Abstract
Introduction
The use of codeine in children aged < 12 years has been restricted in Japan since July 2017 due to safety concerns.
Objective
We aimed to investigate the impacts of two labelling revisions restricting codeine use in children on the trends in its prescriptions in Japan.
Methods
Children aged < 12 years (target group) and 12–17 years (control group) registered in the Japanese health insurance claims database between August 2015 and July 2021 were included. We analyzed the level and trend changes in monthly prescriptions of codeine and non-narcotic antitussives (reference) per 100,000 individuals using an interrupted time-series design across there period: the pre-intervention (before the first revision in July 2017), post-intervention (after the second revision in July 2019), and transitional (between the two revisions) periods.
Results
There was a significant reduction in codeine prescriptions of 67.5% and 36.6% in the first labelling revision in the target and control groups, respectively. In the target group, a downward trend in the level of codeine prescriptions was observed until the end of the transitional period. After the second labelling revision, the level of codeine prescriptions decreased by approximately 95% compared with the last month of the pre-intervention period. For non-narcotic antitussives, there were no significant changes in the level and trend of prescriptions until the transitional period, but the trend significantly changed to negative after the second labelling revision.
Conclusion
Restriction on codeine use in children by the labelling revisions could be effective for reducing codeine prescriptions in patients aged < 12 years.
Similar content being viewed by others
Avoid common mistakes on your manuscript.
The prescription rate of codeine products in children aged < 12 years was significantly reduced after two labelling revisions, especially after the first revision in July 2017. |
This study shows that safety measures such as labelling revisions can effectively reduce pediatric codeine use and promote appropriate prescriptions in healthcare providers in Japan. |
An interrupted time series (ITS) design is useful for evaluating risk minimization activities. |
Introduction
Codeine-containing drugs, including codeine phosphate hydrate and dihydrocodeine phosphate, are used to treat various respiratory diseases, provide analgesia for pain, and improve the symptoms of severe diarrhea. Codeine-containing drugs are metabolized by the hepatic metabolizing enzyme CYP2D6 to the active metabolites morphine and dihydromorphine, which exhibit central depressant (e.g., antitussive and analgesic effects) and peripheral (e.g., antidiarrhea) effects.
In individuals with genetically excessive CYP2D6 activity (ultrarapid metabolizers), blood levels of morphine and dihydromorphine may increase, and respiratory depression and other adverse reactions may occur more frequently [1]. Several revisions to codeine labelling were made in the USA [2] and Europe [3] between 2013 and 2017 to restrict its usage in children because it can be associated with life-threatening adverse events, including severe respiratory depression.
In alignment with warnings and restrictions in the USA and European countries, the Ministry of Health, Labor, and Welfare/Pharmaceuticals and Medical Devices Agency concluded that sufficient precautions should be taken against the use of codeine-containing drugs in Japan, especially among pediatric patients, to prevent the risk of serious outcomes due to respiratory depression [4]. As for the first intervention, the use of codeine-containing prescription drugs in children under 12 years of age was restricted by the labelling revision on 4 July 2017; it was stated in the “Important Precautions” section of the label that, in principle, codeine-containing drugs should not be administered to children under 12 years of age [5]. In the second intervention, on 9 July 2019, after a 2-year transitional period following the first revision, the marketing authorization holders of codeine-containing drugs removed the dosage and administration for pediatric use and revised the labelling described in the “Contraindications” section to administration in children under 12 years of age [6]. These labelling revisions were communicated via “Notification from Director-General of the Ministry of Health, Labour and Welfare.” This procedure includes the Ministry of Health, Labour and Welfare sending notifications to local government at the prefectural level requesting that medical institutions and pharmacies in their area be made aware. Also, all pharmaceutical companies that hold marketing authorization for the drug(s) must notify all medical institutions that purchase the drug(s) within 1 month of issue of the notification. The newly added wording on each label revision is shown in Supplementary Table 1.
