Risk Adjustment Methods Research Articles

NSQIP Quality Benchmarking and Evaluation of Potential Bias Associated with Higher- or Lower-Risk Operation Case Mix.

To ensure validity and acceptance of NSQIP risk-adjusted benchmarking, it is important that adjustments adequately control for hospitals that vary in their proportions of lower- or higher-risk operations (combined risk for procedure and patient). This issue was addressed in separate empirical and simulation studies. For the empirical study, potential miscalibration bias favoring hospitals that do lower-risk operations or disfavoring hospitals that do higher-risk operations was evaluated for 14 modeled outcomes using NSQIP data. A determination was also made as to whether there was a relationship between mean hospital operation risk and benchmarking results (log odds ratio). In the simulation study of the same 14 outcomes, hospital benchmarked performance was evaluated when sampled cases were reconstituted to include either a larger proportion of lower-risk operations or a larger proportion of higher-risk operations. Miscalibration favoring either lower- or higher-risk operations was absent, as were important associations between operative risk and hospital log odds ratios (most model R 2 less than 0.01). In the simulation, there were no substantial changes in log odds ratios when greater percentages of either lower- or higher-risk operations were included in a hospital's sample (nonsignificant p values and effect sizes less than 0.1). These results should enhance NSQIP participants' confidence in the adequacy of NSQIP patient and procedure risk-adjustment methods. NSQIP participants may rely on benchmarking findings, and implement quality improvement efforts based on them, without concern that they are biased by a preponderance of lower or higher risk operations.

The National Outcomes Evaluation Programme in Italy: The Impact of Publication of Health Indicators.

In Italy the National Outcomes Evaluation Programme, (P.N.E.) is the most comprehensive comparative evaluation of healthcare outcomes at the national level. The aim of this report is to describe the P.N.E. and some of the most relevant results achieved. The P.N.E. analysed 184 indicators on quality of care in 2015–2020 period. The data sources are the Italian Health Information Systems. The indicators reported were: proportion of surgery within 2 days after hip fracture in the elderly (HF), 30-day mortality after hospital admission for acute myocardial infarction (AMI), proportion of reoperations within 90 days of breast-conserving surgery and proportion of primary caesarean deliveries. Risk adjustment methods were used to take into account patients’ characteristics. From 2010 to 2020 the proportion of interventions within 2 days after HF increased from 31.3% to 64.6%, the AMI 30-day mortality decreased from 10.4% to 8.3%, the proportion of reinterventions within 90 days of breast-conserving surgery decreased from 12.0% to 5.9% and the proportion of primary caesarean deliveries decreased from 28.4% to 22.7%. Results by area of residence showed heterogeneity of healthcare quality. We observed a general improvement in different clinical areas not always associated with a reduction of heterogeneity among areas of residence.

Modeling Coding Intensity of Procedures in a U.S. Population-Based Hip/Knee Arthroplasty Inpatient Cohort Adjusting for Patient- and Facility-Level Characteristics.

Variations in procedure coding intensity, defined as excess coding of procedures versus industry (instead of clinical) standards, can result in differentials in quality of care for patients and have additional implications for facilities and payors. The literature regarding coding intensity of procedures is limited, with a need for risk-adjusted methods that help identify over- and under-coding using commonly available data, such as administrative claims. Risk-adjusted metrics are needed for quality control and enhancement. We propose a two-step approach to risk adjustment, using a zero-inflated Poisson model, applied to a hip-knee arthroplasty cohort discharged during 2019 (n = 313,477) for patient-level risk adjustment, and a potential additional layer for adjustment based on facility-level characteristics, when desired. A 21.41% reduction in root-mean-square error was achieved upon risk adjustment for patient-level factors alone. Furthermore, we identified facilities that over- and under-code versus industry coding expectations, adjusting for both patient-level and facility-level factors. Excess coding intensity was found to vary across multiple levels: (1) geographically across U.S. Census regional divisions; (2) temporally with marked seasonal components; (3) by facility, with some facilities largely departing from industry standards, even after adjusting for both patient- and facility-level characteristics. Our proposed method is simple to implement, generalizable, it can be used across cohorts with different sets of information available, and it is not limited by the accessibility and sparsity of electronic health records. By identifying potential over- and under-coding of procedures, quality control personnel can explore and assess internal needs for enhancements in their health delivery services and monitor subsequent quality improvements.

