2015 年 38 巻 10 号 p. 1638-1644
There have been concerns that oseltamivir causes neuropsychiatric adverse events (NPAEs). We analyzed the association of age and gender with NPAEs in patients treated with oseltamivir using a logistic regression model. NPAE data were obtained from the U.S. Food and Drug Administration Adverse Event Reporting System (2004 to 2013). The lower limit of the reporting odds ratio (ROR) 95% confidence interval (CI) of “abnormal behavior” in Japan, Singapore, and Taiwan was ≥1. The effects of the interaction terms for oseltamivir in male patients aged 10–19 years were statistically significant. The adjusted ROR of “abnormal behavior” was 96.4 (95% CI, 77.5–119.9) in male patients aged 10–19 years treated with osletamivir. In female patients, the results of the likelihood ratio test for “abnormal behavior” were not statistically significant. The adjusted NPAE RORs were increased in male and female patients under the age of 20 years. Oseltamivir use could be associated with “abnormal behavior” in males aged 10–19 years. After considering the causality restraints of the current analysis, further epidemiological studies are recommended.
Oseltamivir (Tamiflu) is a neuraminidase inhibitor commonly used as an anti-viral in the treatment of influenza.1) Administrative authorities have issued warnings regarding the risk of neuropsychiatric adverse events (NPAEs) following oseltamivir use. In March 2007, the Japanese Ministry of Health, Labour and Welfare (MHLW) warned against the use of oseltamivir in children aged 10–19 years because it could cause abnormal behavior. The U.S. Food and Drug Administration (FDA) added a warning to the oseltamivir label in 2006 to draw attention to the risk of NPAEs. Recent reports suggest that NPAEs may occur after oseltamivir administration.2–9)
In prophylaxis studies, oseltamivir increased the risk of psychiatric adverse events during the combined “on-treatment” and “off-treatment” periods (risk difference 1.06, 0.07 to 2.76%).2) A cohort study indicated a slightly increased risk of delirium with oseltamivir use in patients under 18 years (hazard ratio: 1.51, 95% confidence interval (CI) 0.95–2.40, p=0.084).3) Data from the FDA Adverse Event Reporting System (FAERS) indicates that certain NPAEs are disproportionally reported after oseltamivir use.4) Several cases of oseltamivir-induced NPAEs have been reported, including mania in a Chinese patient5) and depressive episodes in Korean patients.6)
However, a number of epidemiological studies, controlled studies, and commentaries indicate that there is no relationship between oseltamivir use and NPAEs.10–13) Significant controversy exists regarding whether NPAEs result from oseltamivir use. Given this conflicting data, the optimal dosing for patients who require oseltamivir therapy is uncertain.
FAERS is a passive reporting system covering several million case reports on adverse events that is used for pharmacovigilance.14)
Data from the FAERS indicate that certain NPAEs are disproportionately reported after oseltamivir use, especially in Japan.4) To our knowledge, reports from Asian countries other than Japan were not evaluated in the FAERS database. Furthermore, the effect of age and gender on the relationship between oseltamivir and NPAEs has not yet been evaluated using the FAERS database.
Owing to the risk of reporting bias, there are legitimate reasons to doubt the stated benefits and/or risks of FAERS data.14) Despite the limitations inherent to spontaneous reporting, these databases are still the primary tool used for pharmacovigilance, because they reflect the realities of clinical practice.15,16) The pharmacovigilance index, termed the reporting odds ratio (ROR), was developed to detect drug-associated adverse events in the spontaneous reporting database, and is not applicable for evaluating the comparative strength of causalities.17) However, the ROR is applicable to a logistic regression model, which allows for the control of covariates.15,17) We evaluated the possible relationship between oseltamivir use and NPAEs from data available in the FAERS databases using the ROR and logistic regression analysis.
