The Economic Burden of Tenosynovial Giant Cell Tumors Among Employed Workforce in the United States.

OBJECTIVE: To assess the economic burden of tenosynovial giant cell tumor (TGCT) among US employed workforce. METHODS: Patients with TGCT medical claims (N = 1395) and matched controls (1:10) without TGCT claims (N = 13,950) were identified from the OptumHealth Care Solutions, Inc. database (January 1, 1999 to March 31, 2017). Adjusted regression models were used to compare healthcare resource utilization, time lost from work, and associated costs between cohorts. RESULTS: In patients with TGCT, the rates of inpatient admissions, emergency room visits, outpatient visits, and work loss days were 2.8, 1.5, 2.2, and 2.6 times those of matched controls, respectively (all P < 0.001). Total annual all-cause healthcare costs and work loss-related costs were $9368 and $2708 higher for TGCT patients than for matched controls, respectively (all P < 0.001). CONCLUSIONS: TGCT was associated with a significant healthcare and work loss burden on US employers.

Real-world impact of glycated hemoglobin reduction on treatment intensification and glycated hemoglobin goal attainment in type 2 diabetes mellitus patients initiated on a sodium glucose co-transporter 2 (SGLT2) inhibitor (SGLT2i).

Objective: To evaluate the impact of a 0.2% reduction in glycated hemoglobin (HbA1c) on treatment intensification, poor HbA1c control and HbA1c goal attainment in patients with type 2 diabetes mellitus (T2DM) initiated on a sodium glucose co-transporter 2 (SGLT2) inhibitor (SGLT2i).Methods: IQVIATM Health Plan Claims Data - US and IQVIATM Ambulatory EMR Data - US databases (29 October 2012-31 March 2016) were used to identify adults with T2DM initiated on an SGLT2i (index date) who had HbA1c measurements pre- and post-index, and ≥6 months of eligibility pre-index (baseline). HbA1c change was defined as the difference between the first post-index and the last pre-index measurements. Cox regression models were used to assess treatment intensification, poor HbA1c control (i.e. HbA1c > 9%, among patients <9% at baseline) and goal attainment (HbA1c < 7%, <8%; among patients with HbA1c above goal at baseline) adjusting for HbA1c change and baseline characteristics. Patients were observed up to one year after the first HbA1c measurement or end of eligibility. Hazard ratios (HRs) and 95% confidence intervals (CIs) were reported.Results: A total of 938 patients (mean age 54.9, 42.5% female, mean HbA1c 8.5%) were selected. Following SGLT2i initiation, each 0.2% reduction in HbA1c levels was associated with a decreased risk of treatment intensification (HR [95% CI] = 0.90 [0.86-0.92]), a decreased likelihood of reaching HbA1c > 9% (HR [95% CI] = 0.85 [0.79-0.88]) and higher likelihoods of achieving a treatment goal of HbA1c < 7% (HR [95% CI] = 1.17 [1.12-1.21]) and HbA1c < 8% (HR [95% CI] = 1.08 [1.04-1.10]).Conclusions: In T2DM patients, each HbA1c reduction of 0.2% following the initiation of an SGLT2i was associated with a significant positive impact on treatment intensification and HbA1c goal attainment.

Persistent Disease Activity Remains a Burden for Patients with Systemic Lupus Erythematosus.

