Sage Journals: Discover world-class research

Abstract

Background

The present study aimed to investigate prescription patterns for patients aged over 17 years with headaches in the REZULT database.

Methods

We conducted a cross-sectional study (Study 1) of the proportion of over-prescription of acute medications (≥30 tablets/90 days for triptans, combination non-steroidal anti-inflammatory drugs (NSAIDs) and multiple types; ≥45 tablets/90 days for single NSAIDs) among patients with headache diagnosed in 2020. We longitudinally studied (Study 2) patients for >2 years from initial headache diagnosis (July 2010 to April 2022). The number of prescribed tablets was counted every 90 days.

Results

In Study 1, headache was diagnosed in 200,055 of 3,638,125 (5.5%) patients: 13,651/200,055 (6.8%) received acute medication. Single NSAIDs were prescribed to 12,297/13,651 (90.1%) patients and triptans to 1710/13,651 (12.5%). Over-prescription was found in 2262/13,651 (16.6%) patients and 1200/13,651 (8.8%) patients received prophylactic medication. In Study 2, 408,183/6,840,618 (6.0%) patients were first diagnosed with headaches, which persisted for ≥2 years. Over time, the proportion of patients over-prescribed acute medications increased. Over 2 years, 37,617/408,183 (9.2%) patients were over-prescribed acute medications and 29,313/408,183 (7.2%) patients were prescribed prophylaxis at least once.

Conclusions

According to real-world data, prophylaxis remains poorly prescribed, and both acute and prophylactic treatment rates for headaches have increased over time.

Keywords

Acute mediations artificial intelligence medication-overuse headache NSAIDs prophylaxis triptans

Introduction

Primary headaches are a widespread public health problem, with approximately 12% of migraine and up to 50% of tension-type headache (TTH) cases observed in the population (1,2). The International Classification of Headache Disorders, 3rd edition (ICHD-3) includes migraine, TTH and trigeminal autonomic cephalalgias as representative primary headaches, which greatly hinder patients’ daily activities. Primary headache management includes acute medications for headache attacks and prophylactic medications to reduce the frequency and severity of headache attacks (3). Inappropriate management without prophylactic medication for severe headache disorders sometimes increases the risk of chronic migraines, treatment resistance, comorbidities and medication-overuse headache (4,5). However, headache specialists remains few, and general physicians often treat primary headaches worldwide (6 –9). Consequently, inappropriate and over-prescription of acute medications without the prescription of prophylactic medications is a concern.

Recently, National Health Insurance claims data-based research entailing numerous patient datasets as real-world data has been published (10 –15). However, similar research in Japan is rare (16 –18). Because big data research can shed light on inappropriate prescriptions by physicians in Japan, we investigated a health insurance database to clarify treatment patterns for headaches. In particular, using real-world data, we examined the over-prescription of acute medications by physicians. The present study differs from previous studies because of its large sample, cross-sectional evaluation of prescription patterns of acute medications in Japan and longitudinal evaluation of changes in prescription patterns, and the onset of over-prescriptions was investigated over 2 years from the time of first headache diagnosis.

Methods

All residents in Japan are required to have health insurance coverage offered by employee-based and community-based insurers, and the government sets fee schedules for medical services (16). Receipt data of employee-based insurers for >8 million patients have been registered in the REZULT database by Japan System Techniques Co., Ltd. Tokyo, Japan (https://www.jastlab.jast.jp/rezult_data). The REZULT database contains data on more than 420 million receipts from 147 insurance companies, permitting continuous longitudinal evaluation of prescriptions and treatments for patients for up to 12 years. It records the date, disease name for the insurance claim, age and sex of the patient, and medical treatment, including prescription and medical institution details. Each registered patient has an anonymous unique identification number, allowing complete prescription observation over time.

In Japan, combination analgesics and single analgesic agents can be purchased as over-the-counter (OTC) medicines. However, sales of triptans and prophylactic medication for migraine as OTCs are prohibited. In other words, prophylactic treatment and triptans can only be prescribed by physicians to patients diagnosed with migraines. Contrastingly, single and combination non-steroidal anti-inflammatory drugs (NSAIDs) can be prescribed for headaches without a migraine diagnosis. NSAIDs are also prescribed for other chronic pain disorders, including lower back pain. Unfortunately, the database used in the present study does not track the disease for which they were prescribed.

Study 1 (cross-sectional study): prescription pattern for patients with headache

Data were obtained from the REZULT database. We extracted data of patients aged >17 years with one-year continued diagnosis of headache (International Statistical Classification of Diseases and Related Health Problems (ICD)-10 code R51) or other headache syndromes (ICD-10 code G44) in 2020 and those who were prescribed certain medications listed in Table 1 to investigate the number of patients treated with an over-prescription of acute medications. No distinction was made based on whether or not the patient had a diagnosis of migraine (ICD-10 code G43). Drug price list codes are shown in supplementary File 1. Acute treatment was defined as the prescription of a single NSAIDS, a combination of NSAIDs or triptans. Prophylactic treatments included the prescription of lomerizine, propranolol, valproic acid, amitriptyline, verapamil and Japanese herbal Kampo medicine as defined by the Japanese guidelines (3). The number of tablets of acute medications prescribed from January 2020 to December 2020 was counted, and the proportion of over-prescriptions was evaluated. Referring to a previous report on triptan overuse (10), over-prescription was defined according to the ICHD-3 criteria as ‘use of triptans or combination NSAIDs on more than ten days (tablets) a month for three months (≥30 tablets/90 days), that of single NSAIDs on more than 15 days (tablets) a month (≥45 tablets/90 days) or that of any combination of single, combination and/or triptans (≥30 tablets/90 days) as well as one that does not constitute over-prescription for a single drug. As it was not possible to determine the number of days for which acute care drugs were actually used over the study period, ‘number of tablets per 90 days’ was counted instead of the ‘number of days per month when acute medications were used’.

Table 1.

Summary of acute and prophylactic medications.

Treatment type	Drug name
Acute treatment	Combination of non-steroidal anti-inflammatory drugs
	Single non-steroidal anti-inflammatory drugs
	Triptans
Prophylactic treatment	Calcium-channel blockers (lomerizine)
Prophylactic treatment	Beta-blockers (propranolol)
	Anticonvulsants (valproic acid)
	Antidepressants (amitriptyline)
	Calcium-channel blockers (verapamil)
	Japanese herbal kampo medicine

The 90-day period for which the highest number of acute medications were prescribed was selected for each patient, and the prescription pattern was examined. Each table shows the prescribing pattern over the 90-day period when each drug was most prescribed. Prophylactic treatment was defined as ‘more than one tablet prescription during the 90 days’. Most prophylactic medications for headaches have multiple indications, including hypertension, epilepsy and depression.

Study 2 (longitudinal study): over-prescription during the 2 years from the initial diagnosis

We tracked changes in prescription patterns over 2 years from the initial headache diagnosis to determine how many individuals were over-prescribed with too many acute medications. Data for 6,840,618 patients aged over 17 years from the REZULT database who could be followed up for >2 years from the initial headache diagnosis (July 2010 to April 2022) were analysed. The number of prescribed tablets was counted every 90 days from the initial headache diagnosis. We analysed the proportion of patients who experienced over-prescription during the study period. Age, sex, time when patients fell into over-prescription of acute medication, prescription amount of the acute medication during the first 90 days and prescription of prophylactic medication during the first 90 days were investigated. A Mann–Whitney U-test was used to compare the number of medical facilities at which doctors prescribed headache medication, including whether or not the patient was over-prescribed over 2 years. Cox regression analysis was performed to identify factors related to the time for which patients were over-prescribed acute medication. No prescriptions for calcitonin gene-related peptide (CGRP)-related drugs were observed in the data set during this period.

