those with no chronic health problems (the healthy group).
Z tests assessed statistical significance of association
in selected variables. Multiple univariate analyses of the
epileptic population were performed using the Rao Scott
correction for analysis of complex survey design.16
Epilepsy Population (Table
questions were answered by 60,823 subjects (99.3% of the
total sample) of whom 313 described themselves as having
epilepsy. This yielded an adjusted point prevalence rate
of self-reported epilepsy of 5.8 per 1,000, equivalent to
47,000 prevalent cases in Ontario (1990). Prevalence was
highest among those aged 65 years or older, ranging from
3.1/1,000 to 7.2/1,000 in various age groups (Table
2). The age distribution of the epilepsy population
largely reflected that of the general population, with slight
under- and over-representation of the 0-11 and 65 year or
older age groups, respectively. Most self-described epileptic
subjects (88.9%) resided in urban dwellings (Appendix) and
60% lived in single detached homes with two other people
on average. Their mean age was 39 years and gender was equally
represented. Over half of the subjects had no spouse or
common law. An occupation was acknowledged by 131 (42%)
individuals of whom 36% were in managerial, administrative
or professional activities. Epilepsy was the only recognized
chronic health problem for 27% of epileptic subjects. Twenty
percent each had one and two additional health problems,
two gender-related variables exhibited statistically significant
differences. Females had a higher number of health problems
(p=.025) and of hospital admissions within the previous
year (p=.02). Fewer males had low income (Appendix) and
fewer females had primary occupations, eg., farming, factory,
construction, etc., but these were not statistically significant.
geographical regions were identified, ie., north east (NE),
north west (NW), central east (CE), central west (CW), east
(E), and south west (SW). Some significant between-region
differences in the unadjusted analysis lost significance
after correcting for design effect. For example, apparently
significant higher household income and visits to doctors
in the CE, CW and SW (the most densely populated areas),
are not significant after correcting for design effect.
Three variables retain statistical significance:
room use is highest in NE (62%) and NW (51%) and lowest
in E (18%) Ontario as compared with the provincial average
services are used most often in CE (65%) and CW (68%)
than in NW (39%) and NE (44%) regions.
barriers to health care were highest in the CE (36%) and
NW (30%) and lowest in the E (1%) and SW (3%).
significant differences emerged between epilepsy subjects
living in urban and rural areas. The former were more frequent
users of psychological/social work services (16% vs. 4%),
experienced more barriers to health care (25% vs. 7%) and
had poorer QOL (Appendix) scores (all p=.02).
(better) QOL scores were obtained more frequently by epileptic
persons of households with healthy family scores (p=.027).
In addition, higher educational levels were associated with
healthier families (p=.04).
occupation and disability (Table
more epileptic persons (22.1%) had a low annual income (Appendix),
than those with other chronic problems (14.6%), the general
population (13.6%) or healthy subjects (12%) (p <0.001,
all comparisons). The converse was observed for high annual
epileptic subjects (9.5%) were in the service or transport
industry, as compared with the chronically ill or the general
population (16% each).
mean number of days (±95%CI) each person had to stay
in bed or cut down on usual activities during the previous
14 days because of ill health was significantly higher for
epilepsy subjects (1.57 ± 0.21) than for chronically
ill (0.99 ± 0.04), the general population (0.65 ±
0.04), and healthy people (0.02 ± 0.01). The mean annual
number of disability days (±95%CI) in these four groups
is 41 ± 5 , 26 ± 1, 17 ± 1, and 0.5 ±
0.25, respectively. Thus, in Ontario, epilepsy causes from
1.7 million to 2.2 million annual disability days (mean
1.93 million days).
survey probed for traffic and all other accidents in the
previous 12 months that resulted in an injury serious enough
to limit normal activities. The mean number of accidents
was slightly higher in epilepsy subjects than in the general
population but lower than in other chronically ill individuals.
of Life, Family Function, and Social Participation (Table
average, self-reported epileptic subjects' QOL, as measured
by the PGWB scale, was significantly lower (worse) than
any of the three compared groups, including the chronically
ill (p<.0001). More importantly, the proportion of epilepsy
subjects obtaining the lowest (worst) QOL score (<24)
was twice that of the compared groups. Conversely, the proportion
of epileptic persons with high (good) QOL scores (>31)
was almost half that of the group with other chronic illnesses.
