The financial incentives for hospitals to improve care may be weaker if higher insurer payments for adverse conditions offset a portion of hospital costs. The purpose of this study was to simulate incentives for reducing hospital-acquired infections under various payment configurations by Medicare, Medicaid, and private payers.

Matched case-control study.

A large, urban hospital system with 1 community hospital and 2 tertiary-care hospitals.

All patients discharged in 2013 and 2014.

Using electronic hospital records, we identified hospital-acquired bloodstream infections (BSIs) and urinary tract infections (UTIs) with a validated algorithm. We assessed excess hospital costs, length of stay, and payments due to infection, and we compared them to those of uninfected patients matched by propensity for infection.

In most scenarios, hospitals recovered only a portion of excess HAI costs through increased payments. Patients with UTIs incurred incremental costs of $6,238 (P<.01), while payments increased $1,901 (P<.05) at public diagnosis-related group (DRG) rates. For BSIs, incremental costs were $15,367 (P<.01), while payments increased $7,895 (P<.01). If private payers reimbursed a 200% markup over Medicare DRG rates, hospitals recovered 55% of costs from BSI and UTI among private-pay patients and 54% for BSI and 33% for UTI, respectively, across all patients. Under per-diem payment for private patients with no markup, hospitals recovered 71% of excess costs of BSI and 88% for UTI. At 150% markup and per-diem payments, hospitals profited.

Hospital incentives for investing in patient safety vary by payer and payment configuration. Higher payments provide resources to improve patient safety, but current payment structures may also reduce the willingness of hospitals to invest in patient safety.

Infect Control Hosp Epidemiol 2018;39:509–515

While previous work showed that the Centers for Disease Control and Prevention toolkit for carbapenem-resistant Enterobacteriaceae (CRE) can reduce spread regionally, these interventions are costly, and decisions makers want to know whether and when economic benefits occur.

Economic analysis

Orange County, California

Using our Regional Healthcare Ecosystem Analyst (RHEA)-generated agent-based model of all inpatient healthcare facilities, we simulated the implementation of the CRE toolkit (active screening of interfacility transfers) in different ways and estimated their economic impacts under various circumstances.

Compared to routine control measures, screening generated cost savings by year 1 when hospitals implemented screening after identifying ≤20 CRE cases (saving $2,000–$9,000) and by year 7 if all hospitals implemented in a regional coordinated manner after 1 hospital identified a CRE case (hospital perspective). Cost savings was achieved only if hospitals independently screened after identifying 10 cases (year 1, third-party payer perspective). Cost savings was achieved by year 1 if hospitals independently screened after identifying 1 CRE case and by year 3 if all hospitals coordinated and screened after 1 hospital identified 1 case (societal perspective). After a few years, all strategies cost less and have positive health effects compared to routine control measures; most strategies generate a positive cost-benefit each year.

Active screening of interfacility transfers garnered cost savings in year 1 of implementation when hospitals acted independently and by year 3 if all hospitals collectively implemented the toolkit in a coordinated manner. Despite taking longer to manifest, coordinated regional control resulted in greater savings over time.

Infect Control Hosp Epidemiol 2018;39:516–524

To summarize the clinical burden (cumulative incidence, prevalence, case fatality rate and length of stay) and economic burden (healthcare cost) of healthcare-associated infections (HAIs) due to multidrug-resistant organisms (MDROs) among patients in intensive care units (ICUs) in Southeast Asia.

Systematic review.

We conducted a comprehensive literature search in PubMed, EMBASE, CINAHL, EconLit, and the Cochrane Library databases from their inception through September 30, 2016. Clinical and economic burdens and study quality were assessed for each included study.

