M. Alan Brookhart

Healthy user and related biases in observational studies of preventive interventions: a primer for physicians.

The current emphasis on comparative effectiveness research will provide practicing physicians with increasing volumes of observational evidence about preventive care. However, numerous highly publicized observational studies of the effect of prevention on health outcomes have reported exaggerated relationships that were later contradicted by randomized controlled trials. A growing body of research has identified sources of bias in observational studies that are related to patient behaviors or underlying patient characteristics, known as the healthy user effect, the healthy adherer effect, confounding by functional status or cognitive impairment, and confounding by selective prescribing. In this manuscript we briefly review observational studies of prevention that have appeared to reach incorrect conclusions. We then describe potential sources of bias in these studies and discuss study designs, analytical methods, and sensitivity analyses that may mitigate bias or increase confidence in the results reported. More careful consideration of these sources of bias and study designs by providers can enhance evidence-based decision-making.

Instrumental variable methods in comparative safety and effectiveness research

Instrumental variable (IV) methods have been proposed as a potential approach to the common problem of uncontrolled confounding in comparative studies of medical interventions, but IV methods are unfamiliar to many researchers. The goal of this article is to provide a non‐technical, practical introduction to IV methods for comparative safety and effectiveness research. We outline the principles and basic assumptions necessary for valid IV estimation, discuss how to interpret the results of an IV study, provide a review of instruments that have been used in comparative effectiveness research, and suggest some minimal reporting standards for an IV analysis. Finally, we offer our perspective of the role of IV estimation vis‐à‐vis more traditional approaches based on statistical modeling of the exposure or outcome. We anticipate that IV methods will be often underpowered for drug safety studies of very rare outcomes, but may be potentially useful in studies of intended effects where uncontrolled confounding may be substantial. Copyright

Propensity Score Methods for Confounding Control in Nonexperimental Research

Nonexperimental studies are increasingly used to investigate the safety and effectiveness of medical products as they are used in routine care. One of the primary challenges of such studies is confounding, systematic differences in prognosis between patients exposed to an intervention of interest and the selected comparator group. In the presence of uncontrolled confounding, any observed difference in outcome risk between the groups cannot be attributed solely to a causal effect of the exposure on the outcome. Confounding in studies of medical products can arise from a variety of different sociomedical processes.1 The most common form of confounding arises from good medical practice, physicians prescribing medications and performing procedures on patients who are most likely to benefit from them. This leads to a bias known as confounding by indication, which can cause medical interventions to appear to cause events that they prevent.2,3 Conversely, patients who are perceived by a physician to be near the end of life may be less likely to receive preventive medications, leading to confounding by frailty or comorbidity.4–6 Additional sources of confounding bias can result from patients’ health-related behaviors. For example, patients who initiate a preventive medication may be more likely than other patients to engage in other healthy, prevention-oriented behaviors leading to bias known as the healthy user/adherer effect.7–9 Many statistical approaches can be used to remove the confounding effects of such factors if they are captured in the data. The most common statistical approaches for confounding control are based on multivariable regression models of the outcome. To yield unbiased estimates of treatment effects, these approaches require that the researcher correctly models the effect of the treatment and covariates on the outcome. However, correct specification of an outcome model can be challenging, particularly in studies …

Seizure Outcomes Following the Use of Generic versus Brand-Name Antiepileptic Drugs A Systematic Review and Meta-Analysis

AbstractBackground: The automatic substitution of bioequivalent generics for brand-name antiepileptic drugs (AEDs) has been linked by anecdotal reports to loss of seizure control. Objective: To evaluate studies comparing brand-name and generic AEDs, and determine whether evidence exists of superiority of the brand-name version in maintaining seizure control. Data Sources: English-language human studies identified in searches of MEDLINE, EMBASE and International Pharmaceutical Abstracts (1984 to 2009). Study Selection: Randomized controlled trials (RCTs) and observational studies comparing seizure events or seizure-related outcomes between one brand-name AED and at least one alternative version produced by a distinct manufacturer. Data Extraction: We identified 16 articles (9 RCTs, 1 prospective non-randomized trial, 6 observational studies). We assessed characteristics of the studies and, for RCTs, extracted counts for patients whose seizures were characterized as ‘controlled’ and ‘uncontrolled’. Data Synthesis: Seven RCTs were included in the meta-analysis. The aggregate odds ratio (n = 204) was 1.1 (95% CI 0.9, 1.2), indicating no difference in the odds of uncontrolled seizure for patients on generic medications compared with patients on brand-name medications. In contrast, the observational studies identified trends in drug or health services utilization that the authors attributed to changes in seizure control. Conclusions: Although most RCTs were short-term evaluations, the available evidence does not suggest an association between loss of seizure control and generic substitution of at least three types of AEDs. The observational study data may be explained by factors such as undue concern from patients or physicians about the effectiveness of generic AEDs after a recent switch. In the absence of better data, physicians may want to consider more intensive monitoring of high-risk patients taking AEDs when any switch occurs.

