OBJECTIVE: To examine the association between first-trimester angiotensin-converting enzyme (ACE) inhibitor exposure and the risk of overall major congenital, cardiac, and central nervous system malformations.
METHODS: We used a cohort of completed pregnancies linked to liveborn neonates derived from Medicaid claims from 2000 to 2010. We examined the risk of malformations associated with first-trimester exposure to an ACE inhibitor. Propensity score-based methods were used to control for potential confounders including maternal demographics, medical conditions, exposure to other medications, and measures of health care utilization.
RESULTS: The cohort included 1,333,624 pregnancies, of which 4,107 (0.31%) were exposed to ACE inhibitors during the first trimester. The prevalence of overall malformations in the ACE inhibitor-exposed pregnancies was 5.9% compared with 3.3% in the unexposed (unadjusted relative risk, 1.82; 95% confidence interval [CI] 1.61-2.06), of cardiac malformations was 3.4% compared with 1.2% (relative risk 2.95, 95% CI 2.50-3.47), and of central nervous system malformations was 0.27% compared with 0.18% (relative risk 1.46, 95% CI 0.81-2.64). After restricting the cohort to pregnancies complicated by chronic hypertension (both exposed and unexposed) and accounting for other confounding factors, there was no significant increase in the risk of any of the outcomes assessed. Relative risks associated with first-trimester ACE inhibitor exposure were 0.89 (95% CI 0.75-1.06) for overall malformations, 0.95 (95% CI 0.75-1.21) for cardiac malformations, and 0.54 (95% CI 0.26-1.11) for CNS malformations.
CONCLUSIONS: After accounting for confounders, among women with hypertension, exposure to ACE inhibitors during the first trimester was not associated with an increased risk of major congenital malformations.
OBJECTIVE: Given the increasing use and broadening of indications for use of antipsychotic medications in the general population, as well as the paucity of information on the safety of this drug class during pregnancy, the study documented patterns of antipsychotic medication use among pregnant women.
METHODS: Medicaid Analytic eXtract data (2001-2010) from pregnant women who delivered live-born infants were used. Antipsychotic use at both the class and the individual drug level was defined based on dispensed outpatient prescriptions. Users' characteristics, including mental disorder diagnoses, were described. Temporal trends in use, as well as discontinuation patterns and psychotropic polytherapy, during pregnancy were evaluated.
RESULTS: Among 1,522,247 pregnancies, the prevalence of use of second-generation antipsychotics at any time during pregnancy increased threefold, from .4% to 1.3%, over the ten-year period, while the use of first-generation antipsychotics remained stable at around .1%. The increased use of second-generation antipsychotics was largely driven by more frequent use among patients with bipolar disorder. Quetiapine and aripiprazole were the most frequently dispensed drugs, and polytherapy with antipsychotics and antidepressants (65.2%), benzodiazepines (24.9%), and other mood stabilizers (22.0%) was common. More than 50% of women receiving an antipsychotic in the three months prior to pregnancy discontinued the drug during pregnancy.
CONCLUSIONS: A growing number of pregnant women in Medicaid are exposed to second-generation antipsychotics, frequently in combination with other psychotropic medications. This study highlights the importance of documenting the use and safety of these drugs during pregnancy to inform therapeutic decision making for pregnant women with psychiatric disorders.
BACKGROUND: Inflammatory bowel disease (IBD) increases the risk of venous thromboembolism (VTE) by 2 to 3 times. We compared the reduction in risk of incident VTE associated with use of tumour necrosis factor-α (TNF-α) inhibitors versus nonbiologic immunomodulatory agents in patients with IBD.
