The State of Use and Utility of Negative Controls in Pharmacoepidemiologic Studies.

Uses of real-world data in drug safety and effectiveness studies are often challenged by various sources of bias. We undertook a systematic search of the published literature through September 2020 to evaluate the state of use and utility of negative controls to address bias in pharmacoepidemiologic studies. Two reviewers independently evaluated study eligibility and abstracted data. Our search identified 184 eligible studies for inclusion. Cohort studies (115, 63%) and administrative data (114, 62%) were, respectively, the most common study design and data type used. Most studies used negative control outcomes (91, 50%), and for most studies the target source of bias was unmeasured confounding (93, 51%). We identified 4 utility domains of negative controls: 1) bias detection (149, 81%), 2) bias correction (16, 9%), 3) P-value calibration (8, 4%), and 4) performance assessment of different methods used in drug safety studies (31, 17%). The most popular methodologies used were the 95% confidence interval and P-value calibration. In addition, we identified 2 reference sets with structured steps to check the causality assumption of the negative control. While negative controls are powerful tools in bias detection, we found many studies lacked checking the underlying assumptions. This article is part of a Special Collection on Pharmacoepidemiology.

Patient-Informed Value Elements in Cost-Effectiveness Analyses of Major Depressive Disorder Treatment: A Literature Review and Synthesis.

OBJECTIVES: Prior work identified 6 key value elements (attributes of treatment and desired outcomes) for individuals living with major depressive disorder (MDD) in managing their condition: mode of treatment, time to treatment helpfulness, MDD relief, quality of work, interaction with others, and affordability. The objective of our study was to identify whether previous cost-effectiveness analyses (CEAs) for MDD treatment addressed any of these value elements. A secondary objective was to identify whether any study engaged patients, family members, and caregivers in the model development process. METHODS: We conducted a systematic literature review to identify published model-based CEAs. We compared the elements of the published studies with the MDD patient value elements elicited in prior work to identify gaps and areas for future research. RESULTS: Of 86 published CEAs, we found that 7 included patient out-of-pocket costs, and 32 included measures of productivity, which were both priorities for individuals with MDD. We found that only 2 studies elicited measures from patients for their model, and 2 studies engaged patients in the modeling process. CONCLUSIONS: Published CEA models for MDD treatment do not regularly include value elements that are a priority for this patient population nor do they include patients in their modeling process. Flexible models that can accommodate elements consistent with patient experience are needed, and a multistakeholder engagement approach would help accomplish this.

Impact of Private Payer Policies on the Transition to Telemental Health Care Among Privately Insured Patients with Mental Health Disorders.

Introduction: Mental health patients in states without private payer telehealth reimbursement policies before the public health emergency (PHE) may have experienced reduced access to telemental health (TMH). We estimate the association between private payer telehealth policy status in 2019 and the transition to TMH in 2020. Methods: Retrospective cohort study of privately insured individuals 2-64 years old with a mental health disorder and without TMH use in 2019. We examined new telemental use in 2020 by three categories of policy reimbursement status in 2019 (partial parity, full parity vs. no policy), overall (any telemental), and by modality (live video, audio-only, and online assessments) using logistic regression models clustered by state. Results: Among the 34,612 enrollees, 54.7% received TMH for the first time. Relative to no policy states, enrollees in partial or full parity states were equally likely to receive TMH in 2020. However, enrollees in states with a private payer telehealth policy were less likely to receive audio-only (partial parity: odds ratio [OR]: 0.59, 95% confidence interval [CI]: 0.39-0.90; full parity: OR: 0.38, 95% CI: 0.26-0.55), but more likely to receive online assessments (full parity: OR: 2.28, 95% CI: 1.4-4.59). Conclusions: Privately insured enrollees similarly transitioned to TMH across states suggesting a broad impact of the PHE policies on access to this care. The differences in audio-only and online assessments suggest that providers were possibly better prepared to implement TMH care via live video or patient portals in states with telehealth policies.

Impact of Undertreatment of Depression on Suicide Risk Among Children and Adolescents With Major Depressive Disorder: A Microsimulation Study.