OBJECTIVE: Persistent systemic lupus erythematosus (SLE) disease activity is associated with increased morbidity and mortality. In a multicenter cohort of patients with prevalent SLE, we described persistence, patterns, and predictors of change in disease activity over time. METHODS: Based on SLE Disease Activity Index (SLEDAI)-2K scores at cohort entry, patients were classified into 4 groups: low (score < 4; LOW), moderate (4 to < 6; MOD), moderately high (6 to ≤ 10; MHIGH), and very high (> 10; VHIGH). Multivariable linear and longitudinal mixed linear regression models were used to identify predictors of change over time in SLEDAI-2K. RESULTS: There were 2019 participants, with declining followup data over 5 years (1326, 580, 274, 186, and 148 patients, respectively). At cohort entry, mean (± SD) age was 42 (± 17) years, disease duration 11 (± 10) years, and 90% were female. The 4 groups included 44% LOW (n = 891), 20% MOD (n = 400), 22% MHIGH (n = 442), and 14% VHIGH (n = 286); therefore, 36% had clinically important SLE activity. The proportion of patients in the LOW group at entry who moved to a higher activity level varied from 30% (167/557) at 1 year, to 49% (41/83) at 3 years, and 54% (30/56) at 5 years. Among 181 patients with MOD to VHIGH entry activity and 3 years of followup, 116 (64.1%) remained active. In all analyses, only higher SLEDAI-2K at cohort entry remained a significant predictor of higher SLEDAI-2K in subsequent years. CONCLUSION: Higher SLEDAI-2K at study entry was the single major independent predictor of higher SLEDAI-2K over time, reflecting frequent persistence of active disease, even in patients with longstanding disease. This highlights gaps in the optimal treatment of SLE.

Real-world Treatment Patterns Among Patients With Colorectal Cancer Treated With Trifluridine/Tipiracil and Regorafenib.

BACKGROUND: Trifluridine/tipiracil (FTD/TPI) and regorafenib (REG) prolong survival in refractory metastatic colorectal cancer (mCRC) and have similar indications with different side-effect profiles. The present study compared real-world treatment patterns with FTD/TPI and REG for mCRC in a large, representative US claims database. MATERIALS AND METHODS: Retrospective data from the US Symphony Health Solutions' Integrated Dataverse database were analyzed for adult mCRC patients receiving FTD/TPI or REG from October 2014 to July 2016. The index date was the first FTD/TPI or REG prescription date. The observation period spanned from the index date to the end of data collection, end of continuous clinical activity, or treatment switch. Adherence was assessed using the medication possession ratio and proportion of days covered at 3 months. The time to discontinuation was assessed over the observation period with gaps of 45, 60, or 90 days. Outcomes were compared between the cohorts using logistic regression and Cox proportional hazards models adjusting for baseline characteristic differences. RESULTS: A total of 1630 FTD/TPI patients and 1425 REG patients were identified. The FTD/TPI patients were 80% more likely to have a medication possession ratio of ≥ 0.80 compared with the REG patients (odds ratio, 1.80; P < .001) and more than twice as likely to have a proportion of days covered of ≥ 0.80 (odds ratio, 2.66; P < .001) at 3 months. The FTD/TPI patients were 37% less likely to discontinue their treatment compared with the REG patients when using the 60-day gap (hazard ratio, 0.63; P < .001). Similar results were found using the 45- and 90-day gaps. CONCLUSION: mCRC patients taking FTD/TPI were significantly more likely to adhere to and comply with therapy compared with those taking REG.

CMS hospital readmission reduction program and anticoagulants received following a total hip and knee arthroplasty discharge.

OBJECTIVES: To assess association between 30 day readmission rate and treatment received after total hip and knee arthroplasty (THA/TKA) discharge (rivaroxaban vs. warfarin or non-anticoagulant). To subsequently model impact of increasing rivaroxaban use on the Hospital Readmission Reduction Program (HRRP) penalty, which was imposed on hospitals with excess 30 day readmissions after hospitalizations for selected conditions, including THA/TKA. METHODS: The US Truven Health MarketScan Medicare Supplemental database from 1 July 2010 to 30 April 2015 was used. A retrospective claims analysis was conducted to assess the risk of all-cause 30 day readmission among patients receiving either rivaroxaban or warfarin, or no anticoagulation following THA/TKA discharge. Simulations were performed to estimate the impact of post-discharge treatment on the HRRP penalty. RESULTS: The risk-adjusted all-cause 30 day readmission rates were 1.21% (95% confidence interval [95% CI]: 0.94%-1.49%), 1.41% (95% CI: 1.19%-1.58%) and 1.95% (95% CI: 1.81%-2.11%) for rivaroxaban, warfarin and non-anticoagulant cohorts, respectively. Using these rates, simulations illustrated that when switching patients from warfarin or non-anticoagulant to rivaroxaban, annual penalty per hospital would be reduced up to 67% or 88%, respectively. CONCLUSIONS: Rivaroxaban treatment post-THA/TKA discharge reduced the risk of 30 day readmission compared to non-anticoagulants. Simulations illustrated that increasing rivaroxaban use could decrease the HRRP penalty.

