Faculty Bibliography

PURPOSE/OBJECTIVE:To estimate the prevalence and geographic variation of serious mental illness (SMI) among young adults enrolled in Medicaid and to identify individual and community-level factors associated with SMI. METHODS:We used New York Medicaid data for over 1.6 million young adults aged 18-34 years with continuous Medicaid enrollment from April 2021 through March 2022. We merged 3 additional county and zip code-level datasets, including the Distressed Communities Index, Area Health Resource File, and New York homelessness data. We used descriptive analyses, data visualization methods, and multivariable logistic regression to estimate the prevalence of SMI and identify individual- (sex, race and ethnicity, age, disability aid, and homelessness) and community-level (geography, community distress, number of hospital beds, and community mental health centers per capita) factors associated with SMI. RESULTS:The 12-month prevalence of SMI was 8.3% in New York State, with wide geographic variation by zip codes, ranging from 0% to 39%. Young adults identifying as Asian, Black, and Latinx (compared to White), and those living in zip codes with the most community distress were associated with lower odds of SMI, odds ratios (OR) ranged from 0.47 (Asian) (95% confidence interval [CI] [0.46-0.49]) to 0.95 (Latinx) (CI [0.93-0.97]). Young adults identifying as female (compared to male), receiving disability, experiencing homelessness, and rural residence were associated with higher odds of SMI, ORs ranged from 1.64 (female) (95% CI [1.62-1.66]) to 5.49 (homelessness) (95% CI [5.33-5.66]). DISCUSSION/CONCLUSIONS:SMI prevalence varies by individual- and community-level factors, including homelessness and rural residence, signaling the need for specific interventions and policies.

OBJECTIVE/UNASSIGNED:This study assessed the utility and effectiveness of the new general health integration (GHI) framework among community behavioral health organizations designated as certified community behavioral health clinics (CCBHCs) or in the process of applying to become a CCBHC. METHODS/UNASSIGNED:Nineteen licensed community behavioral health clinics, 18 of which had CCBHC status, participated in a 12-month learning collaborative. They used the GHI framework to assess their integration stage for 15 subdomains within eight domains of evidence-based practice. The clinics worked to improve their GHI practices with the support of monthly learning collaborative webinars, individual consultation calls, and technical assistance sessions. Clinics reported on performance quality metrics aligned with national CCBHC standards. Outcome measures included GHI framework scores at baseline and 1-year follow-up, capacity to report quality metrics at baseline and at the end of the collaborative, and average performance on the quality metrics at baseline versus at the end of the collaborative. RESULTS/UNASSIGNED:Clinics showed overall improvement in integration stage over the study period. Of note, higher baseline GHI framework scores demonstrated a significant association with greater-quality performance at baseline (r=0.577, p=0.024) and follow-up (r=0.782, p=0.001). Capacity to track and report quality metrics increased significantly during the learning collaborative, as did average performance on quality metrics. CONCLUSIONS/UNASSIGNED:Community behavioral health clinics using the GHI framework were able to advance their GHI practices with a 12-month learning collaborative project. The framework has the potential to serve as a useful tool for clinics aiming to enhance GHI practices.

