Assessing the environmental and downstream human health impacts of decentralizing cancer care.

Abstract

Listen

Translate article

1522 Background: Greenhouse gas emissions from healthcare are substantial and harm persons with cancer. Emissions from outpatient cancer care visits are not well described, nor are the downstream reduction in human harms that could be obtained through visit “decentralization” (telemedicine and local care when possible). Methods: This life-cycle assessment (LCA)-based study evaluated changes in emissions and downstream health harms associated with (1) telemedicine visits using retrospective observational analysis and (2) fully decentralized visits using counterfactual modeling. The retrospective observational cohort included persons receiving cancer care at Dana-Farber Cancer Institute (DFCI) and 20 affiliated facilities between 5/2015-12/2020. The main outcome was the adjusted per visit-day difference in emissions (in kilograms carbon dioxide equivalents: kgCO2e) between two periods: an in-person care period (5/2015-2/2020; “Pre”) and a telemedicine period (3/2020-12/2020; “Post”). Mixed effects log-linear model assessed emissions changes adjusted for age, sex, race, ethnicity, and cancer type, with random effects on person. The counterfactual model assessed emissions changes between actual in-person visits during the Pre period and a counterfactual with maximal decentralization. Visit-day eligibility for decentralization was obtained by categorizing appointments (e.g., clinician visit, infusion) as DFCI required, local care possible, or telemedicine possible. This cohort was matched to a national population diagnosed with cancer over the same period (Cancer in North America [CiNA] dataset) using mixed-effects linear modeling, through which annual changes in disability-adjusted life-years (DALY) from clinician visit decentralization were estimated using Eckelman’s mortality cost of carbon conversion. Results: There were123,890 unique patients in the DFCI cohort seen over 1.6 million visit-days (Pre N=110,180, Post N=61,691) at a median of 6 visit-days per patient (IQR 2, 15). An estimated 72,554,006 kgCO2e were within scope of the LCA emitted during the study period. In mixed-effects log-linear regression, the adjusted mean absolute reduction in per visit-day emissions between Pre and Post periods was 36.4 kgCO2e (95%CI 36.2,36.6), or a reduction of 81.3% (95%CI 80.8,81.7) compared to the baseline model. In the counterfactual decentralized care model of the Pre period, there was a relative emissions reduction of 33.1% (95%CI 32.9,33.3). When demographically matched to 10.3 million persons in the CiNA dataset, decentralized care would have reduced national emissions by 75.3 million kgCO2e annually; this corresponds to an estimated annual reduction of 15.0-47.7 DALYs. Conclusions: Employing decentralization through telemedicine and local care can substantially reduce emissions during cancer care delivery; this corresponds to a small reduction in human mortality.