The Provision of Public Health Services in the Emergency Department: “Begin with the End in Mind”
In this issue of Academic Emergency Medicine, contrasting opinions regarding the provision of public health services in the emergency department (ED) are provided by Bernstein and Haukoos1 and Kelen.2 We take a middle view, which can be summarized as, “it depends . . .”
Our view is informed by Steven Covey’s admonition to “Begin with the end in mind.”3 It is only when the desired end is defined that a system can be designed to attempt to deliver that end result. The end must be both efficacious and realistic to achieve. In a world without resource constraints, the issue of public health screening in the ED might be moot, the ED could become the “Everything Department,” and the personnel of that ED would be free to provide comprehensive care for all patients without regard to cost or access concerns, with the intent of giving patients the best care possible, whether or not those services align with the traditional mission of the ED. A naive person could review our specialty’s recent literature and conclude, from the many recent articles that advocate adding numerous screening tests in the ED, that this is the direction that our specialty is heading. However, such an “end” is clearly impractical, and strong arguments can be made that it is ill-advised.2,4 Of course in such a utopia, everybody would have easy access to a primary care provider, so there would be no need to provide these services the ED.
The public health services with which we are most concerned are screening, brief interventions, and referral to treatment (SBIRT). For those who believe that public health departments are abdicating their responsibility by not establishing a “branch office” in our EDs to perform SBIRT functions that could clearly enhance the health of the public, we would point out that public health departments are generally so poorly funded that their activities are typically limited to local epidemiology and disease surveillance, screening and treatment of sexually transmitted diseases, and monitoring restaurants for adherence to local sanitary codes. SBIRT activities are hence generally considered the domain of “primary care.”“Primary care” has been defined within a 25-page chapter by the Institute of Medicine5 as, “. . . the provision of integrated, accessible health care services by clinicians who are accountable for addressing a large majority of personal health care needs, developing a sustained partnership with patients, and practicing in the context of family and community” (italics in the original). This definition implies that clinic-based generalists typically diagnose and treat acute minor problems and chronic illness and that SBIRT is part of this mission. The fact that some conditions could be treated in either a clinic or an ED, and that some patients get their care primarily from the ED does not mean that EDs are doing primary care. Unfortunately, some ED patients simply do not receive adequate primary care: this is the underlying problem that the three commentaries in this issue attempt to address.
Public health and primary care interventions that are now carried out, or which have been proposed for the ED, can be divided into five broad categories.
Services that Both Help the Patient and Protect the Public
Such services are exemplified by those profiled by the articles of Ehrenkranz et al.6 and Al-Tayyib et al.7 One might also include screening for problem drinking and violence. Such services not only detect illness in those screened, but also permit the identification, treatment, and education of the diseased patients, so that others whom those patients may contact can potentially benefit. For example, physicians routinely test for sexually transmitted diseases (STDs) when performing pelvic exams even if the final diagnosis is an ovarian cyst (it bears noting, however, that limiting STD screening to these patients has made no sense since the advent of urinary testing8).
Services that Primarily Benefit the Patient
Examples include pneumonia vaccination for indigent elders and the disabled, disease education, evaluation of frail elders for assisted living placement, and smoking cessation. These are useful especially to patients who would not be evaluated or treated elsewhere.
The ED is a logical monitoring site for surveillance for infectious disease epidemics and bioterrorism events. This issue will not be discussed further here, because it represents a function that could be monitored in an automated fashion that would be unlikely to impact patient throughput or significantly consume ED resources.
Services that Make the Health Care System More Convenient For the Patient
Tetanus vaccination is offered to every patient with a laceration, despite the fact that waiting less than 72 hours, until the patient can get to a clinic, would add essentially no risk (unfortunately, when proper care is most crucial, EDs do not perform very well. The vast majority of high-risk patients with highly tetanus-prone wounds do not receive the correct treatment9).
Services that are Routine Parts of the ED Evaluation but have Incidental Screening Value
An example is the obtaining of blood pressure as a part of the vital signs (unfortunately, the vast majority of patients who have their asymptomatic hypertension discovered in the ED do not subsequently receive appropriate treatment10).