IMPORTANCE/UNASSIGNED:Delivery of mental health care through telehealth (telemental health care) increased after the onset of the COVID-19 pandemic. Little is known about the speed of adoption (diffusion) of telemental health in the care in the care of individuals with schizophrenia. OBJECTIVES/UNASSIGNED:To characterize telemental health care diffusion in mental health agencies serving Medicaid beneficiaries with schizophrenia and the beneficiary-level association of telemental health care use with race and ethnicity. DESIGN, SETTING, AND PARTICIPANTS/UNASSIGNED:This retrospective cohort study used New York State Medicaid data from March 1, 2019, to February 29, 2020 (prepandemic period), and from March 11, 2020, to March 31, 2021 (pandemic period), from 261 agencies serving 30 990 beneficiaries with schizophrenia with 1 or more mental health visits during the pandemic period. Statistical analysis was performed from November 2021 through September 2024. EXPOSURE/UNASSIGNED:Agency percentage of patients belonging to racial and ethnic minority groups among all Medicaid-covered patients between March 2019 and February 2020; agency type, categorized as freestanding, hospital affiliated, or state operated; beneficiary-level race and ethnicity, categorized as Asian or other (American Indian or Alaska Native; Native Hawaiian or Other Pacific Islander), Black, Latinx, White, and unknown; and pandemic severity, operationalized as COVID-19 hospitalization rates per 10 000 population in administratively defined catchment areas. MAIN OUTCOMES AND MEASURES/UNASSIGNED:Days to 10% cumulative telemental health care use within agencies, as well as beneficiary-level time to first telemental health care visit and any telemental health care visit in catchment areas in times of varying pandemic severity. RESULTS/UNASSIGNED:In this cohort study of 261 agencies (18 [7%] state operated, 79 [30%] hospital affiliated, and 164 [63%] free standing) and 30 990 beneficiaries with schizophrenia (mean [SD] age, 43 [13] years; 59% male; 7% Asian or other, 38% Black, 20% Latinx, and 25% White), 6 agencies (2%) never adopted telemental health care, and 248 (95%) reached 10% cumulative telemental health care visits in a mean of 18 days. Mean (SD) agency prepandemic shares of beneficiaries belonging to racial or ethnic minority groups (56% [23%]) were not associated with telemental health care diffusion. Diffusion was faster in state-operated vs free-standing agencies (hazard ratio [HR], 2.44 [95% CI, 1.21-4.95]). Relative to White beneficiaries, time to first telemental health care visit was slower in every racial and ethnic minority group (Asian or other: HR, 0.93 [95% CI, 0.88-0.98]; Black: HR, 0.90 [95% CI, 0.87-0.93]; Latinx: HR, 0.95 [95% CI, 0.91-0.99]). Beneficiaries from at least 1 racial or ethnic minority group were less likely than White beneficiaries to have a telemental health care visit regardless of pandemic severity and area; differences narrowed when pandemic severity was higher (eg, in New York City, the odds ratio of Black beneficiaries having a telemental health care visit relative to White beneficiaries when the pandemic severity was high was 0.70 [95% CI, 0.63-0.79] but decreased to 0.59 [95% CI, 0.53-0.67] when the pandemic severity was low). CONCLUSIONS AND RELEVANCE/UNASSIGNED:In this cohort study of Medicaid beneficiaries with schizophrenia, telemental health care diffused rapidly after the onset of the COVID-19 pandemic, particularly in state-operated agencies. Together, agency-level and beneficiary-level race and ethnicity findings suggest within-agency racial and ethnic differences in diffusion of telemental health care. States should monitor the diffusion of innovations across vulnerable populations.

OBJECTIVE/UNASSIGNED:The authors examined the prevalence and correlates of co-occurring opioid use disorder and opioid overdose among individuals receiving psychiatric services. METHODS/UNASSIGNED:This was a cross-sectional study of adults with continuous enrollment in New York State Medicaid who received at least one psychiatric service in 2020 (N=523,885). Logistic regression models were used to examine the correlates of both opioid use disorder and overdose. RESULTS/UNASSIGNED:In the study sample, the prevalence rate of opioid use disorder was 8.1%; within this group, 7.7% experienced an opioid overdose in the study year. Opioid use disorder rates were lower among younger (18-24 years; 2.0%) and older (≥65 years; 3.1%) adults and higher among men (11.1%) and among those residing in rural areas (9.9%). Compared with Whites (9.4%), opioid use disorder rates were lower for Asian Americans (2.0%, adjusted odds ratio [AOR]=0.22) and Blacks (6.8%, AOR=0.76) and higher for American Indians (13.2%, AOR=1.43) and Hispanics (9.6%, AOR=1.29). Individuals with any substance use (24.9%, AOR=5.20), posttraumatic stress (15.7%, AOR=2.34), bipolar (14.9%, AOR=2.29), or anxiety (11.3%, AOR=2.18) disorders were more likely to have co-occurring opioid use disorder; those with conduct (4.5%, AOR=0.51), adjustment (7.4%, AOR=0.88), or schizophrenia spectrum (7.4%, AOR=0.87) disorders were less likely to have opioid use disorder. Those with suicidality (23.9%, AOR=3.83) or economic instability (23.7%, AOR=3.35) had higher odds of having opioid use disorder. Overdose odds were higher among individuals with suicidality (34.0%, AOR=6.82) and economic instability (16.0%, AOR=2.57). CONCLUSIONS/UNASSIGNED:These findings underscore the importance of providing opioid use disorder screening and treatment for patients receiving psychiatric services.

