Epilepsy & Behavior 31 (2014) 307–311
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Healthcare utilization of patients with epilepsy in Yuma County, Arizona: Do disparities exist? Yeeck Sim a, Brandon Nokes b, Seenu Byreddy a, Jenny Chong a, Bruce M. Coull a, David M. Labiner a,c,⁎ a b c
Department of Neurology, University of Arizona, Tucson, AZ, USA College of Medicine, University of Arizona, Tucson, AZ, USA Department of Pharmacy Practice and Science, College of Pharmacy, University of Arizona, Tucson, AZ, USA
a r t i c l e
i n f o
Article history: Received 13 August 2013 Revised 8 October 2013 Accepted 14 October 2013 Available online 5 November 2013 Keywords: Hispanic Epilepsy Cost Border community Patient behavior Seizure Epidemiology
a b s t r a c t The aim of this study was to describe the disparities in healthcare utilization and costs between Hispanic and non-Hispanic patients with seizures or epilepsy. We reviewed the insurance status and healthcare resource utilization data from 2005 to 2008 for all patients with seizures and epilepsy seen at the Yuma Regional Medical Center (YRMC). Charges for medical services provided to Hispanic patients with epilepsy between the ages of 18 and 49 were significantly less than those for non-Hispanic patients with epilepsy (Hispanic: $3167.63 versus non-Hispanic: $5154.36, P b 0.001). Taking into account the differences in insurance status, setting of care, and total number of procedures, we still saw a significant difference in charges between the two groups at the outpatient settings. These data differ from currently available data on national and Eastern US Hispanic patients with epilepsy, suggesting that patients in this border community are somehow different from Hispanics elsewhere in the US. © 2013 Elsevier Inc. All rights reserved.
1. Introduction It is estimated that 1 in 26 individuals in the United States will develop epilepsy or a related seizure disorder by age 75, with a global prevalence between 4 and 6/1000 people [1–3]. Within the US, recent data from the Behavioral Risk Factor Surveillance Survey (BRFSS) report the lifetime prevalence for epilepsy to be between 11.5/1000 and 22/1000 [4]. However, studies show that the prevalence of epilepsy among various ethnic groups, such as Hispanics and non-Hispanic whites, varies by location. Epilepsy has been reported to be more prevalent among Hispanics in New York, whereas in California and Arizona, the prevalence of epilepsy in Hispanics has been reported to be lower than in non-Hispanics [1,5–9]. Regardless of ethnicity, however, epilepsy prevalence among individuals of lower socioeconomic status is consistently high [1,10]. Within the US, individuals in border communities such as Yuma County, Arizona, experience limitations in access to care because of lower socioeconomic status and decreased healthcare facility and provider availability [11]. According to the 2012 US census data, the poverty rate in Yuma is approximately 20% [12]. In a 2010 assessment of border community healthcare access, up to 25% of citizens within border communities, including Yuma County, were uninsured [11], although ⁎ Corresponding author at: Department of Neurology, University of Arizona, 1501 N. Campbell Avenue, PO Box 245023, Tucson, AZ 85724, USA. Fax: +1 520 626 2111. E-mail address: [email protected] (D.M. Labiner). 1525-5050/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yebeh.2013.10.018
patients requiring specialized care such as epilepsy management are less likely to be uninsured [5,13]. The Epilepsy Foundation reported in 2000 that 18% of patients with epilepsy were uninsured [13], although more recent data from California place this number closer to 10% [5]. Underinsured patients tend to have greater medical costs related to poorer disease management, and this is especially true for patients with epilepsy [14]. These greater out-of-pocket costs are related to Emergency Department encounters, epilepsy-related comorbidities such as injuries, as well as indirect costs, such as lost wages and unemployment. A recent study reported that uninsured individuals may have decreased quality of care, indicated by significantly fewer outpatient visits, fewer appointments with neurologists, and greater antiepileptic drug (AED) costs compared with patients with private insurance [15]. The direct costs of epilepsy include costs required to treat the underlying disease, such as physician visits, AEDs, and, for some, surgeries. Indirect costs include lost wages, unemployment, and other medical comorbidities [16]. In general, the costs associated with epilepsy and seizure care are largely associated with the indirect costs of seizure development and are attributable to nonepilepsy-related care [15,17]. Epilepsy care in Hispanics has been relatively understudied, especially with regard to access to care and costs of services within border communities. Historically, it has been reported that within the US, access to care for Hispanics was limited because of lack of insurance and increased out-of-pocket costs [11,15]. However, a number of the studies related to Hispanic epilepsy costs either sampled national trends or were limited to the Eastern US [18–20]. The population characteristics in
308
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
Yuma are unique and unlike those in most parts of the US. In Yuma County, over half (60.1%) of the population is Hispanic, whereas 34.6% of the population is non-Hispanic whites [12]. To our knowledge, no previous studies have looked explicitly at the costs of epilepsy care within border communities such as Yuma County, Arizona. To better elucidate the cost of care for patients with epilepsy and seizures within this border community, administrative data from the Yuma Regional Medical Center (YRMC) were used to investigate the healthcare resource utilization and insurance status of patients with seizure or epilepsy conditions between 2005 and 2008. We also sought to understand whether the underlying cost differences, if any, were attributable to the length of hospital stay or the number of procedures performed either in or out of the hospital. We hypothesized that behavior of patients with epilepsy within this border community and healthcare costs may be different from those previously reported for other populations within the US.
Table 1 Patient demographics.
2. Methods
3. Results
Data between January 2005 and July 2008 were obtained from YRMC. The data contained patient demographics including age, sex, and ethnicity; location of residence by city, state, and zip code; discharge status; setting of care; length of stay during a visit; insurance status; and the total charges in dollars for each encounter. For each encounter, up to five diagnoses were available using the International Classification of Disease, Ninth Revision (ICD-9) codes, and up to three procedure columns were reported using the ICD-9 procedure codes. The ICD-9 codes of interest in this study were 345.0–345.9 (epilepsy); 094.89 (epilepsy and syphilis-related); 123.1 (epilepsy and cysticercosisrelated); 123.9 (epilepsy and parasite-related); 333.2 (progressive familial epilepsy); 347.0 (epilepsy and sleep-related); 907.0 (epilepsy due to traumatic injury); 780.3 and 780.39 (convulsions); and 780.31– 780.32 (febrile convulsions). The ICD-9 procedure codes were examined both overall and within each setting of care. Only patients who reported that they were living in Yuma County were included. Hispanics were self-identified at the time of encounter. Hispanic and non-Hispanic Yuma County cases with seizure and/or epilepsy coded during the visit (hereafter known as the group with epilepsy) were assessed together, while Hispanic and non-Hispanic patients who were not coded with seizure and/or epilepsy codes on any visit served as respective control groups. The data were stratified by age: children (0–17 years), adults (18–64 years), and seniors (65 years and older). In more detailed analyses, the adult group was further divided into three groups: 18–33 years, 34–49 years, and 50– 64 years. Insurance status was categorized into three groups: 1) no insurance/ self-pay; 2) public insurance including Arizona Health Care Cost Containment System (AHCCCS —Arizona's Medicaid program), Tricare/ Military/Veteran's Affairs, and other government health programs, such as Medicare; and 3) private insurance (e.g., commercial plans such as Blue Cross and Worker's Compensation). The emergency department and after-hours clinic visits were analyzed together (ED), while hospital outpatient visits were also analyzed together with outpatient clinic visits (OP). A small number of patients visited numerous “other” sources of care such as the Foothills Clinic (a neighborhood facility where mainly non-Hispanic patients were seen), outpatient surgery, and perinatal services. Information about the specialties of the physicians seeing these patients was not available. The data were analyzed with IBM SPSS version 20.0. Comparisons between the Hispanic and non-Hispanic groups with epilepsy were made in terms of total charges, procedure codes, length of stay, insurance status, and setting of care using independent sample t-tests for continuous variables, and the Pearson chi-square test for comparing frequencies. Differences were considered significant at the P ≤ 0.05 level.