METHODS: This observational cohort study used data from public (Medicaid, 2000-2010; Medicare, 2007-2013) and private (Optum Clinformatics, 2004-2013) health insurance programs in the United States. We included a total of 21 671 patients who had IBD without a prior diagnosis of cancer or VTE. The exposure of interest was treatment initiation with TNF-α inhibitor or nonbiologic (azathioprine, mercaptopurine, methotrexate, cyclosporine). The outcome of interest was admission to hospital with VTE as the principal diagnosis. We used Cox proportional hazard regression models to estimate hazard ratios (HRs) separately for each database after risk adjustment for more than 50 covariables using propensity score fine stratification. We used inverse variance meta-analytic methods to pool the adjusted HRs across the 3 databases.
RESULTS: We included a total of 5173 patients who started TNF-α inhibitor therapy (1439 in the Medicaid database, 1480 in Medicare and 2254 in Optum Clinformatics) and 16 498 who initiated a nonbiologic agent (5041 in Medicaid, 5166 in Medicare, 6291 in Optum Clinformatics). The adjusted pooled HR for VTE risk with TNF-α inhibitor versus a nonbiologic agent was 0.78 (95% confidence interval [CI] 0.60 to 1.02). The HR was lower in patients with Crohn disease (pooled HR 0.62, 95% CI 0.44 to 0.86) and younger patients (18-44 yr; pooled HR 0.55, 95% CI 0.34 to 0.87).
INTERPRETATION: We did not find a statistically significant association between risk of VTE and use of TNF-α inhibitors, relative to nonbiologics, in patients with IBD overall. However, an association was evident for patients younger than 45 years and those with Crohn disease.
STUDY OBJECTIVE: To describe contemporary trends of P2Y12 inhibitor use in patients with acute coronary syndrome (ACS) and comorbid diabetes mellitus (DM) and/or chronic kidney disease (CKD) who have a higher risk of recurring ACS and may benefit from treatment with higher efficacy third-generation agents (prasugrel and ticagrelor).
DESIGN: Observational cohort study.
SETTING: A large U.S. commercial insurance program (2009-2015).
PATIENTS: P2Y12 inhibitor initiated within 2 weeks after an ACS event.
MEASUREMENTS AND MAIN RESULTS: We identified 98,649 P2Y12 inhibitor initiators, of whom 24.5% had comorbid DM (no CKD), 10.5% had CKD (no DM), and 12.6% had DM and CKD. Overall, 85.2% of patients initiated clopidogrel, followed by prasugrel (11.6%) and ticagrelor (3.2%). Prasugrel use decreased over time irrespective of preexisting DM and/or CKD; ticagrelor use increased. In logistic regression models accounting for patient demographics and clinical covariates, preexisting DM alone was not associated with prasugrel or ticagrelor versus clopidogrel treatment initiation; however, having CKD with or without DM significantly reduced the likelihood of receiving prasugrel versus clopidogrel (odds ratio [OR] 0.81, 95% confidence interval [CI] 0.74-0.88 for CKD alone; OR 0.91, 95% CI 0.83-0.98 for DM and CKD). Comorbid DM and CKD reduced the odds of initiating ticagrelor versus clopidogrel (OR 0.80, 95% CI 0.70-0.92).
PRINCIPAL CONCLUSIONS: We observed lower or similar use of prasugrel and ticagrelor compared with clopidogrel in patients with ACS and comorbid DM and/or CKD. Given the potential for worse clinical outcomes with clopidogrel in these patients, our findings highlight the need to investigate the implications of these trends on recurrent ACS and bleeding events.
A propensity score (PS) model's ability to control confounding can be assessed by evaluating covariate balance across exposure groups after PS adjustment. The optimal strategy for evaluating a disease risk score (DRS) model's ability to control confounding is less clear. DRS models cannot be evaluated through balance checks within the full population, and they are usually assessed through prediction diagnostics and goodness-of-fit tests. A proposed alternative is the "dry-run" analysis, which divides the unexposed population into "pseudo-exposed" and "pseudo-unexposed" groups so that differences on observed covariates resemble differences between the actual exposed and unexposed populations. With no exposure effect separating the pseudo-exposed and pseudo-unexposed groups, a DRS model is evaluated by its ability to retrieve an unconfounded null estimate after adjustment in this pseudo-population. We used simulations and an empirical example to compare traditional DRS performance metrics with the dry-run validation. In simulations, the dry run often improved assessment of confounding control, compared with the C statistic and goodness-of-fit tests. In the empirical example, PS and DRS matching gave similar results and showed good performance in terms of covariate balance (PS matching) and controlling confounding in the dry-run analysis (DRS matching). The dry-run analysis may prove useful in evaluating confounding control through DRS models.
