Impact of care coordination on 30-day readmission, mortality, and costs for heart failure.

OBJECTIVES: Although coordination of care has become the main focus of health care reform efforts to improve outcomes and decrease costs, limited information is available concerning the impact of care coordination on 30-day outcomes and costs. We used nationwide, population-based data to examine the influence of care coordination on 30-day readmission, mortality, and costs for heart failure (HF). STUDY DESIGN: We analyzed 20,713 patients with HF 18 years or older discharged from hospitals in 2016 using Taiwan's National Health Insurance Research Database. The coordination of care among a patient's outpatient physicians was measured with care density. METHODS: Multilevel regression models were used after adjustment for patient and hospital characteristics to explore the impact of care density on 30-day readmission, mortality, and costs. RESULTS: Patients with high care coordination had lower odds of 30-day readmission (OR, 0.90; 95% CI, 0.82-0.98) and mortality (OR, 0.83; 95% CI, 0.70-0.99) and lower costs (cost ratio [CR], 0.84; 95% CI, 0.79-0.90) compared with those with low care coordination. Patients with medium care coordination had lower costs (CR, 0.92; 95% CI, 0.86-0.98) than those with low care coordination. CONCLUSIONS: High care coordination is associated with decreased 30-day readmission, mortality, and costs for HF. Enhancing coordination of care has the potential to increase the value of care. It is important to monitor coordination of care and develop strategies to maintain high levels of care coordination for HF.

Cost variation of nonelective surgery for ulcerative colitis across the United States.

BACKGROUND: Although clinical outcomes of surgery for ulcerative colitis (UC) have improved in the modern biologic era, expenditures continue to increase. A contemporary cost analysis of UC operative care is lacking. The present study aimed to characterize risk factors and center-level variation in hospitalization costs after nonelective resection for UC. METHODS: All adults with UC in the 2016-2020 Nationwide Readmissions Database undergoing nonelective colectomy or rectal resection were identified. Mixed-effects models were developed to evaluate patient and hospital factors associated with costs. Random effects were estimated and used to rank hospitals by increasing risk-adjusted center-level costs. High-cost hospitals (HCHs) in the top decile of expenditure were identified, and their association with select outcomes was subsequently assessed. RESULTS: An estimated 10,280 patients met study criteria with median index hospitalization costs of $40,300 (IQR, $26,400-$65,000). Increased time to surgery was significantly associated with a +$2500 increment in costs per day. Compared with low-volume hospitals, medium- and high-volume centers demonstrated a -$5900 and -$8200 reduction in costs, respectively. Approximately 19.2% of variability in costs was attributable to interhospital differences rather than patient factors. Although mortality and readmission rates were similar, HCH status was significantly associated with increased complications (adjusted odds ratio [AOR], 1.39), length of stay (+10.1 days), and nonhome discharge (AOR, 1.78). CONCLUSION: The present work identified significant hospital-level variation in the costs of nonelective operations for UC. Further efforts to optimize time to surgery and regionalize care to higher-volume centers may improve the value of UC surgical care in the United States.

Clinical and Economic Implications of Interventions in Pancreatic Fluid Collections: An Assessment From a National Claims Database.

OBJECTIVE: To compare clinical and economic implications of percutaneous and endoscopic treatment approaches in patients with pancreatic fluid collections (PFCs). MATERIALS AND METHODS: This is a retrospective claims analysis of Medicare beneficiaries who underwent inpatient endoscopic or percutaneous PFC drainage procedures (2016-2020). We performed longitudinal analysis of claims for all-cause mortality and rehospitalization during 180-day follow-up. Main outcome was mortality. Other outcomes were rehospitalization and direct costs. RESULTS: A total of 1311 patients underwent endoscopic (n = 727) or percutaneous (n = 584) drainage. Percutaneous as compared with endoscopic approach was associated with higher mortality (23.08% vs 16.7%, P = 0.004), rehospitalization (58.9% vs 53.3%, P = 0.04), and mean direct hospital costs ($37,107 [SD = $67,833] vs $27,800 [SD = $43,854], P = 0.004). On multivariable analysis, percutaneous drainage (adjusted hazard ratio [HR], 1.38; 95% confidence interval [CI], 1.02-1.86; P = 0.039), older age (hazard ratio [HR], 1.04; 95% CI, 1.01-1.04; P < 0.001), intensive care unit stay (HR, 1.02; 95% CI, 1.01-1.03; P < 0.001), and multiple comorbidities (HR, 1.07; 95% CI, 1.05-1.09; P < 0.001) were significantly associated with mortality. Percutaneous drainage (adjusted odds ratio [OR], 1.30; 95% CI, 1.04-1.63; P = 0.027) and older age (OR, 0.98; 95% CI, 0.97-0.99; P < 0.001) were significantly associated with rehospitalizations. CONCLUSIONS: As percutaneous drainage may be associated with higher mortality, rehospitalization, and costs, when requisite expertise is available, endoscopy should be preferred for treatment of PFC amenable to such an approach. Randomized trials are required to validate these findings.