Transcatheter and Surgical Aortic Valve Replacement in Patients With Bicuspid Aortic Valve Stenosis

BackgroundTranscatheter aortic valve replacement (TAVR) is not widely used in patients with bicuspid aortic valve (BAV) disease and has not yet been studied in randomized clinical trials. We characterized the rate of use and outcomes of TAVR and surgical aortic valve replacement (SAVR) in patients with BAV. MethodsAdults with BAV stenosis receiving SAVR or TAVR procedures were abstracted from the 2012 to 2019 Nationwide Readmissions Database (NRD). Risk-adjusted analyses were performed with NRD-provided weights and inverse probability of treatment weights (IPTW) to examine the association of treatment strategy on inpatient mortality, complications, and hospitalization resource utilization. Nonelective readmissions within 90 days of discharge and reintervention at the first readmission were also examined. ResultsOf an estimated 56 331 patients with BAV requiring aortic valve replacement, 6.8% underwent TAVR. Unadjusted analysis demonstrated higher index hospitalization mortality for TAVR compared with SAVR. Upon risk adjustment using NRD-provided weights, the odds of pacemaker implantation remained significantly higher for TAVR patients compared with SAVR, with no significant difference in mortality. When NRD-provided survey weights were applied, TAVR had higher rates of 90-day readmission. Adjustment with inverse probability of treatment weights resolved these differences between the 2 groups. Regardless of the risk-adjustment method, the odds of reintervention were consistently higher among BAV TAVR patients compared with SAVR. ConclusionsThe present analysis demonstrates comparable in-hospital mortality and morbidity for TAVR and SAVR patients in the moderate-risk era. With increasing TAVR use in BAV, surgeons must further refine selection criteria with consideration of concomitant aortopathy and implications of reintervention.

Developing national benchmarks for antimicrobial resistance–NHSN, 2019

Background: The emergence and spread of drug-resistant pathogens continues to significantly impact patient safety and healthcare systems. Although antimicrobial susceptibility test (AST) results of clinical specimens are used by individual facilities for antimicrobial resistance surveillance, accurate tracking and benchmark comparison of a facility’s antimicrobial resistance using national data requires risk-adjusted methods to be more meaningful. The CDC NHSN Antimicrobial Resistance (AR) Option collects patient-level, deduplicated, isolate information, including AST results, for >20 organisms from cerebrospinal fluid, lower respiratory tract (LRT), blood, and urinary specimens. To provide risk-adjusted national benchmarks, we developed prediction models for incidence of hospital-onset isolates with antimicrobial resistance. Methods: We analyzed AST results of isolates reported through the NHSN AR Option for January through December 2019. Isolates from facilities that had >10% missing AST results for the organism-drug combinations or from hospitals that used outdated breakpoints were excluded. We assessed associations between facility-level factors and incidence rates of hospital-onset (specimen collected 3 days or more after hospital admission) isolates of specific drug-resistant phenotypes from blood, LRT, and urinary specimens. Factors included number of beds, length of stay, and prevalence of community onset isolates of the same phenotype. Drug-resistant phenotypes assessed included methicillin-resistant Staphylococcus aureus (MRSA), multidrug-resistant (MDR) Pseudomonas aeruginosa, carbapenem-resistant Enterobacterales (CRE), fluoroquinolone-resistant Pseudomonas aeruginosa, fluoroquinolone-resistant Enterobacterales, and extended-spectrum cephalosporin-resistant Enterobacterales. Isolates of different phenotypes and from different specimen sources were modeled separately. Negative binomial regression was used to evaluate the factors associated with antimicrobial resistance incidence. Variable entry into the models is based on significance level P Among the models, 1 for each drug-resistant phenotype-specimen type combination, the number of isolates with AST results ranged from 718 (Pseudomonas aeruginosa–fluoroquinolones, blood) to 16,412 (Enterobacterales–fluoroquinolones, urine). The pooled incidence rate was highest for fluoroquinolone-resistant Enterobacterales in urinary specimens (0.2179 isolates per 1,000 patient days) among all phenotype-specimen combinations evaluated (Table 1). The incidence of drug-resistant isolates was consistently associated with community-onset prevalence across models evaluated. Other associated factors varied across phenotype-specimen combinations (Table 2). Conclusions: We developed statistical models to predict facility-level incidence rates of hospital-onset antimicrobial resistant isolates based on community-onset drug-resistant prevalence and facility characteristics. These models will enable facilities to compare antimicrobial resistance rates to the national benchmarks and therefore to inform their antimicrobial stewardship and infection prevention efforts.Funding: NoneDisclosures: None