OHS defined dysfunctional families as those with MFAD scores
> 2.17 (Appendix).15 These were significantly
more frequent among epileptic persons than among other chronically
ill or the general population (p<.001).
participation indices were significantly lower for epileptic
subjects than for other compared groups (p<.001).
to Health Care (Table
barrier to health care for their chronic illness was reported
by 22.75% and 7.45% of epilepsy and chronically ill respondents,
respectively. Overall, specific barriers were minimal and
infrequent. None exceeded 1.5% of epilepsy subjects, eg.,
1.3% did not know an appropriate physician or where to go
and the same proportion experienced difficulty getting an
and Limitations (Table
the surface, epileptic subjects' academic achievement in
Ontario was similar to that of other groups, ie., 28% obtained
only some or complete primary education, 41% initiated secondary
instruction, and 31% engaged in post-secondary education.
However, epileptic subjects were less likely than the general
population to complete secondary and post-secondary education
(odds ratio = 0.8 at each level).
asked whether ill health limited the amount or type of activities
they could do as compared to healthy people of the same
age and sex, one third of epilepsy subjects answered affirmatively.
The corresponding proportions for the chronically ill and
the general population were 11% and 7%, respectively. In
addition, ill health limited work or school in 72% of epilepsy
respondents, as compared to 55% of those with other illnesses
(p<.001). Finally, significantly fewer epileptic subjects
participated in regular physical activities, as compared
with the chronically ill population (p<.001).
of Health Care Resources (Table
was obtained on utilization of health services in the previous
12 months. On average, epileptic persons visited some health
professional 19.3 times, and 20% had >25 professional
visits in one year. Epileptic persons visited their family
physician or specialist an average of 9.1 times and 18%
visited them >13 times. They also used family physician
and specialist services more frequently than control groups.
However, incremental use was larger for specialists than
for family doctors, ie., the ratio of specialist to family
doctor for epileptic subjects, chronically ill and the general
population was 0.59, 0.43 and 0.38, respectively. In addition,
epilepsy respondents sought nursing services twice as frequently
as the general population and other chronically ill. Thirty-five
percent of epilepsy subjects had visited the emergency room
(mean number of times = 2.5, SE .27) and 22% had been admitted
to hospital (mean number of times = 1.6, SE .84), as compared
with 27% and 14% of those with chronic illnesses and 22%
and 12% of the general population, respectively. Fewer epileptic
subjects than other groups visited a dentist. On the other
hand, optometry, pharmacy, and physiotherapy services were
used with similar frequency by all groups. Epilepsy subjects
utilized psychology and/or social work services three times
as frequently as the chronically ill or the general population.
asked about the type of health services received in the
previous 14 days, use of emergency room and telephone advice
was disproportionately high among epilepsy subjects.
finding methods have an impact on estimates of epilepsy
frequency. While it is accepted that under-ascertainment
is pervasive and that no single method will identify all
cases in a population, some methods are more precise than
others. In general, specialist practice-based estimates
may miss up to 80% of cases unless this is the only source
of readily accessible health care.17,18 Using
all available medical records may miss from 7% to 27% of
incident and prevalent cases found by door to door surveys.19
On the other hand, Beran et al. found their population survey
to underestimate previously documented prevalent cases by
23%20 and suggested 20/1,000 as a more accurate
(corrected) prevalence ratio of epilepsy in Australia.21
This remarkably high figure stands out from international
prevalence ratios. The Commission on Epidemiology and Prognosis
(CEP) of the International League against Epilepsy22
does not recommend a specific ascertainment method.
