Comparison of 30-day Readmission Rates and Inpatient Cardiac Procedures for Weekday versus Weekend Hospital Admissions for Heart Failure (2024)

The publisher's final edited version of this article is available at J Card Fail

Design and data Source:

This was a retrospective analysis involving adult hospitalizations for HF in the US from 2010–2019. We extracted the data from National Readmission Database (NRD). The NRD is a part of Healthcare Cost and Utilization Project (HCUP). It is one of the largest publicly available databases including patients from all-payer sources and is sponsored by Agency for Healthcare Research and Quality (AHRQ). The database contains verified deidentified patient linkage numbers that can be used to track a person across hospitals within a State. Hospitals are stratified according to ownership control, number of beds, teaching status, and metropolitan/non-metropolitan location. The NRD contains discharge data from geographically dispersed states within a calendar year. In 2019, NRD was expanded to include 30 U.S states accounting for 61.8% of the total U.S resident population and 60.4% of all U.S hospitalizations. The NRD contains a weighted sample of hospitalizations, and this can be used to derive national estimates. (14) The study was exempted from institutional board review approval as NRD database contains deidentified patient information.

Study population

Eligible patients for this study included US adults age ≥18 years who were admitted with primary diagnosis of HF between 2010 and 2019. Admissions were excluded as an index admission if the hospitalization was elective, had missing data for age, sex or in-hospital mortality, if the patient died during the hospital stay or was transferred to another acute care hospital. In the NRD, patient identifiers cannot be linked across the years, hence patients who had an index hospitalization on December 1 or later in any given year were excluded (Supplemental Figure 1). Time to readmission was calculated by subtracting length of stay of index admission from time between index admission and the readmission. Patients who had at least one readmission within 30 days were included. Planned readmission within 30-days of discharge were excluded.

We grouped patients into those with an admission date on a weekday (Monday–Friday) versus on a weekend (Saturday–Sunday), using the day of the index hospitalization (determined using the ‘AWEEKEND’ variable in NRD). Location of the hospital as urban or rural was determined by HOSP_URCAT4 variable in the NRD. The urban and rural categorization in the NRD is a simplified adaptation of the Urban Influence Codes (UIC). Starting in the 2014 data, the categorization is based on the 2013 version of the UIC. Prior to 2014, the categorization is based on the 2003 version of the UIC.

Variables

Patient demographics including age, sex, primary insurance and median neighborhood household income (income quartiles were identified referred to patients as 1-lowincome, 2-middle income, 3-upper middle income and 4-high income) using the NRD variables. In 2019, quartile 1 reflected household income: ≤$47,999; quartile 2: $48,000–$60,999; quartile 3: $61,000–$81,999; quartile 4: ≥$82,000. (14) We included hospital-specific variables including hospital bed size, hospital teaching status and location. Comorbidities were identified using diagnoses codes from the ICD Ninth Revision and ICD Tenth revision respective to years that were used in the NRD (Supplemental Table 1). The co-morbid conditions were present prior to the hospitalization and were not primary cause of admission. We used Charlson Comorbidity Index (CCI) to assess the severity of comorbid conditions.

Outcome measures

In the primary analysis unplanned (ie, non-elective) readmission occurring within the first 30 days of discharge from the index hospitalization was assessed. If an index hospitalization had more than one readmission within 30 days, we only included only the first readmission. A secondary analysis assessed the relationship between weekday versus weekend hospital admissions for HF and processes of care and cardiovascular testing that occurred during the index HF hospitalization. Pre-specified processes of care and cardiovascular testing that were examined included echocardiography, right heart catheterization, electrical cardioversion, and temporary mechanical support devices (intra-aortic balloon pump, Impella assist device, extracorporeal membrane oxygenation).

Statistical analysis

As per specific Healthcare Cost and Utilization Project recommendations, we performed all our analysis utilizing the Healthcare Cost and Utilization Project STATA survey data analysis packages which incorporate the NRD specific variables including hospital identifiers, stratum, and discharge weights to account for clustering and large survey-weighted data analysis to obtain statistical and variance calculations independent of individual hospital discharge characteristics. The student t-test and the chi-squared test were used to compare continuous and categorical variables, respectively.

