOR, 1,63 [95% CI, 1.51–1.79]), and emergency operation (aOR, 1.43 [95% CI, 1.20–1.72]) for rectal surgery.

For colorectal operations, inclusion of additional variables routinely collected in JANIS can more accurately predict SSI risk than can the NNIS risk index alone.

Infect Control Hosp Epidemiol 2014;35(6):660–666

To evaluate the accuracy of real-time polymerase chain reaction (PCR) for Clostridium difficile–associated disease (CDAD) detection, after hospital CDAD rates significantly increased following real-time PCR initiation for CDAD diagnosis.

Hospital-wide surveillance study following examination of CDAD incidence density rates by interrupted time series design.

Large university-based hospital.

Hospitalized adult patients.

CDAD rates were compared before and after real-time PCR implementation in a university hospital and in the absence of physician and infection control practice changes. After real-time PCR introduction, all hospitalized adult patients were screened for C. difficile by testing a fecal specimen by real-time PCR, toxin enzyme-linked immunosorbent assay, and toxigenic culture.

CDAD hospital rates significantly increased after changing from cell culture cytotoxicity assay to a real-time PCR assay. One hundred ninety-nine hospitalized subjects were enrolled, and 101 fecal specimens were collected. C. difficile was detected in 18 subjects (18%), including 5 subjects (28%) with either definite or probable CDAD and 13 patients (72%) with asymptomatic C. difficile colonization.

The majority of healthcare-associated diarrhea is not attributable to CDAD, and the prevalence of asymptomatic C. difficile colonization exceeds CDAD rates in healthcare facilities. PCR detection of asymptomatic C. difficile colonization among patients with non-CDAD diarrhea may be contributing to rising CDAD rates and a significant number of CDAD false positives. PCR may be useful for CDAD screening, but further study is needed to guide interpretation of PCR detection of C. difficile and the value of confirmatory tests. A gold standard CDAD diagnostic assay is needed.

Infect Control Hosp Epidemiol 2014;35(6):667–673

To explore differences in the prevalence of healthcare-associated infections (HAIs) according to survey methodology.

Repeated point and period prevalence survey strategies.

University-affiliated primary and tertiary care center.

Analysis of data collected from 2006 to 2012 from annual HAI prevalence surveys using definitions proposed by the US Centers for Disease Control and Prevention. The study design allowed the analysis of the same data in the format of a point or a period prevalence survey.

Pooled point and period HAI prevalence was 7.46% and 9.84% (+32%), respectively. This additional 32% was mainly attributable to infections of the lower respiratory tract (2.42% vs 3.20% [+32%]) and the urinary tract (1.76% vs 2.62% [+49%]). Differences in surgical site infections (1.02% vs 1.20% [+19%]) and bloodstream infections (0.76% vs 0.86% [+13%]) were smaller. HAI prevalence for the point and period methodology in acute and long-term care were 7.47% versus 9.38 (+26%) and 8.37% versus 11.89% (+42%), respectively. Differences were stable over time. Focusing on the 4 major HAIs (respiratory tract, urinary tract, surgical site, and bloodstream infections) misses one-quarter of all HAIs.

More HAIs are identified by the period prevalence method, especially those of shorter duration (lower respiratory and urinary tract), which would make this method more suitable to be used in long-term care. Results of the 2 study methods cannot be benchmarked against each other.

Infect Control Hosp Epidemiol 2014;35(6):674–684

To develop and validate a methodology for electronic surveillance of catheter-associated urinary tract infections (CAUTIs).

Diagnostic accuracy study.

A 425-bed university hospital.

A total of 1,695 unique inpatient encounters from November 2009 through November 2010 with a high clinical suspicion of CAUTI.

An algorithm was developed to identify incident CAUTIs from electronic health records (EHRs) on the basis of the Centers for Disease Control and Prevention (CDC) surveillance definition. CAUTIs identified by electronic surveillance were compared with the reference standard of manual surveillance by infection preventionists. To determine diagnostic accuracy, we created 2 × 2 tables, one unadjusted and one adjusted for misclassification using chart review and case adjudication. Unadjusted and adjusted test statistics (percent agreement, sensitivity, specificity, positive predictive value [PPV], negative predictive value [NPV], and κ) were calculated.

Electronic surveillance identified 64 CAUTIs compared with manual surveillance, which identified 19 CAUTIs for 97% agreement, 79% sensitivity, 97% sensitivity, 23% PPV, 100% NPV, and κ of .33. Compared with the reference standard adjusted for misclassification, which identified 55 CAUTIs, electronic surveillance had 98% agreement, 80% sensitivity, 99% specificity, 69% PPV, 99% NPV, and κ of .71.

