Hydroxychloroquine lowers Alzheimer's disease and related dementias risk and rescues molecular phenotypes related to Alzheimer's disease.

We recently nominated cytokine signaling through the Janus-kinase-signal transducer and activator of transcription (JAK/STAT) pathway as a potential AD drug target. As hydroxychloroquine (HCQ) has recently been shown to inactivate STAT3, we hypothesized that it may impact AD pathogenesis and risk. Among 109,124 rheumatoid arthritis patients from routine clinical care, HCQ initiation was associated with a lower risk of incident AD compared to methotrexate initiation across 4 alternative analyses schemes addressing specific types of biases including informative censoring, reverse causality, and outcome misclassification (hazard ratio [95% confidence interval] of 0.92 [0.83-1.00], 0.87 [0.81-0.93], 0.84 [0.76-0.93], and 0.87 [0.75-1.01]). We additionally show that HCQ exerts dose-dependent effects on late long-term potentiation (LTP) and rescues impaired hippocampal synaptic plasticity prior to significant accumulation of amyloid plaques and neurodegeneration in APP/PS1 mice. Additionally, HCQ treatment enhances microglial clearance of Aß1-42, lowers neuroinflammation, and reduces tau phosphorylation in cell culture-based phenotypic assays. Finally, we show that HCQ inactivates STAT3 in microglia, neurons, and astrocytes suggesting a plausible mechanism associated with its observed effects on AD pathogenesis. HCQ, a relatively safe and inexpensive drug in current use may be a promising disease-modifying AD treatment. This hypothesis merits testing through adequately powered clinical trials in at-risk individuals during preclinical stages of disease progression.

Validity of ICD-10-based algorithms to identify patients with influenza in inpatient and outpatient settings.

PURPOSE: To evaluate the validity of ICD-10-CM code-based algorithms as proxies for influenza in inpatient and outpatient settings in the USA. METHODS: Administrative claims data (2015-2018) from the largest commercial insurer in New Jersey (NJ), USA, were probabilistically linked to outpatient and inpatient electronic health record (EHR) data containing influenza test results from a large NJ health system. The primary claims-based algorithms defined influenza as presence of an ICD-10-CM code for influenza, stratified by setting (inpatient/outpatient) and code position for inpatient encounters. Test characteristics and 95% confidence intervals (CIs) were calculated using test-positive influenza as a reference standard. Test characteristics of alternative outpatient algorithms incorporating CPT/HCPCS testing codes and anti-influenza medication pharmacy claims were also calculated. RESULTS: There were 430 documented influenza test results within the study period (295 inpatient, 135 outpatient). The claims-based influenza definition had a sensitivity of 84.9% (95% CI 72.9%-92.1%), specificity of 96.3% (95% CI 93.1%-98.0%), and PPV of 83.3% (95% CI 71.3%-91.0%) in the inpatient setting, and a sensitivity of 76.7% (95% CI 59.1%-88.2%), specificity of 96.2% (95% CI 90.6%-98.5%), PPV of 85.2% (95% CI 67.5%-94.1%) in the outpatient setting. Primary inpatient discharge diagnoses had a sensitivity of 54.7% (95% CI 41.5%-67.3%), specificity of 99.6% (95% CI 97.7%-99.9%), and PPV of 96.7% (95% CI 83.3%-99.4%). CPT/HCPCS codes and anti-influenza medication claims were present for few outpatient encounters (sensitivity 3%-10%). CONCLUSIONS: In a large US healthcare system, inpatient ICD-10-CM codes for influenza, particularly primary inpatient diagnoses, had high predictive value for test-positive influenza. Outpatient ICD-10-CM codes were moderately predictive of test-positive influenza.

Cardiovascular outcomes with SGLT2 inhibitors versus DPP4 inhibitors and GLP-1 receptor agonists in patients with heart failure with reduced and preserved ejection fraction.

