Abstract
General practitioners (GPs) diagnose and treat headache in primary care settings. The objective of this study was to investigate the effect of a 2-day headache education programme for GPs primarily on diagnostic accuracy. The education programme included theoretical lectures and face-to-face patient evaluation with headache specialists. Three GPs evaluated headache patients before and after the programme. Each GP was planned to interview a total of 60 patients (30 before, 30 after the programme). All patients were evaluated by headache specialists following evaluation by the GPs. A total of 189 patients were included in this study. Diagnostic accuracy increased from 56.3% to 81.0% after the headache education programme (P < 0.001), which also significantly improved the choice of proper treatment (P = 0.043). The headache education programme for GPs significantly improved diagnostic accuracy in patients with tension-type headache and the choice of proper treatment. Such education programmes can be standardized and given to GPs working in the primary care setting. These programmes can be arranged locally by the universities and might have a favourable impact on the diagnosis and treatment of headache.
Introduction
Headache is a very common complaint with a life-time prevalence of 95% for women and 90% for men. Migraine has a life-time prevalence of 12–18%, which is age- and gender-dependent, in community-based studies from different parts of the world (1). In Turkey, its life-time prevalence is 16.4% and that of tension-type headache (TTH) is 31.7% (2).
Despite its high frequency (1–4), severity (5) and the economic and social burden on society (3, 6–10), migraine is still an underdiagnosed and undertreated disorder (11–13). In a recent study in France, headaches of more than 40% of very severe migraine sufferers were not diagnosed as migraine (14).
General practitioners (GPs) diagnose and treat headache in primary care settings. Unfortunately, headache education in medical schools is not adequate. Sixty-one percent of medical schools have only a 1-h lecture on headache (15). Headache schools were planned to fill this gap in headache education. Since 2003, a 3-day headache school for GPs, neurology specialists and neurology residents has been held every year by the National Headache Study Group, a study subcommittee of the Turkish Neurological Society. After this experience, the first two authors developed a shorter course education programme for GPs working in primary care settings, including theoretical lectures and face-to-face evaluation of patients with headache specialists.
In this study, we aimed to investigate the impact of a 2-day postgraduate headache education programme on the diagnostic accuracy (particularly on the 1-digit code system) and on treatment and preferences of diagnostic tools in primary headaches.
Methods
Three GPs working in associated primary care settings of Bursa Municipality and the Department of Public Health, School of Medicine, Uludag University were asked to paticipate in this study and all accepted. The GPs were selected among those who had not taken any postgraduate education on headache and were told not to undertake further study of headache before or during the study. Two headache specialists (N.K. and M.Z.) are from the Department of Neurology of the same university, are members of the National Headache Study Group and have been in charge of the headache out-patient clinic for 8 years (N.K.) and 12 years (M.Z.).
The study was in three phases: Phase I, the pre-education phase; Phase II, the education phase; and Phase III, the post-education phase.
In Phase I, the study was explained to patients who had been admitted to the neurology out-patient clinic with headache and had agreed to participate in the study. Each GP interviewed (semistructured interview) a total of 30 patients. GPs were asked to diagnose the headache type according to the 3-digit code system of the International Headache Society (IHS) classification (16). If they had no knowledge of the IHS classification, diagnosis according to their own basic knowledge was accepted. They also recorded diagnostic tests and the proper treatment according to their diagnosis. After the evaluation of the GPs, the same patients were evaluated randomly by one of the two headache specialists. The 3-digit code system of the second edition of International Classification of Headache Disorders (ICHD-2) was used for the case definition criteria by headache specialists (16). Diagnostic tests were performed whenever needed. Treatment was classified into two groups: (i) acute-symptomatic, and (ii) preventive and acute-symptomatic. Medications were classified into basic medication groups: non-steroidal anti-inflammatory drugs (NSAIDs), triptans, β-blockers and antiepileptics. Headache specialists' diagnoses, diagnostic tests and treatment were accepted as the gold standard. The agreement between the GPs and the headache specialists regarding diagnosis, diagnostic tests and treatment choices were determined. Agreement for diagnosis was defined seperately on the basis of 1-, 2- and 3-digit code system of ICHD-2.
