Abstract WP338: The Impact of Social Determinants of Health on 30-Day Readmission After Ischemic Stroke

Objective: Readmission after ischemic stroke presents immense social and financial burden on the patient, family and society. The impact of social determinants of health on post-stroke readmission have not been well studied. This study aimed to examine association of patient social determinants with 30-day readmission risk after ischemic stroke. Methods: We examined patients who were hospitalized for acute ischemic stroke (ICD-9 codes 433.1, 434, and 436) in the state of Florida and New York from February 2012 to November 2013, with a 30-day run-in time and 30-day follow up time. Data were obtained from the State Inpatient Database which was then linked with Census data on social determinants at zip code level. Multivariate logistic regression models were generated to study the association of patient factors with 30-day readmission, after adjusting for patient characteristics, in-hospital infection, hemiplegia, and 16 comorbidities. All statistical analyses were conducted using SAS Version 9.4 software (SAS Institute). Results: A total of 127,290 patients were included in the study. The overall 30-day readmission rate was 23%. The 30-day readmission rates differed by race, insurance but not age, sex or household income at zip code level. The 30-day readmission rate was higher amongst black (27%) and Hispanics (25%), and lower in Native American (17%) and White (22%). Comparable to older age groups, patients who suffered a stroke at young age (<50 years old) had 23% readmission rates at 30 days. In the multivariate logistic regression model, age was not a risk factor for readmission [adjusted odds ratio (OR) 1.0, 95% confidence interval (CI) 1.0-1.0]. Compared to white, black race had higher risk of 30-day readmission (OR1.2, 95% CI 1.1-1.2), as well as Hispanic race (OR1.1, 95% CI 1.1-1.2). Compared to Medicare insurance, patient on Medicaid had higher risk of 30-day readmission (OR 1.1, 95% CI 1.1-1.2). Self-pay patients had lower risk of readmission (OR 0.6, 95% CI 0.5-0.7). Conclusion: Race and medical insurance, not age, sex, or household income, has significant influence in30-day readmission risk after ischemic stroke. This will allow further targeted intervention for readmission reduction.

Download Full-text

Common sampling and modeling approaches to analyzing readmission risk that ignore clustering produce misleading results

10.21203/rs.3.rs-27550/v1 ◽

2020 ◽

Author(s):

Huaqing Zhao ◽

Samuel Tanner ◽

Sherita H. Golden ◽

Susan Fisher ◽

Daniel J. Rubin

Keyword(s):

Logistic Regression ◽

Readmission Rate ◽

Medical Center ◽

Academic Medical Center ◽

Model Performance ◽

Readmission Rates ◽

Modeling Approaches ◽

Multiple Variables ◽

Highly Correlated ◽

Readmission Risk

Abstract Background There is little consensus on how to sample hospitalizations and analyze multiple variables to model readmission risk. The purpose of this study was to compare readmission rates and the accuracy of predictive models based on different sampling and multivariable modeling approaches. Methods We conducted a retrospective cohort study of 17 284 adult diabetes patients with 44 203 discharges from an urban academic medical center between 1/1/2004 and 12/31/2012. Models for all-cause 30-day readmission were developed by four strategies: logistic regression using the first discharge per patient (LR-first), logistic regression using all discharges (LR-all), generalized estimating equations (GEE) using all discharges, and cluster-weighted (CW) GEE using all discharges. Multiple sets of models were developed and internally validated across a range of sample sizes. Results The readmission rate was 10.2% among first discharges and 20.3% among all discharges, revealing that sampling only first discharges underestimates a population’s readmission rate. Number of discharges was highly correlated with number of readmissions (r = 0.87, P < 0.001). Accounting for clustering with GEE and CWGEE yielded more conservative estimates of model performance than LR-all. LR-first produced falsely optimistic Brier scores. Model performance was unstable below samples of 6 000–8 000 discharges and stable in larger samples. GEE and CWGEE performed better in larger samples than in smaller samples. Conclusions Hospital readmission risk models should be based on all discharges as opposed to just the first discharge per patient and utilize methods that account for clustered data.