Association of social determinants of health-related diagnosis codes with postoperative outcomes.

BACKGROUND: The association of an individual's social determinants of health-related problems with surgical outcomes has not been well-characterized. The objective of this study was to determine whether documentation of social determinants of a health-related diagnosis code (Z code) is associated with postoperative outcomes. METHODS: This retrospective cohort study included surgical cases from a single institution's national surgical quality improvement program (NSQIP) clinical registry from October 2015 to December 2021. The primary predictor of interest was documentation of a Z code for social determinants of health-related problems. The primary outcome was 30-day postoperative morbidity. Secondary outcomes included postoperative length of stay, disposition, and 30-day postoperative mortality, reoperation, and readmission. Multivariable regression models were fit to evaluate the association between the documentation of a Z code and outcomes. RESULTS: Of 10,739 surgical cases, 348 patients (3.2%) had a documented social determinants of health-related Z code. In multivariable analysis, documentation of a Z code was associated with increased odds of morbidity (20.7% vs. 9.9%; adjusted odds ratio [aOR], 1.88; 95% confidence interval [CI], 1.39-2.53), length of stay (median, 3 vs. 1 day; incidence rate ratio, 1.49; 95% CI, 1.33-1.67), odds of disposition to a location other than home (11.3% vs. 3.9%; aOR, 2.86; 95% CI, 1.89-4.33), and odds of readmission (15.3% vs. 6.1%; aOR, 1.99; 95% CI, 1.45-2.73). CONCLUSIONS: Social determinants of health-related problems evaluated using Z codes were associated with worse postoperative outcomes. Improved documentation of social determinants of health-related problems among surgical patients may facilitate improved risk stratification, perioperative planning, and clinical outcomes.

Higher Rates Of Emergency Surgery, Serious Complications, And Readmissions In Primary Care Shortage Areas, 2015-19.

Primary care physicians are often the first to screen and identify patients with access-sensitive surgical conditions that should be treated electively. These conditions require surgery that is preferably planned (elective), but, when access is limited, treatment may be delayed and worsening symptoms lead to emergency surgery (for example, colectomy for cancer, abdominal aortic aneurysm repair, and incisional hernia repair). We evaluated the rates of elective versus emergency surgery for patients with three access-sensitive surgical conditions living in primary care Health Professional Shortage Areas during 2015-19. Medicare beneficiaries in more severe primary care shortage areas had higher rates of emergency surgery compared with rates in the least severe shortage areas (37.8 percent versus 29.9 percent). They were also more likely to have serious complications (14.9 percent versus 11.7 percent) and readmissions (15.7 percent versus 13.5 percent). When we accounted for areas with a shortage of surgeons, the findings were similar. Taken together, these findings suggest that residents of areas with greater primary care workforce shortages may also face challenges in accessing elective surgical care. As policy makers consider investing in Health Professional Shortage Areas, our findings underscore the importance of primary care access to a broader range of services.

The impact of different co-morbidities on clinical outcomes and resource utilization in critically ill burn and surgical patients: A population-based analysis of social determinants of health.