BACKGROUND: No study has compared the cardiovascular outcomes for sodium-glucose cotransporter-2 inhibitors (SGLT2i) head-to-head against other glucose-lowering therapies, including dipeptidyl peptidase 4 inhibitor (DDP4i) or glucagon-like peptide-1 receptor agonist (GLP-1RA)-which also have cardiovascular benefits-in patients with heart failure with reduced (HFrEF) or preserved (HFpEF) ejection fraction. METHODS: Medicare fee-for-service data (2013-2019) were used to create four pair-wise comparison cohorts of type 2 diabetes patients with: (1a) HFrEF initiating SGLT2i versus DPP4i; (1b) HFrEF initiating SGLT2i versus GLP-1RA; (2a) HFpEF initiating SGLT2i versus DPP4i; and (2b) HFpEF initiating SGLT2i versus GLP-1RA. The primary outcomes were (1) hospitalization for heart failure (HHF) and (2) myocardial infarction (MI) or stroke hospitalizations. Adjusted hazards ratios (HR) and 95% CIs were estimated using inverse probability of treatment weighting. RESULTS: Among HFrEF patients, initiation of SGLT2i versus DPP4i (cohort 1a; n = 13,882) was associated with a lower risk of HHF (adjusted Hazard Ratio [HR (95% confidence interval)], 0.67 (0.63, 0.72) and MI or stroke (HR: 0.86 [0.75, 0.99]), and initiation of SGLT2i versus GLP-1RA (cohort 1b; n = 6951) was associated with lower risk of HHF (HR: 0.86 [0.79, 0.93]), but not MI or stroke (HR: 1.02 [0.85, 1.22]). Among HFpEF patients, initiation of SGLT2i versus DPP4i (cohort 2a; n = 17,493) was associated with lower risk of HHF (HR: 0.65 [0.61, 0.69]) but not MI or stroke (HR: 0.90 [0.79, 1.02]), and initiation of SGLT2i versus GLP-1RA (cohort 2b; n = 9053) was associated with lower risk of HHF (0.89 [0.83, 0.96]), but not MI or stroke (HR: 0.97 [0.83, 1.14]). Results were robust across range of secondary outcomes (e.g., all-cause mortality) and sensitivity analyses. CONCLUSIONS: Bias from residual confounding cannot be ruled out. Use of SGLT2i was associated with reduced risk of HHF against DPP4i and GLP-1RA, reduced risk of MI or stroke against DPP4i within the HFrEF subgroup, and comparable risk of MI or stroke against GLP-1RA. Notably, the magnitude of cardiovascular benefit conferred by SGLT2i was similar among patients with HFrEF and HFpEF.

Impact of the COVID-19 Pandemic on the Management of Juvenile Idiopathic Arthritis: Analysis of United States Commercial Insurance Data.

BACKGROUND/OBJECTIVE: Given limited information on health care and treatment utilization for juvenile idiopathic arthritis (JIA) during the pandemic, we studied JIA-related health care and treatment utilization in a commercially insured retrospective US cohort. METHODS: We studied rates of outpatient visits, new disease-modifying antirheumatic drug (DMARD) initiations, intra-articular glucocorticoid injections (iaGC), dispensed oral glucocorticoids and opioids, DMARD adherence, and DMARD discontinuation by quarter in March 2018-February 2021 (Q1 started in March). Incident rate ratios (IRR, pandemic vs prepandemic) with 95% confidence intervals (CIs) were estimated using multivariable Poisson or Quasi-Poisson models stratified by diagnosis recency (incident JIA, <12 months ago; prevalent JIA, ≥12 months ago). RESULTS: Among 1294 children diagnosed with JIA, total and in-person outpatient visits for JIA declined during the pandemic (IRR, 0.88-0.90), most markedly in Q1 2020. Telemedicine visits, while higher during the pandemic, declined from 21% (Q1) to 13% (Q4) in 2020 to 2021. During the pandemic, children with prevalent JIA, but not incident JIA, had lower usage of iaGC (IRR, 0.60; 95% CI, 0.34-1.07), oral glucocorticoids (IRR, 0.47; 95% CI, 0.33-0.67), and opioids (IRR, 0.44; 95% CI, 0.26-0.75). Adherence to and discontinuation of DMARDs was similar before and during the pandemic. CONCLUSIONS: In the first year of the pandemic, visits for JIA dropped by 10% to 12% in commercially insured children in the United States, declines partly mitigated by use of telemedicine. Pandemic-related declines in intra-articular glucocorticoids, oral glucocorticoids, and opioids were observed for children with prevalent, but not incident, JIA. These changes may have important implications for disease control and quality of life.