The impacts of similar restrictions on the use of codeine-containing drugs have been evaluated in previous studies in the USA, the UK, France, Germany, Spain, and Taiwan [7,8,9,10,11,12,13,14,15]. Most of these studies showed a decreasing trend in codeine prescriptions coinciding with the timing of the labelling revision [7,8,9,10,11,12,13], although some of these studies concluded that the impacts of labelling revisions were unclear [14, 15]. In addition, some studies showed increases in prescriptions of other antitussive and analgesic medications as alternative treatments, along with a decrease in codeine prescriptions [9, 13, 14]. The impacts of restrictions on the use of codeine-containing drugs for the treatment of tonsillectomy/adenoidectomy, cough, and cold have also been investigated, with some studies showing that codeine is continuously prescribed for postoperative pain in children [16]. However, the impacts of labelling restrictions for codeine use have not yet been evaluated using nationwide Japanese population data. Therefore, the objective of this study was to investigate the impacts of two labelling revisions on restrictions against the use of codeine-containing drugs in July 2017 and July 2019 using large nationwide health insurance claims data and an interrupted time-series (ITS) design.
Methods
Data sources
This study relies on the administrative claims database of employee health insurance coverage provided by JMDC Inc. (JMDC Claims Database). As of 2024, over 17 million people were cumulatively enrolled from over 250 healthcare insurance organizations across Japan [17]. To enable the use of the database for research purposes, personal identifiable information was encrypted by JMDC Inc. to protect patient privacy. Data from August 2015 to July 2021 were used for the main analysis.
Study population
Individuals aged under 18 years with at least a 1-month observation period between August 2015 and July 2021 were included in the study population. The target group was defined as those under 12 years of age, which meets the age limit for the use of codeine-containing drugs according to labelling revisions. Individuals aged 12–17 years were used as the control group [18] for the ITS analysis because they were less likely to be restricted by the labelling revisions.
Measures of codeine and non-narcotic antitussives use (outcome variables)
The outcome variable was the monthly prescription rate of codeine-containing drugs per 100,000 individuals. The monthly prescription rate per 100,000 individuals was calculated by dividing the number of individuals in an age group with a prescription of codeine-containing drugs for any reason (numerator) by the number of individuals in the same age group who were enrolled in JMDC Claims Database during each month (denominator), and then multiplying by 100,000 individuals. In addition, the monthly prescription rate of non-narcotic antitussives per 100,000 individuals was calculated as a reference group because they were not subject to labelling revisions related to prescription restrictions for codeine and may also be used as an alternative treatment for codeine. When individuals were prescribed both codeine-containing drugs and non-narcotic antitussives in the same month, cases were counted separately for both drug groups. Codeine-containing drugs were defined as drugs containing either codeine or dihydrocodeine. Non-narcotic antitussives were defined as drugs containing benproperine, cherry bark extract, clofedanol, cloperastine, dextromethorphan, dimemorfan, eprazinone, noscapine, pentoxyverine, or tipepidine. Combination products containing both codeine/dihydrocodeine and non-narcotic antitussives were also defined as codeine-containing drugs. The drug codes are listed in Supplementary Tables 2 and 3.
Interventions and investigational periods
This study aimed to estimate the impacts of two labelling revisions (interventions) on restrictions on the use of codeine-containing drugs in July 2017 and July 2019. The investigation periods were defined as follows: the pre-intervention period (24 months before the first intervention in July 2017), post-intervention period (24 months after the second intervention in July 2019), and transitional period (24 months between the pre-intervention and post-intervention periods). Generally, 8 or 12 timepoints are recommended before and after each intervention [19, 20] to perform a segmented regression analysis. In this study, 24 months was considered a reasonable intervention period for assessing seasonal variation [19].
Statistical analysis
Descriptive analysis
The demographics of the study population and prescription patterns of codeine-containing drugs or non-narcotic antitussives in January 2016, 2018, and 2020 were shown to represent the three investigational periods because codeine was more likely to be used in the winter. The missing values were counted for each category. The number of individuals enrolled in JMDC Claims Database and those prescribed codeine-containing drugs/non-narcotic antitussives were counted according to four age groups (0–1, 2–5, 6–11, and 12–17 years old).