Adverse events recorded from January 2004 to March 2013 in the FAERS database were obtained from the FDA website (www.fda.gov). The FAERS structure complies with the International Conference on Harmonization and the international safety reporting guidelines, ICH E2B. The adverse events are coded according to the terminology preferred by the Medical Dictionary for Regulatory Activities (MedDRA).18) DrugBank, a reliable drug database, was utilized as a dictionary for the batch conversion and compilation of drug names.19)
In instances of duplicate entries, we adopted the most recent case number in order to identify duplicate reports of the same patient that come from different reporting sources and excluded these from the analysis as is described in the downloadable file ‘Asc_nts.doc’ from the FAERS database website. Patients with NPAEs were grouped by gender, and each gender group was stratified by age: <10 years, 10–19 years, ≥20 years. We extracted those reports that contained the preferred term (PT) “abnormal behavior” from the MedDRA. To further evaluate the effect of oseltamivir in NPAEs, we selected 57 additional PTs related to NPAEs described in the Japanese package insert of oseltamivir (Table 1).
HLTa) | PTb) | PT code | Total | Casesc) | % | RORd) | (95%CIe)) |
---|---|---|---|---|---|---|---|
Total | 222029 | 1347 | |||||
Disturbances in consciousness NECf) | Altered state of consciousness | 10001854 | 3306 | 35 | 2.6 | 5.2 | (3.7–7.2) |
Depressed level of consciousness | 10012373 | 10200 | 86 | 6.4 | 4.1 | (3.3–5.1) | |
Loss of consciousness | 10024855 | 28249 | 92 | 6.8 | 1.6 | (1.3–1.9) | |
Stupor | 10042264 | 939 | 1 | 0.1 | —† | ||
Breath holding | 10006322 | 29 | 0 | — | |||
Staring | 10041953 | 647 | 6 | 0.4 | 4.5 | (2.0–10.0) | |
Regressive behavior | 10054720 | 76 | 1 | 0.1 | —† | ||
Abnormal behavior | 10061422 | 11602 | 489 | 36.3 | 23.5 | (21.3–25.8) | |
Sexually inappropriate behavior | 10070240 | 41 | 0 | — | |||
Deliria | Delirium | 10012218 | 6283 | 164 | 12.2 | 13.3 | (11.4–15.6) |
Delirium tremens | 10012225 | 109 | 0 | — | |||
Delirium febrile | 10059267 | 27 | 8 | 0.6 | 202.4 | (88.5–462.5) | |
Perception disturbances | Hallucination | 10019063 | 12373 | 286 | 21.2 | 12.0 | (10.7–13.6) |
Hallucination, auditory | 10019070 | 3119 | 37 | 2.7 | 5.8 | (4.2–8.0) | |
Hallucination, gustatory | 10019071 | 9 | 0 | — | |||
Hallucination, olfactory | 10019072 | 81 | 0 | — | |||
Hallucination, visual | 10019075 | 3921 | 49 | 3.6 | 6.1 | (4.6–8.1) | |
Hallucination, tactile | 10019074 | 119 | 0 | — | |||
Hallucinations, mixed | 10019079 | 842 | 13 | 1.0 | 7.5 | (4.4–13.0) | |
Somatic hallucination | 10062684 | 20 | 0 | — | |||
Hallucination, synaesthetic | 10062824 | 7 | 0 | — | |||
Delusional perception | 10012258 | 90 | 1 | 0.1 | —† | ||
Narcolepsy and associated conditions | Hypnagogic hallucination | 10020927 | 40 | 1 | 0.1 | —† | |
Hypnopompic hallucination | 10020928 | 15 | 0 | — | |||
Delusional symptoms | Delusion | 10012239 | 2973 | 48 | 3.6 | 7.9 | (6.0–10.6) |
Delusion of reference | 10012244 | 43 | 0 | — | |||
Delusion of replacement | 10012245 | 24 | 0 | — | |||
Delusions, mixed | 10012263 | 2 | 0 | — | |||
Erotomanic delusion | 10015134 | 4 | 0 | — | |||
Jealous delusion | 10023164 | 49 | 0 | — | |||
Persecutory delusion | 10034702 | 537 | 2 | 0.1 | —† | ||
