
==== Front
Ann Am Thorac Soc
Ann Am Thorac Soc
AnnalsATS
Annals of the American Thoracic Society
2329-6933
2325-6621
American Thoracic Society

38820253
202312-1078OC
10.1513/AnnalsATS.202312-1078OC
Original Research
Adult Pulmonary
Impact of COVID-19 Pandemic on Chronic Obstructive Pulmonary Disease Healthcare Use, Exacerbations, and Mortality: A Population Study
https://orcid.org/0000-0002-7366-193X
Lam Grace Y. 1 2 3 7
https://orcid.org/0000-0001-7077-1693
Wen Chuan 8 9
https://orcid.org/0000-0002-3958-7561
Ronksley Paul E. 10 11
https://orcid.org/0000-0002-3658-2554
Bakal Jeffrey A. 8 9
https://orcid.org/0000-0002-1911-4169
Bhutani Mohit 1 2 7
Soril Lesley J. J. 4 6
https://orcid.org/0000-0001-8234-4760
Stickland Michael K. 1 2 6
https://orcid.org/0000-0002-2427-6277
Gross Douglas P. 5
https://orcid.org/0000-0002-0615-4575
Weatherald Jason 1 2 7
1 Division of Pulmonary Medicine, Department of Medicine,
2 Alberta Respiratory Centre,
3 Women and Children’s Health Research Institute,
4 Division of General Internal Medicine, Department of Medicine, and
5 Department of Physical Therapy, University of Alberta, Edmonton, Alberta, Canada;
6 Medicine Strategic Clinical Network – Respiratory Health Section,
7 Alberta Health Services, and
8 Provincial Research Data Services, Alberta Health Services, Edmonton, Alberta, Canada;
9 Alberta Strategy for Patient Oriented Research Unit, Edmonton, Alberta, Canada;
10 Department of Community Health Sciences, Cumming School of Medicine, and
11 O’Brien Institute for Public Health, University of Calgary, Calgary, Alberta, Canada
Correspondence and requests for reprints should be addressed to Grace Y. Lam, M.D., M.Sc., Ph.D., F.R.C.P.C., 3-111C Clinical Sciences Building, 11302 83 Avenue NW, Edmonton, AB, T6G 2G3 Canada. E-mail: glam@ualberta.ca.
1 9 2024
1 9 2024
1 9 2024
21 9 12811288
21 12 2023
29 5 2024
Copyright © 2024 by the American Thoracic Society
2024
https://creativecommons.org/licenses/by-nc-nd/4.0/ This article is open access and distributed under the terms of the Creative Commons Attribution Non-Commercial No Derivatives License 4.0. For commercial usage and reprints, please e-mail Diane Gern.

Rationale

Existing work suggests that patients with chronic obstructive pulmonary disease (pwCOPD) presented less frequently to the emergency department and were less likely to be hospitalized during the coronavirus disease (COVID-19) pandemic, but it is unclear if this was due to improved health and disease management or to increased barriers and/or avoidance of health care.

Objectives

The objective of this study was to determine the impact of the pandemic on inpatient and outpatient healthcare use, disease incidence, and mortality rates in pwCOPD.

Methods

A retrospective population-based analysis using linked administrative datasets from Alberta, Canada 18 months before and after March 12, 2020 was conducted to measure hospitalization, emergency department and outpatient visits, and COPD outpatient exacerbations during these time periods. Mortality data were also analyzed before versus after the pandemic, taking confirmed COVID-19 infection within 30 days into account. Subgroup analysis based on COPD exacerbation risk stratification was undertaken to determine if healthcare use differed based on exacerbation risk. Finally, sex-based analysis of healthcare use during the pandemic was also completed.

Results

Hospitalization or emergency department visits and outpatient treatment for acute exacerbations of COPD dropped, whereas total outpatient COPD visits, including both virtual and in person, increased during the pandemic for pwCOPD. The mortality rate increased even after adjusting for COVID-19–associated deaths. Sex-based subgroup analysis showed a greater drop in acute care use for females, but the rise in mortality was seen for both sexes, with men experiencing a greater rate of mortality than women.

Conclusions

Overall, pwCOPD accessed acute care resources less during the pandemic, which may have contributed to a rise in non–COVID-19 all-cause mortality.

