Inappropriate use of sumatriptan: population based register and interview studyBMJ 1998; 316 doi: http://dx.doi.org/10.1136/bmj.316.7141.1352 (Published 02 May 1998) Cite this as: BMJ 1998;316:1352
- David Gaist, research assistant ()a,
- Ioannis Tsiropoulos, senior registrarb,
- Søren H Sindrup, senior registrara,
- Jesper Hallas, senior registrara,
- Birthe K Rasmussen, senior registrarc,
- Jakob Kragstrup, professord,
- Lars F Gram, professora
- a Department of Clinical Pharmacology, Institute of Medical Biology, Odense University, DK-5000 Odense C, Denmark
- b Department of Neurology, Odense University Hospital, Odense, Denmark
- c Department of Neurology, Hillerød Hospital, Hillerød, Denmark
- d General Practice Research Unit, Odense University, Odense
- Correspondence to: Dr Gaist
- Accepted 31 October 1997
Sumatriptan ranked second in expenditure on drugs for outpatients in Denmark in 1995. The 5% of patients who were heavy users of sumatriptan accounted for nearly 40% of consumption. 1 2 We conducted a population based interview study to evaluate the appropriateness of sumatriptan use.
Subjects, methods, and results
Subjects were recruited through community pharmacies in Funen county, Denmark (population 465 000). Patients who presented prescriptions during two weeks in February 1996 were invited to participate, and relevant data for 1992-6 were retrieved from the county prescription registry.3 Sumatriptan consumption was described as the defined daily dose unit (100 mg for oral sumatriptan and 6 mg for subcutaneous sumatriptan). For each subject, peak dispensing of sumatriptan in any 30 day period was determined from register data. Patients were then classified into three groups: high peak users (≥60 units/30 days), intermediate peak users (30-59 units), and low peak users (<30 units).
After anonymising non-respondents' data, we used register data to evaluate the representativeness of the study population. Participants underwent structured interview by a doctor and were examined by neurologists. Recall was assisted by photographs of drugs and a graph of the patient's monthly sumatriptan use based on register data. Participants completed headache diaries for 30 days.4 Patients' sumatriptan use was evaluated according to criteria defined in the table. The study was approved by the regional ethics committee and the Danish Board of Registers, and patients gave written informed consent.
Of 435 patients eligible for inclusion (83% women, median age 47), 233 (54%) responded. Response rates were 33% (7/21) in the high use group, 47% (30/64) in the intermediate group, and 56% (196/350) in the low use group. Respondents and non-respondents in the low and intermediate groups had comparable age, sex, and drug use. Non-respondents (14 women) in the high use group had consumed more sumatriptan than respondents (4 women, 3 men) (median 1333 v 832 units; P=0.04). All 37 respondents with high or intermediate peak use completed the interview and medical examination; 30 returned completed headache diaries. Of 30 randomly selected patients in the low use group, 29 completed all study phases.
Patients with peak use ≥30 units/30 days reported previous dependence (need to take the drug every day to function normally) on drugs other than sumatriptan more frequently (table); three reported dependence on sumatriptan. Diagnosis of headaches that occurred before chronic use of strong medication for relief of headache was according to the criteria of the International Headache Society.5 We were unable to establish a diagnosis of migraine or cluster headache in nine patients (two high use, three intermediate use, four low use). Headache recurred in 12 (18%) patients with migraine (none from the high use group) within 24 hours in 50% or more of treated episodes; they usually repeated the sumatriptan. Six of seven high use subjects and 22/30 with intermediate use fulfilled one or several of the criteria for inappropriate use of sumatriptan. Chronic use (daily or near daily use for ≥3 consecutive months) of analgesics to relieve headache was common in these patients. They also had drug induced headache frequently (headache for ≥180 days/year and concurrent chronic use of any headache medication other than sumatriptan). Inappropriate use was related to frequent use (≥24 times in past 12 months) in 4/29 low peak users for tension headaches and in another four for drug induced headache.
The relatively low response rate, particularly in high peak users, raises concern about the representativeness of this study. When the higher total consumption of sumatriptan among non-respondents in this group is taken into consideration, this bias could lead to underestimation of sumatriptan overuse. Appropriate heavy use of sumatriptan for cluster headache was rare. We conclude that heavy consumption of sumatriptan generally represents inappropriate use, mainly for tension and drug induced headaches. Inappropriate use may be related to the patient rather than the drug. Patients at greatest risk have generally been excluded from clinical trials conducted before the drug was marketed. Greater awareness of the problem among doctors could lead to more rational use of sumatriptan.
We thank Bente Overgaard Larsen, Lars Clemmensen, and all staff at participating pharmacies for excellent teamwork, and Anne Rosenkrantz for secretarial assistance.
Contributors: DG had the original idea for the study, was principal investigator and study coordinator, interviewed all patients, did statistical analyses, and had main responsibility for writing the article; he is guarantor of this paper. IT and SHS advised on design of the structured interview and criteria for inappropriate use, examined patients, and participated in evaluation of drug use. JH and JK advised on the recruitment part of the study. JH was also responsible for retrieval of prescription data. BKR commented on the part of the structured interview on headaches and the criteria for inappropriate use and was consulted regarding difficult cases. LFG was responsible for funding and advised on the overall design and intrepretation of data. All authors contributed to the writing of the paper.
Funding: Danish Health Science Research Council (grant Nos 12-1970-1 and 9501767).
Conflict of interest: None.