Emergency department visits in Stockholm – a projection ahead

Technical Summary: Emergency Department Visits in Stockholm – A Projection Ahead

Problem Statement
Emergency departments (EDs) in Sweden, particularly in Stockholm, face increasing pressure driven by population growth and demographic shifts. While EDs operate under constant strain due to unpredictable patient inflows, leading to crowding and compromised patient safety, there is a lack of age-specific projections regarding future ED demand. Existing literature addresses short-term fluctuations or international trends, but there are no published estimates quantifying how projected population growth, specifically the aging of the population, will affect the long-term volume of ED visits in the Stockholm region. This study aims to fill that gap by estimating future ED visit trends through 2035 to inform resource planning and ensure equitable healthcare provision.

Methodology
The study utilizes a retrospective and prospective observational design based on aggregated administrative data:

• Data Sources: Age-stratified ED visit counts for Stockholm County were extracted from national registers (National Board of Health and Welfare) covering 2016–2024. Population counts and projections through 2035 were sourced from Statistics Sweden.
• Stratification: Data were aggregated into five age groups: 20–34, 35–49, 50–64, 65–79, and ≥80 years.
• Statistical Modeling: Separate Poisson regression models were fitted for each age group to project utilization from 2025 to 2035. The natural logarithm of population size was included as an offset to account for population variation.
• Temporal Structure: A piecewise linear spline with knots at 2019 and 2021 was specified to capture pre-pandemic trends, the disruption caused by the COVID-19 pandemic, and the subsequent recovery. Models utilized robust (sandwich) standard errors to address potential overdispersion.
• Validation and Sensitivity: Model fit was assessed by comparing predicted counts against observed values (2016–2024). A sensitivity analysis was conducted comparing the main two-knot specification against models with no knots, a single knot at 2020, and a two-knot specification at 2020 and 2022.
• Analysis Tools: All analyses were performed using Stata 18.

Key Results
The study projects a substantial divergence in ED utilization trends based on age:

• Historical Trends (2016–2024): ED visit rates generally declined across all age groups from 2016 to 2022. Between 2022 and 2024, rates increased for the 20–34, 35–49, and 65–79 age groups, while remaining stable for 50–64 and decreasing for those ≥80.
• Projections to 2030: Significant growth is projected for older adults. Visits for the 65–79 age group are expected to increase by 31.8% (95% CI: 5.9, 64.1), and for those ≥80 by 42.6% (95% CI: 16.1, 75.2).
• Projections to 2035: The growth trajectory accelerates. By 2035, ED visits for the 65–79 age group are projected to rise by 71.9%, and for those ≥80 by 69.6%, relative to 2024 baseline levels.
• Stable and Modest Trends: The 50–64 age group is projected to remain stable. Younger groups (20–34 and 35–49) show modest projected increases (37.1% and 18.0% respectively by 2035).
• Uncertainty: Confidence intervals widen notably after 2028, particularly for the oldest age groups, reflecting increasing uncertainty in long-term population forecasts.
• Sensitivity Analysis: Models without knots or with a single knot at 2020 failed to capture the post-2021 rise in visits observed in the data, resulting in flat or declining projections that diverged from reality. The two-knot specification (2019, 2021) provided the best fit for the observed post-pandemic recovery trend.

Significance and Claims
The authors position this study as a pragmatic, policy-relevant contribution to healthcare planning in an urban setting with high care access but complex referral systems.

• Planning Utility: The study claims that while long-term projections (beyond 2030) should be interpreted with caution due to widening confidence intervals, the clear demographic trend of aging provides a predictable basis for planning. The projections serve as "useful planning estimates rather than precise forecasts."
• Systemic Implications: The paper argues that the projected rise in ED demand, driven primarily by an aging population with higher rates of multimorbidity and frailty, threatens patient safety and staff working environments. It highlights that without adaptation, the system risks exacerbating ED crowding, which is already linked to increased mortality and medical errors in Swedish contexts.
• Strategic Recommendations: Based on the findings, the authors claim that maintaining safe and equitable emergency care requires specific adaptations: strengthening geriatric-adapted emergency services, improving coordination across care levels (primary, municipal, and hospital), and expanding alternatives to ED visits. They emphasize that interventions should focus on preventative strategies outside the ED, as hospital-based interventions alone have shown limited impact on reducing utilization.
• Methodological Contribution: The study demonstrates a framework for using publicly available administrative data to support future healthcare planning, a method applicable to other cities facing similar demographic and structural challenges.

The paper concludes that the anticipated exacerbation of ED pressure is a priority for health system planning, necessitating proactive reforms to uphold principles of patient safety and equitable access.