Centering equity in mental health crisis services.

Centering equity in mental health crisis services.

Equity and the Hazard of Veiled Injustice: A Methodological Reflection on Risk Adjustment.

Equity and the Hazard of Veiled Injustice: A Methodological Reflection on Risk Adjustment.

Development of Deep Learning Models for Predicting In-Hospital Mortality Using an Administrative Claims Database: Retrospective Cohort Study.

BackgroundAdministrative claims databases have been used widely in studies because they have large sample sizes and are easily available. However, studies using administrative databases lack information on disease severity, so a risk adjustment method needs to be developed.ObjectiveWe aimed to develop and validate deep learning–based prediction models for in-hospital mortality of acute care patients.MethodsThe main model was developed using only administrative claims data (age, sex, diagnoses, and procedures on the day of admission). We also constructed disease-specific models for acute myocardial infarction, heart failure, stroke, and pneumonia using common severity indices for these diseases. Using the Japanese Diagnosis Procedure Combination data from July 2010 to March 2017, we identified 46,665,933 inpatients and divided them into derivation and validation cohorts in a ratio of 95:5. The main model was developed using a 9-layer deep neural network with 4 hidden dense layers that had 1000 nodes and were fully connected to adjacent layers. We evaluated model discrimination ability by an area under the receiver operating characteristic curve (AUC) and calibration ability by calibration plot.ResultsAmong the eligible patients, 2,005,035 (4.3%) died. Discrimination and calibration of the models were satisfactory. The AUC of the main model in the validation cohort was 0.954 (95% CI 0.954-0.955). The main model had higher discrimination ability than the disease-specific models.ConclusionsOur deep learning–based model using diagnoses and procedures produced valid predictions of in-hospital mortality.

Risk Adjustment Model for Preserved Health Status in Patients With Heart Failure and Reduced Ejection Fraction: The CHAMP-HF Registry.

Health status outcomes are increasingly being promoted as measures of health care quality, given their importance to patients. In heart failure (HF), an American College of Cardiology/American Heart Association Task Force proposed using the proportion of patients with preserved health status as a quality measure but not as a performance measure because risk adjustment methods were not available. We built risk adjustment models for alive with preserved health status and for preserved health status alone in a prospective registry of outpatients with HF with reduced ejection fraction across 146 US centers between December 2015 and October 2017. Preserved health status was defined as not having a ≥5-point decrease in the Kansas City Cardiomyopathy Questionnaire Overall Summary score at 1 year. Using only patient-level characteristics, hierarchical multivariable logistic regression models were developed for 1-year outcomes and validated using data from 1 to 2 years. We examined model calibration, discrimination, and variability in sites' unadjusted and adjusted rates. Among 3932 participants (median age [interquartile range] 68 years [59-75], 29.7% female, 75.4% White), 2703 (68.7%) were alive with preserved health status, 902 (22.9%) were alive without preserved health status, and 327 (8.3%) had died by 1 year. The final risk adjustment model for alive with preserved health status included baseline Kansas City Cardiomyopathy Questionnaire Overall Summary, age, race, employment status, annual income, body mass index, depression, atrial fibrillation, renal function, number of hospitalizations in the past 1 year, and duration of HF (optimism-corrected C statistic=0.62 with excellent calibration). Similar results were observed when deaths were ignored. The risk standardized proportion of patients alive with preserved health status across the 146 sites ranged from 62% at the 10th percentile to 75% at the 90th percentile. Variability across sites was modest and changed minimally with risk adjustment. Through leveraging data from a large, outpatient, observational registry, we identified key factors to risk adjust sites' proportions of patients with preserved health status. These data lay the foundation for building quality measures that quantify treatment outcomes from patients' perspectives.