strictness of case definition would affect prevalence estimates
of active epilepsy; narrow criteria (eg., having a seizure
on the day of survey) yielding lower estimates than broader
criteria (eg., asking do you have epilepsy?). However, its
practical effect on prevalence figures seems minimal in
studies using stricter and looser criteria.23,24
Most studies qualify the definition of epilepsy in some
way, eg., witnessed seizure, afebrile recurrent seizures,
number/recentness of events, use of antiepileptic drugs,
etc. The OHS used no qualifications. Thus, OHS data should
be interpreted as reflecting individuals who perceive themselves
as having epilepsy.
epilepsy self-identification in the OHS reflects the health
of true epileptic patients in the general population is
supported by the congruence of our findings with those in
the literature. For example, Austin et al. found worse QOL
scores in epileptic than asthmatic children25
and Hermann et al. uncovered worse emotional well-being
in epileptic than in hypertensive subjects.26
Similarly, levels of physical fitness have been worse in
epileptic persons than in controls in Steinhoff et al.'s
study.27 The OHS is likely to capture individuals
with active epilepsy, eg., having seizures and/or taking
medication. Accordingly, the observed frequency (5.8/1,000)
may be best interpreted as point prevalence of active epilepsy.
This figure is similar to that of active epilepsy (6/1,000)
found in studies with various case definitions and ascertainment
analysis provides valid health-related information about
people in the general population who consider themselves
as suffering from various illnesses. It also allows for
comparison among different self-perceived conditions, and
for approximation of the burden of illness in the general
population. For example, OHS data have been used to perform
comparisons between Canada and the USA with regard to use
of preventive health care28 and to assess the
impact of socioeconomic/health status on use of physicians'
services.29 It has also provided important information
on the burden of various chronic conditions, including migraine,
cardiovascular and musculoskeletal diseases.30-33
data indicate that, in the non-institutionalized Canadian
population, epilepsy is a marker of poor health, inferior
quality of life, lower income, poorer family function and
social support and higher health care resource utilization.
In most aspects, the burden of illness is greater in self-reported
epileptic persons than in those with one or more common
chronic health problems other than epilepsy. The prevalence
of epilepsy is much higher in the institutionalized (30%),34
mentally retarded (17% to 60%),35 and imprisoned
population (2.5%).36 Failure to include these
segments of the population in the OHS likely underestimates
the frequency and severity of epilepsy in Ontario.
groups under comparison consistently showed a better-to-worse
health status gradient, ie., healthy individuals, the general
population, the chronically ill, and the epilepsy population,
in that order. Indices of health care resource utilization
exhibited a similar gradation. Epilepsy subjects had the
highest level of hospital, emergency, medical and psychosocial
differences among provincial regions merit comment. Perceived
barriers to health care were higher in urban dwellers and
in CE and CW regions (those with higher population densities).
In addition to relatively easier access in less densely
populated areas, this may also reflect epileptic subjects'
satisfaction with access to the available level of care
in rural areas. Overall, the 1990 epilepsy population was
fairly satisfied with access to health care. Higher use
of psychological services by urban dwellers may be due to
easier access to such services in larger communities or
to true differences between rural and urban epileptic subjects.
For example, patients requiring higher levels of support
may migrate to larger communities. The latter concept may
be supported by poorer levels of well-being in urban dwellers.
Finally, the reasons for higher emergency room use in the
Northern regions remain unclear. More severe epilepsy is
an unlikely cause, as physical health, hospital use and
QOL in the region were equal to or better than those in
other areas. Patient education, and administration/allocation
of health services remain plausible explanations.
educational status of Canadian people with epilepsy appears
similar to that of other countries, eg., approximately 60%
in Sillanpaa's Finnish study had basic education.37
However, our data indicate that epileptic persons engaging
in secondary or post-secondary education are less likely
than the general population to complete it successfully.