In Phase II, a 2-day educational programme was given to the GPs by the headache specialists. The theoretical educational programme of the headache school was modified and included face-to-face evaluation of the headache sufferers with the headache specialists (Table 1). On the first day, theoretical lectures were given. On the second day, GPs attended the headache out-patient clinic with the headache specialists. In this part of the progamme, both first-admission and follow-up patients were evaluated.
Details of the 2-day postgraduate headache education programme
The third phase of the study started 1 week after the education programme. GPs interviewed first-admission headache patients. Each GP interviewed patients once in every 3 days, in succession (from GP 1 to GP 3). All patients were evaluated randomly by one of the headache specialists. For migraine, more than four headache days per month was accepted as an indication for preventive treatment. There are no guidelines regarding preventive treatment for TTH, so patient expectations, comorbidity and the clinical impressions of the physician were considered together for indications of preventive treatment in TTH.
The χ2 test was used for agreement analysis between the GPs and headache specialists in both the pre- and post-education phases. For statistical analysis, SPSS for Windows 11.0 was used (SPSS Inc., Chicago, IL, USA). Results were explained to the GPs in privacy after the analysis.
Results
Two male and one female GP participated in this study. The male GPs had graduated in 1990 and 2003, the female GP in 2004, all three from the School of Medicine of Uludag University. The study was performed between November 2004 and April 2005.
A total of 189 (89 pre-education, 100 post-education) patients participated. Two patients from the pre-education phase group were excluded. There were 69 (78.4%) and 83 (83%) female patients in the pre-education and post-education phases, respectively. The mean age [±SD (range)] of the pre-education phase group was 35.7 ± 14.8 years (16–80) and of the post-education phase group 38.1 ± 13.6 years (17–76). There was no significant difference regarding age or gender between the pre- and post-education patient groups.
Data on agreement between GPs and headache specialists on the 1-digit code diagnosis for the pre- and post-education phases are given in Table 2. In the pre-education phase, there was no significant difference between GPs with respect to accurate diagnosis rates. All GPs improved their accurate diagnosis rates in the post-education phase. However, the agreement rate of GP 2 was significantly lower than that of the others (P = 0.006). Statistical analysis showed a significant effect of the headache education programme on 1-digit headache diagnosis (P < 0.001). GPs made a 2-digit code diagnosis in 55.2% of patients in the pre-education phase and in 78% in the post-education phase. Although insignificant (P = 0.052), the increase in agreement rate (from 4.2% to 15.4%) in the 2-digit code diagnosis was nearly fourfold.
The agreement rates between general practitioner and headache specialists for 1- and 2-digit code diagnosis for pre- and post-education phases∗
Data are given as n (%) (total n).
The effect of the education programme was also analysed on the basis of individual headache types, i.e. migraine and TTH. Results for other primary headache types are not given due to small patient numbers. In the pre-education phase, the agreement rate between the headache specialists and GPs was 75% for migraine diagnosis. In the post-education phase, this number increased to 76% (P = 0.34). A total of 86 migraine cases were diagnosed by headache specialists in both phases of the study. The total agreement rate was 72.1% (64/86). For TTH, the rate of accurate diagnosis was 40% in the pre-education phase and 64% in the post-education phase (P = 0.014).
Treatment agreement between the headache specialists and GPs was analysed only if the GPs had made the correct diagnosis. Before education, GPs treated 48 out of 49 correctly diagnosed patients. Analysis revealed a 64.6% treatment agreement rate with headache specialists (31 out of 48 patients). After the programme, the treatment agreement rate increased to 80.8%. These rates showed that the headache education programme significantly improved the choice of proper treatment (P = 0.043). Treatment preferences of the GPs were also changed. Preventive treatment was chosen in 68.6% of the patients before education and in 73.0% after education. Acute treatment was given to 30.2% of patients in the pre-education phase and to 22.0% in the post-education phase. Although insignificant, the reduction in the percentage of acute treatment was >25%.