Keywords

COPD
mortality
hospitalization
healthcare use
pulmonary exacerbations
Canadian Institutes of Health Research 10.13039/501100000024 179875
==== Body
pmcThe coronavirus disease (COVID-19) global pandemic necessitated significant alterations to the way that health care was delivered and accessed. As focus begins to shift toward recovery, more attention is being paid to how the pandemic might have changed the outcomes of patients with chronic conditions. Specifically, patients with pulmonary diseases who are at risk of recurrent exacerbations, such as chronic obstructive pulmonary disease (COPD), were presumed to experience worse outcomes during a respiratory virus pandemic. Consistently, COPD has been identified as a risk factor for poorer COVID-19 outcomes once infected, including greater pneumonia severity leading to a greater need for hospitalization or intensive care unit admission (1). However, a number of retrospective comparative studies from high-income countries have suggested that patients with COPD (pwCOPD) experienced a reduction in COPD-related hospital admissions (2, 3), emergency department visits (4, 5), and primary care visits (5) during the early pandemic period. How the reduction in medical encounters impacted mortality is unclear. Alsallakh and colleagues, using population-level data from the United Kingdom, found no change to the incident rate ratio (IRR) of mortality before compared with after the pandemic, although the impact of COVID-19 infection on mortality was not accounted for (5). Conversely, Fedeli and colleagues examining population-level data from Italy found a reduction in COPD-mediated mortality by 8% during the early pandemic, although mortality of patients with comorbid COPD rose by 14% (6). Here, we used linked population-wide administrative databases to retrospectively compare the healthcare usage, incidence, and mortality outcomes of pwCOPD 18 months before versus after March 12, 2020, the day that the World Health Organization declared a COVID-19 global pandemic. It is hypothesized that although pwCOPD had fewer inpatient or emergency department medical encounters, differences in healthcare access and use during the pandemic may be associated with increased mortality.

The study was approved by the University of Alberta Research Ethics Board (Pro00119420) and conducted in full accordance with the ethics board guidelines. A waiver of consent was granted by the Research Ethics Board, given that the study involved anonymized, deidentified and unidentifiable provincial-level data.

Methods

Study Design and Case Definitions

This retrospective cohort study was conducted in the province of Alberta, Canada (2020 population estimate, 4.4 million). The study compared 18-month periods before and after March 12, 2020. PwCOPD were identified using a validated Canadian algorithm of one or more COPD-related hospitalization or physician visits (see Table E1 in the data supplement) that occurred at age 35 years or older (7, 8). All cases were checked 5 years back from the incident date (to March 31, 2007, because of data availability) to ensure no prior COPD-related encounters were found. Prevalent cases were defined by an incident date before September 12, 2018, which marks the beginning of the prepandemic period. The survived prevalent cohort was defined as the prepandemic prevalent cases of pwCOPD who were alive as of March 12, 2020. Cases with an incident date on or after September 12, 2018 were defined as new incident cases (Figure E1).

Data Sources

As Alberta has a single health system, Alberta Health or Alberta Health Services databases are reflective of provincial-level data, and the following were linked for use in this study: the National Ambulatory Care Reporting System, Inpatient Discharge Abstract Database, Practitioner Claims, Pharmaceutical Information Network, Alberta Health Services Enterprise Data Warehouse schema labeled “COVID-19” (for COVID-19 test results), and the Alberta Population Registry. Linkages were completed using the personal health number. Leveraging these sources, we extracted comprehensive data on hospitalizations, emergency department visits, outpatient clinic (including virtual and in person; specialists and general practitioners) appointments, physician billing claims, pharmacy medication dispensations, COVID-19 testing, and neighborhood-level census data. We limited outpatients’ primary care visits to one primary care provider per patient per day.

Primary Outcomes Measures and Definitions

We assessed general and COPD-related hospitalization, general and COPD-related emergency department visits, general and COPD-related outpatient primary care or specialist visits and all-cause mortality, which were compared among the COPD prevalent, survived COPD prevalent, and new incident cohorts before and after the pandemic. For each outcome, we reported both the aggregate count of events and the total number of affected individuals. We also generated binary indicators to note the occurrence or nonoccurrence of each outcome. For all-cause mortality, we defined COVID-19–associated death as death within 30 days of a documented positive COVID-19 test (either documented polymerase chain reaction or rapid antigen tests administered in healthcare facilities) and excluded these cases for subgroup analysis. Note that during the study period home rapid antigen tests were not yet widely available.

Outpatient episodes of acute exacerbation of COPD (AECOPD) were defined as the episodic dispensation of oral corticosteroid therapy (one consistent dispensation with prescription days ⩾5) or a respiratory antibiotic (Table E2) within 7 days of a physician visit associated with bronchitis, pneumonia, influenza, emphysema, asthma, or other COPD (modified approach from Mapel and colleagues [9]).

Comorbidities

In assessing comorbidities for mortality analysis, we used the Charlson comorbidity index by integrating data from the Inpatient Discharge Abstract Database, National Ambulatory Care Reporting System, and Practitioner Claims databases. The selected data covered a 2-year interval before the commencement of both prepandemic (starting September 12, 2018) and postpandemic (starting March 12, 2020) study durations. For individuals who underwent one or more COVID-19 tests in the postpandemic period, the data collection extended to 2 years before their latest COVID-19 test date.

