Goals of Care During the COVID-19 Pandemic: Implementing DNI, DNR, and DNH Orders in a Skilled Nursing Facility

Abstract

Knowledge of patient treatment preferences concerning resuscitation, intubation, and hospitalization are critical as hospitals and skilled nursing facilities (SNFs) admit older COVID-19 patients facing uncertain disease trajectories that could turn fatal within days. The present study examined the number of Do Not Intubate (DNI), Do Not Resuscitate (DNR), and Do Not Hospitalize (DNH) orders before and after patients were diagnosed with COVID-19 in a New York City SNF, as well as demographic variables related to order implementation. Subjects were 150 SNF patients who tested positive for COVID-19 between March and May 2020. All data, including demographic information, medical orders, and goals of care (GOC) discussions were obtained from the SNF’s electronic medical records. Bivariate correlations and Chi-square analyses were conducted. Patient ages ranged from 42 to 103 years (Mean=79.82, SD=11.78). Ninety-six (64%) patients were female and 48 (32%) were White, 46 (31%) Black, 22 (15%) Hispanic, 2 (1%) Asian, and 32 (21%) unknown. Most (118, 79%) reported English as their primary language. SNF length of stay (LOS) ranged from 6 days to 11 years (Mean=842.67 days, SD=931.18). One hundred and eleven (74%) patients were long-term care (LTC) and 39 (26%) were post-acute (LOS <100 days). Fifty three (35%) patients were enrolled in Optum, a service providing added specialized care by nurse practitioners (NPs). Optum NPs initiate GOC discussions routinely and when life-threatening changes in condition occur. Pre-COVID-19, 79 (53%) patients had DNR orders, 76 (51%) DNI, and 24 (16%) DNH. Overall, 69 (46%) patients did not have any DNI, DNR, or DNH orders. After COVID-19 onset, 99 (66%) of the 150 patients/families had GOC and treatment option discussions with their clinicians. Following these conversations, 96 additional orders (22 DNIs, 22 DNRs, and 52 DNHs) were placed for 57 (58%) patients. Relatedly, the number of patients lacking any medical orders decreased to 47 (31%). Older age (r =.227, p <.01), LTC (X2=7.1, p < .05), and Optum enrollment (X2=15.07, p <.01) were associated with additional medical orders after a COVID-19 diagnosis. No relationship with changes in medical orders were found for gender, race, and primary language. After COVID-19 diagnoses, GOC conversations resulted in a substantial increase in DNR, DNI, and DNH orders. Unsurprisingly, older LTC residents were more likely to implement new DNR, DNI, and DNH orders. As Optum NPs routinely discuss GOC with patients/families, decision-making during critical times may be facilitated.