Psychotropic Medication Prescribing for Children and Adolescents After the Onset of the COVID-19 Pandemic

Key Points Question How did prescribing of psychotropic medications for children and adolescents change in the 2 years following the onset of the COVID-19 pandemic? Findings This cross-sectional study retrieved and analyzed all 8 839 143 psychotropic medication prescriptions dispensed to individuals aged from 6 to 17 years in France between 2016 and 2022. An interrupted time-series analysis showed steady increases in prescription trends for all psychotropic medications after the pandemic onset, with prescription rates of all psychotropic medication classes except psychostimulants higher than expected rates. Meaning Findings of significant increases in prescribing of psychotropic medications for children and adolescents after the pandemic onset underscore the need for future research to identify the underlying determinants.


Introduction
Mental health disorders are identified as one of the most significant challenges in pediatric health care. 1,2Their prevalence has been increasing over the years, and currently almost 15% of children and adolescents are diagnosed with a mental disorder by the age of 18 years. 3In this context, the COVID-19 pandemic has been a major aggravating factor with an unprecedented, large-scale negative impact on mental health.[6] Use of psychotropic medications is often considered a proxy for the mental health status of a population or for the prevalence of specific mental disorders. 72][13] However, data on psychotropic medication use in children and adolescents long after the pandemic onset are currently lacking.The purpose of the present study was to assess the rates and trends of psychotropic medication prescribing for children and adolescents before and in the 2 years after the onset of the COVID-19 pandemic in France.

Study Design and Setting
This cross-sectional study used an interrupted time-series analysis of outpatient drug dispensing data from January 1, 2016, to May 31, 2022.This study was part of the MENTALPED national project, 14 financed in 2021, to evaluate the association between the COVID-19 pandemic and the use of pediatric mental health medical resources by children and adolescents.The project's protocol was approved by the national Ethics and Scientific Committee for Health Research, Studies and Evaluations and by the French Data Protection Authority and is publicly available in the ClinicalTrials.govdatabase. 14The requirement to obtain patient informed consent was waived because the study used deidentified data and no patients were contacted.Findings were reported following the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline. 15

Study Data Sources and Population
Study data originated from the X-ponent database (IQVIA; France), a privately owned database that contains information on prescriptions dispensed from a sample of 14 000 pharmacies, approximately 60% of all retail pharmacies in France, excluding overseas territories. 16 representativeness on a national level.Data are extrapolated through an internal projection algorithm to approximate national estimates of drug prescription fills as detailed elsewhere. 17 extracted all psychotropic medication prescriptions dispensed to pediatric outpatients aged from 6 to 17 years during the study period.Children were defined as individuals aged 6 through 11 years; adolescents, 12 through 17 years.Dispensed prescriptions were identified as all pharmacy sales records containing at least 1 psychotropic medication based on the Anatomical Therapeutic Chemical classification: antipsychotics (N05A), anxiolytics (N05B), hypnotics and sedatives (N05C except N05CH01, melatonin), antidepressants (N06A), and psychostimulants (N06B) (eTable 1 in Supplement 1).For each dispensed prescription, we retrieved the dispensed medication (Anatomical Therapeutic Chemical fifth level) and the patient's age at dispensing.

Outcome Measures and Study Periods
The monthly rates of psychotropic medication prescriptions per 1000 children and adolescents in France were calculated using the counts of dispensed prescriptions (numerator) and appropriate age-specific population estimates derived from the annual dataset of the National Institute of Statistics and Economic Studies (denominator). 18Both the numerator and denominator were aggregated on a monthly level.
The study involved 3 predefined periods: the period before the pandemic (prepandemic) from January 1, 2016, to February 29, 2020, and 2 pandemic periods: the initial pandemic period from March 1 to May 31, 2020 (corresponding to the strict home confinement of the population and responsible for major disruptions in health care), and the period after the pandemic onset from June 1, 2020, to May 31, 2022.

