The findings for Hypothesis 1 are in concurrence with previous literature regarding demographic, chronic illness and mental illness driving ED usage [9
]; however, our analyses offer nuance in distinguishing variables such as chronic illness and mental illness as differential risk factors for type of ED usage. Also, contrary to Anderson’s [33
] supposition that enabling resources and need were the strongest healthcare utilization predictors, this study’s overall model revealed need and predisposing factors to be most influential on non-emergent ED use. In particular, the predisposing characteristics of being of older age, female, or of Spanish/Hispanic origin were associated with an increased likelihood of using the ED for non-emergent reasons [36
]. The need variables concerning chronic disease admission also showed an increased likelihood of non-emergent ED use [10
]. This may be due to individuals with chronic conditions generally being sicker and potentially using more medical resources [28
]. With EMTALA prohibiting disposition of any patient presenting at the ED until they have been medically assessed, regardless of their ability to pay, the ED is always available for patient use [6
A sub-category of chronic disease or illness, mental illness, has previously been shown to be associated with high rates of overall ED use [19
]. However, this study revealed having a mental health condition, including suicide attempts or self-harm, as a reason for emergent ED use. This association may be a reflection of access to mental healthcare, which has been shown to be hindered by distance to a provider, geography, and provider shortages, especially in rural areas [46
]. Mental health patients potentially not having access to a primary-care provider or mental health specialist could result in not having the necessary resources available for regular treatment and, therefore, not seeking care until it is of emergent level in order to avoid costs they cannot afford. Additionally, patients presenting with first onset and/or genuine crisis-type mental health emergencies resulting from manic episodes, psychotic breaks, and suicides or self-harm are often transported via ambulance or law enforcement to an ED for immediate treatment [55
], contributing to the increase in the number of emergent mental health visits occurring in the ED.
Need was most influential in driving non-emergent ED use for Hypothesis 2, followed by enabling resources and predisposing characteristics. Race, which was previously influential in non-emergent ED use in the overall model as in other studies [7
], was no longer a significant predictor among patients with a rural living location. All ethnic minority groups were more likely to use the ED for emergent purposes in rural areas when compared to the original model. This could be the result of cultural or social norms, such as the use of folk remedies, as a preliminary method of healthcare [56
] despite the lack of empirically supported evidence [57
]. Complications and increased morbidity that may result from such cultural norms can require emergent treatment to counteract the effects of the folk remedy administered [56
Need was shown to be more influential in driving non-emergent ED use than the predisposing factors and enabling resources when zero non-emergent care sources were present. Following prior literature [22
], individuals with a chronic disease/illness admission or the presence of a chronic disease/illness were more likely to use the ED for non-emergent reasons. A higher likelihood of individuals of Spanish/Hispanic origin using the ED for non-emergent reasons was also observed, possibly due to the ED being the only accessible and convenient source of healthcare. It has been previously reported that Hispanic persons are more likely to report a difficulty in finding transportation to medical care [58
], conceivably resulting from low SES [59
], a historic disparity for Hispanic persons when compared to non-Hispanics [60
]. Additionally, individuals with a low SES who typically walk or use public transportation to get to medical care are less likely to have a regular source of care [61
]. The non-emergent ED use observed for the Spanish/Hispanic population in this study may also be influenced by the type of health insurance obtained and knowledge regarding where to go for non-emergent care. Upon the Medicaid expansion and implementation of the Patient Protection and Affordable Care Act [61
], some states experienced higher non-emergent ED use potentially due to newly covered individuals who did not necessarily have a regular source of care [6
4.1. Implications for Service Delivery
This study revealed mental health-related ED visits in NYS, including those related to suicide and self-harm, increase the likelihood of emergent ED visits (demonstrating the largest effect on ED usage overall). Since emergency care providers have a high likelihood of dealing with mental health emergencies, it may be beneficial to provide additional mental health training and education for emergency department staff. Such training may help ensure that mental health patients receive the best treatment possible upon their arrival at the ED. The minimal education ED staff receive is likely that provided during medical education or postgraduate training [62
]. Furthermore, ED staff often state that they do not feel adequately educated in assessing and diagnosing mental health diseases; therefore, increasing mental health training concerning clinical assessment and the immediate management of mental health patients in the ED during medical education or postgraduate education may help to decrease any feelings of inadequacy [62
]. This could be done through more coursework or rotations through both the ED and mental health, in order for medical staff to gain more familiarity with what they will likely encounter during practice.
Training for non-mental health healthcare providers has previously been available to help with learning to assess potential at-risk mentally ill populations for conditions such as suicide (i.e., ‘gatekeeper training’) [64
]. Such training has been seen to assist in improving the knowledge, attitudes and skills of individuals likely to encounter at-risk populations [65
] by focusing on learning how to recognize the signs and symptoms of psychological distress, improving communication with at-risk patients, understanding how to manage risk if suicide is a concern, understanding where to refer or bring at-risk patients, and knowing how to refer at-risk patients to specified resources [64
]. Scholars have even called for training and empirical testing on the applicability of gatekeeper training for overall mental illness so that ED staff can recognize these encounters [66
]. As ED providers become more educated on the best way to treat the patient population, ideas for best practices will become the norm. The ideas and methods discovered by ED staff will likely be of value for various levels of policymakers as they evaluate current emergency and mental healthcare policies, or as new policies undergo development that would better serve the patient population. Additionally, as many hospitals contain consultation-liaison services, integrated training with this service for all ED trainees may also be of value for providing healthcare to the patients with mental health concerns seen in the ED.
Historically, medical consultations in the ED from mental health providers are not appropriately used. Training on when to request a mental health consultation during initial medical education or postgraduate education may also assist in standardizing the use of mental health resources in an ED. Having a mental health provider on staff in the ED may also help with better regulating the use of mental health resources [63
]. This would be consistent with recommendations in the literature suggesting that mental health providers move towards new settings and non-traditional career paths [67
One limitation of this study was that all variables except day of the week of visit varied significantly by type of ED use; however, large sample sizes likely account for some of the significant findings. Multiple observations from the same patient were also not accounted for, as this was a visit-level analysis. The capacity to manage the number of patients presenting to non-emergent sources at any given time was not accounted for in this study, as this analysis focused primarily on the total number of non-emergent sources potentially available in a zip code. Additionally, this study was retrospective and examined one year of data, preventing any inference of causal relationships. The study sample also only included those from NYS, which varies demographically from other locations in the U.S., and did not address all racial/ethnic groups, possibly limiting generalizability. Future studies should analyze non-emergent ED use over time in order to explore the possibility of causal relationships, and ED use on public holidays when an individual’s mental health may be more greatly impacted and other sources of care may be less available. Further work can expand study samples to include ED visits from other states, break out the summary variable of chronic disease and mental health to explore the individual impact of each disease, and include all racial/ethnic groups to help increase generalizability.
Further educating ED staff through gatekeeper training may aid in driving changes in how healthcare, especially mental healthcare, is delivered. Currently, the ED serves as a healthcare resource that is available 24 h. The ED will most likely continue to be a non-emergent care ‘safety net’, especially for those without a primary-care provider, unless programs and clinics are developed for non-emergent care. However, these changes require new payment incentives and disease-management strategies to assist in making healthcare more accessible. Improving access to healthcare, including care for mental health conditions such as attempted suicide and self-harm, may be achieved by applying the findings of this study, educating ED staff to better manage the mental healthcare needs of the patients they see, and changing systems for mental healthcare delivery in order to expand care options for patients.