Confounding by indication poses a significant threat to the validity of nonexperimental studies assessing effectiveness and safety of medical interventions. While no different from other forms of ...confounding in theory, confounding by indication often requires specific methods to address the bias it creates in addition to common epidemiological adjustment or restriction methods. Clinical indication influencing treatment prescription is patient‐specific and complex, making it challenging to measure within nonexperimental research. Restriction of the study population to patients with the indication for treatment would effectively mitigate confounding by indication and bring about comparability between exposure and comparator populations with respect to probability of the exposure. Active comparators are often an effective practical solution to restrict the study population in this manner when indication cannot be measured accurately. This article discusses various forms of confounding by indication, the utility of active comparators for nonexperimental studies of treatment effects, and the active comparator, new user (ACNU) study design to implicitly condition on indication. Considerations for selecting active comparators and conducting an ACNU study design are discussed to enable increased adoption of these methods, improve quality of nonexperimental studies, and ultimately strengthen our evidence base for intended and unintended treatment effects in relevant target populations.
Much has been written about real-world evidence (RWE), a concept that offers an understanding of the effects of healthcare interventions using routine clinical data. The reflection of diverse ...real-world practices is a double-edged sword that makes RWE attractive but also opens doors to several biases that need to be minimised both in the design and analytical phases of non-experimental studies. Additionally, it is critical to ensure that researchers who conduct these studies possess adequate methodological expertise and ability to accurately implement these methods. Critical design elements to be considered should include a clearly defined research question using a causal inference framework, choice of a fit-for-purpose data source, inclusion of new users of a treatment with comparators that are as similar as possible to that group, accurately classifying person-time and deciding censoring approaches. Having taken measures to minimise bias ‘by design’, the next step is to implement appropriate analytical techniques (for example propensity scores) to minimise the remnant potential biases. A clear protocol should be provided at the beginning of the study and a report of the results after, including caveats to consider. We also point the readers to readings on some novel analytical methods as well as newer areas of application of RWE. While there is no one-size-fits-all solution to evaluating RWE studies, we have focused our discussion on key methods and issues commonly encountered in comparative observational cohort studies with the hope that readers are better equipped to evaluate non-experimental studies that they encounter in the future.
Graphical abstract
To examine whether dipeptidyl peptidase 4 inhibitors (DPP-4I) increase acute pancreatitis risk in older patients and whether the association varies by age, sex, and history of cardiovascular disease ...(CVD).
We conducted a cohort study of DPP-4I initiators versus thiazolidinedione (TZD) or sulfonylurea initiators using U.S. Medicare beneficiaries, 2007-2014. Eligible initiators were aged 66 years or older without history of pancreatic disease or alcohol-related diseases. Patients were followed up for hospitalization due to acute pancreatitis and censored at 90 days after treatment changes. Weighted Cox models were used to estimate the hazard ratio (HR) for acute pancreatitis. Analyses were performed overall as well as within subgroups defined by age, sex, and CVD history.
We found no increased risk of acute pancreatitis comparing 49,374 DPP-4I initiators to 132,223 sulfonylurea initiators (weighted HR 1.01; 95% CI 0.83-1.24) and comparing 57,301 DPP-4I initiators to 32,612 TZD initiators (weighted HR 1.11; 95% CI 0.76-1.62). Age and sex did not modify the association. Among patients with CVD, acute pancreatitis incidence was elevated in initiators of DPP-4I and sulfonylurea (2.3 and 2.4 per 1,000 person-years, respectively) but not in TZD initiators (1.5). Among patients with CVD, higher risk of acute pancreatitis was observed with DPP-4I compared with TZD (weighted HR 1.84; 95% CI 1.02-3.35) but not compared with sulfonylurea.
Our study provides evidence that DPP-4I is not associated with an increased risk of acute pancreatitis in older adults overall. The positive association observed in patients with CVD could be due to chance or bias but merits further investigation.
