Treatment Adherence and Health Care Utilization After a First Episode of Psychosis

Key Points

Question

What was the rate of antipsychotic medication adherence and its association with health care utilization among young individuals with first-episode psychosis?

Findings

In this cohort study of 11 889 individuals with first-episode psychosis, 19.1% met the adherence threshold (proportion of days covered, ≥0.8). Greater adherence was significantly associated with more behavioral health outpatient visits and fewer emergency department visits.

Meaning

This study suggests that improving adherence to antipsychotic medication may increase the continuity of outpatient care and reduce reliance on emergency services during early psychosis.

Abstract

Importance

Few studies have examined medication adherence immediately after initiation of antipsychotics among young individuals experiencing their first episode of psychosis (FEP).

Objectives

To estimate adherence to antipsychotic medications and examine factors associated with adherence among commercially insured individuals with FEP.

Design, Setting, and Participants

This retrospective cohort study used administrative claims from the Merative MarketScan commercial health insurance database in the US from January 1, 2013, to December 31, 2021. The study included individuals aged 15 to 40 years with 12 months of continuous enrollment who had no prior diagnosis of psychosis and who subsequently experienced FEP and were prescribed antipsychotics. Eligible patients were followed up for 365 days after their first antipsychotic prescription. Data analysis was performed from September to December 2025.

Exposure

The proportion of days covered was used to calculate adherence, defined as the number of covered medication days divided by the 365-day follow-up period.

Main Outcomes and Measures

Health care utilization outcomes included outpatient visits, behavioral health outpatient visits, general hospitalization, psychiatric hospitalization, and emergency department (ED) visits. Log-binomial regressions were used to estimate the association between demographic and clinical factors and adherence rate, and negative binomial regressions were used to examine the association between antipsychotic adherence and health care utilization.

Results

This study included 11 889 individuals who experienced FEP (mean [SD] age, 22.7 [6.8] years; 6708 male [56.4%]). The overall adherence rate to antipsychotics among the FEP population was 19.1% (2270 of 11 889). Compared with other psychotic spectrum disorders, greater adherence to antipsychotics was found among individuals with schizophrenia (prevalence ratio [PR], 1.78; 95% CI, 1.57-2.04; P < .001) and those with mood disorders (PR, 1.21 [95% CI, 1.06-1.39; P = .005]). Moreover, individuals with better adherence tended to have fewer ED visits (incidence rate ratio [IRR], 0.85; 95% CI, 0.81-0.90; P < .001), more behavioral health outpatient visits (IRR, 1.53; 95% CI, 1.47-1.59; P < .001), and more general outpatient visits (IRR, 1.07; 95% CI, 1.02-1.13; P = .003).

Conclusions and Relevance

In this cohort study of individuals with FEP, adherence to antipsychotics was low, with 19.1% of individuals classified as adherent. Moreover, poor adherence to antipsychotics was associated with more ED visits. Low treatment adherence rates may be a hurdle for initiatives aiming to improve psychosis treatment outcomes.

This cohort study examines adherence to antipsychotic medications and the association of adherence with health care utilization among young commercially insured individuals with first-episode psychosis.

Introduction

Psychotic spectrum disorders are a group of severe mental illnesses characterized by delusions, hallucinations, and declines in cognitive, educational, social, and vocational functioning.¹ In the US, the incidence of first-episode psychosis (FEP) is approximately 80.5 in 100 000 people aged 15 to 40 years,² compared with 50.0 per 100 000 in the general population.³ Overall, psychotic spectrum disorders typically emerge in late adolescence to early adulthood and can lead to significant life changes after diagnosis.⁴ People with psychosis often experience psychiatric hospitalization,^5,6 which not only makes psychotic spectrum disorders costly but also substantially affects quality of life and productivity.⁷ Moreover, evidence indicates that adolescents and young adults with FEP had a 6.92-fold higher hazard of all-cause mortality than those in the general outpatient population, underscoring the importance of intervention for this population.⁸