In total, 41 studies met our inclusion criteria; together, 22,876 ICU patients from 7 Southeast Asian countries were included. The cumulative incidence of HAI caused by A. baumannii (AB) in Southeast Asia is substantially higher than has been reported in other regions, especially carbapenem-resistant AB (CRAB; 64.91%) and multidrug-resistant AB (MDR-AB) (58.51%). Evidence of a dose–response relationship between different degrees of drug resistance and excess mortality due to AB infections was observed. Adjusted odds ratios were 1.23 (95% confidence interval [CI], 0.51–3.00) for MDR-AB, 1.72 (95% CI, 0.77–3.80) for extensively drug-resistant AB (XDR-AB), and 1.82 (95% CI, 0.55–6.00) for pandrug-resistant AB (PDR-AB). There is, however, a paucity of published data on additional length of stay and costs attributable to MDROs.

This review highlights the challenges in addressing MDROs in Southeast Asia, where HAIs caused by MDR gram-negative bacteria are abundant and have a strong impact on society. With our findings, we hope to draw the attention of clinicians and policy makers to the problem of antibiotic resistance and to issue a call for action in the management of MDROs.

Infect Control Hosp Epidemiol 2018;39:525–533

To evaluate the impact of universal contact precautions (UCP) on rates of multidrug-resistant organisms (MDROs) in intensive care units (ICUs) over 9 years

Retrospective, nonrandomized observational study

An 800-bed adult academic medical center in New York City

All patients admitted to 6 ICUs, 3 of which instituted UCP in 2007

Using a comparative effectiveness approach, we studied the longitudinal impact of UCP on MDRO incidence density rates, including methicillin-resistant Staphylococcus aureus, vancomycin-resistant enterococci, and carbapenem-resistant Klebsiella pneumoniae. Data were extracted from a clinical research database for 2006–2014. Monthly MDRO rates were compared between the baseline period and the UCP period, utilizing time series analyses based on generalized linear models. The same models were also used to compare MDRO rates in the 3 UCP units to 3 ICUs without UCPs.

Overall, MDRO rates decreased over time, but there was no significant decrease in the trend (slope) during the UCP period compared to the baseline period for any of the 3 intervention units. Furthermore, there was no significant difference between UCP units (6.6% decrease in MDRO rates per year) and non-UCP units (6.0% decrease per year; P=.840).

The results of this 9-year study suggest that decreases in MDROs, including multidrug-resistant gram-negative bacilli, were more likely due to hospital-wide improvements in infection prevention during this period and that UCP had no detectable additional impact.

Infect Control Hosp Epidemiol 2018;39:534–540

To quantify the association between having a prior bed occupant or roommate with a positive blood, respiratory, urine, or wound culture and subsequent infection with the same organism.

Case-control study.

The study included 4 hospitals within an academically affiliated network in New York City, including a community hospital (221 beds), a pediatric acute-care hospital (283 beds), an adult tertiary-/quaternary-care hospital (647 beds), and a pediatric and adult tertiary-/quaternary-care hospital (914 beds).

All 761,426 inpatients discharged from 2006 to 2012 were eligible. Cases included all patients who developed a healthcare-associated infection (HAI) with Staphylococcus aureus, Acinetobacter baumannii, Streptococcus pneumoniae, Pseudomonas aeruginosa, Klebsiella pneumoniae, Enterococcus faecalis, or Enterococcus faecium. Controls were uninfected patients matched by fiscal quarter, hospital, and length of stay. For each bed occupied during the 3–5-day period prior to infection, microbiology results for assigned roommates and the patient who occupied the bed immediately prior to the case were collected. For controls, the day of infection of the matched case served as the reference point.

In total, 10,289 HAIs were identified. In a multivariable analysis controlling for both exposures and patient characteristics, the odds of cases having been exposed to a prior bed occupant with the same organism were 5.83 times that of controls (95% confidence interval [CI], 3.62–9.39), and the odds of cases having been exposed to a roommate with the same organism were 4.82 times that of controls (95% CI, 3.67–6.34).

Infected or colonized roommates and prior occupants do pose a risk, which may warrant enhanced terminal and intermittent cleaning measures.