Adherence to rivaroxaban versus apixaban among patients with non-valvular atrial fibrillation: Analysis of overall population and subgroups of prior oral anticoagulant users.

BACKGROUND: Medication non-adherence can result in poor health outcomes. Understanding differences in adherence rates to non-vitamin K oral anticoagulants (NOACs) could guide treatment decisions and improve clinical outcomes among patients with non-valvular atrial fibrillation (NVAF). OBJECTIVE: To compare adherence to rivaroxaban and apixaban among the overall NVAF population and subgroups of prior oral anticoagulant (OAC) users (e.g., multiple comorbidities, non-adherence risk factors). METHODS: Using healthcare claims from the Truven Health Analytics MarketScan (7/2012-7/2015), adult patients with ≥2 dispensings of rivaroxaban or apixaban ≥ 180 days apart with > 60 days of supply, ≥ 6 months of pre- and post-index eligibility, ≥ 1 atrial fibrillation diagnosis pre- or on the index date, and without valvular involvement were identified. Propensity-score methods adjusting for potential baseline confounders were used to create matched cohorts of rivaroxaban and apixaban patients. Adherence was assessed during the implementation phase using the percentage of patients with proportion of days covered (PDC) ≥0.8 at 6 months. Subgroups of patients with prior OAC use were evaluated; additional subgroups were identified and evaluated by Quan-Charlson Comorbidity index ≥2 and presence of non-adherence risk factors (i.e., mental disorders, stress, isolation, and rheumatoid arthritis). RESULTS: A total of 13,890 NVAF subjects were included in each of the 2 matched cohorts. All baseline characteristics were balanced between cohorts. At 6 months, significantly more rivaroxaban users were adherent to treatment compared to apixaban users (81.8% vs. 78.0%; absolute difference of 3.8%; p<.001). Rivaroxaban users had significantly higher adherence rates in all subgroups examined. CONCLUSION: Rivaroxaban users had consistently higher adherence rates than apixaban users overall and among all NVAF subgroups examined.

Adherence to Rivaroxaban Compared with Other Oral Anticoagulant Agents Among Patients with Nonvalvular Atrial Fibrillation.

BACKGROUND: Adherence to oral anticoagulant (OAC) agents is important for patients with nonvalvular atrial fibrillation (NVAF) to prevent potentially severe adverse events. OBJECTIVE: To compare real-world adherence rates and time to discontinuation for rivaroxaban versus other OACs (apixaban, dabigatran, and warfarin) among patients with NVAF using claims-based data. METHODS: Health care claims from the IMS Health Real-World Data Adjudicated Claims database (July 2012-June 2015) were analyzed. Adherence rate was defined as the percentage of patients with proportion of days covered (PDC) ≥ 0.80 and ≥ 0.90. Discontinuation was defined as a gap of more than 30 days between the end of a dispensing days of supply and the start date of the next fill, if any. Patients were included if they had ≥ 2 dispensings of rivaroxaban, apixaban, dabigatran, or warfarin at least 180 days apart (the first was considered the index date), had > 60 days of supply, had ≥ 6 months of pre-index eligibility, had ≥ 1 atrial fibrillation (AF) diagnosis pre-index or at index date, and had no valvular involvement. A logistic regression model was used to evaluate adherence to OAC therapy, while a Cox model was used to compare time to discontinuation; both models adjusted for baseline confounders. RESULTS: A total of 13,645 rivaroxaban, 6,304 apixaban, 3,360 dabigatran, and 13,366 warfarin patients were identified. A significantly higher proportion of rivaroxaban users (80.1%) was adherent to therapy (PDC ≥ 0.80 at 6 months) versus apixaban (75.8%), dabigatran (69.2%), and warfarin users (64.5%). After adjustment, the proportion of patients adherent to therapy remained significantly higher for rivaroxaban users versus apixaban (absolute difference [AD] = 5.8%), dabigatran (AD = 9.5%), and warfarin users (AD = 13.6%; all P < 0.001). More pronounced differences were found with a PDC ≥0.90. In addition, rivaroxaban users were significantly less likely to discontinue therapy compared with other OACs after adjustments (all P < 0.05). CONCLUSIONS: Among NVAF patients, rivaroxaban was associated with significantly higher adherence rates relative to other OACs whether using either a PDC of > 0.80 or > 0.90. Such differences in adherence could translate into improved patient outcomes and lower health care costs. DISCLOSURES: This research was funded by Janssen Scientific Affairs. Ashton, Crivera, and Schein are employees and stockholders of Janssen Scientific Affairs. Laliberté, Germain, Wynant, and Lefebvre are employees of Analysis Group, a consulting company that received research grants from Janssen Scientific Affairs in connection with this study. McHorney is an employee of Evidera, a consulting company that received research grants from Janssen Scientific Affairs in connection with this study. Peterson received research grants from Janssen Scientific Affairs in connection with this study. All authors contributed to concept and design. The data were collected by Germain, Wynant, Laliberté, and Lefebvre and interpreted primarily by McHorney and Peterson, with the assistance of Lefebvre, Laliberté, Ashton, Crivera, and Schein. The manuscript was written primarily by Laliberté, Germain, and Lefebvre, with the assistance of Wynant. Revisions were made primarily by Ashton, Crivera, McHorney, Schein, and Peterson.