Pandemic vaccine testing: Combining conventional and challenge studies.

Early into COVID, human challenge trials were considered, but usually as alternatives to conventional randomized controlled trials. Instead, assessment of authorized COVID vaccines, of further COVID vaccines, and of vaccines against future pandemics should combine both designs, in five different ways, including a wholly novel one that we elaborate, Viz., combining data from both designs to answer a single question.

Do coronavirus vaccine challenge trials have a distinctive generalisability problem?

Notwithstanding the success of conventional field trials for vaccines against COVID-19, human challenge trials (HCTs) that could obtain more information about these and about other vaccines and further strategies against it are about to start in the UK. One critique of COVID-19 HCTs is their distinct paucity of information on crucial population groups. For safety reasons, these HCTs will exclude candidate participants of advanced age or with comorbidities that worsen COVID-19, yet a vaccine should (perhaps especially) protect such populations. We turn this cliché on its head. The truth is that either an HCT or a field trial has intrinsic generalisability limitations, that an HCT can expedite protection of high-risk participants even without challenging them with the virus, and that an important route to obtaining results generalisable to high-risk groups under either strategy is facilitated by HCTs.

Risk of clinically relevant hyperglycemia with metoprolol compared to carvedilol in older adults with heart failure and diabetes.

BACKGROUND: Although prior literature suggests that metoprolol may worsen glucose control compared to carvedilol, whether this has clinical relevance among older adults with diabetes and heart failure (HF) remains an open question. METHODS: This was a US retrospective cohort study utilizing data sourced from a 50% national sample of Medicare fee-for-service claims of patients with part D prescription drug coverage (2007-2017). Among patients with diabetes and HF, we identified initiators of metoprolol or carvedilol, which were 1:1 propensity score matched on >90 variables. The primary outcome was initiation of a new oral or injectable antidiabetic medication (proxy for uncontrolled diabetes); secondary outcomes included initiation of insulin and severe hyperglycemic event (composite of emergency room visits or hospitalizations related to hyperglycemia). RESULTS: Among 24 239 propensity score-matched pairs (mean [SD] age 77.7 [8.0] years; male [39.1%]), there were 8150 (incidence rate per 100 person-years [IR] = 33.5) episodes of antidiabetic medication initiation among metoprolol users (exposure arm) compared to 8576 (IR = 33.4) among carvedilol users (comparator arm) compared to corresponding to an adjusted hazard ratio (aHR) of 0.97 (95% confidence interval [CI]: 0.94, 1.01). Similarly, metoprolol was not associated with a significant increase in the risk of secondary outcomes including insulin initiation: aHR of 0.98 (95% CI: 0.93, 1.04) and severe hyperglycemic events: aHR of 0.98 (95% CI: 0.93, 1.02). CONCLUSIONS: In this large study of older adults with HF and diabetes, initiation of metoprolol compared to carvedilol was not associated with an increase in the risk of clinically relevant hyperglycemia.

Pharmacoepidemiology: A time for a new multidisciplinary approach to precision medicine.