Infect Control Hosp Epidemiol 2018;39:541–546

To examine the impact of urine culture testing on day 1 of admission on inpatient antibiotic use and hospital length of stay (LOS).

We performed a retrospective cohort study using a national dataset from 2009 to 2014.

The study used data from 230 hospitals in the United States.

Admissions for adults 18 years and older were included in this study. Hospitalizations were matched with coarsened exact matching by facility, patient age, gender, Medicare severity-diagnosis related group (MS-DRG), and 3 measures of disease severity.

A multilevel Poisson model and a multilevel linear regression model were used to determine the impact of an admission urine culture on inpatient antibiotic use and LOS.

Matching produced a cohort of 88,481 patients (n=41,070 with a culture on day 1, n=47,411 without a culture). A urine culture on admission led to an increase in days of inpatient antibiotic use (incidence rate ratio, 1.26; P<.001) and resulted in an additional 36,607 days of inpatient antibiotic treatment. Urine culture on admission resulted in a 2.1% increase in LOS (P=.004). The predicted difference in bed days of care between admissions with and without a urine culture resulted in 6,071 additional bed days of care. The impact of urine culture testing varied by admitting diagnosis.

Patients with a urine culture sent on day 1 of hospital admission receive more days of antibiotics and have a longer hospital stay than patients who do not have a urine culture. Targeted interventions may reduce the potential harms associated with low-yield urine cultures on day 1.

Infect Control Hosp Epidemiol 2018;39:547–554

Surgical site infections (SSIs) following colorectal surgery (CRS) are among the most common healthcare-associated infections (HAIs). Reduction in colorectal SSI rates is an important goal for surgical quality improvement.

To examine rates of SSI in patients with and without cancer and to identify potential predictors of SSI risk following CRS

American College of Surgeons National Surgical Quality Improvement Program (ACS NSQIP) data files for 2011–2013 from a sample of 12 National Comprehensive Cancer Network (NCCN) member institutions were combined. Pooled SSI rates for colorectal procedures were calculated and risk was evaluated. The independent importance of potential risk factors was assessed using logistic regression.

Multicenter study

Of 22 invited NCCN centers, 11 participated (50%). Colorectal procedures were selected by principal procedure current procedural technology (CPT) code. Cancer was defined by International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) codes.

The primary outcome of interest was 30-day SSI rate.

A total of 652 SSIs (11.06%) were reported among 5,893 CRSs. Risk of SSI was similar for patients with and without cancer. Among CRS patients with underlying cancer, disseminated cancer (SSI rate, 17.5%; odds ratio [OR], 1.66; 95% confidence interval [CI], 1.23–2.26; P=.001), ASA score ≥3 (OR, 1.41; 95% CI, 1.09–1.83; P=.001), chronic obstructive pulmonary disease (COPD; OR, 1.6; 95% CI, 1.06–2.53; P=.02), and longer duration of procedure were associated with development of SSI.

Patients with disseminated cancer are at a higher risk for developing SSI. ASA score >3, COPD, and longer duration of surgery predict SSI risk. Disseminated cancer should be further evaluated by the Centers for Disease Control and Prevention (CDC) in generating risk-adjusted outcomes.

Infect Control Hosp Epidemiol 2018;39:555–562

Clostridium difficile infection (CDI) is common in solid organ transplant (SOT) recipients, but few studies have examined long-term outcomes. We studied the impact of CDI after SOT on mortality and transplant organ complication-related hospitalizations (TOH).

SOT recipients ≥18 years of age with at least 1 year of posttransplant data were analyzed using the MarketScan database for 2007–2014. Patients who died within one year of transplant were followed until death. Patients were grouped as early CDI (ie, first occurrence ≤90 days posttransplant), late CDI (ie, first occurrence >90 days posttransplant) and controls (ie, no CDI occurrence during follow-up). The risk of mortality or TOH after CDI was evaluated using Cox and logistic regressions, respectively.