Variable Selection for Propensity Score Models Brookhart, M. Alan; Schneeweiss, Sebastian; Rothman, Kenneth J. ...
American journal of epidemiology,
06/2006, Letnik:
163, Številka:
12
Journal Article
Recenzirano
Odprti dostop
Despite the growing popularity of propensity score (PS) methods in epidemiology, relatively little has been written in the epidemiologic literature about the problem of variable selection for PS ...models. The authors present the results of two simulation studies designed to help epidemiologists gain insight into the variable selection problem in a PS analysis. The simulation studies illustrate how the choice of variables that are included in a PS model can affect the bias, variance, and mean squared error of an estimated exposure effect. The results suggest that variables that are unrelated to the exposure but related to the outcome should always be included in a PS model. The inclusion of these variables will decrease the variance of an estimated exposure effect without increasing bias. In contrast, including variables that are related to the exposure but not to the outcome will increase the variance of the estimated exposure effect without decreasing bias. In very small studies, the inclusion of variables that are strongly related to the exposure but only weakly related to the outcome can be detrimental to an estimate in a mean squared error sense. The addition of these variables removes only a small amount of bias but can increase the variance of the estimated exposure effect. These simulation studies and other analytical results suggest that standard model-building tools designed to create good predictive models of the exposure will not always lead to optimal PS models, particularly in small studies.
IMPORTANCE: Postoperative delirium is associated with decreases in long-term cognitive function in elderly populations. OBJECTIVE: To determine whether postoperative delirium is associated with ...decreased long-term cognition in a younger, more heterogeneous population. DESIGN, SETTING, AND PARTICIPANTS: A prospective cohort study was conducted at a single academic medical center (≥800 beds) in the southeastern United States from September 5, 2017, through January 15, 2018. A total of 191 patients aged 18 years or older who were English-speaking and were anticipated to require at least 1 night of hospital admission after a scheduled major nonemergent surgery were included. Prisoners, individuals without baseline cognitive assessments, and those who could not provide informed consent were excluded. Ninety-day follow-up assessments were performed on 135 patients (70.7%). EXPOSURES: The primary exposure was postoperative delirium defined as any instance of delirium occurring 24 to 72 hours after an operation. Delirium was diagnosed by the research team using the Confusion Assessment Method (CAM). MAIN OUTCOMES AND MEASURES: The primary outcome was change in cognition at 90 days after surgery compared with baseline, preoperative cognition. Cognition was measured using a telephone version of the Montreal Cognitive Assessment (T-MoCA) with cognitive impairment defined as a score less than 18 on a scale of 0 to 22. RESULTS: Of the 191 patients included in the study, 110 (57.6%) were women; the mean (SD) age was 56.8 (16.7) years. For the primary outcome of interest, patients with and without delirium had a small increase in T-MoCA scores at 90 days compared with baseline on unadjusted analysis (with delirium, 0.69; 95% CI, −0.34 to 1.73 vs without delirium, 0.67; 95% CI, 0.17-1.16). The initial multivariate linear regression model included age, preoperative American Society of Anesthesiologists Physical Status Classification System score, preoperative cognitive impairment, and duration of anesthesia. Preoperative cognitive impairment proved to be the only notable confounder: when adjusted for preoperative cognitive impairment, patients with delirium had a 0.70-point greater decrease in 90-day T-MoCA scores than those without delirium compared with their respective baseline scores (with delirium, 0.16; 95% CI, −0.63 to 0.94 vs without delirium, 0.86; 95% CI, 0.40-1.33). CONCLUSIONS AND RELEVANCE: Although a statistically significant association between 90-day cognition and postoperative delirium was not noted, patients with preoperative cognitive impairment appeared to have improvements in cognition 90 days after surgery; however, this finding was attenuated if they became delirious. Preoperative cognitive impairment alone should not preclude patients from undergoing indicated surgical procedures.