Download Full-text

Common sampling and modeling approaches to analyzing readmission risk that ignore clustering produce misleading results

10.21203/rs.3.rs-27550/v2 ◽

2020 ◽

Author(s):

Huaqing Zhao ◽

Samuel Tanner ◽

Sherita H. Golden ◽

Susan Fisher ◽

Daniel J. Rubin

Keyword(s):

Logistic Regression ◽

Readmission Rate ◽

Medical Center ◽

Academic Medical Center ◽

Model Performance ◽

Readmission Rates ◽

Modeling Approaches ◽

Multiple Variables ◽

Highly Correlated ◽

Readmission Risk

Abstract Background: There is little consensus on how to sample hospitalizations and analyze multiple variables to model readmission risk. The purpose of this study was to compare readmission rates and the accuracy of predictive models based on different sampling and multivariable modeling approaches. Methods: We conducted a retrospective cohort study of 17 284 adult diabetes patients with 44 203 discharges from an urban academic medical center between 1/1/2004 and 12/31/2012. Models for all-cause 30-day readmission were developed by four strategies: logistic regression using the first discharge per patient (LR-first), logistic regression using all discharges (LR-all), generalized estimating equations (GEE) using all discharges, and cluster-weighted (CW) GEE using all discharges. Multiple sets of models were developed and internally validated across a range of sample sizes. Results: The readmission rate was 10.2% among first discharges and 20.3% among all discharges, revealing that sampling only first discharges underestimates a population’s readmission rate. Number of discharges was highly correlated with number of readmissions (r=0.87, P<0.001). Accounting for clustering with GEE and CWGEE yielded more conservative estimates of model performance than LR-all. LR-first produced falsely optimistic Brier scores. Model performance was unstable below samples of 6 000-8 000 discharges and stable in larger samples. GEE and CWGEE performed better in larger samples than in smaller samples. Conclusions: Hospital readmission risk models should be based on all discharges as opposed to just the first discharge per patient and utilize methods that account for clustered data.

Download Full-text

Ceramic-on-Polyethylene Bearing Usage in Primary THA Is Associated with Reduced Readmission Risk for the Medicare Population

The Journal of Hip Surgery ◽

10.1055/s-0038-1642056 ◽

2018 ◽

Vol 02 (01) ◽

pp. 022-032 ◽

Cited By ~ 1

Author(s):

Edmund Lau ◽

Doruk Baykal ◽

Bryan Springer ◽

Steven Kurtz

Keyword(s):

Readmission Rate ◽

Propensity Scores ◽

Cox Regression ◽

Readmission Rates ◽

Related Factors ◽

Propensity Score Stratification ◽

The Impact ◽

Strength Of Association ◽

Readmission Risk ◽

Primary Total Hip Arthroplasty

AbstractThe authors hypothesized that unplanned readmissions, which are often caused by infections and dislocation, may be reduced with ceramic bearing usage. They also sought to confirm that the readmission rates for ceramic bearings were associated with the year of surgery. They identified 245,077 elderly patients (65+) who underwent primary total hip arthroplasty (THA) between 2010 and 2015 with known bearing types (ceramic-on-polyethylene [C-PE] ceramic-on-ceramic [COC], and metal-on-polyethylene [M-PE]) from the Medicare 100% inpatient database. Outcomes included relative risk of 30- and 90-day readmission. Propensity scores were developed to adjust for selection bias in the choice of bearing type at index surgery. Cox regression incorporating propensity score stratification (10 levels) was used to evaluate the impact of bearing selection on outcomes, after adjusting for patient-, hospital-, surgeon-related factors, as well as the year of surgery. With C-PE bearings, the unadjusted (crude) 90-day readmission rate decreased from 8.7% in 2010 to 8.3% in 2015. For COC bearings, the crude 90-day readmission rate decreased from 10.5 to 9.1% from 2010 to 2015. After adjustment, year of surgery was associated with reduced readmission risk for both types of ceramic bearings in 30-day readmissions (p < 0.05) and COC in 90-day readmissions (p < 0.001). The authors also found that C-PE bearings were associated with significantly reduced readmission risk relative to M-PE at 30 days (hazard ratio [HR]: 0.91, p < 0.001) and 90 days (HR: 0.93, p < 0.001). In terms of strength of association with 90-day readmission, however, it was ranked the ninth most associated independent factor. To the authors' knowledge, this is the first study to demonstrate an association between THA implant characteristics (in this case C-PE bearing usage) and reduced readmission rates in this context along with patient- and clinical-related factors. The readmission rates for COC were found to be comparable to M-PE.