BACKGROUND: This study aims to establish the significance of social determinants of health and prevalent co-morbidities on multiple indicators for quality of care in patients admitted to the Burn and Surgical Intensive Care Unit (ICU). METHODS: We performed a retrospective analysis of population group data for patients admitted at the Burn and Surgical ICU from January 1, 2016, to November 18, 2019. The primary outcomes were length of hospital stay (LOS), mortality, 30-day readmission, and hospital charges. Pearson's chi-square test for categorical variables and t-test for continuous variables were used to compare population health groups. RESULTS: We analyzed a total of 487 burn and 510 surgical patients. When comparing ICU patients, we observed significantly higher mean hospital charges and length of stay (LOS) in BICU v. SICU patients with a history of mental health ($93,259.40 v. $50,503.36, p = 0.013 and 16.28 v. 9.16 days, p = 0.0085), end-stage-renal-disease (ESRD) ($653,871.05 v. $75,746.35, p = 0.0047 and 96.15 v. 17.53 days, p = 0.0104), sepsis ($267,979.60 v. $99,154.41, p = <0.001 and 39.1 v. 18.42 days, p = 0.0043), and venous thromboembolism (VTE) ($757,740.50 v. $117,816.40, p = <0.001 and 93.11 v. 20.21 days, p = 0.002). Also, higher mortality was observed in burn patients with ESRD, ST-Elevation Myocardial Infarction (STEMI), sepsis, VTE, and diabetes mellitus. 30-day-readmissions were greater among burn patients with a history of mental health, drug dependence, heart failure, and diabetes mellitus. CONCLUSIONS: Our study provides new insights into the variability of outcomes between burn patients treated in different critical care settings, underlining the influence of comorbidities on these outcomes. By comparing burn patients in the BICU with those in the SICU, we aim to highlight how differences in patient backgrounds, including the quality of care received, contribute to these outcomes. This comparison underscores the need for tailored healthcare strategies that consider the unique challenges faced by each patient group, aiming to mitigate disparities in health outcomes and healthcare spending. Further research to develop relevant and timely interventions that can improve these outcomes.

Sprint Team Approach Yields Rapid Improvement in Leapfrog Quality Indicators.

GOAL: Patient safety and quality care are two critical areas that every healthcare organization strives to grow and improve upon. At Scripps Health, specific efforts reviewed for this article were implemented to reduce hospital-acquired conditions and hospital readmissions that are components of Centers for Medicare & Medicaid Services programs and Leapfrog Hospital Survey scores. METHODS: Sprint teams, a novel approach to rapidly develop a checklist for lower-performing care improvement areas, were implemented after an internal review of existing tools and an evidence-based literature review. These areas included catheter-associated urinary tract infections (CAUTIs), central-line associated bloodstream infections (CLABSIs), Clostridioides difficile (C. diff.) and methicillin-resistant Staphylococcus aureus (MRSA) infections, chronic obstructive pulmonary disease (COPD) and heart failure readmissions, surgical site infections and handwashing, bar coding, and the computerized physician order entry components of Leapfrog scoring. The checklist for each area served as a teaching tool for staff and a guideline for case review to ensure that standard work was routinely performed. PRINCIPAL FINDINGS: The sprint teams showed dramatic results in the initial focus areas. From a baseline standardized infection ratio (SIR) of 1.141 for CLABSIs, the sprint team reduced the SIR to 0.885 in Year 1 of the program and to 0.687 in Year 2. For CAUTIs, the SIR decreased from a baseline of 1.391 in Year 1 to 0.720 in Year 2. C. diff. infections fell from 0.422 to 0.315 in Year 1 and to 0.260 in Year 2. While the MRSA SIR did not improve during the first year, the MRSA reduction sprint team showed success in Year 2 with a decrease in the SIR from 0.537 to 0.245. Readmission reduction sprint teams focused on heart failure, COPD, and total hip and knee complications. The teams also achieved positive results in reducing readmissions by following checklists and reviewing each readmission case for justification. PRACTICAL APPLICATIONS: Rapid change can be safely and effectively implemented with multidisciplinary sprint teams. Developed with an evidence-based, case review approach, sprint team checklists can help to standardize processes for the review of any infections or readmissions that occur in the inpatient arena.

Elective Joint Replacement Among Medicaid Beneficiaries: Utilization and Postoperative Adverse Events by Racial and Ethnic Groups.