3.1. Patient demographics
Total patients Patients with epilepsy Mean age without epilepsy (years) Mean age with epilepsy (years) Male Female With insurance
Hispanic
Non-Hispanic
Total
66,663 960 (1.44%) 28 (+/−22) 28 (+/−24) 520 (44.9%) 440 (40.6%) 869 (90.5%)
69,157 1282 (1.85%) 47 (+/−26) 48 (+/−26) 639 (55.1%) 643 (59.4%) 1121 (95.2%)
135,820 2242 135,820 2242 1159 (51.7%) 1083 (48.3%) 2090 (93.2%)
We also conducted a multiple regression analysis with total charges as the dependent variable and ethnicity (Hispanic/non-Hispanic); length of stay; number of procedures performed; number of comorbid diagnoses; and insurance status — public (yes/no), private (yes/no), and self-pay or no insurance (yes/no) as predictors.
Across all 4 years from January 2005 to June 2008, a total of 960 Hispanic patients with epilepsy and 1282 non-Hispanic patients with epilepsy utilized YRMC facilities. Hispanics represented about half (51.9%) of the number of visits, while the non-Hispanic group represented the other half (48.1%). The average age of Hispanic patients was significantly younger than that of non-Hispanic patients; Hispanic patients with epilepsy were also significantly younger than nonHispanic patients with epilepsy (Table 1). 3.2. Total charges The mean total charge for each visit was significantly less for Hispanics than for non-Hispanics with epilepsy. However, on further analysis, this difference was significant only in adults aged 18–64 but not in children or seniors. Furthermore, when the adults were examined in smaller age groups to account for possibly skewed ages, the difference was significant in the 18- to 49-year group but not significant among those between 50 and 64 years. Those not in the group with epilepsy did not show significant charge differences when stratified by age, except for the 65 and over seniors where the Hispanic patients had higher charges compared with non-Hispanic patients (Table 2). 3.3. Insurance status Among those with epilepsy, proportionately more young Hispanic adults (age: 18–33) had private insurance compared with their nonHispanic cohort. Fewer young Hispanic adults in the group with Table 2 Mean total charges in $ US.
Seizure/epilepsy Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+ Age 18–64 Age 18–49 Without seizure/epilepsy Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
Hispanic
Non-Hispanic
Significance
1605.40 2285.52 4979.16 8624.43 10825.81 3279.91 3167.63
1662.80 3660.98 6406.15 10573.02 10386.80 6846.40 5154.36
NS P b 0.001 P b 0.027 NS NS P b 0.001 P b 0.001
1810.00 2869.16 3006.87 4124.87 6089.02
1837.18 2824.67 3033.08 4005.40 5677.51
NS NS NS NS P b 0.001
All ages in years. Mean in $. NS: not significant.
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
epilepsy had public insurance as opposed to the majority of nonHispanics. In the 50- to 64-year-old group with epilepsy, Hispanics were significantly more likely to be enrolled in public insurance compared with non-Hispanics, while a significantly larger proportion of nonHispanics were enrolled in private insurance compared with Hispanics. There were no significant differences regarding insurance status among other age groups or those who had no insurance among Hispanics and non-Hispanics. Hispanics had lower mean total charges (not just those related to epilepsy) compared with non-Hispanics for all three insurance statuses — public, private, or self-pay. Overall, patients with public insurance averaged higher charges compared with those with private insurance.
309
Hispanic adults were admitted to inpatient services compared with the non-Hispanic cohorts. Significant differences were found between Hispanic and nonHispanic older adults aged 50–64 years in the group with epilepsy at the outpatient and inpatient settings. Similar to those aged 18– 49 years, older Hispanic adults were more likely to use hospital outpatient clinics compared with non-Hispanic older adults. Older Hispanic adults were also less likely to be admitted to inpatient services compared with non-Hispanic older adults. Senior adults aged 65 years and over only had a significant difference in the ED setting where more Hispanic seniors sought care compared with non-Hispanic seniors. For the outpatient setting, Hispanics aged 18–49 had significantly lower charges compared with non-Hispanics. Likewise for the “other” sources of care, Hispanics had significantly lower charges compared with non-Hispanics (Table 4).
3.4. Setting of care 3.5. Length of stay For the young adults aged 18–33 years in the group with epilepsy, fewer Hispanics sought care at the ED, compared with non-Hispanics. On the other hand, in the OP setting, significantly more young Hispanic adults received care compared with non-Hispanics. Only a small group of young Hispanic adults were admitted to inpatient services compared with a larger proportion of non-Hispanics. The young Hispanic adults also visited more “other” sources of care compared with non-Hispanics but were less likely to receive outpatient diagnostic imaging services compared with young non-Hispanic adults (Table 3). Adults with epilepsy aged 34–49 years were similar to their younger counterparts (18–33 years) in terms of where they obtained their care. They did not differ in their use of the ED, but significantly more Hispanic adults sought care at OP setting compared with non-Hispanics. However, this does not reflect Hispanics who received their care from private MD offices, which we were not able to measure at this time. Similarly, fewer Table 3 Setting of care. Hispanic
Non-Hispanic
Significance
ED Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
565 (29.9%) 300 (31.1%) 233 (49.1%) 115 (41.1%) 62 (25.6%)
229 (32.7%) 299 (51.6%) 375 (54.7%) 234 (37.3%) 172 (17.7%)
NS P b 0.001 NS NS P b 0.006
OP Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
947 (50.1%) 480 (49.7%) 132 (27.8%) 59 (21.1%) 59 (25.4%)
345 (49.2%) 162 (28.0%) 99 (14.4%) 97 (15.5%) 270 (27.8%)
NS P b 0.001 P b 0.001 P b 0.04 NS
Inpatient Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
164 (8.7%) 60 (6.2%) 72 (15.2%) 78 (27.9%) 102 (42.1%)
57 (8.1%) 60 (10.4%) 150 (21.9%) 238 (38.0%) 446 (45.9%)
NS P b 0.004 P b 0.005 P b 0.003 NS
Outpatient diagnostic imaging Age 0–17 31 (1.6%) Age 18–33 15 (1.6%) Age 34–49 20 (4.2%) Age 50–64 10 (3.6%) Age 65+ 8 (3.3%)
16 (2.3%) 25 (4.3%) 23 (3.4%) 13 (2.1) 26 (2.7%)
NS P b 0.001 NS NS NS
Other Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
54 (7.7%) 33 (5.7%) 39 (5.7%) 45 (7.2%) 58 (6.0%)
NS P b 0.001 NS NS NS
182 (9.6%) 111 (11.5%) 18 (3.8%) 18 (6.4%) 11 (4.5%)
All ages in years, with seizure/epilepsy. NS: not significant. Percentages are within respective age groups.