Finally, an artificial intelligence (AI)-based model was developed as a sensitivity analysis, with the number of prescribed acute and prophylactic tablets in the first 90 days as the ‘explanatory variable’ and whether the patient was over-prescribed acute medications during the 2 years from the initial headache diagnosis as the ‘objective variable’. The model was developed as previously reported (19,20) and in accordance with relevant guidelines (21). The Python code is provided in supplementary File 2. The sensitivity, specificity, and area under the receiver operating characteristic (ROC) curve (AUC) were evaluated, and the accuracy of the model was evaluated. Important variables were evaluated by their gain, the common variables were statistically identified using Cox regression models and a sensitivity analysis was conducted.

Statistical analysis

Normal distribution was confirmed using the Shapiro–Wilk test. Variables with a normal distribution are presented as the mean ± SD, whereas those with a non-normal distribution are presented as the median (interquartile range). A log-minus-log plot was used to check whether the assumption of proportional hazards was satisfied. p < 0.05 (two-tailed) was considered statistically significant. SPS, version 29.0.0 (IBM Corp., Armonk, NY, USA), Python, version 3.9.0 (https://www.python.org/), Pandas, version 2.0.2 (https://pandas.pydata.org/), PyCaret, version 3.1.0 (https://pycaret.org/), scikit-survival, version 0.21.0 (https://scikit-survival.readthedocs.io/en/stable/) and Matplotlib, version 3.5.1 (https://matplotlib.org/) were used.

Ethics

The study was approved by the Itoigawa General Hospital Ethics Committee (approval number: 2022-2, 2022-10). The requirement for written informed consent was waived as a result of the anonymous and retrospective nature of the study. All experiments were performed in accordance with the tenets of the Declaration of Helsinki and conducted according to the Strengthening the Reporting of Observational Studies in Epidemiology guidelines and the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis guidelines.

Results

Study 1 (cross-sectional study): prescription pattern for patients with headache

Among 3,638,125 patients aged >17 years in the 2020 REZULT database, 200,055 (5.50%) were diagnosed with headaches via the health insurance system. Furthermore, 13,651/200,055 (6.82%) patients received at least one prescription of acute medication in 2020; their mean (SD) age was 31.82 ± 12.23 years with median 31 (range 18–74) years and 8889/13,651 (65.12%) were female. Single NSAIDs (acetaminophen and loxoprofen made up the majority) were the most frequently prescribed acute medications (12,297/13,651 patients; 90.08%) and triptans were prescribed to 1710/13,651 (12.53%) patients. Most patients received only acute treatment (12,451/13,651 patients; 91.21%) and 1200/13,651 (8.79%) patients received prophylactic treatment. The number of patients over-prescribed with at least one single NSAIDs (≥45 tablets/90 days), combination NSAIDs, triptans or combination of those (≥30 tablets/90 days) was 2262/13,651 (16.57%) (Table 2). Among 2262 over-prescribed patients, 303/2262 (13.40%) received prophylactic treatment.

Table 2.

Treatment pattern for headache patients.

Characteristics	Overall	%/SD	Patients with acute treatment only	%/SD	Patients with triptans without prophylactic treatment	%/SD	Patients with both acute and prophylactic treatment	%/SD	Patients with triptans and prophylactic treatment	%/SD
Number	13651	100%	12451	91.21%	1274	9.33%	1200	8.79%	436	3.19%
Number of over-prescription (%)	2262	16.57%	1959	15.73%	325	25.51％	303	25.25%	149	34.17%
Age (years), mean, SD	31.82	12.23	31.74	12.31	32.58	10.88	32.67	11.24	33.85	11.07
Sex: female	8889	65.12%	8038	64.56%	617	69.23%	851	70.91%	316	72.48%
Acute treatment
Combination NSAIDs	352	2.58%	306	2.46%	–	–	46	3.83%	–	–
Single NSAIDs	11380	83.36%	10690	85.86%	–	–	690	57.50%	–	–
Triptans	928	6.80%	700	5.62%	700	54.95%	228	19.00%	228	52.29%
Combination and single NSAIDs	209	1.53%	181	1.45%	–	–	28	2.33%	–	–
Combination NSAIDs and triptans	74	0.54%	47	0.38%	47	3.69%	27	2.250%	27	6.19%
Single NSAIDs and triptans	684	5.01%	508	4.08%	508	39.87%	176	14.67%	176	40.37%
All	24	0.18%	19	0.15%	19	1.49%	5	0.42%	5	1.15%
Prophylactic treatment	1200	8.79%	–	–	–	–	1200	100.00%	436	100.00%
Single	1072	7.85%	–	–	–	–	1072	89.33%	346	79.36%
Two types	114	0.84%	–	–	–	–	114	9.50%	81	18.58%
Three types	13	0.10%	–	–	–	–	13	1.08%	8	1.84%
Four types	1	0.01%	–	–	–	–	1	0.08%	1	0.23%
Calcium-channel blockers (lomerizine)	156	1.14%	–	–	–	–	156	13.00%	108	24.77%
Beta-blockers (propranolol)	33	0.24%	–	–	–	–	33	2.75%	14	3.21%
Anticonvulsants (valproic acid)	162	1.19%	–	–	–	–	162	13.50%	97	22.25%
Antidepressants (amitriptyline)	63	0.46%	–	–	–	–	63	5.25%	35	8.03%
Calcium-channel blockers (verapamil)	18	0.13%	–	–	–	–	18	1.50%	13	2.98%
Japanese herbal kampo medicine	640	4.69%	–	–	–	–	640	53.33%	79	18.12%
Lomerizine and beta-blockers	1	0.01%	–	–	–	–	1	0.08%	0	0%
Lomerizine and anticonvulsants	23	0.17%	–	–	–	–	23	1.92%	20	4.59%
Lomerizine and antidepressants	7	0.05%	–	–	–	–	7	0.58%	5	1.15%
Lomerizine and verapamil	2	0.02%	–	–	–	–	2	0.17%	2	0.46%
Lomerizine and kampo medicine	28	0.21%	–	–	–	–	28	2.33%	22	5.05%
Beta-blocker and anticonvulsants	3	0.02%	–	–	–	–	3	0.25%	3	0.69%
Beta-blocker and Antidepressants	1	0.01%	–	–	–	–	1	0.08%	1	0.23%
Beta-blocker and kampo medicine	3	0.02%	–	–	–	–	3	0.25%	1	0.23%
Anticonvulsants and antidepressants	9	0.07%	–	–	–	–	9	0.75%	6	1.38%
Anticonvulsants and verapamil	2	0.02%	–	–	–	–	2	0.17%	2	0.46%
Anticonvulsants and kampo medicine	18	0.13%	–	–	–	–	18	1.00%	12	2.75%
Antidepressants and kampo medicine	14	0.10%	–	–	–	–	14	1.17%	6	1.38%
Verapamil and kampo medicine	3	0.02%	–	–	–	–	3	0.25%	1	0.23%
Lomerizine, antidepressants, kampo	2	0.02%	–	–	–	–	2	0.17%	2	0.46%
Lomerizine, propranolol, kampo	3	0.02%	–	–	–	–	3	0.25%	2	0.46%
Lomerizine, propranolol, anticonvulsants	2	0.02%	–	–	–	–	2	0.17%	0	0%
Lomerizine, anticonvulsants, antidepressants	2	0.02%	–	–	–	–	2	0.17%	1	0.23%
Lomerizine, anticonvulsants, kampo	4	0.03%	–	–	–	–	4	0.33%	3	0.69%
Lomerizine, anticonvulsants, antidepressants, kampo	1	0.01%	–	–	–	–	1	0.08%	1	0.23%

NSAIDs, non-steroidal anti-inflammatory drugs.

