SEARCH

SEARCH BY CITATION

Keywords:

  • Health care utilization;
  • Health survey;
  • Disparities;
  • Asthma;
  • Diabetes;
  • Epilepsy;
  • Migraine

Summary

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Purpose:  (1) To determine whether health resource utilization (HRU) and unmet health care needs differ for individuals with epilepsy compared to the general population or to those with another chronic condition (asthma, diabetes, migraine); and (2) to assess the association among epilepsy status, sociodemographic variables and HRU.

Methods:  Data on HRU were assessed using the 2001–2005 Canadian Community Health Surveys, a nationally representative population-based survey. Weighted estimates of association were produced as adjusted odds ratio with 95% confidence intervals, and logistic regression was used to explore the association between sociodemographic variables and HRU in those with epilepsy. All data on disease status, HRU, and unmet health care needs were self-reported.

Key Findings:  Individuals with epilepsy had the highest rate of hospitalizations and the highest mean number of consultations with physicians. Despite higher rates of consultation with psychologists and social workers compared to the general population, those with epilepsy were significantly more likely to say they had unmet mental health care needs. People with epilepsy were also less likely to use dental services compared to the general population. Epilepsy was a significant predictor of HRU in logistic regression models.

Significance:  Given the prevalence of psychiatric comorbidities in those with epilepsy, it is concerning that this group perceives unmet mental health care needs. It is also troublesome that there was decreased utilization of dental health care resources in those with epilepsy considering that these patients are more likely to have poor oral health. Although individuals with epilepsy use more health care services than the general population, this increase appears to be insufficient to address their health care needs.


Background/Rationale

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Epilepsy is one of the most commonly reported neurologic conditions in primary care, affecting approximately 1% of the population (Murray et al., 1994). Although attempts have been made to determine factors that influence the cost of epilepsy care and patterns of health resource utilization (HRU) in those with epilepsy, such as sociodemographic variables, there have been conflicting results (Bautista et al., 2008; Jetté et al., 2008; Kurth et al., 2010). The discrepancies may in part relate to the different methodologies used in these studies. Ideally, HRU patterns in those with epilepsy should be studied at the population-based level, because such studies have the advantage of using nonselected populations of people with epilepsy. Data relying on patients at a major referral center will be biased as they will include more severely affected patients. Population-based studies allow for comparisons to be made with the general population and with people with other chronic health conditions. The Ontario population health survey found that people with epilepsy have higher HRUs than the general population or the chronically ill (Wiebe et al., 1999). However, national data, Canadian or otherwise, studying HRU patterns in those with epilepsy compared to individuals with other chronic conditions have not been examined. In addition, there is a paucity of evidence available to determine whether people with epilepsy still have unmet health care needs despite increased HRU compared to the general population.

We used a large, national population-based health survey to determine whether HRU and perceived unmet health care needs differ in patients with epilepsy compared to the general population and to those with other chronic health conditions. Comparisons were made to people with asthma and diabetes, as these conditions usually require chronic therapy, as does epilepsy, and migraine because, like epilepsy it is an episodic chronic condition. Our second objective was to determine whether epilepsy and sociodemographic variables were predictors of HRU and patient-perceived unmet health care needs.

Methods

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Data source and study population

The Canadian Community Health Survey (CCHS) is a national population-based cross-sectional survey of Canadians. The CCHS was administered to 400,055 Canadians between the years 2001 and 2005, excluding those who live on Indian reserves, in institutions, in certain remote areas, or who are full-time members of the Canadian Armed Forces (Statistics Canada, 2009). Cycle 1.1 was conducted with individuals age 12 and older, whereas cycles 2.1 and 3.1 only included individuals who were at least 15 years of age (Statistics Canada, 2009). CCHS surveys use a multistage stratified cluster design to randomly select a representative sample of Canadian household residents that represents approximately 98% of Canadians living in the provinces, 97% from the Northwest Territories, 90% from the Yukon, and 71% from Nunavut (Statistics Canada, 2009). Interviews were conducted in-person in the subject’s home or via telephone. In Cycle 1.1, approximately 50% of interviews were completed in the home and 50% completed by telephone (St-Pierre & Béland, 2004). For cycle 2.1, approximately 30% of interviews were conducted in the home and 70% were completed via telephone (St-Pierre & Béland, 2004). For cycle 3.1, approximately 40% of interviews were completed in the home and 60% by telephone. Self-reported information about professionally diagnosed health conditions, utilization of health services, and health-related behaviors were elicited (Statistics Canada, 2009). Proxy responses were obtained from approximately 6% of the sample and were permitted only in cases whereby the interviewee was not able to complete an interview for health reasons or because of language barriers (Statistics Canada, 2009). A comprehensive review of the CCHS is available from the Statistics Canada website (Statistics Canada, 2009) and from Beland’s 2002 methodologic overview (Beland, 2002).

