Effects of health information technology and health information exchanges on readmissions and length of stay

Abstract

IntroductionThis study explored the effects of HIT/HIE on the outcomes and quality of care for diabetes, asthma, acute myocardial infarction, congestive heart failure, and hip and knee fracture. 30 day readmissions and average length of stay were used as measures of quality of care. MethodsThis study used the HIMSS IT database, coupled with the American Hospital Association (AHA) Hospital survey data and the CMS outpatient and skilled nursing facility Medicare claims data from 2006–2012, along with the standard Medicare file of patient claims for 2006–2012 as the control cohort. Using heirarchical clustering analysis methods, we construct HIT scores for measures of HIT and HIE implementation and penetration across 32 measures of HIT adoption. The scores are then used as the coefficient of interest using fixed effects methods and multinomial regressions, controlling for hospital and patient characteristics, as well as time-invariant factors. ResultsThis study finds that those institutions with the highest HIT scores for specific measures of clinical decision support, continuity of care documentation and clinical discharge and summary care documentation showed modest and statistically significant reduction in 30 day readmissions, with no effects on length of stay. DiscussionThis study has major implications for the future of HIT/HIE in coordinating care across disparate entities. Initiatives aimed at improving care coordination, fostering health ecosystems, and reducing costs associated with unnecessary readmissions and length of stay and stand to benefit from this research in that it highlights tangible areas where HIT/HIE participation may be beneficial to achieving improved health.