Undertreatment of depression is common among children and adolescents, but evidence of the impact of undertreatment of depression on risk of suicide is limited due to the low base rate of suicide in the population and lack of sufficient data sources. We developed a microsimulation model that uses evidence from multiple sources to study the impact of different durations of antidepressant treatment on suicide risk in a synthesized sample that is nationally representative of children and adolescents with major depressive disorder. Compared with receiving no treatment, suicide rate and risk of suicide attempt both decreased with increasing duration of antidepressant treatment (for 12 weeks, suicide rate ratios = 0.78 (95% credible interval (CrI): 0.58, 1.15), 36 weeks, 0.65 (95% CrI: 0.44, 0.90), and 52 weeks, 0.63 (95% CrI: 0.45, 0.72); for suicide attempt: 12 weeks, suicide risk ratios = 0.68 (95% CrI: 0.62, 0.69), 36 weeks, 0.56 (95% CrI: 0.52, 0.57), and 52 weeks, 0.55 (95% CrI: 0.51, 0.56). The suicide rate and risk of suicide attempt were lower in children than in adolescents. Males had a lower risk of suicide attempt but higher suicide rate than females. The findings from the microsimulation model show that completion of 12-36 weeks of antidepressant treatment may reduce suicide attempt and suicide among children and adolescents with major depressive disorder.

The association of major adverse limb events and combination stent and atherectomy in patients undergoing revascularization for lower extremity peripheral artery disease.

BACKGROUND: The effectiveness of combined atherectomy and stenting relative to use of each procedure alone for the treatment of lower extremity peripheral artery disease has not been evaluated. AIMS: The objective of this study was to evaluate the short- and long-term major adverse limb event (MALE) following the receipt of stenting, atherectomy, and the combination of stent and atherectomy. METHODS: A retrospective cohort of patients undergoing atherectomy, stent, and combination stent atherectomy for lower extremity peripheral artery disease was derived from the Vascular Quality Initiative (VQI) data set. The primary outcome was MALE and was assessed in the short-term and long-term. Short-term MALE was assessed immediately following the procedure to discharge and estimated using logistic regression. Long-term MALE was assessed after discharge to end of follow-up and estimated using the Fine-Gray subdistribution hazard model. RESULTS: Among the 46,108 included patients, 6896 (14.95%) underwent atherectomy alone, 35,774 (77.59%) received a stent, and 3438 (7.5%) underwent a combination of stenting and atherectomy. The adjusted model indicated a significantly higher odds of short-term MALE in the atherectomy group (OR = 1.35; 95% confidence interval [CI]:1.16-1.57), and not significantly different odds (OR = 0.93; 95% CI:0.77-1.13) in the combination stent and atherectomy group when compared to stenting alone. With regard to long-term MALE, the model indicated that the likelihood of experiencing the outcome was slightly lower (HR = 0.90; 95% CI:0.82-0.98) in the atherectomy group, and not significantly different (HR = 0.92; 95% CI:0.82-1.04) in the combination stent and atherectomy group when compared to the stent group. CONCLUSIONS: Patients in the VQI data set who received combination stenting and atherectomy did not experience significantly different rates of MALE when compared with stenting alone. It is crucial to consider and further evaluate the influence of anatomical characteristics on treatment strategies and potential differential effects of comorbidities and other demographic factors on the short and long-term MALE risks.

Optimizing therapeutic decision-making for off-label medicines use: A scoping review and consensus recommendations for improving practice and research.