Download Full-text

The Impact of the Duration of Antibiotic Therapy in Patients With Community-Onset Pneumonia on Readmission Rates: A Retrospective Cohort Study

Journal of Pharmacy Practice ◽

10.1177/0897190019882260 ◽

2019 ◽

pp. 089719001988226

Author(s):

Daniel M. Parshall ◽

Julia E. Sessa ◽

Kelly M. Conn ◽

Lisa M. Avery

Keyword(s):

Risk Factors ◽

Logistic Regression ◽

Antibiotic Therapy ◽

Retrospective Cohort ◽

Readmission Rate ◽

Multivariate Logistic Regression ◽

Readmission Rates ◽

Antibiotic Duration ◽

The Relationship ◽

Community Onset

Background: Recent publications have confirmed that 70% of hospitalized adults with uncomplicated community-acquired pneumonia and health-care-associated pneumonia are prescribed a duration therapy that exceeds current guideline recommendations. Objective: The primary objective is to evaluate the relationship between antibiotic duration and all-cause 30-day readmission rates. Secondary outcomes include pneumonia-specific 30-day readmission rate and identification of risk factors for readmission. Methods: Patients aged ≥18 years with a primary diagnosis of pneumonia from January 1, 2016, to December 31, 2016, were included in this single-center, retrospective cohort study. Patients were categorized by antibiotic therapy duration of ≤7 days (n = 139) or >7 days (n = 286), and outcomes were analyzed in both bivariate and multivariate models. A multivariate logistic regression was used to assess the relationship between all-cause 30-day readmission and antibiotic days. Results: Baseline characteristics were not significantly different between the 2 groups. All-cause 30-day readmission rates were 15.8% and 15.5% for patients who received ≤7 days versus >7 days of antibiotics, respectively ( P = .95). Pneumonia-specific 30-day readmission occurred in 3.6% of patients who received antibiotics for ≤7 days compared to 3.5% of patients who received antibiotics for >7 days ( P = .95). Multivariate logistic regression showed no statistically significant association between readmission rate and antibiotic duration of >7 days. Statistically significant risk factors for readmission identified by logistic regression include ≥3 hospital admissions within the previous year, a hematocrit <30% at discharge, a history of chronic obstructive pulmonary disorder (COPD), and weight. Conclusion: The use of prolonged antibiotic therapy for the treatment of community-onset pneumonia was not associated with a decrease in all-cause or pneumonia-specific 30-day readmission rates.

Download Full-text

Common sampling and modeling approaches to analyzing readmission risk that ignore clustering produce misleading results

BMC Medical Research Methodology ◽

10.1186/s12874-020-01162-0 ◽

2020 ◽

Vol 20 (1) ◽

Author(s):

Huaqing Zhao ◽

Samuel Tanner ◽

Sherita H. Golden ◽

Susan G. Fisher ◽

Daniel J. Rubin

Keyword(s):

