To develop an automated method for ventilator-associated condition (VAC) surveillance and to compare its accuracy and efficiency with manual VAC surveillance

The intensive care units (ICUs) of 4 hospitals

This study was conducted at Detroit Medical Center, a tertiary care center in metropolitan Detroit. A total of 128 ICU beds in 4 acute care hospitals were included during the study period from August to October 2013. The automated VAC algorithm was implemented and utilized for 1 month by all study hospitals. Simultaneous manual VAC surveillance was conducted by 2 infection preventionists and 1 infection control fellow who were blinded to each another’s findings and to the automated VAC algorithm results. The VACs identified by the 2 surveillance processes were compared.

During the study period, 110 patients from all the included hospitals were mechanically ventilated and were evaluated for VAC for a total of 992 mechanical ventilation days. The automated VAC algorithm identified 39 VACs with sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of 100%. In comparison, the combined efforts of the IPs and the infection control fellow detected 58.9% of VACs, with 59% sensitivity, 99% specificity, 91% PPV, and 92% NPV. Moreover, the automated VAC algorithm was extremely efficient, requiring only 1 minute to detect VACs over a 1-month period, compared to 60.7 minutes using manual surveillance.

The automated VAC algorithm is efficient and accurate and is ready to be used routinely for VAC surveillance. Furthermore, its implementation can optimize the sensitivity and specificity of VAC identification.

Infect. Control Hosp. Epidemiol. 2015;36(9):999–1003

Incidence of catheter-associated urinary tract infection (CAUTI) is a quality benchmark. To streamline conventional detection methods, an electronic surveillance system augmented with natural language processing (NLP), which gathers data recorded in clinical notes without manual review, was implemented for real-time surveillance.

To assess the utility of this algorithm for identifying indwelling urinary catheter days and CAUTI.

Large, urban tertiary care Veterans Affairs hospital.

All patients admitted to the acute care units and the intensive care unit from March 1, 2013, through November 30, 2013, were included. Standard surveillance, which includes electronic and manual data extraction, was compared with the NLP-augmented algorithm.

The NLP-augmented algorithm identified 27% more indwelling urinary catheter days in the acute care units and 28% fewer indwelling urinary catheter days in the intensive care unit. The algorithm flagged 24 CAUTI versus 20 CAUTI by standard surveillance methods; the CAUTI identified were overlapping but not the same. The overall positive predictive value was 54.2%, and overall sensitivity was 65% (90.9% in the acute care units but 33% in the intensive care unit). Dissimilarities in the operating characteristics of the algorithm between types of unit were due to differences in documentation practice. Development and implementation of the algorithm required substantial upfront effort of clinicians and programmers to determine current language patterns.

The NLP algorithm was most useful for identifying simple clinical variables. Algorithm operating characteristics were specific to local documentation practices. The algorithm did not perform as well as standard surveillance methods.

Infect. Control Hosp. Epidemiol. 2015;36(9):1004–1010

To evaluate seasonal variation in the rate of surgical site infections (SSI) following commonly performed surgical procedures.

Retrospective cohort study.

We analyzed 6 years (January 1, 2007, through December 31, 2012) of data from the 15 most commonly performed procedures in 20 hospitals in the Duke Infection Control Outreach Network. We defined summer as July through September. First, we performed 3 separate Poisson regression analyses (unadjusted, multivariable, and polynomial) to estimate prevalence rates and prevalence rate ratios of SSI following procedures performed in summer versus nonsummer months. Then, we stratified our results to obtain estimates based on procedure type and organism type. Finally, we performed a sensitivity analysis to test the robustness of our findings.

We identified 4,543 SSI following 441,428 surgical procedures (overall prevalence rate, 1.03/100 procedures). The rate of SSI was significantly higher during the summer compared with the remainder of the year (1.11/100 procedures vs 1.00/100 procedures; prevalence rate ratio, 1.11 [95% CI, 1.04–1.19]; P=.002). Stratum-specific SSI calculations revealed higher SSI rates during the summer for both spinal (P=.03) and nonspinal (P=.004) procedures and revealed higher rates during the summer for SSI due to either gram-positive cocci (P=.006) or gram-negative bacilli (P=.004). Multivariable regression analysis and sensitivity analyses confirmed our findings.

