ABSTRACT
INTRO: How would universal health coverage change the volume of patient visits, the acuity of encounters, and the structure of residency training in a typical academic ED? To put this question another way, if the current structure of health care funding in the United States is partially responsible for the growth of academic emergency medicine and residency training, how will changes to that funding system impact medical education within the field? In order to enter the national debate regarding access to health services, emergency medicine must understand the current relationship between visit acuity and funding and the potential changes to that relationship given various specific funding policy initiatives. If patient volume and acuity is partially responsible for the increase in the number of EM residency training positions over time, what would the impact on medical education be if that volume and acuity shifts? Is there any reason to expect any shifts in the context of universal health coverage?
MATERIALS/METHODS: This is a retrospective case control study using a computer software database program to conduct a chart review of a large urban hospital ED associated with a university. First, the distribution of visit acuity is described for funded and unfunded patients over a 1 year period. Next, the null hypothesis that there is no difference in those distributions is tested. Finally, multiple regression analysis is used to test the hypothesis that funding status explains the variation in visit acuity and examines how much of the variation in the acuity distributions can be explained by funding status or other variables such as age and race. Finally, these results are discussed in the context of potential new healthcare spending initiatives to address questions of possible changes in patient volume and acuity within an academic ED.
RESULTS:
DISCUSSION: EM has a legal and ethical mandate to treat underfunded and uninsured patients. Therefore, we have unique insight into how health policies affect patients. Thus, we have an obligation to enter the national health care debate.
INTRODUCTION
The goal of this paper is to address the following question: How would universal health coverage change the volume of patient visits, the acuity of encounters, and the structure of residency training in a typical academic ED? To put this question another way, if the current structure of health care funding in the United States is partially responsible for the growth of academic emergency medicine and residency training, how will changes to that funding system impact medical education within the field?
Emergency departments are designed to provide life saving care at all hours of the day and night. However, in reality, many patient encounters are for non-emergent health problems. Numerous hypotheses have been posited to account for this phenomenon and some have suggested that these non-emergent visits likely result from a lack of ambulatory care and underfunding of health coverage in some patient populations (Haywood et al., 1991; Rask et al., 1998; O’Brien et al., 1997). This paper examines the link between the increase in non-emergent ED patient encounters and the number of EM residency training programs. Furthermore, changes to sociopolitical policies and health care affect the demographics of patient visits to the ED. Here, the potential changes that might be expected within some variant of universal health coverage are quantitatively considered.
Past research has focused on differences in the epidemiology and demographic characteristics of frequent versus infrequent users of the ED and a number of researchers have tested hypotheses relating to funding as a causative agent in such differences. In a 2001 study conducted in Sweden by Hansagi et al., the authors found that frequent users of ED services were also frequent users of ambulatory services. The conclusion was that universal coverage does not reduce ED visits. Other studies in Europe have shown similar results regarding the epidemiology of frequent ED users in universal coverage systems (Byrne et al., 2003). However, this conclusion cannot be extrapolated to the United States because many of our patients have no other access to services and we do not know what the ED would like in Sweden sans universal care.
A similar type of study was carried out by Byrne and colleagues in 2003 within a mixed-funding environment (an ED in Massachusetts) and similar results were found regarding the correlation between high frequency ED visits and high frequency outpatient visits. However, these studies have not addressed the types of ED visits that took place or the possible effects of funding changes on those ED visits. In other words, past studies have examined some of the variables within a health services model, but have not examined how those variables co-vary when another variable in that model is changed. Thus, it is critical to explore issues of patient funding, potential changes in coverage, and their estimated effects, on our current system of emergency medicine. In order to enter the national debate regarding access to health services, emergency medicine must understand the current relationship between visit acuity and funding and the potential changes to that relationship given various specific funding policy initiatives. The assumption of this paper is that majority opinion regards the current funding model to be in crisis and that a solution is necessary. Some type of change will arrive and that change might impact the structure of ED visits.
BACKGROUND
Emergency medicine is a growing field. This year, three new allopathic residency training programs have begun accepting new applicants (Florida Emergency Medicine Physicians, Oklahoma State, and UTMB-Galveston) and a number of other programs have increased the allotment of new residents for the subsequent year.