To treat and manage psychotic spectrum disorders, specialized early intervention services that provide comprehensive outpatient care as soon as possible after the first onset of psychotic symptoms have been shown to effectively improve patient health outcomes, such as reduction in hospitalization and relapse rates.⁹ One of the key interventions included in such early intervention services is antipsychotic medication, and previous research has demonstrated that antipsychotic medications are associated with improved quality of life and lowered relapse risk.^10,11,12,13 However, adherence to antipsychotic treatment remains low among patients with FEP.¹⁴ Poor adherence has been associated with younger age, low income, and non-White race and ethnicity, often due to financial burden, stigma, and medication adverse effects.^15,16,17 Although these factors are clearly associated with poor adherence to antipsychotic medications, patients often experience failure with initial treatments, resulting in subsequent medication changes and increased risk of early discontinuation of newly prescribed medication.^11,18,19 Patients who fail to adhere to antipsychotic treatment are at increased risk of relapse of acute psychosis and psychiatric hospitalization.^20,21 Moreover, nonadherence has been associated with greater emergency department (ED) utilization, suicide attempts, and higher all-cause mortality.²² Maintaining adherence to antipsychotic medications is particularly important among adolescents and young individuals, as consistent treatment can improve engagement with care, support long-term disease management, and mitigate the adverse consequences of psychosis.

Previous studies have found a moderate adherence rate to antipsychotic agents among patients with FEP; however, these studies were not based on claims data and were limited by small sample sizes.^{14,23,24,25,26} These studies showed that the adherence rate of antipsychotics in patients ranged from 21.1% to 80%. However, little is known about the adherence rate of antipsychotics among young individuals with commercial health insurance and FEP initiating antipsychotic agents and the factors associated with subsequent health care utilization. This study aimed to (1) examine the adherence rate of young individuals with FEP taking antipsychotic medications and (2) examine the association between the adherence rate and health care utilization among commercial health insurance beneficiaries with FEP in the US.

Methods

Study Design and Data Source

This retrospective cohort study used Merative MarketScan data to examine antipsychotic medication adherence among individuals who experienced a first onset of a psychotic spectrum disorder between January 1, 2013, and December 31, 2021, and who subsequently began treatment with an antipsychotic agent. The MarketScan database contains deidentified individual-level outpatient, inpatient, and pharmacy claims from employees, retirees, and their dependents covered by employer-sponsored health insurance plans across the US. The database includes health care utilization data for approximately 30 million to 50 million commercially insured individuals annually but does not include claims from individuals covered by public insurance programs, such as Medicaid or Medicare.²⁷ Although the MarketScan population is not fully representative of the US population, it is broadly representative of the commercially insured population because the employer-sponsored health insurance plans included in the database are similar to those covering nearly two-thirds of individuals in the US with private health insurance.²⁸ The study was exempted from approval and informed consent by the institutional review board at The Ohio State University because this study used deidentified administrative claims data. We followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

Cohort Selection

We identified individuals aged 15 to 40 years with 12 months of continuous insurance enrollment during which they had no diagnosis of a schizophrenia spectrum disorder, mood disorder with psychotic features, or other psychotic spectrum disorders by International Classification of Diseases, Ninth Revision (ICD-9), or International Statistical Classification of Diseases and Related Health Problems, Tenth Revision (ICD-10), codes (eTable 1 in Supplement 1). ICD-9 codes were converted to ICD-10 codes using the Centers for Medicare & Medicaid Services General Equivalence Mappings.^29,30 Individuals with FEP were identified as having at least 2 psychosis claims recorded on separate days, including the index claim. Claims from either inpatient or outpatient settings were eligible.³¹ We then identified individuals who initiated antipsychotic treatment after receipt of the qualifying diagnosis. A total of 27 592 patients with FEP were identified, of whom 11 889 (43.1%) subsequently received their first prescription of an antipsychotic medication. Figure 1 illustrates the cohort selection process. Antipsychotic medications were identified from outpatient pharmacy claims using National Drug Codes (eTable 2 in Supplement 1). The list of antipsychotic medications was based on the National Committee for Quality Assurance 2022 Health Effectiveness Data and Information Set Medication List Directory.³²

Figure 1. Cohort Selection Flow Diagram.