Determination of Risk of AECOPD

For subgroup analysis based on the severity of AECOPDs, patients were categorized based on their treatment history in the preceding 12 months as either low risk (one or fewer moderate AECOPDs [defined as an outpatient prescription of respiratory antibiotics and prednisone]); or high risk (two or more moderate AECOPDs OR hospitalization or emergency department visit). These definitions are based on the Canadian Thoracic Society 2023 guideline risk of AECOPD classification (10). The rationale for this 12-month analysis was to control for seasonal variations in the rate of exacerbations.

Statistical Analysis

For baseline retrospective cohort comparisons, we used chi-square tests and Mann-Whitney or Wilcoxon rank-sum tests as appropriate. We conducted Cox proportional hazard models for the hazard ratio (HR) of all-cause mortality before and after the pandemic. Given the evidence of overdispersion in our data, we used negative binomial regression models for the IRR of hospitalization and emergency department and outpatient visits. To calculate the adjusted HR (adj-HR) and IRR, we included age, sex, COPD duration, risk of AECOPD based on history of hospitalization (binary outcome) and number of moderate exacerbations from the previous year, rural residency status, income, and material deprivation index as covariates. The Charlson comorbidity score was additionally included to determine adj-HR. We further undertook stratified analyses based on sex. Statistical significance was defined as P < 0.05. All statistical analyses were conducted using SAS v.9.4.

Results

Baseline Demographics

There were 265,589 unique pwCOPD identified in the entire 36-month study period, of whom 216,629 were prevalent cases and 48,960 were incident cases (Figure E1 and Table E3). There were more males represented in both the prevalent and incident cohorts in the pre- and postpandemic study periods, consistent with the well-established male predominance of COPD (11). The average age of the prevalent cohort was 66.8 (standard deviation [SD], 12.5) years, whereas the average age of the incident cohort was 64.6 (SD, 13.8) years. The representation of pwCOPD from a rural setting for the prevalent and incident cohorts was 20.3% and 18.6%, respectively, whereas the average household income was $54,469 (SD, $21,491) and $54,581 (SD, $20,996), respectively. In the prevalent cohort, the age of death was younger in the postpandemic period versus the prepandemic period (76.5 yr vs. 78.0 yr; P < 0.001). There were otherwise no significant differences between the pre- versus postpandemic deceased populations. In the incident cohort, the age of diagnosis was younger in the prepandemic period than in the postpandemic period (64.4 yr vs. 64.9 yr). There were no differences in the proportion of rural residency in the incident cohort before versus after the pandemic, but there was a small increase in average income in the incident cohort after the pandemic.

The baseline demographics of the female and male cohorts in our study demonstrated that females were on average older than men (67.3 yr vs. 66.4 yr; P < 0.0001), with fewer having rural residency (19.8% vs. 20.8%; P < 0.0001) and lower average income ($54,356 vs. $54,576; P = 0.017) (Table E4).

Healthcare Use

Consistent with previously published Alberta provincial data (4), general and COPD-associated emergency department visits declined during the pandemic in the prevalent, survived prevalent, and incident cohorts (Table E5), with an adjusted IRR (adj-IRR) for general and COPD-related emergency department visits of 0.85 (95% confidence interval [CI], 0.84–0.86) and 0.61 (95% CI, 0.59–0.63), respectively (Table 1). Both general and COPD-related hospitalizations were reduced during the pandemic compared with before, with an associated adj-IRR of 0.96 (95% CI, 0.94–0.97) and 0.77 (95% CI, 0.40–0.81), respectively (Table 2). Conversely, COPD-associated outpatient visits (primary care or specialists) increased during the pandemic compared with before, with an associated adj-IRR of 1.08 (95% CI, 1.07–1.09) (Table 3). Although the number of patients who sought outpatient care declined, those who sought care did so much more frequently during the pandemic.

Table 1. Emergency department visits 18 months before and after March 2020 among COPD survived prevalent cohort in Alberta using subgroup analysis by sex

 	Total (n = 191,085)	Female (n = 93,867)	Male (n = 97,218)	
Before COVID-19	After COVID-19	P Value	Before COVID-19	After COVID-19	P Value	Before COVID-19	After COVID-19	P Value	
Emergency department visits	 	 	 	 	 	 	 	 	 	
 Number of cases with ⩾1 emergency department visit, n (%)†	87,986
(46.0)	79,112
(41.4)	<0.001	45,239
(48.2)	40,362
(43.0)	<0.001	42,747
(44.0)*	38,750
(40.0)*	<0.001	
 Number of emergency department visits, n	264,721	224,662	<0.001	137,664	113,925	<0.001	127,057*	110,737*	<0.001	
 Crude IRR (95% CI)	0.85 (0.84–0.86)	<0.001	0.83 (0.81–0.84)	<0.001	0.87 (0.86–0.89)	<0.001	
 Adjusted IRR (95% CI)‡	0.85 (0.84–0.86)	<0.001	0.83 (0.82–0.85)	<0.001	0.87 (0.86–0.89)	<0.001	
COPD-associated emergency department visits	 	 	 	 	 	 	
 Number of cases with ⩾1 emergency department visit, n (%)†	10,190
(5.3)	6,193
(3.2)	<0.001	5,331
(5.7)	3,065
(3.3)	<0.001	4,859
(5.0)*	3,128
(3.2)*	<0.001	
 Number of emergency department visits, n	16,344	9,945	<0.001	8,334	4,861	<0.001	8,010*	5,084§	<0.001	
 Crude IRR (95% CI)	0.61 (0.59–0.63)	<0.001	0.58 (0.55–0.61)	<0.001	0.63 (0.60–0.67)	<0.001	
 Adjusted IRR (95% CI)‡	0.61 (0.59–0.63)	<0.001	0.59 (0.56–0.62)	<0.001	0.64 (0.61–0.67)	<0.001	
Definition of abbreviations: CI = confidence interval; COPD = chronic obstructive pulmonary disease; COVID-19 = coronavirus disease; IRR = incidence rate ratio.