The only significant difference in the average length of stay was between Hispanic and non-Hispanic children with epilepsy from age 0 to 17. There were no differences for adults and seniors with epilepsy. 3.6. Procedures The average number of procedures appears to be underreported or not reported at all. It is also possible that these services were provided by facilities outside of YRMC and, as such, not captured in our data. For the procedures that were reported, the highest frequencies of codes involved are as follows: injection of therapeutic or prophylactic substances (123), injection of antibiotics (97), venous catheterization (91), insertion of endotracheal tubes (86), spinal tap (60), and mechanical ventilation (52) out of a total of 7417 visits. There were only 13 cases of magnetic resonance imaging of the brain reported, and there was only a single case of a computerized tomography scan of the head reported out of 7417 visits. Hispanics in the group with epilepsy had significantly higher number of average procedures during visits for adults aged 18–49 years compared with non-Hispanics. However, Hispanic seniors aged 65 and over had significantly fewer procedures performed, though this is a bit difficult to explain, as the procedures appear to be underreported. Potentially, the Hispanic cohort with epilepsy was receiving cheaper procedures (i.e., more X-rays versus MRI), but the answer is not entirely clear at this point. 3.7. Comorbidities On average, non-Hispanics with epilepsy had more comorbidities compared with their Hispanic counterparts. In the ED setting, the younger non-Hispanic adults with epilepsy (18–33 years and 34– 49 years) had a significantly higher number of comorbidities compared with the Hispanic adults with epilepsy. In the OP setting, the nonHispanic seniors aged 65 years and over had a significantly greater number of comorbidities compared with the Hispanic seniors. In the inpatient care and other care settings, the younger non-Hispanic adults (18–33 years) had a higher average number of comorbidities compared with their cohorts.
Table 4 Mean total charges at setting in $ US.
ED OP Outpatient diagnostic imaging Inpatient Other
Hispanic
Non-Hispanic
Significance
2285.25 695.68 2001.89 17549.24 3956.29
2507.23 888.11 2231.71 18316.57 8294.25
NS P b 0.02 NS NS P b 0.001
Mean in $, age 18–49 years with seizure/epilepsy. NS: not significant.
310
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
Diabetes was the only medical disease that was highly associated with one of the ethnicities (Hispanic). Neither ethnic group showed a higher risk for neurologic comorbidities such as stroke, cognitive impairments, or traumatic brain injury.
3.8. Multiple regression analysis Correlational analyses revealed significantly high and negative correlations between public insurance and private insurance (r = −0.889) and between outpatient and both comorbidity (r = −0.564) and emergency department (r = −0.541), resulting in the public insurance variable and outpatient setting being dropped from the analysis. Results showed that the model was significant with an adjusted R square of 0.723. In other words, the model explained almost three quarters of the variance in the total charges. The main contributors to the total charge are length of stay and the number of procedures performed, with length of stay being the major driver of cost. Hispanics have lower charges. Not being insured (or self-pay) did not have a significant relationship with the total charges (Table 5).
4.3. Differences in comorbidities Although non-Hispanic adults aged 18–49 years had a significantly higher mean number of comorbidities in the ED and inpatient settings, their mean total charge was significantly higher than the Hispanics only in the outpatient setting. The higher number of comorbidities may also play a role in the increased mean total charges incurred by the non-Hispanic patients overall. Whether this implies that the outpatient individuals are somehow different from those in the other settings or whether this is an underreporting of medical comorbidity is unclear. Similarly, the difference in comorbidities may be due to the larger overall number of young Hispanic patients with epilepsy. Future studies should establish a quality-of-care metric to assess whether or not the differences in outpatient care are attributable to the differences in delivery of care or due to the differences in populations served, as we had insufficient data to clarify this. If the Hispanics within this community are receiving the same or better care than their nonHispanic counterparts, this would allow for improved care at reduced costs for all patients with epilepsy within Yuma, regardless of ethnicity.
4. Discussion
4.4. Differences in insurance status
4.1. Differences in preferred location of care
It is also worth noting that more Hispanics with epilepsy had private insurance, particularly in the 18–33 age group, and yet accrued lower charges than the similarly insured non-Hispanic patients with epilepsy of the same age within the outpatient setting. Whether or not this bears relation to their overall healthcare costs is uncertain. Interestingly, patients with public insurance were consistently charged higher fees compared with patients with private insurance, in both Hispanic and non-Hispanic patients. This is not due to contracted rates as our data analysis utilized actual charges prior to contractual adjustments. This is in contrast to previous analysis of the Medical Expenditure Panel Survey (MEPS), in which uninsured and underinsured have greater out-of-pocket expenses [15]. Also of interest, the preferred setting of care and insurance status of Hispanic patients with epilepsy and seizures in Yuma is starkly different from that of Hispanics within previous national and Eastern US studies, in which Hispanics were more likely to be uninsured and utilize the ED for medical treatment [15,18,19]. This point stresses the lack of generalizability of the Hispanic patient population and suggests that the behavior of Hispanic patients with epilepsy/seizures in this border community is different from the behavior of Hispanic patients with epilepsy elsewhere. It is certainly possible that some care (particularly that associated with out-ofpocket costs) could be obtained across the border for Hispanic patients, and this could manifest as lower cost in Yuma County, while the actual costs borne by the patient would be greater.
The principal finding in this study was that Hispanic patients with epilepsy or seizures had significantly lower charges compared with non-Hispanic patients in Yuma County, Arizona. This may be, in part, attributable to the Hispanic patients' preferred setting of care (OP as opposed to ED). Hispanic patients with epilepsy displayed a healthcare utilization behavior of more outpatient visits and less ED, outpatient diagnostic imaging, and inpatient visits. Although this could account for overall lower charges/costs for Hispanic patients, even within the outpatient clinic and “other” settings, Hispanics still had significantly lower average charges compared with non-Hispanics. The lower outpatient charges are difficult to explain and may be attributable to variability (or disparity) in outpatient patient care.
4.2. Differences in procedures and length of stay The regression analysis indicated that length of stay and mean number of procedures were the main contributors to the total cost in both Hispanics and non-Hispanics. The influence of these variables is not apparent by univariate or multivariate analysis but may play a significant role when other potential factors are accounted for within the analysis.
Table 5 Multiple regression analysis.
Possible confounders
Insurance Site of care
Constant Hispanic Age Length of stay Comorbidities Procedures Private ins. Self-pay/no ins. Inpatient Other sites ODI ED
Unstandardized coefficients
Standardized coefficients
Beta
Standard error
Beta
−301.032 −468.118 16.648 2957.250 −171.734 4763.268 669.161 488.681 3296.985 2512.274 1832.586 2049.973
203.246 145.745 3.334 36.366 62.860 119.213 162.271 315.200 285.735 279.399 437.408 192.753
−.021 .038 .645 −.026 .270 .026 .009 .116 .059 .025 .087
t-Stat
−1.481 −3.212 4.993 81.318 −2.732 39.956 4.124 1.550 11.539 8.992 4.190 10.635
Sig.
NS .001 .000 .000 .006 .000 .000 NS .000 .000 .000 .000
CI: confidence interval. ED: emergency department. Ins: insurance. ODI: outpatient diagnostic imaging. Sig: significant. NS: not significant.
