Previous studies have identified a “weekend effect” in terms of a poor outcome for patients hospitalized with various acute medical conditions. The aim of our study was to investigate whether weekend admissions for atrial fibrillation (AF) result in worse outcomes than those admitted on weekdays. In the Nationwide Inpatient Sample 2008 database, we identified a total of 86,497 discharges with a primary discharge diagnosis of AF. The use of a cardioversion procedure for AF on weekends was lower than that on a weekday (7.9% vs 16.2%; p <0.0001; odds ratio 0.5, 95% confidence interval 0.45 to 0.55, p <0.0001). After adjusting for patient and hospital characteristics and disease severity, the adjusted in-hospital mortality odds were greater for weekend admissions (odds ratio 1.23, 95% confidence interval 1.03 to 1.51; p <0.0001). The length of stay was significantly longer for weekend admissions. In conclusion, patients admitted with AF on weekends had lower odds of undergoing a cardioversion procedure and greater odds of dying.
The outcome of conditions such as acute myocardial infarction is relatively poor when patients are hospitalized over a weekend versus a weekday. A likely contributing factor to this observation is the limited number of hospital staff and availability of in-house expertise for invasive coronary procedures on the weekends. Data on the “weekend effect” for atrial fibrillation (AF)-related hospitalizations and their outcomes are not available. Because AF is the most common sustained cardiac rhythm disturbance, even a small difference in mortality between the weekday and weekend admissions of patients would result in a substantial number of increased deaths in the population by virtue of its high incidence. We investigated the differences in outcomes for AF-related hospitalization on the weekend versus weekdays using a large national hospitalization database.
Methods
The Nationwide Inpatient Sample (NIS) for 2008 is the largest all-payer database of hospital inpatient stays available in United States. Data from the NIS have been used to identify, track, and analyze national trends in healthcare usage, patterns of major procedures, access, disparity of care, trends in hospitalizations, charges, quality, and outcomes. The 2008 NIS contains all discharge data from 1,044 hospitals located in 40 States, approximating a 20% stratified sample of United States community hospitals. The sampling frame for the 2008 NIS is a sample of hospitals that comprises approximately 90% of all hospital discharges in the United States. Each individual hospitalization is de-identified and maintained in the NIS as a unique entry with 1 primary discharge diagnosis and ≤14 secondary diagnoses during that hospitalization. Each entry also carries information on demographic details, insurance status, co-morbidities, primary and secondary procedures, hospitalization outcome, and length of stay.
The NIS contains the clinical and resource use information included in a typical discharge abstract, with safeguards to protect the privacy of individual patients, physicians, and hospitals (as required by the data sources). The 2008 database was the most recent release from the Healthcare Cost and Utilization Project. NIS data have been used to explore the outcomes for discharges for various medical and surgical diagnoses.
Our primary interest group was hospitalizations with a primary diagnosis of AF. All patients with International Classification of Diseases, 9th Revision, Clinical Modification, code 427.31 or 427.32 as the principal diagnosis were included. In alignment with previous studies, weekend admissions were defined as admissions between midnight on Friday through midnight on Sunday.
Outcomes
Our main outcomes were the use of cardioversion procedures, length of stay, hospital mortality, and total hospitalization charges.
All analyses were performed using SAS, version 9.2 (SAS Institute, Cary, North Carolina). Survey procedures available within the SAS were applied in the analysis to account for design features of the complex sample survey, such as clustering, stratification, and sampling weights. Therefore, the resulting estimates should be representative of the national hospital inpatient admissions. Descriptive statistics, including proportions, mean values, and standard errors, were generated for the individual and hospital characteristics for both weekend and weekday admissions. Univariate tests were applied to compare the equality of the mean or proportions for the mentioned outcomes between the weekday and weekend admissions. These tests consisted of the Rao-Scott chi-square test for categorical outcomes and simple linear regression analysis for continuous outcomes. Finally, multivariate models were applied to test the adjusted associations between the outcomes and weekend versus weekday admissions. For the number of procedures, days to the procedure, length of stay, and total charges, we applied multivariate linear regression models. Adjusted parameter estimates for weekend versus weekday admission were collected. For in-hospital mortality, a logistic regression model was used. The level of significance (α) was set at 5%.
Results
There were 86,497 discharges with AF as the primary diagnosis, yielding a national estimate of 425,744 hospitalizations, with the number adjusted for the entire United States population. Of these, 16,949 were characterized as weekend admissions and 69,548 as weekday admissions. The baseline characteristics of the patients as a whole and in each group (weekday admission group and the weekend admission group) are listed in Table 1 .
| Characteristic | Weekday Hospitalizations (n = 69,548) | Weekend Hospitalizations (n = 16,949) | Total Hospitalizations (n = 86,497) |
|---|---|---|---|
| Age (years) | 70.3 ± 0.1 | 70.5 ± 0.1 | 70.3 ± 0.1 |
| Gender | |||
| Male | 48% | 46% | 46% |
| Female | 52% | 54% | 54% |
| Race | |||
| White | 84% | 82% | 84% |
| Black | 6% | 7% | 6% |
| Hispanic | 5% | 6% | 5% |
| Asian | 1% | 1% | 1% |
| Native American | 1% | 1% | 1% |
| Other | 3% | 3% | 3% |
| Primary payer | |||
| Medicare | 65% | 65% | 65% |
| Medicaid | 3% | 4% | 3% |
| Private | 27% | 25% | 27% |
| Self | 2% | 3% | 2% |
| No charge | 1% | 1% | 1% |
| Other | 2% | 2% | 2% |
| Hospital region | |||
| Northeast | 20% | 20% | 20% |
| Midwest | 26% | 26% | 26% |
| South | 38% | 39% | 38% |
| West | 16% | 15% | 16% |
| Hospital teaching station | |||
| Nonteaching | 58% | 62% | 59% |
| Teaching | 42% | 38% | 41% |
| Hospital location | |||
| Rural | 14% | 16% | 15% |
| Urban | 86% | 84% | 85% |
| Hospital bed size | |||
| Small | 14% | 14% | 14% |
| Medium | 23% | 24% | 23% |
| Large | 63% | 62% | 63% |
We compared the in-hospital mortality among patients admitted on weekends and those admitted on a weekday. We found that the patients admitted on weekends experienced a greater proportion of in-hospital mortality than those admitted on weekdays (1.1% vs 0.9%; p = 0.01). Patients hospitalized on a weekend for AF were 1.24 times more likely to die compared to those hospitalized on weekdays (odds ratio 1.24, 95% confidence interval 1.02 to 1.51; p = 0.0029) after the data were adjusted for patient characteristics, co-morbidities, and hospital characteristics.
The number of inpatient cardioversion procedures, interval to procedure, length of stay, and cost of hospital stay before and after adjustment for patient characteristics are listed in Tables 2 and 3 , respectively. We found that those admitted on the weekend for AF underwent fewer cardioversion procedures than those hospitalized on a weekday (7.9% vs 16.2%, p <0.0001). The time to cardioversion was significantly longer for patients with a weekend admission, and this resulted in a longer length of stay. The weekend AF admission was associated with a $2,500 lower total charge on average, after adjustment for other covariates. This difference remained significant after adjustment for patient characteristics, disease severity, and hospital characteristics (odds ratio 0.5, 95% confidence interval 0.45 to 0.55, p <0.0001).
