Hospital Readmissions to Nonindex Hospitals: Patterns and Determinants Following the Medicare Readmission Reduction Penalty Program

Centers for Disease Control (CDC) Wound Classification is Prognostic of 30-Day Readmission Following Surgery

BACKGROUND

The goal of this study was to investigate factors associated with 30-day readmission in a multivariate model, including the CDC wound classes "clean," "clean/contaminated," "contaminated," and "dirty/infected."

METHODS

The 2017-2020 American College of Surgeons-National Surgical Quality Improvement Program (ACS-NSQIP) database was queried for all patients undergoing total hip replacement, coronary artery bypass grafting, Ivor Lewis esophagectomy, pancreaticoduodenectomy, distal pancreatectomy, pneumonectomy, and colectomies. ACS-defined wound classes were concordant with CDC definitions. Multivariate linear mixed regression was used to determine risk factors for readmission while adjusting for type of surgery as a random intercept.

RESULTS

477,964 cases were identified, with 38,734 (8.1%) patients having experienced readmission within 30 days of surgery. There were 181,243 (37.9%) cases classified as wound class "clean", 215,729 (45.1%) cases classified as "clean/contaminated", 40,684 cases (8.5%) classified as "contaminated", and 40,308 (8.4%) cases classified as "dirty/infected". In the multivariate generalized mixed linear model adjusting for type of surgery, sex, body mass index, race, American Society of Anesthesiologists class, presence of comorbidity, length of stay, urgency of surgery, and discharge destination, "clean/contaminated" (p < .001), "contaminated" (p < .001), and "dirty/infected" (p < .001) wound classes (when compared to "clean") were significantly associated with 30-day readmission. Organ/space surgical site infection and sepsis were among the most common reasons for readmission in all wound classes.

CONCLUSIONS

Wound classification was strongly prognostic for readmission in multivariable models, suggesting that it may serve as a marker of readmissions. Surgical procedures that are "non-clean" are at significantly greater risk for 30-day readmission. Readmissions may be due to infectious complications; optimizing antibiotic use or source control to prevent readmission are areas of future study.

Failure to Rescue in Emergency General Surgery: Impact of Fragmentation of Care

OBJECTIVE

Compare EGS patient outcomes after index and nonindex hospital readmissions, and explore predictive factors for nonindex readmission.

BACKGROUND

Readmission to a different hospital leads to fragmentation of care. The impact of nonindex readmission on patient outcomes after EGS is not well established.

METHODS

The Nationwide Readmissions Database (2017) was queried for adult patients readmitted after an EGS procedure. Patients were stratified and propensity-matched according to readmission destination: index versus nonindex hospital. Outcomes were failure to rescue (FTR), mortality, number of subsequent readmissions, overall hospital length of stay, and total costs. Hierarchical logistic regression was performed to account for clustering effect within hospitals and adjusting for patient- and hospital-level potential confounding factors.

RESULTS

A total of 471,570 EGS patients were identified, of which 79,127 (16.8%) were readmitted within 30 days: index hospital (61,472; 77.7%) versus nonindex hospital (17,655; 22.3%). After 1:1 propensity matching, patients with nonindex readmission had higher rates of FTR (5.6% vs 4.3%; P < 0.001), mortality (2.7% vs 2.1%; P < 0.001), and overall hospital costs [in $1000; 37 (27-64) vs 28 (21-48); P < 0.001]. Nonindex readmission was independently associated with higher odds of FTR [adjusted odds ratio 1.18 (1.03-1.36); P < 0.001]. Predictors of nonindex readmission included top quartile for zip code median household income [1.35 (1.08-1.69); P < 0.001], fringe county residence [1.08 (1.01-1.16); P = 0.049], discharge to a skilled nursing facility [1.28 (1.20-1.36); P < 0.001], and leaving against medical advice [2.32 (1.81-2.98); P < 0.001].

CONCLUSION

One in 5 readmissions after EGS occur at a different hospital. Nonindex readmission carries a heightened risk of FTR.

LEVEL OF EVIDENCE

Level III Prognostic.

STUDY TYPE

Prognostic.

In-hospital interventions for reducing readmissions to acute care for adults aged 65 and over: An umbrella review

PURPOSE

The aim of this umbrella review was to synthesize existing systematic review evidence on the effectiveness of in-hospital interventions to prevent or reduce avoidable hospital readmissions in older people (≥65 years old).

