Pediatric Hospital Discharges to Home Health and Postacute Facility Care: A National Study

Acute care hospitals are challenged to provide efficient, high-quality care to children who have medically complex conditions and may require weeks or months for recovery. Although the use of home health care (HHC) and facility-based postacute care (PAC) after discharge is well documented for adults, to our knowledge, little is known for children.

To assess the national prevalence of, characteristics of children discharged to, and variation in use across states of HHC and PAC for children.

Retrospective analysis of 2,423,031 US acute care hospital discharges in 2012 for patients ages 0 to 21 years from the nationally representative Agency for Healthcare Research and Quality Kids' Inpatient Database.

Discharges to HHC (eg, visiting or private-duty home nursing) and PAC (eg, rehabilitation facility) were identified from Centers for Medicare and Medicaid Services Discharge Status Codes. We compared children's characteristics (eg, race/ethnicity and number of chronic conditions) by discharge type using generalized linear regression.

The median age of participants was 3 years (interquartile range, 0-13 years), and 45.6% were female. Of 2,423,031 US acute care hospital discharges in 2012 for patients ages 0 to 21 years, 122,673 discharges (5.1%) were to HHC and 26,282 (1.1%) were to PAC facilities. Neonatal care was the most common reason (44.5%, n = 54,589) for acute care hospitalization with discharge to HHC. Nonneonatal respiratory, musculoskeletal, and trauma-related problems, collectively, were the most common reasons for discharge to PAC (42.9%, n = 11,275). When compared with PAC, more discharges to HHC had no chronic condition (34.4% vs 18.0%, P < .001) and fewer discharges to HHC had 4 or more chronic conditions (22.5% vs 37.7%, P < .001). In multivariable analysis, Hispanic children were less likely to use PAC (0.8% vs 1.1%; odds ratio [OR], 0.9 [95% CI, 0.8-0.9]) or HHC (3.3% vs 5.5%; OR, 0.8 [95% CI, 0.7-0.8]) compared with other children. Children with 4 or more chronic conditions compared with no chronic conditions had a higher likelihood of HHC use (11.0% vs 4.4%; OR, 2.9 [95% CI, 2.8-3.0]) and PAC (3.9% vs 0.8%; OR, 4.5 [95% CI, 4.3-4.9]). After case-mix adjustment, there was significant (P < .001) variation across states in HHC (range, 0.4%-24.5%) and PAC (range, 0.4%-4.9%) use.

Home health care and PAC use after discharge for hospitalized children is infrequent, even for children with multiple chronic conditions. It varies significantly by race/ethnicity and across states. Further investigation is needed to assess reasons for this variation and to determine for which children HHC and PAC are most effective.