Incidence, risk factors, and mortality of pulmonary embolism in the Netherlands (2015-22): sex differences and shifts during the coronavirus disease 2019 pandemic

Abstract

Background and aims: Epidemiology of pulmonary embolism (PE) may have shifted since the coronavirus disease 2019 (COVID-19) pandemic. This study aimed to describe temporal trends in PE epidemiology in the Netherlands since 2015.

Methods: Using nationwide data from Statistics Netherlands, all Dutch inhabitants (>16 million) without a history of PE were dynamically identified on 1 January of each year to assemble eight cohorts of PE-free Dutch inhabitants in 2015-22. They were individually followed until the end of that respective year to determine 1-year risk of PE (identified by hospital diagnoses/primary cause of death) and establish relevant risk factors. The PE cases were subsequently studied to determine 1-year all-cause mortality following PE. Multivariable logistic regression with cluster-robust standard errors and robust Poisson regression were respectively employed to evaluate relative differences in PE incidence and mortality between years.

Results: Pulmonary embolism incidence in the Dutch population decreased from 2015 to 2019 but markedly increased by 23% (95% confidence interval 20%-26%), 52% (48%-56%), and 7% (4%-9%) in 2020-22 (vs. 2019), respectively. Most traditional PE risk factors remained associated with PE in 2020-22 but generally with a weaker association. Pulmonary embolism mortality was stable until 2019 but then increased by 10% (6%-14%) in 2020 and 9% (6%-13%) in 2021, while the increase [2% (-1% to 6%)] was insignificant in 2022. The above-mentioned changes since 2020 were generally greater in males than females.

Conclusions: The seemingly favourable pre-pandemic temporal trends in PE epidemiology in the Netherlands reversed during the COVID-19 pandemic but appear to revert to pre-pandemic levels after 2022.

Keywords: COVID-19; Epidemiology; Pulmonary embolism; Sex characteristics; Temporal trend.

structured graphical abstract

Figure 1

Study design and flow diagram of the study cohorts. ^†The source population should be considered dynamic, with details presented in Supplementary data online, Figure S1. PE, pulmonary embolism

Figure 2

Temporal trends in pulmonary embolism incidence in the pulmonary embolism-free Dutch inhabitants. The first date (i.e. 1 January) of a calendar year was considered the index date (i.e. baseline) of the cohort of pulmonary embolism-free Dutch inhabitants identified in that calendar year. The bar charts show the 1-year cumulative incidences (and 95% confidence intervals, estimated by the Clopper–Pearson exact method) of pulmonary embolism (identified by a hospital diagnosis or primary cause of death). The points and error bars show the Model 3-adjusted odds ratios (and 95% confidence intervals, estimated by cluster-robust standard errors) for pulmonary embolism between cohorts with the pulmonary embolism-free Dutch inhabitants in 2019 as the reference. CI, confidence interval; OR, odds ratio; PE, pulmonary embolism

Figure 3

Temporal trends in pulmonary embolism incidence by age group and sex in the pulmonary embolism-free Dutch inhabitants. The first date (i.e. 1 January) of a calendar year was considered the index date (i.e. baseline) of the cohort of pulmonary embolism-free Dutch inhabitants identified in that calendar year. The points show the 1-year cumulative incidences of pulmonary embolism (identified by a hospital diagnosis or primary cause of death), and the error bars show the 95% confidence intervals (estimated by the Clopper–Pearson exact method). PE, pulmonary embolism