Diagnosis of chronic conditions

Data on chronic disease status were obtained from the following question: “Now I’d like to ask you about certain chronic health conditions which you may have. We are interested in long-term conditions that are expected to last, or have already lasted, 6 months or more and that have been diagnosed by a health professional.” A list of chronic conditions followed, and respondents were asked to respond yes or no.

Variables studied

The variables studied included: (1) demographics (age, gender, marital status, income, education, urban or rural residence); (2) health resource use (having a regular family doctor, consultations with health professionals including allied health specialists and hospitalizations in the past year, length of stay in hospital); and (3) perceived unmet health care needs. Data on unmet health care needs were obtained for the following two questions. The first question asked: “During the past 12 months, was there ever a time when you felt that you needed care but you did not receive it?” If the individual responded yes, they were asked the following question: “Again, thinking of the most recent time, what was the type of care that was needed?” Response options for this question included: (1) treatment of a physical health problem, (2) treatment of an emotional or mental health problem, (3) a regular checkup (including regular prenatal care), (4) care of an injury, (5) other. Individuals were classified as having an unmet health care need if they responded “yes” to the first question. It is not possible to assess from this dataset if this unmet health care need represents a single occasion where care was not received or multiple occasions where care was needed but was not received.

Details on CCHS questions and variables are available from Statistics Canada (2009).

Statistical methods

Microdata from cycles 1.1 (2001), 2.1 (2003), and 3.1 (2005) of the CCHS were pooled to create a new dataset, and weighted estimates were constructed to examine the proportion of individuals using specific health resources and perceived unmet health care needs. Statistics Canada’s recommendations for merging surveys were followed (Thomas & Wannell, 2009). Because these are cross-sectional surveys, it is unlikely that any one individual would have been included in multiple cycles; however, no personally identifying information was available so we are unable to assess the likelihood of this. However, given that all questions asked about behaviors and resource utilization occurred in the past year, and surveys were conducted 2 years apart, this is less likely to introduce clustering or recall bias into the results. Descriptive statistics were calculated and backwards multivariable logistic regression was used to construct four separate models examining whether epilepsy or other factors were associated with consulting a health professional in the past year, having a regular medical doctor, being hospitalized overnight, and having a perceived unmet health care need. All variables were initially entered into the model, and variables that did not achieve statistical significance (p < 0.05) were removed one at a time, until all remaining variables were statistically significant. Because missing data were minimal for most predictor variables, list-wise deletion was used to address missing data.

Ethical approval

This study was approved by the Conjoint Health Research Ethics Board at the University of Calgary.

Results

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Demographics

A response rate of 85% was achieved for cycles 1.1, 2.1, and 3.1, with 400,055 individuals enrolled in the study. The weighted prevalence was 0.6% [95% confidence interval (CI) 0.6–0.6] for active epilepsy, 4.5% (95% CI 4.5–4.6) for diabetes, 10.1% (95% CI 9.6–10.3) for migraine, and 8.4% (95% CI 8.2–8.5) for asthma. Demographic characteristics are listed in Table 1 and the Appendix. Of all the groups studied, those with epilepsy were least likely to be married, more likely to be low income earners, and less likely to have completed postsecondary education.

Table 1.   Demographic characteristics of sample
 General populationAsthmaDiabetesEpilepsyMigraine
  1. Data for age are presented as mean ± standard deviation. Data for female sex, marital status, income, residence, and education are reported as percentage of respondents with their 95% confidence intervals.

