More Hospital Beds, More Avoidable Hospitalizations: China’s Primary Care Gap

Every health system faces a version of the same paradox: build more hospital capacity, and the hospitals fill up. A new longitudinal study from China now provides some of the clearest quantitative evidence yet that this is not merely an anecdote but a measurable, system-wide phenomenon — and that the missing ingredient is primary care. Researchers at Sun Yat-sen University in Guangzhou analyzed eight years of administrative data covering more than 126 million residents across 121 county-level jurisdictions in Guangdong Province, and found that every expansion of hospital bed supply was consistently accompanied by higher rates of hospitalizations that should, in principle, have been preventable.

The study, published in BMC Health Services Research, focused on what health services researchers call ambulatory care sensitive conditions, or ACSCs. These are illnesses — diabetes, hypertension, chronic heart failure, chronic obstructive pulmonary disease, and asthma — for which good outpatient management, early intervention, and continuous follow-up can usually keep patients out of the hospital. High hospitalization rates for these conditions are widely used as a sentinel indicator of primary care weakness: when the front line of the health system works, fewer people with unstable blood sugar, fluid overload, or wheezing end up in an inpatient bed. When it does not, admissions climb.

Between 2016 and 2023, the mean hospital bed density across the Guangdong jurisdictions rose by 13.98 percent. Primary care physician and assistant physician density, by contrast, grew by only 6.00 percent over the same period, reaching just 0.53 per 1,000 population in 2023. That asymmetry — inpatient capacity expanding roughly twice as fast as the outpatient workforce — is the structural backdrop against which the study’s central finding emerges. Using generalized estimating equations, a statistical framework designed for repeated measures over time, the team calculated incidence rate ratios linking bed supply to hospitalization rates for each of the five conditions.

The results were strikingly uniform. Higher bed density was associated with increased hospitalization rates across all five ACSCs, with incidence rate ratios ranging from 1.04 to 1.13. In practical terms, a jurisdiction with greater bed availability experienced measurably more admissions for diabetes, heart failure, COPD, asthma, and hypertension than a comparable jurisdiction with fewer beds per resident — even after accounting for the repeated-measures structure of the panel data. The finding echoes a proposition known since the 1970s as Roemer’s Law: a built hospital bed tends to get filled. What makes the new study notable is that it quantifies the effect across a vast population and then goes a step further, asking what would blunt it.

The answer, the researchers show, is primary care staffing — but at densities far beyond anything currently on the ground. The team modeled the interaction between bed supply and the density of primary care physicians and assistant physicians, and then computed what they call the offset threshold: the workforce density at which the association between bed supply and hospitalization rates would be reduced to null. For diabetes, that threshold came out at 2.88 per 1,000 population; for chronic heart failure, 1.56 per 1,000. Primary care supply significantly moderated the bed-supply effect for these two conditions, meaning the statistical interaction was real and protective in direction.

Here is the sobering part: those thresholds are approximately three to five times higher than the current density of primary care physicians and assistant physicians in the study area, which stood at 0.53 per 1,000 in 2023. Even for heart failure, the condition with the lower threshold, the existing workforce would need to nearly triple to reach the level at which primary care could fully neutralize the demand induced by hospital bed supply. For diabetes, the gap is wider still. In other words, primary care in Guangdong is not merely underfunded relative to hospital expansion — it is structurally insufficient to perform the gatekeeping function that health policy assumes it performs.

The technical design of the study deserves attention, because the strength of the conclusion rests on it. The researchers used aggregated administrative data at the county level, a longitudinal panel spanning 2016 to 2023, and generalized estimating equations to model hospitalization rates while accounting for within-jurisdiction correlation across years. Model specification was guided by the quasi-likelihood under independence model criterion, a standard tool for selecting among candidate correlation structures in GEE analysis. The outcomes were hospitalization rates for the five ACSCs, identified using the International Statistical Classification of Diseases, Tenth Revision. Because the analysis used de-identified, aggregated data, the institutional review board of Sun Yat-sen University’s School of Public Health waived the requirement for informed consent, and the protocol was approved in accordance with the Declaration of Helsinki.

Why does excess bed supply translate into avoidable admissions? The mechanisms are well understood in health services research. Beds create the physical capacity for admission; hospital payment systems and professional norms create the incentive to use it; and patients with poorly controlled chronic disease provide the clinical raw material. A patient whose diabetes is destabilized because no one is monitoring it, or whose heart failure decompensates without timely diuretic adjustment, becomes an admission once a bed is available. Primary care physicians and assistant physicians interrupt this chain at its origin — through medication management, early recognition of deterioration, and continuity of follow-up — which is precisely why their density moderates the bed-supply effect. But when their numbers are too thin, the interruption never happens, and the built capacity absorbs the demand.

The policy implications reach well beyond Guangdong. China’s health reforms have invested heavily in hospital infrastructure, and the study’s finding that bed expansion is consistently associated with higher avoidable hospitalizations suggests that this investment may be generating utilization that adds cost without adding health. The authors argue that health reforms should prioritize parallel investments in primary care staffing to bridge the gap between existing resources and the thresholds required for sustainable population health management. The arithmetic they provide makes the scale of that investment concrete: a workforce that must grow roughly threefold to fivefold, depending on the condition, before gatekeeping becomes statistically effective.

For a health system of more than 126 million residents in the study area alone — and for the many other systems, in China and elsewhere, that have expanded inpatient capacity faster than outpatient capacity — the message is hard to ignore. Hospital beds, the study shows, are not neutral infrastructure: their availability is measurably associated with admissions that better primary care could prevent. The research was supported by the National Natural Science Foundation of China, and the authors report no competing interests. As China and other rapidly expanding health systems confront aging populations and rising chronic disease burden, the study suggests that the most consequential decision may not be how many beds to build, but how many primary care physicians and assistant physicians to deploy alongside them — at densities that current policy has yet to approach.

Subject of Research: Association of hospital bed and primary care supply with avoidable hospitalizations in China: a longitudinal panel study, 2016–2023

Article Title: Association of hospital bed and primary care supply with avoidable hospitalizations in China: a longitudinal panel study, 2016–2023

Article References: Chen, S., Zhu, Y., Li, Z., Wang, Y., Chen, L., & Huang, Y. (2026). Association of hospital bed and primary care supply with avoidable hospitalizations in China: a longitudinal panel study, 2016–2023. BMC Health Services Research. https://doi.org/10.1186/s12913-026-15663-z

Image Credits: AI Generated

DOI: 10.1186/s12913-026-15663-z

Keywords: Association, hospital, primary, care, supply, avoidable, hospitalizations, China, longitudinal, panel, scientific research