Application of Propensity Score – an overview with the help of r software package

Experimental studies or randomized clinical trail in health care setting are usually the preferred type of research when we want to compare to treatments or two groups or test procedures. When the experimental or clinical trials are not feasible or unethical to conduct then observational studies are preferred over the experimental studies to compare the two groups out of which one is received the treatment under consideration and another one is not the received the treatment or placebo group. One of the main drawbacks of the observational studies is that the experimental units are not randomized among the two groups which results in imbalance in the base line covariates which ultimately affect the outcome under consideration which is not the case in experimental or clinical trials. Propensity score[1,2] helps us to overcome the imbalance in the two groups by assigning propensity scores to treatment groups and non-treatment groups and the members of the two groups with equal propensity scores are matched which makes the two groups comparable with respect to the covariates. This paper provides an overview of propensity score, its application and computing the propensity score with the use of R statistical package which is an open source statistical package.