Number400,05535,16622,4322,55539,797
Age45.4 ± 20.239.5 ± 19.262.9 ± 14.743 ± 17.741.1 ± 16.8
Female50.7 (50.5–51.0)58.9 (58.1–59.7)46.9 (45.9–48.0)50.9 (47.9–53.8)71.4 (70.7–72.1)
Marital status     
 Married/common law58.4 (58.2–58.6)50.3 (49.5–51.1)67.6 (66.7–68.6)48.0 (45.0–50.9)59.7 (58.9–60.4)
 Widowed/separated/divorced11.7 (11.6–11.8)12.5 (12.0–12.9)22.6 (21.8–23.4)13.6 (11.8–15.4)12.1 (11.6–12.5)
 Single29.9 (29.7–30.1)37.2 (36.4–38.0)9.8 (9.2–10.3)38.4 (35.6–41.2)28.2 (27.6–28.9)
Income quintiles (household)     
 Lowest14.1 (13.9–14.3)17.6 (17.0–18.2)26.2 (25.3–27.1)27.1 (24.4–29.7)15.6 (15.1–16.1)
 Low middle17.2 (17.0–17.4)17.3 (16.7–18.0)21.7 (20.8–22.6)18.4 (15.9–20.8)16.6 (16.0–17.2)
 Middle21.3 (21.1–21.6)20.5 (19.8–21.2)18.6 (17.6–19.5)17.7 (15.0–20.3)21.2 (20.5–21.9)
 Upper middle23.2 (22.9–23.4)22.0 (21.2–22.7)13.8 (12.9–14.6)16.9 (14.3–19.5)23.3 (22.6–24.0)
 Highest24.2 (23.9–24.4)22.6 (21.9–23.4)19.7 (18.8–20.6)20.0 (17.4–22.6)23.3 (22.6–24.0)
Residence     
 Urban81.6 (81.5–81.8)82.2 (81.7–82.8)80.1 (79.4–80.8)80.9 (78.8–83.1)82.0 (81.5–82.5)
 Rural18.4 (18.2–18.5)17.8 (17.2–18.3)19.9 (19.2–20.6)19.1 (16.9–21.2)18.0 (17.5–18.5)
Highest level of education completed     
 Did not complete high school26.6 (26.4–26.9)32.4 (31.7–33.2)38.5 (37.5–39.5)38.4 (35.5–41.3)24.4 (23.7–25.0)
 Completed high school17.3 (17.3–17.5)15.4 (14.8–16.0)15.8 (15.0–16.5)18.5 (16.2–20.8)17.7 (17.1–18.3)
 Some postsecondary8.3 (8.1–8.4)9.3 (8.8–9.8)5.9 (5.4–6.4)8.1 (6.5–9.8)8.9 (8.4–9.3)
 Completed postsecondary47.8 (47.5–48.0)42.8 (42.0–43.6)39.8 (38.8–40.9)35.0 (32.1–37.9)49.1 (48.3–49.9)

Health resource utilization

In this sample, 85.2% (95% CI 85.0–85.4) of the general population reported having a regular medical doctor (Fig. 1). This proportion was higher in those with asthma (88.4%, 95% CI 87.9–89.0), diabetes (96.0%, 95% CI: 95.6–96.4), epilepsy (93.4%, 95% CI: 92.0–94.8), or migraine (89.4%, 95% CI: 88.9–89.9). Those with a chronic condition were also more likely to have consulted any health professional in the past year (general population: 94.2%, 95% CI 94.1–94.3; asthma 96.9%, 95% CI 96.6–97.1; diabetes 98.3%, 95% CI 98.0–98.6; epilepsy 98.0%, 95% CI 97.3–98.6; migraine 96.8%, 95% CI 96.6–97.1) (Fig. 1).

image

Figure 1.   Contacts with physicians in the past 12 months. *p < 0.05 compared to the general population.

Download figure to PowerPoint

As seen in Fig. 1, 77.7% (95% CI 77.5–77.9) of the general population reported visiting their family doctor in the last year, with a mean of 3.2 (standard deviation, SD 5.9) consultations over the last 12 months. Those with asthma (4.8, SD 7.7), epilepsy (5.5, SD 9.1), and migraine (5.1, SD 8.4) each had a higher mean number of consultations with their family physician in the preceding year.