PURPOSE: Off-label medicines use is a common and sometimes necessary practice in many populations, with important clinical, ethical and financial consequences, including potential unintended harm or lack of effectiveness. No internationally recognized guidelines exist to aid decision-makers in applying research evidence to inform off-label medicines use. We aimed to critically evaluate current evidence informing decision-making for off-label use and to develop consensus recommendations to improve future practice and research. METHODS: We conducted a scoping review to summarize the literature on available off-label use guidance, including types, extent and scientific rigor of evidence incorporated. Findings informed the development of consensus recommendations by an international multidisciplinary Expert Panel using a modified Delphi process. Our target audience includes clinicians, patients and caregivers, researchers, regulators, sponsors, health technology assessment bodies, payers and policy makers. RESULTS: We found 31 published guidance documents on therapeutic decision-making for off-label use. Of 20 guidances with general recommendations, only 35% detailed the types and quality of evidence needed and the processes for its evaluation to reach sound, ethical decisions about appropriate use. There was no globally recognized guidance. To optimize future therapeutic decision-making, we recommend: (1) seeking rigorous scientific evidence; (2) utilizing diverse expertise in evidence evaluation and synthesis; (3) using rigorous processes to formulate recommendations for appropriate use; (4) linking off-label use with timely conduct of clinically meaningful research (including real-world evidence) to address knowledge gaps quickly; and (5) fostering partnerships between clinical decision-makers, researchers, regulators, policy makers, and sponsors to facilitate cohesive implementation and evaluation of these recommendations. CONCLUSIONS: We provide comprehensive consensus recommendations to optimize therapeutic decision-making for off-label medicines use and concurrently drive clinically relevant research. Successful implementation requires appropriate funding and infrastructure support to engage necessary stakeholders and foster relevant partnerships, representing significant challenges that policy makers must urgently address.

Caregiver Burden in Caregivers of Children With Special Health Care Needs and Association With Chronic Pain.

BACKGROUND/OBJECTIVES: Chronic noncancer pain (CNCP) affects millions of individuals in the United States but evidence of its prevalence among caregivers of children with special health care needs is sparse. We sought to estimate the prevalence of CNCP and its association with caregiver burden, in a nationally representative sample. METHODS: Retrospective cross-sectional study using pooled Medical Expenditure Panel Survey data for 2010-2015. Within interviewed households, family groups consisting of at least 1 parent and 1 child (0-17 y) were identified. CNCP was identified by one or more International Classification of Diseases, Ninth Revision (ICD-9)-CM codes utilizing previously published approaches. Level of caregiver burden was defined using a validated screener questionnaire identifying children with high burden of care (ie, special health care needs), for example, high or low burden. We estimated prevalence of CNCP as a function of caregiver burden, as well as the association of risk factors with CNCP, including parent sociodemographic features, clinical diagnoses, and family level characteristics. RESULTS: We identified 46,525 caregivers of whom 3.6% reported experiencing high caregiving burden. The prevalence of CNCP was 25.5% and 14.0% among parents with high compared with low caregiving burden, respectively. Odds of CNCP were higher among parents with high compared to those with lower caregiver burden (odds ratio=1.29, 95% confidence interval=1.06-1.55). Being obese, experiencing disability, and having a mental health diagnosis were associated with higher odds of CNCP. CONCLUSIONS: Chronic pain is more common among caregivers with high caregiver burden. Our findings highlight the need to further explore the nature and impact of risk factors on caregiver health and disability.

Naloxone Prescribing Associated With Reduced Emergency Department Visits in the Military Health System.

PURPOSE: The aim was to determine the association between the receipt of naloxone and emergency department (ED) visits within 60 days after the receipt of an opioid. METHODS: A retrospective cohort of individuals 18 years of age or above, enrolled in TRICARE, and were dispensed an opioid at any time from January 1, 2019, through September 30, 2020 was identified within the United States Military Health System. Individuals receiving naloxone within 5 days of the opioid dispensing date were propensity score matched with individuals receiving opioids only. A logistic regression was used to estimate the odds of an ED visit in the 60-day follow-up period after the index opioid dispense event among those co-dispensed naloxone and those receiving opioids only. RESULTS: Of the 2,136,717 individuals who received an opioid prescription during the study period, 800,071 (10.1%) met study inclusion criteria. Overall, 5096 (0.24%) of individuals who received an opioid prescription were co-dispensed naloxone. Following propensity score matching, those who received naloxone had a significantly lower odds of ED utilization in the 60 days after receiving an opioid prescription (odds ratio: 0.74, 95% CI: 0.68-0.80, P<0.001). CONCLUSION: This study highlights the importance of expanding access to naloxone in order to reduce ED utilization. Future research is needed to examine additional outcomes related to naloxone receipt and develop programs that make naloxone prescribing a routine practice.

Evaluating the association between antidepressant dose trajectories and treatment augmentation in pediatric depression.

