Managerial Epidemiology

In preparation for writing a text on managerial epidemiology more than a decade ago, I asked one well-respected hospital CEO whether he thought epidemiology was important for hospital managers; his response was “we have a nurse that does that!” I am not sure that the value of epidemiology as a set of tools that can improve decision making among health care executives has increased all that much in the past decade, and this notwithstanding the requirement that epidemiology be incorporated in Master of Health Administration curricula for many years. Managerial epidemiology can be defined as the “application of the tools and principles of epidemiology to the decision-making process” ¹ within health care settings.

In our first text, we tried to advance the argument that health care managers should “cultivate an epidemiologic perspective.” The decision-making process for each of the functional domains of a heath care manager (ie, planning, controlling, staffing, directing, organizing) would be improved by such a perspective. For example, strategic planning and needs assessment must consider the present and future burden of disease (measured by what epidemiologists call “prevalence”) and the burden of risk factors, which can translate into subsequent disease, by a factor that epidemiologists call relative risk. I consider it virtually impossible to develop a comprehensive strategic plan without incorporating estimates of the prevalence of disease.

The health care landscape has undergone significant change over the past century, from the introduction of insurance, to the landmark Medicare and Medicaid legislation, to the prospective payment reimbursement system to the parade of health care organization structures, including health maintenance organizations, preferred provider organizations, health insurance exchanges, and many others. The underlying sickness model, however, remains the same. People do not remain healthy forever and eventually succumb to various acute or chronic diseases. The severity of disease with which they access the health care system, whether it be to primary care providers or at acute care settings, will doubtless depend on the extent to which they have insurance coverage. The overall burden of disease, particularly severe disease, should therefore depend on the access to care across population groups. Less access probably means more severe disease, because the “effective” price of seeking care is much higher among those without insurance or among those who face serious barriers to care. The tools of epidemiology provide critical information for managers and planners seeking to predict future demand for services amid the current insurance markets; public payer beneficiaries are struggling to find doctors willing to work for less, and private insurance clients face more cost sharing in terms of higher monthly premiums and copayments.

We have seen a trend in the past decade or two in relying on the “evidence” to make decisions. So evidence-based medicine, which some have referred to as clinical epidemiology, would have clinicians making diagnostic and therapeutic decisions, and providing prognostic estimates based on evidence from the literature with a goal of balancing the “art” and “science” of medicine; with the latter presumably based on epidemiologic and clinical evidence from the literature. Much of this in encapsulated into clinical practice guidelines, such as the Physician Data Inquiry guidelines of the National Cancer Institute and those guidelines assembled by the National Comprehensive Cancer Network. “Evidence-based management” presumes that managers, even health care managers, make decisions based on solid evidence. Such evidence can include the dollars and cents associated with finance and accounting, and the strategies associated with diplomacy or leadership. I would argue that managers must also base their decisions on sound and insightful epidemiologic evidence.

In primary care settings, such evidence can come from a variety of sources. For clinicians, there are practice care patterns and guidelines promulgated by professional organizations such as the American Academy of Family Physicians. Managerial epidemiologic research, such as that which uses epidemiologic studies designs such as the case–control and cohort studies to engage in health services research and look at issues of cost, quality and access to care are a bit more difficult to find. The paucity of such research relates partly to the fact that most primary care physicians are not trained to be researchers or even epidemiologists for that matter, and partly to the lack of professional or financial motivation to engage in such activity. There is some evidence that epidemiologic or quasi-experimental study designs can work well in primary care settings such as pre–post studies to evaluate quality improvement interventions or case–control studies looking at specific determinants of disease.

A number of primary care research networks reach across the United States. Large integrated systems with many primary care affiliates such as Mayo Clinic, Cleveland Clinic, Geisinger Health System, and Kaiser Permanente should be urged to become not only consumers of evidence but also producers inasmuch as their research units could contribute substantially to the primary care evidence base. Practiced-based research networks, such as the Kentucky Ambulatory Network with more than 280 community-based clinicians from 68 Kentucky counties, provide another wonderful opportunity to facilitate primary care research and promulgate the evidence on which practicing clinicians rely. Even moderate sized group practices can conduct small cross-sectional, case–control, or cohort studies. The challenge is to match eager researchers with ambitious, innovative, and visionary primary care physicians to make such research happen.

So how do we make this all happen? From the consumer side, we can encourage primary care physicians to become consumers of epidemiologic research in their diagnostic and therapeutic decision making, and embrace the tenets of evidence-based medicine. The electronic medical record will make this process much easier. From the producer side, we need to consider how to match researchers with willing physicians around researchable topics. Such topics can arise from either practicing clinicians or researchers. For example, we have found that people in Appalachian Kentucky without any comorbid illness report proportionately less guideline concordant colonoscopy than those with some morbidities. A single primary care physician or group practice could have reached a similar conclusion through a rather simple record review. In either case, the challenge would be to assemble teams of researcher and clinicians to carry out this simple research in practice settings where physicians already have many competing demands on their time.