Individuals with any of the chronic diseases studied were more likely than members of the general population (27.2%, 95% CI 27.0–27.4) to report consulting a medical doctor other than their family physician in the past year, with the highest proportion found in those with epilepsy (44.5%, 95% CI 41.5–47.5) (Fig. 1). However, individuals with diabetes (46.4%, 95% CI 45.4–47.5) and epilepsy (55.9%, 95% CI 52.9–58.8) were less likely to report consulting a dentist in the past year than the general population (62.6%, 95% CI 62.3–62.8) (Fig. 2). Although those with asthma (41.8%, 95% CI 41.0–42.6), diabetes (61.8%, 95% CI 60.8–62.9), and migraine (41.6%, 95% CI 40.9–42.4) were more likely to have seen an optometrist in the past year than the general population (39.1%, 95% CI 38.9–39.4), a significant difference was not found for individuals with epilepsy (42.1%, 95% CI 39.2–45.0). Other commonly consulted health professionals included complementary and alternative health care (CAM) practitioners, chiropractors, nurses, and physiotherapists (Fig. 2). Although 8.0% (95% CI 7.8–8.1) of the general population reported being hospitalized overnight in the past year, the proportion of people with diabetes (17.2%, 95% CI 16.5–18.0) or epilepsy (17.9%, 95% CI 15.8–20.0) who were hospitalized was more than double this rate (Fig. 3). Individuals with asthma (12.3%, 95% CI 11.8–12.8) and migraine (11.0%, 95% CI 10.6–11.5) were also more likely to report being hospitalized than the general population. Individuals with diabetes (13.7%, 95% CI 11.3–16.1) and epilepsy (14.3%, 95% CI 8.6–20.0) were also more likely to report being admitted to the hospital for the treatment of an injury than the general population (7.8%, 95% CI 7.4–8.2) (Fig. 3). On average, individuals with any of the chronic conditions studied reported a longer mean length of hospital stay (asthma: 8.9, SD 22.0 days; diabetes: 14.3, SD 31.5 days; epilepsy: 13.7, SD 30.9 days; migraine: 10.1, SD 14.0 days) than the general population (8.6, SD 22.3 days).

image

Figure 2.   Contacts with other health professionals in the past year. *p < 0.05 compared to the general population.

Download figure to PowerPoint

image

Figure 3.   Hospital admissions in the past year. *p < 0.05 compared to the general population.

Download figure to PowerPoint

Perceived unmet health care needs

Despite reporting significantly higher HRU, individuals with asthma (17.5%, 95% CI 16.9–18.1), epilepsy (17.9%, 95% CI 15.6–20.2), and migraine (21.0%, 95% CI 20.4–21.6) were also more likely to report that they had unmet health care needs than the general population (11.6%, 95% CI 11.5–11.8) (Fig. 4). Of individuals perceiving unmet health care needs, individuals with asthma were more likely than the general population to report that they did not receive care for a physical problem (75.4%, 95% CI 73.8–77.0); whereas individuals with epilepsy (14.5%, 95% CI 9.8–19.1) and migraine (11.1%, 95% CI 10.1–12.0) were more likely to report that they did not receive care for a mental health problem (Fig. 4). Although not reporting a higher rate of unmet health care needs overall compared to the general population, individuals with diabetes were still more likely to report that in the past year they did not receive care for a physical problem (77.2%, 95% CI 74.7–79.8) (Fig. 4).

image

Figure 4.   Perceived unmet health care needs. The individuals who reported not receiving care for a physical problem, mental health care problem, not receiving a checkup, not receiving care for an injury, or not receiving care for another problem are reported as a percentage of the individuals perceiving an unmet health care need. *p < 0.05 compared to the general population.