* P < 0.001; denotes sex-based comparisons before versus after COVID-19.

† Percentage of survived prevalent.

‡ Adjusted for age, sex, rural residency, COPD duration, income, and material deprivation index.

§ Not significant; denotes sex-based comparisons before versus after COVID-19.

Table 2. Comparison of hospitalization rate before and after the pandemic in the COPD survived prevalent cohort analyzed by sex

 	Total (n = 191,085)	Female (n = 93,867)	Male (n = 97,218)	
 Before COVID-19	After COVID-19	P Value	 Before COVID-19	After COVID-19	P Value	 Before COVID-19	After COVID-19	P Value	
General hospitalization	 	 	 	 	 	 	 	 	 	
 Number of cases with ⩾1 hospitalization, n (%)	34,984
(18.3)	32,752
(17.1)	<0.001	17,634
(18.8)	16,210
(17.8)	<0.001	17,350
(17.8)*	16,542
(17.0)†	<0.001	
 Number of hospitalization events	60,184	57,334	<0.001	30,457	28,179	<0.001	29,727*	29,155†	0.07	
 Crude IRR (95% CI)	0.95 (0.94–0.97)	<0.001	0.93 (0.90–0.95)	<0.001	0.98 (0.96–1.01)	NS	
 Adjusted IRR (95% CI)‡	0.96 (0.94–0.97)	<0.001	0.93 (0.90–0.95)	<0.001	0.98 (0.96–1.01)	NS	
COPD-associated hospitalization	 	 	 	 	 	 	
 Number of cases with ⩾1 hospitalization, n (%)	6,155
(3.2)	4,529
(2.4)	—	3,230
(3.4)	2,260
(2.4)	—	2,925
(3.0)*	2,269
(2.3)†	—	
 Number of hospitalizations	8,885	6,755	—	4,631	3,324	—	4,254*	3,431†	—	
 Crude IRR (95% CI)	0.76 (0.73–0.79)	<0.001	0.72 (0.67–0.76)	<0.001	0.81 (0.76–0.86)	<0.001	
 Adjusted IRR (95% CI)‡	0.77 (0.74–0.81)	<0.001	0.73 (0.68–0.77)	<0.001	0.82 (0.77–0.87)	<0.001	
Definition of abbreviations: CI = confidence interval; COPD = chronic obstructive pulmonary disease; COVID-19 = coronavirus disease; IRR = incidence rate ratio; NS = not significant.

* P < 0.001; denotes sex-based comparisons.

† Not significant; denotes sex-based comparisons.

‡ Adjusted for age, sex, rural residency, COPD duration, income, and material deprivation index.

Table 3. Outpatient clinic (general practitioner and specialist; in-person and virtual) visits before and after the pandemic in the COPD survived prevalent cohort analyzed by sex

 	Total (n = 191,085)	Female (n = 93,867)	Male (n = 97,218)	
Before COVID-19	After COVID-19	P Value	Before COVID-19	After COVID-19	P Value	Before COVID-19	After COVID-19	P Value	
Number of cases with ⩾1 clinic visits, n (%)	179,581
(94.0)	176,675
(92.5)	<0.001	89,194
(95.0)	87,938
(93.7)	<0.001	90,886
(93.5)*	88,737
(91.3)*	<0.001	
Number of clinic visits, n	3,617,952	3,908,514	<0.001	1,932,842	2,111,468	<0.001	1,685,110*	2,335,591*	<0.001	
Crude IRR (95% CI)	1.08 (1.07–1.09)	<0.001	1.09 (1.08–1.10)	<0.001	1.07 (1.06–1.08)	<0.001	
Adjusted IRR (95% CI)†	1.08 (1.07–1.08)	<0.001	1.09 (1.08–1.10)	<0.001	1.06 (1.05–1.07)	<0.001	
Definition of abbreviations: CI = confidence interval; COPD = chronic obstructive pulmonary disease; COVID-19 = coronavirus disease; IRR = incidence rate ratio.