Multivariable Logistic regression was used to estimate ORs for 30-day readmission and the processes of care across weekend versus weekday (reference) hospitalizations. Multivariable Poisson regression analysis was performed for yearly trends with year as the independent variable and readmission as the dependent variable. All the outcomes were adjusted for age, sex, hospital characteristics (location, bed-size), income, insurance and Charlson comorbidity index. Adjusted readmission rates were obtained using marginal effects following multivariable regression analysis. Trends in readmission rate between groups (HF weekday or HF weekend admission) were compared using a group-year interaction term in a multivariable regression analysis with readmission event as the dependent variable. A 2-sided P<0.05 was considered to represent statistical significance. All analyses were performed using STATA version 16.

Weekday versus Weekend Hospital Admissions

Overall, 8,270,717 index HF hospitalizations were included, of which, 6,302,775 were admitted on a weekday and 1,967,942 were admitted on a weekend. Overall, 7,268,860 index hospitalizations occurred at urban hospitals (5,541,928 on weekday and 1,726,932 on weekend) and 1,001,856 index hospitalization occurred at rural hospitals (760,847 on weekday and 241,009 on weekend).

Baseline Characteristics

Baseline patient and hospital characteristics of HF hospitalizations based on the day of admission are presented in Table 1. Mean age among HF patients admitted on a weekday was 71.9 years (± 0.05) compared with 72.3 years (± 0.05) among HF patients admitted on a weekend (p<0.001). Male patients comprised 51.2% of weekday HF hospitalizations compared with 50.2% of weekend HF hospitalizations (p<0.001). 18.5% of weekday HF hospitalizations had CCI score of 2 and 68.7% had CCI score of ≥3, whereas 18.7% of weekend HF hospitalizations had CCI score of 2 and 68.5% had CCI score of ≥3 (all p<0.001).

Mean length of stay was longer for weekday heart failure index hospitalizations (5.4 days, 95% CI 5.3 – 5.4) compared to the weekend heart failure hospitalizations (5.1 days, 95% CI 5.0 – 5.1 days), p<0.001.

30-day readmission:

Among weekday hospitalizations, the 30-day all-cause readmission rate was 19.8% and the 30-day HF-specific readmission rate was 8.1%. Corresponding 30-day readmission rates for weekend hospitalizations were 20.3% and 8.4%, respectively. After statistical adjustment, weekend admissions were independently associated with higher odds of 30-day all-cause (aOR 1.04, 95% CI 1.03 – 1.05, p<0.001) and HF-specific readmissions (aOR 1.04, 95% CI 1.03 – 1.05, p<0.001) as shown in Figure 1.

Subgroup analysis demonstrated a similar relationship between weekend admissions and odds of readmission at both urban (aOR for all-cause readmission 1.04, 95% CI 1.03 – 1.05, p<0.001; aOR for HF-specific readmissions 1.04, 95% CI 1.03 – 1.05, p<0.001) and rural hospitals (aOR for all-cause readmission 1.04, 95% CI 1.02 – 1.06, p=0.001; aOR for HF-specific readmissions 1.05, 95% CI 1.02 – 1.09, p=0.001). Subgroup p-interaction for all-cause readmission was 0.136, and the sub-group p-interaction for HF-specific readmission was 0.060 (Figure 1).

Linear trends in 30-day readmission:

Linear trend for all-cause readmissions based on admission day showed readmission rate for weekday heart failure admissions decreased from 18.5% in 2010 to 18.2% in 2019 (trend p<0.001), whereas all-cause readmission rate for weekend admissions remained consistent from 2010 through 2019 (trend p=0.280), p-interaction <0.001 (Figure 2, Panel A).