Download figure to PowerPoint

Predictors of health resource utilization and perceived unmet health care needs

Epilepsy remained a significant predictor for all measures studied after adjusting for various sociodemographic factors (Table 2). Having epilepsy increased the odds of consulting a health professional in the past year, having a regular medical doctor, being hospitalized overnight, and having unmet health care needs. The same trend was observed for female patients and for individuals with asthma, diabetes, or migraines. However, no differences were found between the chronic conditions. Individuals who lived in a rural area were less likely to have consulted a health professional in the past year and more likely to have been hospitalized than those who lived in urban areas. Those who were not married or in a common-law relationship were less likely to have a regular medical doctor or to have consulted a health professional in the past year. Those of high income were more likely than those of low income to have a regular medical doctor, and were the least likely to be hospitalized overnight and to have perceived unmet health care needs.

Table 2.   Logistic regression models
Predictor variableMeasure
Consulted a health professional OR (95% CI)Has a regular medical doctor OR (95% CI)Was hospitalized overnight OR (95% CI)Has a perceived unmet health care need OR (95% CI)
  1. CI, confidence interval; OR, adjusted odds ratio; ref, reference standard.

  2. Models were constructed using backwards multivariable logistic regression. Variables not achieving statistical significance (p < 0.05) were removed one at a time until all remaining variables were statistically significant.

Asthma    
 0 = NoRefRefRefRef
 1 = Yes1.91 (1.71–2.12)1.37 (1.28–1.45)1.59 (1.50–1.68)1.56 (1.48–1.64)
Diabetes    
 0 = NoRefRefRefRef
 1 = Yes3.65 (3.04–4.39)2.97 (2.65–3.32)1.92 (1.80–2.05)1.20 (1.11–1.30)
Epilepsy    
 0 = NoRefRefRefRef
 1 = Yes3.23 (2.19–4.76)2.44 (1.91–3.13)2.36 (1.99–2.80)1.56 (1.30–1.86)
Migraine    
 0 = NoRefRefRefRef
 1 = Yes1.69 (1.53–1.88)1.45 (1.36–1.53)1.41 (1.33–1.49)1.89 (1.81–1.98)
Sex    
 1 = MaleRefRefRefRef
 2 = Female2.44 (2.31–2.57)1.93 (1.87–2.00)1.44 (1.38–1.49)1.24 (1.20–1.28)
Age group    
 0 = 25–44RefRefRefRef
 1 = 12–241.63 (1.49–1.78)1.17 (1.11–1.24)1.05 (0.97–1.14)0.86 (0.80–0.91)
 2 = 45–641.36 (1.28–1.44)1.93 (1.85–2.01)0.83 (0.79–0.87)0.81 (0.78–0.85)
 3 = ≥652.57 (2.37–2.78)4.66 (4.37–4.97)1.57 (1.48–1.65)0.48 (0.45–0.51)
Highest level of education    
 1 = Didn’t graduate from high schoolRefRefRefRef
 2 = High school graduate1.02 (0.95–1.10)0.88 (0.84–0.93)0.93 (0.88–0.99)1.18 (1.11–1.25)
 3 = Some postsecondary1.19 (1.07–1.31)0.79 (0.74–0.84)0.96 (0.89–1.03)1.66 (1.56–1.78)
 4 = Completed postsecondary1.49 (1.39–1.59)0.74 (0.71–0.78)0.93 (0.88–0.98)1.47 (1.40–1.55)
Residence    
 1 = UrbanRefRefRefNot included in final model
 2 = Rural0.83 (0.79–0.87)1.05 (1.01–1.09)1.08 (1.04–1.13) 
Marital status    
 1 = Married/common lawRefRefRefRef
 2 = Widowed/separated/divorced0.87 (0.81–0.93)0.88 (0.84–0.93)0.94 (0.89–0.99)1.16 (1.10–1.22)
 3 = Single0.80 (0.74–0.85)0.68 (0.65–0.71)0.60 (0.56–0.64)1.03 (0.98–1.08)
Household income quintiles    
 1 = LowestRefRefRefRef
 2 = Low middle1.23 (1.14–1.31)1.32 (1.26–1.39)0.81 (0.77–0.86)0.80 (0.76–0.84)
 3 = Middle1.71 (1.58–1.84)1.61 (1.53–1.69)0.67 (0.64–0.72)0.71 (0.67–0.75)
 4 = Upper middle2.49 (2.28–2.71)1.94 (1.83–2.05)0.59 (0.55–0.63)0.63 (0.60–0.67)
 5 = Highest2.21 (2.03–2.39)1.98 (1.88–2.09)0.60 (0.56–0.64)0.65 (0.61–0.69)

