Introduction

Headache is a prevalent condition worldwide. Epidemiological data from Japan revealed that approximately 36% of participants self-identified as ‘headache sufferers’ in a telephone interview.1 Primary headache negatively affects the quality of life (QOL) of patients.2 According to the Global Burden of Disease (GBD) 2015 report, headache disorders, including migraine, tension-type headache and medication overuse headache (MOH), account for 51% of disability-adjusted life years attributed to neurological disorders.3 Among these headache disorders, migraine is a prominent cause of years lived with disability, ranking second worldwide and fourth in Japan according to the 2016 GBD findings.4

Physician diagnosis is crucial for identifying life-threatening secondary headache and for improving headache management tailored to specific headache types. Recent advancements in headache management, particularly for migraine, are expected to enhance the quality of healthcare for people with headache. However, several barriers hinder individuals with headache from seeking medical attention or accessing appropriate treatment. Previous research conducted in Japan revealed that approximately 90% of individuals with tension-type headache5 and 60%–70% of those with migraine have never sought medical attention.1 5 6 Approximately half the patients managed their symptoms using over-the-counter (OTC) medications.1 5 It is suggested that a lack of awareness regarding headache contributes to the low rate of physician consultation.1 5 6 Many individuals also believe that their headache is not severe enough to warrant medical attention5 6 or that the pain would subside.6 To address the needs of ‘silent sufferers’, identifying the characteristics of individuals who experience headache but do not seek medical attention is essential.

To understand the current state of headache treatment and the impact of migraine, we conducted a comprehensive study combining an online survey with a medical claims database. Our previous reports demonstrated the prevalence, characteristics, clinical features, status of medical attention seeking, medication use and the overall impact of migraine7 and other types of headache.8 Our findings demonstrated that across all headaches, including migraine, tension-type headache, cluster headache and other headache types, only a small percentage of individuals sought regular medical attention, whereas the majority had never consulted a doctor specifically for their headache (migraine, 81.0%7; tension-type headache, 92.0%8; cluster headache, 57.1%8; and other headache types, 90.7%8).

This study focuses on physician consultation for headache to describe the characteristics of individuals who sought medical attention for their headache and those who did not, by examining participant characteristics, status of medical attention seeking, medication use and the impact of headache.

MethodsStudy design and population

This observational study used pre-existing anonymised data from a cross-sectional online survey and linked medical claims data. The methodologies are described in detail in our previous reports.7 8 In summary, the dataset used was provided by DeSC Healthcare Inc (DeSC) and comprised medical claims and survey data provided by multiple Society-Managed Employment-Based Health Insurance associations, under a contract with DeSC and an agreement on the secondary usage of data. It is important to note that Society-Managed Employment-Based Health Insurance associations primarily provide coverage for employees and their dependents in large Japanese companies. These associations obtained consent for the secondary usage of medical data from their subscribers, on whom DeSC conducted the survey. The survey data were linked with claims data before data anonymisation. Therefore, we used only the data that had already been linked and anonymised before the study.

Between November 1 and November 30, 2020, an online survey was conducted using the kencom health promotion application service provided by DeSC. The survey was distributed to all kencom registrants aged 19–74 years who were members of Society-Managed Employment-Based Health Insurance associations. The survey consisted of 69 questions; a summary of the outcome measures is described in the subsequent section. In addition, we extracted prelinked medical claims data for the survey respondents over 3 years from December 1, 2017, to November 30, 2020. To ensure the reliability of survey response data, we compared self-reported sex and age information in survey data with the corresponding information in claims data and included respondents with identical data in both (n=21 480). Here, we focused on survey respondents who reported experiencing headache (n=7311) (online supplemental figure 1).

Ethics statement

This study used the data anonymously processed and provided to DeSC by the Society-Managed Employment-Based Health Insurance associations before the study commencement. This strict anonymisation process ensured that individuals could not be identified under any circumstances. Therefore, obtaining new individual-level consent for this study was not necessary.

Measures

Information regarding the participant characteristics, status of medical attention seeking, medication use and impact of headache were extracted.

Participant characteristics

To describe participant characteristics, the following data were extracted from survey data: sex, age, age at onset of periodic headache, job category, socioeconomic status including annual household income and the number of days with headache (past 3 months and 30 days). Medical claims codes related to headache recorded within the past 6 months, and any existing comorbidities were extracted from claims data.

