Loading field map
Health services research (HSR) asks a deceptively simple question: how can health care be organized, financed, and delivered to produce the best possible outcomes for populations? The question is simple, but the answers have generated decades of productive disagreement. Since the 1960s, researchers have approached the problem from different angles—some focusing on the social structure of hospitals, others on economic incentives, still others on the measurement of quality or the reduction of medical errors. Each framework emerged because earlier approaches left something important unexamined. The result is a field that remains deeply multidisciplinary, held together by a shared commitment to improving systems rather than just treating individual patients.
The modern history of HSR begins with Medical Sociology and Organizational Theory, a framework that treated health care organizations as social systems. Drawing on the work of sociologists like Eliot Freidson and organizational theorists like Paul Lawrence and Jay Lorsch, this perspective examined how hospitals, clinics, and professional networks functioned as complex institutions with their own norms, hierarchies, and power structures. It asked questions that clinical medicine ignored: how do professional relationships shape decision-making? What happens when organizational goals conflict with patient needs? This framework established the foundational insight that health care delivery is not just a technical process but a social one—an insight that later frameworks would build on, narrow, or, in some cases, forget.
Just a few years later, Avedis Donabedian introduced the Structure-Process-Outcome Model (1966), which offered a radically different way to think about quality. Instead of analyzing organizations as social systems, Donabedian proposed a simple, measurable framework: quality could be assessed by examining the structures of care (facilities, staffing, equipment), the processes of care (what clinicians actually do), and the outcomes of care (mortality, morbidity, patient satisfaction). This model did not reject organizational theory so much as narrow its focus. Where Medical Sociology asked how power and culture shaped institutions, the Structure-Process-Outcome Model asked how specific inputs and activities produced measurable results. It became the dominant framework for quality assessment and accreditation, and it remains influential today in performance measurement and hospital rating systems.
At nearly the same moment, Health Economics (1970) introduced a very different set of concerns. Drawing on microeconomic theory, this framework treated health care as a market—albeit a highly imperfect one—and asked how incentives, prices, and regulations affected behavior. Health economists studied insurance design, physician payment models, hospital competition, and the cost-effectiveness of treatments. The tension with earlier frameworks was immediate and lasting. Where Medical Sociology saw professional norms and organizational culture, Health Economics saw principal-agent problems and moral hazard. Where the Structure-Process-Outcome Model focused on clinical quality, Health Economics focused on efficiency and resource allocation. This tension between quality and cost, between clinical judgment and market incentives, became the central dialectic of the field.
Practice Variation and Appropriateness Research (1970) emerged from a different kind of dissatisfaction. Researchers like John Wennberg and Jack Wennberg (no relation) used small-area analysis to show that medical practice varied dramatically across geographic regions, even after adjusting for patient characteristics. A patient in one town might be twice as likely to receive a prostatectomy or a hysterectomy as a patient in a neighboring town, with no apparent difference in health outcomes. This finding destabilized the assumption that clinical decisions were driven by science. If practice varied so widely, then much of medicine was not evidence-based but rather driven by local custom, physician supply, or financial incentives. Practice variation research did not directly challenge Health Economics or the Structure-Process-Outcome Model, but it exposed a problem that neither had fully addressed: the gap between what clinicians did and what the evidence supported.
By the 1990s, the problem of unwarranted variation had become impossible to ignore. Health Technology Assessment (HTA, 1990) emerged as a direct institutional response. HTA aimed to systematically evaluate the clinical effectiveness, cost-effectiveness, and broader social implications of medical technologies—drugs, devices, procedures, and even organizational interventions. Where Practice Variation research had documented the problem, HTA tried to build the evidence base that would reduce it. It drew heavily on the methods of Health Economics (cost-effectiveness analysis) and clinical epidemiology (systematic reviews and meta-analyses), but it also incorporated ethical and social considerations. HTA did not replace earlier frameworks; it absorbed their methods and redirected them toward a practical goal: informing coverage decisions, clinical guidelines, and health policy.