Discussion

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

We used a large Canadian population-based health survey to determine whether HRU and perceived unmet health care needs differ in patients with epilepsy compared to the general population and to those with other chronic health conditions. Many aspects of HRU were significantly higher in all of the chronic conditions examined (epilepsy, asthma, diabetes, and migraine) compared to the general population. Notable exceptions were that those with epilepsy had the highest rate of hospitalizations and the highest mean number of consultations with physicians, but were less likely to use dental health care resources.

Despite significantly higher HRU, people with epilepsy, asthma, and migraine were more likely to perceive unmet health care needs than the general population. People with epilepsy in particular were most likely to report that they did not receive care for a mental health problem. Using logistic regression models we also determined that having epilepsy is a significant independent predictor for consulting a health professional, having a regular medical doctor, being hospitalized overnight, and having self-perceived unmet health care needs.

Health resource utilization

Many aspects of HRU were increased in people with epilepsy, which is not surprising given that this population has a poorer pattern of health-related behaviors and poorer health status compared to the general population (Hinnell et al., 2010). People with any of the chronic conditions examined in this study were significantly more likely to have a regular family doctor than people in the general population. They were also significantly more likely to consult both their family doctor and another medical doctor in the preceding 12 months. For most of the chronic conditions, including epilepsy, increased contact was also reported with other health care professionals such as nurses, physiotherapists, psychologists, social workers, speech pathologists, and with self-help groups. Some of this increased HRU appears to be generalizable to people with chronic health conditions. However, individuals with epilepsy were more likely to consult with a social worker or speech pathologist than individuals with the other chronic conditions studied. Increased likelihood of consultation with a social worker may be due to the increased social stigma of epilepsy (Kilinç & Campbell, 2009; Smith et al., 2009) as well as the impact this condition has on activities such as driving and employment (McCagh et al., 2009). People with epilepsy, particularly focal epilepsies, may also be more likely to have speech disturbances either related to the epilepsy itself, to other associated neurological conditions, or to previous epilepsy surgery.

On the other hand, people with epilepsy were significantly less likely to report consulting a dentist in the past year compared with the general population. Although visits with physicians and most other health care professionals in Canada are covered at no cost to the patient through universal health care, visits with dentists are not. Therefore, it is not surprising that those with epilepsy, who are the lowest income earners overall, are less likely to consult a dentist. The low self-reported rate of contact with a dentist is concerning, given that many people with epilepsy are treated with phenytoin, which can lead to gingival hyperplasia (see Meraw & Sheridan, 1998), and because people with epilepsy have worse oral health and dental status compared to age-matched controls (Károlyházy et al., 2003).

People with all of the chronic conditions studied here reported an increased rate of overnight hospitalizations compared with the general population. Although people with epilepsy may indeed have more medical events that require overnight admission (Kurth et al., 2010), it is also possible that this group is more likely to be admitted to hospital overnight for a similar problem due to a lack of social support at home (Charyton et al., 2009). People with epilepsy or diabetes were also more likely to be admitted for treatment of an injury. This is not surprising considering that people with epilepsy are more likely than the general population to sustain an injury leading them to seek medical attention (Kwon et al., 2010).

