Workforce Attachment After Patent Foramen Ovale Closure

Highlights

• •Cryptogenic stroke after patent foramen ovale (PFO) is commonly believed to be linked to mild symptoms.
• •This study assessed workforce attachment after PFO closure in stroke patients.
• •71% of patients worked or studied 1 year after PFO closure, up from 39% at closure.
• •The proportion on sick leave dropped from 51% at closure to 15% after 1 year.
• •Disability pension use rose steadily, reaching 11% 5 years after PFO closure.

Patent foramen ovale (PFO) enables right-to-left shunting, facilitating paradoxical embolism and ischemic stroke.¹ Transcatheter PFO closure plus antithrombotic therapy reduces stroke recurrence more than antithrombotic therapy alone and is the standard care for patients <60 years with PFO-attributable stroke or transient ischemic attack (TIA). Although prior studies have assessed recurrence risk and postprocedural atrial fibrillation, workforce outcomes remain unexamined. Despite the common perception that cryptogenic strokes are mild, data on return to work are lacking. We assessed workforce attachment following PFO closure for cryptogenic stroke at 2 years.

We conducted a population-based cohort study using nationwide Danish registry data.² Individual-level data linkage was enabled by a unique personal identifier assigned to all residents. Danish health care is publicly funded and free at the point of care, with residents eligible for public benefits regardless of employment status.² Stroke care is centralized in public units with standardized protocols. The Danish Stroke Registry collects data on all stroke admissions, including severity, treatments, and quality metrics.³ Diagnostic validity in the registry is high, with positive predictive values of 90% for ischemic stroke and 79% for TIA. Using the Danish National Patient Registry, we identified patients 18–60 years with PFO closure during 2014–2021. Included patients had ischemic stroke or TIA recorded in the Stroke Registry ≤1 year before PFO closure and were employed or receiving education-related benefits 1 year before the procedure. We obtained weekly data on public transfer payments from the Danish DREAM registry, which records all payments nationwide and is valid.⁴ Individuals were categorized into seven mutually exclusive weekly states: (1) active work/study (employment, parental leave, and education grants); (2) unemployment (non–health-related); (3) sick leave; (4) other health-related benefits (e.g., flexible job schemes); (5) disability pension; (6) voluntary early retirement (available from age 50 with qualifying contributions); and (7) death. Workforce attachment included categories 1–4, while permanent exit from the workforce included categories 5–7. We estimated the 5-year cumulative incidence of permanent workforce exit and its components using Aalen–Johansen methods. Average time in the workforce and in active work/study was estimated by integrating the relevant area under the curve values, with 95% confidence intervals (CIs) derived via percentile bootstrapping (500 iterations).

The study included 684 patients (median age, 47 years [interquartile range (IQR), 40–53]; 62% were males [44% in 2020–2021]). Ischemic stroke was the most common index event (80%), with a median time to closure of 102 days (IQR, 64–153). The median Scandinavian Stroke Scale score was 58 (IQR, 52–58), with 88% of strokes classified as mild. Thrombolysis was administered to 29% of patients. Comorbidity burden was low, with 54% having a Danish Comorbidity Index for Acute Myocardial Infarction (DANCAMI) score of 0% and 32% scoring 1–3. Common comorbidities included hypertension (17%), prior stroke/TIA (14%), and migraine (10%). Atrial fibrillation (2%) and ischemic heart disease (4%) were rare. At 1 year before closure, 88% were in active work/study, and 3% were on sick leave (Figure 1). At the time of closure, 39% were in active work/study, and 51% were on sick leave. By 1, 2, and 5 years postclosure, sick leave decreased to 15%, 8%, and 8%, whereas active work/study increased to 71%, 72%, and 61%, respectively. The use of other health-related benefits and disability pension increased from 7% and 0.1% at closure to 15% and 11% at 5 years. Fewer than five patients died, and no patients retired early. The 5-year cumulative incidence of disability pension was 7.4% (95% CI, 3.1–11.6). On average, patients spent 4.8 years (95% CI, 4.7–4.8) in the workforce, including 3.4 years (95% CI, 3.3–3.5) in active work/study.

Figure 1.

Weekly prevalences of workforce attachment and death from 1 year before PFO closure until 5 years after.

Abbreviation: PFO, patent foramen ovale.

Although a direct comparison with non-PFO stroke patients was not feasible, a prior study of working-aged stroke patients reported 64% in active work/study or unemployment and 12% on disability pension at 2 years.⁵ In contrast, our study found 72% in active work/study, 4% unemployed, and 3% on disability pension at 2 years after PFO closure. These differences are likely attributable to the younger age (47 versus 51 years) and milder stroke severity in our cohort. The decline in workforce attachment began before PFO closure, suggesting that the stroke (rather than the closure) was the primary factor. The increase in other health-related benefits over time suggests that some patients may have transitioned to flexible job schemes, which may have been underrepresented in our "active work/study" estimates. One limitation is that short-term sick leave periods (<4 weeks) are not captured in the DREAM registry, potentially underestimating sick leave prevalence. However, this would not likely impact post-PFO closure estimates. Another limitation is the lack of data on job type and specific stroke etiology (e.g., Trial of Org 10172 in Acute Stroke Treatment [TOAST] classification), although all cryptogenic strokes were presumed to be embolic since the patients were referred to PFO closure.

Workforce attachment was slightly higher in patients with cryptogenic stroke treated with PFO closure compared to the general working-aged stroke population, likely due to the milder severity of stroke in this cohort. However, 30% were not actively working or studying 1 year postclosure, underscoring the substantial impact of cryptogenic stroke on patients' ability to engage in daily work activities.

Funding

Kasper Bonnesen is supported by the Danish Diabetes and Endocrine Academy (grant no. NNF22SA0079901). Nils Skajaa is supported by the Platform for Inequality Research at Aarhus University. Morten Schmidt is supported by the Novo Nordisk Foundation (grant no. NNF190C0054908). The funding sources played no part in the design and conduct of the study, analyses, or interpretation of the data.

Ethics Statement

Not applicable.

Disclosure Statement

The Department of Clinical Epidemiology at Aarhus University and Aarhus University Hospital receives funding from private and public institutions in the form of research grants to (and administered by) Aarhus University. None of these grants has any relation to this study. Nils Skajaa and Claus Z. Simonsen report receiving speaker honoraria from Pfizer outside the submitted work. Asger Andersen reports receiving speaker honoraria from ABBOTT, Gore Medical, Angiodynamics, EPS Vascular, and Jannsen, and he is a consultant for Inari Medical and Magneto Thrombectomy Solutions. Nils Skajaa was affiliated with the Department of Clinical Epidemiology at Aarhus University and Aarhus University Hospital during the initiation of this study, but is now an employee of Novo Nordisk A/S.

The other authors had no conflicts to declare.