Regional Impact of a Specialist Pediatric Palliative Care Program on Health Care Utilization and Cost: A Population-Level Difference-in-Differences Analysis
David L. Lysecki, James Lee, Myanca Rodrigues, Sumit Gupta, Rahul Chanchlani, Carolyn Wilson, Lehana Thabane, Danielle Zucchet, Richard PerezBackground:
Access to pediatric palliative care (PPC) is a universal human right, yet most children-in-need globally go without. Costs remain a primary barrier to program development, yet research regarding the economic impact of programs is inconsistent and inconclusive.
Objectives:
To examine health care utilization and costs among regional pediatric decedents during the development of a specialist PPC program.
Design:
Population-based retrospective cohort study using linked administrative databases and a difference-in-differences analytic approach.
Setting/Subjects:
Ontario, Canada, 2010–2018. 3,892 pediatric decedents (aged 31 days to 18 years) across Ontario, including 737 from Southcentral Ontario (SCO), the catchment wherein a specialist PPC program was initiated in 2016.
Measurements:
Primary outcomes were total health care days and total costs in the last year of life. Secondary outcomes included acute care days and costs, intensive care unit (ICU) days, and location of death. Difference-in-differences regression estimated incident rate ratios (IRRs) for days, mean relative differences (MRDs) for costs, and odds ratios for location of death. Counterfactual models estimated region-level impact.
Results:
Relative to the remainder of Ontario, the SCO cohort experienced decreased total costs (MRD 0.63; 95% confidence intervals [CIs] 0.49–0.81), acute costs (MRD 0.62; 95% CI: 0.48–0.80), and ICU days (IRR 0.56; 95% CI: 0.38–0.83) following initiation of the program. Counterfactual analysis estimated a cumulative regional cost reduction of $15.6 million CAD over three years, representing 0.87% of all regional pediatric expenditures.
Conclusions:
A specialist PPC program was associated with lower end-of-life costs at the population level, supporting investment in and evaluation of scalable PPC models.