Elucidating the association between regional variation in diagnostic frequency with risk-adjusted mortality through analysis of claims data of medicare inpatients: a cross-sectional study

ObjectiveThe validity of risk-adjustment methods based on administrative data has been questioned because hospital referral regions with higher diagnosis frequencies report lower case-fatality rates, implying that diagnoses do not track...

Hyperoxemia Is Associated With Mortality in Critically Ill Children.

Multiple studies among adults have suggested a non-linear relationship between arterial partial pressure of oxygen (PaO2) and clinical outcomes. Meta-analyses in this population suggest that high levels of supplemental oxygen resulting in hyperoxia are associated with mortality. This mini-review focuses on the non-neonatal pediatric literature examining the relationship between PaO2 and mortality. While only one pilot pediatric randomized-controlled trials exists, over the past decade, there have been at least eleven observational studies examining the relationship between PaO2 values and mortality in critically ill children. These analyses of mixed-case pediatric ICU populations have generally reported a parabolic (“u-shaped”) relationship between PaO2 and mortality, similar to that seen in the adult literature. However, the estimates of the point at which hyperoxemia becomes deleterious have varied widely (300–550 mmHg). Where attempted, this effect has been robust to analyses restricted to the first PaO2 value obtained, those obtained within 24 h of admission, anytime during admission, and the number of hyperoxemic blood gases over time. These findings have also been noted when using various methods of risk-adjustment (accounting for severity of illness scores or complex chronic conditions). Similar relationships were found in the majority of studies restricted to patients undergoing care after cardiac arrest. Taken together, the majority of the literature suggests that there is a robust parabolic relationship between PaO2 and risk-adjusted pediatric ICU mortality, but that the exact threshold at which hyperoxemia becomes deleterious is unclear, and likely beyond the typical target value for most clinical indications. Findings suggest that clinicians should remain judicious and thoughtful in the use of supplemental oxygen therapy in critically ill children.

Möglichkeiten und Grenzen der Risikoadjustierung von Qualitätsindikatoren aus Krankenhausabrechnungsdaten - ein Werkstattbericht

The quality indicators of the Initiative Qualitätsmedizin e.V. (IQM) have been developed as triggers to examine treatment processes for opportunities for improvement. Published quality results have partly been used for external quality comparisons in the media. Therefore, member hospitals of IQM demanded to investigate if methods of risk adjustment should be applied in the calculation of the quality indicators. After a hearing of experts had been held, a task force was founded to conduct test calculations on risk adjustment methods. Specific risk adjustment models for mortality in myocardial infarction, heart failure, stroke, pneumonia, and colectomy in colorectal cancer were developed in the database of national German DRG data of the year 2016. These models were used to calculate standardized mortality ratios (SMR) per indicator in a sample of 172 member hospitals of IQM based on the data of the year 2018. Median SMR per indicator were compared to median SMR based on a standardization by age and gender, which is the standard procedure in IQM. Correlations between the different SMR were calculated. Quality of care was judged by two different approaches: a) a descriptive discrepancy of |0.1| from the SMR value of 1, and b) a significant discrepancy from 1 using the 95% confidence limits. The effect of using the specific risk adjustment in relation to the standard procedure was investigated for both approaches (a and b). The specific risk adjustment methods showed an area under the curve between 0.72 and 0.84. The median differences between the SMR based on standardization by age and gender and the SMR based on specific risk adjustment were small (between 0 and 0.4); Spearman's correlations were between 0.90 and 0.99. Changes in the judgement of quality of care in comparison to the national average occurred in 3.9% (mortality from pneumonia) to 20.6% of the hospitals (mortality from heart failure) in descriptive comparisons. When the judgement was based on confidence limits changes were observed in 1.6% (mortality after colectomy) to 17.4% of the hospitals (mortality from heart failure). Implementing specific risk adjustment models had only minor effects on the distribution of risk-adjusted mortality compared to the standard procedure, but the judgement of quality of care could change for a fifth of the hospitals in individual indicators. Concerning methodological and practical reasons, the task force recommends further development of risk adjustment methods for selected indicators. This should be accompanied by studies on the validity of inpatient administrative data for quality management as well as by efforts to improve the usefulness of these data for such purposes.