Perceived unmet health care needs

Although many aspects of HRU were increased in people with epilepsy, asthma, and migraine, these populations still reported more unmet health care needs compared with the general population. Approximately 18% of individuals with epilepsy reported unmet health care needs. Within this 18% of the epilepsy population, approximately 70% perceived they did not receive care for a physical problem, 14% perceived they did not receive care for a mental health problem, 8% did not receive a checkup, 7% did not receive care for an injury, and 10% felt they did not receive care for another problem. Therefore, some patients perceived unmet health care needs for more than one type of problem. In particular, people with epilepsy reported unmet care for mental health problems compared to the general population with unmet health care needs. Given the increased prevalence of psychiatric comorbidities in people with epilepsy compared to the general population, including increased suicidal ideation (Tellez-Zenteno et al., 2007), the fact that patients perceive their mental health care needs as being unmet is concerning. Our results support previous studies, which have found patients with partial epilepsies have unique unmet treatment needs including a high incidence of depression and lack of care for suicidality (Schmitz et al., 2010). Despite the high prevalence, depression in patients with epilepsy is often unrecognized or untreated (Kanner et al., 2000; Wiegartz et al., 1999). Across the world, the epilepsy treatment gap, meaning the proportion of patients with epilepsy who require but are not receiving treatment, ranges from 10% in high income countries to as high as 75% in low income countries (Meyer et al., 2009). Although we found that psychologists and social workers are consulted at an increased rate in people with epilepsy, this study did not directly assess consultations with a psychiatrist. It may be that despite increased HRU in other areas, patients with epilepsy have less contact with a psychiatrist and perceived mental health care needs were not met for this reason. The more likely explanation is that due to the increased rate of psychiatric comorbidities in this group of patients, even the increased contact they have with mental health professionals is not perceived as being adequate. Future studies that address the content of mental health care, in addition to the use of care, are warranted in order to determine whether this is indeed the case so that this problem can be better addressed. Comorbid depression increases HRU in people with epilepsy, particularly when it is not being treated (Cramer et al., 2004).

This study identified individuals with unmet health care needs if they responded “yes” to the survey question “During the past 12 months, was there ever a time when you felt that you needed care but you did not receive it?” We cannot extrapolate whether the individual perceived unmet health care needs on only one occasion or whether it was a continuous or recurring problem. However, these unmet needs are clinically important, even if they only occur once a year. Depression in patients with epilepsy significantly affects quality of life (Gülpek et al., 2011), and untreated or unrecognized depression can lead to disastrous outcomes, such as suicide (Lonnqvist et al., 1995; Henriksson et al., 2001). Therefore, ensuring that mental health care needs are adequately met may help improve HRU in this population.

Sociodemographic factors as predictors of HRU and perceived unmet health care needs

A previous U.S.-based study found that variables such as age, marital status, educational level, and household income were not significantly associated with HRU (Bautista et al., 2008), although an effect of age, sex, and Aboriginal status has been reported in another Canadian study (Jetté et al., 2008). However, it must be noted that the methodologies differed, relying on surveys of patients with epilepsy at a major referral center (Bautista et al., 2008) versus administrative databases (Jetté et al., 2008). It is also possible that a mixing of effects is observed due to the heterogeneity of health care services studied here, whereby those with high education levels may be visiting specialists and those with low incomes may be visiting social workers.

Strengths and limitations

This study has many advantages over other studies that assess HRU in people with epilepsy. Because data were obtained from a national, population-based survey, they were not reliant on a tertiary care center and included people with epilepsy that do not necessarily see a neurologist. The high response rate of 85% for the survey is greater than that in many large scale surveys, such as the Behavioral Risk Factor Surveillance Studies (CDC, 2001; Kobau et al., 2008), and also helps ensure a more representative sample. In addition, results were compared to those with other chronic conditions, including migraine, and not just the general public, making it easier to differentiate which aspects of HRU are specific to people with epilepsy. Furthermore, the universal health care system in Canada ensures that most of the HRU studied was not affected by insurance and income.

There are also some limitations that must be addressed. First, use of a population survey limited us to the questions that were included in the CCHS, where only a few disease-specific questions were asked. For example, no distinctions were made between epilepsy subtypes. In a U.S.-based study of HRU utilizing insurance databases, grand mal status, epilepsia partialis continua, and infantile spasms, all had distinct patterns of HRU (Kurth et al., 2010). Seizure control and frequency were also not evaluated, which may be important because patients with higher seizure frequency (Bautista et al., 2008), refractory epilepsy (Sancho et al., 2008), or antiepileptic drug nonadherence (Faught et al., 2009) consume a disproportionately higher percentage of health resources. In addition, we cannot distinguish if HRU was specifically related to the care of epilepsy itself or to other medical problems. It has been previously shown that non–epilepsy-related HRU is the strongest contributor to HRU variation in patients with active epilepsy (Kurth et al., 2010). In addition, although there should be a great deal of external validity given the population-based nature of the study, the CCHS did not include people with epilepsy in institutions, on reserves, in prison, or in the armed forces. Children younger than the age of 12 years with epilepsy were also not represented in this survey, and the sample between the ages of 12 and 14 was very small. Finally, many patients with epilepsy, especially focal epilepsy, have memory impairment (Hoppe et al., 2007), and there is some question as to the validity of self-reporting (Elgar & Stewart, 2008). Self-reporting has been assessed in a large sample of patients with epilepsy and found to be of moderate to moderately high accuracy (>80%), although there are still concerns of underreporting (Corey et al., 2009). On the other hand, similar methods have been used in the Behavioral Risk Factor Surveillance Studies (CDC, 2001; Kobau et al., 2008).