Logistic Regression ◽

Readmission Rate ◽

Medical Center ◽

Academic Medical Center ◽

Model Performance ◽

Readmission Rates ◽

Modeling Approaches ◽

Multiple Variables ◽

Highly Correlated ◽

Readmission Risk

Abstract Background There is little consensus on how to sample hospitalizations and analyze multiple variables to model readmission risk. The purpose of this study was to compare readmission rates and the accuracy of predictive models based on different sampling and multivariable modeling approaches. Methods We conducted a retrospective cohort study of 17,284 adult diabetes patients with 44,203 discharges from an urban academic medical center between 1/1/2004 and 12/31/2012. Models for all-cause 30-day readmission were developed by four strategies: logistic regression using the first discharge per patient (LR-first), logistic regression using all discharges (LR-all), generalized estimating equations (GEE) using all discharges, and cluster-weighted (CWGEE) using all discharges. Multiple sets of models were developed and internally validated across a range of sample sizes. Results The readmission rate was 10.2% among first discharges and 20.3% among all discharges, revealing that sampling only first discharges underestimates a population’s readmission rate. Number of discharges was highly correlated with number of readmissions (r = 0.87, P < 0.001). Accounting for clustering with GEE and CWGEE yielded more conservative estimates of model performance than LR-all. LR-first produced falsely optimistic Brier scores. Model performance was unstable below samples of 6000–8000 discharges and stable in larger samples. GEE and CWGEE performed better in larger samples than in smaller samples. Conclusions Hospital readmission risk models should be based on all discharges as opposed to just the first discharge per patient and utilize methods that account for clustered data.

Download Full-text

Impact of Michigan’s new opioid prescribing laws on spine surgery patients: analysis of the Michigan Spine Surgery Improvement Collaborative

Journal of Neurosurgery Spine ◽

10.3171/2020.7.spine20729 ◽

2020 ◽

pp. 1-6

Author(s):

Paul Park ◽

Victor Chang ◽

Hsueh-Han Yeh ◽

Jason M. Schwalb ◽

David R. Nerenz ◽

...

Keyword(s):

Spine Surgery ◽

Spinal Surgery ◽

Readmission Rate ◽

Rating Scale ◽

Short Term ◽

Readmission Rates ◽

Opioid Prescribing ◽

Number Of Patients ◽

Before And After ◽

The Impact

OBJECTIVEIn 2017, Michigan passed new legislation designed to reduce opioid abuse. This study evaluated the impact of these new restrictive laws on preoperative narcotic use, short-term outcomes, and readmission rates after spinal surgery.METHODSPatient data from 1 year before and 1 year after initiation of the new opioid laws (beginning July 1, 2018) were queried from the Michigan Spine Surgery Improvement Collaborative database. Before and after implementation of the major elements of the new laws, 12,325 and 11,988 patients, respectively, were treated.RESULTSPatients before and after passage of the opioid laws had generally similar demographic and surgical characteristics. Notably, after passage of the opioid laws, the number of patients taking daily narcotics preoperatively decreased from 3783 (48.7%) to 2698 (39.7%; p < 0.0001). Three months postoperatively, there were no differences in minimum clinically important difference (56.0% vs 58.0%, p = 0.1068), numeric rating scale (NRS) score of back pain (3.5 vs 3.4, p = 0.1156), NRS score of leg pain (2.7 vs 2.7, p = 0.3595), satisfaction (84.4% vs 84.7%, p = 0.6852), or 90-day readmission rate (5.8% vs 6.2%, p = 0.3202) between groups. Although there was no difference in readmission rates, pain as a reason for readmission was marginally more common (0.86% vs 1.22%, p = 0.0323).CONCLUSIONSThere was a meaningful decrease in preoperative narcotic use, but notably there was no apparent negative impact on postoperative recovery, patient satisfaction, or short-term outcomes after spinal surgery despite more restrictive opioid prescribing. Although the readmission rate did not significantly increase, pain as a reason for readmission was marginally more frequently observed.

Download Full-text

High Risk of One-year Stroke Recurrence in Patients with Younger Age and Prior History of Ischemic Stroke

Current Neurovascular Research ◽

10.2174/1567202616666190618164528 ◽

2019 ◽

Vol 16 (3) ◽

pp. 250-257 ◽

Cited By ~ 1

Author(s):

Jiann-Der Lee ◽

Ya-Han Hu ◽

Meng Lee ◽

Yen-Chu Huang ◽

Ya-Wen Kuo ◽

...