Baseline characteristics and treatment patterns of patients with schizophrenia initiated on once-every-three-months paliperidone palmitate in a real-world setting.

OBJECTIVES: Since May 2015, adult patients with schizophrenia adequately treated with once monthly paliperidone palmitate (PP1M) may be transitioned to once-every-three-months paliperidone palmitate (PP3M). This study aims to describe baseline characteristics and treatment patterns of patients with schizophrenia initiated on PP3M in a real-world setting. METHODS: Pharmacy and medical claims from May 2014 to September 2016 for adult patients with schizophrenia initiated on PP3M (index date) in the Symphony Health Solutions database were analyzed. The cohort consisting of all patients and the one restricted to those transitioning from PP1M as per prescribing guideline recommendations were considered. Baseline characteristics were assessed during the 12 month baseline period. PP1M treatment patterns, proportion of days covered (PDC) by mental-health-related medications, and healthcare resource utilization (HRU) patterns were evaluated for each baseline quarter. PP3M treatment patterns were assessed post-index. RESULTS: Among the 1545 adult patients initiated on PP3M who formed the first cohort, 68.8% transitioned from PP1M based on prescribing guidelines and on an adaptation of the strict clinical trial protocol for PP1M to PP3M transition, forming the second cohort. In both cohorts, the proportion of patients with a PDC ≥80% for antipsychotics, antidepressants, anxiolytics, and mood stabilizers increased while the proportion of patients with ≥1 emergency room, inpatient, or outpatient visit decreased in baseline quarters closer to PP3M initiation. Among patients with ≥4 months of follow-up after the first dose, 85-88% had a second dose. Similarly, among those with ≥4 months of follow-up after the second dose, 87-90% received a third dose. CONCLUSIONS: Patients initiated on PP3M demonstrated decreased HRU and increased adherence in quarters closer to PP3M initiation, and were persistent on their PP3M treatment.

Validation of the alternating conditional estimation algorithm for estimation of flexible extensions of Cox's proportional hazards model with nonlinear constraints on the parameters.

Standard optimization algorithms for maximizing likelihood may not be applicable to the estimation of those flexible multivariable models that are nonlinear in their parameters. For applications where the model's structure permits separating estimation of mutually exclusive subsets of parameters into distinct steps, we propose the alternating conditional estimation (ACE) algorithm. We validate the algorithm, in simulations, for estimation of two flexible extensions of Cox's proportional hazards model where the standard maximum partial likelihood estimation does not apply, with simultaneous modeling of (1) nonlinear and time-dependent effects of continuous covariates on the hazard, and (2) nonlinear interaction and main effects of the same variable. We also apply the algorithm in real-life analyses to estimate nonlinear and time-dependent effects of prognostic factors for mortality in colon cancer. Analyses of both simulated and real-life data illustrate good statistical properties of the ACE algorithm and its ability to yield new potentially useful insights about the data structure.

