| Abstrakt: |
Background: Medicare's Readmission Reduction program fines hospitals with readmission rates that exceed the national average; however, standardized risk adjustment models do not take into account how far people live from an institution. Larger academic medical centers may be disadvantaged by this oversight. The purpose of this study was to determine if distance travelled could be predicted by length of stay (LOS) and comorbidity. Methods: The Bridging the Discharge Gap Effectively (BRIDGE) was accessed to obtain ZIP code, LOS and to calculate the Charlson Comorbidity Index for patients referred in 2012. Mean LOS was compared to national data from the HCUP database for acute coronary syndrome (ACS), heart failure (HF), and atrial fibrillation (AF). Distance from the medical center (MC) to the center of each ZIP code was evaluated for correlation with LOS and comorbidity. Local and Distant patients were defined as patients who lived within or outside a 40 mile radius of the MC. Results: A total of 785 were included in this analysis; mean age was 66.0 + 14.3 years and 60.9% were male. Overall, ACS patients had an average LOS 1 day less than the national average, whereas HF and AF patients stayed respectively 1.6 and 1.3 days longer than the national average. There was a significant difference in mean LOS for both ACS and HF patients who lived greater than 40 miles from the MC, (p=0.014, p=0.046). Although insignificant, Distant AF patients also had a mean LOS longer than that of local patients (p=0.12). For all patients who participated in the study, and for ACS patients in particular, there was a weak, but significant, relationship between LOS and distance from the MC (All: r=.18, n=785, p< .001; ACS: r=0.25, n=217, p<0.001), Conclusions: In this sample, LOS is positively correlated with distance from the MC. Because our institution is a large referral center, those referred from a distance are those who likely cannot be managed near their homes. As seen here, more complex HF and AF patients travel greater than 40 miles to receive treatment. Programs designed to decrease readmissions, like BRIDGE, may be less effective in these referral populations who may be less likely to travel for follow-up appointments. Further study is warranted to ascertain if distance should be part of standardized risk adjusting calculations. [ABSTRACT FROM AUTHOR] |
