Abstract
Migraine and tension-type headaches are the most frequent types of headaches in Santiago. The purpose of this paper is to describe the patterns of health care utilization in migraine and compare them with tension-type headache sufferers in a community-based study of the prevalence of headache in Santiago, Chile. A two-step questionnaire was administered to an age- and gender-representative sample of 1540 eligible subjects who were aged 15 years or older. Of these, 1385 (89.9%) responded. Cases were defined according to the most frequent type of headache experienced in the last year. Migraine and tension-type headaches were diagnosed according to a questionnaire-based algorithm constructed using the IHS criteria. Consultation rates in migraneurs were significantly (P < 0.01) higher (63%) than in tension-type headache suffers (39%). Migraneurs were more likely to be female, younger and less educated, and had more severe attacks than those with tension-type headache. Female sex, younger age, fewer years of education in migraneurs and younger age and moderate or severe headache in tension-type headache sufferers were found to be independently associated with the likelihood of ever consulting, respectively. Migraneurs are more likely to consult than tension-type headache suffers in this population.
Introduction
Migraine and tension-type headache represented together almost 92% of all IHS headaches reported in a representative sample of adults of Santiago Chile (1, 2).
The use of healthcare services by headache sufferers reflects the direct costs of disease and is a measure of the societal burden of disease (3). The examination of the factors that influence the patients' inclination to consult a general practitioner over a specialist or to use over-the-counter medication as opposed to specific drugs, may help in the organization of health care delivery systems (4).
Population-based cross-sectional samples provide unbiased information of a representative sample of subjects. All headache sufferers are captured regardless of pain diagnosis (migraine, tension-type headache), pain severity, demographics, access to health care, treatment status and beliefs. Few population-based studies have addressed these issues and fewer still in the developing world (5, 6).
The purpose of this paper is to describe the patterns of health care utilization in migraine and compare them with tension-type headache sufferers in a community-based study of the prevalence of headache in a representative sample of adults in the metropolitan area of Santiago, Chile.
Methods
A face-to-face questionnaire was administered to a sample of the adult population of Santiago. Households were selected as a probability sample of the adult, urban population of the province of Santiago with regard to gender and age. A two-step block design was used. Initially, 34 blocks, one in each of the 34 municipalities of Santiago, were selected at random. Subsequently, 500 households were randomly selected (15 from each block) and recruited by response to an initial questionnaire to the head of the household. Details of the survey have been previously reported (1, 2)
A two-step questionnaire was administered to an age- and gender-representative sample of 1540 eligible subjects who were aged 15 years or older. Of these, 1385 (89.9%) responded after at least three attempts to contact them. Initially the heads of household were interviewed regarding headache sufferers in the household and subsequently all headache sufferers identified were interviewed using a structured validated questionnaire by 20 trained lay interviewers. Before each interview the objective of the study was described and verbal informed consent was obtained. Demographic and socioeconomic information was obtained for each headache sufferer. Detailed questions were asked regarding the most frequent headache type. Detailed questions were asked about clinical characteristics of pain and impact on work and family activities. Specific questions regarding use of health services included if the subject had ever consulted, who had been consulted, list of medications used, effectiveness of medications, non-pharmacological ways of dealing with pain and beliefs about the causes of headache. Cases were defined according to the most frequent type of headache experienced in the last year. Migraine and tension-type headaches were diagnosed according to a questionnaire-based algorithm constructed using the IHS criteria (7).
Analyses
Univariate comparisons between migraine and tension-type headache sufferers were performed with the chi-square test for categorical variables (gender, age groups, socioeconomic groups, years of education, attacks per month and severity of attacks). Crude odds ratios were derived as the ratio of a characteristic in migraine sufferers vs. tension-type headache sufferers. A logistic regression model was used to investigate the association between consulting behaviour and several covariates (gender, age groups, socioeconomic groups, years of education, attacks per month and severity of attacks). Age and years of education were modelled as continuous variables. This was done independently for migraine sufferers and tension-type headache sufferers.
Results
Study variables were not distributed equally among migraine and tension-type headache sufferers. Migraneurs were more likely to be female, younger, less educated and have more severe attacks than tension-type headache sufferers (Table 1).
Sociodemographic features and pain characteristics in migraine and tension type headache sufferers in Santiago, 1993
The odds of ever consulting were significantly higher among migraneurs (Table 2). Of those who ever consulted, the proportion seeking care from general practitioners or specialists was similar between both headache diagnoses.
Consulting behaviour among migraine and tension-type headache sufferers
Could have consulted to more than one alternative.