PURPOSE: To identify antidepressant dose trajectories in the first 6-months of antidepressant initiation and to evaluate the association between antidepressant dose trajectories and augmentation with another psychotropic medication. METHODS: Using the IQVIA PharMetrics® Plus database, we identified 5655 commercially insured youth (3-18 years) with depression who newly initiated an antidepressant anytime from January 2007 to June 2015. No antidepressant use within 1 year prior to the index prescription defined new use. Latent class growth analysis of antidepressant dosing in the 6 months after initiation defined the exposure groups. The outcome was any regimen change, that is, antidepressant augmentation with another psychotropic or discontinuation of the antidepressant, with and without switching to another psychotropic. Baseline covariates measured in the 6 months before antidepressant initiation included demographic factors, psychiatric comorbidities, and health service use. Multinomial logistic regression tested the association between antidepressant dose trajectories and the odds of an antidepressant regimen change. RESULTS: Five dose trajectories were sharp decline (n = 897; 16%), slow decline (n = 1029; 18%), stable minimum dose (n = 1397; 25%), stable maximum dose (n = 1783; 32%), and increasing high dose (n = 549; 10%). Relative to the stable minimum dose group, the sharp and slow decline groups were more likely to discontinue the antidepressant, either switch to another psychotropic (OR [odds ratio]: 5.91; 95%CI: 3.23-10.80 and OR: 1.67; 95%CI: 1.04-2.68, respectively) or stop all psychotropic medication (OR: 6.64; 95%CI: 4.24-10.39 and OR: 1.62; 95%CI: 1.22-2.13, respectively). However, the stable maximum and increasing high-dose groups were less likely to discontinue, either switch (OR: 0.38; 95%CI: 0.24-0.61 and OR: 0.30; 95%CI: 0.16-0.59, respectively) or stop all psychotropic medications (OR: 0.15; 95%CI: 0.12-0.20 and OR: 0.02; 95%CI: 0.01-0.03 respectively) than augment with another psychotropic. CONCLUSIONS: The findings from this cross-sectional study demonstrate an association between antidepressant dose trajectories within 6 months of initiating treatment and the odds of augmenting with another psychotropic.

Trends in antipsychotic use for youth with attention-deficit/hyperactivity disorder and disruptive behavior disorders.

PURPOSE: To examine trends in off-label antipsychotic use for youth with attention-deficit/hyperactivity disorder with and without a comorbid disruptive behavior disorder. METHOD: This cross-sectional study of annual trends from 2007 through 2015 used the IQVIA PharMetrics® Plus for Academics data. We identified 165 794 commercially-insured youth 3-18-year-old who had a diagnosis of attention-deficit/hyperactivity disorder and classified them into subgroups with and without disruptive behavior disorders comorbidities. Antipsychotic use, with or without a stimulant, was the primary dependent outcome. Logistic regression estimated the odds of antipsychotic use associated with comorbid attention-deficit/hyperactivity disorder and disruptive behavior disorders, adjusting for age, sex, study year, and other psychotropic use. RESULTS: Over 70% of the 165 794 youth with attention-deficit/hyperactivity disorder were 5-14-year-old and male, and 12% had disruptive behavior disorders. Antipsychotic prevalence, with or without a stimulant, was 4.4% in 2007 and 3.4% in 2015. Stimulants with antipsychotics increased significantly from 2007 to 2015 for females (19.5%-28.7%) and youth 15-18-year-old (25.9%-32.7%). Adjusting for age, sex, study year, and other psychotropic use, youth with a comorbid disruptive behavior had a 2.5 (95% CI: 2.3, 2.7) higher likelihood of receiving an antipsychotic than youth with attention-deficit/hyperactivity disorder and no comorbidities. CONCLUSIONS: Antipsychotic use was associated with comorbid disruptive behaviors in youth with attention-deficit/hyperactivity disorder. Off-label antipsychotic use has increased for females and older adolescents.

Repetitive Traumatic Brain Injury Among Older Adults.