Single NSAIDs were prescribed to 12,297/13,651 (90.08%) patients with headache (Table 3). The mean ± SD and median number of tablets of a single NSAIDs during 90 days were 27.63 ± 34.88 and 18, respectively. Prophylactic treatment was prescribed to 899/12,297 (7.31%) patients. The number of patients over-prescribed with a single NSAIDs (≥45 tablets/90 days) was 1739/12,297 (14.14%); however, the mean ± SD number of tablets was 87.44 ± 56.22. Among those patients, 174/1739 (10.01%) received prophylactic treatment.

Table 3.

Treatment pattern of single NSAIDs.

Characteristics	Overall	%/SD	0–11 tablets/90 days	%/SD	12–29 tablets/90 days	%/SD	30–44 tablets/90 days	%/SD	≥45 tablets/90 days	%/SD
n	12297	100%	4159	33.82%	4491	36.52%	1908	15.53%	1739	14.14%
Number of tablets/90 days	27.63	34.88	7.72	2.60	18.95	4.49	36.11	5.12	87.44	56.22
Age (years), mean, SD	31.62	12.29	30.96	12.80	31.02	11.51	31.59	11.42	33.12	12.69
Sex: female	7978	64.88%	2570	61.79%	2860	63.68%	1310	68.66%	1238	71.19%
Acute treatment
Single NSAIDs	11380	92.54%	3875	93.17%	4165	92.74%	1760	92.24%	1580	90.86%
Combination and single NSAIDs	209	1.70%	68	1.64%	75	1.67%	25	1.31%	41	2.36%
Single NSAIDs and triptans	684	5.56%	206	4.95%	239	5.32%	122	6.39%	117	6.73%
All	24	0.20%	10	0.24%	12	0.27%	1	0.05%	1	0.06%
Prophylactic treatment	899	7.31%	293	7.05%	296	6.59%	136	7.13%	174	10.01%
None	11398	92.69%	3866	92.96%	4195	93.41%	1772	92.87%	1565	89.99%
Single	825	6.71%	279	6.71%	264	5.88%	126	6.60%	156	8.97%
Two types	67	0.55%	14	0.34%	30	0.67%	9	0.47%	14	0.81%
Three types	6	0.05%	0	0%	2	0.05%	1	0.05%	3	0.17%
Four types	1	0.01%	0	0%	0	0%	0	0%	1	0.06%
Calcium-channel blockers (lomerizine)	80	0.65%	25	0.60%	18	0.40%	14	0.73%	23	1.32%
Beta-blockers (propranolol)	27	0.22%	5	0.12%	12	0.27%	6	0.31%	4	0.23%
Anticonvulsants (valproic acid)	92	0.75%	36	0.87%	21	0.47%	10	0.52%	25	1.44%
Antidepressants (amitriptyline)	41	0.33%	10	0.24%	9	0.20%	9	0.47%	13	0.75%
Calcium-channel blockers (verapamil)	9	0.07%	2	0.05%	5	0.11%	2	0.11%	0	0%
Japanese herbal kampo medicine	576	4.68%	201	4.83%	199	4.43%	85	4.46%	91	5.23%
Lomerizine and beta-blockers	1	0.01%	1	0.02%	0	0%	0	0%	0	0%
Lomerizine and anticonvulsants	10	0.08%	1	0.02%	7	0.16%	0	0%	2	0.12%
Lomerizine and antidepressants	4	0.03%	0	0%	4	0.09%	0	0%	0	0 %
Lomerizine and verapamil	2	0.02%	0	0%	0	0%	0	0%	2	0.12%
Lomerizine and kampo medicine	16	0.13%	4	0.10%	5	0.11%	3	0.16%	4	0.23%
Beta-blocker and anticonvulsants	2	0.02%	0	0%	2	0.05%	0	0%	0	0%
Beta-blocker and kampo medicine	2	0.02%	0	0%	1	0.02%	0	0%	1	0.06%
Anticonvulsants and antidepressants	5	0.04%	1	0.02%	3	0.07%	1	0.05%	0	0%
Anticonvulsants and kampo medicine	11	0.09%	6	0.14%	2	0.05%	1	0.05%	2	0.12%
Antidepressants and kampo medicine	12	0.10%	1	0.02%	6	0.13%	3	0.16%	2	0.12%
Verapamil and kampo medicine	2	0.02%	0	0%	0	0%	1	0.05%	1	0.06%
Lomerizine, propranolol, kampo	3	0.02%	0	0%	0	0%	0	0%	3	0.17%
Lomerizine, anticonvulsants, antidepressants	2	0.02%	0	0%	1	0.02%	1	0.05%	0	0%
Lomerizine, anticonvulsants, kampo	1	0.01%	0	0%	1	0.02%	0	0%	0	0%
Lomerizine, anticonvulsants, antidepressants, kampo	1	0.01%	0	0%	0	0%	0	0%	1	0.06%

NSAIDs, non-steroidal anti-inflammatory drugs.

Combination NSAIDs were prescribed to 659 (4.83%) of 13,651 patients with headaches (see supplementary Table 1). The number of patients treated with over-prescriptions of combination NSAIDs (≥30 tablets/90 days) was 156/659 (23.67%). Among them, 24/156 (15.39%) patients received prophylactic treatment.

Triptans were prescribed to 1710 of 13,651 (12.53%) patients with headache (Table 4) and 106/1710 (6.20%) were over-prescribed with triptans (≥30 tablets/90 days). Among them, 44/106 (41.51%) patients received prophylactic treatment.

Table 4.

Treatment pattern of triptans.