Overall, 96 patients had early CDI, 97 patients had late CDI, and 5,913 patients were used as controls. The risk for death was significantly higher in the early CDI group than the control group (hazard ratio [HR],1.92; 95% confidence interval [CI], 1.12–3.29; P=.018); there was no significant difference between the late CDI group and the control group (HR, 0.86; 95% CI, 0.38–1.94; P=.717). Both the early CDI group (odds ratio [OR], 2.19; 95% CI, 1.45–3.31; P<.001) and the late CDI group (OR, 4.36; 95% CI, 2.84–6.71; P<.001) had higher risk for TOH than the control group. For those patients who survived >90 days posttransplant, both the early CDI group (n=89) and the late CDI group (n=97) had increased risk for death or TOH during follow-up than the control group (n=5,734).

Though our study could not prove causality, both early and late CDI occurrence in SOT recipients were associated with worse future outcomes than for SOT recipients without CDI.

Infect Control Hosp Epidemiol 2018;39:563–570

Multidrug-resistant organisms (MDROs) are increasingly reported in residential care homes for the elderly (RCHEs). We assessed whether implementation of directly observed hand hygiene (DOHH) by hand hygiene ambassadors can reduce environmental contamination with MDROs.

From July to August 2017, a cluster-randomized controlled study was conducted at 10 RCHEs (5 intervention versus 5 nonintervention controls), where DOHH was performed at two-hourly intervals during daytime, before meals and medication rounds by a one trained nurse in each intervention RCHE. Environmental contamination by MRDOs, such as methicillin-resistant Staphylococcus aureus (MRSA), carbapenem-resistant Acinetobacter species (CRA), and extended-spectrum β-lactamse (ESBL)–producing Enterobacteriaceae, was evaluated using specimens collected from communal areas at baseline, then twice weekly. The volume of alcohol-based hand rub (ABHR) consumed per resident per week was measured.

The overall environmental contamination of communal areas was culture-positive for MRSA in 33 of 100 specimens (33%), CRA in 26 of 100 specimens (26%), and ESBL-producing Enterobacteriaceae in 3 of 100 specimens (3%) in intervention and nonintervention RCHEs at baseline. Serial monitoring of environmental specimens revealed a significant reduction in MRSA (79 of 600 [13.2%] vs 197 of 600 [32.8%]; P<.001) and CRA (56 of 600 [9.3%] vs 94 of 600 [15.7%]; P=.001) contamination in the intervention arm compared with the nonintervention arm during the study period. The volume of ABHR consumed per resident per week was 3 times higher in the intervention arm compared with the baseline (59.3±12.9 mL vs 19.7±12.6 mL; P<.001) and was significantly higher than the nonintervention arm (59.3±12.9 mL vs 23.3±17.2 mL; P=.006).

The direct observation of hand hygiene of residents could reduce environmental contamination by MDROs in RCHEs.

Infect Control Hosp Epidemiol 2018;39:571–577

To assess general medical residents’ familiarity with antibiograms using a self-administered survey

Cross-sectional, single-center survey

Residents in internal medicine, family medicine, and pediatrics at an academic medical center

Participants were administered an anonymous survey at our institution during regularly scheduled educational conferences between January and May 2012. Questions collected data regarding demographics, professional training; further open-ended questions assessed knowledge and use of antibiograms regarding possible pathogens, antibiotic regimens, and prescribing resources for 2 clinical vignettes; a series of directed, closed-ended questions followed. Bivariate analyses to compare responses between residency programs were performed.

Of 122 surveys distributed, 106 residents (87%) responded; internal medicine residents accounted for 69% of responses. More than 20% of residents could not accurately identify pathogens to target with empiric therapy or select therapy with an appropriate spectrum of activity in response to the clinical vignettes; correct identification of potential pathogens was not associated with selecting appropriate therapy. Only 12% of respondents identified antibiograms as a resource when prescribing empiric antibiotic therapy for scenarios in the vignettes, with most selecting the UpToDate online clinical decision support resource or The Sanford Guide. When directly questioned, 89% reported awareness of institutional antibiograms, but only 70% felt comfortable using them and only 44% knew how to access them.