95.0% CI for beta Lower bound
Upper bound
−699.448 −753.817 10.111 2885.963 −294.955 4529.580 351.069 −129.194 2736.871 1964.578 975.152 1672.128
97.383 −182.420 23.184 3028.538 −48.513 4996.956 987.254 1106.555 3857.100 3059.969 2690.020 2427.818
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
4.5. Less epilepsy care taking place within the US — the healthy immigrant hypothesis De Jesus and Chenyang reported that according to the Pew Hispanic Survey, Hispanics are more likely to seek care in Mexico owing to lack of continuous insurance coverage and a perceived lack of quality of care within the US [21]. A potential reason for less Hispanic epilepsy management in the US is a lack of understanding within the US Hispanic community about epilepsy [22]. Hispanics may be less likely to pursue chronic epilepsy management. Alternatively, the patients may instead be receiving chronic care in Mexico after their initial seizure episode, but this seems less likely, considering their better insurance status compared with the non-Hispanics. Others have reported that Hispanics have less chronic illnesses due to the healthy immigrant hypothesis [23,24]. Among those Hispanics who do have chronic illnesses, some studies have reported that they seek management in Mexico [21]. Similarly, Hispanic patients with epilepsy may not need as much inpatient or outpatient care because of their overall younger age and potentially overall better health status, but our study still shows significant differences in the outpatient setting compared with the nonHispanic young adults with epilepsy. 4.6. Limitations There are several limitations worth considering for this study. Principally, there are a number of factors not readily discernible from health records, such as degree of education, socioeconomic status, time within the US, and even the means of diagnosis. Further, our study design is retrospective in nature and does not describe this patient population as well as a prospective study would. It is also worth considering that a number of epilepsy patients within this border community may not be presenting for care, though it is difficult to quantitatively define the percentage of the Yuma population with epilepsy who did not do so during the measured time frame. However, the relative prevalence of epilepsy detected within this study is roughly concordant with a previous study we conducted along the Arizona border [9]. Lastly, although YRMC is the only major hospital in Yuma, not every outpatient clinic, including those provided by Indian Health Services, is accounted for. As such, future studies should consider including as many institutions as possible in order to better describe the patient population with epilepsy within Yuma. 4.7. Conclusion This study highlights the differences in patient behavior in regard to ED and OP care preferences as well as overall charge differences between border communities and previously reported Hispanic patients from the eastern US. Despite overall cost differences, there does not appear to be limitations in access to care between Hispanic and non-Hispanic patients with epilepsy. We hope that by identifying the differences in epilepsy and seizure care within the border community, quality of care could be improved for this patient population through further elucidation of these cost differences. In light of our findings, we will continue to assess patient behavior within border communities, with the consideration that the behavior of Hispanics within these communities may be different and not applicable to that of other Hispanics within the US.
311
Acknowledgments We thank and acknowledge Yuma Regional Medical Center for allowing us to use and analyze their patient data and Dr. William Johnson of the Center for Health Information and Research at Arizona State University for providing us with the YRMC database. References [1] Bartolini E, Bell GS, Sander JW. Multicultural challenges in epilepsy. Epilepsy Behav 2011;20:428–34. [2] Cardarelli WJ, Smith BJ. The burden of epilepsy to patients and payers. Am J Manag Care 2010;16:S331–6. [3] England MJ, Liverman CT, Schultz AM, Strawbridge LM. Epilepsy across the spectrum: promoting health and understanding. A summary of the Institute of Medicine report. Epilepsy Behav 2012;25:266–76. [4] Kobau R, Zahran H, Grant D, Thurman DJ, Price PH, Zack MM. Prevalence of active epilepsy and health-related quality of life among adults with self-reported epilepsy in California: California Health Interview Survey, 2003. Epilepsia 2007;48:1904–13. [5] Kobau R, Zahran H, Grant D, Thurman DJ, Price PH, Zack MM. Prevalence of active epilepsy and health-related quality of life among adults with self-reported epilepsy in California: California Health Interview Survey, 2003. Epilepsia 2007;48:1904–13. [6] Epilepsy in adults and access to care — United States, 2010. MMWR morbidity and mortality weekly report, 61; 2012. p. 909–13. [7] Kelvin EA, Hesdorffer DC, Bagiella E, Andrews H, Pedley TA, Shih TT, et al. Prevalence of self-reported epilepsy in a multiracial and multiethnic community in New York City. Epilepsy Res 2007;77:141–50. [8] Chong J, Drake K, Atkinson PB, Ouellette E, Labiner DM. Social and family characteristics of Hispanics with epilepsy. Seizure 2012;21:12–6. [9] Chong J, Hesdorffer DC, Thurman DJ, Lopez D, Harris RB, Hauser WA, et al. The prevalence of epilepsy along the Arizona-Mexico border. Epilepsy Res 2013;105: 206–15. [10] Kaiboriboon PB K, Lhatoo S, Koroukian S. Incidence & prevalence of epilepsy among poor health & low income Americans: a longitudinal cohort study (1992–2006). American Epilepsy Society; 2012. [11] Robinson KL, Ernst KC, Johnson BL, Rosales C. Health status of southern Arizona border counties: a Healthy Border 2010 midterm review. Rev Panam Salud Publica 2010;28:344–52. [12] U.S. Census Bureau: State and County QuickFacts. U.S. Census Bureau; 2012 [Accessed 2013, at.]. [13] Fisher RS, Vickrey BG, Gibson P, Hermann B, Penovich P, Scherer A, et al. The impact of epilepsy from the patient's perspective I. Descriptions and subjective perceptions. Epilepsy Res 2000;41:39–51. [14] Manjunath R, Paradis PE, Parise H, Lafeuille MH, Bowers B, Duh MS, et al. Burden of uncontrolled epilepsy in patients requiring an emergency room visit or hospitalization. Neurology 2012;79:1908–16. [15] Halpern MT, Renaud JM, Vickrey BG. Impact of insurance status on access to care and out-of-pocket costs for U.S. individuals with epilepsy. Epilepsy Behav 2011;22:483–9. [16] Begley CE, Famulari M, Annegers JF, Lairson DR, Reynolds TF, Coan S, et al. The cost of epilepsy in the United States: an estimate from population-based clinical and survey data. Epilepsia 2000;41:342–51. [17] Frost FJ, Hurley JS, Petersen HV, Gunter MJ, Gause D. A comparison of two methods for estimating the health care costs of epilepsy. Epilepsia 2000;41:1020–6. [18] Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy in Rochester, Minnesota: 1940–1980. Epilepsia 1991;32:429–45. [19] Hauser WA, Annegers JF, Rocca WA. Descriptive epidemiology of epilepsy: contributions of population-based studies from Rochester, Minnesota. Mayo Clin Proc 1996;71:576–86. [20] Begley CE, Basu R, Reynolds T, Lairson DR, Dubinsky S, Newmark M, et al. Sociodemographic disparities in epilepsy care: Results from the Houston/New York City health care use and outcomes study. Epilepsia 2009;50:1040–50. [21] De Jesus MX, Chenyang. Cross-border health care utilization among the Hispanic population in the United States: implications for closing the health care access gap. Ethn Health 2012:1–18. [22] Sirven JI, Lopez RA, Vazquez B, Van Haverbeke P. Que es la Epilepsia? Attitudes and knowledge of epilepsy by Spanish-speaking adults in the United States. Epilepsy Behav 2005;7:259–65. [23] Palloni A, Arias E. Paradox lost: explaining the Hispanic adult mortality advantage. Demography 2004;41:385–415. [24] Hummer RA. Adult mortality differentials among Hispanic subgroups and nonHispanic whites. Soc Sci Q 2000;81:459–76.