* P < 0.001; denotes sex-based comparisons.

† Adjusted for age, sex, rural residency, COPD duration, income, and material deprivation index.

All-Cause Mortality

Despite the reduction in acute care use, there was a statistically significant increase in the crude HR of mortality after COVID-19 compared with before COVID-19 of 1.1 (95% CI, 1.07–1.13), which remained significantly increased after adjusting for covariates (1.21; 95% CI, 1.18–1.24; Table 4). Seven percent of records were missing from the material deprivation index data. A sensitivity analysis was conducted with the missing data excluded. No meaningful difference in adj-HR for post- compared with prepandemic mortality was identified.

Table 4. Hazard ratio of all-cause mortality and non–coronavirus disease all-cause mortality in the COPD disease all prevalent cohort

 	Crude HR	95% CI	P Value	Adjusted HR	95% CI	P Value	
All-cause mortality* (after COVID-19 relative to before COVID-19)	 	 	 	 	 	 	
 Both sexes	1.1	1.07,1.13	<0.0001	1.21	1.18–1.24	<0.0001	
 Female	1.07	1.03–1.11	0.0003	1.19	1.15–1.23	<0.0001	
 Male	1.12	1.09–1.16	<0.0001	1.23	1.19–1.28	<0.0001	
Non–COVID-19-related all-cause mortality* (after COVID-19 relative to before COVID-19)	 	 	 	 	 	 	
 Both sexes	1.01	0.99–1.04	NS	1.11	1.09–1.14	<0.0001	
 Female	0.98	0.95–1.02	NS	1.09	1.05–1.13	0.0001	
 Male	1.04	1.0–1.07	0.045	1.14	1.10–1.18	<0.0001	
Sex-based comparison of all-cause mortality† (male relative to female)	 	 	 	 	 	 	
 Before COVID-19	1.09	1.05–1.13	<0.0001	1.14	1.10–1.18	<0.0001	
 After COVID-19	1.15	1.11–1.19	<0.0001	1.19	1.15–1.24	<0.0001	
 After COVID-19, non–COVID-19-related all-cause mortality	1.15	1.11–1.19	<0.0001	1.20	1.15–1.24	<0.0001	
Definition of abbreviations: CI = confidence interval; COPD = chronic obstructive pulmonary disease; COVID-19 = coronavirus disease; HR = hazard ratio; NS = not significant.

A total of 7% of entries were missing material deprivation index data.

* Adjusted for age, sex, Charlson comorbidity index, income, material deprivation index, COPD duration and exacerbation risk, and COPD hospitalization 1 year before March 2018.

† Adjusted for age, Charlson comorbidity index, income, material deprivation index, COPD duration and exacerbation risk, and COPD hospitalization 1 year before March 2018.

Given that pwCOPD were at increased risk of death when infected with COVID-19, mortality was also considered in the context of COVID-19–associated deaths. Of the 204,073 survived prevalent pwCOPD, 90,309 underwent COVID-19 testing, of whom 8,413 individuals tested positive (Table E6). COVID-19 test positivity conferred a 38% increased risk of all-cause mortality for pwCOPD within 30 days of testing. Repeat mortality analysis excluding COVID-19–associated deaths demonstrated no difference in crude all-cause mortality in pwCOPD before versus after the pandemic (1.01; 95% CI, 0.99–1.04; Table 4), but, after adjusting for covariates, the adj-HR for all-cause mortality increased after compared with before the pandemic (1.11; 95% CI, 1.09–1.14).

Pulmonary Exacerbations

Rates of AECOPD before versus after the pandemic were next examined in the context of COPD reexacerbation risk. The large majority of pwCOPD who were at low risk of AECOPD remained at low risk (Table E7), whereas, among those who were at high risk of AECOPD, there was a 78.8% reduction in high-risk classification during the pandemic. We next examined these categories over three epochs: April 2017–March 2018, April 2018–March 2019, and April 2020–March 2021, which showed a similar shift in the proportion of those at high risk of AECOPD to low risk during the pandemic period (Figure 1).

Figure 1. Sankey diagram illustrating the distribution of patients with chronic obstructive pulmonary disease (pwCOPD) at low versus high risk of acute exacerbation of chronic obstructive pulmonary disease (AECOPD) over three 12-month epochs among surviving COPD prevalence in Alberta. Patients in April 2019–March 2020 were designated low risk (green; defined as one or fewer moderate exacerbations in the preceding year) versus high risk of AECOPD (red; defined as two or more moderate exacerbations [i.e., event requiring course of oral prednisone or respiratory antibiotic] OR a hospitalization or emergency department visit in the preceding year). The portion of pwCOPD in each risk category is represented by the percentage of all pwCOPD, with the absolute number reported in parentheses. The membership of patients in the high-risk category dropped during the pandemic compared with the preceding 2 years, whereas membership of patients in the low-risk category increased during the pandemic. A significant number of patients who previously belonged in the high-risk category now met the definition of low risk during the pandemic.