Partnership Forum: The Role of Research in the Transformation of Veterans Affairs Community Care.

Partnership Forum: The Role of Research in the Transformation of Veterans Affairs Community Care.

Potential of community-based risk estimates for improving hospital performance measures and discharge planning

BackgroundRisk-adjusted rates of hospital readmission are a common indicator of hospital performance. There are concerns that current risk-adjustment methods do not account for the many factors outside the hospital setting...

Treatment plan quality control using multivariate control charts

Statistical process control tools such as control charts were recommended by the American Association of Physicists in Medicine (AAPM) Task Group 218 for radiotherapy quality assurance. However, the tools needed to analyze multivariate, correlated data that is often encountered in treatment plan quality measures, are lacking. In this study, we develop quality control tools that can model multivariate plan quality measures with correlations and account for patient-specific risk-factors, without adding a significant burden to clinical workflow. A multivariate, quality control chart is developed that includes a risk-adjustment model, Hotelling's T2 statistic, and principal component analysis (PCA). PCA accounts for correlations among a set of organ-at-risk (OAR) dose-volume histogram (DVH) points that serves as proxies for plan quality. Risk-adjustment models estimate the principal components from PCA using a set of patient- and treatment-specific risk factors. The resulting residuals from the risk-adjustment models are used to compute the Hotelling's T2 statistic; the corresponding multivariate control chart is then plotted based on the beta distribution followed by the statistic. Further, the box-cox transformation is used to account for non-normality in DVH points. We investigate the application of the proposed methodology via three multivariate control charts-a conventional chart that ignores risk-adjustment and PCA, a risk-adjusted chart ignoring PCA, and a PCA-based, risk-adjusted chart. These control charts are evaluated on 69 head-and-neck cases. The conventional multivariate control chart fails to account for important patient-specific risk factors, including volumes and cross-sectional areas of the tumor and OARs and distances in-between. This failure leads to a larger number of false alarms. While the multivariate risk-adjusted control chart is able to reduce false alarms, it fails to account for correlations in DVH points. The multivariate PCA-based, risk-adjusted control chart can detect unusual plans after accounting for the correlations. By replanning, improvements are shown on an unusual plan identified by both risk-adjusted methods. The multivariate risk-adjusted control chart developed here enables quality control of plans prior to delivery. This methodology is generic and can be readily applied for other radiotherapy quality assurance protocols, such as gamma analysis pass-rates.

One-year Mortality after Lung Transplant is Not a Reliable Indicator of Lung Transplant Center Performance