Keyword(s):

Logistic Regression ◽

Ischemic Stroke ◽

Multivariate Logistic Regression Analysis ◽

Stroke Recurrence ◽

Prior History ◽

Multivariate Logistic Regression ◽

Cart Analysis ◽

Younger Age ◽

History Of ◽

One Year

Background and Purpose: Recurrent ischemic strokes increase the risk of disability and mortality. The role of conventional risk factors in recurrent strokes may change due to increased awareness of prevention strategies. The aim of this study was to explore the potential risk factors besides conventional ones which may help to affect the advances in future preventive concepts associated with one-year stroke recurrence (OSR). Methods: We analyzed 6,632 adult patients with ischemic stroke. Differences in clinical characteristics between patients with and without OSR were analyzed using multivariate logistic regression and classification and regression tree (CART) analyses. Results: Among the study population, 525 patients (7.9%) had OSR. Multivariate logistic regression analysis revealed that male sex (OR 1.243, 95% CI 1.025 – 1.506), age (OR 1.015, 95% CI 1.007 - 1.023), and a prior history of ischemic stroke (OR 1.331, 95% CI 1.096 – 1.615) were major factors associated with OSR. CART analysis further identified age and a prior history of ischemic stroke were important factors for OSR when classified the patients into three subgroups (with risks of OSR of 8.8%, 3.8%, and 12.5% for patients aged > 57.5 years, ≤ 57.5 years/with no prior history of ischemic stroke, and ≤ 57.5 years/with a prior history of ischemic stroke, respectively). Conclusions: Male sex, age, and a prior history of ischemic stroke could increase the risk of OSR by multivariate logistic regression analysis, and CART analysis further demonstrated that patients with a younger age (≤ 57.5 years) and a prior history of ischemic stroke had the highest risk of OSR.

Download Full-text

The second pregnancy has no effect in the incidence of macrosomia: a cross-sectional survey in two western Chinese regions

Journal of Health Population and Nutrition ◽

10.1186/s41043-021-00244-z ◽

2021 ◽

Vol 40 (1) ◽

Author(s):

Li Luo ◽

Huan Zeng ◽

Mao Zeng ◽

Xueqing Liu ◽

Xianglong Xu ◽

...

Keyword(s):

Logistic Regression ◽

Western China ◽

Fetal Macrosomia ◽

Multivariate Logistic Regression Model ◽

Multivariate Logistic Regression ◽

Cross Sectional ◽

Second Child ◽

Child Policy ◽

Significant Difference ◽

The Impact

Abstract Background After the implementation of the universal two-child policy in China, the increase in parity has led to an increase in adverse pregnancy outcomes. The impact of one and two fetuses on the incidence of fetal macrosomia has not been fully confirmed in China. This study aimed to explore the differences in the incidence of fetal macrosomia in first and second pregnancies in Western China after the implementation of the universal two-child policy. Methods A total of 1598 pregnant women from three hospitals were investigated by means of a cross-sectional study from August 2017 to January 2018. Participants were recruited by convenience and divided into first and second pregnancy groups. These groups included 1094 primiparas and 504 women giving birth to their second child. Univariate and multivariate logistic regression analyses were performed to discuss the differences in the incidence of fetal macrosomia in first and second pregnancies. Results No significant difference was found in the incidence of macrosomia in the first pregnancy group (7.2%) and the second pregnancy group (7.1%). In the second-time pregnant mothers, no significant association was found between the macrosomia of the second child (5.5%) and that of the first child (4.7%). The multivariate logistic regression model showed that mothers older than 30 years are not likely to give birth to children with macrosomia (odds ratio (OR) 0.6, 95% confidence interval (CI) 0.4,0.9). Conclusions The incidence of macrosomia in Western China is might not be affected by the birth of the second child and is not increased by low parity.

Download Full-text

Impact of oral anticoagulants on 30-day readmission: a study from a single academic centre

BMJ evidence-based medicine ◽

10.1136/bmjebm-2018-111026 ◽

2018 ◽

Vol 24 (1) ◽

pp. 10-14 ◽

Cited By ~ 3

Author(s):

Mukul Bhattarai ◽

Tamer Hudali ◽

Robert Robinson ◽

Mohammad Al-Akchar ◽

Carrie Vogler ◽

...