Incidence

There was a 63.6% persistent reduction in the incidence of COPD during the pandemic (Figure E2). The incidence rate remained stable in both pre- and postpandemic periods of observation (incidence rate per 1,000 patient-years 1,345 vs. 1,330, respectively), such that the incidence rate difference was not statistically significant (−0.016; 95% CI, −0.04 to 0.026).

Sex-based Subgroup Comparisons of Healthcare Use, Incidence, and Mortality

Healthcare use for both sexes was reduced during the pandemic compared with before (Tables E1–E3). General and COPD-associated emergency department visits were significantly reduced for females and males before compared with after the pandemic, with an IRR of 0.83 (95% CI, 0.82–0.84) and 0.57 (95% CI, 0.54–0.59); and 0.86 (95% CI, 0.85–0.87) and 0.64 (95% CI, 0.61–0.66), respectively (Table 1). Before the pandemic, more females than males presented to the emergency department for general and COPD-related care. After the pandemic, more females than males presented to the emergency department for general care although there were no sex differences in the rate of COPD-related emergency department visits. Similarly, there were fewer general hospitalizations in both females (IRR, 0.91; 95% CI, 0.9–0.93) and males (IRR, 0.97; 95% CI, 0.95–0.98), with a greater relative reduction in females (Table 2). There were no significant differences in hospitalization rates between the sexes after the pandemic as there were before the pandemic. A similar pattern of reduction was seen for COPD-related hospitalizations. Regarding outpatient visits, a greater relative reduction during the pandemic was seen in females (IRR, 1.06; 95% CI, 1.05–1.07) compared with males (IRR, 1.09; 95% CI, 1.08–1.10) (Table 3). Sex-based analysis of all-cause mortality revealed an increase for both males and females during the pandemic, reflected in the increase in adj-HR 1.23 (95% CI, 1.19–1.28) and 1.19 (95% CI, 1.15–1.23), respectively (Table 4). After accounting for COVID-19–associated deaths, the adj-HR for non–COVID-19-associated all-cause mortality was increased after the pandemic for both sexes (males: 1.14; 95% CI, 1.10–1.18; females: 1.09; 95% CI, 1.05–1.13). Finally, there were reductions in COPD incidence for both males (by 62.7%) and females (by 64.5%) after the pandemic, with no difference in incident rate difference for either sex between the pre- or postpandemic periods (males: −0.022; 95% CI, −0.056 to 0.011; females: −0.009; 95% CI, −0.045 to 0.026).

Discussion

The COVID-19 pandemic has had a profound direct and indirect impact on the lives of people with pulmonary conditions. In this study, we attempted to unravel the impact of the pandemic on the healthcare use of pwCOPD before and after the start of the pandemic. By linking large population-level administrative databases, we demonstrated that acute care use (hospitalization and emergency department visits) dropped during the pandemic, consistent with literature from other high-income countries (3, 6, 12). However, there was an increase in total outpatient in-person and virtual visits. We further added to the literature by conducting a stringent assessment of the adjusted risk of mortality for pwCOPD during the pandemic by taking into account not only the known risk factors of COPD mortality but also the confounder of COVID-19 infection. COPD is a known risk factor for more severe COVID-19 outcomes and mortality (1), although it is not a risk factor for acquiring COVID-19 infection itself (13–15). Here, we have provided clear evidence that the pandemic is associated with an 11% increase in adjusted risk for non–COVID-19-associated all-cause mortality for pwCOPD, providing some analytical clarity around the conflicting findings in the mortality data (5, 6, 12) (Table 4). We suspect the ethnic diversity in our provincial data is reasonably representative of the experience of pwCOPD in most high-income countries. Collectively, these findings raise concerns that the increase in non–COVID-19-associated mortality for pwCOPD could be secondary to reduced acute care encounters, despite an increase in outpatient clinic visits, although the contribution of other changes brought on by the pandemic, such as lockdown policies or temporary holds on preventative disease screening protocols, could not be accounted for in this study.

We further critically examined the relationship of reduced acute care use and increased mortality seen during the pandemic by interrogating if this was a sex-dependent phenomenon. Our work demonstrated for the first time that there was a consistent reduction in acute care use by both sexes during the pandemic, with females demonstrating a greater relative reduction in acute care use (Tables 1–3). Sex disparities in healthcare access, as a result of a multitude of system barriers uniquely experienced by females, have worsened during the pandemic (16–18) (i.e., ability to take time off work to attend clinics, ability to secure childcare to attend clinics, affordability of medications, etc.). Intriguingly, the sex disparity flips when outpatient visits are examined, raising the possibility that outpatient care, and perhaps virtual care, may have removed some access barriers for women. Further research into the complexities of sex- and gender-based healthcare inequities is required.