Nonlinear association between betel quid chewing and oral cancer: Implications for prevention.

PURPOSE: Betel quid chewing is a major oral cancer risk factor and the human papillomaviruses (HPV) may play an aetiological role in these cancers. However, little is known about the shape of the dose-response relationship between the betel quid chewing habit and oral cancer risk in populations without HPV. We estimate the shape of this dose-response relationship, and discuss implications for prevention. METHODS: Cases with oral squamous cell carcinoma (350) and non-cancer controls (371) were recruited from two major teaching hospitals in South India. Information on socio-demographic and behavioral factors was collected using a questionnaire and the life grid technique. The effect of daily amount of use and duration of the habit were estimated jointly as risk associated with cumulative exposure (chew-years). The shape of the dose-response curve was estimated using restricted cubic spline transformation of chew-years in a conditional logistic regression model. Risk estimates for low dose combinations of daily amount and duration of the habit were computed from flexible regression. RESULTS: Most (72%) oral cancer cases were betel quid chewers in contrast to only 18% of controls. A nonlinear dose-response relationship was observed; the risk increased steeply at low doses and plateaued at high exposures to betel quid (>425 chew-years). A threefold increase in risk (OR=3.92, 95%CI: 1.87-8.21) was observed for the lowest dose; equivalent to the use of one quid per day for one year. CONCLUSION: Our findings may be used to counsel people to refrain from even low betel quid chewing.

Flexible estimation of survival curves conditional on non-linear and time-dependent predictor effects.

Prognostic studies often estimate survival curves for patients with different covariate vectors, but the validity of their results depends largely on the accuracy of the estimated covariate effects. To avoid conventional proportional hazards and linearity assumptions, flexible extensions of Cox's proportional hazards model incorporate non-linear (NL) and/or time-dependent (TD) covariate effects. However, their impact on survival curves estimation is unclear. Our primary goal is to develop and validate a flexible method for estimating individual patients' survival curves, conditional on multiple predictors with possibly NL and/or TD effects. We first obtain maximum partial likelihood estimates of NL and TD effects and use backward elimination to select statistically significant effects into a final multivariable model. We then plug the selected NL and TD estimates in the full likelihood function and estimate the baseline hazard function and the resulting survival curves, conditional on individual covariate vectors. The TD and NL functions and the log hazard are modeled with unpenalized regression B-splines. In simulations, our flexible survival curve estimates were unbiased and had much lower mean square errors than the conventional estimates. In real-life analyses of mortality after a septic shock, our model improved significantly the deviance (likelihood ratio test = 84.8, df = 20, p < 0.0001) and changed substantially the predicted survival for several subjects.

Flexible modeling of disease activity measures improved prognosis of disability progression in relapsing-remitting multiple sclerosis.

OBJECTIVES: To illustrate the advantages of updating time-varying measures of disease activity and flexible modeling in prognostic clinical studies using the example of the association between the frequency of past relapses and occurrence of ambulation-related disability in multiple sclerosis (MS). STUDY DESIGN AND SETTING: Longitudinal population-based study of 288 patients from Burgundy, France, diagnosed with relapsing-remitting MS in 1990-2003. The end point was a nonreversible moderate MS disability (European Database for Multiple Sclerosis score ≥ 3.0 derived from Extended Disability Status Scale). Alternative time-varying measures of attacks frequency included (1) conventional number of early MS attacks in the first 2 years after diagnosis; and two new measures, continuously updated during the follow-up; (2) cumulative number of past attacks; and (3) number of recent attacks, during the past 2 years. Multivariate analyses used Cox proportional hazards model and its flexible generalization, which accounted for time-dependent changes in the hazard ratios (HRs) for different attack frequency measures. RESULTS: HRs for all measures decreased significantly with increasing follow-up time. The proposed updated number of recent attacks improved model's fit to data, relative to alternative measures of attack frequency, and was associated with a statistically significantly increased hazard of developing ambulation-related MS disability in the next 2 years during the entire follow-up period. CONCLUSION: Updated measures of recent disease activity, such as frequency of recent attacks and modeling of their time-dependent effects, may substantially improve prognosis of clinical outcomes, such as development of MS disability.