<h3>Purpose</h3> Measuring performance in lung transplantation (LTx) is complex, but in the United States the most common performance indicator is one-year mortality. This indicator is shared with centers in quarterly reports and is used by the United Network for Organ Sharing (UNOS) to identify programs for review and trigger penalties including loss of accreditation. Reliability adjustment, a novel technique for quantifying and removing statistical "noise" from quality rankings, is becoming more widely used outside surgery. We sought to evaluate the reliability of 1-year mortality outcomes in LTx. <h3>Methods</h3> We used the Standard Transplant Analysis and Research (STAR) files form UNOS to identify LTx recipients from 2013-2018 in the United States. We first used standard risk adjustment methods to estimate expected 1-year mortality rates at the hospital level. We then examined the effect of adjusting for reliability using empirical Bayes techniques. <h3>Results</h3> We identified 12,769 recipients in 69 centers over the study period. Reliability adjustment greatly diminished apparent variation in hospital outcomes. For risk-adjusted 1-year mortality, there was a large range (range 0 to 89%), which decreased to a 2.5-fold difference (6 to 15%) after reliability adjustment. Reliability adjustment had a large impact on hospital mortality rankings. For example, with rankings based on mortality, 22% (4 hospitals) of the "best" hospitals (top 25%) were reclassified after reliability adjustment . Similarly, 18% (3 hospitals) of the "worst" hospitals (bottom 25%) were reclassified after reliability adjustment. The overall reliability of 1-year mortality was 0.42. <h3>Conclusion</h3> Conclusions: The reliability of 1-year mortality benchmarking in lung transplantation is low. Accounting for reliability yields more accurate measures of LTx center performance.

Novel method of calculating adjusted antibiotic use by microbiological burden.

To determine the usefulness of adjusting antibiotic use (AU) by prevalence of bacterial isolates as an alternative method for risk adjustment beyond hospital characteristics. Retrospective, observational, cross-sectional study. Hospitals in the southeastern United States. AU in days of therapy per 1,000 patient days and microbiologic data from 2015 and 2016 were collected from 26 hospitals. The prevalences of Pseudomonas aeruginosa, extended-spectrum β-lactamase (ESBL)-producing bacteria, methicillin-resistant Staphylococcus aureus (MRSA), and vancomycin-resistant enterococci (VRE) were calculated and compared to the average prevalence of all hospitals in the network. This proportion was used to calculate the adjusted AU (a-AU) for various categories of antimicrobials. For example, a-AU of antipseudomonal β-lactams (APBL) was the AU of APBL divided by (prevalence of P. aeruginosa at that hospital divided by the average prevalence of P. aeruginosa). Hospitals were categorized by bed size and ranked by AU and a-AU, and the rankings were compared. Most hospitals in 2015 and 2016, respectively, moved ≥2 positions in the ranking using a-AU of APBL (15 of 24, 63%; 22 of 26, 85%), carbapenems (14 of 23, 61%; 22 of 25; 88%), anti-MRSA agents (13 of 23, 57%; 18 of 26, 69%), and anti-VRE agents (18 of 24, 75%; 15 of 26, 58%). Use of a-AU resulted in a shift in quartile of hospital ranking for 50% of APBL agents, 57% of carbapenems, 35% of anti-MRSA agents, and 75% of anti-VRE agents in 2015 and 50% of APBL agents, 28% of carbapenems, 50% of anti-MRSA agents, and 58% of anti-VRE agents in 2016. The a-AU considerably changes how hospitals compare among each other within a network. Adjusting AU by microbiological burden allows for a more balanced comparison among hospitals with variable baseline rates of resistant bacteria.

The Society of Thoracic Surgeons Congenital Heart Surgery Database: 2020 Update on Outcomes and Research

The Society of Thoracic Surgeons (STS) Congenital Heart Surgery Database (CHSD) continues to be a highly regarded, comprehensive clinical outcomes database that captures more than 90% of all congenital heart surgery cases in the United States and has more than 90% of all congenital heart surgery centers as participants. This report includes aggregate information on clinical outcomes evaluated at the aggregate and The Society of Thoracic Surgeons-European Association for Cardio-Thoracic Surgery (STAT) category levels for the period July 1, 2015, through June 30, 2019. It also includes the published research activities that use data from the STS CHSD. Also included is information on the STS CHSD auditing function, a summary of the significant revisions to the data, which is collected on each patient, and an update on efforts to update the risk-adjustment methods for evaluation of the outcomes.