Somatic delusion | 10041317 | 76 | 0 | — | |||
Depressive delusion | 10063033 | 15 | 0 | — | |||
Delusion of grandeur | 10012241 | 120 | 0 | — | |||
Seizures and seizure disorders NECf) | Convulsion | 10010904 | 33234 | 174 | 12.9 | 2.6 | (2.2–3.0) |
Convulsion neonatal | 10010911 | 79 | 0 | — | |||
Drug withdrawal convulsions | 10013752 | 164 | 0 | — | |||
Tonic convulsion | 10043994 | 311 | 2 | 0.1 | —† | ||
Convulsion in childhood | 10052391 | 4 | 0 | — | |||
Clonic convulsion | 10053398 | 213 | 1 | 0.1 | —† | ||
Convulsions local | 10010920 | 36 | 0 | — | |||
Febrile convulsion | 10016284 | 295 | 6 | 0.4 | 10.0 | (4.4–22.4) | |
Anxiety symptoms | Agitation | 10001497 | 17046 | 84 | 6.2 | 2.4 | (1.9–3.0) |
Tremor (excl congenital) | Tremor | 10044565 | 30757 | 52 | 3.9 | 0.8 | (0.6–1.1) |
Essential tremor | 10015496 | 143 | 1 | 0.1 | —† | ||
Intention tremor | 10022520 | 159 | 1 | 0.1 | —† | ||
Orthostatic tremor | 10069917 | 7 | 0 | — | |||
Resting tremor | 10071390 | 33 | 0 | — | |||
Psychogenic tremor | 10072377 | 0 | 0 | — | |||
Action tremor | 10072413 | 1 | 0 | — | |||
Postural tremor | 10073211 | 1 | 0 | — | |||
Parasomnias | Nightmare | 10029412 | 6557 | 58 | 4.3 | 4.3 | (3.3–5.6) |
Abnormal dreams | 10000125 | 7066 | 13 | 1.0 | 0.9 | (0.5–1.5) | |
Neurological signs and symptoms NECf) | Dizziness | 10013573 | 79088 | 90 | 6.7 | 0.5 | (0.4–0.7) |
Dizziness exertional | 10013576 | 59 | 0 | — | |||
Dizziness postural | 10013578 | 1145 | 3 | 0.2 | 1.3 | (0.4–3.9) | |
Apocrine and eccrine gland disorders | Night sweats | 10029410 | 4285 | 8 | 0.6 | 0.9 | (0.4–1.8) |
a) HLT: High level terms in MedDRA. b) PT: Preferred terms in MedDRA. c) Patients who reported adverse events taking oseltamivir. d) ROR: Reporting odds ratio. e) CI: Confidence interval. f) NEC: Not elsewhere classified. † Number of cases ≤2.
Using established pharmacovigilance indices, “cases” were defined as patients who reported NPAEs, while “non-cases” consisted of patients who reported other side effects.15) The ROR is the ratio of the odds of reporting an adverse event versus all other events for oseltamivir compared to the odds of reporting an adverse event for all other drugs present in the database.14,15,20) RORs were expressed as point estimates with a 95% CI. A signal is considered an event when the lower limit of the ROR 95% CI is ≥1,20) and at least 3 cases were required to define the signal. We refined the signal with a dedicated correction to detect possible confounding factors present in the database. The RORs were adjusted for age in the male and female subset groups using logistic regression analysis. To construct the logistic model, we included the terms for reporting year, oseltamivir, age-stratified group, and the co-occurrence of oseltamivir use within the age-stratified groups. The logistic model used to calculate the adjusted ROR was as follows:
The FAERS database contains 4746890 reports from January 2004 to March 2013. After excluding duplicates according to the FDA recommendation, there were 3522995 reports that were usable. We excluded a further 1265093 reports because the age and gender were not indicated and finally analyzed 2257902. The total number of reports citing NPAEs and oseltamivir were 222029 and 4696, respectively. In 1347 cases, the NPAEs corresponded to the 57 PTs associated with oseltamivir use.