The field of EM was born approximately 40 years ago with the founding of ACEP in 1968 and the subsequent movement toward board specialty status. Herbert Flessa started the first EM residency training program at the University of Cincinnati with Bruce Janiak as the first resident in 1970. By 1975, there were 31 residency programs (http://www.emra.org). In 2001, there were 124 allopathic sites (Lathrop et al., 2001) and AMA-FREIDA lists 141 allopathic EM residency training programs as of October 2007. Furthermore, there were 4,957 approved training spots according to the ACGME in the 2003-2004 season. However, EM, like all fields of medicine, is affected by sociopolitical policies and changes. In the ED, the demographics of our patient encounters may be most reflective of shifting lifestyle choices or social-epidemiological risk factors more so than anywhere else in medicine.
On the eve of the 2008 elections, a number of candidates have outlined some form of mandatory universal health coverage. Furthermore, the recent presidential veto to expand S-CHIP has renewed the debate regarding a broader universal health care system in the United States for both children and adults. While health care expenditures continue to rise as a percent of our GDP, health outcomes have not improved substantially. Many health care professionals now accept the fact that change is necessary to insure solvency and better care for our patients.
In order to remain ahead of potential shifts in health policy, it is important to consider how such changes to coverage and patient funding might affect the volume and types of patient encounters in the ED. Our field is still relatively young and has not yet witnessed multiple shifts in policy and political cycles. It is critical to examine potential scenarios and affects of those scenarios on the field of residency training and medical education in emergency medicine.
The 2006 IOM report on EM notes that, over the past decade, the number of emergency rooms has decreased while the number of ED visits has increased. During this same period, the overall number of training positions has also increased. The volume of patient visits is not the only variable that accounts for the growth of our field. Certainly the spread of the requirement that hospitals staff EDs with ABEM certified physicians has also created a shortage of well-trained doctors that is likely to persist for some time, even if residency training continues to expand. However, it is important to consider all variables that lead to growth of a particular industry and, surely, patient volume must be one of those factors that provides a basis for increasing the total number of EM residency positions. The question is, to what extent does volume matter and, more importantly, is there even any reason to expect a volume decline in the context of improved patient funding and better access to primary care? The goals of this paper are to elucidate the relationships between funding, volume, and acuity of patients visiting a typical academic ED in order to provide more insight into how the field might change a future scenario with 100% access to ambulatory physicians and preventive services.
MATERIALS AND METHODS
In order to answer questions regarding changes to the distribution of visit types and volume, it is necessary to first examine descriptive data for a typical academic ED in an urban setting under the current model of coverage. What is the distribution of acuity in ED visits among insured patients? What is the distribution of acuity in ED visits among unfunded patients?
Next, I will test the null hypothesis that there is no difference in these distributions. If the null hypothesis if false, a prediction is that, if unfunded patients seek ED care for primary health concerns, the distribution should be right skewed (towards lower acuity visits with Level 4 and Level 5 triage designations)*. However, it is also possible that unfunded patients will have a left-skewed distribution toward lower acuity visits if (1) they rely on the ED for ambulatory care at roughly the same volume as the normal population visits primary care physicians and (2) if the occurrence of high acuity visits is infrequent, or at least equal to that of the funded group so as to not inflate the mean due to ordinal numbers of higher magnitude. If there is a right-shift distribution in the unfunded patients, how does this represent the overall distribution of ED visits in a given year?
If there is a statistically significant difference between funded and unfunded low-acuity visits, that difference may represent a pool of patients that would seek care with a primary care physician if funding were equal between the two groups. If this pool of patients is lost, what would the impact be on the ED in terms of lost patient encounters, teaching experiences, residency training and medical education? In other words, how much primary care/ambulatory care do we really see in the ED and how much of that would likely be lost if all patients could visit a primary care physician?
Finally, it is important to examine the state of EM and the demographics of patient encounters within academic EDs in other countries that have different types of funding models. For example, the demographics of an academic ED in Canada or other industrialized democracies may provide some insight into expected changes within our own departments.
