ABSTRACT
Background: Many hospitals have established goals-of-care (GOC) programs in response to the COVID- 19 pandemic;however, few have reported their outcomes. MD Anderson Cancer Center launched a multicomponent interdisciplinary GOC (myGOC) program in March 2020 that involved risk stratification, team huddles to discuss care planning, oncologist-initiated GOC discussions, communication training, palliative care involvement, rapid-response GOC team deployment, and daily monitoring with immediate feedback. We examined the impact of this myGOC program among medical inpatients. Methods: This single-center study with a quasi-experimental design included consecutive adult patients with cancer admitted to medical units at MD Anderson Cancer Center, Texas during an 8-month pre-implementation (May 1, 2019 to December 31, 2019) and post-implementation period (May 1, 2020 to December 31, 2020). The primary outcome was intensive care unit (ICU) mortality. Secondary outcomes included ICU length of stay, hospital mortality, and proportion/timing of patients with inhospital do-not-resuscitate (DNR) orders, medical power of attorney (MPOA), living will (LW) and outof- hospital DNR (OOHDNR). Propensity score weighting was used to adjust for differences in potential covariates, including age, sex, cancer diagnosis, race/ethnicity, and Sequential Organ Failure Assessment (SOFA) Score. With a sample size of 600 ICU patients over each time period and a baseline ICU mortality of 28%, we had 80% power to detect a 5% reduction in mortality using a two-tailed test at 5% significance level. Results: This study involved 12,941 hospitalized patients with cancer (Pre n = 6,977;Post n = 5,964) including 1365 ICU admissions (Pre n = 727;Post n = 638). After myGOC initiation, we observed a significant reduction in ICU mortality (28.2% vs. 21.9%;change -6.3%, 95% CI -9.6, -3.1;P = 0.0001). We also observed significant decreases in length of ICU stay (mean change -1.4 days, 95% CI -2.0, -0.7 days;P < 0.0001) and in-hospital mortality (7% vs. 6.1%, mean change -0.9%, 95% CI -1.5%, -0.3%;P = 0.004). The proportion of hospitalized patients with an inhospital DNR order increased significantly from 14.7% to 19.6% after implementation (odds ratio [OR] 1.4, 95% CI 1.3, 1.5;P < 0.0001) and DNR was established earlier (mean difference -3.0 d, 95% CI -3.9 d, -2.1 d;P < 0.0001). OOHDNR (OR 1.3, 95% CI 1.1, 1.6, P < 0.0007) also increased post-implementation but not MPOA and LW. MPOA, LW and OOHDNR were documented significantly earlier relative to the index hospitalization in the post-implementation period (P < 0.005 for all). Conclusions: This study showed improvement in hospital outcomes and care plan documentation after implementation of a system-wide, multicomponent GOC intervention. Our findings may have implications for GOC programs during the pandemic and beyond.