Measurement of Adherence to Antipsychotics

The main outcome was adherence to antipsychotic medications, measured over the 12-month period after the first antipsychotic prescription (index prescription) using the proportion of days covered (PDC) method. The first antipsychotic prescription was used to define the index date and the start of follow-up. PDC was calculated as the proportion of days covered by antipsychotic medications during the 365-day follow-up period.³³ For individuals prescribed multiple antipsychotic medications simultaneously, if a patient refilled the initial antipsychotic before the days’ supply of the previous fill was exhausted, overlapping days’ supply was counted only once. PDC ranged from 0 to 1, and a higher PDC value represents better medication adherence. We also measured adherence as a binary outcome, with individuals with PDC values of 0.8 or greater considered to be adherent.^34,35

We measured health care utilization within 12 months after the first prescription date using the number of behavioral health outpatient visits, general outpatient visits, psychiatric hospitalizations, general hospitalizations, and ED visits. Behavioral health outpatient visits were categorized by practice place and clinician type.

Statistical Analysis

Data analysis was performed from September to December 2025. The adherence rate was estimated by descriptive analyses using means and SDs for continuous variables and frequencies and percentages for categorical variables. We used a multivariable log-binomial regression to examine factors associated with the categorical measure of adherence to antipsychotics and estimated prevalence ratios (PRs) and 95% CIs. The model included age at the first psychosis diagnosis (categorized as 15-17, 18-25, and 26-40 years), sex (male and female), case type of psychoses (schizophrenia spectrum disorders, mood disorder, and other psychotic spectrum disorders), geographic regions (Northeast, North Central, South, West, and Unknown), Charlson Comorbidity Index (CCI) score (0, 1, 2, and ≥3), and behavioral health conditions, including anxiety disorders, autism spectrum disorders, mood disorders without psychotic features, posttraumatic stress disorder, and substance use disorder in the baseline period.³⁶ These behavioral health conditions were identified using ICD-9, Clinical Modification, and ICD-10, Clinical Modification, diagnosis codes (eTable 3 in Supplement 1). The CCI score was developed to predict 1-year mortality based on risk factors (eg, age and myocardial infarction). The cumulative score increases with a higher burden of comorbid conditions.³⁷ In addition, the association between the binary outcome of adherence and health care utilization, including number of behavioral health outpatient visits, general outpatient visits, and ED visits, was examined by negative binomial regression and estimated incidence rate ratios (IRRs) and 95% CIs. The analytic variables included in this study did not contain missing values because the MarketScan database contains fully adjudicated and standardized medical and pharmacy claims, and annual database releases occur after nearly all claims have been processed, minimizing incomplete claims records and improving data reliability.³⁸ To provide a more comprehensive assessment of medication use, we conducted a sensitivity analysis that estimated adherence using major psychotropic medication classes, including antipsychotics, antidepressants, and mood stabilizers (eTable 4 in Supplement 1). Analyses were conducted using R, version 4.5.2 (R Project for Statistical Computing), with conventional statistical significance thresholds of 2-sided P < .05.

Results

This national sample included 11 889 individuals who experienced FEP and initiated antipsychotic treatment (mean [SD] age, 22.7 [6.8] years; 6708 male [56.4%] and 5181 female [43.6%]) (Table 1). A total of 2118 patients (17.8%) resided in the Northeastern US and 9110 patients (76.6%) had a CCI score of 0. The most common psychosis spectrum disorder in our study population was schizophrenia spectrum disorder (5111 [43.0%]), followed by mood disorders with psychotic features (5058 [42.5%]) and other psychotic spectrum disorders (1720 [14.5%]). At baseline, anxiety disorders were the most common behavioral health condition (3949 [33.2%]), followed by substance use disorders (1700 [15.0%]), posttraumatic stress disorder (394 [3.3%]), and autism spectrum disorder (175 [1.5%]).

Table 1. Sample Characteristics at Baseline.

Characteristic	No. (%) (N = 11 889)
Age, mean (SD), y	22.7 (6.8)
Age group, y
15-17	2517 (21.2)
18-25	6465 (54.4)
26-40	2907 (24.5)
Sex
Male	6708 (56.4)
Female	5181 (43.6)
Diagnosis group
Schizophrenia spectrum disorder	5111 (43.0)
Mood disorders with psychotic features	5058 (42.5)
Other psychotic spectrum disorders	1720 (14.5)
Region
Northeast	2118 (17.8)
North Central	2681 (22.6)
South	5153 (43.3)
West	1883 (15.8)
Unknown	54 (0.5)
Baseline CCI scorea
0	9110 (76.6)
1	2183 (18.4)
2	442 (3.7)
≥3	154 (1.3)
Behavioral health conditions
Anxiety disorders	3949 (33.2)
Autism spectrum disorders	175 (1.5)
Mood disorders without psychotic features	1724 (14.5)
Posttraumatic stress disorder	394 (3.3)
Substance use disorders	1700 (15.0)

Abbreviation: CCI, Charlson Comorbidity Index.