Hip and knee replacement have been marked by racial and ethnic disparities in both utilization and postoperative adverse events among Medicare beneficiaries, but limited knowledge exists regarding racial and ethnic differences in joint replacement care among Medicaid beneficiaries. To close this gap, this study used Medicaid claims in 2018 and described racial and ethnic differences in the utilization and postoperative adverse events of elective joint replacements among Medicaid beneficiaries. Among the 2,260,272 Medicaid beneficiaries, 5987 had an elective joint replacement in 2018. Asian (0.05%, 95% confidence interval [CI]: 0.03%-0.07%) and Hispanic beneficiaries (0.12%, 95% CI: 0.07%-0.18%) received joint replacements less frequently than American Indian and Alaska Native (0.41%, 95% CI: 0.27%-0.55%), Black (0.33%, 95% CI: 0.19%-0.48%), and White (0.37%, 95% CI: 0.25%-0.50%) beneficiaries. Black patients demonstrated the highest probability of 90-day emergency department visits (34.8%, 95% CI: 32.7%-37.0%) among all racial and ethnic groups and a higher probability of 90-day readmission (8.0%, 95% CI: 6.9%-9.0%) than Asian (3.4%, 95% CI: 0.7%-6.0%) and Hispanic patients (4.4%, 95% CI: 3.4%-5.3%). These findings indicate evident disparities in postoperative adverse events across racial and ethnic groups, with Black patients demonstrating the highest probability of 90-day emergency department visits. This study represents an initial exploration of the racial and ethnic differences in joint replacement care among Medicaid beneficiaries and lay the groundwork for further investigation into contributing factors of the observed disparities.

Racial Disparities in 30-day Readmission After Orthopaedic Surgery: A 5-year National Surgical Quality Improvement Program Database Analysis.

BACKGROUND: Readmission rate after surgery is an important outcome measure in revealing disparities. This study aimed to examine how 30-day readmission rates and causes of readmission differ by race and specific injury areas within orthopaedic surgery. METHODS: The American College of Surgeon-National Surgical Quality Improvement Program database was queried for orthopaedic procedures from 2015 to 2019. Patients were stratified by self-reported race. Procedures were stratified using current procedural terminology codes corresponding to given injury areas. Multiple logistic regression was done to evaluate associations between race and all-cause readmission risk, and risk of readmission due to specific causes. RESULTS: Of 780,043 orthopaedic patients, the overall 30-day readmission rate was 4.18%. Black and Asian patients were at greater (OR = 1.18, P < 0.01) and lesser (OR = 0.76, P < 0.01) risk for readmission than White patients, respectively. Black patients were more likely to be readmitted for deep surgical site infection (OR = 1.25, P = 0.03), PE (OR = 1.64, P < 0.01), or wound disruption (OR = 1.45, P < 0.01). For all races, all-cause readmission was highest after spine procedures and lowest after hand/wrist procedures. CONCLUSIONS: Black patients were at greater risk for overall, spine, shoulder/elbow, hand/wrist, and hip/knee all-cause readmission. Asian patients were at lower risk for overall, spine, hand/wrist, and hip/knee surgery all-cause readmission. Our findings can identify complications that should be more carefully monitored in certain patient populations.

Serious mental illness is associated with elevated risk of hospital readmission in veterans with heart failure.

OBJECTIVE: Adults with serious mental illness (SMI) have high rates of cardiovascular disease, particularly heart failure, which contribute to premature mortality. The aims were to examine 90- and 365-day all-cause medical or surgical hospital readmission in Veterans with SMI discharged from a heart failure hospitalization. The exploratory aim was to evaluate 180-day post-discharge engagement in cardiac rehabilitation, an effective intervention for heart failure. METHODS: This study used administrative data from the Veterans Health Administration (VHA) and Centers for Medicare & Medicaid Services between 2011 and 2019. SMI status and medical comorbidity were assessed in the year prior to hospitalization. Cox proportional hazards models (competing risk of death) were used to evaluate the relationship between SMI status and outcomes. Models were adjusted for VHA hospital site, demographics, and medical characteristics. RESULTS: The sample comprised 189,767 Veterans of which 23,671 (12.5%) had SMI. Compared to those without SMI, Veterans with SMI had significantly higher readmission rates at 90 (16.1% vs. 13.9%) and 365 (42.6% vs. 37.1%) days. After adjustment, risk of readmission remained significant (90 days: HR: 1.07, 95% CI: 1.03, 1.11; 365 days: HR: 1.10, 95% CI: 1.07, 1.12). SMI status was not significantly associated with 180-day cardiac rehabilitation engagement (HR: 0.98, 95% CI: 0.91, 1.07). CONCLUSIONS: Veterans with SMI and heart failure have higher 90- and 365-day hospital readmission rates even after adjustment. There were no differences in cardiac rehabilitation engagement based on SMI status. Future work should consider a broader range of post-discharge interventions to understand contributors to readmission.

Use of interrupted time-series analyses in evaluating health economic outcomes following implementation of multilayer water-tight wound closure in a primary total joint arthroplasty population.