A different kind of response came from Patient Safety and Quality Improvement Science (1999), which was galvanized by the Institute of Medicine's landmark report To Err Is Human. This framework revived the systems-thinking of 1960s organizational theory but applied it to a specific problem: medical errors. Where earlier organizational theory had studied hospitals as social systems, Patient Safety research analyzed them as high-risk systems prone to failure. It borrowed concepts from human factors engineering, aviation safety, and cognitive psychology to understand why errors occurred and how to prevent them. Root cause analysis, checklists, and simulation training became its signature tools. The framework coexisted with HTA and the Structure-Process-Outcome Model, but it shifted the focus from measuring quality to actively improving safety—a distinction that mattered deeply to clinicians and patients.
Comparative Effectiveness Research (CER, 2000) emerged from a growing recognition that clinical trials, while rigorous, often failed to answer the questions that mattered most in real-world practice. A drug might work in a tightly controlled trial but perform differently in a diverse patient population with multiple chronic conditions. CER directly addressed the gap that Practice Variation research had exposed: the need for evidence that could guide decisions in actual clinical settings. It used pragmatic trials, observational studies, and patient-reported outcomes to compare treatments head-to-head in real-world populations. CER did not reject HTA or Health Economics, but it broadened the evidence base beyond the narrow focus on cost-effectiveness. It asked not just "Is this treatment worth the cost?" but "Which treatment works best for which patient, under what circumstances?"
The Learning Health System (LHS, 2000) represents the most ambitious integrative vision yet. Proposed by the Institute of Medicine, the LHS framework imagines a health care system that continuously generates, analyzes, and applies evidence from its own routine operations. Every patient encounter becomes a data point; every clinical decision becomes a potential source of learning. The LHS draws on all the earlier frameworks: the organizational theory insight that systems are social and complex; the Structure-Process-Outcome model's emphasis on measurement; Health Economics' concern with efficiency; HTA's commitment to evidence synthesis; Patient Safety's focus on improvement; and CER's insistence on real-world relevance. It is less a replacement for these frameworks than an attempt to integrate them into a single, self-improving system.
Today, the leading frameworks in HSR—Health Economics, Comparative Effectiveness Research, and the Learning Health System—coexist in a state of productive tension. They agree on several fundamentals: that health care systems should be studied empirically, that evidence should guide policy, and that outcomes matter more than processes. But they disagree on what counts as the most important outcome. Health Economics prioritizes efficiency and cost-effectiveness, often measured in quality-adjusted life years (QALYs). CER prioritizes real-world effectiveness and patient-centered outcomes, which may not align neatly with cost-effectiveness thresholds. The Learning Health System prioritizes continuous improvement and system integration, which requires infrastructure and data-sharing that neither Health Economics nor CER has fully solved.
The deepest disagreement remains the one that has shaped the field since the 1970s: the tension between economic efficiency and clinical quality. Health economists argue that resources are finite and that every dollar spent on one treatment is a dollar not spent on another. Quality and safety researchers argue that the goal of health care is to improve health, not to minimize costs, and that a narrow focus on efficiency can undermine patient trust and clinical judgment. The Learning Health System attempts to bridge this divide by embedding both perspectives into a continuous feedback loop, but whether such a system can reconcile the competing demands of cost control, quality improvement, and patient autonomy remains an open question.
What is clear is that no single framework has won. Medical Sociology and Organizational Theory still inform studies of professional culture and institutional change. The Structure-Process-Outcome Model remains the backbone of accreditation and performance measurement. Health Economics continues to shape payment reform and insurance design. Practice Variation research has become institutionalized in programs like the Dartmouth Atlas. HTA guides coverage decisions in many countries. Patient Safety and Quality Improvement Science has transformed hospital protocols worldwide. CER is reshaping clinical research. And the Learning Health System is the aspirational horizon toward which many health systems are working. The field's strength lies not in consensus but in the ongoing, productive friction among these frameworks—each one illuminating a different dimension of the complex problem of delivering better health care.