Figure 4

Temporal trends in associations of comorbidities/medical history with pulmonary embolism by sex in the pulmonary embolism-free Dutch inhabitants. The first date (i.e. 1 January) of a calendar year was considered the index date (i.e. baseline) of the cohort of pulmonary embolism-free Dutch inhabitants identified in that calendar year. The points present age-adjusted odds ratios (and 95% confidence intervals) for 1-year risk of pulmonary embolism (identified by a hospital diagnosis or primary cause of death) among individuals from the pulmonary embolism-free Dutch inhabitants with a comorbidity/medical history at baseline vs. those from the same cohort but without the comorbidity/medical history at baseline. Comorbidity/medical history was identified in a 5-year lookback period before index date. For the comorbidity/medical history venous thromboembolism, by study design, it only referred to deep vein thrombosis, and/or other types of venous thromboembolism, without including pulmonary embolism. ATE, arterial thromboembolism; CI, confidence interval; COPD, chronic obstructive pulmonary disease; CTD, systemic connective tissue disorders; GERD, gastrooesophageal reflux disease; OR, odds ratio; PE, pulmonary embolism; RMS/MHV, rheumatic mitral stenosis/mechanical heart valves; VTE, venous thromboembolism

Figure 5

Weekly incidence of pulmonary embolism and coronavirus disease 2019 in the pulmonary embolism-free Dutch inhabitants. The first date (i.e. 1 January) of a calendar year was considered the index date (i.e. baseline) of the cohort of pulmonary embolism-free Dutch inhabitants identified in that calendar year, which was also considered as the first day of the first week in that calendar year. This figure only presents events occurred in the first 52 weeks (i.e. 364 days) in each calendar year. COVID-19, coronavirus disease 2019

Figure 6

Temporal trends in 1-year all-cause mortality in the pulmonary embolism cohorts. The pulmonary embolism cohorts were individuals from the cohorts of pulmonary embolism-free Dutch inhabitants who were diagnosed with pulmonary embolism (by a hospital diagnosis or primary cause of death) within 1 year after the first date (i.e. 1 January) of the calendar years, and hence, the same calendar years were used to index the pulmonary embolism cohorts. The bar charts show the 1-year cumulative incidences (and 95% confidence intervals, estimated by the Clopper–Pearson exact method) of all-cause mortality. The points and error bars show the Model 3-adjusted risk ratios and 95% confidence intervals (estimated by Poisson regression model with a robust error variance) for all-cause mortality between cohorts with the pulmonary embolism cohort 2019 being the reference. CI, confidence interval; PE, pulmonary embolism; RR, risk ratio

Figure 7

Temporal trends in 1-year all-cause mortality by age and sex in the pulmonary embolism cohorts. The pulmonary embolism cohorts were individuals from the cohorts of pulmonary embolism-free Dutch inhabitants who were diagnosed with pulmonary embolism (by a hospital diagnosis or primary cause of death) within 1 year after the first date (i.e. 1 January) of the calendar years, and hence, the same calendar years were used to index the pulmonary embolism cohorts. The points show the 1-year cumulative incidence (and 95% confidence intervals, i.e. the error bars, estimated by the Clopper–Pearson exact method) of all-cause mortality following pulmonary embolism. PE, pulmonary embolism

Figure 8

Temporal trends in proportion of primary cause of death in individuals from the pulmonary embolism cohorts who died within 1 year following pulmonary embolism. The pulmonary embolism cohorts were individuals in the cohorts of pulmonary embolism-free Dutch inhabitants who were diagnosed with pulmonary embolism (by a hospital diagnosis or primary cause of death) within 1 year after the first date (i.e. 1 January) of the calendar years, and hence, the same calendar years were used to index the pulmonary embolism cohorts. The cause of deaths included (1) pulmonary embolism; (2) coronavirus disease 2019; (3) malignant neoplasms; (4) diseases of the circulatory system (except for pulmonary embolism); (5) diseases of the respiratory system; (6) external causes of morbidity and mortality; (7) diseases of the digestive system; (8) symptoms, signs, and abnormal clinical and laboratory findings, not elsewhere classified; (9) certain infectious and parasitic diseases; (10) diseases of the nervous system; (11) in situ/benign neoplasms, or neoplasms of uncertain or unknown behaviour; (12) diseases of the genitourinary system; (13) mental and behavioural disorders; (14) endocrine, nutritional, and metabolic diseases; (15) diseases of the musculoskeletal system and connective tissue; and (16) others. COVID-19, coronavirus disease 2019; PE, pulmonary embolism