DATA SOURCES

A comprehensive database search was conducted in May 2019 through MEDLINE, EMBASE, CINAHL, the JBI Database of Systematic Reviews, DARE and Epistemonikos.

STUDY SELECTION

Systematic reviews and other research syntheses, including meta-analyses, exploring the effectiveness of hospital-based interventions to reduce readmissions for people aged 65 and older, irrespective of gender or clinical condition, were included for review. If a review did not exclusively focus on this age group, but data for this group could be extracted, then it was considered for inclusion. Only reviews in English were included.

DATA EXTRACTION

Data extracted for each review included the review objective, participant details, setting and context, type of studies, intervention type, comparator and findings.

RESULTS OF DATA SYNTHESIS

Twenty-nine reviews were included for analysis. Within these reviews, 11 intervention types were examined: in-hospital medication review, discharge planning, comprehensive geriatric assessment, early recovery after surgery, transitional care, interdisciplinary team care, in-hospital nutrition therapy, acute care geriatric units, in-hospital exercise, postfall interventions for people with dementia and emergency department-based palliative care. Except for discharge planning and transitional care, none of the interventions significantly reduced readmissions among older adults.

CONCLUSION

There is limited evidence to support the effectiveness of existing hospital-based interventions to reduce readmissions for people aged 65 and older.

Care Discontinuity in Emergency General Surgery: Does Hospital Quality Matter?

BACKGROUND

Changes in care providers and hospitals after emergency general surgery (EGS) (care discontinuity) are associated with increased morbidity and mortality. The cause of these worse outcomes is unknown. Our goal was to determine if hospital quality is associated with mortality after readmissions independent of continuity in care.

STUDY DESIGN

This was a retrospective analysis of Medicare inpatient claims (2007 to 2015). All inpatients older than 65 years of age who underwent 1 of 7 EGS procedures shown to represent 80% of EGS volume, complications, and mortality nationally, were included. Care discontinuity was defined as readmission within 30 days to a nonindex hospital. Hospital quality was determined by hospital-level, risk-adjusted mortality rates by EGS procedure and categorized into high quality (HQ) and low quality (LQ). The primary outcome was overall mortality. Multivariate logistic regression analysis was used to determine the association of discontinuity and mortality.

RESULTS

There were 882,929 EGS patients, 87,232 of whom were readmitted within 30 days of discharge. Care discontinuity was independently associated with mortality (odds ratio [OR] 1.23; 95% CI 1.17 to 1.29). When readmitted patients were stratified by quality of index and readmitting hospital, mortality was associated with the quality of both the index hospital and the readmitting hospital. The highest mortality rate was observed in patients with index admission at low-quality hospitals and readmission to a different low-quality hospital.

CONCLUSIONS

Both care discontinuity and hospital quality are independently associated with mortality in EGS patients. These data support maintaining continuity of care, even at low performing hospitals.

Robust policy evaluation from large-scale observational studies

Under the current policy decision making paradigm we make or evaluate a policy decision by intervening different socio-economic parameters and analyzing the impact of those interventions. This process involves identifying the causal relation between interventions and outcomes. Matching method is one of the popular techniques to identify such causal relations. However, in one-to-one matching, when a treatment or control unit has multiple pair assignment options with similar match quality, different matching algorithms often assign different pairs. Since all the matching algorithms assign pairs without considering the outcomes, it is possible that with the same data and same hypothesis, different experimenters can reach different conclusions creating an uncertainty in policy decision making. This problem becomes more prominent in the case of large-scale observational studies as there are more pair assignment options. Recently, a robust approach has been proposed to tackle the uncertainty that uses an integer programming model to explore all possible assignments. Though the proposed integer programming model is very efficient in making robust causal inference, it is not scalable to big data observational studies. With the current approach, an observational study with 50,000 samples will generate hundreds of thousands binary variables. Solving such integer programming problem is computationally expensive and becomes even worse with the increase of sample size. In this work, we consider causal inference testing with binary outcomes and propose computationally efficient algorithms that are adaptable for large-scale observational studies. By leveraging the structure of the optimization model, we propose a robustness condition that further reduces the computational burden. We validate the efficiency of the proposed algorithms by testing the causal relation between the Medicare Hospital Readmission Reduction Program (HRRP) and non-index readmissions (i.e., readmission to a hospital that is different from the hospital that discharged the patient) from the State of California Patient Discharge Database from 2010 to 2014. Our result shows that HRRP has a causal relation with the increase in non-index readmissions. The proposed algorithms proved to be highly scalable in testing causal relations from large-scale observational studies.