OBJECTIVE: To determine the incidence of and assess risk factors for repetitive traumatic brain injury (TBI) among older adults in the United States. DESIGN: Retrospective cohort study. SETTING: Administrative claims data obtained from the Centers for Medicare & Medicaid Services' Chronic Conditions Data Warehouse. PARTICIPANTS: Individuals 65 years or older and diagnosed with TBI between July 2008 and September 2012 drawn from a 5% random sample of US Medicare beneficiaries. MAIN MEASURES: Repetitive TBI was identified as a second TBI occurring at least 90 days after the first occurrence of TBI following an 18-month TBI-free period. We identified factors associated with repetitive TBI using a log-binomial model. RESULTS: A total of 38 064 older Medicare beneficiaries experienced a TBI. Of these, 4562 (12%) beneficiaries sustained at least one subsequent TBI over up to 5 years of follow-up. The unadjusted incidence rate of repetitive TBI was 3022 (95% CI, 2935-3111) per 100 000 person-years. Epilepsy was the strongest predictor of repetitive TBI (relative risk [RR] = 1.44; 95% CI, 1.25-1.56), followed by Alzheimer disease and related dementias (RR = 1.32; 95% CI 1.20-1.45), and depression (RR = 1.30; 95% CI, 1.21-1.38). CONCLUSIONS: Injury prevention and fall-reduction interventions could be targeted to identify groups of older adults at an increased risk of repetitive head injury. Future work should focus on injury-reduction initiatives to reduce the risk of repetitive TBI as well as assessment of outcomes related to repetitive TBI.

Opioid prescribing and risks among commercially insured women undergoing pelvic organ prolapse repair.

PURPOSE: Opioid use after surgical repair for pelvic organ prolapse (POP) is intended for short-term post-operative pain. This study compared the incidence of opioid prescribing in women undergoing POP transabdominal repair with mesh and transvaginal native tissue repair. METHODS: A retrospective cohort of women undergoing POP transabdominal repair with mesh or transvaginal native tissue repair, was derived from a 10% random sample of enrollees from 2007 to 2015 within the IQVIA PharMetrics® Plus Database. Primary outcomes were any prescription of opioids and cumulative days of opioids prescribed in the 14- 180 days following surgical intervention. Inverse probability of treatment weights controlled for observed baseline confounders. Any opioid prescription was estimated using logistic regression and generalized linear regression for cumulative days of opioids prescribed. RESULTS: The cohort of 49 052 women who underwent POP surgical repair included 46 813 women with transvaginal native tissue repair and 2239 women with transabdominal repair with mesh. Women with a transabdominal repair with mesh had a 1.19 (95%CI: 1.09-1.31) significantly higher odds of receiving an opioid prescription than women with transvaginal native tissue repair. Post-operatively, over 29% of women received opioid prescriptions. Mean cumulative days of post-surgical opioid prescribing was 32.2 (SD = 43.1), and was not statistically different between groups. Thirteen percent of women were prescribed opioids for 90 days or more. CONCLUSIONS: Women undergoing POP with transabdominal mesh are more likely to receive prescriptions for opioids after surgery compared to transvaginal native tissue repair. Treatment plans that address pain while mitigating the risks associated with prolonged opioid prescribing should be employed.

Spillover effects of state medicaid antipsychotic prior authorization policies in US commercially insured youth.

PURPOSE: To evaluate spillover effects of Medicaid antipsychotic prior authorization (PA) policies among commercially insured youth. METHODS: Commercially insured youth residing in nine US states that implemented PA exclusively for antipsychotics in 2011 or 2012 were identified using a 10% random sample of enrollees in the IQVIA PharMetrics Plus database spanning 2007 to 2015. Youth were included if they were ≤18 years, met the age criteria of the PA at the time of dispensing, and had at least 1 month of prescription drug coverage from 2007 to 2015. The primary outcome of interest was the monthly prevalence of antipsychotics. We implemented segmented regression of interrupted time series analysis to estimate changes in the monthly prevalence of targeted medications, overall and stratified by age. Trends were compared in the 4-year period before and the 3-year period after implementation of PA policies. RESULTS: Antipsychotics prescribing significantly decreased 6.74/10 000 (95% CI, -9.04 to -4.44) enrollees per month immediately after PA implementation. However, PA was not associated with significant long-term trend changes (-0.06; 95% CI, -0.16 to 0.03). Antipsychotic prescribing in children <12 years-old significantly decreased 0.14/10 000 (95% CI, -0.21 to -0.07) enrollees per month after PA implementation, while prescribing in adolescents 12 to 18 years-old significantly increased 0.32/10 000 (95% CI, 0.16 to 0.47) enrollees per month. CONCLUSION: While Medicaid PA polices for antipsychotic oversight did not affect overall prescribing, there were spillover effects in U.S. commercially insured children <12 years-old. This suggests that state-level Medicaid policies intended to improve the quality of care and safe use of antipsychotics can have broad reach.