Aim: Total joint arthroplasty (TJA) with multi-layer, watertight closure (MLWC) using knotless barbed suture and 2-octyl cyanoacrylate plus polymer mesh tape was compared with conventional closure (CC) using Vicryl™ sutures and staples. Patients & methods: Electronic medical records of patients undergoing TJA (1574: total knee arthroplasty; 580: total hip arthroplasty; 13: unknown) from a single surgeon at a US hospital (CC 2011 to 2013; MLWC 2015 to 2020) were reviewed. Outcomes were length of stay (LOS), discharge to skilled nursing facility (SNF), 90-day surgical site infection (SSI) and 90-day readmission. Logistic regression controlled for baseline characteristics. Adjusted interrupted time series (ITS) analyses accounted for decreasing trends in LOS and SNF discharge over time. Results: Among 2167 TJA cases (mean [standard deviation] age 66.0 [9.7] years, 53.3% female), 906 received CC and 1261 received MLWC. Bivariate analysis showed no statistically significant differences in 90-day SSI rates; however, MLWC patients had 60% lower 90-day readmission rates (1.5 vs 3.8%, p < 0.05), 44% lower LOS (1.4 vs 2.5 days, p < 0.05) and 40% lower discharge rates to a skilled care facility (8.5 vs 14.1%, p < 0.05). Multivariable analyses showed CC patients were 2.45-times more likely to be readmitted within 90 days, 1.88-times more likely to be discharged to SNF and had 1.67-times longer LOS compared with MLWC. ITS analyses showed a sharp decline in LOS (0.9 days) and discharge to SNF (5.6% incidence) after implementation of MLWC, followed by no further changes for the remainder of the study period. Conclusion: MLWC was associated with ≥40% reduction in 90-day readmission, LOS and SNF discharge compared with TJA CC. LOS and discharge rate to SNF declined sharply after the implementation of MLWC.

Same day discharge colon surgery: is it financially worth it?

AIM: Same day discharge (SDD) for colorectal surgery shows increasing promise in the era of enhanced recovery after surgery protocols and minimally invasive surgery. It has become increasingly relevant due to the constraints posed by the COVID-19 pandemic. The aim of this study was to compare SDD and postoperative day 1 (POD1) discharge to understand the clinical outcomes and financial impact on factors such as cost, charge, revenue, contribution margin and readmission. METHOD: A retrospective review of colectomies was performed at a single institution over a 2-year period (n = 143). Two populations were identified: SDD (n = 51) and POD1 (n = 92). Patients were selected by International Statistical Classification of Diseases and Related Health Problems-10 (ICD-10) and Diagnosis Related Grouper (DRG) codes. RESULTS: There was a statistically significant difference favouring SDD in total hospital cost (p < 0.0001), average direct costs (p < 0.0001) and average charges (p < 0.0016). SDD average hospital costs were $8699 (values in USD throughout) compared with $11 652 for POD 1 (p < 0.0001), and average SDD hospital charges were $85 506 compared with $97 008 for POD1 (p < 0.0016). The net revenue for SDD was $22 319 while for POD1 it was $26 173 (p = 0.14). Upon comparison of contribution margins (SDD $13 620 vs. POD1 $14 522), the difference was not statistically significant (p = 0.73). There were no identified statistically significant differences in operating room time, robotic console time, readmission rates or surgical complications. CONCLUSIONS: Amidst the pandemic-related constraints, we found that SDD was associated with lower hospital costs and comparable contribution margins compared with POD1. Additionally, the study was unable to identify any significant difference between operating time, readmissions, and surgical complications when performing SDD.

Impact of electronic information exchange on repeat imaging during 30-day readmissions among Medicare beneficiaries.

OBJECTIVES: We examined the association between electronic health information sharing and repeat imaging in readmissions among older adults with and without Alzheimer disease (AD). STUDY DESIGN: Cohort study using national Medicare data. METHODS: Among Medicare beneficiaries with 30-day readmissions in 2018, we examined repeat imaging on the same body system during the readmission. This was evaluated between fragmented and nonfragmented (same-hospital) readmissions and across categories of electronic information sharing via health information exchanges (HIEs) in fragmented readmissions: admission and readmission hospitals share the same HIE, admission and readmission hospitals participate in different HIEs, one or both do not participate in HIE, or HIE data missing. This relationship was evaluated using unadjusted and adjusted logistic regression. RESULTS: Overall, 14.3% of beneficiaries experienced repeat imaging during their readmission. Compared with nonfragmented readmissions, fragmented readmissions were associated with 5% higher odds of repeat imaging on the same body system in older adults without AD. This was not mitigated by the presence of electronic information sharing: Fragmented readmissions to hospitals that shared an HIE had 6% higher odds of repeat imaging (adjusted OR, 1.06; 95% CI, 1.00-1.13). There was no difference seen in the odds of repeat imaging for older adults with AD. CONCLUSIONS: Despite substantial investment, HIEs as currently deployed and used are not associated with decreased odds of repeat imaging in readmissions.