This is a retrospective case control study using a computer software database program to conduct a chart review of a large urban hospital ED associated with a university. X number of patients will be identified that had no existing funding on presentation. Next, X number of control patients will be identified who have some source of funding (private insurance, Medicare, Medicaid, other state or local agency coverage – such as the Hillsborough County Health Care Plan). [Or should I look at all cases over a given time period – e.g. 6 months, 1 year, etc – depends on how easy it is to pull out info from the database??]. The cases and controls will be assigned a unique identifier. Given the retrospective and anonymous nature of this study, a request for informed consent was [will be] approved by the USF IRB and other aspects of the study were done with appropriate review. [will this likely receive exemption from the IRB people?]
The cases and controls will be compared to ensure that any observed differences in acuity level are explained by variation in funding status, as opposed to age, race, or sex. Thus, a chi-square analysis is conducted between the two groups for each of those three variables. Any significant differences in the groups will be addressed and controlled for in further statistical analyses (i.e. age and/or race may predict funding status and -- especially age -- may also predict level of acuity of a given ED visit).
For the cases and the controls, the level of acuity will be assigned a number from 1-5 based on the initial triage assessment [or should I base this on the discharge ICD-9 code and correlate that to acuity???]. Statistical analysis is used to test the hypothesis that there is no difference in the distribution of acuity level between groups. If there is a difference in acuity level, the hypothesis that the difference is explained by funding status will be further explored.
To test the hypothesis that there is no difference in the distributions of acuity level in the funded and unfunded group, a two-tailed T-test is employed. If statistically significant at p < style=""> Finally, an odds ratio can be calculated for the likelihood of a high acuity visit [Level 4 and 5?] between the unfunded and funded groups. A 95% CI is calculated for the odds ratio. An odds ratio is useful because the results of the ratio calculation can go in either direction and still provide relevant data. In other words, is it more likely that an unfunded patient relies on the ED for all primary care, thus, comes in for a wider range of ED visits with a likely lower mean-acuity score? Or, is it possible that the unfunded patient only uses the ED in a catastrophic event, simply living without medical care for most health concerns?
RESULTS
DISCUSSION
Leadership is seeing problems before they exist and knowing possible answers to questions that have not been asked. EM is a unique position to see the impact of our country’s healthcare policies first hand. Through federal legislation (EMTALA) and the words of our current president (“they can just go to the ER [if they don’t have health insurance]”) we are also the safety net for the underfunded and uninsured. Thus, we have an ethical mandate to play some role in the national health care spending debate.
Currently, we are nowhere near universal health coverage in this country. However, as our patients continue to suffer, vulnerable to the vetoes and philosophical debates of legislation, we must be willing to discuss realistic approaches to improving patient care. If our overall goal is to improve the health of our patients and keep them out of the ED, it is critical to understand the projected impacts of a wide range of potential solutions and improvements to our current healthcare crisis. While the idea of keeping customers away may seem to be a paradoxically positive outcome in the business of EM, this is balanced by the prospect of 100% reimbursement rates in the context of future health care spending policies.
References
1. http://www.emra.org
2. http://www.acgme.org/acWebsite/CMS/resPopData_specialty03-04.pdf
3. http://www.cjaonline.net/Communities/FL_Hillsborough.htm
4. Hayward RA, Bernard AM, Freeman HE, et al. Regular source of ambulatory care and access to health services. Am J Public Health. 1991;81:434-438.
5. O’Brien GM, Stein MD, Zierler S, et al. Use of the ED as a regular source of care: associate factors beyond lack of health insurance. Ann Emerg Med. 1997;30:286-291.
6. Rask KJ, Williams MV, McNagny SE, et al. Ambulatory health care use by patients in a public hospital emergency department. J Gen Intern Med. 1998;13:614-620.
7. IOM. Hospital based emergency care – at the breaking point. 2006.
Notes
* However, this is complicated by the possibility that unfunded patients waiting to seek care may present with more high acuity complaints due to delay in care.
**In Hillsborough County, there is an award-winning and nationally recognized comprehensive health care plan for indigent members of the community. Up to 28,000 residents are eligible for this plan at up to 100% poverty level. The plan emphasizes preventative services, early intervention, health education, and the coordination of health and social service. Thus, enrollment previous to recorded ED visit will be considered “funded” for this paper. However, enrollment during ED visit will be considered “unfunded” even if retroactive funding covers the particular visit since that patient would not have benefited from the primary care inherent to the plan previous to the ED visit.
***If there are more than funded vs. unfunded (e.g. underfunded) – conduct a trend analysis to examine the relationship between acuity and level of funding