Keyword(s):

Logistic Regression ◽

Risk Factor ◽

Length Of Stay ◽

Hospital Readmission ◽

Readmission Rate ◽

Warfarin Therapy ◽

Oral Anticoagulants ◽

Emergency Department Visits ◽

Day Hospital ◽

The Impact

Researchers are extensively searching for modifiable risk factors including high-risk medications such as anticoagulation to avoid rehospitalisation. The influence of oral anticoagulant therapy on hospital readmission is not known. We investigated the impact of warfarin and direct oral anticoagulants (DOACs) on all cause 30-day hospital readmission retrospectively in an academic centre. We study the eligible cohort of 1781 discharges over 2-year period. Data on age, gender, diagnoses, 30-day hospital readmission, discharge medications and variables in the HOSPITAL score (Haemoglobin level at discharge, Oncology at discharge, Sodium level at discharge, Procedure during hospitalisation, Index admission, number of hospital Admissions, Length of stay) and LACE index (Length of stay, Acute/emergent admission, Charlson comorbidity index score, Emergency department visits in previous 6 months), which have higher predictability for readmission were extracted and matched for analysis. Warfarin was the most common anticoagulant prescribed at discharge (273 patients) with a readmission rate of 20% (p<0.01). DOACs were used by 94 patients at discharge with a readmission rate of 4% (p=0.219). Multivariate logistic regression showed an increased risk of readmission with warfarin therapy (OR 1.36, p=0.045). Logistic regression did not show DOACs to be a risk factor for hospital readmission. Our data suggests that warfarin therapy is a risk factor for all-cause 30-day hospital readmission. DOAC therapy is not found to be associated with a higher risk of hospital readmission. Warfarin anticoagulation may be an important target for interventions to reduce hospital readmissions.

Download Full-text

Association Between Diagnosis Code Expansion and Changes in 30‐Day Risk‐Adjusted Outcomes for Cardiovascular Diseases

Journal of the American Heart Association ◽

10.1161/jaha.120.020668 ◽

2021 ◽

Author(s):

Lauren Gilstrap ◽

Rishi K. Wadhera ◽

Andrea M. Austin ◽

Stephen Kearing ◽

Karen E. Joynt Maddox ◽

...

Keyword(s):

Logistic Regression ◽

International Classification Of Diseases ◽

Fee For Service ◽

Readmission Rates ◽

Difference In Differences ◽

Diagnosis Codes ◽

Classification Of Diseases ◽

Changes In Risk ◽

The Impact

BACKGROUND In January 2011, Centers for Medicare and Medicaid Services expanded the number of inpatient diagnosis codes from 9 to 25, which may influence comorbidity counts and risk‐adjusted outcome rates for studies spanning January 2011. This study examines the association between (1) limiting versus not limiting diagnosis codes after 2011, (2) using inpatient‐only versus inpatient and outpatient data, and (3) using logistic regression versus the Centers for Medicare and Medicaid Services risk‐standardized methodology and changes in risk‐adjusted outcomes. METHODS AND RESULTS Using 100% Medicare inpatient and outpatient files between January 2009 and December 2013, we created 2 cohorts of fee‐for‐service beneficiaries aged ≥65 years. The acute myocardial infarction cohort and the heart failure cohort had 578 728 and 1 595 069 hospitalizations, respectively. We calculate comorbidities using (1) inpatient‐only limited diagnoses, (2) inpatient‐only unlimited diagnoses, (3) inpatient and outpatient limited diagnoses, and (4) inpatient and outpatient unlimited diagnoses. Across both cohorts, International Classification of Diseases, Ninth Revision ( ICD‐9 ) diagnoses and hierarchical condition categories increased after 2011. When outpatient data were included, there were no significant differences in risk‐adjusted readmission rates using logistic regression or the Centers for Medicare and Medicaid Services risk standardization. A difference‐in‐differences analysis of risk‐adjusted readmission trends before versus after 2011 found that no significant differences between limited and unlimited models for either cohort. CONCLUSIONS For studies that span 2011, researchers should consider limiting the number of inpatient diagnosis codes to 9 and/or including outpatient data to minimize the impact of the code expansion on comorbidity counts. However, the 2011 code expansion does not appear to significantly affect risk‐adjusted readmission rate estimates using either logistic or risk‐standardization models or when using or excluding outpatient data.

Download Full-text