^a The cumulative score increases with a higher burden of comorbid conditions.

We found that individuals with FEP adhered to their antipsychotic agents for 43.3% of days (158 of 365 days) in the 12 months after antipsychotic initiation. Table 2 demonstrates the distribution of adherence rates to antipsychotics and is classified by diagnostic group. The overall adherence rate to antipsychotics was 19.1% (2270 of 11 889), with substantial variation by type of psychosis. The adherence rates were 23.1% (1181 of 5111) for individuals with schizophrenia spectrum disorder, 16.9% (857 of 5058) for individuals with mood disorders with psychotic features, and 13.5% (232 of 1720) for individuals with other psychotic spectrum disorders.

Table 2. Distribution of Adherence to Antipsychotic Agents Across Different Diagnostic Groups.

Characteristic	Schizophrenia spectrum disorder (n = 5111)	Mood disorders with psychotic features (n = 5058)	Other psychotic spectrum disorders (n = 1720)	All individuals (N = 11 889)
PDC, mean (SD)	0.5 (0.3)	0.4 (0.3)	0.3 (0.3)	0.4 (0.3)
Adherence rate, No. (%)a	1181 (23.1)	857 (16.9)	232 (13.5)	2270 (19.1)

Abbreviation: PDC, proportion of days covered.

^a Adherence rate was defined as a PDC of 0.8 or more.

Figure 2 presents PRs from a multivariable log-binomial regression of antipsychotic medications. Compared with individuals with other psychotic spectrum disorders, individuals with schizophrenia spectrum disorders (PR, 1.78; 95% CI, 1.57-2.04; P < .001) and mood disorders with psychotic features (PR, 1.21; 95% CI, 1.06-1.39; P = .005) had higher prevalences of adherence to antipsychotics. No significant difference was found for sex, CCI score, autism spectrum disorders, and posttraumatic stress disorder.

Figure 2. Dot Plot of Association Between Sample Characteristics and Antipsychotic Medication Adherence Using Log-Binomial Regression Models.

The prevalences of adherence to antipsychotics and demographic and clinical factors are shown. Compared with individuals with other psychotic spectrum disorders, individuals with schizophrenia spectrum disorder and mood disorder with psychotic features had higher prevalences of adherence to antipsychotics. CCI indicates Charlson Comorbidity Index.

^aThe reference category for each behavioral health condition is the absence of that condition (ie, the reference for anxiety disorder is no anxiety disorder).

Figure 3 shows that individuals who were adherent to antipsychotics (PDC, ≥0.8) experienced fewer ED visits (IRR, 0.85; 95% CI, 0.81-0.90; P < .001), more behavioral health outpatient visits (IRR, 1.53; 95% CI, 1.47-1.59; P < .001), and more general outpatient visits (IRR, 1.07; 95% CI, 1.02-1.13; P = .003). We found no statistically significant association between adherence to antipsychotics and psychiatric hospitalization (IRR, 0.99; 95% CI, 0.83-1.17; P = .88) or general hospitalization (IRR, 0.96; 95% CI, 0.88-1.05; P = .39).

Figure 3. Dot Plots of Health Care Utilization Outcomes Associated With Antipsychotic Adherence .

Incidence rate ratios (IRRs) for health care utilization outcomes associated with antipsychotic adherence are shown. The results showed that individuals who adhered to antipsychotic medications had fewer emergency department visits, more behavioral outpatient visits, and more general outpatient visits. No statistically significant associations were observed between antipsychotic adherence and psychiatric or general hospitalization. CCI indicates Charlson Comorbidity Index.

^aThe reference category for each behavioral health condition is the absence of that condition (ie, the reference for anxiety disorder is no anxiety disorder).

The sensitivity analysis that estimated adherence using only major psychotropic medication classes, including antipsychotics, antidepressants, and mood stabilizers (eTable 4 in Supplement 1), found that the adherence rate increased from 19.1% (2270 of 11 889) with antipsychotics only to 32.7% (3891 of 11 889) with major psychotropic medication classes. In addition, 11 337 of 11 889 individuals (95.4%) in this analytic cohort received oral antipsychotics exclusively, while 59 of 11 889 (0.5%) received long-acting injectable (LAI) formulations exclusively and 461 of 11 889 (3.9%) used both oral and LAI formulations (eTable 5 in Supplement 1).