Modeling of segmented regression
An ITS design was used to assess the impacts of the interventions on the monthly prescription rates of codeine-containing drugs and non-narcotic antitussives. The level and trend changes in monthly prescription rates of codeine-containing drugs or non-narcotic antitussives were estimated using segmented regression models [19] constructed using the PROC AUTOREG procedure in SAS/ETS Software, which estimates the parameters of the ordinary least squares method by correcting the error terms that are auto correlated with the time-series data [20,21,22]. The details of the segmented regression modeling are shown in the Supplementary Methods section. We calculated the 95% confidence interval (CI) of the estimates. P values < 0.05 were considered statistically significant. All analyses were performed using SAS version 9.4 and SAS/ETS version 15.2 (SAS Institute Inc., Cary, NC, USA).
Stratified and sensitivity analyses
Stratified analysis by age group was performed, and the target group (0–11 years) was classified into three strata (0–1, 2–5, and 6–11 years old). Three sensitivity analyses were performed (sensitivity analyses 1, 2, and 3): sensitivity analysis 1 was performed to assess the impact of the coronavirus disease 2019 (COVID-19) pandemic using a dataset censoring the period after April 2020, after the first emergency declaration for the COVID-19 on 7 April 2020, which may have resulted in changes in medical behavior. Sensitivity analysis 2 omitted the month of label revision implementation (July 2017 and July 2019) and the following month (August 2017 and August 2019) from the investigational period, given the lag effect due to the dissemination of label revisions to healthcare providers. In sensitivity analysis 3, patients who underwent post-tonsillectomy or adenoidectomy were excluded from the study population because the labelling revision in July 2019 included a contraindication to prescribing codeine for patients under 18 years of age who used it for analgesic purposes after tonsillectomy or adenoidectomy (the claim code is shown in Supplementary Table 4).
Results
Characteristics of the study population
Table 1 shows the characteristics of the denominator population (all insured persons aged < 18 years in the JMDC Claims Database) for each month during the pre-intervention (January 2016), transitional (January 2018), and post-intervention (January 2020) periods. The total number of insured persons aged < 18 years increased during the transitional and post-intervention periods compared with the pre-intervention period, with 1,575,289, 1,831,513, and 988,659 insured persons, respectively (Table 1), which is consistent with the expansion of insured persons in the JMDC Claims Database. As for the disease classes and drug types, the percentage of “Influenza and pneumonia” and “Antivirals for systemic use” were higher during the transitional period than during the pre- and post-intervention periods. The distributions of age, sex, receipt type, physician specialty, hospital size, top five most common disease classes, and top five most common drug types were generally comparable across the study periods.
Characteristics of codeine-containing drug users
Table 2 shows the characteristics of the numerator population (i.e., codeine-containing drug users) by age group during the pre-, transitional, and post-intervention periods (characteristics not displayed in Table 2 are shown in Supplementary Table 5). In the target group (0- to 11-year-old), the number of codeine-containing drug users decreased over time from the pre-intervention period (6251 patients; 0.94% of insured persons) to the transitional (3621 patients; 0.35%) and post-intervention (607 patients; 0.05%) periods, although the total number of insured persons increased with time (Table 2). In the control group (12- to 17-year-old), the number of codeine-containing drug users during the pre- and post-intervention periods was almost the same (2704 patients; 0.83% and 3097 patients; 0.50%, respectively), whereas it was higher during the transitional period (6698 patients; 1.26%). In both age groups, 28–44% of the total codeine-containing drug users received a co-prescription of non-narcotic antitussives during each period. In terms of the type of codeine prescription, more than 85% of the codeine prescriptions were multiple-ingredient products.
ITS analysis
Figure 1 shows the monthly time series of codeine and non-narcotic antitussive prescription rates per 100,000 individuals. Overall, codeine and non-narcotic antitussive prescriptions showed seasonality, with more prescriptions in winter, coinciding with the cold and flu seasons.