ED patients are drawn from a group with important differences from the population at large. ED patients are more likely to be poor and less likely to be insured. This is important because income relates directly to health outcomes.11 Emergency patients may be appropriate targets for preventative interventions because many of them cannot or will not negotiate the health care system of their communities in the manner envisioned by public health authorities. Health care is fragmented in the United States. Follow-up referrals must be made while mindful of the patient’s insurance status and type (if any), geographic residence, social factors, and local resources. This task is difficult even for the well-off. Referral patterns will be different for the abused elder with Medicare coverage than for the commercially insured youth with chlamydia. However, the obtaining of appropriate office-based care is particularly complex for socially marginalized indigent patients without a primary care home. Public health workers correctly argue that if certain services are not provided in the ED, they will not be received by the patient anywhere—even if they are available in other settings.
The amount of time that could be devoted to the public health intervention of prevention is daunting. The Agency for Healthcare Research and Quality currently recommends 70 preventative services.12 Yarnall13 analyzed a previous 34-item list published by The U.S. Preventative Services Task Force14 and estimated that it would take a primary care provider approximately 45 minutes per patient per year to deliver this smaller list of services under optimal circumstances. Clearly, this cannot be done in the ED with current staffing.
We believe decisions regarding which public health/primary care interventions should be done in EDs should be based on research and informed by economic considerations. The economic question is what will be the cost per outcome? In other words, what is the direct and indirect cost per quality- or disability-adjusted life-year (QUALY or DALY), or more simply, what is the true cost/benefit ratio measured as a change in patient health status? This answer will depend heavily upon the screening system used and the identity of the screener. In most departments, the emergency physicians are both the most scarce and the most costly resource. Patient kiosks and physician extenders are a less expensive and potentially better resource than emergency physicians. Most physicians, for instance, are grateful when social workers can negotiate for psychiatric placement on behalf of their patients, which increases efficiency by permitting the physician to diagnose and treat the other patients in the department. Cost must not be viewed narrowly. Although some physician advice can be integrated into the current ED workup, physicians should be cautious when told that this multitasking will have no effect on the care of the emergent problem—the brain is simply not good at doing two things at once.15
Perhaps the solution will be to have public health or primary care providers (which may include but not be limited to other physicians) in or near to the ED. The question of financing, and how such financing would impact the provision of other health care, should also be addressed. Since evaluation and management coding for the emergency physician’s time and effort makes it difficult to generate significant additional revenue from SBIRT and other public health activities, having a separate provider carry out these functions would allow additional billing under counseling codes. This might make public health interventions in the ED self-funding, but it might also engender conflict with primary care providers. Last, ED screening has no benefit if patients do not get treated. For most conditions, this requires a robust, accessible, and effective outpatient primary care system.
Currently, the selection of public health services provided in EDs is governed more by tradition, provider interest, government mandates, and committee recommendations than by logic or by cost/benefit considerations. There is a desperate need for better evidence and more consistent application of logic to guide us. Fundamentally we need to practice not only evidence-based emergency medicine, but also evidence-based emergency department public health, in a manner that measurably improves patient outcomes at an acceptable cost. Discovering what this is will require a major research effort.
At the outset, we stated our belief that the answer to the question of what public health services should be provided in the ED is, “It depends . . .” The appropriate course of action should depend upon what evidence tells us regarding which interventions delivered in the emergency setting are cost effective for which patients and for society, when integrated with value judgments (such as Item 1, below) about such services. This engenders a complex set of questions that may have different answers for different settings and populations, but answering them is crucial to make rational decisions about what nonacute health care interventions should be done in the ED. As an example, recent investigations have evaluated SBIRT effectiveness for alcohol use.16 This kind of effort would be enhanced by adding cost data to better inform the analysis.
This is the “bottom line,” as we see it:
- 1Public health interventions that benefit society by preventing others from falling victim to illness suffered by a current ED patient, or from suffering a preventable injury potentially caused by a current ED patient, should probably have highest priority.
- 2Public health and primary care providers should be encouraged to partner with EDs for problems that lie in the domain of each. It should not be wholly up to EDs and their personnel to address this set of problems that a public health department is best suited to address.
- 3Systems must be developed so that medical problems identified by ED screening are appropriately addressed and followed-up.
- 4ED care, especially these public health initiatives, cannot be viewed as separate from the overall care available to the patient.
- 5More research, using resource expenditures and patient outcomes as the endpoints, is needed to provide cost/benefit data to inform the choices to be made between public health services offered and not offered in an ED.