Discussion

This study used a comprehensive national database to investigate the adherence rate to antipsychotic medications among individuals with FEP and examined the association between adherence to antipsychotics and health care utilization. We found that individuals aged 15 to 40 years with FEP adhered to their antipsychotic agents for 43.3% of days in the 12 months after antipsychotic initiation, and only 19.1% met the threshold for adherence (PDC, ≥0.8) for the entire 12-month follow-up. In contrast, prior studies using the same PDC measure reported adherence rates of 35.4% to 52.5% among individuals with chronic schizophrenia.^17,33 Moreover, among our sample, better adherence was associated with more behavioral health outpatient visits and general outpatient visits and fewer ED visits.

One possible explanation for this low adherence rate is the younger age and early disease stage of our study population. Previous studies based on Medicare and Medicaid databases reported higher antipsychotic adherence rates than those observed in our study.^17,33 However, participants in those studies were substantially older, with mean ages ranging from 42.9 to 55 years, whereas the mean age in our cohort was 22.7 years. Younger individuals may have limited insight into illness severity, greater uncertainty about long-term medication treatment, and lower perceived necessity for sustained medication use or follow-up care, all of which may contribute to failure to initiate or continue antipsychotic treatment.^39,40 Moreover, concerns about medication-related adverse effects may further discourage treatment initiation or continuation among young people with FEP. Antipsychotic medications are associated with several adverse effects, including weight gain, sexual adverse effects, sedation, altered glucose metabolism, and increased long-term cardiovascular disease risk,³⁵ which may be particularly concerning during adolescence and young adulthood.⁴¹ These potential adverse effects may outweigh the perceived benefits of symptom control, especially when psychotic symptoms are intermittent or less severe.

Another explanation for the low adherence is that our study population represents a different population compared with previous studies. Prior studies used Medicare and Medicaid data,^17,33 which include individuals with more comorbidities, lower socioeconomic status, and longer engagement within the health care system. These individuals are more likely to be connected with ongoing care and familiar with available services. In contrast, our study focused on individuals with a new diagnosis of FEP who were initiating antipsychotic treatment for the first time within a commercially insured population and who may have limited familiarity with the health care system and fewer established connections with mental health services.⁴² As a result, they may face greater challenges in sustaining treatment, which could contribute to the lower adherence rates.

To provide a more comprehensive assessment of medication use, we conducted a sensitivity analysis that estimated adherence using major psychotropic medication classes, including antipsychotics, antidepressants, and mood stabilizers (eTable 4 in Supplement 1). The adherence rate increased from 19.1% with antipsychotics only to 32.7% with major psychotropic medication classes, suggesting that low antipsychotic adherence does not necessarily reflect complete disengagement from medication treatment. This finding may reflect changes in medication use during the early course of psychosis, including transitions across psychotropic medication classes in response to evolving symptoms or adverse effects.⁴³ Future studies should further characterize longitudinal medication trajectories during early psychosis to identify opportunities for interventions that improve long-term treatment continuity and guideline-concordant care.

In addition to patient-level factors, treatment modality may also contribute to the low adherence. LAI formulations of antipsychotics have been shown to improve adherence and reduce relapse rates.^44,45 However, only 0.5% of our cohort received LAI formulations exclusively and 3.9% used both oral and LAI formulations (eTable 5 in Supplement 1), indicating that LAI antipsychotic use at the time of treatment initiation among individuals with FEP remained uncommon. This small sample limited the analysis of the use of LAI antipsychotics, thereby restricting our ability to conduct robust multivariable analyses. Most individuals with FEP relied on oral antipsychotic formulations rather than LAIs, which may represent 1 potential factor associated with the low adherence rates observed. Previous research found that psychiatrists may assume that individuals with FEP are reluctant to accept LAI antipsychotics and offer this option less frequently, partly because some patients perceive LAI formulations as more coercive and stigmatizing.⁴⁶ Although prior studies have shown that LAI antipsychotics are significantly associated with reduced readmission rates and longer time to hospitalization compared with oral formulations,^{47,48,49,50,51,52} our findings show a low uptake of LAI antipsychotics nationwide for individuals with FEP. Given the consistent evidence supporting the clinical benefits associated with LAI formulations, this underutilization highlights a potential gap in early treatment strategies. Greater consideration of the use of LAI antipsychotics during the initial treatment phase may improve adherence and reduce reliance on acute care services.