In summary, the results of this large, national population-based survey confirm that patients with active epilepsy have increased HRU for many aspects of health care. Despite the increased visits with many types of health care professionals we have identified a decreased utilization of dental health resources, which is alarming considering that this population has poor oral health. Further studies are needed to address this issue to ensure that people with epilepsy are receiving the oral health care they need. Despite increased reports of visits with social workers and psychologists, people with epilepsy also perceived themselves as having unmet mental health care needs, and given the high degree of psychiatric comorbidities in this population future studies are warranted to address the quality of mental health care that this population is receiving.

Acknowledgments

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Although the research and analysis are based on data from Statistics Canada, the opinions expressed do not represent the views of Statistics Canada. This project was in part funded by Alberta Innovates Health Solutions; both N. Jetté and S. Patten hold salary support awards from this agency. N. Jetté is the holder of a Canada Research Chair Tier 2 in Neuroscience Population Health and Health Services Research. N. Jetté sits on the Editorial Board of Epilepsia. S. Wiebe is funded by the Hopewell Professorship of Clinical Neurosciences Research. A. Metcalfe has a doctoral research award in Genetics (Ethics, Law and Society) and a STIRRHS training grant in Maternal Fetal Newborn Health and Genetics, Child Development, and Health from CIHR.

Disclosure

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

None of the authors has any conflict of interest to disclose. We confirm that we have read the Journal’s position on issues involved in ethical publication and affirm that this report is consistent with those guidelines.

References

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Appendix

  1. Top of page
  2. Summary
  3. Background/Rationale
  4. Methods
  5. Results
  6. Discussion
  7. Acknowledgments
  8. Disclosure
  9. References
  10. Appendix

Appendix. Weighted cell counts for subgroups

 General populationAsthmaDiabetesEpilepsyMigraine
Sex     
 1 = Male197,06813,7569,6461,12911,535
 2 = Female202,98719,6888,5271,17028,858
 MissingNo observationsNo observationsNo observationsNo observationsNo observations
Age group     
 0 = 25–44141,94111,2172,24680817,609
 1 = 12–2482,7699,5653414327,638
 2 = 45–64118,1818,4307,79576412,502
 3 = ≥6557,1634,2327,7902952,643
 MissingNo observationsNo observationsNo observationsNo observationsNo observations
Highest level of education     
 1 = Didn’t graduate from high school104,49810,6616,7738559,677
 2 = High school graduate67,9505,0622,7754127,017
 3 = Some post-secondary32,3893,0591,0391813,520
 4 = Completed post-secondary187,29914,0737,00978019,499
 Missing7,91859057770680
Residence     
 1 = Urban326,49227,49814,5511,86133,112
 2 = Rural73,5635,9463,6214387,281
 MissingNo observationsNo observationsNo observationsNo observationsNo observations
Marital status     
 1 = Married/common law233,35316,81112,2831,10224,085
 2 = Widowed/separated/divorced46,8014,1714,1023134,878
 3 = Single119,51012,4421,77288311,403
 Missing3922115Insufficient cell size to be released26
Household income quintiles     
 1 = Lowest46,9104,8833,9415105,308
 2 = Low middle57,4054,8063,2703465,641
 3 = Middle71,0335,6992,7963337,213
 4 = Upper middle77,1836,0952,0753197,926
 5 = Highest80,4606,2802,9663777,929
 Missing67,0655,6813,1254146,376