Status of medical attention seeking

The current status of medical attention seeking was defined based on the response to the survey question: ‘Are you currently visiting hospitals for headache or migraine (within the past 6 months)’? Those indicating ‘visiting regularly’ or ‘visiting but not regularly’ were classified as seeking medical attention for headache, whereas those who selected ‘not visiting’ were classified as not seeking medical attention for headache. While there may be recall bias, survey responses are believed to broadly reflect individual’s perceptions regarding the need for physician consultation for headache. Additionally, we extracted survey data on temporal and financial burden associated with doctor visits and the reasons for current and past visits (within the last 3 years for those who did not seek medical attention).

The reasons for their initial doctor visit among respondents who sought medical attention were extracted from the survey data. Among those who did not seek medical attention, the following reasons were extracted from the survey data, based on the pattern of past visits: (1) reasons for their initial doctor visit (respondents who sought medical attention in the past 3 years but not thereafter), (2) reasons for their initial visit to the doctor and for not seeking further medical attention (respondents who had only sought medical attention once in the past 3 years but not thereafter) and (3) reasons for not visiting a doctor in the past 3 years (respondents who never sought medical attention in the past 3 years).

Medication use

We extracted survey data on whether respondents were currently taking medication for their headache, current use of prescription drugs, reasons for no longer taking previously prescribed medications and the number of OTC analgesics they had used.

Based on survey and claims data in the past 6 months, we assessed the percentage of individuals currently using headache medication, by their medication usage patterns of OTC and/or prescription drugs. During the same 6-month period, we examined the claims data to identify the types of acute and prophylactic drug prescriptions and the total days of supply of acetaminophen or non-steroidal anti-inflammatory drug (NSAID) prescribed.

Impact of headache

Information on the impact of headache was derived from survey data of Migraine-Specific QOL (MSQ) questionnaire V.2.1,9–11 Work Productivity and Activity Impairment (WPAI) Questionnaire-General Health (WPAI-GH),12 and headache intensity and daily activity impairment. MSQ questionnaire measures the functional limitations of migraine on daily activities, work and social activities (role function-restrictive domain); the prevention of daily activities, work and social activities (role function-preventive domain); and emotional well-being (emotional function domain). The converted scores range from 0 to 100, with higher scores indicating better QOL.10

WPAI-GH assesses the impact of respondent health problems on four domains: absenteeism (missed work time), presenteeism (impairment at work), overall work impairment (overall productivity loss at work) and activity impairment (impairment of non-work-related daily activities), yielding four domain scores. The scores are expressed as impairment percentages, with higher percentages indicating greater impairments and reduced productivity.

The headache intensity and daily activity impairment were each measured in the survey using a 5-point scale: no, mild, moderate, moderately severe and severe.

Statistical analysis

Data were summarised for two groups: individuals who sought medical attention for headache and those who did not. All analyses were descriptive. Since the aim was to describe the characteristics of individuals who sought medical attention and those who did not, statistical tests were not conducted, except for the analysis of MSQ and WPAI-GH scores, where a two-sample t-test was conducted assuming equal variance between groups (a two-sided significance level of p<0.05) as a reference to aid the interpretation of the data. Therefore, the description and interpretation of the results are based on the summary of descriptive data. Continuous data are presented as median and mean±SD, 1st and 3rd quartiles, and minimum or maximum values. Categorical data are presented as absolute frequencies and percentages. As a post hoc analysis, we compiled reasons for not consulting a doctor within a subgroup of participants who experienced ≥15 headache-days per month but did not seek medical attention. To assess the reliability of self-reported status of medical attention seeking, we examined the subgroup prescribed acute and prophylactic treatments among those claiming they did not seek medical attention for headache and aggregated the number and percentage of respondents who had diagnostic codes for headache.

All statistical analyses were performed using SAS Release V.9.4 (SAS Institute Inc, Cary, NC, USA).

Patient and public involvement

None.

ResultsAnalysis population and participant characteristics

Among the 7311 respondents who self-reported experiencing headache, 735 (10.1%) sought medical attention for headache, while 6576 (89.9%) did not (online supplemental figure 1).

The distributions of demographic and socioeconomic status variables were generally comparable between the two groups (table 1, online supplemental table 1).

