Emergency Department use for Dental Conditions: Trends over 10 years Introduction More than a decade ago, the Surgeon General s report on Oral health highlighted the importance of oral health, and the disparities that existed in oral health status in the US (1). About 40% of US adults do not have dental insurance based on 2001 BRFSS whereas only 14% lack health insurance (2). Dental diseases are largely considered preventable but the costs of delaying dental care could be enormous. It may lead to significant pain, poor quality of life, systemic health problems, personal financial burden as well as societal financial burden by increasing healthcare costs which needs to be subsidized for low income vulnerable populations. Access to dental care remains a public health problem due to large proportion of population lacking dental insurance and the limited participation of dental providers in public insurance programs. These contribute to lack of regular dental care for the uninsured and public program insured individuals which puts them at risk for poor oral health status. Emergency departments (ED) in hospitals are intended to serve medical and dental emergencies, however increasing healthcare costs have often been attributed to inappropriate use of EDs (3,4). Patients may seek care for dental conditions in the ED for one of the following situations: dental emergencies causing pain/swelling and needing immediate attention or for non-urgent dental condition when they have either no dental care provider or the wait times are too long. Hence, EDs may serve as a safety net for those with limited access to dental care. 1
Use of EDs for non-emergent conditions not only increases the costs but also distracts ED personnel from the actual emergencies. Also, EDs provide palliative treatment which is often not enough to resolve the dental condition. So, regular comprehensive dental care is still required for patients who present to EDs for dental conditions. Hence ED use for dental conditions, especially non- emergent dental conditions, underlines an unmet need of access to regular dental care in ambulatory settings. The magnitude of this unmet need can be estimated by assessing the prevalence of ED visits for dental conditions and it could also be used to identify the vulnerable populations in this regard. The existing literature on use of EDs for dental conditions is limited, especially the studies that look at such use nationally and over multiple years. One such study that examined the use of EDs nationally used the 1997 to 2000 National Hospital Ambulatory Medical Care Survey data and found that dental visits formed 0.7% of the total ED visits and found no significant time trends over four years (5). They found that younger adults, especially those without private insurance were at a greater risk for using the ED for a dental problem and recommend that EDs should be staffed and equipped to provide appropriate triage, diagnose, provide basic treatment and ensure follow up for dental care. Another study assessed 8 year trends using ED visit data from New Hampshire and found a slight increase in ED use for non-traumatic dental conditions from 2001 to 2008. They also reported that young adults and self paying individuals were most likely to have such visits (8). As the current multi-year studies come from individual institutions or cities, it was decided to conduct a ten year analysis of nationally representative ED visits for dental conditions (8,9,10). Also, none of the existing literature describes the pattern of care that takes place during the ED 2
visit, including but not limited to medications prescribed, procedures performed and follow up planned. These characteristics of the visit are important to understand the level of care provided and that still needed by patients who present to the ED. The aim of the study is to understand the trends in ED use for dental conditions over 10 years nationally and identify the characteristics of patients who seek care at EDs for potentially non-emergent dental conditions. Based on review of literature, it was hypothesized that: H 1 : There is an increase in the use of EDs for dental conditions from 2000 to 2009. H 2 : A majority of dental visits to the ED result in prescription for a medication and referral for follow up each year. H 3 : A majority of dental visits to the ED do not result in a medical procedure being performed each year. H 4 : More uninsured/ publically insured people had dental visits to the ED than privately insured population each year. Methods National Hospital Ambulatory Medical Care Survey (NHAMCS) is a national probability sample survey of hospital outpatient and emergency department visits that is conducted by the National Center for Health Statistics, Centers for Disease Control and Prevention. The survey is conducted at about 500 nationally representative hospitals over 4 weeks of time at each hospital (6). Ten years of the emergency department component of NHAMCS which is publically available on the CDC website was utilized from 2000 to 2009. 3
