Comparison of Sex-Based Differences in Home or Nonhome Discharge Utilization of Rehabilitative Services and Outcomes Following Transcatheter Aortic Valve Implantation in the United States.

Sex-based differences in outcomes have been shown to affect caregiving in medical disciplines. Increased spending due to postacute care transfer policies has led hospitals to further scrutinize patient outcomes and disposition patterns after inpatient admissions. We examined sex-based differences in rehabilitative service utilization after transcatheter aortic valve implantation (TAVI). We queried all TAVI discharges in the National Inpatient Sample database from 2012 to 2014 (n = 40,900). Thirteen thousand eight hundred fifteen patients were discharged to home and 12,175 patients were discharged to rehabilitation facility; those not discharged routinely or to a rehabilitation facility were excluded. Patients with nonhome discharges were older (83.3 vs 79.0 years) and female (58.3% vs 37.7%) with a greater number of chronic conditions (9.91 vs 9.03) and number of Elixhauser co-morbidities (6.5 vs 5.8, all p < 0.05). Nonhome discharge patients also had a significantly longer length of stay (LOS) (11.3 days vs 5.3 days) and higher hospitalization costs ($66,246 vs $48,710, all p < 0.001) compared to home-discharged patients. Overall in-hospital mortality for female patients who underwent TAVI was higher compared to males (4.6% vs 3.6%, p < 0.05). On multivariable logistic regression, female sex was an independent predictor for disposition to rehabilitation facilities after TAVI (odds ratio 2.17; 95% confidence interval: 1.88 to 2.50; p < 0.001). Other independent predictors for females discharged to rehabilitation included the presence of rheumatoid arthritis and collagen vascular disease, body mass index greater than 30 kg/m2, depression, and sum of Elixhauser co-morbidities (all p < 0.001). In conclusion, nonhome discharge TAVI patients added LOS and hospital costs compared to home discharge TAVI patients, and female sex was one of the major predictors despite the lower co-morbidities.