In this MiniReview, we provide general considerations for the planning and conduct of pharmacoepidemiological studies of associations between drug use and cancer development. We address data sources, ...study design, assessment of drug exposure, ascertainment of cancer outcomes, confounder adjustment and future perspectives. Aspects of data sources include assessment of complete history of drug use and data on dose and duration of drug use, allowing estimates of cumulative exposure. Outcome data from formal cancer registries are preferable, but cancer data from other sources, for example, patient or pathology registries, medical records or claims are also suitable. The two principal designs for observational studies evaluating drug–cancer associations are the cohort and case–control designs. A key challenge in studies of drug–cancer associations is the exposure assessment due to the typically long period of cancer development. We present methods to examine early and late effects of drug use on cancer development and discuss the need for employing ‘lag‐time’ in order to avoid reverse causation. We emphasize that a new‐user study design should always be considered. We also underline the need for ‘dose–response’ analyses, as drug–cancer associations are likely to be dose‐dependent. Generally, studies of drug–cancer associations should explore risk of site‐specific cancer, rather than cancer overall. Additional differentiation may also be crucial for organ‐specific cancer with various distinct histological subtypes (e.g., lung or ovary cancer). We also highlight the influence of confounding factors and discuss various methods to address confounding, while emphasizing that the choices of methods depend on the design and specific objectives of the individual study. In some studies, use of active comparator(s) may be preferable. Pharmacoepidemiological studies of drug–cancer associations are expected to evolve considerably in the coming years, due to the increasing availability of long‐term data on drug exposures and cancer outcomes, the increasing conduct of multinational studies, allowing studies of rare cancers and subtypes of cancer, and methodological improvements specifically addressing cancer and other long‐term outcomes.
Confounding can cause substantial bias in nonexperimental studies that aim to estimate causal effects. Propensity score methods allow researchers to reduce bias from measured confounding by ...summarizing the distributions of many measured confounders in a single score based on the probability of receiving treatment. This score can then be used to mitigate imbalances in the distributions of these measured confounders between those who received the treatment of interest and those in the comparator population, resulting in less biased treatment effect estimates. This methodology was formalized by Rosenbaum and Rubin in 1983 and, since then, has been used increasingly often across a wide variety of scientific disciplines. In this review article, we provide an overview of propensity scores in the context of real‐world evidence generation with a focus on their use in the setting of single treatment decisions, that is, choosing between two therapeutic options. We describe five aspects of propensity score analysis: alignment with the potential outcomes framework, implications for study design, estimation procedures, implementation options, and reporting. We add context to these concepts by highlighting how the types of comparator used, the implementation method, and balance assessment techniques have changed over time. Finally, we discuss evolving applications of propensity scores.
Abstract
To extend previous simulations on the performance of propensity score (PS) weighting and trimming methods to settings without and with unmeasured confounding, Poisson outcomes, and various ...strengths of treatment prediction (PS c statistic), we simulated studies with a binary intended treatment T as a function of 4 measured covariates. We mimicked treatment withheld and last-resort treatment by adding 2 “unmeasured” dichotomous factors that directed treatment to change for some patients in both tails of the PS distribution. The number of outcomes Y was simulated as a Poisson function of T and confounders. We estimated the PS as a function of measured covariates and trimmed the tails of the PS distribution using 3 strategies (“Crump,” “Stürmer,” and “Walker”). After trimming and reestimation, we used alternative PS weights to estimate the treatment effect (rate ratio): inverse probability of treatment weighting, standardized mortality ratio (SMR)-treated, SMR-untreated, the average treatment effect in the overlap population (ATO), matching, and entropy. With no unmeasured confounding, the ATO (123%) and “Crump” trimming (112%) improved relative efficiency compared with untrimmed inverse probability of treatment weighting. With unmeasured confounding, untrimmed estimates were biased irrespective of weighting method, and only Stürmer and Walker trimming consistently reduced bias. In settings where unmeasured confounding (e.g., frailty) may lead physicians to withhold treatment, Stürmer and Walker trimming should be considered before primary analysis.