National Estimates of Short- and Longer-Term Hospital Readmissions After Major Surgery Among Community-Living Older Adults.

Importance: Nationally representative estimates of hospital readmissions within 30 and 180 days after major surgery, including both fee-for-service and Medicare Advantage beneficiaries, are lacking. Objectives: To provide population-based estimates of hospital readmission within 30 and 180 days after major surgery in community-living older US residents and examine whether these estimates differ according to key demographic, surgical, and geriatric characteristics. Design, Setting, and Participants: A prospective longitudinal cohort study of National Health and Aging Trends Study data (calendar years 2011-2018), linked to records from the Centers for Medicare & Medicaid Services (CMS). Data analysis was conducted from April to August 2023. Participants included community-living US residents of the contiguous US aged 65 years or older who had at least 1 major surgery from 2011 to 2018. Data analysis was conducted from April 10 to August 28, 2023. Main Outcomes and Measures: Major operations and hospital readmissions within 30 and 180 days were identified through data linkages with CMS files that included both fee-for-service and Medicare Advantage beneficiaries. Data on frailty and dementia were obtained from the annual National Health and Aging Trends Study assessments. Results: A total of 1780 major operations (representing 9 556 171 survey-weighted operations nationally) were identified from 1477 community-living participants; mean (SD) age was 79.5 (7.0) years, with 56% being female. The weighted rates of hospital readmission were 11.6% (95% CI, 9.8%-13.6%) for 30 days and 27.6% (95% CI, 24.7%-30.7%) for 180 days. The highest readmission rates within 180 days were observed among participants aged 90 years or older (36.8%; 95% CI, 28.3%-46.3%), those undergoing vascular surgery (45.8%; 95% CI, 37.7%-54.1%), and persons with frailty (36.9%; 95% CI, 30.8%-43.5%) or probable dementia (39.0%; 95% CI, 30.7%-48.1%). In age- and sex-adjusted models with death as a competing risk, the hazard ratios for hospital readmission within 180 days were 2.29 (95% CI, 1.70-3.09) for frailty and 1.58 (95% CI, 1.15-2.18) for probable dementia. Conclusions and Relevance: In this nationally representative cohort study of community-living older US residents, the likelihood of hospital readmissions within 180 days after major surgery was increased among older persons who were frail or had probable dementia, highlighting the potential value of these geriatric conditions in identifying those at increased risk.

The association of the Sequential Organ Failure Assessment score at intensive care unit discharge with intensive care unit readmission in the cardiac intensive care unit.

AIMS: Unplanned intensive care unit (ICU) readmissions contribute to increased morbidity, mortality, and healthcare costs. The severity of patient illness at ICU discharge may predict early ICU readmission. Thus, in this study, we investigated the association of cardiac ICU (CICU) discharge Sequential Organ Failure Assessment (SOFA) score with unplanned CICU readmission in patients admitted to the CICU. METHODS AND RESULTS: We retrospectively reviewed the hospital medical records of 4659 patients who were admitted to the CICU from 2012 to 18. Sequential Organ Failure Assessment scores at CICU admission and discharge were obtained. The predictive performance of organ failure scoring was evaluated by using area under the receiver operating characteristic (AUROC) curves. The primary outcome was unplanned CICU readmission. Of the 3949 patients successfully discharged from the CICU, 184 (4.7%) had an unplanned CICU readmission or they experienced a deteriorated condition but died without being readmitted to the CICU (readmission group). The readmission group had significantly higher rates of organ failure in all organ systems at both CICU admission and discharge than the non-readmission group. The AUROC of the discharge SOFA score for CICU readmission was 0.731, showing good predictive performance. The AUROC of the discharge SOFA score was significantly greater than that of either the initial SOFA score (P = 0.020) or the Acute Physiology and Chronic Health Evaluation II score (P < 0.001). In the multivariable regression analysis, SOFA score, overweight or obese status, history of heart failure, and acute heart failure as reasons for ICU admission were independent predictors of unplanned ICU readmission during the same hospital stay. CONCLUSION: The discharge SOFA score may identify patients at a higher risk of unplanned CICU readmission, enabling targeted interventions to reduce readmission rates and improve patient outcomes.