BACKGROUND: Biologic disease-modifying antirheumatic drugs (DMARDs) are increasingly used for rheumatoid arthritis (RA) treatment. However, little is known based on contemporary data about the factors associated with DMARDs and patterns of use of biologic DMARDs for initial and subsequent RA treatment.
METHODS: We conducted an observational cohort study using claims data from a commercial health plan (2004-2013) and Medicaid (2000-2010) in three study groups: patients with early untreated RA who were naïve to any type of DMARD and patients with prevalent RA with or without prior exposure to one biologic DMARD. Multivariable logistic regression models were used to examine the effect of patient demographics, clinical characteristics and healthcare utilization factors on the initial and subsequent choice of biologic DMARDs for RA.
RESULTS: We identified a total of 195,433 RA patients including 78,667 (40%) with early untreated RA and 93,534 (48%) and 23,232 (12%) with prevalent RA, without or with prior biologic DMARD treatment, respectively. Patients in the commercial insurance were 87% more likely to initiate a biologic DMARD versus patients in Medicaid (OR = 1.87, 95% CI = 1.70-2.05). In Medicaid, African-Americans had lower odds of initiating (OR = 0.59, 95% CI = 0.51-0.68 in early untreated RA; OR = 0.71, 95% CI = 0.61-0.74 in prevalent RA) and switching (OR = 0.71, 95% CI = 0.55-0.90) biologic DMARDs than non-Hispanic whites. Prior use of steroid and non-biologic DMARDs predicted both biologic DMARD initiation and subsequent switching. Etanercept, adalimumab, and infliximab were the most commonly used first-line and second-line biologic DMARDS; patients on anakinra and golimumab were most likely to be switched to other biologic DMARDS.
CONCLUSIONS: Insurance type, race, and previous use of steroids and non-biologic DMARDs were strongly associated with initial or subsequent treatment with biologic DMARDs.
BACKGROUND: There has been concern that exposure to lithium early in pregnancy may be associated with a marked increase in the risk of Ebstein's anomaly (a right ventricular outflow tract obstruction defect) in infants and overall congenital cardiac defects, but data are conflicting and limited.
METHODS: We conducted a cohort study involving 1,325,563 pregnancies in women who were enrolled in Medicaid and who delivered a live-born infant between 2000 and 2010. We examined the risk of cardiac malformations among infants exposed to lithium during the first trimester as compared with unexposed infants and, in secondary analyses, with infants exposed to another commonly used mood stabilizer, lamotrigine. Risk ratios and 95% confidence intervals were estimated with control for psychiatric and medical conditions, medications, and other potential confounders.
RESULTS: Cardiac malformations were present in 16 of the 663 infants exposed to lithium (2.41%), 15,251 of the 1,322,955 nonexposed infants (1.15%), and 27 of the 1945 infants exposed to lamotrigine (1.39%). The adjusted risk ratio for cardiac malformations among infants exposed to lithium as compared with unexposed infants was 1.65 (95% confidence interval [CI], 1.02 to 2.68). The risk ratio was 1.11 (95% CI, 0.46 to 2.64) for a daily dose of 600 mg or less, 1.60 (95% CI, 0.67 to 3.80) for 601 to 900 mg, and 3.22 (95% CI, 1.47 to 7.02) for more than 900 mg. The prevalence of right ventricular outflow tract obstruction defects was 0.60% among lithium-exposed infants versus 0.18% among unexposed infants (adjusted risk ratio, 2.66; 95% CI, 1.00 to 7.06). Results were similar when lamotrigine-exposed infants were used as the reference group.