The electronic surveillance methodology had a high NPV and a low PPV compared with the reference standard, indicating a role of the electronic algorithm in screening data sets to exclude cases. However, the PPV markedly improved compared with the reference standard adjusted for misclassification, suggesting a future role in surveillance with improvements in EHRs.

Infect Control Hosp Epidemiol 2014;35(6):685–691

With growing demands to track and publicly report and compare infection rates, efforts to utilize automated surveillance systems are increasing. We developed and validated a simple algorithm for identifying patients with clinical methicillin-resistant Staphylococcus aureus (MRSA) infection using microbiologic and antimicrobial variables. We also estimated resource savings.

Patients who had a culture positive for MRSA at any of 5 acute care Veterans Affairs hospitals were eligible. Clinical infection was defined on the basis of manual chart review. The electronic algorithm defined clinical MRSA infection as a positive non-sterile-site culture with receipt of MRSA-active antibiotics during the 5 days prior to or after the culture.

In total, 246 unique non-sterile-site cultures were included, of which 168 represented infection. The sensitivity (43.4%–95.8%) and specificity (34.6%–84.6%) of the electronic algorithm varied depending on the combination of antimicrobials included. On multivariable analysis, predictors of algorithm failure were outpatient status (odds ratio, 0.23 [95% confidence interval, 0.10–0.56]) and respiratory culture (odds ratio, 0.29 [95% confidence interval, 0.13–0.65]). The median cost was $2.43 per chart given 4.6 minutes of review time per chart.

Our simple electronic algorithm for detecting clinical MRSA infections has excellent sensitivity and good specificity. Implementation of this electronic system may streamline and standardize surveillance and reporting efforts.

Infect Control Hosp Epidemiol 2014;35(6):692–698

Healthcare workers experience occupational risk of infection and may transmit infections to patients. Vaccination provides an efficient means of protecting workers and patients, but uptake may be low. We sought to identify factors influencing vaccine acceptance by healthcare workers in order to obtain insights leading to more effective vaccination programs in this population.

Systematic review and meta-analysis.

We searched Medline, Embase, and CINAHL databases to identify studies published up to May 2012. Factors influencing vaccination acceptance were devised a priori. Random-effects meta-analysis was performed to generate summary estimates of effect. Heterogeneity and publication bias were explored using statistical tools.

Thirty-seven studies evaluating a variety of vaccines (against influenza, pertussis, smallpox, anthrax, and hepatitis B) were included. Homogeneous effects on vaccine acceptance were identified with desire for self-protection (odds ratio [OR], 3.42 [95% confidence interval (CI), 2.42–4.82]) and desire to protect family and friends (OR, 3.28 [95% CI, 1.10–9.75]). Concern that vaccine transmits the illness it was meant to prevent decreased acceptance (OR, 0.42 [95% CI, 0.30–0.58]). Differences in physician and nurse acceptance of immunization were seen between Asian and non-Asian studies.

Consideration of self-protection (rather than absolute disease risk or protection of patients) appears the strongest and most consistent driver of healthcare workers’ decisions to accept vaccination, though other factors may also be impactful, and reasons for between-study divergence in effects is an important area for future research. This finding has important implications for the design of programs to enhance healthcare worker vaccine uptake.

Infect Control Hosp Epidemiol 2014;35(6):699–708

The mass media plays an important role in public health behavior.

The objective of the present study was to investigate the effect of mass media coverage of the H1N1 pandemic on the number of emergency department (ED) visits and hospital admission rates.

An ecological study of ED visits to 8 general Israeli hospitals due to influenza-like illness during the period June–October 2009 was performed. Data on the number of visits per day for children and adults and daily hospitalization rates were analyzed. Associations with the estimated value of H1N1-related publications and weekly reports from nationwide sentinel clinics were assessed. The analysis was performed in 2012–2013.

There were 55,070 ED visits due to influenza-like illness during the study period. The overall number of media reports was 1,812 (14.3% radio broadcasts, 9.8% television broadcasts, 27.5% newspaper articles, and 48.5% major website reports). The overall estimated value of advertising of publications was $16,399,000, excluding the Internet. While H1N1 incidence recorded by Israeli sentinel clinics showed no association with mass media publications, peaks of media reports were followed by an increase in the number of ED visits, usually with a delay of 3 days (P = .005). This association was noted in children (P < .001) but not in adults (P > .1), with a corresponding decrease in hospital admission rates. Publications’ framing had no association with ED visits.

During the 2009 H1N1 influenza outbreak in Israel, an increase in mass media coverage was associated with an increase in pediatric ED visits.

Infect Control Hosp Epidemiol 2014;35(6):709–716