We next explored whether the reduction in acute healthcare use could be driven by the larger subgroup of pwCOPD who were at low risk of AECOPD, masking the same or increased acute care usage by the smaller but higher-risk subgroup. Our novel observation that exacerbation rates dropped regardless of exacerbation risk confirms that all pwCOPD, regardless of their exacerbation risk, experienced a reduction in exacerbation during the pandemic (Figure 1 and Table E7).

It is unclear if the drop in acute care use is the result of care avoidance out of fears of viral exposure or other systemic barriers to healthcare access. It may be hypothesized that more assessments and treatments were taking place in the outpatient setting to minimize exposure risk. Thus, we attempted to address this theory by triangulating outpatient visits to general practitioners and specialists with outpatient medication dispensation data, generating a complete picture of outpatient healthcare use and treatment of AECOPDs. Given that outpatient visits increased together with a drop in dispensed courses of outpatient treatment of AECOPDs, the increased all-cause mortality may be because outpatient visits could not compensate for the need for acute care. If the increased in all-cause mortality is reflected in COPD-related mortality (data which we do not have in our database), then this may hint that outpatient care, and perhaps virtual care in particular, could have resulted in missed diagnoses or treatment opportunities for moderate-to-severe AECOPDs.

The reduced incidence of COPD comparing 18 months before and after the start of the pandemic was expected. Given the transition of in-person to virtual care during the pandemic, it may be logical to assume that in-person testing might have similarly decreased because of test deferment to mitigate viral exposure risk. Correspondingly, there was a reduction in new cases of COPD from March 2020 onward (Figure E2). It remains to be seen if the incidence of COPD will increase back to prepandemic levels later in the pandemic.

Limitations

There are a few limitations that should be considered within this study. First, with the use of provincial-level administrative data, we do not have more granular detail on the severity of COPD. As forced expiratory volume in 1 second (FEV1) is associated with COPD mortality and morbidity (19–21), it is possible that the experience of patients with low versus high FEV1 could have differed during the pandemic. We tried to address this limitation with a subgroup analysis of AECOPD exacerbation risk, which demonstrated reduction in exacerbations regardless of risk, suggesting that those with mild or severe COPD likely experienced similar outcomes during the pandemic. Another limitation of our study is the definition of outpatient diagnosis of COPD exacerbation from administrative data. Because COPD billing codes are not specific for visits that involved treatment of AECOPD versus a routine COPD follow-up visit, the use of drug-dispensing data adds further certainty of a treated exacerbation. However, this is limited to cases where medications are dispensed, thus missing cases of nonadherence. We also defined an outpatient AECOPD based on a concurrent prescription for steroids and respiratory antibiotics (i.e., moderate AECOPD). Thus, mild AECOPDs, where only inhaled therapies were adjusted, or noninfective exacerbations without the need for antibiotics would not be captured. Finally, the accessible provincial mortality database primarily collects all-cause mortality data. Consequently, we are unable to directly determine rates of COPD-related deaths. Although we attempted to address this limitation in part by adjusting for COVID-19–associated mortality, some uncertainty remains.

Conclusions

Our findings demonstrate that outpatient COPD-related encounters increased during the pandemic, possibly to compensate for reduced COPD-related acute care use in pwCOPD, but was associated with an increase in adjusted non–COVID-19–associated deaths for pwCOPD. This raises significant policy implications, as virtual care, because it comprised the majority of outpatient visits, might be inadequate to provide preventative care for pwCOPD. More research is needed to understand what contributed to this increase in mortality and what changes to clinical policy can be made to improve this outcome even beyond the pandemic.

Supported by Canadian Institutes of Health Research grant number 179875.

Author Contributions: G.Y.L. was responsible for conceptualization, development of methodology, funding acquisition, supervision, and manuscript writing. C.W. was responsible for data acquisition and data analysis. J.A.B. was responsible for supervision and data acquisition. M.B., L.J.J.S., and D.P.G. were responsible for manuscript editing and review. M.K.S. and P.E.R. were responsible for development of methodology, manuscript editing, and review. J.W. was responsible for development of methodology, funding acquisition, manuscript editing, and review. All authors read and approved the final manuscript.

Availability of data and materials: The datasets used and analyzed in the current study are available from the corresponding author upon reasonable request in accordance with the regulations set forth by the University of Alberta REB and Alberta Health Services.

This article has a data supplement, which is accessible at the Supplements tab.