Statistical Analysis
We assessed changes in the monthly rate (level) and trend (slope) of psychotropic medication prescriptions before and after the pandemic onset using a quasi-Poisson regression accounting for autocorrelation and prepandemic trends. 19,20Seasonal variations in psychotropic medication dispensing may be due to seasonal fluctuations of clinical symptoms, prescribing patterns, and practical considerations.Therefore, seasonality was addressed by fitting calendar months as a categorical variable.Residual autocorrelation in the errors was assessed through visual inspection of correlograms (autocorrelation and partial autocorrelation functions plots) and was corrected, when present, by using the Newey-West method. 21 hypothesized that there would be an immediate change in the prescription rate (level change) at the beginning of every pandemic period and a progressive change (slope change) in the period after the pandemic onset.The interrupted time-series model, therefore, included time as a continuous variable to assess the baseline trend, 2 binary variables indicating the 2 pandemic periods to measure the level changes, and an interaction term between time and the postpandemic binary variable to measure the slope change.The validity of all regression models was assessed by visual inspection of correlograms and appropriate residual analysis.Model parameters were interpreted as percentage changes with 95% CIs.A linear combination of the trend parameters (prepandemic slope and estimated slope change) was implemented to obtain the trend after the onset of the pandemic with 95% CIs. 20The monthly prescription rates estimated by the model during the 2-year period after the onset of the pandemic were also compared with the expected rates forecast by the model based on prepandemic data only.Results were expressed as rate ratios (RRs) with corresponding 95% CIs calculated by bootstrapping (n = 10 000 replications) (eMethods in Supplement 1).

Results
Between

Changes in Overall Psychotropic Medication Prescription Rates
In January 2016, the estimated rate of monthly psychotropic medication prescriptions was 9.9 per 1000 children and adolescents in France, and the prepandemic monthly prescription rate was increasing by 0.4% per month (95% CI, 0.3%-0.4%)(

Changes in the Prescription Rates of Antipsychotics
In January 2016, the estimated rate of antipsychotic prescriptions was 2.6 per 1000 children and adolescents, and the prepandemic rate was increasing by 0.3% per month (95% CI, 0.2%-0.4%)(Table , Figure 3).The monthly rate of prescriptions did not significantly change in the initial pandemic period nor after the pandemic onset.However, the trend significantly changed, and the antipsychotic medication prescription rate increased by 1.0% per month (95% CI, 0.9%-1.2%),reaching 4.1 per 1000 children and adolescents in May 2022.Compared with the expected rates, a relative increase of 11% (RR, 1.11 [95% CI, 1.09-1.13])was observed.