Contents lists available at ScienceDirect
Epilepsy & Behavior journal homepage: www.elsevier.com/locate/yebeh
Healthcare utilization of patients with epilepsy in Yuma County, Arizona: Do disparities exist? Yeeck Sim a, Brandon Nokes b, Seenu Byreddy a, Jenny Chong a, Bruce M. Coull a, David M. Labiner a,c,⁎ a b c
Department of Neurology, University of Arizona, Tucson, AZ, USA College of Medicine, University of Arizona, Tucson, AZ, USA Department of Pharmacy Practice and Science, College of Pharmacy, University of Arizona, Tucson, AZ, USA
a r t i c l e
i n f o
Article history: Received 13 August 2013 Revised 8 October 2013 Accepted 14 October 2013 Available online 5 November 2013 Keywords: Hispanic Epilepsy Cost Border community Patient behavior Seizure Epidemiology
a b s t r a c t The aim of this study was to describe the disparities in healthcare utilization and costs between Hispanic and non-Hispanic patients with seizures or epilepsy. We reviewed the insurance status and healthcare resource utilization data from 2005 to 2008 for all patients with seizures and epilepsy seen at the Yuma Regional Medical Center (YRMC). Charges for medical services provided to Hispanic patients with epilepsy between the ages of 18 and 49 were significantly less than those for non-Hispanic patients with epilepsy (Hispanic: $3167.63 versus non-Hispanic: $5154.36, P b 0.001). Taking into account the differences in insurance status, setting of care, and total number of procedures, we still saw a significant difference in charges between the two groups at the outpatient settings. These data differ from currently available data on national and Eastern US Hispanic patients with epilepsy, suggesting that patients in this border community are somehow different from Hispanics elsewhere in the US. © 2013 Elsevier Inc. All rights reserved.
1. Introduction It is estimated that 1 in 26 individuals in the United States will develop epilepsy or a related seizure disorder by age 75, with a global prevalence between 4 and 6/1000 people [1–3]. Within the US, recent data from the Behavioral Risk Factor Surveillance Survey (BRFSS) report the lifetime prevalence for epilepsy to be between 11.5/1000 and 22/1000 [4]. However, studies show that the prevalence of epilepsy among various ethnic groups, such as Hispanics and non-Hispanic whites, varies by location. Epilepsy has been reported to be more prevalent among Hispanics in New York, whereas in California and Arizona, the prevalence of epilepsy in Hispanics has been reported to be lower than in non-Hispanics [1,5–9]. Regardless of ethnicity, however, epilepsy prevalence among individuals of lower socioeconomic status is consistently high [1,10]. Within the US, individuals in border communities such as Yuma County, Arizona, experience limitations in access to care because of lower socioeconomic status and decreased healthcare facility and provider availability [11]. According to the 2012 US census data, the poverty rate in Yuma is approximately 20% [12]. In a 2010 assessment of border community healthcare access, up to 25% of citizens within border communities, including Yuma County, were uninsured [11], although ⁎ Corresponding author at: Department of Neurology, University of Arizona, 1501 N. Campbell Avenue, PO Box 245023, Tucson, AZ 85724, USA. Fax: +1 520 626 2111. E-mail address: [email protected] (D.M. Labiner). 1525-5050/$ – see front matter © 2013 Elsevier Inc. All rights reserved. http://dx.doi.org/10.1016/j.yebeh.2013.10.018
patients requiring specialized care such as epilepsy management are less likely to be uninsured [5,13]. The Epilepsy Foundation reported in 2000 that 18% of patients with epilepsy were uninsured [13], although more recent data from California place this number closer to 10% [5]. Underinsured patients tend to have greater medical costs related to poorer disease management, and this is especially true for patients with epilepsy [14]. These greater out-of-pocket costs are related to Emergency Department encounters, epilepsy-related comorbidities such as injuries, as well as indirect costs, such as lost wages and unemployment. A recent study reported that uninsured individuals may have decreased quality of care, indicated by significantly fewer outpatient visits, fewer appointments with neurologists, and greater antiepileptic drug (AED) costs compared with patients with private insurance [15]. The direct costs of epilepsy include costs required to treat the underlying disease, such as physician visits, AEDs, and, for some, surgeries. Indirect costs include lost wages, unemployment, and other medical comorbidities [16]. In general, the costs associated with epilepsy and seizure care are largely associated with the indirect costs of seizure development and are attributable to nonepilepsy-related care [15,17]. Epilepsy care in Hispanics has been relatively understudied, especially with regard to access to care and costs of services within border communities. Historically, it has been reported that within the US, access to care for Hispanics was limited because of lack of insurance and increased out-of-pocket costs [11,15]. However, a number of the studies related to Hispanic epilepsy costs either sampled national trends or were limited to the Eastern US [18–20]. The population characteristics in
308
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
Yuma are unique and unlike those in most parts of the US. In Yuma County, over half (60.1%) of the population is Hispanic, whereas 34.6% of the population is non-Hispanic whites [12]. To our knowledge, no previous studies have looked explicitly at the costs of epilepsy care within border communities such as Yuma County, Arizona. To better elucidate the cost of care for patients with epilepsy and seizures within this border community, administrative data from the Yuma Regional Medical Center (YRMC) were used to investigate the healthcare resource utilization and insurance status of patients with seizure or epilepsy conditions between 2005 and 2008. We also sought to understand whether the underlying cost differences, if any, were attributable to the length of hospital stay or the number of procedures performed either in or out of the hospital. We hypothesized that behavior of patients with epilepsy within this border community and healthcare costs may be different from those previously reported for other populations within the US.
Table 1 Patient demographics.
2. Methods
3. Results
Data between January 2005 and July 2008 were obtained from YRMC. The data contained patient demographics including age, sex, and ethnicity; location of residence by city, state, and zip code; discharge status; setting of care; length of stay during a visit; insurance status; and the total charges in dollars for each encounter. For each encounter, up to five diagnoses were available using the International Classification of Disease, Ninth Revision (ICD-9) codes, and up to three procedure columns were reported using the ICD-9 procedure codes. The ICD-9 codes of interest in this study were 345.0–345.9 (epilepsy); 094.89 (epilepsy and syphilis-related); 123.1 (epilepsy and cysticercosisrelated); 123.9 (epilepsy and parasite-related); 333.2 (progressive familial epilepsy); 347.0 (epilepsy and sleep-related); 907.0 (epilepsy due to traumatic injury); 780.3 and 780.39 (convulsions); and 780.31– 780.32 (febrile convulsions). The ICD-9 procedure codes were examined both overall and within each setting of care. Only patients who reported that they were living in Yuma County were included. Hispanics were self-identified at the time of encounter. Hispanic and non-Hispanic Yuma County cases with seizure and/or epilepsy coded during the visit (hereafter known as the group with epilepsy) were assessed together, while Hispanic and non-Hispanic patients who were not coded with seizure and/or epilepsy codes on any visit served as respective control groups. The data were stratified by age: children (0–17 years), adults (18–64 years), and seniors (65 years and older). In more detailed analyses, the adult group was further divided into three groups: 18–33 years, 34–49 years, and 50– 64 years. Insurance status was categorized into three groups: 1) no insurance/ self-pay; 2) public insurance including Arizona Health Care Cost Containment System (AHCCCS —Arizona's Medicaid program), Tricare/ Military/Veteran's Affairs, and other government health programs, such as Medicare; and 3) private insurance (e.g., commercial plans such as Blue Cross and Worker's Compensation). The emergency department and after-hours clinic visits were analyzed together (ED), while hospital outpatient visits were also analyzed together with outpatient clinic visits (OP). A small number of patients visited numerous “other” sources of care such as the Foothills Clinic (a neighborhood facility where mainly non-Hispanic patients were seen), outpatient surgery, and perinatal services. Information about the specialties of the physicians seeing these patients was not available. The data were analyzed with IBM SPSS version 20.0. Comparisons between the Hispanic and non-Hispanic groups with epilepsy were made in terms of total charges, procedure codes, length of stay, insurance status, and setting of care using independent sample t-tests for continuous variables, and the Pearson chi-square test for comparing frequencies. Differences were considered significant at the P ≤ 0.05 level.