Author disclosures are available with the text of this article at www.atsjournals.org.
==== Refs
References

1 Gerayeli FV Milne S Cheung C Li X Yang CWT Tam A et al. COPD and the risk of poor outcomes in COVID-19: a systematic review and meta-analysis EClinicalMedicine 2021 33 100789 33758801
2 Alqahtani JS Oyelade T Aldhahir AM Mendes RG Alghamdi SM Miravitlles M et al. Reduction in hospitalised COPD exacerbations during COVID-19: a systematic review and meta-analysis PLoS One 2021 16 e0255659 34343205
3 So JY O’Hara NN Kenaa B Williams JG deBorja CL Slejko JF et al. Population decline in COPD admissions during the COVID-19 pandemic associated with lower burden of community respiratory viral infections Am J Med 2021 134 1252 1259.e3 34126098
4 Rennert-May E Leal J Thanh NX Lang E Dowling S Manns B et al. The impact of COVID-19 on hospital admissions and emergency department visits: a population-based study PLoS One 2021 16 e0252441 34061888
5 Alsallakh MA Sivakumaran S Kennedy S Vasileiou E Lyons RA Robertson C et al. EAVE II Collaborators Impact of COVID-19 lockdown on the incidence and mortality of acute exacerbations of chronic obstructive pulmonary disease: national interrupted time series analyses for Scotland and Wales BMC Med 2021 19 124 33993870
6 Fedeli U Barbiellini Amidei C Marcon A Casotto V Grippo F Grande E et al. Mortality related to chronic obstructive pulmonary disease during the COVID-19 pandemic: an analysis of multiple causes of death through different epidemic waves in Veneto, Italy Int J Environ Res Public Health 2022 19 12844 36232144
7 Gershon AS Wang C Guan J Vasilevska-Ristovska J Cicutto L To T Identifying individuals with physcian diagnosed COPD in health administrative databases COPD 2009 6 388 394 19863368
8 Gershon AS Wang C Wilton AS Raut R To T Trends in chronic obstructive pulmonary disease prevalence, incidence, and mortality in Ontario, Canada, 1996 to 2007: a population-based study Arch Intern Med 2010 170 560 565 20308643
9 Mapel DW Roberts MH Sama S Bobbili PJ Cheng WY Duh MS et al. Development and validation of a healthcare utilization-based algorithm to identify acute exacerbations of chronic obstructive pulmonary disease Int J Chron Obstruct Pulmon Dis 2021 16 1687 1698 34135580
10 Bourbeau J Bhutani M Hernandez P Aaron SD Beauchesne MF Kermelly SB et al. 2023 Canadian Thoracic Society guideline on pharmacotherapy in patients with stable COPD Chest 2023 164 1159 1183 37690008
11 Halbert RJ Natoli JL Gano A Badamgarav E Buist AS Mannino DM Global burden of COPD: systematic review and meta-analysis Eur Respir J 2006 28 523 532 16611654
12 Kendzerska T Zhu DT Pugliese M Manuel D Sadatsafavi M Povitz M et al. Trends in all-cause mortality and inpatient and outpatient visits for ambulatory care sensitive conditions during the first year of the COVID-19 pandemic: a population-based study J Hosp Med 2022 17 726 737 35929531
13 Argenziano MG Bruce SL Slater CL Tiao JR Baldwin MR Barr RG et al. Characterization and clinical course of 1000 patients with coronavirus disease 2019 in New York: retrospective case series BMJ 2020 369 m1996 32471884
14 Alqahtani JS Oyelade T Aldhahir AM Alghamdi SM Almehmadi M Alqahtani AS et al. Prevalence, severity and mortality associated with COPD and smoking in patients with COVID-19: a rapid systematic review and meta-analysis PLoS One 2020 15 e0233147 32392262
15 Guan WJ Ni ZY Hu Y Liang WH Ou CQ He JX et al. China Medical Treatment Expert Group for COVID-19 Clinical characteristics of coronavirus disease 2019 in China N Engl J Med 2020 382 1708 1720 32109013
16 Guerrina R Borisch B Callahan LF Howick J Reginster JY Mobasheri A Health and gender inequalities of the COVID-19 pandemic: adverse impacts on women’s health, wealth and social welfare Front Glob Women’s Health 2021 2 670310 34816222
17 Aleksanyan Y Weinman JP Women, men and COVID-19 Soc Sci Med 2022 294 114698 34999529
18 Bambra C Albani V Franklin P COVID-19 and the gender health paradox Scand J Public Health 2021 49 17 26 33308006
19 Bikov A Lange P Anderson JA Brook RD Calverley PMA Celli BR et al. FEV1 is a stronger mortality predictor than FVC in patients with moderate COPD and with an increased risk for cardiovascular disease Int J Chron Obstruct Pulmon Dis 2020 15 1135 1142 32547001
20 Duong M Islam S Rangarajan S Leong D Kurmi O Teo K et al. PURE investigators Mortality and cardiovascular and respiratory morbidity in individuals with impaired FEV1 (PURE): an international, community-based cohort study Lancet Glob Health 2019 7 e613 e623 31000131
21 Ching S-M Chia YC Lentjes MAH Luben R Wareham N Khaw KT FEV1 and total cardiovascular mortality and morbidity over an 18 years follow-up population-based prospective EPIC-NORFOLK study BMC Public Health 2019 19 501 31053065