Codeine prescriptions
In the target group (0- to 11-year-old), a negative trend in codein prescriptions was observed during the pre-intervention period. The baseline monthly prescription rate per 100,000 individuals during the pre-intervention period in the target group was 1339.8 (95% CI 1268.7–1410.8), showing a significant decreasing trend in the number of prescriptions of − 17.7 (95% CI − 22.8 to − 12.7; P < 0.0001). The prescription rate substantially decreased 67.5% after the first month of the transitional period compared with the last month of the pre-intervention period [− 608.8 (95% CI − 710.3 to − 507.4); P < 0.0001] (Table 3, Fig. 1a). Although a downward trend was observed during the transitional period, the degree of the decrease was smaller than that observed during the pre-intervention period. During the post-intervention period, there was no significant immediate reduction (level change) in the prescription rate [− 64.6 (95% CI − 165.8 to 36.6); P = 0.207] (Table 3, Fig. 1a). The number of prescriptions immediately after the second intervention (August 2019) was as low as 46.1, which was approximately 95% lower than that immediately before the first intervention (June 2017).
In the control group (12–17 year-old), the trend of the monthly prescription rates was similar during the pre-intervention [1.5 (95% CI − 5.2 to 8.2); P = 0.66] and transitional [− 0.6 (95% CI − 8.7 to 7.4); P = 0.874] periods (Table 3, Fig. 1b), but there was a significant immediate decrease in the prescription rate (level change) in the first month of the transitional period compared with the rate of the last month of pre-intervention period [− 335.8 (95% CI − 469.7 to − 201.9); P < 0.0001; 36.6% decrease] as well as the first month of the post-intervention period compared with the rate of the last month of transitional period [− 171.1 (95% CI − 307.1 to − 35.1); P = 0.015; 31.9% decrease]. After the second intervention, the trend significantly changed to negative [− 21.0 (95% CI − 29.0 to − 13.0); P < 0.0001].
Non-narcotic antitussive prescriptions
In the target group, the baseline prescription rate of non-narcotic antitussive prescriptions was 16,180.1 (95% CI 13,843.5–18,516.8), with a consistent pre-intervention trend. While there was no significant change in either the level or the trend during the transitional period, the point estimate shows a shift from a slight negative trend during the pre-intervention period [− 18.4 (95% CI − 183.5 to 146.8); P = 0.825] to an upward trend during the transitional period [113.1 (95% CI − 102.7 to 328.9); P = 0.299]. After the second intervention, the trend significantly changed to negative [− 290.6 (95% CI − 507.6 to − 73.7); P = 0.010), indicating a decrease in the response variable over time.
In the control group, there was no significant change in the trend during the pre-intervention and transitional periods and in the level of prescriptions after the two interventions. However, after the second intervention, the trend significantly changed to negative [− 154.6 (95% CI − 241.1 to − 68.1); P = 0.001].
Age stratified analysis
A stratified analysis was conducted for the target group (0- to 11-year-old), which was divided into three strata (0–1-, 2–5-, and 6–11 year olds) to identify differences in codeine prescription restrictions in the target group (Supplementary Table 6 and Supplementary Fig. 1). In each age group, compared to all populations aged 0- to 11-year-old, there was no notable change in the overall trend; there was a decreasing trend during the pre-intervention period and a significant decrease in the level after the first month of the transitional period. The decreasing trend persisted during the transitional period, and the prescription rate decreased to nearly zero after the second intervention. The percentage of prescription decrease (level change) after the first intervention was the highest in the age group of 0–1 year old (71.6% decrease), followed by 2–5 year old (70.0% decrease) and 6–11 year old (60.2% decrease).
Sensitivity analysis
Sensitivity analysis 1, excluding the period of the COVID-19 pandemic, was performed to evaluate the impact of changes in medical behavior during the COVID-19 pandemic (Supplementary Table 7 and Supplementary Fig. 2). For codeine prescriptions in the target group, there was no significant change in the estimate compared with the primary analysis. For non-narcotic antitussive prescription in the target group, the significant decrease in the level change after the second intervention (β4) in the primary analysis disappeared in the sensitivity analysis (Supplementary Table 6). In the control group, significant changes in the level and trend in the primary analysis were no longer observed in the sensitivity analysis for both codeine and non-narcotic antitussive prescriptions.
