Jay Berry1,2, Karen Wilson3, Helene Dumas2, Edwin Simpser4, Jane O'Brien1,2, Kathleen Whitford5, Rachna May6, Vineeta Mittal7, Nancy Murphy8, David Steinhorn9, Rishi Agrawal10, Kris Rehm11, Michelle Marks5, Christine Traul5, Michael Dribbon12, Christopher Haines12, Matt Hall13. 1. Boston Children's Hospital, Harvard Medical School, Boston, Massachusetts, USA. 2. Franciscan Children's Hospital, Boston, Massachusetts, USA. 3. Children's Hospital Colorado, University of Colorado, Aurora, Colorado, USA. 4. St. Mary's Healthcare System for Children, Bayside, New York, USA. 5. Cleveland Clinic Children's Hospital for Rehabilitation, Cleveland, Ohio, USA. 6. The Children's Hospital, University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, USA. 7. UTSW Medical Center & Children's Medical Center Dallas, Dallas, Texas, USA. 8. Primary Children's Hospital, University of Utah School of Medicine, Salt Lake City, Utah, USA. 9. Children's National Medical Center, Washington, D.C., USA. 10. Lurie Children's Hospital, Northwestern University Feinberg School of Medicine, Chicago, Illinois, USA. 11. Monroe Carroll Jr., Children's Hospital, Vanderbilt University School of Medicine, Nashville, Tennessee, USA. 12. Children's Specialized Hospital, New Brunswick, New Jersey, USA. 13. Children's Hospital Association, Overland Park, Kansas, USA.
Abstract
BACKGROUND: Recovery from respiratory illness (RI), a common reason for hospitalization, can be protracted for some children because of high illness severity or underlying medical complexity. OBJECTIVE: We assessed which children hospitalized with RI are the most likely to use post-acute facility care (PAC) for recovery. METHODS: Retrospective analysis of 609,800 hospitalizations for patients in 43 US children's hospitals between 2010- 2015 for RI, identified with the Agency for Healthcare Research and Quality Clinical Classification System. Discharge to PAC was identified using Centers for Medicare & Medicaid Services Discharge Status Codes. We compared patient characteristics by PAC use with generalized estimating equations. RESULTS: There were 2660 (0.4%) RI hospitalizations resulting in PAC transfer (n = 2660, 0.4%). Discharges to PAC had greater percentages of technology assistance (83.2% vs 15.1%), neuromuscular chronic condition (57.5% vs 8.9%), and mechanical ventilation (52.7% vs 9.1%), 𝑃 < 0.001 for all. The highest likelihood of PAC use occurred with ≥11 vs no chronic conditions (odds ratio [OR] 11.7 [95% CI, 8.0- 17.2]), ≥9 vs no therapeutic medication classes (OR 4.8 [95% CI, 1.8-13.0]), and existing tracheostomy (OR 3.0, 95% confidence interval [CI], 2.6-3.5). Median (interquartile range [IQR]) acute-care length of stay (LOS) for children most likely to use PAC was 19 (8-56) days; LOS remained long (median 13 [6-41] days) for children with the same attributes (n = 9448) not transferred to PAC. CONCLUSIONS: Children with RI who are most likely to use PAC have a high prevalence of multiple chronic conditions, multiple medications, and medical technology. Future investigations should assess the supply of PAC against the demand of hospitalized children with RI who might need it.
BACKGROUND: Recovery from respiratory illness (RI), a common reason for hospitalization, can be protracted for some children because of high illness severity or underlying medical complexity. OBJECTIVE: We assessed which children hospitalized with RI are the most likely to use post-acute facility care (PAC) for recovery. METHODS: Retrospective analysis of 609,800 hospitalizations for patients in 43 US children's hospitals between 2010- 2015 for RI, identified with the Agency for Healthcare Research and Quality Clinical Classification System. Discharge to PAC was identified using Centers for Medicare & Medicaid Services Discharge Status Codes. We compared patient characteristics by PAC use with generalized estimating equations. RESULTS: There were 2660 (0.4%) RI hospitalizations resulting in PAC transfer (n = 2660, 0.4%). Discharges to PAC had greater percentages of technology assistance (83.2% vs 15.1%), neuromuscular chronic condition (57.5% vs 8.9%), and mechanical ventilation (52.7% vs 9.1%), 𝑃 < 0.001 for all. The highest likelihood of PAC use occurred with ≥11 vs no chronic conditions (odds ratio [OR] 11.7 [95% CI, 8.0- 17.2]), ≥9 vs no therapeutic medication classes (OR 4.8 [95% CI, 1.8-13.0]), and existing tracheostomy (OR 3.0, 95% confidence interval [CI], 2.6-3.5). Median (interquartile range [IQR]) acute-care length of stay (LOS) for children most likely to use PAC was 19 (8-56) days; LOS remained long (median 13 [6-41] days) for children with the same attributes (n = 9448) not transferred to PAC. CONCLUSIONS:Children with RI who are most likely to use PAC have a high prevalence of multiple chronic conditions, multiple medications, and medical technology. Future investigations should assess the supply of PAC against the demand of hospitalized children with RI who might need it.