Trends in the Incidence, Demographics, and Outcomes of End-Stage Renal Disease Due to Lupus Nephritis in the US From 1995 to 2006

OBJECTIVE This study was undertaken to investigate whether recent advances in lupus nephritis treatment have led to changes in the incidence of end-stage renal disease (ESRD) secondary to lupus nephritis, or in the characteristics, treatments, and outcomes of patients with lupus nephritis ESRD. METHODS Patients with incident lupus nephritis ESRD (1995-2006) were identified in the US Renal Data System. Trends in sociodemographic and clinical characteristics were assessed. We tested for temporal changes in standardized incidence rates (SIRs) for sociodemographic groups using Poisson regression. Changes in rates of waitlisting for kidney transplant, kidney transplantation, and all-cause mortality were examined using crude and adjusted time-to-event analyses. RESULTS We identified 12,344 incident cases of lupus nephritis ESRD. Mean age at ESRD onset was 41 years; 81.6% of the patients were women and 49.5% were African American. SIRs for lupus nephritis ESRD among those who were ages 5-39 years, African American, or lived in the southeastern US increased significantly from 1995 to 2006. Increases in body mass index and in the prevalence of both diabetes mellitus and hypertension were detected. Mean serum hemoglobin level at ESRD onset increased, while that of serum creatinine decreased over time. More patients received hemodialysis and fewer received peritoneal dialysis. There was a slight increase in the frequency of preemptive kidney transplantation at ESRD onset, but kidney transplantation rates within the first 3 years of ESRD declined. Mortality did not change over the 12 years of study. CONCLUSION Our findings indicate that the characteristics of patients with lupus nephritis ESRD and initial therapies have changed in recent years. While SIRs rose in younger patients, among African Americans, and in the South, outcomes did not improve in over a decade of evaluation.

Infection Risk with Bolus versus Maintenance Iron Supplementation in Hemodialysis Patients

Intravenous iron may promote bacterial growth and impair host defense, but the risk of infection associated with iron supplementation is not well defined. We conducted a retrospective cohort study of hemodialysis patients to compare the safety of bolus dosing, which provides a large amount of iron over a short period of time on an as-needed basis, with maintenance dosing, which provides smaller amounts of iron on a regular schedule to maintain iron repletion. Using clinical data from 117,050 patients of a large US dialysis provider merged with data from Medicares ESRD program, we estimated the effects of iron dosing patterns during repeated 1-month exposure periods on risks of mortality and infection-related hospitalizations during the subsequent 3 months. Of 776,203 exposure/follow-up pairs, 13% involved bolus dosing, 49% involved maintenance dosing, and 38% did not include exposure to iron. Multivariable additive risk models found that patients receiving bolus versus maintenance iron were at increased risk of infection-related hospitalization (risk difference [RD], 25 additional events/1000 patient-years; 95% confidence interval [CI], 16 to 33) during follow-up. Risks were largest among patients with a catheter (RD, 73 events/1000 patient-years; 95% CI, 48 to 99) and a recent infection (RD, 57 events/1000 patient-years; 95% CI, 19 to 99). We also observed an association between bolus dosing and infection-related mortality. Compared with no iron, maintenance dosing did not associate with increased risks for adverse outcomes. These results suggest that maintenance iron supplementation may result in fewer infections than bolus dosing, particularly among patients with a catheter.

