Study design and data source
The study was conducted as a fact-finding survey using the unique combination of medical claims receipt data and the results of an online survey on migraine and headaches conducted among members of health insurance associations contracted by DeSC Healthcare Inc (DeSC)., and individuals registered with the health promotion support service application provided by DeSC. This combined use of data sources is thought to be unique and a particular strength of this study design.
The data sources used in this study are medical claims receipt data and online survey data provided by DeSC. This data source is anonymously processed information comprising medical claims receipts and survey data provided by the Society-Managed Employment-Based Health Insurance prior to the start of this study. Therefore, the research that was conducted in this this study used only anonymously processed information that had already been created. The study subjects included in the data used are as follows. Employees (approximately 550,000 people) aged 19-74 who work for large companies on a nationwide scale, including subscribers (approximately 200,000 people) who registered for the mHealth web service “kencom” provided by DeSC. In addition, this study included subscribers who participated in two surveys implemented by DeSC. These were conducted in November 2020 and December 2020 respectively, with a month response period.
The DeSC database contains individual-level demographic information (e.g., age, sex) and dated medical information for inpatient and outpatient service provided by health care organizations and pharmacies (e.g., start of treatment, name of procedure, name of prescription and disease coded in the International Statistical Classification of Diseases and Related Health Problems, 10th revision, name of medical service, cost, name of medical institution). Provided that the employee did not leave the insurance program (e.g. retiring, changing jobs), it was possible to trace the medical and treatment history from multiple institutions.
Ethics statement
As this survey used only anonymized data, and because Otsuka Pharmaceutical Co., Ltd., Clinical Study Support Inc. (CSS)., and the medical experts did not possess or receive data correspondence sheets, it was impossible to identify any individual. In addition, DeSC does not have a correspondence table for the data provided to Otsuka Pharmaceutical Co., Ltd, and it is, therefore, impossible to identify individuals from this data. Therefore, no new individual level consent was obtained for the use of the data in this study. However, the study protocol was approved by the ethics committee of the Research Institute of Healthcare Data Science (approval No.: RI2020012). Additionally, this survey was, however, conducted in consideration of the Declaration of Helsinki (revised October 2013) by the World Medical Association and the Ethical Guidelines for Medical Research Involving Human Subjects.
Study population
We extracted online survey response data and medical claims receipt data from the database for the past three years including the month in which the survey was conducted from 1 December 2017 to 30 November 2020. The study population included in the study was defined as “all individuals whose response data were available”.
Outcome measures
Primary outcomes:
The primary outcomes (based on the online survey response) were:
- The overall number and proportion of people with migraine.
- The overall prevalence of migraine, stratified by age and sex.
The definition of migraine (Supplementary Table 1) was based on the structured survey response, including internal diagnostic criteria, as migraine with and without aura, including probable migraine according to the International Classification of Headache Disorders, 3rd edition (ICHD-3) [13]. Additionally, prevalence of tension-type headache and cluster headache which were also classified according to ICHD-3 were also examined. Individuals not classified in any of these headache types are included in other headache types.
Secondary outcomes:
Use of medical care (based on the online survey response)
The use of medical care was categorized as follows: The number of individuals with migraine who made hospital visits for headache (regularly visited, not regularly visited, not visited) within 6 months before answering the survey, overall and stratified by age and sex; frequency of hospital visits (<once/month, once/month, once/2 months, once/3 months, <once/3 months); reasons for visiting a hospital (unable to tolerate headaches, worried about other brain diseases, increased headache frequency, and OTC drugs no longer effective; and reasons for not visiting a doctor OTC drugs effective, used to having headache, spontaneously resolving after endurance, or pain not sufficiently severe)
Clinical features and symptoms of headache (based on the online survey response)
Clinical features/symptoms were classified as follows: symptoms of headache (nausea or vomiting, stiff shoulder, and neck pain); site of pain (unilateral, bilateral, frontal, occipital, periorbital, other locations); time of day of headache onset (Upon waking, morning, afternoon, evening, other, no particular time of day); headache triggers (fatigue physical or mental stress, bad weather such as the time of typhoon, lack of sleep, turning points of the seasons, sunny or rainy days, work or housework, and menstruation); activities that were refrained from or reduced by headache (Operating a computer or smart phone, drinking alcohol, exercising such as playing sports or walking, going to crowded places, driving a car, housework (excluding grocery shopping, laundry, and cooking), cooking, socializing with friends and playing with children, going grocery shopping, and taking public transportation)
Medication use (based on the online survey response and medical claims receipt data) and comorbidity (medical claims receipt data)
Medical use was classified as follows: current medication use (OTC and prescription drugs, prescription drugs only (both acute and prophylactic medications), OTC drugs only, and none); number of OTC drug class use (1 or ≥2 types); prescription drugs for prophylactic treatments (antidepressants, anti-epileptics, calcium channel blockers, angiotensin-receptor blockers/ angiotensin converting enzyme inhibitors, Beta blockers, and others); prescription drugs for acute treatments (acetaminophen, NSAIDs, triptans, ergotamine, and antiemetic drugs); and comorbidity (hypertension, cardiovascular disease, cerebrovascular disease, gastrointestinal disorder, psychiatric and psychosomatic disorder, depression, epilepsy, asthma, allergy, and autoimmune disorder)
Activity impairment, MS-QOL, and Work Productivity and Activity Impairment WPAI score (based on the online survey response)
Activity impairment was classified as follows: severity of migraine when taking medicines or not taking medicines (severe, moderate, mild); impairment in daily activities (severe, moderate, mild); Migraine-Specific Quality of Life (MS-QOL) estimated using the MSQ version 2.1, which is a 14-item questionnaire measuring the impact of migraine across 3 domains during the past 4 weeks: role function-restrictive (RR) that measures functional limitations on daily, work, and social activities (7 items); role function-preventive (RP) that measures functional prevention on daily, work, and social activities (4 items); and emotional function (EF) that measures the impact on emotion (3 items) [14,15,16]. The source data responses were scaled to range from 0 to 100; the higher score indicating better quality of life. Work Productivity Activity Impairment (WPAI) scores were estimated using the WPAI Questionnaire-General Health for the last 7 days before questionnaire response as follows: 1) percentage of work time missed in the last week due to health conditions (absenteeism); 2) percentage of impairment while working due to health conditions (presenteeism); 3) percentage of overall work impairment due to health conditions; and 4) percentage of activity impairment due to health conditions [17].
Statistical analysis
The analysis population included all individuals who had response data for the surveys conducted by DeSC. Demographic and clinical characteristics were descriptively summarized for overall individuals with migraine. For continuous variables, the mean ± standard deviation and median (minimum, maximum) were presented. For categorical variables, the number and percentage were presented. Post-hoc analysis of other headaches was conducted using criteria from the ID-Migraine [18] and the 4-item simple migraine screener [19]. Additionally, a sensitivity analysis of the health outcomes of kencom users was conducted, to evaluate whether kencom users showed greater health conscious behaviors than other patients. All statistical analyses were performed in SAS Release 9.4 (SAS Institute, Inc., NC, USA).