The advent of the genomic age has created a rapid increase in complexity for the development and selection of drug treatments. A key component of precision medicine is the use of genetic information to improve therapeutic effectiveness of drugs and prevent potential adverse drug reactions. Pharmacoepidemiology, as a field, uses observational methods to evaluate the safety and effectiveness of drug treatments in populations. Pharmacoepidemiology by virtue of its focus, tradition, and research orientation can provide appropriate study designs and analysis methods for precision medicine. The objective of this manuscript is to demonstrate how pharmacoepidemiology can impact and shape precision medicine and serve as a reference for pharmacoepidemiologists interested in contributing to the science of precision medicine. This paper depicts the state of the science with respect to the need for pharmacoepidemiology and pharmacoepidemiological methods, tools and approaches for precision medicine; the need for and how pharmacoepidemiologists use their skills to engage with the precision medicine community; and recommendations for moving the science of precision medicine pharmacoepidemiology forward. We propose a new integrated multidisciplinary approach dedicated to the emerging science of precision medicine pharmacoepidemiology.

[Validation of the geriatrie-check for identification of geriatric patients in emergency departments]. / Validierung des Geriatrie-Checks zur Identifikation geriatrischer Patienten in der Notaufnahme.

BACKGROUND: The geriatric check was developed for identification of geriatric patients in emergency departments (ED) as part of the concept for geriatric care in Baden-Württemberg. AIM: Determination of convergent and predictive validity of the geriatric check with respect to identification and outcome prediction of geriatric patients in the ED. MATERIAL AND METHODS: A prospective cohort study between November 2015 and April 2016 including 146 patients older than 70 years in the internal medicine ED of Ulm University Hospital. Separate assessment by physicians and nursing staff of the following: identification of seniors at risk (ISAR), geriatric check, additional cognitive and functional assessments and for outcome: change in care index, Barthel index, living arrangements. RESULTS: The ISAR classified 117 patients as geriatric patients and the geriatric check 107. The correlation was 78.1%. With ISAR as the gold standard the geriatric check showed a sensitivity of 82.0% and a specificity of 62.1%. The positive and negative predictive values were 89.7% and 46.1%, respectively. The identification by simple estimation was better for nurses than for doctors (sensitivity 70.5% vs. 58%, specificity 88.9% vs. 83.3%). The predictive validity 5 months after admission with respect to the abovementioned outcome parameters was best for nurses and doctors (especially regarding specificity). Both tests were very sensitive but not very specific. DISCUSSION: The geriatric check is comparable to the ISAR. The convergent validity showed little difference. Both the ISAR and geriatric check were slightly more sensitive than doctors and nurses. Regarding predictive validity, doctors and nurses were superior to both scores. An algorithm starting with ISAR or geriatric check and followed by an estimation by doctors or nurses could be most suitable for optimal resource allocation.

Risk of sudden sensorineural hearing loss in adults using phosphodiesterase type 5 inhibitors: Population-based cohort study.

PURPOSE: The objective of the study was to determine the risk of sudden sensorineural hearing loss (SNHL) associated with use of phosphodiesterase type 5 (PDE5) inhibitors. METHODS: We conducted a retrospective cohort study in the MarketScan Commercial Claims and Encounters Database including adult men who initiated a PDE5 inhibitor (n = 377,722) and 1,957,233 nonusers between 1998 and 2007. Periods of drug exposure were assessed on a weekly basis based on pharmacy billing records, assuming use of 1 dose per week (current use). Incident sudden SNHL was defined based on inpatient or outpatient visits with International Classification of Diseases, Ninth Revision, Clinical Modification codes 389.1x, 389.2x, or 388.2 plus ≥2 procedure codes for audiometric hearing testing within ±30 days of sudden SNHL diagnosis. We used age- and propensity score-adjusted Cox proportional hazards model to evaluate the risk of sudden SNHL during periods of current or recent use compared with that of nonuse. We conducted sensitivity analyses by varying the assumed drug utilization frequency and sudden SNHL case definition. RESULTS: We evaluated 1233 sudden SNHL cases, resulting in an incidence of 4.35, 5.58, and 2.38 per 10,000 person-years for current, recent, and nonuse of PDE5 inhibitors, respectively. Compared with nonuse, the adjusted hazard ratio was 1.25 (1.01-1.55) for current use with a risk difference of 1.97 (1.12-2.82) per 10,000 person-years. For recent use, the adjusted hazard ratio was 1.60 (1.33-1.94) and risk difference was 3.19 (2.24-4.14). Estimates were consistent across the sensitivity analyses. CONCLUSIONS: Use of PDE5 inhibitors is associated with a small but significantly increased risk of sudden SNHL.

