Impact of COVID19-related non-pharmacologic interventions on healthcare utilization for other virally-triggered respiratory illnesses

Résumé

Background: Acute and chronic respiratory illnesses are a leading cause of morbidity and mortality in Canada. While non-pharmacological interventions (NPIs) such as masking and physical distancing have effectively stemmed the spread of COVID19, the efficacy of NPIs in preventing other virally-triggered respiratory illnesses (VRIs) is less well understood. As the world moves into what may be the endemic phase of the COVID19 pandemic, better evidence is needed to inform rapidly-evolving public policy recommendations on the role of NPIs in infection control.

Methods: This study assessed the impact of NPI implementation on VRI-related healthcare utilization during the COVID19 pandemic. Following ethics approval from the Conjoined Health Research Ethics Board of Alberta (CHREB), long-term retrospective tableau data was extracted from the Alberta Health Services (AHS) data analytics enterprise data warehouse. International classification of disease (ICD-10) codes were used to identify patients who presented to an acute care facility in Alberta during the pandemic (Mar 2020-2021) and pre-pandemic (Feb 2015-2020) periods with a primary complaint of asthma, community-acquired pneumonia, influenza, or chronic obstructive pulmonary disease (COPD). Heart failure (HF) and acute appendicitis (AA) served as controls. The study team consisted of a medical student, an undergraduate student, a principal investigator from the Department of Emergency Medicine at the University of Calgary, members of the AHS provincial research data services team, and collaborating faculty members.

Evaluation Methods: The final study dataset comprised 585,809 ED visits and 175,456 hospitalizations. The primary outcome of interest was the change in ED visits and hospitalizations between the pandemic and pre-pandemic period for VRIs and controls. This was evaluated using quasi-experimental interrupted time-series analyses. A secondary outcome of interest was the cost-reduction associated with NPI implementation, for which multivariable regression models were constructed. These evaluation methods aimed to identify whether NPI implementation can (i) improve patient outcomes by preventing VRI-related ED visits and hospitalizations (ii) alleviate the strain on an already-constrained healthcare system by reducing VRI-associated healthcare spending.

Results: Triage acuity and comorbidity index scores were similar between the two periods. While a substantial decrease in healthcare utilization was observed in the early months of the pandemic for both VRIs and controls, a rapid rebound towards pre-pandemic caseloads was observed only for controls, while VRI-related health utilization remained consistently low. Overall, there was a 43-62% and 41-84% decrease in weekly ED visits and hospitalizations for individual VRIs during the pandemic period (all P<0.001). ED visits and hospitalizations for HF declined by a small magnitude of 6% (P=0.002) and 8% (P<0.001), respectively. In contrast, an 11% increase in ED visits (P<0.001) and 3% increase in hospitalizations (P=0.046) was observed for AA. The decrease in VRI-related healthcare utilization resulted in $121 million in cost reduction. Surprisingly, even after accounting for COVID19, there was a significant decrease of 19,391 ED visits and 1,524 hospitalizations for respiratory illnesses during the pandemic period (P<0.001).

Advice and Lessons Learned:

• NPI implementation was followed by a substantial decrease in healthcare utilization for VRIs. This resulted in substantial decrease in healthcare utilization costs and likely prevented significant patient morbidity and mortality.
• The greater magnitude decrease for VRIs than controls as well as the fact that acuity/comorbidity scores did not increase indicates that the observed decrease I healthcare utilization was primary driven by NPI implementation rather than an avoidance of healthcare settings due to fears of nosocomial COVID19 acquisition.
• NPIs appear to be an effective method of reducing the perennial burden of common respiratory illnesses. These findings provide a strong foundation for public policy recommendations on NPI use and establish the rationale for randomized studies on NPI use for preventing VRIs.