Evaluating Algorithmic Bias in 30-Day Hospital Readmission Models: Retrospective Analysis.

BACKGROUND: The adoption of predictive algorithms in health care comes with the potential for algorithmic bias, which could exacerbate existing disparities. Fairness metrics have been proposed to measure algorithmic bias, but their application to real-world tasks is limited. OBJECTIVE: This study aims to evaluate the algorithmic bias associated with the application of common 30-day hospital readmission models and assess the usefulness and interpretability of selected fairness metrics. METHODS: We used 10.6 million adult inpatient discharges from Maryland and Florida from 2016 to 2019 in this retrospective study. Models predicting 30-day hospital readmissions were evaluated: LACE Index, modified HOSPITAL score, and modified Centers for Medicare & Medicaid Services (CMS) readmission measure, which were applied as-is (using existing coefficients) and retrained (recalibrated with 50% of the data). Predictive performances and bias measures were evaluated for all, between Black and White populations, and between low- and other-income groups. Bias measures included the parity of false negative rate (FNR), false positive rate (FPR), 0-1 loss, and generalized entropy index. Racial bias represented by FNR and FPR differences was stratified to explore shifts in algorithmic bias in different populations. RESULTS: The retrained CMS model demonstrated the best predictive performance (area under the curve: 0.74 in Maryland and 0.68-0.70 in Florida), and the modified HOSPITAL score demonstrated the best calibration (Brier score: 0.16-0.19 in Maryland and 0.19-0.21 in Florida). Calibration was better in White (compared to Black) populations and other-income (compared to low-income) groups, and the area under the curve was higher or similar in the Black (compared to White) populations. The retrained CMS and modified HOSPITAL score had the lowest racial and income bias in Maryland. In Florida, both of these models overall had the lowest income bias and the modified HOSPITAL score showed the lowest racial bias. In both states, the White and higher-income populations showed a higher FNR, while the Black and low-income populations resulted in a higher FPR and a higher 0-1 loss. When stratified by hospital and population composition, these models demonstrated heterogeneous algorithmic bias in different contexts and populations. CONCLUSIONS: Caution must be taken when interpreting fairness measures' face value. A higher FNR or FPR could potentially reflect missed opportunities or wasted resources, but these measures could also reflect health care use patterns and gaps in care. Simply relying on the statistical notions of bias could obscure or underplay the causes of health disparity. The imperfect health data, analytic frameworks, and the underlying health systems must be carefully considered. Fairness measures can serve as a useful routine assessment to detect disparate model performances but are insufficient to inform mechanisms or policy changes. However, such an assessment is an important first step toward data-driven improvement to address existing health disparities.

Hospital Variation in Skilled Nursing Facility Use After Coronary Artery Bypass Graft Surgery.

BACKGROUND: Over 20% of patients are discharged to a skilled nursing facility (SNF) after coronary artery bypass graft surgery, but little is known about specific drivers for postdischarge SNF use. The purpose of this study was to evaluate hospital variation in SNF use and its association with postoperative outcomes after coronary artery bypass graft. METHODS AND RESULTS: A retrospective study design utilizing Medicare Provider Analysis and Review files was used to evaluate SNF use among 70 509 beneficiaries undergoing coronary artery bypass graft, with or without valve procedures, between 2016 and 2018. A total of 17 328 (24.6%) were discharged to a SNF, ranging from 0% to 88% across 871 hospitals. Multilevel logistic regression models identified significant patient-level predictors of discharge to SNF including increasing age, comorbidities, female sex, Black race, dual eligibility, and postoperative complications. After adjusting for patient and hospital factors, 15.6% of the variation in hospital SNF use was attributed to the discharging hospital. Compared with the lower quartile of hospital SNF use, hospitals in the top quartile of SNF use had lower risk-adjusted 1-year mortality (12.5% versus 8.6%, P<0.001) and readmission (59.9% versus 49.8%, P<0.001) rates for patients discharged to a SNF. CONCLUSIONS: There is high variability in SNF use among hospitals that is only partially explained by patient characteristics. Hospitals with higher SNF utilization had lower risk-adjusted 1-year mortality and readmission rates for patients discharged to a SNF. More work is needed to better understand underlying provider and hospital-level factors contributing to SNF use variability.