Thus, their academic progress is slower or truncated.
have others, we found family dysfunction, a known contributor
to psychiatric, emotional and behavioural problems, to be
worse in those with epilepsy than in controls.38
This analysis supports the notion that subjective well-being
correlates importantly with physical health.39
Both were poorest in the epilepsy group than in all other
groups, including the chronically ill.
may be argued that comparing epilepsy with a broad range
of chronic illnesses may be less meaningful than comparisons
with neurological illnesses affecting similar age and gender
groups. Comparison with published OHS data for migraine
sufferers, a group demographically similar to epilepsy,
is telling. Migraneurs' education was superior (primary
or lower in only 12.8%, post secondary in 38%), their income
was higher (only 11% had low income) and their hospitalization
rate was lower (14%). This confirms self-reported epilepsy
as a marker of poorer health, lower education and income,
and higher per patient health care resource use.
of productivity loss due to disability days in the epilepsy
population can be approximated by considering 250 working
days per year, an average Ontarian annual income of CAN
$25,902 (Revenue Canada, 1991) and an average epileptic
person's two-week disability of 1.6 days. Correcting for
the fraction of working days in two weeks (10/14), the number
of yearly disability days is 28.6 per individual and 1,342,857
provincially. The corresponding yearly productivity loss
is $2,962.3 per individual and $139.23 million for the provincial
epileptic population. Studies that include unemployment
arrive at much higher cost estimates of productivity loss.3
Imputing average earnings to disability days as an approximation
of the opportunity cost of time lost due to ill health (human
capital approach) has been successfully used in previous
analyses of epilepsy and other neurological illnesses.3,31,40
However, some caveats apply to this method, eg., it is assumed,
among other things, that all disability occurs in productive
individuals.3,41 Similarly, we do not account
for excess epilepsy-related mortality, estimated at 0.38%
to 0.56% in some studies.3,42 Its inclusion would
result in higher estimates of productivity loss. Moreover,
surveys may underestimate outpatient health care utilization
by as much as 20%.43
of their scope, method and purpose, omnibus population surveys
often lack the disease-specific depth desired by clinicians.
Thus, the OHS contains limited diagnostic and therapeutic
information about any given condition, including epilepsy.
Nonetheless, the current population-based analysis of the
epilepsy population in Canada is useful in several ways.
First, it demonstrates the physical and psychosocial ill
health of the epilepsy population. Second, it begs important
questions about these patients' health related needs and
the optimum way of dealing with them. For example, efficiency
may be gained by devoting more resources to structured counselling
or liaison services.44 Exploring variables that
determine small area variations in care may be beneficial,
eg., educational efforts targeting emergency room and hospital
admissions. Canadian epilepsy patients use telephone contact
with health professionals extensively. Thus, readily accessible
telephone services offering judicious epilepsy advice and
triaging may help provide better and more efficient care.
Finally, the observed pattern of ill health and resource
use may assist health professionals and administrators in
formulating future epilepsy-specific surveys for the Canadian
Function: General Functioning Subscale of the McMaster
Family Assessment Device (MFAD), a 12-item, reliable and
valid measure of family function.14 Higher scores
indicate worse family function.
of Life: Dupuy's Psychological General Well-Being scale
(PGWB)13 is a 22-item, 6-subscale, widely used,
valid and reliable instrument. Using 6-point Likert scales,
individuals rate their level of energy, control of emotions,
state of morale, interest in life, perceived stress and
health status, and satisfaction with relationships during
the previous 12 months. Higher scores indicate better quality
Support Index: Designed specifically for the OHS, this
instrument explores the number of friends/relatives the
respondent felt close to, proportion of leisure time spent
with others, satisfaction with social life, availability
of a confidant and participation in voluntary organizations.