The data stratified by age and gender are summarized in Table 2. “Abnormal behavior” was primarily reported in people under 20 years of age. The RORs (95%CI) of “abnormal behavior” in Japan, Singapore, and Taiwan were 30.6 (27.0–34.7), 94.4 (18.3–486.8), and 15.3 (4.5–52.5), and those of “NPAEs” were 3.6 (3.3–3.9), 11.1 (5.3–23.4), and 1.3 (0.6–3.1), respectively (Table 3).
Total | Casesa) | RORb) | (95%CIc)) | |
---|---|---|---|---|
Abnormal behavior | 11602 | 489 | 23.7 | (21.6–26.1) |
Male | 6396 | 335 | 36.9 | (32.8–41.5) |
0–9 years | 1434 | 167 | 73.4 | (61.4–87.7) |
10–19 years | 1239 | 126 | 89.4 | (72.3–110.4) |
≥20 years | 3723 | 42 | 7.5 | (5.5–10.2) |
Female | 5206 | 154 | 12.6 | (10.7–14.8) |
0–9 years | 623 | 79 | 48.4 | (37.8–62.0) |
10–19 years | 618 | 27 | 22.7 | (15.2–33.6) |
≥20 years | 3965 | 48 | 5.1 | (3.8–6.7) |
Neuropsychiatric adverse events (NPAEs) | 222029 | 1347 | 3.8 | (3.5–4.0) |
Male | 88880 | 784 | 5.3 | (4.9–5.8) |
0–9 years | 5449 | 327 | 10.5 | (8.9–12.3) |
10–19 years | 6148 | 238 | 13.3 | (10.9–16.2) |
≥20 years | 77283 | 219 | 2.2 | (1.9–2.6) |
Female | 133149 | 563 | 2.6 | (2.4–2.9) |
0–9 years | 3614 | 186 | 8.1 | (6.6–9.9) |
10–19 years | 5877 | 90 | 4.9 | (3.8–6.4) |
≥20 years | 123658 | 287 | 1.6 | (1.4–1.9) |
a) Patients who reported adverse events taking oseltamivir. b) ROR: Reporting odds ratio. c) CI: Confidence interval.
RORa) | (95%CIb)) | Casesc) | Reporting number | |||
---|---|---|---|---|---|---|
n | (%) | Oseltamivir | Adverse events | |||
Abnormal behavior | ||||||
FAERS | 22.2 | (20.4–24.3) | 551 | (8.5) | 6508 | 15113 |
China | —† | 1 | (1.6) | 61 | 21 | |
Japan | 30.6 | (27.0–34.7) | 424 | (17.5) | 2421 | 1220 |
Korea | —† | 2 | (1.9) | 104 | 20 | |
Singapore | 94.4 | (18.3–486.8) | 3 | (9.1) | 33 | 6 |
Taiwan | 15.3 | (4.5–52.5) | 3 | (5.5) | 55 | 27 |
Neuropsychiatric adverse events (NPAEs) | ||||||
FAERS | 3.3 | (3.1–3.5) | 1528 | (23.5) | 6508 | 299707 |
China | 0.6 | (0.2–1.9) | 3 | (4.9) | 61 | 1065 |
Japan | 3.6 | (3.3–3.9) | 707 | (29.2) | 2421 | 12731 |
Korea | 0.4 | (0.2–1.2) | 4 | (3.8) | 104 | 683 |
Singapore | 11.1 | (5.3–23.4) | 11 | (33.3) | 33 | 133 |
Taiwan | 1.3 | (0.6–3.1) | 6 | (10.9) | 55 | 549 |
a) ROR: Reporting odds ratio. b) CI: Confidence interval. c) Patients who reported adverse events taking oseltamivir. † Number of cases ≤2.