All ED visits were included for ten years under study and then a new binary variable was created to differentiate the visits with a primary diagnosis of dental disease from other diagnosis. It was decided to use primary diagnosis as the differentiating criteria as we were interested in the extent of ED use for primarily dental conditions, and not dental complaints in which case, reason for visit could be used as a criterion. International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) codes were used to classify the visits. As it was decided apriori to include only dental diseases, diagnosis related to salivary glands, temporomandibular joint and other related diagnosis were excluded. ICD-9-CM codes of 521.00 to 525.9 were included to indicate a visit due to a dental condition. To assess the extent of ED visits that were made due to a dental condition, numbers of ED visits due to a dental condition were calculated as a proportion of that year's total ED visits. A linear trend analysis was then performed to assess if the percent of ED visits due to dental condition had been linearly increasing over the 10 years studied. As the data under study spanned over ten years, substantial changes had been made to the way variables were collected, coded and named. These yearly changes in variables and sampling were studied and appropriate recoding done to allow comparisons across all 10 years. Trends over time were calculated using Spearman correlation to account for the data that were not distributed normally. Bivariate analysis was conducted between explanatory variables across ten years and having an ED visit for a dental condition. Significant differences between years were assessed using Chisquared or t-tests or their non-parametric analogs, as appropriate. 4
Finally, multivariable analysis was conducted using logistic regression to determine the factors that are associated with increased likelihood of having an ED visit for dental condition. Variables that were included in both bivariate and multivariate analysis were restricted to include only those variable that were established apriori to have a potential association with the outcome variable or those that were thought to be of interest otherwise. Weights were applied in all analysis to get nationally representative estimates each year. All statistical analysis was performed using SAS version 9.2 and significance was determined at 0.05 level. Results Total number of ED visits represented by 10 years of survey data was 1159.5 million which included about 14 million ED visits due to dental conditions. Yearly, the number of ED visits with a primary diagnosis of dental condition ranged from 1.07 million in year 2000 to 1.94 million in year 2009. This represented 0.99% of total ED visit in 2000 and 1.43% of total ED visits in 2009 (Table 1). The percent of ED visits that were due to dental conditions demonstrated a significantly increasing linear trend over the ten years examined (Spearman Correlation=0.915, p-value=0.0002) (Figure 1). More than 50% of the primary diagnosis each year except one was 'other diseases/conditions of teeth' (525.), followed by 'diseases of the pulp and periapical tissues' (522.), 'diseases of the hard tissues of teeth' (521.), 'diseases of gingival and periodontal tissues' (523.) and finally 'dentofacial anomalies including malocclusion' (524.). 5
About 69% of all ED visits for dental conditions presented with symptoms related to teeth and gums, 7% presented with nonspecific site pain, 2.2% presented with earaches and about 2% presented with head, neck or face injury. The mean age of patients presenting to the ED with dental conditions was about 36 years with 19-44 age group forming the majority of those who went to ED for dental conditions. The age groups and gender did not vary significantly across years. However, race and expected payment source (proxy measure for insurance status) were significantly different across ten years (Figure 2). Medicaid enrollees seem to visit ED for dental conditions more often in 2009, when they formed 31.5% of all payer types than in 2000, when they we only 19% of all payer types. Racial minorities are also seeking dental care at EDs more often in 2009 than they were in 2000 (30.6% in 2009 compared to 24.7% in 2000) (Figure 2a). A total of 8.4% of all ED dental visits were injury related, with some variation across the years that was significant (Spearman correlation=0.782, p-value=0.0075). Maximum injury related visits occurred in 2005 that constituted 14.5% of total ED dental visits and minimum in 2002 when it comprised of just 3.7% (Figure 3). The severity of pain varied significantly across years with 58% patients presenting with severe pain and only 2.5% presenting with no pain in 2009 whereas in 2000, about 26% presented with severe pain and 43% presented with no pain. On an average, about 3% of all ED dental visits were made by patients who had already been seen in the same ED in past 72 hours. Another variable that indicates past use of ED for health issues was number of ED visits in past 12 months. About 17% of patients with ED dental visits had been to the ED once in last 12 months, 8% had been to ED twice, 4% had been to the ED 6