CONCLUSIONS: Maternal use of lithium during the first trimester was associated with an increased risk of cardiac malformations, including Ebstein's anomaly; the magnitude of this effect was smaller than had been previously postulated. (Funded by the National Institute of Mental Health.).
OBJECTIVE: To evaluate whether psychostimulants used to treat attention-deficit/hyperactivity disorder (ADHD) are associated with risk of adverse placental-associated pregnancy outcomes including preeclampsia, placental abruption, growth restriction, and preterm birth.
METHODS: We designed a population-based cohort study in which we examined a cohort of pregnant women and their liveborn neonates enrolled in Medicaid from 2000 to 2010. Women who received amphetamine-dextroamphetamine or methylphenidate monotherapy in the first half of pregnancy were compared with unexposed women. We considered atomoxetine, a nonstimulant ADHD medication, as a negative control exposure. To assess whether the risk period extended to the latter half of pregnancy, women who continued stimulant monotherapy after 20 weeks of gestation were compared with those who discontinued. Risk ratios and 95% CIs were estimated with propensity score stratification to control for confounders.
RESULTS: Pregnancies exposed to amphetamine-dextroamphetamine (n=3,331), methylphenidate (n=1,515), and atomoxetine (n=453) monotherapy in early pregnancy were compared with 1,461,493 unexposed pregnancies. Among unexposed women, the risks of the outcomes were 3.7% for preeclampsia, 1.4% for placental abruption, 2.9% for small for gestational age, and 11.2% for preterm birth. The adjusted risk ratio for stimulant use was 1.29 for preeclampsia (95% CI 1.11-1.49), 1.13 for placental abruption (0.88-1.44), 0.91 for small for gestational age (0.77-1.07), and 1.06 for preterm birth (0.97-1.16). Compared with discontinuation (n=3,527), the adjusted risk ratio for continuation of stimulant use in the latter half of pregnancy (n=1,319) was 1.26 for preeclampsia (0.94-1.67), 1.08 for placental abruption (0.67-1.74), 1.37 for small for gestational age (0.97-1.93), and 1.30 for preterm birth (1.10-1.55). Atomoxetine was not associated with the outcomes studied.
CONCLUSION: Psychostimulant use during pregnancy was associated with a small increased relative risk of preeclampsia and preterm birth. The absolute increases in risks are small and, thus, women with significant ADHD should not be counseled to suspend their ADHD treatment based on these findings.
BACKGROUND: Many countries lack fully functional pharmacovigilance programs, and public budgets allocated to pharmacovigilance in industrialized countries remain low due to resource constraints and competing priorities.
OBJECTIVE: Using 3 case examples, we sought to estimate the public health and economic benefits resulting from public investment in active pharmacovigilance programs to detect adverse drug effects.
RESEARCH DESIGN: We assessed 3 examples in which early signals of safety hazards were not adequately recognized, resulting in continued exposure of a large number of patients to these drugs when safer and effective alternative treatments were available. The drug examples studied were rofecoxib, cerivastatin, and troglitazone. Using an individual patient simulation model and the health care system perspective, we estimated the potential costs that could have been averted by early systematic detection of safety hazards through the implementation of active surveillance programs.
RESULTS: We found that earlier drug withdrawal made possible by active safety surveillance would most likely have resulted in savings in direct medical costs of $773-$884 million for rofecoxib, $3-$10 million for cerivastatin, and $38-$63 million for troglitazone in the United States through the prevention of adverse events. By contrast, the yearly public investment in Food and Drug Administration initiated population-based pharmacovigilance activities in the United States is about $42.5 million at present.
CONCLUSION: These examples illustrate a critical and economically justifiable role for active adverse effect surveillance in protecting the health of the public.
OBJECTIVE: To assess whether first-trimester exposure to pregabalin is associated with an increased risk of major congenital malformations, as recently suggested in a pregnancy registry study.