Trends in the Use and Outcomes of Implantable Cardioverter-Defibrillators in Patients Undergoing Dialysis in the United States

BACKGROUND Sudden cardiac death constitutes the leading cause of death in patients receiving dialysis. Little is known about the trends in implantable cardioverter-defibrillator (ICD) use and the outcomes of such device placement. STUDY DESIGN Retrospective cohort study. SETTING & PARTICIPANTS US long-term dialysis patients who received an ICD in 1994-2006. PREDICTORS, OUTCOMES, & MEASUREMENTS: ICD utilization rates and incident rates of all-cause mortality, device infections, and other device-related procedures were measured. We compared mortality between recipients and otherwise similar patients who did not receive such a device using high-dimensional propensity score matching. We also examined the associations of demographics, dialysis type, baseline comorbid conditions, cardiovascular events at the time of admission, and recent infection with the study outcomes. RESULTS 9,528 patients received an ICD in 1994-2006, with >88% placed after 2000. Almost all ICD use in the 1990s was for secondary prevention, however, half the patients received ICDs for apparent primary prevention in 2006. Mortality rates after implantation were high (448 deaths/1,000 patient-years) and most deaths were cardiovascular. Postimplantation infection rates were high, especially in the first year after implantation (988 events/1,000 patient-years) and were predicted by diabetes and recent infection. Patients receiving ICDs for secondary prevention had an overall 14% (95% CI, 9%-19%) lower mortality risk compared with propensity-matched controls, but these benefits seemed to be restricted to the early postimplantation time. LIMITATIONS Lack of clinical data, especially for laboratory and heart function studies. Residual confounding by indication. CONCLUSIONS ICD use in dialysis patients is increasing, but rates of all-cause and cardiovascular mortality remain high in dialysis patients receiving these devices. Device infections are common, particularly in patients with recent infections. Randomized trials of ICDs are needed to determine the efficacy, safety, and risk-benefit ratio of these devices in dialysis patients.

Effectiveness and Safety of Dabigatran and Warfarin in Real‐World US Patients With Non‐Valvular Atrial Fibrillation: A Retrospective Cohort Study

Background The recent availability of dabigatran, a novel oral anticoagulant, provided a new treatment option for stroke prevention in atrial fibrillation beyond warfarin, the main therapy for years. Little is known about their real‐world comparative effectiveness and safety, even less among patient demographic and clinical subgroups. Methods and Results Using a cohort of non‐valvular AF patients initiating anticoagulation from October 2010 to December 2012 drawn from a large US database of commercial and Medicare supplement claims, we applied propensity score weights to Cox proportional hazards regression to assess the comparative effectiveness and safety of dabigatran versus warfarin. Analyses were repeated among clinical and demographic subgroups using stratum‐specific propensity scores as an exploratory analysis. Of the 64 935 patients initiating anticoagulation, 32.5% used dabigatran. Compared with warfarin, dabigatran was associated with a lower risk of ischemic stroke or systemic embolism (composite adjusted Hazard Ratio [aHR], 95% CI: 0.86, 95% CI: 0.79 to 0.93), hemorrhagic stroke (aHR: 0.51, 0.40 to 0.65), and acute myocardial infarction (aHR: 0.88, 95% CI: 0.77 to 0.99), and no relation was seen between dabigatran and the composite harm outcome (aHR: 0.94, 95% CI: 0.87 to 1.01). However, dabigatran was associated with a higher risk of gastrointestinal bleeding (aHR: 1.11, 95% CI: 1.02 to 1.22). Estimates of effectiveness and safety appeared to be mostly similar across subgroups. Conclusions Dabigatran could be a safe and potentially more effective alternative to warfarin in patients with atrial fibrillation managed in routine practice settings.

The role of the c-statistic in variable selection for propensity score models

The applied literature on propensity scores has often cited the c‐statistic as a measure of the ability of the propensity score to control confounding. However, a high c‐statistic in the propensity model is neither necessary nor sufficient for control of confounding. Moreover, use of the c‐statistic as a guide in constructing propensity scores may result in less overlap in propensity scores between treated and untreated subjects; this may require the analyst to restrict populations for inference. Such restrictions may reduce precision of estimates and change the population to which the estimate applies. Variable selection based on prior subject matter knowledge, empirical observation, and sensitivity analysis is preferable and avoids many of these problems. Copyright

The implications of propensity score variable selection strategies in pharmacoepidemiology – an empirical illustration

To examine the effect of variable selection strategies on the performance of propensity score (PS) methods in a study of statin initiation, mortality, and hip fracture assuming a true mortality reduction of < 15% and no effect on hip fracture.