The rate of SSI following commonly performed surgical procedures was higher during the summer compared with the remainder of the year. Summer SSI rates remained elevated after stratification by organism and spinal versus nonspinal surgery, and rates did not change after controlling for other known SSI risk factors.

Infect. Control Hosp. Epidemiol. 2015;36(9):1011–1016

To evaluate the quality of gastrointestinal endoscope reprocessing and discuss the advantages of microbiological surveillance testing of these endoscopes.

Retrospective analysis of the results of endoscope sampling performed from October 1, 2006, through December 31, 2014, in a gastrointestinal endoscopy unit of a teaching hospital equipped with 89 endoscopes and 3 automated endoscope reprocessors, with an endoscopy quality assurance program in place. The compliance rate was defined as the proportion of the results classified at target or alert levels according to the French guidelines. A multivariate analysis (logistic regression) was used to identify the parameters influencing compliance.

A total of 846 samples were taken. The overall compliance rate was 86% and differed significantly depending on the sampling context (scheduled or not scheduled), the type of endoscope, and the season. No other parameter was associated with compliance. A total of 118 samples carried indicator microorganisms such as Pseudomonas aeruginosa, Stenotrophomonas maltophilia, Enterobacteriaceae, and Candida sp.

The systematic use of an automated endoscope reprocessor does not provide totally satisfactory compliance. Microbiological surveillance is indispensable to monitor reprocessing, reinforce good practices (endoscopes, reprocessing units), and detect endoscopes requiring early technical maintenance.

Infect. Control Hosp. Epidemiol. 2015;36(9):1017–1023

Standard estimates of the impact of Clostridium difficile infections (CDI) on inpatient lengths of stay (LOS) may overstate inpatient care costs attributable to CDI. In this study, we used multistate modeling (MSM) of CDI timing to reduce bias in estimates of excess LOS.

A retrospective cohort study of all hospitalizations at any of 120 acute care facilities within the US Department of Veterans Affairs (VA) between 2005 and 2012 was conducted. We estimated the excess LOS attributable to CDI using an MSM to address time-dependent bias. Bootstrapping was used to generate 95% confidence intervals (CI). These estimates were compared to unadjusted differences in mean LOS for hospitalizations with and without CDI.

During the study period, there were 3.96 million hospitalizations and 43,540 CDIs. A comparison of unadjusted means suggested an excess LOS of 14.0 days (19.4 vs 5.4 days). In contrast, the MSM estimated an attributable LOS of only 2.27 days (95% CI, 2.14–2.40). The excess LOS for mild-to-moderate CDI was 0.75 days (95% CI, 0.59–0.89), and for severe CDI, it was 4.11 days (95% CI, 3.90–4.32). Substantial variation across the Veteran Integrated Services Networks (VISN) was observed.

CDI significantly contributes to LOS, but the magnitude of its estimated impact is smaller when methods are used that account for the time-varying nature of infection. The greatest impact on LOS occurred among patients with severe CDI. Significant geographic variability was observed. MSM is a useful tool for obtaining more accurate estimates of the inpatient care costs of CDI.

Infect. Control Hosp. Epidemiol. 2015;36(9):1024–1030

To determine the effect of interhospital patient sharing via transfers on the rate of Clostridium difficile infections in a hospital.

Retrospective cohort.

Using data from the Healthcare Cost and Utilization Project California State Inpatient Database, 2005–2011, we identified 2,752,639 transfers. We then constructed a series of networks detailing the connections formed by hospitals. We computed 2 measures of connectivity, indegree and weighted indegree, measuring the number of hospitals from which transfers into a hospital arrive, and the total number of incoming transfers, respectively. Next, we estimated a multivariate model of C. difficile infection cases using the log-transformed network measures as well as covariates for hospital fixed effects, log median length of stay, log fraction of patients aged 65 or older, and quarter and year indicators as predictors.

We found an increase of 1 in the log indegree was associated with a 4.8% increase in incidence of C. difficile infection (95% CI, 2.3%–7.4%) and an increase of 1 in log weighted indegree was associated with a 3.3% increase in C. difficile infection incidence (1.5%–5.2%). Moreover, including measures of connectivity in our models greatly improved their fit.

Our results suggest infection control is not under the exclusive control of a given hospital but is also influenced by the connections and number of connections that hospitals have with other hospitals.