Characteristics	Overall	%/SD	0–11 tablets/90 days	%/SD	12–29 tablets/90 days	%/SD	30–44 tablets/90 days	%/SD	≥45 tablets/90 days	%/SD
n	1710	100%	1217	71.17%	387	22.63%	68	3.98%	38	2.22%
Number of tablets/90 days	11.44	11.01	6.31	3.04	18.08	4.54	35.80	5.00	58.93	16.02
Age (years), mean, SD	32.83	10.94	32.01	10.64	34.67	11.43	34.93	12.05	32.62	12.05
Sex: female	1198	70.06%	843	69.27%	274	70.80%	51	75.00%	30	78.95%
Acute treatment
Triptans	928	54.27%	670	55.05%	203	52.46%	37	54.41%	18	47.37%
Combination NSAIDs and Triptans	74	4.33%	55	4.52%	17	4.39%	2	2.94%	0	0%
Single NSAIDs and Triptans	684	40.00%	469	38.54%	167	43.15%	28	41.18%	20	52.63%
All	24	1.40%	23	1.89%	0	0%	1	1.47%	0	0%
Prophylactic treatment	436	25.50%	253	20.79%	139	35.92%	31	45.59%	13	34.21%
None	1274	74.50%	964	79.21%	248	64.08%	37	54.41%	25	65.79%
Single	346	20.23%	204	16.76%	114	29.46%	20	29.41%	8	21.05%
Two types	81	4.74%	45	3.70%	21	5.43%	10	14.71%	5	13.16%
Three types	8	0.47%	3	0.25%	4	1.03%	1	1.471%	0	0%
Four types	1	0.06%	1	0.08%	0	0%	0	0%	0	0%
Calcium-channel blockers (lomerizine)	108	6.32%	57	4.68%	40	10.34%	9	13.24%	2	5.26%
Beta-blockers (propranolol)	14	0.82%	9	0.74%	5	1.29%	0	0%	0	0%
Anticonvulsants (valproic acid)	97	5.67%	59	4.85%	33	8.53%	4	5.88%	1	2.63%
Antidepressants (amitriptyline)	35	2.05%	15	1.23%	18	4.65%	1	1.47%	1	2.63%
Calcium-channel blockers (verapamil)	13	0.76%	7	0.58%	4	1.03%	1	1.47%	1	2.63%
Japanese herbal kampo medicine	79	4.62%	57	4.68%	14	3.62%	5	7.35%	3	7.90%
Lomerizine and anticonvulsants	20	1.17%	9	0.74%	10	2.58%	0	0%	1	2.63%
Lomerizine and antidepressants	5	0.29%	3	0.25%	1	0.26%	1	1.47%	0	0%
Lomerizine and verapamil	2	0.12%	2	0.16%	0	0%	0	0%	0	0%
Lomerizine and kampo medicine	22	1.29%	11	0.90%	3	0.78%	6	8.82%	2	5.26%
Beta-blocker and anticonvulsants	3	0.18%	2	0.16%	1	0.26%	0	0%	0	0%
Beta-blockers and antidepressants	1	0.06%	1	0.08%	0	0%	0	0%	0	0%
Beta-blocker and kampo medicine	1	0.06%	0	0%	1	0.26%	0	0%	0	0%
Anticonvulsants and antidepressants	6	0.35%	2	0.16%	2	0.52%	1	1.47%	1	2.63%
Anticonvulsants and verapamil	2	0.12%	1	0.08%	0	0%	1	1.47%	0	0%
Anticonvulsants and kampo medicine	12	0.70%	7	0.58%	3	0.78%	1	1.47%	1	2.63%
Antidepressants and kampo medicine	6	0.35%	6	0.49%	0	0%	0	0%	0	0%
Verapamil and kampo medicine	1	0.06%	1	0.08%	0	0%	0	0%	0	0%
Lomerizine, antidepressants, kampo	2	0.12%	0	0%	1	0.26%	1	1.47%	0	0%
Lomerizine, propranolol, kampo	2	0.12%	0	0%	2	0.52%	0	0%	0	0%
Lomerizine, anticonvulsants, antidepressants	1	0.06%	1	0.08%	0	0%	0	0%	0	0%
Lomerizine, anticonvulsants, kampo	3	0.18%	2	0.16%	1	0.26%	0	0%	0	0%
Lomerizine, anticonvulsants, antidepressants, kampo	1	0.06%	1	0.08%	0	0%	0	0%	0	0%

NSAIDs, non-steroidal anti-inflammatory drugs.

Any combination of single, combination, and/or triptans, as well as one that does not constitute over-prescription for a single drug, was over-prescribed (≥30 tablets/90 days) to 296 of 13,651 (2.17%) patients with headache (see supplementary Table 2). Among them, 70/296 (23.65%) patients received prophylactic treatment.

Study 2 (longitudinal study): over-prescription during the study period

Among 6,840,618 patients from the REZULT database, 408,183 (5.97%) were first diagnosed with headaches and this diagnosis persisted for at least two consecutive years during the studied period (July 2010 to April 2022). The mean ± SD age was 35.23 ± 10.71 years, and 271,846 (66.60%) patients were women. During the first 90 days after headache diagnosis, 3258/408,183 (0.80%) patients received >45 tablets of a single NSAIDs, 188/408,183 (0.05%) received >30 tablets of combination NSAIDs and 91/408,183 (0.02%) received >30 tablets of triptans. At 22–24 months after the initial visit, the number of patients over-prescribed acute medications (≥30 or 45 tablets/90 days) increased (7733/408,183 patients, 1.89%) (Figure 1a and Table 5). Throughout the two years, 37,617/408,183 (9.22%) patients were over-prescribed acute medications in 90 days at least once.

Figure 1.

Prescription pattern during the two years from the initial diagnosis. (a) Percentage of over-prescriptions with acute medications. Two years after the initial visit, the number of patients over-prescribed with acute medications (≥30 or 45 tablets/90 days) increased with time. During the two years, 37,617/408,183 (9.22%) patients were over-prescribed with many acute medications in 90 days on at least one occasion and (b) Percentage of prescriptions for prophylactic medication. During the two years, 29,313/408,183 (7.18%) patients were prescribed prophylactic medication on at least one occasion.

Table 5.

Changes in prescribing patterns over time.

Variables (n = 408,183)	1–3 months	4–6 months	7–9 months	10–12 months	13–15 months	16–18 months	19–21 months	22–24 months
At least one of follows	1.69%	1.18%	1.45%	1.49%	1.60%	1.72%	1.82%	1.89%
Single NSAIDs ≥45 tablets/90 days	0.80%	0.59%	0.75%	0.78%	0.86%	0.91%	0.96%	1.02%
Combination NSAIDs ≥30 tablets/90 days	0.05%	0.05%	0.05%	0.06%	0.06%	0.07%	0.07%	0.07%
Triptans ≥30 tablets/90 days	0.02%	0.02%	0.02%	0.04%	0.04%	0.05%	0.05%	0.06%
Overprescribed any combination of single, combination, and/or triptans as well as that do not constitute over-prescription for a single drug	0.83%	0.53%	0.64%	0.63%	0.66%	0.70%	0.76%	0.76%
Prophylactic treatment	1.32%	1.08%	1.27%	1.37%	1.44%	1.55%	1.70%	1.77%
Calcium-channel blockers (lomerizine)	0.11%	0.11%	0.12%	0.14%	0.16%	0.18%	0.19%	0.20%
Beta-blockers (propranolol)	0.03%	0.03%	0.03%	0.04%	0.04%	0.04%	0.05%	0.06%
Anticonvulsants (valproic acid)	0.11%	0.14%	0.19%	0.21%	0.24%	0.27%	0.29%	0.33%
Antidepressants (amitriptyline)	0.05%	0.05%	0.06%	0.08%	0.09%	0.10%	0.12%	0.11%
Calcium-channel blockers (verapamil)	0.03%	0.02%	0.02%	0.03%	0.03%	0.03%	0.03%	0.03%
Japanese herbal kampo medicine	1.04%	0.78%	0.93%	0.96%	0.97%	1.04%	1.13%	1.15%
One type	1.28%	1.04%	1.21%	1.29%	1.36%	1.46%	1.60%	1.66%
Two types	0.04%	0.04%	0.05%	0.07%	0.07%	0.08%	0.09%	0.09%
Three types	0.01%	0.01%	0.01%	0.01%	0.01%	0.01%	0.01%	0.01%
Four types	0%	0%	0%	0.001%	0.001%	0.001%	0.001%	0.001%

NSAIDs, non-steroidal anti-inflammatory drugs.