Receipt of Treatment for Depression Following Traumatic Brain Injury.

OBJECTIVE: Lack of evidence for efficacy and safety of treatment and limited clinical guidance have increased potential for undertreatment of depression following traumatic brain injury (TBI). METHODS: We conducted a retrospective cohort study among individuals newly diagnosed with depression from 2008 to 2014 to assess the impact of TBI on receipt of treatment for incident depression using administrative claims data. We created inverse probability of treatment-weighted populations to evaluate the impact of TBI on time to receipt of antidepressants or psychotherapy following new depression diagnosis during 24 months post-TBI or matched index date (non-TBI cohort). RESULTS: Of 10 428 individuals with incident depression in the TBI cohort, 44.7% received 1 or more antidepressants and 20.0% received 1 or more psychotherapy visits. Of 10 463 in the non-TBI cohort, 41.2% received 1 or more antidepressants and 17.6% received 1 or more psychotherapy visits. TBI was associated with longer time to receipt of antidepressants compared with the non-TBI cohort (average 39.6 days longer than the average 126.2 days in the non-TBI cohort; 95% confidence interval [CI], 24.6-54.7). Longer time to psychotherapy was also observed among individuals with TBI at 6 months post-TBI (average 17.1 days longer than the average 47.9 days in the non-TBI cohort; 95% CI, 4.2-30.0), although this association was not significant at 12 and 24 months post-TBI. CONCLUSIONS: This study raises concerns about the management of depression following TBI.

Future of Patients in Healthcare Evaluation: The Patient-Informed Reference Case.

The "Reference Case" was developed to facilitate comparability among published cost-effectiveness analyses intended to contribute to decisions about the broad allocation of healthcare resources. Although the societal perspective is recommended for Reference Case analyses, empirical estimations rarely adequately represent the patient perspective, and more often, healthcare system or payer perspectives are used. In this commentary, we discuss the evolution of the Reference Case over the past 20 years and how it now needs to further evolve. This should begin with a patient-informed societal perspective. A realignment of the societal perspective to better include patient perspectives in CEA creates a conduit for patient inclusion. Engaging patients to both derive patient-informed value elements and prioritize value elements using stated preference methods will lead to patient inclusion in the societal perspective and a patient-informed Reference Case analysis.

Engaging stakeholders in pharmacoepidemiology research: Current state and recommendations.

PURPOSE: Given current efforts to enhance patient-centered care and shared decision-making, the International Society of Pharmacoepidemiology Workgroup on Patient Engagement assessed patient and other stakeholder engagement in pharmacoepidemiology research and provides recommendations for the field. METHODS: A systematic review used MEDLINE and EMBASE to identify published literature from 2005 to 2016 addressing how stakeholders-patients, caregivers, and others-assisted researchers conducting pharmacoepidemiologic research. Three pairs of Workgroup members screened titles and abstracts to select articles for full-text review and analysis. Two Workgroup members abstracted the following data: research focus, characterization and role of stakeholders, and type(s) of engagement strategy employed. Data were summarized descriptively. RESULTS: We identified 5717 references for abstract screening. Of these, 69 met the criteria for full-text screening, and 11 were selected for data abstraction. Of these 11 studies, seven focused on the development of a research agenda and eight had stakeholders react or advise on an aspect of the study. Although patients were the most commonly identified stakeholders, advocacy groups and health care professionals were also frequently identified. Some studies reported the engagement of other stakeholders, including local government or policy experts. Engagement strategies varied, with five studies using more than one strategy. Studies often did not indicate the involvement of stakeholders in developing the study design or with implementation. CONCLUSIONS: Currently, few pharmacoepidemiology publications mention patient or other stakeholder engagement in the design, analysis, or reporting of research. This suggests that there are opportunities to expand stakeholder engagement and/or increase the transparency of reporting stakeholder engagement.

