Physician Workforce and Population Mortality: A Globally Validated Age-Structured Modeling Approach

Youngmin Oh, Eun‐Ju LeemedRxiv 2025 · DOI

TL;DR

This paper introduces a novel mathematical model that links physician density directly to age-specific mortality rates within a dynamic population framework. By applying this to South Korea, the study reveals that in well-resourced healthcare systems, doubling the physician intake yields only marginal improvements in population survival, suggesting that policy should shift from increasing the quantity of doctors to improving their efficiency and distribution.

What Problem Does This Paper Address?

Healthcare workforce planning is notoriously difficult. Traditional models usually rely on "utilization-based" metrics (how many services people currently use) or "needs-based" metrics (subjective estimates of health requirements). Both are prone to error over long horizons because they ignore how population dynamics and physician supply interact over time.

The authors argue that mortality is a more objective, stable, and globally comparable metric for assessing health system performance. The challenge they tackle is creating a robust, age-structured model that can predict how changes in the "doctor-to-population" ratio actually translate into lives saved across different age groups over several decades.

Methods

The researchers developed a unified model that integrates physician density into the Lotka–McKendrick framework, a standard mathematical approach for modeling age-structured population dynamics.

  • The Linkage: They modeled age-specific mortality rates ($M_p$) as an exponential function of physician density ($\rho$).
  • Endogenous Dynamics: Unlike traditional models that treat health outcomes as external factors, this model allows physician supply and population mortality to interact. As the population ages or shrinks, the physician density changes, which in turn feeds back into the mortality rates.
  • Occam’s Razor: To maintain tractability, the model makes several simplifying assumptions: physicians retire at 65, migration effects are negligible, and the "law of large numbers" averages out individual physician productivity.
  • Simulation: They simulated four policy scenarios for South Korea, ranging from a baseline intake of 3,058 new physicians per year to an aggressive expansion of 7,058 per year, projected through 2065.

Datasets

  • WHO Mortality Database: Used for global validation across six high-income countries: Japan, the United States, the United Kingdom, France, Germany, and Canada.
  • KOSIS (Korean Statistical Information Service): Provided the specific demographic, fertility, and physician workforce data for the South Korean case study.

Key Insights and Ideas

  • Diminishing Returns in Mature Systems: In countries that already have high physician accessibility, the "low-hanging fruit" of mortality reduction has already been picked. Adding more doctors results in increasingly smaller gains in life expectancy.
  • The "100+ Age Group" Paradox: Interestingly, the model showed that higher physician density is associated with higher mortality in the 100+ age group. This isn't a failure of care; rather, it's a sign of success. Better healthcare allows more people to survive to extreme old age, where mortality is naturally and inevitably higher, leading to a "concentration" of deaths in that bracket.
  • Density Increases Naturally: In South Korea, even under the baseline scenario (no expansion), physician density is projected to rise. This is due to the combination of existing training pipelines and a shrinking total population.
  • Paradigm Shift: The results suggest that for well-resourced nations, the debate shouldn't be about how many doctors we have, but where they are and how they work (e.g., urban vs. rural distribution, primary care vs. specialization).

Key Results

  • Global Validation: The model’s fit was statistically significant across all age groups in all six validated countries (adjusted $p < 0.05$).
  • Marginal Mortality Gains: Under the most aggressive expansion scenario in South Korea (7,058 intake), the reduction in mortality by 2065 was negligible for most:
  • Under 65: < 0.27% reduction.
  • Ages 85–94: < 7.31% reduction.
  • Ages 95–99: < 12.6% reduction.
  • Population Impact: There was no statistically significant difference in the total projected population size between the baseline and the aggressive expansion scenarios ($p > 0.07$). Essentially, doubling the number of new doctors does not significantly change the demographic trajectory of the country.

Limitations and Future Work

  • Mortality vs. Morbidity: The model focuses strictly on mortality (death). It does not account for "quality of life" or the management of non-fatal chronic diseases, which are increasingly important in aging societies.
  • Technological Progress: The model assumes a stable relationship between physician density and mortality, potentially overlooking how AI or new medical technologies might change physician productivity.
  • Migration: The study assumes a "closed" population. In countries with high rates of medical "brain drain" or significant immigration, the model would need to be adjusted.
  • Future Directions: The authors suggest that future iterations could incorporate morbidity indicators or focus on the redistributive efficiency of the workforce rather than just total numbers.