Linear trend for HF-specific readmissions based on admission day showed readmission rate for weekday heart failure admissions decreased from 8.4% in 2010 to 8.3% in 2019 (trend p<0.001), whereas HF-specific readmission rate for weekend admissions decreased from 8.8% in 2010 to 8.7% in 2019 (trend p<0.001), p-interaction <0.001 (Figure 2, Panel B).

Cardiovascular Testing and Procedures

Observed rates of in-hospital cardiovascular tests and procedures are displayed in Figure 3. After multivariable adjustment, weekend HF admissions were less likely to undergo echocardiography (aOR 0.95, 95% CI 0.94 – 0.96, p<0.001), right heart catheterization (aOR 0.80, 95%CI 0.79 – 0.81, p<0.001), electrical cardioversion (aOR 0.90, 95% CI 0.88 – 0.93, p<0.001) and temporary mechanical support devices (aOR 0.84, 95% CI 0.79 – 0.89, p<0.001) during index hospitalization.

In subgroup analysis, weekend HF admissions in urban hospitals were less likely to undergo echocardiography, right heart catheterization, electrical cardioversion and temporary mechanical support devices. Weekend HF admissions in rural hospitals were as likely to undergo echocardiography, electrical cardioversion and temporary mechanical support devices; and less likely to undergo right heart catheterization (Figure 3).

Limitations

Limitations of this study should be noted. First, these observational data cannot determine cause-effect relationships and residual or unmeasured confounding may remain. Second, the NRD does not include information regarding hospital follow-up in primary care or subspecialty clinics, which may have a significant effect on the rates of readmissions and mortality. Third, during the study period, ICD coding changed from the Ninth Revision to the Tenth Revision in 2015. Though this change may have affected the total number of hospitalizations included in our final analysis, it is unlikely to affect the calculated 30-day readmission rate as this rate was reported as a proportion of the index hospitalizations within each calendar year. Fourth, the NRD data on mortality is limited only to in-hospital mortality. Thus, examining associations between weekend versus weekday admissions and post-discharge mortality was not possible. Fifth, the NRD does not include data regarding race, which precluded assessment of the potential interaction between race and the results seen here. Sixth, after-hours or holiday admissions were not considered in our study, and the relationships between these times of admission and hospital performance and patient outcome could not be assessed. Seventh, left ventricular ejection fraction data cannot be captured in the NRD. Thus, whether relations between day of admission, readmission, and cardiovascular procedures are consistent among patients with HFrEF and HFpEF could not be examined. Eighth, the NRD does not contain data about medical care delivered in emergency departments (ED). Hence, it is not possible to account for procedures that may have occurred in the ED or to determine whether differences in care delivered in EDs may have had an impact on 30-day readmission rates for patients admitted for HF. Ninth, though HF was the most common reason for 30-day readmission, the majority of readmissions were for reasons other than HF. Narrowing gaps in HF-related procedures during index hospitalization would not be expected to impact rates of non-HF readmission. Lastly, the NRD does not make it possible to determine the day of hospital discharge and whether that the patient is discharged on weekend or weekday. Thus, we were unable to study the potential contribution of the weekend versus weekday status of the discharge date to the observed associations with admission date seen here.

​

Online Supplement

Central illustration created with BioRender.com

1. Tsao CW, Aday AW, Almarzooq ZI, Alonso A, Beaton AZ, Bittencourt MS, et al. Heart Disease and Stroke Statistics—2022 Update: A Report From the American Heart Association. Circulation. 2022;145(8):e153–639. 10.1161/CIR.0000000000001052. [PubMed] [CrossRef] [Google Scholar]

2. Arora S, Patel P, Lahewala S, Patel N, Patel NJ, Thakore K, et al. Etiologies, Trends, and Predictors of 30-Day Readmission in Patients With Heart Failure. Am J Cardiol. 2017;119(5):760–9. 10.1016/j.amjcard.2016.11.022. [PubMed] [CrossRef] [Google Scholar]

3. Afzal A, Ananthasubramaniam K, Sharma N, Ai-Malki Q, Ali AS, Jacobsen G, et al. Racial differences in patients with heart failure. Clin Cardiol. 1999;22(12):791–4. 10.1002/clc.4960221207. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