Emerging Good Practices for Transforming Value Assessment: Patients' Voices, Patients' Values.

BACKGROUND: Patient engagement is a transformative strategy for improving value assessment. US value framework developers have increased engagement activities, but more needs to be learned about how to best achieve meaningful patient engagement in value assessment. The objective was to glean good practices in patient engagement emerging from patient community experiences, to be used in value assessment. METHODS: The National Health Council Value Workgroup conducted a survey and held a focus group with its member advocacy organizations to gather experiences with value framework developers and views on emerging good practices. RESULTS: Ten of 13 organizations completed the survey; reporting 13 interactions with four framework developers. Most rated experiences as "good" to "very good." Emerging good practices included (1) engage early; (2) engage a range of patients; (3) leverage patient-provided information, data resources, and outreach mechanisms; (4) be transparent; and (5) appreciate and accommodate resource constraints. Twelve of 13 organizations participated in the focus group, and this produced 30 emerging good practices in four areas: (1) timing; (2) methodology and data; (3) partnering; and (4) characterizing engagement. DISCUSSION: Patient engagement was limited in early development of value frameworks but has increased in the past few years. Patient groups report positive experiences that can serve as emerging good practices. These groups also reported experienced challenges in their interactions and recommended good practices to mitigate those challenges. CONCLUSIONS/RECOMMENDATIONS: The growing pool of patient engagement experiences can be translated into good practices to advance a patient-centered, value-driven health care ecosystem. Lessons learned from these early experiences can help establish recommend emerging good practices that can eventually result in best practices and standards in the field.

Impact of Coordinated Behavioral Health Management on Quality Measures of Antipsychotic Use.

A state Care Management Entity (CME) using the wraparound practice model provided intensive care coordination for youth with severe mental illness, those most likely to receive antipsychotics. The model has led to improved clinical/functional outcomes, but little is known about the impact on antipsychotic prescribing and safety monitoring. A pre-post study was conducted to evaluate antipsychotic dosing, concomitant antipsychotic use, and metabolic monitoring among CME-enrolled and non-CME-enrolled comparison groups. CME-enrolled youth had greater decrease in concomitant antipsychotic use than non-CME-enrolled youth, but no difference in dosing or metabolic monitoring. More education of prescribing antipsychotics and team-based engagement in care coordination are needed.

Tree-based Claims Algorithm for Measuring Pretreatment Quality of Care in Medicare Disabled Hepatitis C Patients.

BACKGROUND: To help broaden the use of machine-learning approaches in health services research, we provide an easy-to-follow framework on the implementation of random forests and apply it to identify quality of care (QC) patterns correlated with treatment receipt among Medicare disabled patients with hepatitis C virus (HCV). METHODS: Using Medicare claims 2006-2009, we identified 1936 patients with 6 months continuous enrollment before HCV diagnosis. We ran a random forest on 14 pretreatment QC indicators, extracted the forest's representative tree, and aggregated its terminal nodes into 4 QC groups predictive of treatment. To explore determinants of differential QC receipt, we compared patient-level and county-level (linked AHRF data) characteristics across QC groups. RESULTS: The strongest predictors of treatment included "liver biopsy," "HCV genotype testing," "specialist visit," "HCV viremia confirmation," and "iron overload testing." High QC [n=360, proportion treated (pt)=33.3%] was defined for patients with at least 2 from the above-mentioned metrics. Good QC patients (n=302, pt=12.3%) had either "HCV genotype testing" or "specialist visit," whereas fair QC (n=282, pt=7.1%) only had "HCV viremia confirmation." Low QC patients (n=992, pt=2.5%) had none of the selected metrics. The algorithm accuracy of predicting treatment was 70% sensitivity and 78% specificity. HIV coinfection, drug abuse, and residence in counties with higher supply of hospitals with immunization and AIDS services correlated with lower QC. CONCLUSIONS: Machine-learning techniques could be useful in exploring patterns of care. Among Medicare disabled HCV patients, the receipt of more QC indicators was associated with higher treatment rates. Future research is needed to assess determinants of differential QC receipt.