Comparative environmental impact analysis of screening tests for colorectal cancer: a single-centre life cycle assessment
Vivek A Rudrapatna, Tzu An Wang, Parsia Vazirnia, Kaiyi Wang, Nathan Alhalel, Shadera Azzam, Gunnar Mattson, Amy Becker, Ching-Ying Oon, Shan Wang, William Karlon, Scott Pasternak, Cassandra L Thiel, Seema Gandhi, Sean A WoolenBackground
Healthcare is a major contributor to greenhouse gases. One of the most widely used healthcare services in the USA is colorectal cancer (CRC) screening, indicated for 134 million adults. Recommended screening includes annual faecal immunochemical tests (FITs), CT colonographies every 5 years or colonoscopies every 10 years.
Objective
To compare the environmental impacts of these tests for CRC screening.
Design
We conducted a comparative life cycle assessment of three CRC screening strategies at UCSF. We performed on-site audits to document the resources used for each screening test. We estimated the environmental impacts of these procedures, measured by global warming potential (GWP) and damage to human health. We computed 10-year impacts of each screening strategy using a Markov model. We accounted for uncertainty using hierarchical Monte Carlo simulations.
Results
FITs had the lowest environmental impacts, roughly 20 percentage points superior to colonoscopies, and this was robust on sensitivity analyses. Across tests, the biggest cause of environmental harm was car-based transportation of patients and staff. Prioritising FITs over colonoscopies in the USA could enhance population health by 5.2 million disability-adjusted life-years per decade. Transitioning to electric vehicles could reduce the GWP of all screening tests by 15%–20%.
Conclusions
Given the similar efficacy and safety of these tests, payors should probably prioritise FITs for low-risk patients. Governments should decarbonise transportation and mandate environmental product declarations. We call for a closer look at resource-intensive preventative health strategies, which could result in more harm than good if applied to a low-risk population.