Separate versions for individuals younger or older than
60 years were used. High indices indicate more participation
in social support systems.15
annual household income: Income from all sources <$12,000
regardless of family size, $12,000 to $19,999 if household
size two or more, or $20,000 to $29,999 if household size
four or more.
area: Census agglomeration or metropolitan areas with
populations > 10,000.
area: Dwellings not included in urban area.
health problems: Operationally defined as long-term,
permanent or recurring physical health problems. Nineteen
specific illnesses were explored by interview as follows:
1) skin allergies and other skin diseases, 2) hay fever
or other allergies, 3) serious trouble with back pain, 4)
arthritis or rheumatism, 5) other serious joint/bone problems,
6) paralysis or speech problems due to stroke, 7) asthma,
8) emphysema or chronic bronchitis or persistent cough,
9) epilepsy, 10) high blood pressure or hypertension, 11)
circulatory problems, 12) heart disease, 13) diabetes, 14)
urinary problems or kidney disease, 15) stomach ulcer, 16)
other digestive problems, 17) goitre or thyroid trouble,
18) eye problems, for example glaucoma, cataract, 19) cancer
Zena Pellegatta assisted with manuscript preparation. Dr.
Warren Blume provided insightful comments on an earlier
CJ, Lopez AD, Jamison DT. The global burden of disease
in 1990: summary results, sensitivity analysis and
future directions. Bull World Health Organ 1994; 72:
WA, Hesdorffer DC. Incidence and Prevalence. In: Hauser
WA Hesdorffer DC. eds. Epilepsy: frequency, causes
and consequences. New York: Demos, 1990: 1-51.
OC, Hart YM, Sander JWAS, et al. The cost of epilepsy
in the United Kingdom: An estimation based on the
results of two population-based studies. Epilepsy
Res 1994; 18: 249-260.
GA, Jacoby A, Buck D, et al. Quality of life of people
with epilepsy: a European study. Epilepsia 1997; 38:
JM, Shapiro E. Self-rated health: a predictor of mortality
among the elderly. Am J Public Health 1982; 72: 800-808.
RW, Manga P, Binder DA, et al. The use of physician
services under a national health insurance scheme.
An examination of the Canada Health Survey. Med Care
1983; 21: 1037-1054.
EM, Goel V. Factors related to emergency department
use: results from the Ontario Health Survey 1990.
Ann Emerg Med 1994; 24: 1083-1091.
Health Survey 1990. Highlights. 1992; (Abstract)
Canada. Annual demographic statistics. 1997; 91-213-XPB:(Abstract)
Health Survey 1990. User's Guide. Documentation. 1993;
Health of Canadians, Report of the Canada Health Survey.
M. Sources and Rationalization of Questions Used in
the Quebec Health Survey. Technical Manual 87-03.
HJ. The psychological general well-being (PGWB) index.
In: Wenger NK, Mattson ME, Furberg CD, et al. Assessment
of Quality of Life in Clinical Trials of Cardiovascular
Therapy. New York: Le Jacq, 1984: 170-183.
J, Byrne C, Boyle MH, et al. Ontario Child Health
Study: reliability and validity of the general functioning
subscale of the McMaster Family Assessment Device.
Fam Process 1988; 27: 97-104.
Health Survey 1990. User's Guide. Microdata Manual.
1993; Vol. 2:(Abstract)
JN, Scott AJ. A simple method for the analysis of
clustered binary data. Biometrics 1992; 48: 577-585.
Graaf AS. Epidemiological aspects of epilepsy in Norway.
Epilepsia 1974; 15: 291-299.
HM. Transient ischemic attacks, Parkinson's disease
and epilepsy: results from the 1980 National Ambulatory
Medical Care Survey. Public Health 1986; 100: 368-374.
AF, Anderson DW, Schoenberg BS, et al. Prevalence
and clinical features of epilepsy in a biracial United
States population. Epilepsia 1986; 27: 66-75.
RG, Michelazzi J, Hall L, et al. False negative response
rate in epidemiologic studies to define prevalence
ratios of epilepsy. Neuroepidemiology 1985; 4: 82-85.
RG, Hall L, et al. Michelazzi J. An accurate assessment
of the prevalence ratio of epilepsy adequately adjusted
by influencing factors. Neuroepidemiology 1985; 4:
on Epidemiology and Prognosis: International League
Against Epilepsy. Guidelines for Epidemiologic Studies
on Epilepsy. Epilepsia 1993; 34: 592-596.