Comparative adherence to diabetes drugs: An analysis of electronic health records and claims data.

Non-adherence to medications is a major challenge in diabetes care. The objective of this brief report is to compare adherence rates for 6 major classes of diabetes medications: metformin, sulfonylurea, thiazolidinedione, basal insulin, DPP-4 inhibitors, and GLP-1 receptor agonists. We used a data source that linked electronic prescriptions with insurance claims to assess whether new electronic prescriptions for diabetes medications were followed by dispensing claims consistent with that prescription. After one year of follow-up, the daily medication possession probability (MPP) - a measure of overall adherence - at one year for sulfonylurea was 0.49 and for metformin was 0.46. Thiazolidinediones and basal insulin had a similar final daily MPP at 0.36 and 0.39, respectively, which was significantly lower than that for sulfonylurea or metformin (P < .05). GLP-1 receptor agonists and DPP-4 inhibitors were also comparable to one another at a final daily MPP of .30 and .21, respectively (P < .05 compared to any of the aforementioned drug classes). In summary, the rates at which diabetes drugs are prescribed, and the rates at which patients actually take them, differ substantially. Physicians should be aware of potentially significant challenges concerning adherence to newer agents.

Impact of the black triangle label on prescribing of new drugs in the United Kingdom: lessons for the United States at a time of deregulation.

PURPOSE: Newly approved novel drugs in Europe receive a black triangle label to promote pharmacovigilance. With growing momentum for earlier drug approvals and reliance on real-world evidence, we studied if the black triangle label promotes more judicious prescribing. METHODS: We examined whether general practitioners prescribed escitalopram, tadalafil, and vardenafil with a black triangle more cautiously than the same or similar drugs without a black triangle in The Health Improvement Network (UK). We performed interrupted time-series analyses to estimate changes in new prescription rates and nested case-control studies to compare characteristics of new users before and after removal of a black triangle. RESULTS: Prescribing rates to the 33 441 new users of these new drugs were highest shortly after initial approval and declined subsequently; there were no increases in rates of new prescriptions after a black triangle's removal (new prescriptions/million/month postlabel: escitalopram -1.5 [95% CI, -1.9 to -1.2]; tadalafil and vardenafil: -0.1 [95% CI, -0.6 to 0.4]). Among drugs in the same class, loss of a patent had more impact on prescribing rates than loss of a black triangle. People who began taking black triangle drugs were less likely to be young or to have multiple comorbidities or recent hospitalization compared with those starting the same drugs after the label's removal. However, these differences generally reflected secular trends seen also in similar, unlabeled medicines. CONCLUSIONS: Accelerated drug approvals could cause more uncertainty about drug effectiveness and safety, but specific labeling of newly approved medicines is unlikely to promote more judicious prescribing.

Agreement and validity of electronic health record prescribing data relative to pharmacy claims data: A validation study from a US electronic health record database.