Prophylactic medication was initiated in the first 90 days in 5388/408,183 (1.32%) patients, which increased to 7208/408,183 (1.77%) patients after 2 years (Figure 1b). Throughout the 2 years, 29,313/408,183 (7.18%) patients with headaches were prescribed prophylactic medications on at least one occasion. The details of the prescription pattern for the 2 years from the initial diagnosis are shown in Table 5.

Among 408,183 patients, 37,617 (9.22%) were over-prescribed medications during the 2-year observational period. The mean ± SD age was 36.58 ± 10.71 years, and 25,039/37,617 (66.56%) patients were female. The mean ± SD duration of the first instance of over-prescription was 4.29 ± 2.02 terms of 90 days, conprising approximately 12–15 months (Figure 2a). The mean ± SD number of medical facilities at which doctors prescribed headache medication without over-prescription was 1.14 ± 0.45 (range 1–8) and those facilities that over-prescribed (1.52 ± 0.85, range 1–11) differed significantly (p < 0.001). Sub-analysis revealed that the number of patients over-prescribed with a single NSAIDs (≥45 tablets/90 days), combination NSAIDs, triptans, or any combination of single, combination and/or triptans, as well as that not constituting over-prescription for a single drug (≥30 tablets/90 days), was 19,820/408,183 (4.86%), 1,101/408,183 (0.27%), 842/408,183 (0.21%) and 20,108/408,183 (4.93%), respectively.

Figure 2.

Cox regression analysis for over-prescription during the two years from the initial diagnosis. (a) Trends in over-prescription in patients with a persisting headache diagnosis for two years from the initial diagnosis. (b) Patients prescribed prophylactic medication in the first three months were more likely to be over-prescribed. Sub-analyses for (c, d) single non-steroidal anti-inflammatory drugs (NSAIDs), (e, f) combination NSAIDs, (g, h) triptans and (i, j) any combination of single, combination and/or triptans, as well as any that do not constitute over-prescription for a single drug. x-axis, time; y-axis, proportion of over-prescription.

According to the sub-analysis, 3064/408,183 (0.75%) patients remained over-prescribed with a single NSAIDs (≥45 tablets/90 days) for two terms (4–6 months), 350/408,183 (0.09%) patients for four terms (10–12 months) and 18/408,183 (0.01%) patients for eight terms (22–24 months). Then, 286/408,183 (0.07%) patients remained over-prescribed with combination NSAIDs (≥30 tablets/90 days) for two terms (4–6 months), 59/408,183 (0.01%) patients for four terms (10–12 months) and 4/408,183 (0.001%) patients for eight terms (22–24 months). Also, 136/408,183 (0.03%) patients remained over-prescribed with triptans (≥30 tablets/90 days) for two terms (4–6 months), 15/408,183 (0.01%) patients for four terms (10–12 months) and no patients for eight terms (22–24 months). Finally, the number of patients who remained over-prescribed with any combination of single, combination and/or triptans, as well as that not constituting over-prescription for a single drug (≥30 tablets/90 days), was 877/408,183 (0.21%) for two terms (4–6 months), 11/408,183 (0.01%) for four terms (10–12 months) and zero for eight terms (22–24 months).

Cox regression analysis revealed that older age, more prescriptions of single, combination NSAIDs and triptans in the first three months, and prescription of prophylactic medication during the first three months were related to early over-prescription in the two years (all p < 0.001) (Table 6). Early use of prophylactic medication during the first three months (n = 5388/408,183 patients; 1.32%) was related to early over-prescription (Figure 2b). Sub-analyses for single NSAIDs, combination NSAIDs, triptans or any combination of single, combination and/or triptans, as well as that not constituting over-prescription for a single drug, showed similar trends (Figure 2c–h and Table 6). Log-minus-log plots confirmed that the assumption of proportional hazards was satisfied (see supplementary Figure 1).

Table 6.

Hazard ratio for overprescription using Cox regression.

Total	B	SE	Wald	p-value	HR	95% confidence interval (lower)	(upper)
Female	0.01	0.01	0.92	0.338	1.01	0.99	1.03
Age (years)	0.01	0.00	653.28	<0.001	1.01	1.01	1.01
Single NSAIDs during the first 90 days (tablets/90 days)	0.00	0.00	1987.22	<0.001	1.00	1.00	1.00
Combination NSAIDs during the first 90 days (tablets/90 days)	0.01	0.00	936.612	<0.001	1.01	1.01	1.01
Triptans during the first 90 days (tablets/90 days)	0.07	0.00	2304.32	<0.001	1.08	1.07	1.08
Presence of prophylactic treatment prescription during the first 90 days	0.73	0.03	522.83	<0.001	2.08	1.95	2.21
Single NSAIDs
Female	0.03	0.02	4.32	0.038	1.03	1.00	1.06
Age (years)	0.02	0.00	1040.69	<0.001	1.02	1.02	1.02
Single NSAIDs during the first 90 days (tablets/90 days)	0.00	0.00	2118.15	<0.001	1.00	1.00	1.00
Combination NSAIDs during the first 90 days (tablets/90 days)	0.01	0.00	28.95	<0.001	1.01	1.00	1.01
Triptans during the first 90 days (tablets/90 days)	0.03	0.01	66.16	<0.001	1.03	1.02	1.04
Presence of prophylactic treatment prescription during the first 90 days	0.81	0.04	345.29	<0.001	2.24	2.06	2.44
Combination NSAIDs
Female	0.44	0.07	38.96	<0.001	1.55	1.35	1.78
Age (years)	0.03	0.01	135.99	<0.001	1.03	1.03	1.04
Single NSAIDs during the first 90 days (tablets/90 days)	0.00	0.00	17.36	<0.001	1.00	1.00	1.00
Combination NSAIDs during the first 90 days (tablets/90 days)	0.02	0.00	1979.28	<0.001	1.02	1.02	1.02
Triptans during the first 90 days (tablets/90 days)	0.05	0.01	49.77	<0.001	1.05	1.04	1.07
Presence of prophylactic treatment prescription during the first 90 days	1.40	0.14	103.33	<0.001	4.05	3.09	5.31
Triptans
Female	0.29	0.08	13.80	<0.001	1.34	1.15	1.57
Age (years)	0.03	0.01	65.99	<0.001	1.03	1.02	1.03
Single NSAIDs during the first 90 days (tablets/90 days)	0.00	0.00	0.37	0.541	1.00	1.00	1.00
Combination NSAIDs during the first 90 days (tablets/90 days)	0.01	0.00	3.67	0.055	1.01	1.00	1.01
Triptans during the first 90 days (tablets/90 days)	0.13	0.00	2663.43	<0.001	1.14	1.13	1.14
Presence of prophylactic treatment prescription during the first 90 days	1.78	0.11	246.34	<0.001	5.91	4.73	7.37
Overprescribed any combination of single, combination, and/or triptans as well as that do not constitute over-prescription for a single drug
Female	–0.01	0.02	0.37	0.544	0.99	0.96	1.02
Age (years)	0.01	0.01	69.11	<0.001	1.01	1.00	1.01
Single NSAIDs during the first 90 days (tablets/90 days)	0.00	0.00	172.98	<0.001	1.00	1.00	1.00
Combination NSAIDs during the first 90 days (tablets/90 days)	0.01	0.01	17.91	<0.001	1.01	1.00	1.01
Triptans during the first 90 days (tablets/90 days)	0.05	0.00	519.71	<0.001	1.05	1.05	1.06
Presence of prophylactic treatment prescription during the first 90 days	0.60	0.05	174.35	<0.001	1.83	1.67	2.00

B, slope; HR, hazard ratio; NSAIDs, non-steroidal anti-inflammatory drugs.