The number of cases and the ROR (95%CI) for each PT are summarized in Table 1. The number of cases of “abnormal behavior,” “hallucination,” “convulsion,” and “delirium” were 489 (36.3%), 286 (21.2%), 174 (12.9%), and 164 (12.2%), respectively. The RORs (95%CI) for “abnormal behavior,” “hallucination,” “convulsion,” and “delirium” were 23.5 (21.3–25.8), 12.0 (10.7–13.6), 2.6 (2.2–3.0), and 13.3 (11.4–15.6), respectively. If the patient used oseltamivir, the lower limit of the ROR 95%CI for “abnormal behavior” was ≥1.
The adjusted RORs for “abnormal behavior” and “NPAEs” stratified by age and gender are summarized in Table 4. A likelihood ratio test was used to evaluate the effect of adding the term of oseltamivir* 10–19 years for abnormal behavior in male patients. Adding this interaction term had a statistically significant effect (p=0.0006). The adjusted ROR (95%CI) for “abnormal behavior” when considering oseltamivir* 10–19 years was 96.4 (77.5–119.9). When considering NPAEs, the likelihood ratio tests for oseltamivir* 0–9 years and oseltamivir* 10–19 years were also statistically significant (p<0.0001). The adjusted RORs (95%CI) for oseltamivir* 0–9 years and oseltamivir* 10–19 years were 10.7 (9.2–12.5) and 13.4 (11.0–16.3), respectively.
Abnormal behavior | Neuropsychiatric adverse events (NPAEs) | |||||
---|---|---|---|---|---|---|
Likelihood ratio test | Adjusted RORa) | (95%CIb)) | Likelihood ratio test | Adjusted ROR | (95%CI) | |
Male subgroup (N=887157) | ||||||
Oseltamivir | <0.0001* | 8.3 | (6.1–11.4) | <0.0001* | 2.3 | (2.0–2.6) |
Reporting year | <0.0001* | 0.9 | (0.9–0.9) | <0.0001* | 0.9 | (0.9–0.9) |
AGE | ||||||
0–9 years | <0.0001* | 8.7 | (8.2–9.3) | <0.0001* | 1.8 | (1.7–1.8) |
10–19 years | <0.0001* | 6.1 | (5.7–6.5) | <0.0001* | 1.6 | (1.6–1.7) |
Interaction term oseltamivir* AGE | ||||||
Oseltamivir* 0–9 years | 0.5751 | 80.8 | (67.2–97.1) | <0.0001* | 10.7 | (9.2–12.5) |
Oseltamivir* 10–19 years | 0.0006* | 96.4 | (77.5–119.9) | <0.0001* | 13.4 | (11.0–16.3) |
Oseltamivir* ≥20 years (as reference) | 1.0 | 1.0 | ||||
Female subgroup (N=1370745) | ||||||
Oseltamivir | <0.0001* | 8.5 | (6.3–11.3) | <0.0001* | 1.7 | (1.5–1.9) |
Reporting year | <0.0001* | 0.9 | (0.9–0.9) | <0.0001* | 0.9 | (0.9–0.9) |
AGE | ||||||
0–9 years | <0.0001* | 7.6 | (6.9–8.3) | <0.0001* | 1.6 | (1.5–1.6) |
10–19 years | <0.0001* | 4.0 | (3.7–4.4) | <0.0001* | 1.3 | (1.2–1.3) |
Interaction term oseltamivir* AGE | ||||||
Oseltamivir* 0–9 years | 0.2341 | 81.3 | (63.6–103.9) | <0.0001* | 8.3 | (6.8–10.2) |
Oseltamivir* 10–19 years | 0.7107 | 37.7 | (25.3–56.0) | <0.0001* | 5.0 | (3.8–6.5) |
Oseltamivir* ≥20 years (as reference) | 1.0 | 1.0 |
a) ROR: Reporting odds ratio. b) CI: Confidence interval. * Statistically significant.