Variation in long-term antipsychotic polypharmacy and high-dose prescribing across physicians and hospitals.

OBJECTIVES: This study had two aims: to measure the prevalence of long-term prescribing of high doses of antipsychotics and antipsychotic polypharmacy in a large Canadian province and to estimate the relative contributions of patient-, physician-, and hospital-level factors. METHODS: Government hospital discharge, physician, and pharmaceutical claims data were linked to identify individuals with schizophrenia who in 2004 had antipsychotics available to them for at least 11 months. Individuals on a high dose throughout that period, as well as individuals on multiple concurrent antipsychotics (polypharmacy), were identified. Logistic and generalized linear mixed models using patient-, physician-, and hospital-level predictors were estimated. RESULTS: Among the 12,150 individuals identified, 11.9% were on a high dose and 10.4% on antipsychotic polypharmacy continually, with 3.7% in both groups. After adjustment for potential confounders, analyses showed that systematic propensity for physicians to prescribe high doses accounted for 10.9% of the remaining unexplained variance, and physicians as a group who prescribed high doses across a hospital or psychiatry department accounted for 3.0%. For antipsychotic polypharmacy the corresponding percentages were 9.7% and 6.2%. Even after adjustment, the variation in high-dose prescribing and antipsychotic polypharmacy remained substantial. CONCLUSIONS: Long-term high-dose and antipsychotic polypharmacy prescribing appeared partly driven by some physicians' and some hospitals' propensities to prescribe in this way independently of patient characteristics. Given the weight of the evidence against high-dose prescribing and antipsychotic polypharmacy, measures addressed to physicians and hospitals most likely to prescribe high doses, antipsychotic polypharmacy, or both should be considered.

Impact of the model-building strategy on inference about nonlinear and time-dependent covariate effects in survival analysis.

Cox's proportional hazards (PH) model assumes constant-over-time covariate effects. Furthermore, most applications assume linear effects of continuous covariates on the logarithm of the hazard. Yet, many prognostic factors have time-dependent (TD) and/or nonlinear (NL) effects, that is, violate these conventional assumptions. Detection of such complex effects could affect prognosis and clinical decisions. However, assessing the effects of each of the multiple, often correlated, covariates in flexible multivariable analyses is challenging. In simulations, we investigated the impact of the approach used to build the flexible multivariable model on inference about the TD and NL covariate effects. Results demonstrate that the conclusions regarding the statistical significance of the TD/NL effects depend heavily on the strategy used to decide which effects of the other covariates should be adjusted for. Both a failure to adjust for true TD and NL effects of relevant covariates and inclusion of spurious effects of covariates that conform to the PH and linearity assumptions increase the risk of incorrect conclusions regarding other covariates. In this context, iterative backward elimination of nonsignificant NL and TD effects from the multivariable model, which initially includes all these effects, may help discriminate between true and spurious effects. The practical importance of these issues was illustrated in an example that reassessed the predictive ability of selected biomarkers for survival in advanced non-small-cell lung cancer. In conclusion, a careful model-building strategy and flexible modeling of multivariable survival data can yield new insights about predictors' roles and improve the validity of analyses.

Gestational age-dependent risk factors for preterm birth: associations with maternal education and age early in gestation.