Our findings align with previous research demonstrating that individuals who adhere to antipsychotic medications had lower use of ED services and reduced acute health care utilization.^53,54,55 Mental health–related ED visits are often viewed as indicators of symptom exacerbation or relapse, particularly among individuals experiencing worsening psychotic symptoms.^56,57,58 Care delivered in ED settings typically prioritizes acute stabilization and short-term symptom relief rather than long-term disorder management.⁵⁹ Over the long term, sustained behavioral health outpatient engagement and consistent adherence to antipsychotic medications are essential for effective symptom management, relapse prevention, and reduced reliance on emergency services.⁴⁴ Decreased adherence therefore appears to be associated with a shift in the site of care from outpatient services toward more intensive acute settings such as ED visits.^60,61 Continued reliance on ED-based care may be challenging and costly and could adversely affect patients’ quality of life.⁶² This is because emergency services are not structured to provide ongoing management; repeated reliance on ED care may contribute to fragmented treatment, limit continuity with outpatient clinicians, and disrupt educational, vocational, and social stability. Recurrent acute encounters may also hinder the development of sustained therapeutic relationships and delay long-term recovery planning.⁶³ These challenges may be further compounded by limitations in the availability of inpatient psychiatric services. Recent evidence indicates that psychiatric bed capacity in short-term acute care hospitals has declined over time and remains unavailable in many communities despite persistently high mental health needs.⁶⁴ These system-level limitations highlight the importance of optimizing outpatient management, including sustained antipsychotic adherence and continuity of behavioral health care, to reduce reliance on emergency services and improve outcomes among individuals with FEP.

Strengths and Limitations

Our study has several strengths. First, previous studies focused on Medicare, Medicaid, and Veterans Affairs populations, which comprised older individuals with schizophrenia. Although several studies have examined antipsychotic adherence early in the course of a psychotic disorder, no study to our knowledge has systematically examined antipsychotic medication adherence among individuals who initiated antipsychotic treatment after a diagnosis of FEP. As a result, our study addressed this gap and included adolescents and young adults with FEP newly receiving prescriptions, which is a group at high risk of poor adherence and relapse. Second, this study draws on a large, national commercial claims database with private-payer claims from approximately 230 million unique deidentified patients (30 million-50 million covered annually).²⁷ Third, our study examined the association between adherence to antipsychotics and health care utilization, which provides a deeper understanding of clinical consequences of adherence among young people with FEP.

This study also has some limitations. First, our study used claims data, so we could not completely capture medications excluded from insurance claims or administered in hospitalization settings. Claims data also do not reflect patients’ medication-taking behaviors that may impact their adherence to antipsychotics. Moreover, PDC is based on prescription fills and represents an upper limit of adherence, as having medication available does not guarantee that it was taken as prescribed. Second, although the MarketScan database includes individuals with commercial health insurance and reflects health plans similar to those covering nearly two-thirds of US individuals with private health insurance, our findings may not be generalizable to individuals covered by public health insurance programs.

Conclusions

In this cohort study of individuals with FEP, the adherence rate to antipsychotic medication was 19.1%. People with schizophrenia and mood disorders tended to have higher adherence rates, and better adherence to antipsychotic medication was associated with more behavioral health outpatient visits and fewer ED visits. These findings could underscore the importance of early prescribing to prevent worse psychosis-related prognoses that are costly and associated with substantial productivity loss. It suggests that the importance of early prescribing and improving outpatient continuity may be effective in reducing acute care utilization and long-term outcomes.

Supplement 1.

eTable 1. Diagnostic Codes for First-Episode Psychosis

eTable 2. List of Antipsychotic Medications

eTable 3. Diagnostic Codes for Behavioral Health Conditions

eTable 4. Comparison of Adherence Estimates Using Antipsychotics Alone vs Major Psychotropic Medication Classes

eTable 5. Distribution of Antipsychotic Administration Routes

Supplement 2.

Data Sharing Statement