METHODS: We performed a cohort study nested in the US Medicaid Analytic eXtract (MAX). The study population included 1,323,432 pregnancies resulting in a live-born infant between 2000 and 2010. We examined the risk of major congenital malformations among infants born to women exposed to pregabalin during the first trimester compared with women unexposed to anticonvulsants. We used propensity score fine stratification to control for >50 potential confounders, and we estimated relative risks (RRs) and 95% confidence intervals (CIs) in generalized linear models. The analyses were replicated in the Truven Health MarketScan Commercial Database (MarketScan). Pooled estimates based on the adjusted RR produced in MAX, MarketScan, and the previous registry study were calculated.
RESULTS: Of 477 infants exposed to pregabalin during the first trimester in MAX, 28 (5.9%) had malformations compared to 3.3% in nonexposed infants. The crude RR of major congenital malformations for pregabalin was 1.80 (95% CI 1.26-2.58). After propensity score adjustment, the RR moved to 1.16 (95% CI 0.81-1.67). Restriction to pregabalin monotherapy and sensitivity analyses produced similar results. The adjusted RR for major congenital malformations for the 174 infants exposed in MarketScan was 1.03 (95% CI 0.56-1.90). The pooled RR was 1.33 (95% CI 0.83-2.15) for pregabalin any use and 1.02 (95% CI 0.69-1.51) for pregabalin monotherapy.
CONCLUSIONS: Findings did not confirm the suggested teratogenic effects of pregabalin, although they cannot rule out the possibility of a small effect.
BACKGROUND: When exposure is infrequent, propensity-score matching results in reduced precision because it discards a large proportion of unexposed patients. To our knowledge, the relative performance of propensity-score stratification in these circumstances has not been examined.
METHODS: Using an empirical example of the association of first trimester statin exposure (prevalence = 0.04%) with risk of congenital malformations and 1,000 simulated cohorts (n = 20,000) with eight combinations of exposure prevalence (0.5%, 1%, 5%, 10%) and outcome risk (3.5%, 10%), we compared four propensity-score-based approaches to confounding adjustment: (1) matching (1:1, 1:5, full), (2) stratification in 10, 50, and 100 strata by entire cohort propensity-score distribution, (3) stratification in 10, 50, and 100 strata by exposed group propensity-score distribution, (4) standardized mortality ratio (SMR) weighting. Weighted generalized linear models were used to derive effect estimates after weighting unexposed according to the distribution of the exposed in their stratum for the stratification approaches.
RESULTS: In the empirical example, propensity-score stratification (cohort) approaches resulted in greater imbalances in covariate distributions between statin-exposed and unexposed compared with propensity-score stratification (exposed) and matching. In simulations, propensity-score stratification (exposed) resulted in smaller relative bias than the cohort approach with 10 and 50 strata, and greater precision than matching and SMR weighting at 0.5% and 1% exposure prevalence, but similar performance at 5% and 10%.
CONCLUSION: For exposures with prevalence under 5%, propensity-score stratification with fine strata, based on the exposed group propensity-score distribution, produced the best results. For more common exposures, all approaches were equivalent.
To assess the impact of in utero co-exposure to psychotropic medications and opioids on the incidence and severity of neonatal drug withdrawal. Observational cohort study. Nationwide sample of pregnancies in publicly insured women in the US, nested in the Medicaid Analytic eXtract (2000-10). 201 275 pregnant women with public insurance who were exposed to opioids around the time of delivery and their liveborn infants. In utero exposure to psychotropic medications, in particular antidepressants, atypical antipsychotics, benzodiazepines, gabapentin, and non-benzodiazepine hypnotics (Z drugs), with prescriptions filled within the same time window as prescriptions for opioids. Diagnosis of neonatal drug withdrawal in infants exposed in utero to opioids and psychotropic medications compared with opioids alone. The absolute risk for neonatal drug withdrawal ranged from 1.0% in infants exposed in utero to prescription opioids alone to 11.4% for those exposed to opioids co-prescribed with gabapentin. Among neonates exposed in utero to prescription opioids, the relative risk adjusted for propensity score was 1.34 (95% confidence interval 1.22 to 1.47) with concomitant exposure to antidepressants, 1.49 (1.35 to 1.63) with benzodiazepines, 1.61 (1.26 to 2.06) with gabapentin, 1.20 (0.95 to 1.51) with antipsychotics, and 1.01 (0.88 to 1.15) with Z drugs. In utero exposure to two or more psychotropic medications along with opioids was associated with a twofold increased risk of withdrawal (2.05, 1.77 to 2.37). The severity of the withdrawal seemed increased in neonates exposed to both opioids and psychotropic medications compared with opioids alone. During pregnancy, the use of psychotropic medications in addition to prescription opioids is common, despite a lack of safety data. The current findings suggest that these drugs could further increase the risk and severity of neonatal drug withdrawal.