OBJECTIVES: Preterm birth (PTB) before 37 weeks can occur over a wide range of gestational ages, but few studies have assessed if associations between risk factors and PTB vary over the duration of gestation. We sought to evaluate if associations between two major risk factors (maternal education and age) and PTB depend on gestational age at delivery. STUDY DESIGN: We estimated hazard ratios of PTB for education and age in a time-to-event analysis using a retrospective cohort of 223,756 live singleton births from the province of Québec, Canada for the years 2001-2005. Differences in hazards of maternal education and age with PTB were assessed over gestational age in a Cox proportional hazards model using linear and nonlinear time interaction terms, adjusting for maternal characteristics. RESULTS: Associations of PTB with lower (vs. higher) education and older (vs. younger) age strengthened progressively at earlier gestational ages, such that the risk of PTB for maternal education and age was not constant over the course of gestation. CONCLUSIONS: Associations of PTB with risk factors such as maternal low education and older age may be stronger early in gestation. Models that capture the time-dependent nature of PTB may be useful when the goal is to assess associations at low gestational ages, and to avoid masked or biased associations early in gestation.

Underprescribing of clozapine and unexplained variation in use across hospitals and regions in the Canadian province of Québec.

BACKGROUND: Clozapine remains the antipsychotic of choice for people who, having met the criteria for a diagnosis of schizophrenia or a related psychotic disorder, do not respond adequately to other antipsychotic medications. Utilization rates appear highly variable across jurisdictions, with an overall tendency toward underuse. This paper describes patterns of clozapine use in the province of Québec, Canada. METHODS: Individuals with a diagnosis of schizophrenia were identified using linked government medical claims and hospitalization records for 2003 and 2004. Linked data on their filled prescriptions in 2004 were then used to determine clozapine-use rates at the level of the province, the region, and the hospital at which individuals received most of their services. Individual predictors of clozapine use were identified using logistic regression. RESULTS: Only 6.7% of the 29,155 individuals identified with schizophrenia received clozapine for six months or longer in 2004. Utilization rates ranged from 3.9 to 9.0% among regions with 1,000 or more people with schizophrenia. Over 8% of 61 hospitals did not prescribe clozapine at all. People with schizophrenia taking clozapine experienced 3.4 fewer days of hospitalization per year than those not taking clozapine-representing a cost offset of about $1,800 per year. Medication costs were higher, however, by about $3,000 per year. CONCLUSIONS: Given the increasingly clear benefits of clozapine for people who do not respond to other antipsychotics, measures to increase access to clozapine for people who can benefit from it are likely to be cost effective and are urgently needed.

Patterns of health services use prior to a first diagnosis of psychosis: the importance of primary care.

PURPOSE: The observed association between treatment delay and poor outcomes in first-episode psychosis has led to an interest in the topography of symptom development preceding the onset of psychosis and associated help-seeking behaviors. We estimated the extent to which socio-demographic, clinical, and health service indicators are associated with patterns of service use for mental health reasons preceding a first diagnosis of psychosis. METHODS: Population-based administrative data from physician billings, hospitalizations, and public health clinics were used to identify incident cases of schizophrenia-spectrum psychosis among individuals aged 14-25 years in Montréal. Mental health contacts in the 4 years preceding the index diagnosis were analyzed. RESULTS: Thirty-two percent of cases had no contact with services for a mental health reason preceding the index diagnosis, and nearly 50% received the index diagnosis of psychosis in the emergency department. Individuals in contact with primary care had a reduced likelihood of contact with the emergency department and inpatient services (OR = 0.15, 0.06-0.39) and of receiving the index diagnosis in the emergency department (OR = 0.36, 0.24-0.54), but also had a longer time to contact with a psychiatrist (HR = 0.32, 0.23-0.45). CONCLUSIONS: Improving access to primary care may reduce the burden on emergency departments and inpatient units; however, primary care providers may need additional training in the symptoms of early psychosis and referral protocols. Given the limitations associated with using clinical samples from specialized services, population-based administrative data are an important source of information for understanding patterns of health services use preceding a first diagnosis of psychosis.

Occupational exposure to lead and lung cancer: results from two case-control studies in Montreal, Canada.