For sensitivity analysis, we developed an AI-based model predicting the occurrence of over-prescriptions during the two years from the initial headache diagnosis based on the prescription amount in the initial 90 days. A gradient-boosting classifier was used. The confusion matrix is shown in Figure 3a. Sensitivity and specificity were 20.36% and 99.97%, respectively. The micro- and macro-average AUC of the ROC were 0.95 (95% confidence interval (CI) = 0.92–0.98) and 0.79 (95% CI = 0.70–0.88), respectively (Figure 3b). The important variables and their gains are shown in Figure 3c. Single NSAIDs, triptans, combination NSAIDs, Japanese herbal Kampo medicine, lomerizine, verapamil, propranolol, amitriptyline, valproic acid (tablets or packets/90 days for all) and sex were significant in that order. The learning curve and calibration plot are shown in Figure 3d,e. Of the survival prediction model using random survival forest, Harrell’s concordance index (22) was 0.80. In other words, this AI-based predictive model was highly accurate, and its important variables (Figure 3c) were similar to those that were identified as important in Cox regression analysis (Table 6).

Figure 3.

Performance of the artificial intelligence (AI)-based model to predict over-prescription. (a) Confusion matrix. (b) Micro- and macro-average area under the curve (AUC) of the receiver operating characteristic curve (ROC). (c) The important variables and their gains. (d) Learning curve and (e) Calibration plot.

Discussion

We used the big health insurance REZULT database to perform two investigations on prescription patterns of headache medications. First, in the cross-sectional study, among 13,651 patients diagnosed with headaches and treated with prescribed medication in 2020, only 1200/13,651 (8.79%) received prophylactic medication. The number of patients over-prescribed with at least one single NSAIDs, combination NSAIDs or triptans was 2262/13,651 (16.57%). Second, in the longitudinal study, we evaluated 408,183 patients diagnosed with headaches, which persisted for at least two consecutive years. Throughout these two years, 37,617/408,183 (9.22%) patients were over-prescribed acute medications in 90 days on at least one occasion, and 29,313/408,183 (7.18%) patients were prescribed prophylactic medications on at least one occasion. Patients with over-prescription were prescribed headache medications by multiple healthcare providers compared with those that were not over-prescribed. Notably, early use of prophylactic medication during the first three months (5388/408,183 patients, 1.32%) was related to early acute medication over-prescription. Younger age, more prescriptions of single, combination NSAIDs and triptans in the first three months were also related factors.

Investigation of foreign health insurance claims databases

In the present study, we describe recent reports on triptan over-prescription because, unlike NSAIDs, triptans cannot be prescribed without a diagnosis of migraine in Japan. Analysis of the Dutch Health Care Insurance Board Database of 2005, comprising data for 46% of the total population, revealed that triptans were used by 1.3%, and 10.4% of these were suspected overusers. They also found that almotriptan tended to be overused, whereas sumatriptan was not (10). An investigation of the French General Health Insurance System in 2010 and 2011 revealed that 2243 of 99,450 patients (2.3%) prescribed with triptans could be overusers. Older age and the use of prophylactic medications, benzodiazepines and antidepressants were risk factors for triptan overuse (15). A 2012 study conducted in two provinces in Italy showed that approximately 10% of triptan users were frequent users (≥10 tablets/month). Approximately two-thirds of the frequent users maintained their frequent triptan use for an additional three months (14). An Austrian research of 5,918,487 people found that 0.56% were triptan users and 6.0% were overusers. Overusers were older than non-overusers (12).

Compared with results of reports worldwide, our results showed similar trends. Triptans were prescribed to 1710 of 13,651 (12.53%) patients with headaches; 106/1710 (6.20%) were suspected overusers. Therefore, awareness of the appropriate use of triptans and prevention of overuse/over-prescription are needed worldwide.

Previous Japanese studies using health insurance claims data

Meyers et al. (16) investigated patients with migraines and their treatment patterns using data from the Japan Medical Data Centre database. In three years, among 16,433 patients with migraines, 9873 (60.1%) received only acute medication, 3022 (18.4%) received prophylactic medications and 3548 (21.6%) were not prescribed any medication. After 61.2 ± 65.3 days of initial prophylactic treatment, 62.2% patients discontinued it; among them, 15.2% reinitiated the initial treatment and 7.0% switched the treatment regimen within one year. Therefore, the use rate of prophylactic medications was low, with a high discontinuation rate following a brief treatment period. Additionally, we observed lower prevalence (1.4%) of migraines than previous prevalence studies (23). This underestimation may be because 69.4% of patients with migraines had never consulted a physician for headaches, and only 11.6% of patients were aware that their headaches were migraines (23). Calcium channel blockers and anticonvulsants were the most common prophylactic medications reported in the database. In the present study, although the prevalence of migraine could not be directly compared with the study by Meyers et al. (16) because of the different sampling methods, the lower prevalence of migraine compared with that in previous epidemiological studies (23,24) could be for the same reasons; patients with migraine may have refrained from seeing a doctor (2) or those who did may not have been prescribed prophylactic medications (25). Furthermore, Japanese herbal Kampo medicines (4) were most commonly prescribed; however, besides that, the trends appeared to be similar. By contrast to the findings of Meyer et al. (16), in the present study, the rate of prescriptions for preventive medications increased over time (Figure 1 and Table 5). Our patients retained their headache diagnosis for two years; therefore, they may have been treated longer than patients in Mayer’s study, which may have caused titration of acute and prophylactic medications.

Sakai et al. (17) investigated anonymised online survey data coupled with medical claims data of individuals aged 19–74 years. Migraine prevalence was 3.2% (691/21,480) and was the highest among 30–39-year-olds. Among them, 81.0% (560/691) of patients did not consult a doctor and 57.8% (362/691) used OTC drugs only. However, only 6.1% (38/691) of patients used prescription medicines. The findings of an epidemiological study conducted 20 years ago (23), as well as those of another study (17) suggested that, even with the advent of triptans and CGRP-related drugs and the paradigm shift that has occurred, a lack of appropriate consultation behaviour and use of prophylactic treatment persist (26).

Hirata et al. (18) analysed prevalence and burden according to headache subtypes in 21,480 patients. They noted that the irrational behaviour of refraining from seeing a doctor was observed in all headache subtypes. By combining an epidemiological and a questionnaire survey, the researchers reported that patients with headaches are still burdened by the disease, refrain from seeing their doctors, and do not receive prophylactic treatment. The few patients for whom prophylactic treatment was prescribed was similarly noted in our study.

Timing of initial consultation

Generally, prophylactic medications should decrease the frequency and severity of headaches and reduce the frequency of acute care medications (3). However, in our longitudinal study, the group receiving prophylactic treatment within the first 90 days experienced over-prescription of acute medications earlier than the group that did not receive prophylactic medications. This suggests that patients who potentially require follow-up for >2 years may have had their first hospital consultation when the migraine had already progressed and become intractable.