Table 1

Participant characteristics

The number of days with headache in the past 3 months was numerically higher among respondents who sought medical attention (mean,17, and median, 10) than among those who did not (mean, 8, and median, 5). Similarly, the number of headache days in the past 30 days was slightly higher among those who sought medical attention (mean, 7, and median, 4) than among those who did not (mean, 3, and median, 2). Most respondents experienced fewer than 8 headache-days per month, while 115 (15.6%) respondents who sought medical attention and 239 (3.6%) respondents who did not seek medical attention reported experiencing ≥15 headache-days.

Numerically, the percentage of common comorbidities was generally higher among individuals who sought medical attention than in those who did not (gastrointestinal disorders, 69.1% (508/735) and 56.9% (3744/6576); psychiatric or psychosomatic disorders, 26.3% (193/735) and 10.1% (663/6,576); allergies, 25.3% (186/735) and 15.4% (1010/6576); and hypertension, 20.1% (148/735) and 10.7% (706/6576), respectively).

Status of medical attention seeking

Among the individuals who sought medical attention for headache, 28.4% (209/735) of the respondents reported visiting a doctor regularly. Some respondents who sought medical attention perceived the frequency of their visits as a temporal (32.4% (238/735)) or financial (41.6% (306/735)) burden. Common reasons for initially seeking medical attention for headache were as follows: unable to tolerate headache (36.5% (268/735)), increased headache frequency (24.6% (181/735)), worried about other brain diseases (23.7% (174/735)) and OTC analgesics no longer effective (14.6% (107/735)) (table 2).

Table 2

Reasons for seeking and not seeking medical attention for headache

Among the respondents who did not seek medical attention for headache in the past 6 months, a subset had visited the doctor for headache within the past 3 years. The reasons provided for their previous visits included being unable to tolerate headache (4.5% (298/6576)), worried about other brain diseases (4.5% (299/6576)) and increased headache frequency (2.0% (130/6576)). However, 3.5% (227/6576) of those who did not seek medical attention in the past 6 months stopped seeking further medical attention after a single visit because they were relieved not to have a life-threatening brain disease. Additionally, 1.3% (85/6576) found hospital visits bothersome, 1.2% (81/6576) cited improved symptoms after a doctor visit, and 0.8% (50/6576) reported having no time as the reason for discontinuing their visits.

Furthermore, among the respondents who did not seek medical attention for headache in the past 6 months, the reasons for having never visited a doctor for headache in 3 years were pain not sufficiently severe (35.3% (2323/6576)), OTC analgesics effective (34.9% (2293/6576)) and the belief that the pain would resolve if endured for some time (28.0% (1844/6576)).

When the analysis was restricted to respondents who experienced ≥15 headache-days per month and did not seek medical attention in the past 6 months (online supplemental figure 2), the most common reasons for not seeking medical attention were OTC analgesics effective (38.7%, 67/173), used to having a headache (31.8%, 55/173) and pain would resolve if endured for some time (30.1%, 52/173). Additionally, 17.9% (31/173) did not have time to go to a hospital and did not know which hospital or department to attend, 16.2% (28/173) did not think headache was a disease to be treated at a hospital, and 15.6% (27/173) saw it as a financial burden. These percentages were higher in this subgroup compared with that in the overall group of individuals who did not seek medical attention.

Medication use

Taking medication to treat headache was common in both groups (651 (88.6%) respondents of those who sought medical attention and 4341 (66.0%) respondents of those who did not) (table 3).

Table 3

Medication use for headache among people who experience headache (n=7311)

In the subgroup of respondents who answered that they were currently taking medication for their headache, OTC drugs were commonly used (solely or with prescription drugs) by those who sought medical attention (82.0%, 534/651) and those who did not (88.9%, 3858/4341) (table 3). Additionally, prescription drugs were commonly used by those who sought medical attention, with 45.0% (331/735) reporting current use of prescribed medications.

According to claims data, a greater number of individuals who sought medical attention were prescribed both acute and prophylactic drugs than those who did not seek medical attention. Among the prescribed medications, acetaminophens/NSAIDs were the most prescribed (50.6% (372/735) among those who sought medical attention and 25.4% (1672/6576) among those who did not). These medications were prescribed for ≥15 days per month (3-month average) in 9.9% (73/735) of individuals who sought medical attention and 2.4% (161/6576) of those who did not.