To compare the risk of serious infections associated with use of systemic steroids, non-biologic agents, or tumor necrosis factor α (TNF) inhibitors in pregnancy. Observational cohort study. Public (Medicaid, 2001-10) or private (Optum Clinformatics, 2004-15) health insurance programs in the US. 4961 pregnant women treated with immunosuppressive drugs for rheumatoid arthritis, systemic lupus erythematosus, ankylosing spondylitis, psoriatic arthritis, or inflammatory bowel disease. Exposure was classified into steroid, non-biologic, or TNF inhibitors on first filled prescription during pregnancy. Because TNF inhibitors are not used to treat systemic lupus erythematosus, patients with this condition were excluded from comparisons involving TNF inhibitors. The main outcome was occurrence of serious infections during pregnancy, defined by hospital admission for bacterial or opportunistic infections. Hazard ratios were derived using Cox proportional hazard regression models after adjustment for confounding with propensity score fine stratification. A logistic regression model was used to conduct a dose-response analysis among women filling at least one steroid prescription. 71 out of 4961 pregnant women (0.2%) treated with immunosuppressive agents experienced serious infections. The crude incidence rates of serious infections per 100 person years among 2598 steroid users, 1587 non-biologic users, and 776 TNF inhibitors users included in this study were 3.4 (95% confidence interval 2.5 to 4.7), 2.3 (1.5 to 3.5), and 1.5 (0.7 to 3.0), respectively. No statistically significant differences in the risk of serious infections during pregnancy were observed among users of the three immunosuppressive drug classes: non-biologics steroids, hazard ratio 0.81 (95% confidence interval 0.48 to 1.37), TNF inhibitors steroids 0.91 (0.36 to 2.26), and TNF inhibitors non-biologics 1.36 (0.47 to 3.93). In the dose-response analysis, higher steroid dose was associated with an increased risk of serious infections during pregnancy (coefficient for each unit increase in average prednisone equivalent mg daily dose=0.019, P=0.02). Risk of serious infections is similar among pregnant women with systemic inflammatory conditions using steroids, non-biologics, and TNF inhibitors. However, high dose steroid use is an independent risk factor of serious infections in pregnancy.
BACKGROUND: Ten biologic disease-modifying antirheumatic drugs (bDMARDs) are available as treatment for rheumatoid arthritis (RA), but relatively little is known about population-level time trends in the use of these agents.
OBJECTIVE: To describe time trends in the use of bDMARDs in RA patients with private or public insurance in the United States.
METHODS: Claims data from private (Optum Clinformatics, 2004-2015) and public (Medicaid Analytic eXtract [MAX], 2000-2010) insurance programs were used. Patients with RA diagnosis codes and continuous health plan enrollment for 1-year baseline and 1-year follow-up periods were identified into 2 separate cohorts: (1) patients not using any bDMARD or (2) patients using a single bDMARD during the baseline period. Initiation of the first bDMARD from group 1 and switch to a second bDMARD from group 2 was identified as the outcome of interest during the 1-year follow-up period. Using mixed-effects regression models, we calculated yearly rates of initiation and switch for bDMARDs, adjusted for case-mix. We also described the proportion of all initiations and switches accounted for by each agent.