Pharmacy, other health care professional.
A total of 31% (4/13) of men and 68% (60/88) of women with migraine ever consulted. Similarly 32% (37/116) of men and 42% (109/257) of women with tension-type headache ever consulted. The variables found to be associated with the odds of ever consulting in univariate analysis in migraneurs compared with tension-type headache sufferers were: female sex, age groups 30–39 and 50–59, medium socioeconomic group, fewer than 12 years of education, fewer than five attacks per month and moderate severity of pain (Table 3).
Univariate analysis of consultation rates in migraine and tension-type headache
Variables found to be independently associated with the odds of ever consulting in migraneurs were: female sex, younger age and fewer years of education (Table 4). Variables found to be independently associated with the odds of ever consulting among tension-type headache sufferers were: young age and moderate or severe headache (Table 5).
Variables independently associated with the odds of consultation in migraneurs
Variables independently associated with the odds of consultation in tension-type headache suffers
Discussion
This is the first paper to report on the consulting behaviour of a sample of the adult population in a Latin American country. Our data support previous findings that migraneurs are more likely to seek medical consultation than those with tension-type headache. Edmeads et al. in Canada found that 64% of migraneurs and 45% of those with tension-type headache had sought care from a physician at some time (5). Similarly, of those seeking medical care in our sample, 40% of migraneurs and 36% of those with tension-type headache were seen by a specialist. In Canada, Edmeads found that 41% and 32% of migraneurs and tension-type headache sufferers, respectively, were further referred to a specialist (5).
A significant gender difference in consultation rates was found in our sample in migraine sufferers; one-third of males compared with two-thirds of females had ever sought medical care. In fact female sex was found to be independently associated with the odds of consulting. This is similar to that reported by Stewart et al. in the American migraine study, where female migraine sufferers were 30% more likely to be diagnosed than their male counterparts (6). It has been stated that this could reflect gender-related patterns of physician consultation, which has also been found in Chile, though in this case it would apply only to migraine sufferers. This gender difference was much less significant among tension-type headache sufferers in our sample, where female-sex was not an independent covariate for consultation.
Younger age was found to be independently associated with the odds of consultation in both migraine and tension-type headache sufferers in our sample. This differs from the findings in the American migraine study (6) and could reflect several facts, for example patients seek care at the beginning of their headaches rather than later on, especially if they are severe because of the anxiety of having a brain disorder. Younger patients have less concomitant disease and thus headache is a major health problem. Younger patients are active both at work and socially and thus headaches could be more disabling.
Higher socioeconomic groups were more likely to consult in our sample, in both migraine and tension-type headache sufferers, but in neither headache type was this an independent predictor of consultation. These findings could be biased in our sample due to the under-representation of the higher socioeconomic groups, but could also reflect the fact that access to primary health and specialist care for the lower socioeconomic groups in the public sector, even though it is free, has more limitations than the private sector. Previous studies have found an increased probability of physician diagnosis of migraine among migraine sufferers with higher income (6). Surprisingly, in our sample individuals with fewer years of education were found to have higher consultation rates and this was an independent factor in migraine headache sufferers. This could reflect different beliefs about headaches, more severe and frequent headaches in the less educated or different patterns of access to health care, such as over-the-counter medication. In Chile the less educated are more likely to use the public health care system that is free and has open access to medical care.
Consultation rates were higher among those with moderate or severe pain intensity in both headache types and were an independent predictor of consultation in tension-type headache sufferers. This has been reported previously by several authors in population-based studies, especially in migraine headache sufferers (6, 8). Interestingly in our sample, higher consultation rates in both migraine and tension-type headaches were found among those with less frequent attacks. This was not an independent predictor of consultation but contradicts previous findings where the likelihood of physician consultation increases with the frequency of migraine attacks (8, 9). This could be explained by the fact that in nearly half of the cases in both headache types the information about headache frequency was not obtained.
Co-existing migraine and tension-type headache has been reported in previous population-based studies and could be a factor contributing to consultation (8). We did not investigate this variable as our cases were defined according to the most frequent headache type in the last year. Nevertheless, it is probable that this variable is not independent of headache severity.
Our data reflect the consulting behaviour of a sample of adults in Santiago, Chile, suffering from recurrent headaches (migraine and tension-type). The findings differ from previous reports in some aspects and reflect a pattern of health care utilization that is useful for health planners in the organization of care and education for subjects suffering from chronic headaches in Santiago, Chile.
Footnotes
Acknowledgements
This work was supported by a grant from Glaxo-Wellcome Farmaceutica Chile Ltda.