G. Epilepsy in Iceland. A clinical and epidemiological
investigation. Acta Neurol Scand 1966; 43:Suppl 25:1-124.
JM, Brody JA, et al. Brink E. Convulsions among the
Chamorro people of Guam, Mariana islands. I. Seizure
disorders. Am J Epidemiol 1972; 95: 292-298.
JK, Smith MS, Risinger MW, et al. Childhood epilepsy
and asthma: comparison of quality of life. Epilepsia
1994; 35: 608-615.
BP, Vickrey B, Hays RD, et al. A comparison of health-related
quality of life in patients with epilepsy, diabetes
and multiple sclerosis. Epilepsy Res 1996; 25: 113-118.
BJ, Neususs K, Thegeder H, et al. Leisure time activity
and physical fitness in patients with epilepsy. Epilepsia
1996; 37: 1221-1227.
SJ, Hofer TP. Socioeconomic disparities in preventive
care persist despite universal coverage. Breast and
cervical cancer screening in Ontario and the United
States. JAMA 1994; 272: 530-534.
SJ, Hofer TP, et al. Manning WG. Physician use in
Ontario and the United States: the impact of socioeconomic
status and health status. Am J Public Health 1996;
EM, Rasooly I, et al. Webster GK. Relative importance
of musculoskeletal disorders as a cause of chronic
health problems, disability, and health care utilization:
findings from the 1990 Ontario Health Survey. J Rheumatol
1994; 21: 505-514.
T, Wu K. Health care utilization and disability of
migraine: the Ontario Health Survey. Can J Public
Health 1995; 86: 195-199.
T, Pomerleau J, Speechley M, et al. Correlates of
body mass index in the 1990 Ontario Health Survey.
Can Med Assoc J 1995; 152: 1811-1817.
Gardner R, Steven D. An analysis of the Ontario Health
Survey from a cardiovascular perspective. Can J Cardiovasc
Nurs 1994; 5: 7-14.
E, Ferini-Strambi L, Sala M, et al. Epilepsy in institutionalized
patients with encephalopathy: clinical aspects and
nosological considerations. Am J Ment Retard 1993;
C, Hosking G. Epilepsy in school children with intellectual
impairments in Sheffield: the size and nature of the
problem and the implications for service provision.
J Ment Defic Res 1989; 33: 511-514.
S, Coleman TE, Patmon C, et al. Epilepsy in prison:
elevated prevalence and no relationship to violence.
Neurology 1984; 34: 775-782.
M. Social functioning and seizure status of young
adults with onset of epilepsy in childhood. An epidemiological
20-year follow-up study. Acta Neurol Scand 1983; 96:
PJ, Upton D. Quality of life in family members of
persons with epilepsy. In: Trimble MR Dodson WE. eds.
Epilepsy and Quality of Life. New York: Raven Press
Ltd. 1994: 19-31.
CE, Hayes D. Exercise and psychologic well-being in
the community. Am J Epidemiol 1988; 127: 762-771.
CE, Annegers JF, Lairson DR, et al. Cost of epilepsy
in the United States: a model based on incidence and
prognosis. Epilepsia 1994; 35: 1230-1243.
R. Costs and cost-minimisation analysis. Br Med J
1993; 307: 726-728.
P, Cole TB, Annegers JF, et al. Cohort study of incidence
of sudden unexplained death in persons with seizure
disorder treated with antiepileptic drugs in Saskatchewan,
Canada. Epilepsia 1995; 36: 29-36.
JB, White AA, Kelley CL, et al. Recall strategies
and memory for health-care visits. Milbank Q 1990;
MP, Readman S, Hague B, et al. A district epilepsy
service, with community-based specialist liaison nurses
and guidelines for shared care. Seizure 1994; 3: 121-127.
J. Neurol. Sci. 1999; 26: 263-270