When selecting empiric antibiotics, many residents are not comfortable using antibiograms as part of treatment decisions. Efforts to improve antibiotic use may benefit from residents being given additional education on both infectious diseases pharmacotherapy and antibiogram utilization.

Infect Control Hosp Epidemiol 2018;39:578–583

To characterize trends in outpatient antibiotic prescriptions in the United States

Retrospective ecological and temporal trend study evaluating outpatient antibiotic prescriptions from 2013 to 2015

National administrative claims data from a pharmacy benefits manager PARTICIPANTS. Prescription pharmacy beneficiaries from Express Scripts Holding Company

Annual and seasonal percent change in antibiotic prescriptions

Approximately 98 million outpatient antibiotic prescriptions were filled by 39 million insurance beneficiaries during the 3-year study period. The most commonly prescribed antibiotics were azithromycin, amoxicillin, amoxicillin/clavulanate, ciprofloxacin, and cephalexin. No significant changes in individual or overall annual antibiotic prescribing rates were found during the study period. Significant seasonal variation was observed, with antibiotics being 42% more likely to be prescribed during February than September (peak-to-trough ratio [PTTR], 1.42; 95% confidence interval [CI], 1.39–1.61). Similar seasonal trends were found for azithromycin (PTTR, 2.46; 95% CI, 2.44–3.47), amoxicillin (PTTR, 1.52; 95% CI, 1.42–1.89), and amoxicillin/clavulanate (PTTR, 1.78; 95% CI, 1.68–2.29).

This study demonstrates that annual national outpatient antibiotic prescribing practices remained unchanged during our study period. Furthermore, seasonal peaks in antibiotics generally used to treat viral upper respiratory tract infections remained unchanged during cold and influenza season. These results suggest that inappropriate prescribing of antibiotics remains widespread, despite the concurrent release of several guideline-based best practices intended to reduce inappropriate antibiotic consumption; however, further research linking national outpatient antibiotic prescriptions to associated medical conditions is needed to confirm these findings.

Infect Control Hosp Epidemiol 2018;39:584–589

Hospitalized patients with suspected tuberculosis (TB) are placed in airborne isolation until 3 sputum smear samples are negative for acid-fast bacilli (AFB). The Xpert MTB/RIF assay (“Xpert”) nucleic acid amplification test (NAAT) to identify Mycobacterium tuberculosis DNA and resistance to rifampicin is superior to AFB sputum smear microscopy for the diagnosis of TB.

To compare the performance of a single Xpert to AFB smear microscopy for time to airborne infection isolation (AII) discontinuation.

Consecutive patients over 17 years of age in AII for suspected pulmonary TB between October 1, 2014, and March 31, 2016, with leftover respiratory AFB samples were enrolled in this study. A single Xpert was performed on the first available sample. Demographic, clinical, and microbiological data were recorded for each patient. We compared the duration of AII using a single Xpert to AFB smear microscopy under multiple theoretical scenarios using Kaplan-Meier cumulative incidence curves and the log-rank test.

In total, 131 samples were included in our performance analysis of the Xpert, and 114 samples were included in our AII analysis. Overall, 81 patients (65%) were immunosuppressed, of whom 46 (37%) were positive for human immunodeficiency virus (HIV). The sensitivity and specificity of Xpert for diagnosis of M. tuberculosis infection were 67% and 100%, respectively. Xpert was negative in all cases of nontuberculous mycobacteria. Use of a single Xpert reduced AII duration from a median of 67 hours per patient to 42 hours with usual reporting, to 26 hours with direct communication, and to 12 hours with immediate testing.

A single negative Xpert result can reduce AII duration compared to the AFB smear microscopy technique under multiple theoretical scenarios.

Infect Control Hosp Epidemiol 2018;39:590–595