Among female patients, the likelihood ratio test for “abnormal behavior” was not statistically significant. When considering NPAEs, the likelihood ratio tests using oseltamivir* 0–9 years and oseltamivir* 10–19 years were statistically significant (p<0.0001). The adjusted RORs (95%CI) for oseltamivir* 0–9 years and oseltamivir* 10–19 years were 8.3 (6.8–10.2) and 5.0 (3.8–6.5), respectively.
The ROR was developed to detect unexpected adverse events.14) The ROR of “abnormal behavior” in Singapore, and Taiwan was ≥1 at 95%CI in Table 3. The ROR of Japan and Singapore were higher than those from other countries. Japan uses more than 75% of the oseltamivir manufactured.12) Pediatric use of oseltamivir to treat influenza is higher in Japan than in the United States.10) Thus, it is not surprising that most NPAE reports are from Japan. One plausible reason for this difference could be inflated numbers of spontaneous event reports due to “notoriety bias,” which was caused by media attention and publicity after regulatory action in Japan.21) Genetic profile among Asian populations may reveal similarities, and the observation of NPAEs in Japanese patients treated with oseltamivir may be relevant in other Asian countries. Since prescribing oseltamivir has become more widespread, further epidemiological studies, such as case-control and follow-up studies, are necessary in Asian populations.
In Table 4, we demonstrated a possible relationship between oseltamivir use and “abnormal behavior” in male patients between the ages of 10 and 19 years using the adjusted ROR and logistic regression analysis. Previous reports using the FAERS database discussed only a small number of reported cases where abnormal behavior was increased by oseltamivir, as assessed using the RORs.3) The ROR does not allow for the evaluation of comparative strength of causality and offers only a rough indication of the strength of the signal.22) The adjusted RORs were increased in male and female patients under the age of 20 years. As for “abnormal behavior,” the adjusted RORs for oseltamivir in the male patients aged 10–19 years had the highest values. The effects of the interaction terms for oseltamivir in the male patients aged 10–19 years were statistically significant, and the estimated values for this term in the logistic regression exceed 0. We refined the covariates for female patients the likelihood ratio test for “abnormal behavior” was not statistically significant in female patient. “Abnormal behavior” may not be influenced by oseltamivir in females.
In contrast, some NPAEs may be associated with oseltamivir use in both genders. This may be because the PTs for NPAEs were determined using the Japanese package insert for oseltamivir, which contains a number of symptoms related to “abnormal behavior” in clinical practices. Thus, it may be crucial to address NPAEs in a clinical setting among patients under 20 years of age.
Toovey et al. reported that the largest category of NPAEs in the FAERS database from 2007 to 2010 was abnormal behavior (25.3%).10) NPAEs were also more common in children, representing 59.4% of adverse events in people under the age of 16 years.10) Most NPAEs were reported in patients less than 20 years of age (Table 2). Our study supports previous observations that the FAERS database contains a disproportionately high number of reports involving NPAEs linked to oseltamivir.
The relationship between oseltamivir use and NPAEs is poorly understood, since similar NPAEs were observed in the absence of oseltamivir.10,12,13) Influenza is associated with a variety of neurological and behavioral symptoms, including hallucinations, delirium, and abnormal behavior. These events may occur in the setting of encephalitis or encephalopathy.23)
The risk of NPAEs associated with oseltamivir has been controversial, and has been predominantly indicated in case reports, but not in controlled studies and preclinical research. A number of publications suggested that there is no evidence, nor plausible mechanism of action, to link oseltamivir with NPAEs.10,12,13) They concluded that the risk of abnormal behavior was increased by influenza, and not by oseltamivir use. Urushihara et al. reported that the number of abnormal behavior reports in Japan was consistently dependent on the size of the population treated with antivirals except for the reporting for oseltamivir in fiscal year 2007.8) They considered that potential biases introduced in the spontaneous adverse events reporting system might be substantial, making results difficult to interpret.