Achieving equity: patient demographics and outcomes after surgical and non-surgical procedures in South Australia, 2022.

BACKGROUND: Although modern Australian healthcare systems provide patient-centred care, the ability to predict and prevent suboptimal post-procedural outcomes based on patient demographics at admission may improve health equity. This study aimed to identify patient demographic characteristics that might predict disparities in mortality, readmission, and discharge outcomes after either an operative or non-operative procedural hospital admission. METHODS: This retrospective cohort study included all surgical and non-surgical procedural admissions at three of the four major metropolitan public hospitals in South Australia in 2022. Multivariable logistic regression, with backwards selection, evaluated association between patient demographic characteristics and outcomes up to 90 days post-procedurally. RESULTS: 40 882 admissions were included. Increased likelihood of all-cause, post-procedure mortality in-hospital, at 30 days, and 90 days, were significantly associated with increased age (P < 0.001), increased comorbidity burden (P < 0.001), an emergency admission (P < 0.001), and male sex (P = 0.046, P = 0.03, P < 0.001, respectively). Identification as ATSI (P < 0.001) and being born in Australia (P = 0.03, P = 0.001, respectively) were associated with an increased likelihood of 30-day hospital readmission and decreased likelihood of discharge directly home, as was increased comorbidity burden (P < 0.001) and emergency admission (P < 0.001). Being married (P < 0.001) and male sex (P = 0.003) were predictive of an increased likelihood of discharging directly home; in contrast to increased age (P < 0.001) which was predictive of decreased likelihood of this occurring. CONCLUSIONS: This study characterized several associations between patient demographic factors present on admission and outcomes after surgical and non-surgical procedures, that can be integrated within patient flow pathways through the Australian healthcare system to improve healthcare equity.

Effects of Social Determinants and Pharmacogenetic Medication Interactions on 90-Day Hospital Readmissions.

Context: The present study builds on our prior work that demonstrated an association between pharmacogenetic interactions and 90-day readmission. Objective: Evaluate aggregate contribution of social determinants, comorbidity, and gene-x-drug interactions to moderate 90-day hospital readmission. Study Design and Analysis: Non-concurrent cohort study; Multivariable logistic regression Setting: Hospital/integrated healthcare delivery system in northern Illinois Population Studied: 19,999 adults tracked from 2010 through 2020 who underwent testing with a 13-gene pharmacogenetic panel Outcome Measure: 90-day hospital readmission (primary outcome) Results: Univariate logistic regression analyses demonstrated that strongest associations with 90 day hospital readmissions were the number of medications prescribed within 30 days of a first hospital admission that had Clinical Pharmacogenomics Implementation Consortium (CPIC) guidance (CPIC medications) (5+ CPIC medications, odds ratio (OR) = 7.66, 95% confidence interval 5.45-10.77) (p < 0.0001), major comorbidities (5+ comorbidities, OR 3.36, 2.61-4.32) (p < 0.0001), age (65 + years, OR = 2.35, 1.77-3.12) (p < 0.0001), unemployment (OR = 2.19, 1.88-2.64) (p < 0.0001), Black/African-American race (OR 2.12, 1.47-3.07) (p < 0.0001), median household income (OR = 1.63, 1.03-2.58) (p = 0.035), male gender (OR = 1.47, 1.21-1.80) (p = 0.0001), and one or more gene-x-drug interaction (defined as a prescribed CPIC medication for a patient with a corresponding actionable pharmacogenetic variant) (OR = 1.41, 1.18-1.70). Health insurance was not associated with risk of 90-day readmission. Race, income, employment status, and gene-x-drug interactions were robust in a multivariable logistic regression model. The odds of 90-day readmission for patients with one or more identified gene-x-drug interactions after adjustment for these covariates was attenuated by 10% (OR = 1.31, 1.08-1.59) (p = 0.006). Although the interaction between race and gene-x-drug interactions was not statistically significant, White patients were more likely to have a gene-x-drug interaction (35.2%) than Black/African-American patients (25.9%) who were not readmitted (p < 0.0001). Conclusions: These results highlight the major contribution of social determinants and medical complexity to risk for hospital readmission, and that these determinants may modify the effect of gene-x-drug interactions on rehospitalization risk.