Loading field map
Health services research (HSR) is driven by a persistent tension: whose priorities should shape how we study health care—the clinician's, the patient's, the payer's, or the public's? Over the past seven decades, the field has produced a series of frameworks that each redefined what counts as a relevant question, what methods are trustworthy, and whose outcomes matter. The story of HSR is not a linear march toward consensus but a series of competing and complementary ways of seeing the same system.
The first generation of HSR frameworks emerged from sociology and economics, disciplines that offered very different lenses on health care. Structural-Functional Analysis, dominant from the 1950s through the 1970s, treated hospitals and clinics as social systems whose parts worked together to maintain stability. It asked how roles, norms, and organizational structures contributed to the functioning of the whole. This perspective was soon challenged by Health Economics, which, beginning with Kenneth Arrow's landmark 1963 article on uncertainty and welfare, framed health care not as a social system but as a market with peculiar features—uncertainty, information asymmetry, and non-profit motives. Where structural-functionalists saw equilibrium, health economists saw inefficiency and misaligned incentives. By the late 1970s, economic reasoning had largely displaced structural-functional analysis in HSR, not because the sociological questions were wrong, but because the economic toolkit offered more precise predictions about behavior under constraints like insurance and reimbursement.
Running alongside this shift, Health Care Organizational Theory (1950–present) and Medical Sociology (1951–present) continued to examine how organizations and social contexts shape care. Health Care Organizational Theory drew on management and organizational behavior to study hospital governance, professional autonomy, and institutional change. Medical Sociology focused on the patient–clinician relationship, the social determinants of illness, and the professionalization of medicine. Both frameworks remained active, coexisting with health economics rather than being absorbed by it. They provided the conceptual infrastructure for later work on disparities and patient experience.
Two enduring models were born in the 1960s. Donabedian's Quality-of-Care Model (1966) broke quality into structure, process, and outcome, giving researchers a simple but powerful way to evaluate care. It did not compete with economics or sociology; instead, it offered a measurement framework that could be used by any discipline. Andersen's Behavioral Model of Health Services Use (1968) explained why people seek care through predisposing, enabling, and need factors. It absorbed insights from medical sociology about social determinants while remaining compatible with economic analyses of access. Both models are still widely used today, though they have been revised to include contextual and equity dimensions.
By the 1970s, HSR had begun to confront the gap between what medicine claimed to know and what was actually delivered. Health Technology Assessment (HTA, 1970–present) emerged to evaluate the clinical effectiveness and cost-effectiveness of medical devices, drugs, and procedures. HTA borrowed heavily from health economics—especially cost-effectiveness analysis—but added systematic review and evidence synthesis methods. It narrowed the scope of health economics by focusing on specific technologies rather than whole systems, and it created an institutional home for evidence-based coverage decisions.
At the same time, Practice Variation and Appropriateness Research (1973–present) revealed that medical practice varied dramatically across small geographic areas without any corresponding difference in patient health. John Wennberg's studies of small-area variations showed that the rate of procedures like tonsillectomy or prostate surgery depended more on where a patient lived than on their clinical condition. This finding challenged the assumption that physicians uniformly applied scientific evidence. Practice variation research coexisted with HTA—HTA asked what worked on average; practice variation asked why the average was so rarely achieved. Together, they shifted HSR's attention from ideal care to actual care.
The late 1980s and 1990s saw a three-way expansion of HSR's concerns. Outcomes and Comparative Effectiveness Research (CER, 1989–present) grew out of the recognition that clinical trials often studied narrow populations under ideal conditions. CER compared interventions in real-world settings, using observational data, registries, and pragmatic trials. It competed with Patient-Centered Outcomes and Experience Research (PCOR, 1995–present), which argued that effectiveness was not enough—research should measure outcomes that patients themselves value, such as functional status, symptom burden, and quality of life. PCOR was partly a reaction to CER's emphasis on clinical endpoints; it insisted that the patient's perspective should define the research question, not just the outcome list. The two frameworks remain in productive tension: CER provides population-level evidence, while PCOR ensures that evidence reflects what matters to individuals.
Health Care Disparities and Equity Research (1990–present) introduced a normative dimension that health economics had largely ignored. The 2003 Institute of Medicine report Unequal Treatment documented racial and ethnic disparities in care even after controlling for insurance and income. Disparities research directly challenged health economics' assumption that market forces would eventually equalize access; it showed that discrimination, bias, and systemic racism produced persistent inequities. The competition between equity and efficiency remains one of HSR's deepest disagreements: health economists tend to prioritize cost-effectiveness, while disparities researchers argue that equity should constrain efficiency.
The turn of the century brought two frameworks that reoriented HSR toward system-level improvement. Patient Safety and Quality Improvement Science (1999–present) was galvanized by the Institute of Medicine report To Err Is Human, which estimated that medical errors caused tens of thousands of deaths annually. This framework absorbed Donabedian's structure-process-outcome model but added a systems-engineering perspective: errors were not individual failures but consequences of poorly designed systems. It narrowed Donabedian's broad quality concept to focus on preventable harm, while coexisting with CER and PCOR by emphasizing safety as a distinct dimension of quality.
Implementation Science and Dissemination Research (2006–present) addressed the evidence-to-practice gap that practice variation research had exposed. Even when effective interventions exist, they are often not adopted in routine care. Implementation science studies the strategies—training, audit-and-feedback, organizational change—that promote uptake. It complements CER and PCOR by asking not just what works, but how to make it work in real settings. It also links directly to the sibling subfield of Implementation and Dissemination Research, which shares many of its frameworks.
Today, HSR is a pluralistic field. The most influential frameworks are Health Economics, CER, PCOR, Disparities and Equity Research, and Implementation Science. They agree on several fundamentals: evidence should be empirical, rigorous, and relevant to real-world decisions; health systems are complex and require multi-level analysis; and no single discipline can answer all questions. But they disagree sharply on priorities. Health economics and CER tend to privilege aggregate efficiency and clinical effectiveness, while PCOR and disparities research insist that patient values and equity must be primary. Implementation science sits between them, concerned with how to achieve change regardless of which outcome is prioritized. Institutional forces sustain this pluralism: funding agencies like AHRQ and PCORI explicitly support multiple frameworks, and the field's journals publish work from all traditions. No single framework has achieved dominance, and the productive friction among them continues to drive HSR forward.