The post hoc analysis (online supplemental table 2) demonstrated that all respondents who did not seek medical attention but were prescribed triptans had at least one diagnostic record of a headache (12/12). For those prescribed antiepileptics, 16.1% (10/62) had diagnostic codes for headache, while the percentages for other prescriptions were <10%.

Impact of headache

Overall, the mean (SD) MSQ scores were lower in respondents who sought medical attention than in those who did not (total: 77.1 (18.1) vs 87.6 (13.0)), indicating a lower QOL in those who sought medical attention. This difference was observed across all domain scores with statistical significance (p<0.0001) (figure 1A).

Figure 1

Migraine-Specific Quality of Life questionnaire (A) and Work Productivity and Activity Impairment: General Health scores (B) *among individuals who sought medical attention and those who did not. A two-sample t-test was used to compare the mean scores or percentages between the groups, with a two-sided significance level of p<0.05. *The analysis of the Work Productivity and Activity Impairment included respondents who answered the questions (643 of those who sought medical attention for headache and 5859 of those who did not) for all variables. However, all individuals were analysed for the Activity Impairment.

Similarly, across all the WPAI-GH domains, including absenteeism, presenteeism, overall work impairment, and activity impairment, the percentage of impairment was higher (indicating greater impairment) in individuals who sought medical attention than in those who did not, with statistical significance (p<0.0001) (figure 1B).

Among those who sought medical attention, 41.2% (303/735) experienced moderately severe or severe headache when not taking medication for their headache, almost doubling the 19.0% (1252/6576) reported by those who did not seek medical attention (figure 2A). When taking medication, the percentage of individuals with moderately severe or severe headache decreased to 7.2% (47/651) for those who sought medical attention and 7.6% (329/4341) for those who did not.

Figure 2

Headache intensity (A) and daily activity impairment (B) by healthcare-seeking status and medication use in individuals with headache (n=7311). Number of individuals who responded (n, %). Denominator was those who responded the corresponding questions.

The percentage of individuals with moderately severe or severe impairment of daily activities when not taking medication was also higher in those who sought medical attention (24.6%, 181/735) than in those who did not (7.3%, 478/6576) (figure 2B). When taking medication, the percentage decreased to 4.5% (29/651) in those who sought medical attention and 2.4% (103/4341) in those who did not.

Discussion

Using data from the survey and linked medical claims, we described the characteristics of individuals who sought medical attention for headache and those who did not. Consequently, statistical tests were not conducted, and interpretation of the results is based on the summary of the descriptive data. Those who sought medical attention exhibited higher comorbidity rates, more frequent and intense headache, lower QOL and greater WPAI, and greater daily activity impairment when not taking medication than those who did not seek medical attention. Regardless of their current healthcare-seeking status, most respondents with headache self-managed their pain using OTC drugs.

Individuals seeking medical attention for headache exhibited severe conditions in terms of frequency, headache intensity and impact of headache. A recent survey on migraine epidemiology in Japan also suggested that elevated frequency of migraine episodes was accompanied by an increased percentage of patients visiting doctors.13 The primary reasons for never visiting a doctor for headache were pain was not sufficiently severe and OTC analgesics were effective, aligning with previous reports.5 6 13 14 Conversely, the most common reasons for the initial visit were inability to tolerate headache and increased headache frequency. These results showed that many individuals with headache in Japan initially relied on OTC analgesics or tolerated the pain, regarding their condition as mild, and did not seek medical attention until their condition worsened to an unbearable level and started to negatively affect their lives. However, if the condition is left untreated, some patients may increase the frequency of OTC analgesics without knowing that frequent use of analgesics and acute treatment induces a shift from episodic to chronic migraine, potentially leading to MOH.15 Disseminating MOH risk information among self-medicating individuals, particularly those who were unattended by physicians, may be important.

The higher proportion of individuals who sought medical attention used prescribed and OTC drugs, compared with those who did not. Furthermore, a higher proportion of participants who sought medical attention had comorbidities. Although we did not evaluate medications used for comorbidities, these participants may have used medications, such as analgesics, for their comorbidity alongside medications for headache. Considering these results with the trend of more intense and frequent headache in those who sought medical attention, we infer that this group might use more analgesics than those who did not seek medical attention. Analgesic overuse, even for conditions other than headache, predisposes individuals with primary headache to a higher risk of MOH.16 Although the amount or duration of OTC drug use was unavailable in the database, clinical experience suggests that self-medication with OTC drugs, without guidance from healthcare professionals and in conjunction with prescribed medications, may elevate the risk for MOH. While the available data did not allow precise identification of individuals who met the MOH criteria, the results revealed that a higher percentage of those seeking medical attention were prescribed acetaminophen/NSAIDs for an average of ≥15 days per month, heightening their risk of developing MOH if the prescriptions persisted at the same or higher frequency.