An epidemiologic study was conducted in Japan in 2006–2007 to evaluate the relationship between oseltamivir and NPAEs.11) The research group also concluded that no positive associations were detected between oseltamivir and abnormal behaviors. In contrast, Yorifuji et al. disagreed and proposed correct analysis (based on the person-time approach) and recalculated the estimated crude rate ratio of 1.57 (95%CI 1.34–1.83).7,24,25) Fujita et al. indicated a slightly increased risk of NPAEs in the cohort study.3) Hoffman et al. reported that FAERS data suggest disproportionally elevated reporting of certain NPAEs linked to oseltamivir in 2013.4) In 2014, members of the Cochrane Collaboration and others reported that NPAEs were increased in the combined on- and off-treatment periods in oseltamivir prophylaxis studies.2,26) Against such a background, we provide new information to the on-going controversy regarding NPAEs related to oseltamivir. Despite the limitations, we hope these findings offer an update for clinicians.
As for possible mechanism of neuropsychiatric adverse events in patients treated with oseltamivir, one plausible mechanism of action might be based on an increase in the penetration rate of the active metabolite (oseltamivir carboxylate) from the plasma to the central nervous system (CNS). Toovey et al. indicated that the penetration into the CNS of both oseltamivir and active metabolite was low in adults, and that they were and no specific CNS/behavioral effects after administration of doses corresponding to>or=100 times the clinical dose in animal studies.12) On the other hand, it has recently been reported that developmental changes in p-glycoprotein function in the blood brain barrier of rhesus monkeys using positron emission tomography (PET) with R-(11)C-oseltamivir, and this change may be closely related to the observed difference in drug responses in the brains of children and adult humans.27) Another group, using PET and autoradiography, radioactivity observed in the brains of infant, juvenile and adult rats after injection of [(11)C] oseltamivir.28) The highest radioactivity was found in the infant brain and the radioactivity level decreased with age.28) Furthermore, in acute encephalopathies and encephalitis, there is usually brain edema with evidence of damage to the blood brain barrier, which might be considered to cause an increase in the penetration of oseltamivir into the CNS.29) To the best of our knowledge, there have been no reports on the relationship between the concentrations of the active metabolite in the plasma (or in the CNS) and NPAEs in children. Wherever feasible, retrospective or prospective research using the data from oseltamivir clinical trials in children should be undertaken in the future.
We do not have a conclusive explanation for the gender difference in NPAEs in patients treated with oseltamivir. Hoshino et al. reported that no gender difference was noted in acute encephalopathy.30) Several reports suggest that in patients with attention deficit hyperactivity disorder (ADHD), boys are likely to be more hyperactive and impulsive and to have more comorbid externalizing disorder.31,32) The objection will no doubt be raised that ADHD and NPAEs in patients treated with oseltamivir are different; however, the gender difference in ADHD might be a thought-provoking observation in the interpretation of our results.
Our disproportionality analysis has several limitations. Because of common errors within spontaneous reporting databases, such as reporting bias, the cases reported in the FAERS databases do not always contain sufficient information to properly evaluate an event. We partially mitigate the confounding factors using the logistic regression method. Potential signals of adverse events are likely to be small and difficult to detect. The marked under-reporting that is common in FAERS may have resulted in ROR being underestimated. Further, it might be difficult to evaluate causality between oseltamivir and NAPEs, since there is a lack of information on dose response or withdrawal in patients. Despite the limitations, our results indicate that oseltamivir influences “abnormal behavior” in male patients between the ages of 10 and 19 years. Furthermore, oseltamivir may influence NPAEs in patients of both genders under the age of 20 years. Thus, young male patients should be closely monitored for the occurrence of “abnormal behavior” or “NPAEs” when they are prescribed oseltamivir. We hope that these data will enhance the information available to clinicians and be useful for improving the management of influenza.
This research was partially supported by JSPS KAKENHI Grant Number, 24390126.
JA is an employee of Medical Database. The rest of the authors have no conflict of interest.