2014/03/07 by Sara Geneletti, Geneletti, Sara, Aidan G. O’Keeffe +7 · 1 citation
Economics, Econometrics and Finance · Mathematics · #Advanced Causal Inference Techniques #FOS: Computer and information sciences #Health Systems, Economic Evaluations, Quality of Life #Methodology (stat.ME) #Statistical Methods in Clinical Trials
paper · pdf · doi:10.48550/arxiv.1403.1806
openalex publication_date 2014/03/07 · openalex created_date 2025/10/24 · openalex updated_date 2026/07/28
The regression discontinuity (RD) design is a quasi-experimental design that\nestimates the causal effects of a treatment by exploiting naturally occurring\ntreatment rules. It can be applied in any context where a particular treatment\nor intervention is administered according to a pre-specified rule linked to a\ncontinuous variable. Such thresholds are common in primary care drug\nprescription where the RD design can be used to estimate the causal effect of\nmedication in the general population. Such results can then be contrasted to\nthose obtained from randomised controlled trials (RCTs) and inform prescription\npolicy and guidelines based on a more realistic and less expensive context. In\nthis paper we focus on statins, a class of cholesterol-lowering drugs, however,\nthe methodology can be applied to many other drugs provided these are\nprescribed in accordance to pre-determined guidelines. NHS guidelines state\nthat statins should be prescribed to patients with 10 year cardiovascular\ndisease risk scores in excess of 20%. If we consider patients whose scores are\nclose to this threshold we find that there is an element of random variation in\nboth the risk score itself and its measurement. We can thus consider the\nthreshold a randomising device assigning the prescription to units just above\nthe threshold and withholds it from those just below. Thus we are effectively\nreplicating the conditions of an RCT in the area around the threshold, removing\nor at least mitigating confounding. We frame the RD design in the language of\nconditional independence which clarifies the assumptions necessary to apply it\nto data, and which makes the links with instrumental variables clear. We also\nhave context specific knowledge about the expected sizes of the effects of\nstatin prescription and are thus able to incorporate this into Bayesian models\nby formulating informative priors on our causal parameters.\n