Infect. Control Hosp. Epidemiol. 2015;36(9):1031–1037

To report on the prevalence and incidence of Clostridium difficile infection (CDI) from 2009 to 2013 among Veterans Healthcare Administration patients

A retrospective descriptive analysis of data extracted from a large electronic medical record (EMR) database

Data were acquired from VHA healthcare records from 2009 to 2013 that included outpatient clinical visits, long-term care, and hospitalized care as well as pharmacy and laboratory information.

In 2009, there were 10,207 CDI episodes, and in 2013, there were 12,143 CDI episodes, an increase of 19.0%. The overall CDI rate increased by 8.4% from 193 episodes per 100,000 patient years in 2009 to 209 episodes per 100,000 patient years in 2013. Of the CDI episodes identified in 2009, 58% were identified during a hospitalization, and 42% were identified in an outpatient setting. In 2013, 44% of the CDI episodes were identified in an outpatient setting.

This is one of the largest studies that has utilized timely EMR data to describe the current CDI epidemiology at the VHA. Despite an aging population with greater burden of comorbidity than the general US population, our data show that VHA CDI rates stabilized between 2011 and 2013 following increases likely attributable to the introduction of the more sensitive nucleic acid amplification tests (NAATs). The findings in this report will help establish an accurate benchmark against which both current and future VA CDI prevention initiatives can be measured.

Infect. Control Hosp. Epidemiol. 2015;36(9):1038–1045

To determine the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) nasal colonization in hemodialysis patients and to analyze the cost-effectiveness of our screening approach compared with an alternative strategy.

Screening study and cost-effectiveness analysis.

Analysis of twice-yearly MRSA prevalence studies conducted in the hemodialysis unit of a 950-bed tertiary care hospital from January 1, 2004, through December 31, 2013. For this purpose, nasal swab samples were cultured on MRSA screening agar (mannitol-oxacillin biplate).

There were 20 mass screenings during the 10-year study period. We identified 415 patients participating in at least 1 screening, with an average of 4.5 screenings per patient. Of 415 screened patients, 15 (3.6%) were found to be MRSA carriers. The first mass screening in 2004 yielded the highest percentage of MRSA (6/101 [6%]). Only 7 subsequent screenings revealed new MRSA carriers, whereas 4 screenings confirmed previously known carriers, and 8 remained negative. None of the carriers developed MRSA bacteremia during the study period. The total cost of our screening approach, that is, screening and isolation costs, was US $93,930. The total cost of an alternative strategy (ie, no mass screening administered) would be equivalent to costs of isolation of index cases and contact tracing was estimated to be US $5,382 (difference, US $88,548).

In an area of low MRSA endemicity (<5%), regular nasal screenings of a high-risk population yielded a low rate of MRSA carriers. Twice-yearly MRSA screening of dialysis patients is unlikely to be cost-effective if MRSA prevalence is low.

Infect. Control Hosp. Epidemiol. 2015;36(9):1046–1049

To estimate the frequency of methicillin-resistant Staphylococcus aureus (MRSA) transmission to gowns and gloves worn by healthcare workers (HCWs) interacting with nursing home residents to better inform infection prevention policies in this setting

Observational study

Participants were recruited from 13 community-based nursing homes in Maryland and Michigan

Residents and HCWs from these nursing homes

Residents were cultured for MRSA at the anterior nares and perianal or perineal skin. HCWs wore gowns and gloves during usual care activities. At the end of each activity, a research coordinator swabbed the HCW’s gown and gloves.

A total of 403 residents were enrolled; 113 were MRSA colonized. Glove contamination was higher than gown contamination (24% vs 14% of 954 interactions; P<.01). Transmission varied greatly by type of care from 0% to 24% for gowns and from 8% to 37% for gloves. We identified high-risk care activities: dressing, transferring, providing hygiene, changing linens, and toileting the resident (OR >1.0; P<.05). We also identified low-risk care activities: giving medications and performing glucose monitoring (OR<1.0; P<.05). Residents with chronic skin breakdown had significantly higher rates of gown and glove contamination.

MRSA transmission from MRSA-positive residents to HCW gown and gloves is substantial; high-contact activities of daily living confer the highest risk. These activities do not involve overt contact with body fluids, skin breakdown, or mucous membranes, which suggests the need to modify current standards of care involving the use of gowns and gloves in the nursing home setting.