Exclusion of the month of intervention (sensitivity analysis 2) and tonsillectomy or adenoidectomy (sensitivity analysis 3) did not significantly change the overall trend (Supplementary Figs. 3 and 4), with only a slight change in the prescription rates (estimates) compared with those in the main analysis (Supplementary Table 7).
Discussion
This is the first study to investigate the impact of a two-stage labelling revision of codeine on its use in the pediatric population in Japan, using a robust quasi-experimental design and a nationwide health insurance claims database. Overall, the number of prescriptions of codeine-containing drugs showed > 90% reduction among pediatric patients aged < 12 years who were the target of prescription restrictions through a two-stage labelling revision. This result indicates that the labelling revisions could have affected doctors’ decisions on the prescription of codeine-containing drugs in children.
Descriptive analysis of the overall population and codeine-containing drug users
In the present study, population characteristics were generally comparable during the pre-, transitional, and post-intervention periods. Therefore, the assumption of the ITS that the composition of the population under study does not change across periods was considered verified [19, 20]. One difference observed in the comparison of the characteristics of the periods was that “influenza and pneumonia” were more common during the transitional period (January 2018) in both the denominator and codeine-user populations, consistent with the evidence of a higher incidence of these infectious diseases during the winter season of 2018 [23] compared with the other 2 years.
Regarding the number of codeine-containing drug prescriptions in the target group from the age-stratified analysis, the population in the lower age groups (0–1 and 2–5 years) especially decreased after the two-stage labelling revision compared to the age group of 6–11 years, indicating that the labelling changes had an impact on children, especially at lower ages.
Findings on prescribing trends for codeine and non-narcotic antitussives from ITS analysis
The present study shows a significant decrease in the prescription rate of codeine-containing drugs for patients aged < 12 years after the first intervention. Changes in prescription trends were temporally compatible with the first labelling revision. Although the two revisions of labelling in Japan were conducted in a stepwise manner, the regulatory authority had already announced at the time of the first revision that codeine would be contraindicated in that population within 2 years. Therefore, the prescription rate was significantly reduced during the first revision. Additionally, in the present study, codeine-containing drugs were mainly prescribed in winter, and upper respiratory inflammation and influenza were listed among the most common diseases in the claims. Therefore, codeine-containing drugs are considered to be mainly used for acute cough suppressive therapy due to cold or influenza, and easy switching of the drugs may have also contributed to the substantial decrease in prescriptions. The results of this study are in line with other studies conducted in other countries [7,8,9,10,11,12,13].
Prescriptions of codeine in the target group had already shown a downward trend before the first intervention, whereas there was no downward trend in the control group of codeine or both groups of non-narcotic antitussives. A downward trend before the intervention (revision of the labelling restricting use to children) was also observed in other countries [11]. This indicates that physicians were aware of concerns regarding the use of codeine in patients under 12 years of age through alerts other than labelling revisions. For example, the PRAC recommendation regarding restricting prescription of codeine for children in 2013 was included in the 2013 National Institute of Health Sciences (NIHS) report [24].
Regarding the downward trend during the transitional period, the decrease was smaller than during the pre-intervention period. This may have been because the number of prescriptions decreased after the first intervention. The decrease in the prescription rate after the second intervention was difficult to evaluate because the target population with prescriptions was already very low.
Our study found a spillover effect among children aged 12–17 years. This spillover effect is an unintended outcome. It results from the complex features of healthcare systems and delivery and can either increase or decrease the overall effectiveness of interventions [25]. In our study, even though this age group was not the subject of restrictions in the labelling revisions (hence used as a control group in the present study), there was a sharp drop in the levels after the first intervention. One possible reason for this decrease could be the establishment of restrictions on the use of codeine during tonsillectomy or adenoidectomy in patients < 18 years of age, which was implemented simultaneously with the first intervention in this study. However, our sensitivity analysis indicated that the impact of restrictions on tonsillectomy or adenoidectomy might be small because the number of patients who used codeine for tonsillectomy or adenoidectomy during the study period was very small. Therefore, it is suggested that the spillover effect on children aged 12–17 years contributed to this decrease. These results align with previous studies that showed a decrease in prescriptions in older children aged > 12 years [7, 10, 14].