4. Schmulewitz L, Proudfoot A, Bell D. The impact of weekends on outcome for emergency patients. Clin Med. 2005;5(6):621–5. 10.7861/clinmedicine.5-6-621. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

5. Angus Derek C., et al. “Critical Care Delivery in the United States: Distribution of Services and Compliance with Leapfrog Recommendations.” Critical Care Medicine, vol. 34, no. 4, Apr. 2006, pp. 1016–24. 10.1097/01.CCM.0000206105.05626.15. [PubMed] [CrossRef] [Google Scholar]

6. Petersen LA, Brennan TA, O’Neil AC, Cook EF, Lee TH. Does Housestaff Discontinuity of Care Increase the Risk for Preventable Adverse Events?Ann Intern Med. 1994;121(11):866. 10.7326/0003-4819-121-11-199412010-00008. [PubMed] [CrossRef] [Google Scholar]

7. Hoshijima H, Takeuchi R, Mihara T, Kuratani N, Mizuta K, Wajima Z, et al. Weekend versus weekday admission and short-term mortality. Medicine. 2017;96(17):e6685. 10.1097/MD.0000000000006685. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

8. Martin GP, Kwok CS, Spall HGCV, Volgman AS, Michos E, Parwani P, et al. Readmission and processes of care across weekend and weekday hospitalisation for acute myocardial infarction, heart failure or stroke: an observational study of the National Readmission Database. Bmj Open. 2019;9(8):e029667. 10.1136/bmjopen-2019-029667. [PMC free article] [PubMed] [CrossRef] [Google Scholar]

9. Shah M, Patnaik S, Patel B, Arora S, Patel N, Lahewala S, et al. The day of the week and acute heart failure admissions: Relationship with acute myocardial infarction, 30-day readmission rate and in-hospital mortality. Int J Cardiol. 2017;249:292–300. 10.1016/j.ijcard.2017.09.003. [PubMed] [CrossRef] [Google Scholar]

10. Fonarow GC, Abraham WT, Albert NM, Stough WG, Gheorghiade M, Greenberg BH, et al. Day of Admission and Clinical Outcomes for Patients Hospitalized for Heart Failure. Circulation Hear Fail. 2008;1(1):50–7. 10.1161/CIRCHEARTFAILURE.107.748376. [PubMed] [CrossRef] [Google Scholar]

11. Hamaguchi S, Kinugawa S, Tsuchihashi-Makaya M, Goto D, Tsutsui H. Weekend versus weekday hospital admission and outcomes during hospitalization for patients due to worsening heart failure: a report from Japanese Cardiac Registry of Heart Failure in Cardiology (JCARE-CARD). Heart Vessels. 2014;29(3):328–35. 10.1007/s00380-013-0359-5. [PubMed] [CrossRef] [Google Scholar]

12. Meaco*ck R, Doran T, Sutton M. What are the Costs and Benefits of Providing Comprehensive Seven-day Services for Emergency Hospital Admissions?Health Econ. 2015;24(8):907–12. 10.1002/hec.3207. [PubMed] [CrossRef] [Google Scholar]

13. Hatten JM, and Connerton RE. “Urban and Rural Hospitals: How Do They Differ?” Health Care Financing Review, vol. 8, no. 2, 1986, pp. 77–85. [PMC free article] [PubMed] [Google Scholar]

14. HCUP Nationwide Readmissions Database (NRD). Healthcare Cost and Utilization Project (HCUP). 2010–2019. Agency for Healthcare Research and Quality, Rockville, MD. www.hcup-us.ahrq.gov/nrdoverview.jsp. [Google Scholar]

15. Kwok CS, Abramov D, Parwani P, Ghosh RK, Kittleson M, Ahmad FZ, et al. Cost of inpatient heart failure care and 30-day readmissions in the United States. Int J Cardiol. 2021;329:115–22. 10.1016/j.ijcard.2020.12.020 [PubMed] [CrossRef] [Google Scholar]