Given the chronic nature of schizophrenia, it is important to examine age-specific prevalence and incidence to understand the scope of the burden of schizophrenia across the lifespan. Estimates of lifetime prevalence of schizophrenia have varied widely and have often relied upon community-based data estimates from over two decades ago, while more recent studies have shown considerable promise by leveraging pooled datasets. However, the validity of measures of schizophrenia, particularly new onset schizophrenia, has not been well studied in these large health databases. The current study examines prevalence and validity of incidence measures of new diagnoses of schizophrenia in 2019 using two U.S. administrative health databases: MarketScan, a national database of individuals receiving employer-sponsored commercial insurance (N = 16,365,997), and NYS Medicaid, a large state public insurance program (N = 4,414,153). Our results indicate that the prevalence of schizophrenia is over 10-fold higher, and the incidence two-fold higher, in the NYS Medicaid population compared to the MarketScan database. In addition, prevalence increased over the lifespan in the Medicaid population, but decreased in the employment based MarketScan database beginning in early adulthood. Incident measures of new diagnoses of schizophrenia had excellent validity, with positive predictive values and specificity exceeding 95%, but required a longer lookback period for Medicaid compared to MarketScan. Further work is needed to leverage these findings to develop robust clinical outcome predictors for new onset of schizophrenia within large administrative health data systems.

School-based mental health clinics are the ideal venue to provide mental health services for youth due to their accessibility and lack of stigma compared with other community treatment centers. There were challenges associated with the abrupt shift to remote education and clinic services caused by COVID-19, but some school-based mental health clinics excelled in their implementation of telehealth services. This study of New York schools found four main facilitators to implementing telehealth services among school-based mental health clinics: (1) strong collaborative relationships between the clinic and school setting, (2) active and responsive leadership, (3) provider experience in conducting telehealth, and (4) provider flexibility to accommodate the needs of children and families. Factors were identified at every level of the school-based mental health clinic and system that helped to facilitate exemplary telehealth service delivery and implementation.

Importance/UNASSIGNED:Individuals with serious mental illness are at increased risk of severe COVID-19 infection. Several psychotropic medications have been identified as potential therapeutic agents to prevent or treat COVID-19 but have not been systematically examined in this population. Objective/UNASSIGNED:To evaluate the associations between the use of psychotropic medications and the risk of COVID-19 infection among adults with serious mental illness receiving long-term inpatient psychiatric treatment. Design, Setting, and Participants/UNASSIGNED:This retrospective cohort study assessed adults with serious mental illness hospitalized in a statewide psychiatric hospital system in New York between March 8 and July 1, 2020. The final date of follow-up was December 1, 2020. The study included 1958 consecutive adult inpatients with serious mental illness (affective or nonaffective psychoses) who received testing for SARS-CoV-2 by reverse transcriptase-polymerase chain reaction or antinucleocapsid antibodies and were continuously hospitalized from March 8 until medical discharge or July 1, 2020. Exposures/UNASSIGNED:Psychotropic medications prescribed prior to COVID-19 testing. Main Outcomes and Measures/UNASSIGNED:COVID-19 infection was the primary outcome, defined by a positive SARS-CoV-2 reverse transcriptase-polymerase chain reaction or antibody test result. The secondary outcome was COVID-19-related death among patients with laboratory-confirmed infection. Results/UNASSIGNED:Of the 2087 adult inpatients with serious mental illness continuously hospitalized during the study period, 1958 (93.8%) underwent testing and were included in the study; 1442 (73.6%) were men, and the mean (SD) age was 51.4 (14.3) years. A total of 969 patients (49.5%) had laboratory-confirmed COVID-19 infection that occurred while they were hospitalized; of those, 38 (3.9%) died. The use of second-generation antipsychotic medications, as a class, was associated with decreased odds of infection (odds ratio [OR], 0.62; 95% CI, 0.45-0.86), whereas the use of mood stabilizers was associated with increased odds of infection (OR, 1.23; 95% CI, 1.03-1.47). In a multivariable model of individual medications, the use of paliperidone was associated with decreased odds of infection (OR, 0.59; 95% CI, 0.41-0.84), and the use of valproic acid was associated with increased odds of infection (OR, 1.39; 95% CI, 1.10-1.76). Clozapine use was associated with reduced odds of mortality in unadjusted analyses (unadjusted OR, 0.25; 95% CI, 0.10-0.62; fully adjusted OR, 0.43; 95% CI, 0.17-1.12). Conclusions and Relevance/UNASSIGNED:In this cohort study of adults hospitalized with serious mental illness, the use of second-generation antipsychotic medications was associated with decreased risk of COVID-19 infection, whereas the use of valproic acid was associated with increased risk. Further research is needed to assess the mechanisms that underlie these findings.

