Diagnosis and treatment of influenza based on health insurance claims between the 2010–2011 and 2019–2020 influenza seasons in Japan

Abstract Medical practices for influenza virus infection vary among countries. In Japan, treatment with anti‐influenza drugs is recommended for patients diagnosed with influenza. This health claims database study provides quantitative information aimed at describing the actual medical practices, including diagnostic testing and medication use, for managing influenza in Japan. Most patients diagnosed with influenza underwent diagnostic tests and were prescribed anti‐influenza drugs. Meanwhile, the majority of patients prescribed anti‐influenza drugs had undergone diagnostic testing. However, an increase in the percentage of anti‐influenza prescriptions without diagnostic testing was observed during the 2019–2020 influenza season, which may be associated with the COVID‐19 pandemic.


Although influenza virus infection is self-limiting in most patients, it is
an important disease that affects many people and causes severe symptoms in some cases. In Japan, the epidemic occurs mainly during winter, and the cumulative seasonal number of influenza patients is approximately 5%-15% of the population. 1 Influenza can cause complications such as influenza-associated encephalopathy in infants and young children and pneumonia in the elderly. 2,3 Medical practices for influenza vary among countries, although rapid diagnosis and clinical management reduce the incidence of severe illness and complications. 4 In Japan, treatment with anti-influenza drugs is recommended for patients who were confirmed or suspected of having influenza by the Japanese Association for Infectious Diseases (JAID). 5 A rapid diagnostic test is usually performed for suspected cases, but the sensitivity varies depending on time between symptom onset and presentation. 6 Consequently, influenza diagnosis may not always be based solely on test results but can instead be comprehensively determined based on symptoms and disease prevalence.
Quantitative information about medical practices for influenza, including diagnostic testing, diagnosis, and treatment, is not well provided in Japan, although the epidemic status of seasonal influenza has been reported by surveillances, 2,7,8 and a high prescription rate (98.4%) for hospitalized child patients during the 2009 influenza A (H1N1) pandemic has been reported by a chart review study. 9 Furthermore, medical practices may have changed due to the coronavirus disease 2019 (COVID-19) pandemic. Therefore, this study analyzed insurance claims databases to show the quantitative information about medical practices for influenza between the 2010-2011 and 2019-2020 seasons and better understand the medical handling of influenza in Japan.

| METHODS
A claims database of Japanese health insurance associations, 10 provided by JMDC Inc., contains data of employees and their family members obtained from companies belonging to these associations, which corresponded to approximately 6% of the Japanese population as of August 2020. The percentage of males is higher than that of females, people aged ≥65 years are few (those aged 70-74 are <1% of the members), and those aged ≥75 years are excluded. The database includes the daily records of medical practices, including diagnostic tests, treatments, and monthly diagnostic records. All records associated with remuneration can be traced for each individual, regardless of medical institution type. Results of diagnostic tests are not included.
This study used descriptive statistics. We analyzed patient numbers and the distribution of combination patterns of diagnostic tests (hereafter tests), diagnoses, and prescriptions for anti-influenza medications throughout each season. We also analyzed weekly numbers of tests and prescriptions. The study period was from September 2010 to April 2020 and was divided by each influenza season, starting from

| RESULTS
The percentage of patients with records of tests, diagnoses, or prescriptions for influenza was 11.0% of the population in 2012/2013the lowest during the study period (Table 1). In 2017/2018 and 2018/2019, the percentage was the highest, almost 20% in both the seasons. Among the combination patterns, the test + diagnosis + prescription combination was the most frequent, followed by test alone until 2018/2019, but test alone was the most frequent (55.2%) in 2019/2020 ( Figure 1A). Among patients with a diagnosis, ≥90% took the test ( Figure 1B) and ≥90% received prescriptions ( Figure 1C  highest in the other weeks (Figures 2 and S1). Prescription alone was <5% in most weeks. In particular, the percentage of prescription alone was low during the weeks when the number of patients was small with a high percentage of test alone, and the percentage was high when the number of patients was large with a high percentage of the test + prescription combination. However, in 2019/2020, the percentage of prescription alone was high and remained approximately 6%-8% from the 29th to the 35th week, although the patient number was small (Figures 2 and S1). for the COVID-19 pandemic. 13 Later, in August 2020, proposals from the same association released the recommendation of diagnostic tests for both influenza and COVID-19. 14 Therefore, this observation may have changed again in late 2020.

| DISCUSSION
This study had several limitations. First, diagnosis codes may have been intentionally assigned along with diagnostic tests or prescription drugs. In addition, patients who visit medical institutions are usually willing to take treatment. These might contribute to high percentages of the combinations of diagnoses + tests and/or prescriptions. Second, the database included few elderly people, which did not accurately reflect the Japanese population ratio.

PATIENT CONSENT STATEMENT
Informed consent was not required because this study used a database that contained anonymized data collected for secondary use.

PERMISSION TO REPRODUCE MATERIAL FROM OTHER SOURCES
This manuscript does not include any previously published material.

PEER REVIEW
The peer review history for this article is available at https://publons. com/publon/10.1111/irv.12977.

DATA AVAILABILITY STATEMENT
The data that support the findings of this study are available from JMDC Inc. Restrictions apply on the availability of these data, which were used under license for this study. Data are available from the authors with the permission of JMDC Inc.