3.1. Patient demographics
Total patients Patients with epilepsy Mean age without epilepsy (years) Mean age with epilepsy (years) Male Female With insurance
Hispanic
Non-Hispanic
Total
66,663 960 (1.44%) 28 (+/−22) 28 (+/−24) 520 (44.9%) 440 (40.6%) 869 (90.5%)
69,157 1282 (1.85%) 47 (+/−26) 48 (+/−26) 639 (55.1%) 643 (59.4%) 1121 (95.2%)
135,820 2242 135,820 2242 1159 (51.7%) 1083 (48.3%) 2090 (93.2%)
We also conducted a multiple regression analysis with total charges as the dependent variable and ethnicity (Hispanic/non-Hispanic); length of stay; number of procedures performed; number of comorbid diagnoses; and insurance status — public (yes/no), private (yes/no), and self-pay or no insurance (yes/no) as predictors.
Across all 4 years from January 2005 to June 2008, a total of 960 Hispanic patients with epilepsy and 1282 non-Hispanic patients with epilepsy utilized YRMC facilities. Hispanics represented about half (51.9%) of the number of visits, while the non-Hispanic group represented the other half (48.1%). The average age of Hispanic patients was significantly younger than that of non-Hispanic patients; Hispanic patients with epilepsy were also significantly younger than nonHispanic patients with epilepsy (Table 1). 3.2. Total charges The mean total charge for each visit was significantly less for Hispanics than for non-Hispanics with epilepsy. However, on further analysis, this difference was significant only in adults aged 18–64 but not in children or seniors. Furthermore, when the adults were examined in smaller age groups to account for possibly skewed ages, the difference was significant in the 18- to 49-year group but not significant among those between 50 and 64 years. Those not in the group with epilepsy did not show significant charge differences when stratified by age, except for the 65 and over seniors where the Hispanic patients had higher charges compared with non-Hispanic patients (Table 2). 3.3. Insurance status Among those with epilepsy, proportionately more young Hispanic adults (age: 18–33) had private insurance compared with their nonHispanic cohort. Fewer young Hispanic adults in the group with Table 2 Mean total charges in $ US.
Seizure/epilepsy Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+ Age 18–64 Age 18–49 Without seizure/epilepsy Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
Hispanic
Non-Hispanic
Significance
1605.40 2285.52 4979.16 8624.43 10825.81 3279.91 3167.63
1662.80 3660.98 6406.15 10573.02 10386.80 6846.40 5154.36
NS P b 0.001 P b 0.027 NS NS P b 0.001 P b 0.001
1810.00 2869.16 3006.87 4124.87 6089.02
1837.18 2824.67 3033.08 4005.40 5677.51
NS NS NS NS P b 0.001
All ages in years. Mean in $. NS: not significant.
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
epilepsy had public insurance as opposed to the majority of nonHispanics. In the 50- to 64-year-old group with epilepsy, Hispanics were significantly more likely to be enrolled in public insurance compared with non-Hispanics, while a significantly larger proportion of nonHispanics were enrolled in private insurance compared with Hispanics. There were no significant differences regarding insurance status among other age groups or those who had no insurance among Hispanics and non-Hispanics. Hispanics had lower mean total charges (not just those related to epilepsy) compared with non-Hispanics for all three insurance statuses — public, private, or self-pay. Overall, patients with public insurance averaged higher charges compared with those with private insurance.
309
Hispanic adults were admitted to inpatient services compared with the non-Hispanic cohorts. Significant differences were found between Hispanic and nonHispanic older adults aged 50–64 years in the group with epilepsy at the outpatient and inpatient settings. Similar to those aged 18– 49 years, older Hispanic adults were more likely to use hospital outpatient clinics compared with non-Hispanic older adults. Older Hispanic adults were also less likely to be admitted to inpatient services compared with non-Hispanic older adults. Senior adults aged 65 years and over only had a significant difference in the ED setting where more Hispanic seniors sought care compared with non-Hispanic seniors. For the outpatient setting, Hispanics aged 18–49 had significantly lower charges compared with non-Hispanics. Likewise for the “other” sources of care, Hispanics had significantly lower charges compared with non-Hispanics (Table 4).
3.4. Setting of care 3.5. Length of stay For the young adults aged 18–33 years in the group with epilepsy, fewer Hispanics sought care at the ED, compared with non-Hispanics. On the other hand, in the OP setting, significantly more young Hispanic adults received care compared with non-Hispanics. Only a small group of young Hispanic adults were admitted to inpatient services compared with a larger proportion of non-Hispanics. The young Hispanic adults also visited more “other” sources of care compared with non-Hispanics but were less likely to receive outpatient diagnostic imaging services compared with young non-Hispanic adults (Table 3). Adults with epilepsy aged 34–49 years were similar to their younger counterparts (18–33 years) in terms of where they obtained their care. They did not differ in their use of the ED, but significantly more Hispanic adults sought care at OP setting compared with non-Hispanics. However, this does not reflect Hispanics who received their care from private MD offices, which we were not able to measure at this time. Similarly, fewer Table 3 Setting of care. Hispanic
Non-Hispanic
Significance
ED Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
565 (29.9%) 300 (31.1%) 233 (49.1%) 115 (41.1%) 62 (25.6%)
229 (32.7%) 299 (51.6%) 375 (54.7%) 234 (37.3%) 172 (17.7%)
NS P b 0.001 NS NS P b 0.006
OP Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
947 (50.1%) 480 (49.7%) 132 (27.8%) 59 (21.1%) 59 (25.4%)
345 (49.2%) 162 (28.0%) 99 (14.4%) 97 (15.5%) 270 (27.8%)
NS P b 0.001 P b 0.001 P b 0.04 NS
Inpatient Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
164 (8.7%) 60 (6.2%) 72 (15.2%) 78 (27.9%) 102 (42.1%)
57 (8.1%) 60 (10.4%) 150 (21.9%) 238 (38.0%) 446 (45.9%)
NS P b 0.004 P b 0.005 P b 0.003 NS
Outpatient diagnostic imaging Age 0–17 31 (1.6%) Age 18–33 15 (1.6%) Age 34–49 20 (4.2%) Age 50–64 10 (3.6%) Age 65+ 8 (3.3%)
16 (2.3%) 25 (4.3%) 23 (3.4%) 13 (2.1) 26 (2.7%)
NS P b 0.001 NS NS NS
Other Age 0–17 Age 18–33 Age 34–49 Age 50–64 Age 65+
54 (7.7%) 33 (5.7%) 39 (5.7%) 45 (7.2%) 58 (6.0%)
NS P b 0.001 NS NS NS
182 (9.6%) 111 (11.5%) 18 (3.8%) 18 (6.4%) 11 (4.5%)
All ages in years, with seizure/epilepsy. NS: not significant. Percentages are within respective age groups.