Summary updated

Abstract

Summary Background Evaluating long-term health workforce planning requires robust and globally valid indicators. While many existing metrics, such as patient satisfaction or healthcare utilization, are inherently uncertain over long horizons, mortality is an objective and stable metric. To analyze how physician supply impacts population health, we propose a model that links physician density to age-specific mortality rates and incorporate it to the age-sturctured dynamics. Methods We developed a unified model that links physician workforce density to age-structured mortality rates. Subsequently, we applied this model within a Lotka-McKendrick framework to simulate physician supply expansion scenarios in South Korea. The system simulates future age-specific mortality outcomes under varying physician supply scenarios, including a baseline intake of 3,058 physicians per year and an expanded intake of up to 7,058 physicians per year, projected through 2065. Findings Our model was validated using WHO mortality data including Japan, the United States, and the United Kingdom. Its validity of our model holds across all age groups in each country, as confirmed by statistical analysis with false discovery rate correction (maximum adjusted p &lt; 0.05). Using the age-structured dynamics with the model in South Korea, we confirm that future physician density increases even under the baseline scenario (3,058 physicians per year), and that projected population sizes under the baseline and the aggressive expansion scenario (7,058 physicians per year) are not statistically different ( p &gt; 0.07). Moreover, under the aggressive expansion scenario, the projected reductions in age-specific mortality rates by 2065 remain marginal: less than 0.27% for those under 65, less than 0.69% for ages 65–75, less than 2.75% for ages 75–84, less than 7.31% for ages 85–94, and less than 12.6% for ages 95–99. Interpretation In well-resourced health systems facing aging populations and persistently low fertility rates, further expansion of physician supply alone offers limited mortality benefits. Our findings suggest a paradigm shift: from quantity-driven to efficiency-focused workforce strategies. The proposed method is readily adaptable to other contries, offering a policy-relevant and outcome-oriented tool for long-term health workforce planning. Funding None. The funder had no role in study design, data collection, data analysis, data interpretation, or writing of the report. Research in context Evidence before this study We searched PubMed, Web of Science, and Google Scholar for studies published between 2000 and 2024 using terms such as “physician workforce,” “healthcare workforce planning,” “medical workforce density,” and “age-structured dynamics.” Most physician workforce planning models are either utilization-based (projected service demand) or needs-based (estimated population health requirements). These models often extrapolate current health service utilization rates or population health needs into the future, typically assuming that such values remain constant over time, despite the acknowledged difficulty of forecasting long-term demand dynamics. Additionally, methodologies differ widely across countries, which hampers direct comparisons of outcomes. Some metrics, such as patient needs or preferences, are also subjective and difficult to measure consistently. In contrast, more objective indicators like mortality rates are more readily available and comparable across settings. As far as we are aware, no published model has dynamically integrated physician density—endogenously derived from workforce supply—into an age-structured framework for projecting population mortality. Added value of this study To the best of our knowledge, this study is the first to integrate a model linking physician workforce density with age-structured mortality rates into the Lotka–McKendrick framework, capturing the interaction between physicians and populations across age groups. By modeling how changes in doctor-to-population ratios affect mortality, our approach moves beyond traditional models that treat health outcomes as exogenous. This allows novel analyses of policy scenarios, such as estimating life expectancy gains from increasing physician supply. Our model provides a unified framework to project both physician workforce dynamics and population health outcomes simultaneously. Implications of all available evidence Our findings underscore the importance of explicitly linking age-structured population dynamics with physician density and mortality outcomes. Traditional models often overlook these interdependencies, risking inaccurate forecasts and limiting the relevance of cross-national comparisons. By capturing how physician supply influences mortality across age groups, our approach enables more precise, outcome-driven workforce planning.
BibTeX
@misc{oh2025physician,
  author = {Youngmin Oh and Eun‐Ju Lee},
  title = {{Physician Workforce and Population Mortality: A Globally Validated Age-Structured Modeling Approach}},
  year = {2025},
  doi = {10.1101/2025.06.16.25329730},
  howpublished = {medRxiv},
}
Figure