Changes in the Prescription Rates of Anxiolytics
In January 2016, the estimated rate of anxiolytic prescriptions was 2.7 per 1000 children and adolescents, and the prepandemic rate was decreasing by 0.2% per month (95% CI, −0.3% to −0.1%) (Table, Figure 3).After an immediate drop in the monthly prescription rate during the initial pandemic period of 13.6% (95% CI, −17.7% to −9.3%), the trend substantially changed after the pandemic onset, and the anxiolytic prescription rate increased by 0.9% per month (95% CI, 0.7%-  2).Our findings are consistent with a rise in psychotropic medication prescribing for children and adolescents during the 2 years after the pandemic onset compared with prepandemic rates and trends.
The most substantial increases in rates and trends of prescriptions concerned anxiolytic, hypnotic and sedative, and antidepressant medications.5][36] Before the pandemic onset, anxiolytics and hypnotics and sedatives were the only psychotropic medication classes with decreasing prescription trends in the pediatric population in France.This decrease probably accounts for the substantial relative increases observed after the pandemic onset even though prescription rates remained low overall, especially those of hypnotics and sedatives.Conversely, prepandemic rates of antidepressant prescriptions were already increasing, especially in adolescents, and this trend increase accelerated in the years following the pandemic.
We observed only moderate increases in prescription rates and trends for antipsychotics, as has been reported in previous studies. 10,37Antipsychotic medications are usually prescribed for the treatment of psychotic and conduct disorders and only as second-line treatments for major behavioral and bipolar disorders. 38,39Finally, despite a more steeply increasing trend after the pandemic onset, prescriptions of psychostimulants did not rise above expected rates in France, as opposed to other countries. 32,33,36This finding should be interpreted with caution, as the acceleration of the prescription trend may also be associated with a modification of methylphenidate prescribing conditions in France.Indeed, initiation of methylphenidate treatment was reserved for hospital pediatric specialists only until September 2021, when it was extended to include pediatric specialists in private practice. 40Despite this, psychostimulant prescription rates have not yet exceeded the expected rates because of a significant decrease-the largest among all of the psychotropic medication classes-in prescriptions in the initial pandemic period.Hence, prescribing of psychostimulant medications should continue to be monitored in the coming years.
It is unlikely that the observed increases in psychotropic medication prescribing were the consequence of increased screening and management of mood and anxiety disorders in children and adolescents after the pandemic, as availability of mental health services did not increase despite the development of telemedicine. 41Likewise, during the study period, there were no new psychotropic medications with a pediatric marketing authorization in France.Therefore, our findings likely reflect the actual mental health needs of children and adolescents after the onset of the pandemic, whether pre-existing or newly emerged.Indeed, the direct physical impact of the COVID-19 pandemic in children was low, but the indirect societal burden of the pandemic was considerable. 24,42Social isolation and school closures have been cited as main triggers of youth distress 23 ; in France, however, school closures were mandated only in the first few months of the pandemic.Undeniably, the pandemic was associated with boosted social media growth and online peer interactions. 43,44The pandemic was also associated with an important economic fallout and a rise in domestic violence. 45The negative association of these stressors in adolescents, who already experience important physiological and psychological turmoil, is obvious but should not be neglected for younger children. 43,46In our study, increases in anxiolytic, hypnotic and sedative, and antidepressant prescriptions were also important in this population and underline the vulnerability of young children to stressful environments, with potential major repercussions on their future psychological trajectories.

Strengths and Limitations
Our study has important strengths.First, we used a retail pharmacy dispensing database that readily extrapolates data at a national level.Furthermore, given that the rate of pediatric psychotropic medication prescriptions is known to be increasing in recent years, [47][48][49] our findings were based on a time-series statistical analysis accounting for these prepandemic rates and trends.This study has some limitations.First, previous studies have shown significant differences between men and women in the association between the pandemic and mental health and psychotropic medication use. 36,37,50The X-ponent database did not reliably record this information

Figure 1 .
Figure 1.Changes in Rates and Trends of Monthly Psychotropic Medication Prescriptions for Children and Adolescents in France

Figure 2 .
Figure 2. Rate Ratios (RRs) and 95% CIs of Estimated and Expected Monthly Rates of Psychotropic Medication Prescriptions for Children and Adolescents in the 2 Years Following Pandemic Onset (1.00-1.04)0.97 (0.95-0.98 ) Estimated monthly prescription rates were fitted by the model based on observed data, and expected monthly rates were forecast by the model based only on prepandemic observed data.Vertical dotted line indicates no relative change between estimated and expected rates (RR = 1).
Pediatric Psychotropic Medication Prescriptions and the COVID-19 PandemicTable.Changes in Rates and Trends of Psychotropic Medication PrescriptionsSimilar patterns of level and trend changes were observed in children and adolescents, although estimates were higher in adolescents than in children both before and after the pandemic onset (Table; eFigure 3 in Supplement 1).Monthly rates of anxiolytic prescriptions exceeded the expected rates by 30% (RR, 1.30 [95% CI, 1.25-1.35]) in adolescents and 11% (RR, 1.11 [95% CI, 1.09-1.12]) in children (Figure 1.1%), reaching 3.2 per 1000 children and adolescents in May 2022.Compared with the expected rates, a relative increase of 25% (RR, 1.25 [95% CI, 1.21-1.29])was observed.JAMA Network Open | Psychiatry a Level corresponds to the monthly prescription rate per 1000 children and adolescents estimated by the model.b Model estimates for trend and changes in level and trend are presented as percentage changes per month (95% CI).JAMA Network Open.2024;7(4):e247965.doi:10.1001/jamanetworkopen.2024.7965(Reprinted) April 23, 2024 5/14 Downloaded from jamanetwork.comby guest on 04/25/2024