BACKGROUND: Granular clinical and laboratory data available in electronic health record (EHR) databases provide researchers the opportunity to conduct investigations that would not be possible in insurance claims databases; however, for pharmacoepidemiology studies, accurate classification of medication exposure is critical. OBJECTIVE: The aim of this study was to evaluate the validity of classifying medication exposure using EHR prescribing (EHR-Rx) data. METHODS: We conducted a retrospective cohort study among patients with linked claims and EHR data in OptumLabs™ Data Warehouse. The agreement between EHR-Rx data and pharmacy claims (PC-Rx) data (for 40 medications) was determined using the positive predictive value (PPV) and medication possession ratio (MPR)-calculated in 1- and 12-month medication exposure periods (MEPs). Secondary analyses were restricted to incident vs prevalent EHR-Rxs, age ≥65 vs <65, white vs black race, males vs females, and number of EHR-Rxs. RESULTS: The validity metrics varied substantially among the 40 medications assessed. Across all medications, the period PPV and MPR were 62% and 63% in the 1-month MEP. They were 78% and 43% in the 12-month MEP. Overall, PPV and MPR were higher for patients with a prevalent EHR-Rx and age <65. CONCLUSIONS: Despite substantial variability among different medications, there was very good agreement between EHR-Rx data and PC-Rx data. To maximize the validity of classifying medication exposure with EHR prescribing data, researchers may consider using longer MEPs (eg, 12 months) and potentially require multiple EHR-Rxs to classify baseline medication exposure.

Lithium treatment and risk for dementia in adults with bipolar disorder: population-based cohort study.

BackgroundLithium inhibits glycogen synthase kinase-3, an enzyme implicated in the pathogenesis of dementia.AimsTo examine the association of lithium and dementia risk in a large claims-based US cohort of publicly insured older adults with bipolar disorder.MethodThe cohort included individuals ≥50 years diagnosed with bipolar disorder who did not receive dementia-related services during the prior year. Each follow-up day was classified by past-year cumulative duration of lithium use (0, 1-60, 61-300 and 301-365 days). Dementia diagnosis was the study outcome. Anticonvulsants commonly used as mood stabilisers served as a negative control.ResultsCompared with non-use, 301-365 days of lithium exposure was associated with significantly reduced dementia risk (hazard ratio (HR) = 0.77, 95% CI 0.60-0.99). No corresponding association was observed for shorter lithium exposures (HR = 1.04, 95% CI 0.83-1.31 for 61-300 days; HR = 1.07, 95% CI 0.67-1.71 for 1-60 days) or for any exposure to anticonvulsants.ConclusionsContinuous lithium treatment may reduce dementia risk in older adults with bipolar disorder.

Prevalence and correlates of antipsychotic polypharmacy in children and adolescents receiving antipsychotic treatment.

Antipsychotic polypharmacy (APP), which is common in adults with psychotic disorders, is of unproven efficacy and raises safety concerns. Although youth are increasingly prescribed antipsychotics, little is known about APP in this population. We performed a systematic PubMed search (last update 26 January 2013) of studies reporting the prevalence of APP in antipsychotic-treated youth. Summary statistics and statistical tests were calculated at the study level and not weighted by sample size. Fifteen studies (n = 58 041, range 68-23 183) reported on APP in youth [mean age = 13.4 ± 1.7 yr, 67.1 ± 10.2% male, 77.9 ± 27.4% treated with second-generation antipsychotics (SGAs)]. Data collected in these studies covered 1993-2008. The most common diagnoses were attention-deficit hyperactivity disorder (ADHD; 39.9 ± 23.5%) and conduct disorder/oppositional defiant disorder (CD/ODD; 33.6 ± 24.8). In studies including predominantly children (mean age = <13 yr, N = 5), the most common diagnosis were ADHD (50.6 ± 25.4%) and CD/ODD (39.5 ± 27.5%); while in studies with predominantly adolescents (mean age = ⩾13 yr, N = 7) the most common diagnoses were schizophrenia-spectrum disorders (28.6 ± 23.8%), anxiety disorders (26.9 ± 14.9%) and bipolar-spectrum disorders (26.6 ± 7.0%), followed closely by CD/ODD (25.8 ± 17.7). The prevalence of APP among antipsychotic-treated youth was 9.6 ± 7.2% (5.9 ± 4.5% in child studies, 12.0 ± 7.9% in adolescent studies, p = 0.15). Higher prevalence of APP was correlated with a bipolar disorder or schizophrenia diagnosis (p = 0.019) and APP involving SGA+SGA combinations (p = 0.0027). No correlation was found with APP definition [⩾1 d (N = 10) vs. >30-⩾90 d (N = 5), p = 0.88]. Despite lacking safety and efficacy data, APP in youth is not uncommon, even in samples predominantly consisting of non-psychotic patients. The duration, clinical motivations and effectiveness of this practice require further study.