The proportion of individuals experiencing moderately severe or severe headache without medication was twice as high among those who sought medical attention as it was among those who did not. This suggests that those with severe headache are more likely to seek medical attention than those with milder headache. When taking medication, the proportion of individuals experiencing such severe headache decreased by a greater extent in those who sought medical attention than in those who did not. Additionally, among those not currently seeking medical attention, some had previously visited doctors but discontinued their visits because their symptoms improved after a doctor’s consultation. While we cannot establish causality between the physician visit and headache improvement, these findings suggest that if patients seek medical attention to receive proper treatment and guidance for medication use, their headache may be alleviated. Given the more severe conditions, substantial impact of headache and frequent use of analgesics among those who sought medical attention, it is plausible that if these participants visited doctors before their headache became too severe, the impact on their lives and the reliance on analgesics might have been reduced, at least partially.

Interestingly, approximately 4% of individuals who did not seek medical attention had headache for ≥15 days in the past 30 days. Conversely, only one-third of all respondents who had headache for ≥15 days per month sought medical attention for headache. Moreover, the maximum number of headache days in the past 30 days and 3 months was 30 and 92 days, respectively, for both groups. This indicates that some individuals did not seek medical attention despite experiencing headache almost every day. Further examination of the reasons within a subgroup of individuals with ≥15 headache-days who did not seek medical attention suggested that some individuals were aware of their symptoms and the need for professional treatment. However, they were hindered by a lack of knowledge, such as not knowing which hospital or department to attend, and constraints, such as time limitations or financial resources. Among the participants who did not seek medical attention despite having frequent headaches for ≥15 days, patients with chronic migraine, chronic tension-type headaches and MOH were potentially included. Therefore, we consider it particularly important to raise awareness about this disease and its treatment and improve access to care for these patients.

Approximately 10% of individuals who did not seek medical attention did not consider headache as a disease that necessitated medical treatment. This perspective may lead to a disregard for the potential risk of serious underlying diseases associated with headache. Moreover, some respondents reported initially visiting the doctor because of concerns regarding brain diseases, while others discontinued their visits after receiving assurance that their condition was not life-threatening. Respondents might also believe that merely experiencing headache may not warrant a hospital visit and that only serious, life-threatening diseases require professional treatment.

Collectively, these findings highlight the importance of raising awareness among individuals who do not seek medical attention for headache. It is crucial to convey the message that headache should not be disregarded or underestimated but require proper diagnosis and treatment by physicians. Increasing awareness of available treatment options empowers individuals to make informed decisions about seeking appropriate care for headache.

This study employed a unique combination of online surveys and medical claims data to complement the limitations associated with each data source. The self-administered survey provided insight into the perspectives of individuals with headache, which is valuable considering the low rate of physician consultations and frequent OTC drug use. However, the survey might be subject to recall bias, demonstrated by instances where some patients reported not seeking medical attention for headache despite being prescribed analgesics and having headache diagnostic codes.

Notably, medical claims data can entail potential risks of misclassification, as some diagnostic codes may have only been recorded for claims purposes and not necessarily reflect the actual clinical diagnosis. Furthermore, the therapeutic indication for the prescription was unidentifiable based on medical claims data alone.

As acknowledged in our previous reports,7 8 another limitation of this study is the limited generalisability of our findings, which can be attributed to the nature of the database used, which consists of health insurers contracted by the DeSC. The database population consists of employees and their dependents in large Japanese companies, which may explain the similar socioeconomic status, particularly the annual household income, between the two groups. Furthermore, our study population represented a subset of individuals who used the kencom service, indicating an assumed higher awareness of health-related matters, further limiting the generalisability of these results. To address this concern, we compared the proportion of patients who sought medical attention between kencom users and non-users using the medical claims data and found no notable differences between the two groups (data not shown). Further longitudinal studies examining the impact of self-medication and delayed medical attention seeking are required. Such studies would validate our findings, providing valuable insights into patient decision-making and empowering healthcare professionals to better engage with patients.