It has been proposed that migraine is a progressive disease (27). Episodic migraine is estimated to progress to chronic migraine at a rate of 2.5% per year and may be underestimated as a result of the diagnostic criteria for chronic migraine, which remains under debate (28). Factors associated with migraine progression include a high frequency of migraine attacks, medication overuse, comorbid pain syndromes and obesity (27). We observed a similar trend: patients with prescriptions of combination NSAIDs and triptans in the first three months were related to early over-prescribing of acute medications.

Hirata et al. (18) found that 14–38% of patients with headaches did not visit a hospital because of the effectiveness of OTC medicines. By contrast, some patients first visited a doctor because OTC had become ineffective (48/691, 6.9%) or their headaches had increased in frequency (55/691, 8.0%). Prophylactic treatment can be terminated relatively early after migraine remission (3,4,17). Prophylactic treatment is not well-performed in Japan (1,25). We hypothesised that patients who encourage their physicians to prescribe prophylactic treatments and those for whom early prophylactic treatment is prescribed may have already begun to develop chronicity from the initial consultation. If so, further educational efforts (2,29) are needed, and patients should consult doctors during the period when they have relatively mild episodic migraine.

A similar trend is suspected in other countries. One study showed that approximately one in three migraine patients take acute prescription medications for a migraine attack, whereas less than one in five patients take preventive prescription medications (9). Additionally, misdiagnosis and delayed diagnosis of chronic migraine are problematic (30). Although the situation differs from country to country regarding accessibility to medical care, differences in advertising methods and different drugs available through insurance, similar trends may be applicable worldwide, given that reports similar to Japan's have been published from various countries (31,32). Large-scale longitudinal surveys are expected in the future.

Number of medical facilities visited

In clinical practice, patient education is important; however, obtaining a long-term perspective is difficult, and patients often drop out and avail of multiple medical services (33). Our results showed that patients were over-prescribed headache medications, and these prescriptions may have been provided by multiple healthcare providers. This may cause a lack of trust in doctors and healthcare facilities. In the future, it will be necessary to build trust between doctors and patients and plan treatment based on shared decision-making (34).

Limitations

This study had some limitations. First, it has a sampling bias. In cross-sectional Study 1, we selected a 90-day period during which the highest number of acute medications (tablet number) were prescribed. Additionally, because it was impossible to determine the number of days for which the acute medications were actually used over the study period, ‘number of tablets per 90 days’ was counted instead of the ‘number of days per month when acute medications were used’. Although this calculation method was based on previously established methodology (10), accurate recording using headache diaries or other methods is desirable. In longitudinal Study 2, we selected patients who continued to be labelled with headaches for two years after their first diagnosis. This may have resulted in a large number of patients with relatively severe headaches. Additionally, severely ill patients may have been included when their migraines had already begun to progress. Moreover, because NSAIDs were not only prescribed for headaches, patients who were regularly prescribed analgesics for other types of chronic pain, including back pain, could not be excluded. Therefore, for studying migraines specifically, it may be better to focus only on triptans in future studies.

Second, a retrospective analysis of administrative claims data for billing was performed. Patients were identified only based on information available in the database, and we could not determine migraine severity, the actual number of medications taken per day, and OTC medicine use. Some migraine treatments have multiple indications (e.g. anticonvulsants and antidepressants) and patients may have been incorrectly considered to have received treatment for migraine, when it was actually intended for another comorbid condition.

Third, we could not track those patients who used insurance systems not included in the REZULT database. Fourth, it was impossible to identify patients with only migraine because all types of headaches were put under one category; reanalysis of only migraines will be further considered. The apparent prevalence of headache disorders may have increased because the disease was named for the sake of insurance claims, and the disease labels were kept or forgotten to be removed. Fifth, 21.6% of patients with migraines were not prescribed medications in a previous health insurance-based study (16) and such patients could not be evaluated. Sixth, it is undeniable that a large percentage of medication-overuse headaches attributed to multiple drug classes not individually overused (ICHD-3 code 8.2.6) patients may be overcounted as a result of the method of counting the number of tablets. Finally, we will continue to track how this prescribing pattern changes with headache awareness (2) and the introduction of new drugs, including CGRP-related drugs and lasmiditan.

Conclusions

We used REZULT, a big health insurance database, to investigate prescription patterns in headaches and found that only 8.79% (1200/13,651) of patients diagnosed with headaches received prophylactic medication. Among patients who received acute medications (NSAIDs or triptans), 16.57% (2262/13,651) were over-prescribed. In the two-year longitudinal study, 10.76% (37,617/408,183) of patients were over-prescribed acute medications and 29,313/408,183 (7.18%) patients with headaches were prescribed prophylactic medications in 90 days on at least one occasion. Early use of prophylactic medication during the first three months, younger age, more prescriptions of single, combination NSAIDs and triptans in the first three months were related to early over-prescription.

Clinical implications

Among 13,651 patients with headache and with prescriptions, 16.6% were over-prescribed acute medications (NSAIDs or triptans).

In 408,183 patients with headaches for two consecutive years, 10.8% were over-prescribed acute medications during 90 days on at least one occasion.

Throughout the two-year period, 7.2% of the 408,183 patients with headache were prescribed prophylactic medications on at least one occasion.

Supplemental Material

sj-xlsx-1-cep-10.1177_03331024231226177 - Supplemental material for Treatment patterns and characteristics of headache in patients in Japan: A retrospective cross-sectional and longitudinal analysis of health insurance claims data

Supplemental material, sj-xlsx-1-cep-10.1177_03331024231226177 for Treatment patterns and characteristics of headache in patients in Japan: A retrospective cross-sectional and longitudinal analysis of health insurance claims data by Masahito Katsuki, Yasuhiko Matsumori, Taisuke Ichihara, Yuya Yamada, Shin Kawamura, Kenta Kashiwagi, Akihito Koh, Tetsuya Goto, Kazuma Kaneko, Naomichi Wada and Fuminori Yamagishi in Cephalalgia

Supplemental Material

sj-pdf-2-cep-10.1177_03331024231226177 - Supplemental material for Treatment patterns and characteristics of headache in patients in Japan: A retrospective cross-sectional and longitudinal analysis of health insurance claims data

Supplemental material, sj-pdf-2-cep-10.1177_03331024231226177 for Treatment patterns and characteristics of headache in patients in Japan: A retrospective cross-sectional and longitudinal analysis of health insurance claims data by Masahito Katsuki, Yasuhiko Matsumori, Taisuke Ichihara, Yuya Yamada, Shin Kawamura, Kenta Kashiwagi, Akihito Koh, Tetsuya Goto, Kazuma Kaneko, Naomichi Wada and Fuminori Yamagishi in Cephalalgia

Footnotes

Acknowledgements

We are thankful for the help provided by the medical staff.

Data availability

The datasets generated and/or analysed during the current study are available from the corresponding author upon reasonable request.

Declaration of conflicting interests

Masahito Katsuki received personal consultancy fees from Amgen Astellas BioPharma K.K., Daiichi Sankyo Company, Limited, Eli Lilly Japan K.K. and Otsuka Pharmaceutical Co., Ltd. during the conduct of the study.

Yasuhiko Matsumori received personal consultancy fees from Amgen Astellas BioPharma K.K., Daiichi Sankyo Company, Limited, Eli Lilly Japan K.K. and Otsuka Pharmaceutical Co., Ltd. during the conduct of the study.