Variations in Concurrent Central-Line Use Among Central-Line–Associated Bloodstream Infection (CLABSI) Patients by National Healthcare Safety Network (NHSN) Location Type

Background: Current NHSN denominator reporting for central-line–associated bloodstream infection (CLABSI) counts each patient day with n central lines as 1 central-line day. The NHSN does not directly adjust for potential increased risk of CLABSI from concurrent central lines, but the current NHSN standardized infection ratio (SIR) methods may account for differences in concurrence by adjusting for location type. Objective: We examined differences in central-line concurrence by NHSN location type among CLABSI patients. Methods: In a retrospective cohort of adults with CLABSI at 4 hospitals from 2012 to 2017, we linked central-line data to encounter and CLABSI data. Central lines were considered concurrent if they overlapped for &gt;1 day. We calculated proportion of patients with concurrence at both NHSN location and SIR group levels; risk ratios for concurrence between NHSN location types within each SIR group (ie,, locations defined by SIR models as equal “risk”) were determined. Results: In total, 930 CLABIs were identified from 19 NHSN-defined locations that map to 7 SIR groups. Most CLABSIs occurred in locations mapped to either of 2 SIR groups: wards (227, 16% concurrence) and ICUs (294, 33% concurrence). The ward group had 3 NHSN locations (median, 78 CLABSIs) with concurrence range 8% (medical-surgical ward) to 20% (surgical ward). The ICU group had 6 NHSN locations (median, 47.5 CLABSIs) and concurrence ranged from 20% (neurosurgical ICU) to 39% (medical ICU). Despite the noted variations, no risk ratio was statistically different within each SIR group (Table 1). Conclusions: In patients with CLABSIs, the frequency of concurrence varied up to 2-fold between location types within the current NHSN SIR groups, though not statistically significantly. Assessing whether this difference in magnitude persists in all patients with central lines is an important next step in refining risk adjustment methods to account for concurrent central-line use.Funding: NoneDisclosures: Scott Fridkin reports that his spouse receives consulting fees from the vaccine industry.

Hospital Microbiologic Culture Results to Predict the Use of Anti–methicillin-Resistant Staphylococcus aureus (MRSA)

Background: To provide a standardized, risk-adjusted method for summarizing antimicrobial use (AU), the Centers for Disease Control and Prevention developed the standardized antimicrobial administration ratio, an observed-to-predicted use ratio in which predicted use is estimated from a statistical model accounting for patient locations and hospital characteristics. The infection burden, which could drive AU, was not available for assessment. To inform AU risk adjustment, we evaluated the relationship between the burden of drug-resistant gram-positive infections and the use of anti-MRSA agents. Methods: We analyzed data from acute-care hospitals that reported ≥10 months of hospital-wide AU and microbiologic data to the National Healthcare Safety Network (NHSN) from January 2018 through June 2019. Hospital infection burden was estimated using the prevalence of deduplicated positive cultures per 1,000 admissions. Eligible cultures included blood and lower respiratory specimens that yielded oxacillin/cefoxitin–resistant Staphylococcus aureus (SA) and ampicillin-nonsusceptible enterococci, and cerebrospinal fluid that yielded SA. The anti-MRSA use rate is the total antimicrobial days of ceftaroline, dalbavancin, daptomycin, linezolid, oritavancin, quinupristin/dalfopristin, tedizolid, telavancin, and intravenous vancomycin per 1,000 days patients were present. AU rates were modeled using negative binomial regression assessing its association with infection burden and hospital characteristics. Results: Among 182 hospitals, the median (interquartile range, IQR) of anti-MRSA use rate was 86.3 (59.9–105.0), and the median (IQR) prevalence of drug-resistant gram-positive infections was 3.4 (2.1–4.8). Higher prevalence of drug-resistant gram-positive infections was associated with higher use of anti-MRSA agents after adjusting for facility type and percentage of beds in intensive care units (Table 1). Number of hospital beds, average length of stay, and medical school affiliation were nonsignificant. Conclusions: Prevalence of drug-resistant gram-positive infections was independently associated with the use of anti-MRSA agents. Infection burden should be used for risk adjustment in predicting the use of anti-MRSA agents. To make this possible, we recommend that hospitals reporting to NHSN’s AU Option also report microbiologic culture results.Funding: NoneDisclosures: None