thrice and 3.6% had been to the ED four times in the past 12 months. However, this variable was only available for 2007, 2008 and 2009; hence previous years could not be compared. A majority of patients with ED visits for dental conditions did not receive any medical procedure, the proportion who did receive a medical procedure ranged from 8% to 17% over the years. However, there was no significant discernable trend. Contrary to this, a majority of patients were prescribed medications each year, the proportion ranged from 86% to 95%. Main categories of first drug that was prescribed were either antibiotics, non-steroidal antiinflammatory drugs or opioid analgesics. Follow up was planned for vast majority of patients who had an ED dental visit, which ranged from 89% to almost 96%. However, there was no linear trend in the pattern of follow up planning over years (Spearman correlation=0.55, p value=0.09). Final model was selected using the stepwise selection method and significant predictors of having an ED visit due to dental condition were metropolitan statistical area status, expected source of payment and immediacy as determined by triage. Those who were most likely to have an ED visit for dental condition were from a non-metropolitan statistical area, self-pay as expected source of payment and those with immediacy of needed attention after more than 2 hours (Odds Ratios presented in Table 2). Discussion The current study is the only study to the best of our knowledge that has assessed the use of ED for dental visits over a period of 10 years using nationally representative data. A study that used the same data source and looked at four years of data from 1997 to 2000 did not find any time trends (5). They reported an average of 738,000 visits each year which is significantly lower than 7
1.2 million ED dental visits from our study. However, they used reason for visit to classify dental ED visits whereas we used primary diagnosis, and this may partly account for the difference observed. Another study that assessed secular trends in ED use for dental conditions assessed data from Kansas City, Missouri from 2001-2006. They reported a significant time trend and found an average of 1.7% of ED visits made due to dental conditions, which is slightly higher than our estimate (7). However, this study included temporomandibular disorders as dental conditions, which can explain their higher estimates. The trends in racial composition and expected source of payment are especially interesting. It demonstrates further widening of the disparities that existed and were referenced in Surgeon General's Report on Oral Health. Based on the results, it can be interpreted that more vulnerable populations are becoming increasingly dependent on EDs for seeking dental care. It should be noted that trauma is included under 'other diseases/conditions of teeth', as well as limited ability of ED personnel to accurately diagnose dental diseases may be a reason for a majority of the visits being coded as other diseases. No plausible reasons can explain the drastic change in pain levels among patients from 2000 to 2009, and it could reflect a change in thoroughness of pain assessment and coding. The 3% of patients who had already been to the ED in past 72 hours are of concern. Although it cannot be confirmed that the past visit was related to dental condition, it can be assumed that a majority of those past visits could be due to the same problem that made them seek care in the ED again. Similarly, about 40% of all patients with ED dental visits had had at least one visit to 8
the ED in past 12 months, which reflects the chronic use of and reliance on emergency departments for healthcare. As indicated by the frequency of medical procedures and medication prescriptions for patients who have ED visit for dental conditions, it is clear that the main line of management for such patients is through medications in the ED. Rarely are hospital EDs staffed with a dental personnel and most often, no definitive procedures are performed during the ED visit. The aim is pain and discomfort management, usually accomplished through medications, and then follow ups are planned for further resolution of the dental problem. However, it would be of interest for future research to follow patients who had an ED dental visit to assess if they later had a visit to dental office for the complete resolution of the problem that brought them to the ED. There are several limitations to this study. Although many variables were recoded, recreated and renamed to allow for 10 year comparisons, certain data was either not collected or the format in which it was collected certain years did not permit it to be directly compared with other years. This limited us in comparing only a few years rather than all ten on few variables. Certain variables, such as metropolitan statistical area that was a significant predictor of ED visits for dental conditions, refers to MSA status of the hospital and not the patients. Hence, it should be recalled that patients may be from a different area than the hospital to which they go, and get coded same as the hospital. As dental diseases are largely preventable, and could be either prevented or treated in early stages through regular dental care, use of emergency departments for dental conditions can be argued to be potentially inappropriate. As the time trends are increasing in ED visits for dental, it would be of interest to perform a cost- of illness type of analysis to ascertain the added economic 9
burden. However, as no cost or charge data is available in these datasets, such analysis could not be performed. Conclusion Significant increasing trend in proportion of ED dental visits, especially by racial minorities and uninsured and Medicaid enrollees highlights the increasing disparities in access to oral healthcare, even after a decade when they were nationally recognized by the US surgeon General's report in 2000. Only palliative treatment, most often only medications are provided at EDs for dental problems and patient would need to still seek definitive care. Future research should investigate whether these patients ultimately receive complete resolution from the dental problem that caused them to visit the ED. 10