Use of virtual care near the end of life before and during the COVID-19 pandemic: A population-based cohort study

Abstract

Background and aims: The expanded use of virtual care may worsen pre-existing disparities in use and delivery of end-of-life care among certain groups of people. We measured the use of virtual care in the last three months of life before and after the introduction of virtual care fee codes that funded care delivery at the start of COVID-19 on March 14, 2020, and identified changes in the characteristics of people using it.

Methods: We used linked clinical and administrative datasets to study use of virtual care in the last three months of life among 411,564 adults who died between January 25, 2018, and November 30, 2022. Modified Poisson regression was used to measure the association of the use of virtual care in the last three months of life with the pandemic study period and its association with each person- and physician-level factor.

Results: 14,261 people (8%) used virtual care in the last three months of life before the pandemic, and 161,000 people (69%) used it during the pandemic (relative risk [RR] 8.76; 95% CI 8.48-9.05). Several individual patient characteristics were associated with statistically significant increases in the use of virtual care after March 14, 2020 (following the introduction of virtual care fee codes), compared to before such as among older adults, ethnic minorities, multiple chronic comorbid health conditions and higher frailty groups.

Conclusions: The introduction of new fee codes broadening technology and funding for end-of-life care at the start of pandemic combined with pandemic-related effects was associated with a substantial increase in the use of virtual care near the end of life among certain groups and a general leveling of pre-existing disparities in its use. Virtual end-of-life care delivery may strengthen person-centredness for individuals with limited ability to attend in-person appointments and by providers who may not have previously engaged in such care.

Fig 1. Construction of the study cohort.

Fig 2

A and B. Forest plot of the adjusted relative risk of receiving virtual end-of-life care before (blue squares, pre-pandemic) and after (red circles, pandemic) the introduction of reimbursable virtual care physician fee codes on March 14, 2020 according to person-level (Panel A) and most responsible physician-level (Panel B) characteristics among adults in their last 3 months of life who died in Ontario between 2018 and 2022. An RR >1 indicates an increased and a RR < indicates a decreased likelihood of receiving virtual end-of-life care. Reference groups were age (vs. 18–29 years), female sex (vs. male), rural residence (vs. urban), neighbourhood income quintile (vs. 1^st quintile), household index quintile (vs. 5^th quintile), material resources index quintile (vs. 5^th quintile), age & labor index quintile (vs. 5^th quintile), racialized and newcomers index quintile (vs. 5^th quintile), surname-based ethnicity (vs. general population), chronic conditions (yes vs. no), hospital frailty risk score (vs. not hospitalized), prior hospitalization/palliative care/end-of-life designation (yes vs. no). COPD–chronic obstructive pulmonary disease; PallC–palliative care; EOL–end-of-life.