RESULTS: There were 113,031 RA patients with public insurance and 97,751 RA patients with private insurance who were included in the study. The rates of initiation of bDMARDs (per 100 patients) increased significantly over time in Medicaid data for incident RA patients (from 1.1 to 3.1, P = 0.0006) and prevalent RA patients (from 4.6 to 10.9, P = 0.008). In Optum Clinformatics data, the rates were stable, with 7.7 to 8.3 per 100 incident RA patients (P = 0.10) and 11.0 to 11.5 per 100 prevalent RA patients (P = 0.12). The rates of switching (per 100 patients) increased over time from 6.4 to 16.0 (P = 0.04) in Medicaid data and 9.1 to 17.0 (P = 0.00003) in Optum Clinformatics data. Use of etanercept as the most common first-line agent was stable at approximately 50% of all biologic initiations, but use of infliximab decreased and the use of newer agents increased.
CONCLUSIONS: More RA patients used bDMARDs in recent years, and use of newer agents, including certolizumab, golumumab, and tocilizuamab, is rising, which highlights a need for further comparative safety and effectiveness research of these agents to better guide evidence-based decision making.
DISCLOSURES: This study was supported by an investigator-initiated research grant from Pfizer. The study was conducted by the authors independent of the sponsor. The sponsor was given the opportunity to make nonbinding comments on a draft of the manuscript, but the authors retained the right of publication and determined the final wording. Solomon receives salary support through research support to his hospital from Amgen, Pfizer, AstraZeneca, Genentech, Lilly, Bristol-Myers Squibb, and CORRONA. Kim receives research grants from AstraZeneca, Lilly, Pfizer, Bristol-Myers Squibb, and Genentech. Desai receives research grants from Merck. Study concept and design were contributed by Desai and Kim. Liu took the lead in data collection, along with Jin, Desai, and Kim. All authors contributed equally in the interpretation of the results. The manuscript was written and revised primarily by Desai, along with Kim and the other authors.
PURPOSE OF REVIEW: Large healthcare databases, which contain data collected during routinely delivered healthcare to patients, can serve as a valuable resource for generating actionable evidence to assist medical and healthcare policy decision-making. In this review, we summarize use of large healthcare databases in rheumatology clinical research.
RECENT FINDINGS: Large healthcare data are critical to evaluate medication safety and effectiveness in patients with rheumatologic conditions. Three major sources of large healthcare data are: first, electronic medical records, second, health insurance claims, and third, patient registries. Each of these sources offers unique advantages, but also has some inherent limitations. To address some of these limitations and maximize the utility of these data sources for evidence generation, recent efforts have focused on linking different data sources. Innovations such as randomized registry trials, which aim to facilitate design of low-cost randomized controlled trials built on existing infrastructure provided by large healthcare databases, are likely to make clinical research more efficient in coming years.
SUMMARY: Harnessing the power of information contained in large healthcare databases, while paying close attention to their inherent limitations, is critical to generate a rigorous evidence-base for medical decision-making and ultimately enhancing patient care.
AIM: To assess heterogeneity in adherence to medications in two example comparative effectiveness research studies.
PATIENTS & METHODS: We analyzed data from commercially insured patients initiating a statin or anticoagulant during 2005-2012. We calculated the cross-validated R from a series of hierarchical linear models to assess variation in 1-year adherence.
RESULTS: There was less heterogeneity in adherence in the statin cohort compared with the anticoagulant cohort, where patient characteristics explained 7.2% of variation in adherence, and adding therapy and provider characteristics increased the proportion of variation explained to 8.0 and 8.5%, cumulatively. Random effects provided essentially no explanatory power, even in the statin cohort with large numbers of patients clustered within each pharmacy, prescriber and provider.
CONCLUSION: The dependence of adherence on the healthcare system was stronger when the healthcare system influenced treatment choice and patient access to medication and when indications for treatment were strong.