OBJECTIVES: We investigated the association between workplace lead exposure and lung cancer risk, separately for organic lead and for inorganic lead, from either engine emissions or from other sources. METHODS: Two population-based case-control studies were carried out in Montreal (1979-1986 and 1996-2002) to investigate occupational factors in relation to lung cancer among 1593 men with histologically confirmed incident lung cancer, and 1426 controls from the general population. Interviews elicited information on sociodemographic characteristics, lifetime smoking and occupational history. Chemists translated each job into potential chemical exposures. Cumulative indices of exposure were derived and classified into non-substantial and substantial exposure. ORs adjusted for several potential confounders including smoking, and 95% CIs were estimated by logistic regression. RESULTS: Lifetime prevalences of exposure in Study I were 3% for organic lead, 40% for inorganic lead from engine emissions and 17% for inorganic lead from other sources; corresponding prevalences in Study II were 4%, 19% and 16%, respectively. No associations were observed when comparing ever to never exposed subjects in pooled analyses (organic lead, OR=1.39, 95% CI 0.77 to 2.52; inorganic lead from engine emissions: OR=0.89, 95% CI 0.72 to 1.09; inorganic lead from other sources: OR=0.99, 95% CI 0.76 to 1.29). Nor were these exposures associated with lung cancer in subjects with substantial cumulative exposure. CONCLUSIONS: In this large study, using a blinded expert-based assessment of lifetime occupational exposure and adjustment for several potential confounders, we observed no increased risk of lung cancer with exposure to lead compounds.

Immediate and past cumulative effects of oral glucocorticoids on the risk of acute myocardial infarction in rheumatoid arthritis: a population-based study.

OBJECTIVES: To determine the effect of glucocorticoids (GCs) on acute myocardial infarction (MI) risk in patients with RA. METHODS: Using administrative health data, we conducted a population-based cohort study of 8384 incident RA cases (1997-2006). Primary exposure was incident GC use. MI events were ascertained using hospitalization and vital statistics data. We used Cox proportional-hazards models and modelled GC use as four alternative time-dependent variables (current use, current dose, cumulative dose and cumulative duration), adjusting for demographics, comorbidities, cardiovascular drug use, propensity score and RA characteristics. Sensitivity analyses explored potential effects of unmeasured confounding. RESULTS: Within 50 238 person-years in 8384 RA cases, we identified 298 incident MI events. Multivariable models showed that current GC use was associated with 68% increased risk of MI [Hazard ratio (HR) = 1.68, 95% CI 1.14, 2.47]. Similarly, separate multivariable models showed that current daily dose (HR = 1.14, 95% CI 1.05, 1.24 per each 5 mg/day increase), cumulative duration of use (HR = 1.14, 95% CI 1.00, 1.29 per year of GC use) and total cumulative dose (HR = 1.06, 95% CI 1.02, 1.10 per gram accumulated in the past) were also associated with increased risk of MI. Furthermore, in the same multivariable model, current dose and cumulative use were independently associated with an increased risk of MI (10% per additional year on GCs and 13% per 5 mg/day increase). CONCLUSION: GCs are associated with an increased risk of MI in RA. Our results suggest a dual effect of GCs on MI risk, an immediate effect mediated through current dosage and a long-term effect of cumulative exposure.

Extreme maternal education and preterm birth: time-to-event analysis of age and nativity-dependent risks.

PURPOSE: Increasing numbers of women achieve extremely high education, but the association with preterm birth (PTB) is poorly understood, especially over the life course. We sought to determine how very high educational attainment is associated with PTB, and to assess differences by maternal age and nativity. METHODS: Data included singleton live births to mothers aged ≥ 20 years in metropolitan areas of Québec, Canada, from 1995 to 2005 (n = 537,525). Hazard ratios of PTB (<37 gestational weeks) were estimated over the continuous range of education (0-30 years) according to maternal age (20-24, 25-29, 30-34, ≥ 35 years) and nativity in a flexible survival model. RESULTS: The relationship between education and PTB was not linear, but suggested that extremely high education was not as protective against PTB as slightly lower education. Education thresholds that offered maximum protection increased with maternal age, and were lower for Canadian-born (17-21 years of education) than foreign-born (22-25 years of education) mothers. CONCLUSIONS: Extremely high education did not confer more protection against PTB than slightly lower education, and associations varied over the life course. The threshold number of years of education most protective against PTB: (1) increased with maternal age, especially for Canadian-born mothers, and (2) was higher for foreign-born mothers.