There was no significant change in non-narcotic antitussive prescriptions after codeine prescription was restricted in the first intervention. Although the difference was not significant, the point estimate showed that the decreasing trend before the first intervention changed to an increasing trend. This may indicate that the restriction of codeine prescriptions led to an increase in the prescription of alternative antitussive drugs (i.e., non-narcotic antitussive drugs); however, no definitive conclusions can be drawn from the current study.
Potential factors influencing the outcomes
In all the investigated groups, a negative trend was observed after the second intervention. Because of the COVID-19 pandemic since March 2020, the government has been requesting children to refrain from going out and taking measures for school and kindergarten/nursery suspension in Japan, which may have reduced contact between children and the infection itself. Sensitivity analysis, excluding the COVID-19 period, revealed that some estimates that were significantly different in the main analyses were no longer significant in the sensitivity analysis. It is more reasonable to consider that this change in the trend after the second labelling revision was due to the influence of the government’s request to stay at home, including the closure of schools, nurseries, and kindergartens, and the promotion of general preventive actions for infections such as hand washing, wearing a mask, and avoiding close contact due to the COVID-19 pandemic [26]. No other events, such as changes in reimbursements or guidelines, were considered to have influenced the outcome.
Limitation
Our study had some limitations. First, because the JMDC Claims Database contains health insurance claims data obtained from a large-scale company’s health insurance society and includes a large amount of data on Japanese families with relatively high incomes, caution should be exercised when generalizing this result to the entire Japanese population. Second, the JMDC Claims Database contains potential measurement biases that claims database naturally have, such as inaccuracies in coding and underreporting. For example, data on comorbidities refer to disease codes added for reimbursement claims and may not be actual medically confirmed disease names. Furthermore, we were unable to obtain patient characteristics such as disease severity of the study population, and these could not be adjusted for in the analysis model. Third, data on over-the-counter (OTC) drugs are not included in the JMDC Claims Database. Although codeine-containing OTC drugs whose label were also revised at the same timing as its prescription drugs are available in Japan, we could not investigate the impact of labelling changes in these OTC drugs. Fourth, the focus of this analysis was on the monthly prescription rate in the population, and changes in prescriptions at an individual level (e.g., dose reduction and shortening of the prescription period) could not be examined. Finally, a limitation of the study design is the vulnerability of concurrent events. When other interventions or events occur around the time of the interventions that may affect the outcome, it becomes difficult to evaluate the impact of the targeted intervention itself. To improve the validity of the study, characteristic-based control (population aged 12–18 years), and reference drug groups were used. To the best of our knowledge, there were no other co-occurring policy interventions or events during the investigation period that could have a significant impact on codeine use found in our study.
Conclusion
Using a large-scale claims database and quasi-experimental design, this study shows that the restriction on the use of codeine by two labelling revisions was sufficiently effective in reducing prescriptions in patients aged < 12 years in Japan.
References
Dean L, Kane M. Codeine therapy and CYP2D6 genotype. In: Pratt VM, Scott SA, Pirmohamed M, Esquivel B, Kattman BL, Malheiro AJ, editors. Medical genetics summaries. Bethesda: National Center for Biotechnology Information; 2012.
US Food and Drug Administration. Safety review update of codeine use in children; new Boxed Warning and Contraindication on use after tonsillectomy and/or adenoidectomy. February 20, 2013. https://www.fda.gov/downloads/Drugs/DrugSafety/UCM339116.pdf. Accessed 29 Aug 2024.
European Medicines Agency. Codeine not to be used in children below 12 years for cough and cold EMA/249413/2015. April 24 2015. https://www.ema.europa.eu/en/medicines/human/referrals/codeine-containing-medicinal-products-treatment-cough-or-cold-paediatric-patients. Accessed 29 Aug 2024.
Pharmaceutical and Medical Devices Agency. Report on the investigation results. https://www.pmda.go.jp/files/000218870.pdf. Accessed 29 Aug 2024.
Pharmaceutical and Medical Devices Agency. Revisions of PRECAUTIONS July 4, 2017. Revisions of PRECAUTIONS (FY2017). https://www.pmda.go.jp/safety/info-services/drugs/calling-attention/revision-of-precautions/0319.html. Accessed 29 Aug 2024.