Antipsychotic medication use in nursing homes: a proposed measure of quality.

OBJECTIVE: The potential misuse of antipsychotic medications (APMs) is an ongoing quality concern in nursing homes (NHs), especially given recent black box warnings and other evidence regarding the risk of APMs when used in NH populations. One mechanism regulators could use is public reporting of APM use by NHs; however, there is currently no agreed-upon measure of guideline-inconsistent APM use. In this paper, we describe a proposed measure of quality of APM use that is based on Centers for Medicare and Medicaid Services (CMS) Interpretive Guidelines, Food and Drug Administration (FDA) indications for APMs, and severity of behavioral symptoms. METHODS: The proposed measure identifies NH residents who receive an APM but do not have an approved indication for APM use. We demonstrate the feasibility of this measure using data from Medicaid-eligible long-stay residents aged 65 years and older in seven states. Using multivariable logistic regressions, we compare it to the current CMS Nursing Home Compare quality measure. RESULTS: We find that nearly 52% of residents receiving an APM lack indications approved by CMS/FDA guidelines compared with 85% for the current CMS quality measure. APM guideline-inconsistent use rates vary significantly across resident and facility characteristics, and states. Only our measure correlates with another quality indicator in that facilities with higher deficiencies have significantly higher odds of APM use. Predictors of inappropriate use are found to be consistent with other measures of NH quality, supporting the validity of our proposed measure. CONCLUSION: The proposed measure provides an important foundation to improve APM prescribing practices without penalizing NHs when there are limited alternative treatments available.

Validation of a claims-based antipsychotic polypharmacy measure.

Purpose Given the metabolic and neurologic side effects of antipsychotics and concerns about the increased risks associated with concomitant use, antipsychotic polypharmacy is a quality concern. This study assessed the operating characteristics of a Medicaid claims-based measure of antipsychotic polypharmacy. Methods A random sample from 10 public mental health clinics and 312 patients met criteria for this study. Medical record extractors were blind to measure status. We examined the prevalence, sensitivity, specificity, and positive predictive value (PPV) in Medicaid claims, testing nine different definitions of antipsychotic polypharmacy, including >14, >60, or >90 days concurrent use of ≥2 antipsychotic agents, each with allowable gaps of up to 0, 14, or 32 days in days' supply of antipsychotic medications. Results All Medicaid claims measure definitions tested had excellent specificity and PPV (>91%). Good to excellent sensitivity was dependent upon use of a 32-day gap allowance, particularly as duration of concurrent antipsychotic use increased. The proposed claims-based measure (90-day concurrent use of ≥2 or more antipsychotics, allowing for a 32-day gap) had excellent specificity (99.1%, 95%CI: 98.2-99.6) and PPV (90.9%, 95%CI: 83.1-95.7) with good sensitivity (79.4%, 95%CI: 70.4-86.6). The overall level of concordance between claims and medical record-based categorization of antipsychotic polypharmacy was high (96.4%, n = 301/312 clients, Cohen's K = 84.7, 95%CI: 75.9-93.5). Discrepant cases were reviewed, and implications are discussed. Conclusions Administrative claims data can be used to construct valid measures of antipsychotic polypharmacy.