Taisuke Ichihara and Yuya Yamada are employees of Japan System Techniques Co. Ltd.

The other authors report no conflicts of interest concerning the materials or methods used in this study or the findings presented in this paper. This article and contents of this study were not published nor presented previously.

Funding

The study was funded by Bio-Medicines, Japan Drug Development & Medical Affairs at Eli Lilly Japan K.K.

ORCID iD

Masahito Katsuki

Supplemental material

Supplemental material for this article is available online.

References

Matsumori

Ueda

Komori

, et al. Burden of Migraine in Japan: Results of the ObserVational Survey of the Epidemiology, tReatment, and Care Of MigrainE (OVERCOME [Japan]) Study. Neurol Ther 2022; 11: 205–222.

Katsuki

Matsumori

Kawahara

, et al. Headache education by leaflets distribution during COVID-19 vaccination and school-based on-demand E-learning-Itoigawa Geopark Headache Awareness Campaign. Headache 2023; 63: in press.

Headache Clinical Practice Guideline Development Committee C. [Clinical Practice Guideline for Headache Disorders 2021] (Japanese). Tokyo: Igaku-Shoin, 2021.

Katsuki

Kashiwagi

Kawamura

, et al. The efficacy of Japanese herbal kampo medicine as an acute and prophylactic medication to treat chronic daily headache and medication overuse headache: -Single arm retrospective study. Cureus 2022; 14: e25419.

Katsuki

Kawahara

Matsumori

, et al. Questionnaire-based survey during COVID-19 vaccination on the prevalence of elderly’s migraine, chronic daily headache, and medication-overuse headache in one Japanese city – Itoigawa Hisui Study. J Clin Med 2022; 11: 4707.

Roxas

Quiles

Wang

S-J.

Delivery of care for migraine in the Asian Oceanian region: a cross-sectional study. Cephalalgia 2021; 41: 1348–1358.

Woldeamanuel

YW.

Headache in resource-limited settings. Curr Pain Headache Rep 2017; 21: 51.

Katsarava

Mania

Lampl

, et al. Poor medical care for people with migraine in Europe – evidence from the Eurolight study. J Headache Pain 2018; 19: 10.

Lipton

Nicholson

Reed

, et al. Diagnosis, consultation, treatment, and impact of migraine in the US: Results of the OVERCOME (US) study. Headache 2022; 62: 122–140.

10.

Dekker

Wiendels

De Valk

, et al. Triptan overuse in the Dutch general population: a nationwide pharmaco-epidemiology database analysis in 6.7 million people. Cephalalgia 2011; 31: 943–952.

11.

Huber

Agosti

Näpflin

, et al. Treatment patterns in patients using triptan and prophylactic medication: an analysis of clinical practice prior to the introduction of CGRP antagonists. J Pain Res 2019; 12: 2211–2221.

12.

Zebenholzer

Gall

Wöber

Use and overuse of triptans in Austria – a survey based on nationwide healthcare claims data. J Headache Pain 19. Epub ahead of print 1 December 2018. DOI: 10.1186/s10194-018-0864-0.

13.

Scuteri

Adornetto

Rombolà

, et al. Pattern of triptans use: a retrospective prescription study in Calabria, Italy. Neural Regen Res 2020; 15: 1340–1343.

14.

Da Cas

Nigro

Terrazzino

, et al. Triptan use in Italy: insights from administrative databases. Cephalalgia 2015; 35: 619–626.

15.

Braunstein

Donnet

Pradel

, et al. Triptans use and overuse: a pharmacoepidemiology study from the French health insurance system database covering 4.1 million people. Cephalalgia 2015; 35: 1172–1180.

16.

Meyers

Davis

Lenz

, et al. Treatment patterns and characteristics of patients with migraine in Japan: a retrospective analysis of health insurance claims data. Cephalalgia 2019; 39: 1518–1534.

17.

Sakai

Hirata

Igarashi

, et al. A study to investigate the prevalence of headache disorders and migraine among people registered in a health insurance association in Japan. J Headache Pain 2022; 23: 70.

18.

Hirata

Sano

Kondo

, et al. Clinical characteristics, medication use, and impact of primary headache on daily activities: an observational study using linked online survey and medical claims data in Japan. BMC Neurol 2023; 23: 80.

19.

Katsuki

Shimazu

Kikui

, et al. Developing an artificial intelligence-based headache diagnostic model and its utility for non-specialists’ diagnostic accuracy. Cephalalgia 2023; 43: 033310242311569.

20.

Katsuki

Matsumori

Kawamura

, et al. Developing an artificial intelligence-based diagnostic model of headaches from a dataset of clinic patients’ records. Headache 2023; 63: 1097–1108.

21.

Luo

Phung

Tran

, et al. Guidelines for developing and reporting machine learning predictive models in biomedical research: a multidisciplinary view. J Med Internet Res 2016; 18: e323.

22.

Harrell

Lee

Mark

DB.

Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996; 15: 361–87.

23.

Sakai

Igarashi

Prevalence of migraine in Japan: a nationwide survey. Cephalalgia 1997; 17: 15–22.

24.

Katsuki

Yamagishi

Matsumori

, et al. Questionnaire-based survey on the prevalence of medication-overuse headache in Japanese one city-Itoigawa study. Neurol Sci 2022; 43: 3811–3822.

25.

Takeshima

Ueda

Komori

, et al. Potential unmet needs in acute treatment of migraine in Japan: Results of the OVERCOME (Japan) study. Adv Ther 2022; 39: 5176–5190.

26.

Ueda

Lombard

, et al. Real-world treatment patterns and patient-reported outcomes in episodic and chronic migraine in Japan: analysis of data from the Adelphi migraine disease specific programme. J Headache Pain 2019; 20. Epub ahead of print 7 June 2019. DOI: 10.1186/s10194-019-1012-1.

27.

Rattanawong

Rapoport

Srikiatkhachorn

Neurobiology of migraine progression. Neurobiol Pain 2022; 12: 100094.

28.

Ishii

Schwedt

Dumkrieger

, et al. Chronic versus episodic migraine: the 15-day threshold does not adequately reflect substantial differences in disability across the full spectrum of headache frequency. Headache 2021; 61: 992–1003.

29.

Carlsen

Westergaard

Bisgaard

, et al. National awareness campaign to prevent medication-overuse headache in Denmark. Cephalalgia 2018; 38: 1316–1325.

30.

Pavlovic

Silberstein

, et al. Development of a claims-based algorithm to identify potentially undiagnosed chronic migraine patients. Cephalalgia 2019; 39: 465–476.

31.

Wells-Gatnik

Ambat

FDF

Martelletti

Hesitancies in primary headaches treatments. Expert Rev Neurother 2022; 22: 733–735.

32.

Smith

Nicholson

Banks

JW.

Migraine education improves quality of life in a primary care setting. Headache 2010; 50: 600–12.

33.

Lau

JTF

Cheung

JCK

, et al. Studies on common illnesses and medical care utilization patterns of adolescents in Hong Kong. J Adolesc Health 2000; 27: 443–452.

34.

Urtecho

Wagner

Wang

, et al. A qualitative evidence synthesis of patient perspectives on migraine treatment features and outcomes. Headache 2023; 63: 185–201.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.18 MB

0.01 MB