In the pre-education phase, GPs made incorrect diagnoses in 38 (43.7%) patients. Of these, migraine was diagnosed in 16 (42.1%) patients. Specialist interview revealed that 14 (87.5%) of 16 misdiagnosed migraine patients were TTH sufferers. GPs misdiagnosed seven (18.4%) patients' headaches as TTH. In four (57.1%) of these patients, migraine without aura was the accurate headache diagnosis. In the post-education phase, 19 (19%) headache sufferers were misdiagnosed. Similar misdiagnosis rates were recorded for both migraine (n = 3, 15.8%) and TTH (n = 3, 15.8%). Headache specialists diagnosed two TTH and two migraine cases in both groups. The ratio of migraine among misdiagnosed headache cases dropped from 42.1% to 15.8% after the education programme (P = 0.047).
There was a slight, insignificant improvement in the diagnostic test agreement rate between the headache specialists and GPs after the education.
Discussion
The headache education programme was based on the experiences of the headache school and European Headache Federation (EHF) guidelines for the organization of an educational programme on headache (17).
Our programme contained practice with the headache specialists. EHF guidelines offer only lessons in theory (17). We believe that evaluating headache patients with headache specialists is more effective than theoretical lectures or video case reports. We therefore added face-to-face patient evaluation to the programme and decreased theoretical lesson hours.
Our results showed that the 2-day headache education programme significantly increased the overall diagnostic accuracy of GPs. The improvement rate was significant (P < 0.001). The increase in agreement rate in 2-digit diagnosis was nearly fourfold but insignificant. Low agreement rates in 2-digit diagnosis showed that more time should have been spent on subtypes of primary headaches during the education programme. Emphasizing the differences between subtypes of primary headaches would help to solve this problem. GPs made no 3-digit diagnoses. The 3-digit diagnosis is used in headache studies, but it is very difficult for GPs to make a diagnosis based on this system. Therefore, this result was expected. All GPs improved their accuracy in diagnosis, although significantly so in only two out of three GPs. Although insignificantly, GP 2 also showed improvement in accuracy of diagnosis.
The diagnostic agreement rate between the headache specialists and GPs in TTH increased significantly. However, there was no increase in this rate for migraine. The improvements in the treatment of acute migraine attacks, particularly those related to triptans, have caused drug companies to increase their visits to doctors and to make education programmes about migraine. However, TTH does not receive the attention it deserves, either from drug companies or researchers. The efforts of the drug companies have increased the awareness and knowledge of migraine. This might be the reason for the significant effect of the education programme in accuracy of diagnosis of TTH, but not migraine, in our study. Our pre-education diagnosis rate for migraine was better than that found by De Diego and Lanteri-Minet (14). This finding may have limited the benefit of the education programme in terms of the diagnosis of migraine. Overdiagnosis of migraine in the pre-education phase might be another reason for the ineffectiveness of the education programme in the diagnosis of migraine. Approximately 42% of the misdiagnosed cases were diagnosed as migraine. Of those, 87.5% were TTH cases, showing that GPs had diagnosed TTH as migraine. In the post-education phase, the migraine rate dropped significantly to 15.8% among misdiagnosed headaches. This significant decrease in the migraine diagnosis rate among misdiagnosed headaches was another favourable effect of the education.
The treatment agreement rate between the headache specialists and GPs was analysed and it was shown that education had significantly improved the choice of proper treatment for patients whose conditions were diagnosed correctly (P = 0.043). The change in the diagnostic test agreement rate between the headache specialists and GPs was insignificant. The emphasis on the diagnosticalgorithm of the headache might have been inadequate in the education programme. On the other hand, lack of self-confidence about the neurological examination may have forced GPs to request unnecessary imaging tests.
This study has some limitations. Both GP and patient numbers were small. Larger groups might have better demonstrated the efficacy of the programme. Large patient numbers would also have increased the number of other primary headache types and helped to evaluate the effect of education on other headache types. The education time for some headache subtypes was not adeaquate. Future education studies should include more GPs and patients. Awareness of headache can also be evaluated after education in workplace settings.
In our opinion, headache education programmes can be standardized and can be given to GPs working in the primary care setting. Practice with headache specialists adds considerably to the education. Therefore, we believe that at least 1 day should be dedicated to practice with headache specialists in such programmes. Education groups may contain 10–20 GPs. Local programmes for small groups also have the advantage of closer relations between those attending and the faculty both during and after the education programme. Therefore, we suggest that local universities or tertiary care facilities should arrange local headache education programmes. For standardization and global programming, the guidelines of the IHS and EHF can be used or modified (17).