The only significant difference in the average length of stay was between Hispanic and non-Hispanic children with epilepsy from age 0 to 17. There were no differences for adults and seniors with epilepsy. 3.6. Procedures The average number of procedures appears to be underreported or not reported at all. It is also possible that these services were provided by facilities outside of YRMC and, as such, not captured in our data. For the procedures that were reported, the highest frequencies of codes involved are as follows: injection of therapeutic or prophylactic substances (123), injection of antibiotics (97), venous catheterization (91), insertion of endotracheal tubes (86), spinal tap (60), and mechanical ventilation (52) out of a total of 7417 visits. There were only 13 cases of magnetic resonance imaging of the brain reported, and there was only a single case of a computerized tomography scan of the head reported out of 7417 visits. Hispanics in the group with epilepsy had significantly higher number of average procedures during visits for adults aged 18–49 years compared with non-Hispanics. However, Hispanic seniors aged 65 and over had significantly fewer procedures performed, though this is a bit difficult to explain, as the procedures appear to be underreported. Potentially, the Hispanic cohort with epilepsy was receiving cheaper procedures (i.e., more X-rays versus MRI), but the answer is not entirely clear at this point. 3.7. Comorbidities On average, non-Hispanics with epilepsy had more comorbidities compared with their Hispanic counterparts. In the ED setting, the younger non-Hispanic adults with epilepsy (18–33 years and 34– 49 years) had a significantly higher number of comorbidities compared with the Hispanic adults with epilepsy. In the OP setting, the nonHispanic seniors aged 65 years and over had a significantly greater number of comorbidities compared with the Hispanic seniors. In the inpatient care and other care settings, the younger non-Hispanic adults (18–33 years) had a higher average number of comorbidities compared with their cohorts.
Table 4 Mean total charges at setting in $ US.
ED OP Outpatient diagnostic imaging Inpatient Other
Hispanic
Non-Hispanic
Significance
2285.25 695.68 2001.89 17549.24 3956.29
2507.23 888.11 2231.71 18316.57 8294.25
NS P b 0.02 NS NS P b 0.001
Mean in $, age 18–49 years with seizure/epilepsy. NS: not significant.
310
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
Diabetes was the only medical disease that was highly associated with one of the ethnicities (Hispanic). Neither ethnic group showed a higher risk for neurologic comorbidities such as stroke, cognitive impairments, or traumatic brain injury.
3.8. Multiple regression analysis Correlational analyses revealed significantly high and negative correlations between public insurance and private insurance (r = −0.889) and between outpatient and both comorbidity (r = −0.564) and emergency department (r = −0.541), resulting in the public insurance variable and outpatient setting being dropped from the analysis. Results showed that the model was significant with an adjusted R square of 0.723. In other words, the model explained almost three quarters of the variance in the total charges. The main contributors to the total charge are length of stay and the number of procedures performed, with length of stay being the major driver of cost. Hispanics have lower charges. Not being insured (or self-pay) did not have a significant relationship with the total charges (Table 5).
4.3. Differences in comorbidities Although non-Hispanic adults aged 18–49 years had a significantly higher mean number of comorbidities in the ED and inpatient settings, their mean total charge was significantly higher than the Hispanics only in the outpatient setting. The higher number of comorbidities may also play a role in the increased mean total charges incurred by the non-Hispanic patients overall. Whether this implies that the outpatient individuals are somehow different from those in the other settings or whether this is an underreporting of medical comorbidity is unclear. Similarly, the difference in comorbidities may be due to the larger overall number of young Hispanic patients with epilepsy. Future studies should establish a quality-of-care metric to assess whether or not the differences in outpatient care are attributable to the differences in delivery of care or due to the differences in populations served, as we had insufficient data to clarify this. If the Hispanics within this community are receiving the same or better care than their nonHispanic counterparts, this would allow for improved care at reduced costs for all patients with epilepsy within Yuma, regardless of ethnicity.
4. Discussion
4.4. Differences in insurance status
4.1. Differences in preferred location of care
It is also worth noting that more Hispanics with epilepsy had private insurance, particularly in the 18–33 age group, and yet accrued lower charges than the similarly insured non-Hispanic patients with epilepsy of the same age within the outpatient setting. Whether or not this bears relation to their overall healthcare costs is uncertain. Interestingly, patients with public insurance were consistently charged higher fees compared with patients with private insurance, in both Hispanic and non-Hispanic patients. This is not due to contracted rates as our data analysis utilized actual charges prior to contractual adjustments. This is in contrast to previous analysis of the Medical Expenditure Panel Survey (MEPS), in which uninsured and underinsured have greater out-of-pocket expenses [15]. Also of interest, the preferred setting of care and insurance status of Hispanic patients with epilepsy and seizures in Yuma is starkly different from that of Hispanics within previous national and Eastern US studies, in which Hispanics were more likely to be uninsured and utilize the ED for medical treatment [15,18,19]. This point stresses the lack of generalizability of the Hispanic patient population and suggests that the behavior of Hispanic patients with epilepsy/seizures in this border community is different from the behavior of Hispanic patients with epilepsy elsewhere. It is certainly possible that some care (particularly that associated with out-ofpocket costs) could be obtained across the border for Hispanic patients, and this could manifest as lower cost in Yuma County, while the actual costs borne by the patient would be greater.
The principal finding in this study was that Hispanic patients with epilepsy or seizures had significantly lower charges compared with non-Hispanic patients in Yuma County, Arizona. This may be, in part, attributable to the Hispanic patients' preferred setting of care (OP as opposed to ED). Hispanic patients with epilepsy displayed a healthcare utilization behavior of more outpatient visits and less ED, outpatient diagnostic imaging, and inpatient visits. Although this could account for overall lower charges/costs for Hispanic patients, even within the outpatient clinic and “other” settings, Hispanics still had significantly lower average charges compared with non-Hispanics. The lower outpatient charges are difficult to explain and may be attributable to variability (or disparity) in outpatient patient care.
4.2. Differences in procedures and length of stay The regression analysis indicated that length of stay and mean number of procedures were the main contributors to the total cost in both Hispanics and non-Hispanics. The influence of these variables is not apparent by univariate or multivariate analysis but may play a significant role when other potential factors are accounted for within the analysis.
Table 5 Multiple regression analysis.
Possible confounders
Insurance Site of care
Constant Hispanic Age Length of stay Comorbidities Procedures Private ins. Self-pay/no ins. Inpatient Other sites ODI ED
Unstandardized coefficients
Standardized coefficients
Beta
Standard error
Beta
−301.032 −468.118 16.648 2957.250 −171.734 4763.268 669.161 488.681 3296.985 2512.274 1832.586 2049.973
203.246 145.745 3.334 36.366 62.860 119.213 162.271 315.200 285.735 279.399 437.408 192.753
−.021 .038 .645 −.026 .270 .026 .009 .116 .059 .025 .087
t-Stat
−1.481 −3.212 4.993 81.318 −2.732 39.956 4.124 1.550 11.539 8.992 4.190 10.635
Sig.
NS .001 .000 .000 .006 .000 .000 NS .000 .000 .000 .000
CI: confidence interval. ED: emergency department. Ins: insurance. ODI: outpatient diagnostic imaging. Sig: significant. NS: not significant.