OBJECTIVE:This study examined the impact of Pathway Home™ (PH) transition services for high utilizers of psychiatric hospitalization on inpatient days and outpatient engagement post-hospital discharge. METHODS:This case series study of forty PH graduates (5/22/2015-8/31/2018) used Medicaid claims to assess psychiatric inpatient days-per-month, average proportion of months with psychiatric emergency room, outpatient, and health home care management services. T-tests compared three time periods: the year prior, during, and after enrollment. RESULTS:Graduates had significantly fewer psychiatric inpatient days/month during (M = 1.84, p < 0.001) and after PH enrollment (M = 1.88, p < 0.001) compared to prior to enrollment (M = 7.1), while emergency services were stable. Outpatient visits increased from 45% prior to 76% during enrollment (p < 0.001) and was sustained on follow-up (67%, p = 0.008). A similar pattern emerged for health home services (32%, 60%, and 50%). CONCLUSION/CONCLUSIONS:PH is a promising approach for improving outcomes for high utilizers of psychiatric inpatient services, with sustained impact on follow-up.

BACKGROUND:Behavioral health outpatients are at risk for self-harm. Identifying individuals or combination of risk factors could discriminate those at elevated risk for self-harm. METHODS:The study population (N=248,491) included New York State Medicaid-enrolled individuals aged 10 to 64 with mental health specialty clinic visits 11/1/15-11/1/16. Self-harm episodes were defined using ICD-10 codes from emergency department and inpatient visits. Multi-predictor logistic regression models were fit on a subsample of the data and compared to a testing sample based on discrimination performance (Area Under the Curve or AUC). RESULTS:Of N=248,491 patients, 4,224 (1.70%) had an episode of intentional self-harm. Factors associated with increased self-harm risk were age17-25, being female and having recent diagnoses of depression (AOR=4.3, 95%CI: 3.6-5.0), personality disorder (AOR=4.2, 95%CI: 2.9-6.1), or substance use disorder (AOR=3.4, 95%CI: 2.7-4.3) within the last month. A multi-predictor logistic regression model including demographics and new psychiatric diagnoses within 90 days prior to index date had good discrimination and outperformed competitor models on a testing sample (AUC=0.86, 95%CI:0.85-0.87). LIMITATIONS/CONCLUSIONS:New York State Medicaid data may not be generalizable to the entire U.S population. ICD-10 codes do not allow distinction between self-harm with and without intent to die. CONCLUSIONS:Our results highlight the usefulness of recency of new psychiatric diagnoses, in predicting the magnitude and timing of intentional self-harm risk. An algorithm based on this finding could enhance clinical assessments support screening, intervention and outreach programs that are at the heart of a Zero Suicide prevention model.

OBJECTIVE:Many public health approaches to suicide prevention emphasize connecting at-risk individuals to professional treatment. However, it is unclear to what degree the outpatient mental health workforce has the requisite knowledge and skills to provide the evidence-based care needed to help those at risk. In this project, prior to the implementation of a statewide suicide prevention initiative, we assessed the baseline suicide prevention training and clinical practices of the New York State outpatient mental health workforce, a group likely representative of the broader U.S. clinical workforce. METHOD:A workforce survey of suicide prevention training and clinical practices was administered to 2,257 outpatient clinicians, representing 169 clinics serving approximately 90,000 clients. Clinicians were asked to complete the survey online, and all responses were confidential. RESULTS:Clinicians reported substantial gaps in their suicide prevention knowledge and training. The vast majority reported moderate self-efficacy working with suicidal clients and endorsed using evidence-based assessment procedures, but varied in utilization of recommended intervention practices. CONCLUSIONS:This study highlights gaps in clinicians' training and clinical practices that need to be overcome to provide evidence-based suicide care. Promisingly, positive associations were found between training and clinician knowledge, self-efficacy, and use of evidence-based practices.