Infect. Control Hosp. Epidemiol. 2015;36(9):1050–1057

To use whole genome sequencing to describe the likely origin of an outbreak of Pseudomonas aeruginosa in a neonatal unit.

Outbreak investigation.

The neonatal intensive care unit service of a major obstetric tertiary referral center.

Infants admitted to the neonatal unit who developed P. aeruginosa colonization or infection.

We undertook whole genome sequencing of P. aeruginosa strains isolated from colonized infants and from the neonatal unit environment.

Eighteen infants were colonized with P. aeruginosa. Isolates from 12 infants and 7 environmental samples were sequenced. All but one of the clinical isolates clustered in ST253 and no differences were detected between unmapped reads. The environmental isolates revealed a variety of sequence types, indicating a large diverse bioburden within the unit, which was subsequently confirmed via enterobacterial repetitive intergenic consensus–polymerase chain reaction typing of post-outbreak isolates. One environmental isolate, obtained from a sink in the unit, clustered within ST253 and differed from the outbreak strain by 9 single-nucleotide polymorphisms only. This information allowed us to focus infection control activities on this sink.

Whole genome sequencing can provide detailed information in a clinically relevant time frame to aid management of outbreaks in critical patient management areas. The superior discriminatory power of this method makes it a powerful tool in infection control.

Infect. Control Hosp. Epidemiol. 2015;36(9):1058–1064

To understand the professional and psychosocial factors that influence physician antibiotic prescribing habits in the inpatient setting.

We conducted semi-structured interviews with 30 inpatient physicians. Interviews consisted of open-ended questions and flexible probes based on participant responses. Interviews were audio recorded, transcribed, de-identified, and reviewed for accuracy and completeness. Data were analyzed using emergent thematic analysis.

Two teaching hospitals in Indianapolis, Indiana

A total of 30 inpatient physicians (10 physicians-in-training, 20 supervising staff) were enrolled in this study.

Participants recognized that antibiotics are overused, and many admitted to prescribing antibiotics even when the clinical evidence of infection was uncertain. Overprescription was largely driven by anxiety about missing an infection, whereas potential adverse effects of antibiotics did not strongly influence decision making. Participants did not routinely disclose potential adverse effects of antibiotics to inpatients. Physicians-in-training were strongly influenced by the antibiotic prescribing behavior of their supervising staff physicians. Participants sometimes questioned their colleagues’ antibiotic prescribing decisions, but they frequently avoided providing direct feedback or critique. These physicians cited obstacles of hierarchy, infrequent face-to-face encounters, and the awkwardness of these conversations.

A physician-based culture of prescribing antibiotics involves overusing antibiotics and not challenging the decisions of colleagues. The potential adverse effects of antibiotics did not strongly influence decision making in this sample. A better understanding of these factors could be leveraged in future efforts to improve antibiotic prescribing practices in the inpatient setting.

Infect. Control Hosp. Epidemiol. 2015;36(9):1065–1072

Estimates of the excess length of stay (LOS) attributable to healthcare-associated infections (HAIs) in which total LOS of patients with and without HAIs are biased because of failure to account for the timing of infection. Alternate methods that appropriately treat HAI as a time-varying exposure are multistate models and cohort studies, which match regarding the time of infection. We examined the magnitude of this time-dependent bias in published studies that compared different methodological approaches.

We conducted a systematic review of the published literature to identify studies that report attributable LOS estimates using both total LOS (time-fixed) methods and either multistate models or matching patients with and without HAIs using the timing of infection.

Of the 7 studies that compared time-fixed methods to multistate models, conventional methods resulted in estimates of the LOS to HAIs that were, on average, 9.4 days longer or 238% greater than those generated using multistate models. Of the 5 studies that compared time-fixed methods to matching on timing of infection, conventional methods resulted in estimates of the LOS to HAIs that were, on average, 12.6 days longer or 139% greater than those generated by matching on timing of infection.

Our results suggest that estimates of the attributable LOS due to HAIs depend heavily on the methods used to generate those estimates. Overestimation of this effect can lead to incorrect assumptions of the likely cost savings from HAI prevention measures.

Infect. Control Hosp. Epidemiol. 2015;36(9):1089–1094