95.0% CI for beta Lower bound
Upper bound
−699.448 −753.817 10.111 2885.963 −294.955 4529.580 351.069 −129.194 2736.871 1964.578 975.152 1672.128
97.383 −182.420 23.184 3028.538 −48.513 4996.956 987.254 1106.555 3857.100 3059.969 2690.020 2427.818
Y. Sim et al. / Epilepsy & Behavior 31 (2014) 307–311
4.5. Less epilepsy care taking place within the US — the healthy immigrant hypothesis De Jesus and Chenyang reported that according to the Pew Hispanic Survey, Hispanics are more likely to seek care in Mexico owing to lack of continuous insurance coverage and a perceived lack of quality of care within the US [21]. A potential reason for less Hispanic epilepsy management in the US is a lack of understanding within the US Hispanic community about epilepsy [22]. Hispanics may be less likely to pursue chronic epilepsy management. Alternatively, the patients may instead be receiving chronic care in Mexico after their initial seizure episode, but this seems less likely, considering their better insurance status compared with the non-Hispanics. Others have reported that Hispanics have less chronic illnesses due to the healthy immigrant hypothesis [23,24]. Among those Hispanics who do have chronic illnesses, some studies have reported that they seek management in Mexico [21]. Similarly, Hispanic patients with epilepsy may not need as much inpatient or outpatient care because of their overall younger age and potentially overall better health status, but our study still shows significant differences in the outpatient setting compared with the nonHispanic young adults with epilepsy. 4.6. Limitations There are several limitations worth considering for this study. Principally, there are a number of factors not readily discernible from health records, such as degree of education, socioeconomic status, time within the US, and even the means of diagnosis. Further, our study design is retrospective in nature and does not describe this patient population as well as a prospective study would. It is also worth considering that a number of epilepsy patients within this border community may not be presenting for care, though it is difficult to quantitatively define the percentage of the Yuma population with epilepsy who did not do so during the measured time frame. However, the relative prevalence of epilepsy detected within this study is roughly concordant with a previous study we conducted along the Arizona border [9]. Lastly, although YRMC is the only major hospital in Yuma, not every outpatient clinic, including those provided by Indian Health Services, is accounted for. As such, future studies should consider including as many institutions as possible in order to better describe the patient population with epilepsy within Yuma. 4.7. Conclusion This study highlights the differences in patient behavior in regard to ED and OP care preferences as well as overall charge differences between border communities and previously reported Hispanic patients from the eastern US. Despite overall cost differences, there does not appear to be limitations in access to care between Hispanic and non-Hispanic patients with epilepsy. We hope that by identifying the differences in epilepsy and seizure care within the border community, quality of care could be improved for this patient population through further elucidation of these cost differences. In light of our findings, we will continue to assess patient behavior within border communities, with the consideration that the behavior of Hispanics within these communities may be different and not applicable to that of other Hispanics within the US.
311
Acknowledgments We thank and acknowledge Yuma Regional Medical Center for allowing us to use and analyze their patient data and Dr. William Johnson of the Center for Health Information and Research at Arizona State University for providing us with the YRMC database. References [1] Bartolini E, Bell GS, Sander JW. Multicultural challenges in epilepsy. Epilepsy Behav 2011;20:428–34. [2] Cardarelli WJ, Smith BJ. The burden of epilepsy to patients and payers. Am J Manag Care 2010;16:S331–6. [3] England MJ, Liverman CT, Schultz AM, Strawbridge LM. Epilepsy across the spectrum: promoting health and understanding. A summary of the Institute of Medicine report. Epilepsy Behav 2012;25:266–76. [4] Kobau R, Zahran H, Grant D, Thurman DJ, Price PH, Zack MM. Prevalence of active epilepsy and health-related quality of life among adults with self-reported epilepsy in California: California Health Interview Survey, 2003. Epilepsia 2007;48:1904–13. [5] Kobau R, Zahran H, Grant D, Thurman DJ, Price PH, Zack MM. Prevalence of active epilepsy and health-related quality of life among adults with self-reported epilepsy in California: California Health Interview Survey, 2003. Epilepsia 2007;48:1904–13. [6] Epilepsy in adults and access to care — United States, 2010. MMWR morbidity and mortality weekly report, 61; 2012. p. 909–13. [7] Kelvin EA, Hesdorffer DC, Bagiella E, Andrews H, Pedley TA, Shih TT, et al. Prevalence of self-reported epilepsy in a multiracial and multiethnic community in New York City. Epilepsy Res 2007;77:141–50. [8] Chong J, Drake K, Atkinson PB, Ouellette E, Labiner DM. Social and family characteristics of Hispanics with epilepsy. Seizure 2012;21:12–6. [9] Chong J, Hesdorffer DC, Thurman DJ, Lopez D, Harris RB, Hauser WA, et al. The prevalence of epilepsy along the Arizona-Mexico border. Epilepsy Res 2013;105: 206–15. [10] Kaiboriboon PB K, Lhatoo S, Koroukian S. Incidence & prevalence of epilepsy among poor health & low income Americans: a longitudinal cohort study (1992–2006). American Epilepsy Society; 2012. [11] Robinson KL, Ernst KC, Johnson BL, Rosales C. Health status of southern Arizona border counties: a Healthy Border 2010 midterm review. Rev Panam Salud Publica 2010;28:344–52. [12] U.S. Census Bureau: State and County QuickFacts. U.S. Census Bureau; 2012 [Accessed 2013, at.]. [13] Fisher RS, Vickrey BG, Gibson P, Hermann B, Penovich P, Scherer A, et al. The impact of epilepsy from the patient's perspective I. Descriptions and subjective perceptions. Epilepsy Res 2000;41:39–51. [14] Manjunath R, Paradis PE, Parise H, Lafeuille MH, Bowers B, Duh MS, et al. Burden of uncontrolled epilepsy in patients requiring an emergency room visit or hospitalization. Neurology 2012;79:1908–16. [15] Halpern MT, Renaud JM, Vickrey BG. Impact of insurance status on access to care and out-of-pocket costs for U.S. individuals with epilepsy. Epilepsy Behav 2011;22:483–9. [16] Begley CE, Famulari M, Annegers JF, Lairson DR, Reynolds TF, Coan S, et al. The cost of epilepsy in the United States: an estimate from population-based clinical and survey data. Epilepsia 2000;41:342–51. [17] Frost FJ, Hurley JS, Petersen HV, Gunter MJ, Gause D. A comparison of two methods for estimating the health care costs of epilepsy. Epilepsia 2000;41:1020–6. [18] Hauser WA, Annegers JF, Kurland LT. Prevalence of epilepsy in Rochester, Minnesota: 1940–1980. Epilepsia 1991;32:429–45. [19] Hauser WA, Annegers JF, Rocca WA. Descriptive epidemiology of epilepsy: contributions of population-based studies from Rochester, Minnesota. Mayo Clin Proc 1996;71:576–86. [20] Begley CE, Basu R, Reynolds T, Lairson DR, Dubinsky S, Newmark M, et al. Sociodemographic disparities in epilepsy care: Results from the Houston/New York City health care use and outcomes study. Epilepsia 2009;50:1040–50. [21] De Jesus MX, Chenyang. Cross-border health care utilization among the Hispanic population in the United States: implications for closing the health care access gap. Ethn Health 2012:1–18. [22] Sirven JI, Lopez RA, Vazquez B, Van Haverbeke P. Que es la Epilepsia? Attitudes and knowledge of epilepsy by Spanish-speaking adults in the United States. Epilepsy Behav 2005;7:259–65. [23] Palloni A, Arias E. Paradox lost: explaining the Hispanic adult mortality advantage. Demography 2004;41:385–415. [24] Hummer RA. Adult mortality differentials among Hispanic subgroups and nonHispanic whites. Soc Sci Q 2000;81:459–76.