Pharmaceutical and Medical Devices Agency. Revisions of PRECAUTIONS July 9, 2019. Revisions of PRECAUTIONS (FY2019). https://www.pmda.go.jp/safety/info-services/drugs/calling-attention/revision-of-precautions/0349.html. Accessed 29 Aug 2024.
Lin CW, Wang CH, Huang WI, Ke WM, Chao PH, Chen WW, et al. Impact of safety-related regulations on codeine use in children: a quasi-experimental study using Taiwan’s National Health Insurance Research Database. Drug Saf. 2017;40(7):615–27. https://doi.org/10.1007/s40264-017-0524-3.
Chua KP, Shrime MG, Conti RM. Effect of FDA investigation on opioid prescribing to children after tonsillectomy/adenoidectomy. Pediatrics. 2017;140(6): e20171765. https://doi.org/10.1542/peds.2017-1765.
Lawrence A, Cooper JN, Deans KJ, Minneci PC, Wrona SK, Chisolm DJ. Effects of the FDA codeine safety investigation on racial and geographic disparities in opioid prescribing after pediatric tonsillectomy and/or adenoidectomy. Glob Pediatr Health. 2021;8:2333794X20987444. https://doi.org/10.1177/2333794X20987444.
Renny MH, Jent V, Townsend T, Cerda M. Impact of the 2017 FDA drug safety communication on codeine and tramadol dispensing to children. Pediatrics. 2022. https://doi.org/10.1542/peds.2021-055887.
Hedenmalm K, Blake K, Donegan K, Macia MA, Gil M, Williams J, et al. A European multicentre drug utilisation study of the impact of regulatory measures on prescribing of codeine for pain in children. Pharmacoepidemiol Drug Saf. 2019;28(8):1086–96. https://doi.org/10.1002/pds.4836.
Goldman JL, Ziegler C, Burckardt EM. Otolaryngology practice patterns in pediatric tonsillectomy: the impact of the codeine boxed warning. Laryngoscope. 2018;128(1):264–8. https://doi.org/10.1002/lary.26719.
Van Cleve WC. Pediatric posttonsillectomy analgesia before and after the black box warning against codeine use. JAMA Otolaryngol Head Neck Surg. 2017;143(10):1052–4. https://doi.org/10.1001/jamaoto.2017.1089.
Oliveira J, Silva L, Anderson JL, Bellolio F, Campbell RL, Jeffery MM. Codeine dispensing for privately insured children in the USA: a retrospective database study. BMJ Paediatr Open. 2021. https://doi.org/10.1136/bmjpo-2021-001321.
Chua KP, Conti RM. Prescriptions for codeine or hydrocodone cough and cold dedications to US children and adolescents following US Food and Drug Administration Safety Communications. JAMA Netw Open. 2021;4(11): e2134142. https://doi.org/10.1001/jamanetworkopen.2021.34142.
Kohler JE, Cartmill RS, Kalbfell E, Schumacher J. Continued prescribing of periprocedural codeine and tramadol to children after a Black Box Warning. J Surg Res. 2020;256:131–5. https://doi.org/10.1016/j.jss.2020.06.023.
JMDC Inc. JMDC Claims Database. 2023. https://www.jmdc.co.jp/en/jmdc-claims-database/. Accessed 29 Aug 2024.
Lopez Bernal J, Cummins S, Gasparrini A. The use of controls in interrupted time series studies of public health interventions. Int J Epidemiol. 2018;47(6):2082–93. https://doi.org/10.1093/ije/dyy135.
Wagner AK, Soumerai SB, Zhang F, Ross-Degnan D. Segmented regression analysis of interrupted time series studies in medication use research. J Clin Pharm Ther. 2002;27(4):299–309. https://doi.org/10.1046/j.1365-2710.2002.00430.x.
Penfold RB, Zhang F. Use of interrupted time series analysis in evaluating health care quality improvements. Acad Pediatr. 2013;13(6 Suppl):S38-44. https://doi.org/10.1016/j.acap.2013.08.002.
SAS Institute Inc. SAS/ETS User's Guide 14.2. The AUTOREG Procedure. https://support.sas.com/documentation/onlinedoc/ets/142/autoreg.pdf. Accessed 29 Aug 2024.
Slavova S, Costich JF, Luu H, Fields J, Gabella BA, Tarima S, et al. Interrupted time series design to evaluate the effect of the ICD-9-CM to ICD-10-CM coding transition on injury hospitalization trends. Inj Epidemiol. 2018;5(1):36. https://doi.org/10.1186/s40621-018-0165-8.
Ministry of Health, Labour and Welfare. Incidence of Influenza [in Japanese]. https://www.mhlw.go.jp/stf/seisakunitsuite/bunya/kenkou_iryou/kenkou/kekkaku-kansenshou01/houdou.html. Accessed 29 Aug 2024.
Natinal Institute of Health Sciences. PRAC recommends restricting the use of codeine when used for pain relief in children referral. NIHS Overseas Drug Safety Information vol. 11 No15(2013/07/18) [in Japanese]. https://www.nihs.go.jp/dig/sireport/weekly11/15130718.pdf. Accessed 29 Aug 2024.
Francetic I, Meacock R, Elliott J, Kristensen SR, Britteon P, Lugo-Palacios DG, et al. Framework for identification and measurement of spillover effects in policy implementation: intended non-intended targeted non-targeted spillovers (INTENTS). Implement Sci Commun. 2022;3(1):30. https://doi.org/10.1186/s43058-022-00280-8.
Wakabayashi T, Sasaoka Y, Sakai Y, Hinotsu S, Kawasaki Y. Decrease in unscheduled pediatric outpatient visits due to SARS-CoV-2. Pediatr Int. 2022;64(1): e14748. https://doi.org/10.1111/ped.14748.
Acknowledgements
We thank Shingo Kuroda and Shuji Sumino (employees of Takeda Pharmaceutical Company Limited) for reviewing the SAS programming code and data analysis. We also thank Megumi Sugawara, Yoshiko Maeda, Mari Shibata, Masaya Hiraizumi, and Youko Ueda (employees of Takeda Pharmaceutical Company Limited) for their constructive comments.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Funding
This work was funded by Takeda Pharmaceutical Company Limited, Japan.
Conflict of interest
All authors are full-time employees of Takeda Pharmaceutical Company Limited, a marketing authorization holder of codeine phosphate in Japan.
Ethics approval
Ethics approval was not applicable. Since the data used in this study are aggregate drug prescription rates and patient background, the privacy and confidentiality of the data are well-protected. Although patients can be identified via a unique identifier by JMDC Inc., the secondary database used for research purposes has de-identified data only.
Consent to participate and publication
The need for informed consent and consent to publish was waived due to the retrospective nature of the study and the anonymity of the patient database.
Availability of data and material
The data used in this study were provided on a contract basis between the sourcing vendor (JMDC Inc.) and Takeda. Upon our request, JMDC Inc. agreed to provide the data for this study without restriction. The data cannot be shared publicly due to the aforementioned contracts of both companies.
Code availability
The code used for this study are available from the corresponding author upon reasonable request.
Author contributions
Y.S. and Y.O. contributed to conceptualization, formal analysis, methodology, writing—original draft, and equally contributed to this work. M.H. contributed to conceptualization, validation, methodology, and writing—review and editing. M.S. contributed to conceptualization, funding acquisition, supervision, resource, and writing—review and editing. N.N. contributed to conceptualization, formal analysis, methodology, project administration, and writing—original draft. All authors read and approved the final version.
Supplementary Information
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial 4.0 International License, which permits any non-commercial use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc/4.0/.
About this article
Cite this article
Sakakibara, Y., Ogino, Y., Hasegawa, M. et al. Impacts of labelling revisions on pediatric use of codeine: interrupted time-series analysis using the Japanese nationwide claims database. Drugs Ther Perspect (2024). https://doi.org/10.1007/s40267-024-01110-8
Accepted:
Published:
DOI: https://doi.org/10.1007/s40267-024-01110-8