Estimating Effects of Nursing Intervention via Propensity Score Analysis

Abstract

Background-Lack of randomization of nursing intervention in outcome effectiveness studies may lead to imbalanced covariates. Consequently, estimation of nursing intervention effect can be biased as in other observational studies. Propensity score analysis is an effective statistical method to reduce such bias and further derive causal effects in observational studies.

The Impact of Nursing Surveillance on Failure to Rescue

“…Propensity scores were calculated based on variables the researcher believed to be confounders, those variables thought to influence the dependent variable (i.e., failure to rescue), and the amount of surveillance received by the patient. Hospitalizations were then matched on those propensity scores, thereby creating two groups that were essentially equivalent in terms of those confounders (Qin et al, 2008). Even though randomization was not possible in this observational study, the treatment effect of surveillance is more clearly understood because of the use of propensity scores, which helped create two equivalent groups that received different "doses" of nursing surveillance.…”

“…Variables included in the main regression included the dichotomous surveillance treatment variable (i.e., high or low surveillance use) and other independent variables related to the dependent variable. As Figure 1 indicates, if a variable was used to calculate the propensity scores, it was not used again in the main regression to avoid correlations between the propensity scores and the variable (Qin et al, 2008). Variables were selected to enter the main regression if they were thought to impact the outcome of failure to rescue.…”

“…Propensity scores are used to reduce the estimation bias of the treatment effect between subjects that received treatment and those subjects that did not (Austin, Grootendorst, & Anderson, 2007;D'Agostino, 1998). The propensity score, calculated for each subject, can be used in one of three ways: to match subjects on, to stratify subjects into like groups, or covarianceadjustment in a regression (D'Agostino, 1998;Qin, Titler, Shever, & Kim, 2008).…”

Improving sleep quality interventions among menopausal women with sleep disturbances in Taiwan: a preliminary study

“…Propensity scores can be used to correct the bias caused by confounding factors in the observed results. By making adjustments using the propensity score, the dependent variable distribution allowed for random distribution, reducing the interference that covariance had on the research results, which established a causal relationship between the variables in the study (Quin, 2008).…”

Neurocognition, Social Cognition and Symptoms in Schizophrenia

“…Propensity score analysis is an effective statistical method to reduce selection bias and further derive causal effects in observational studies. It also provides an alternative approach to the classical multivariate regression, stratification, and matching techniques for examining the effects of nursing interven with a large number of confounding covariates in the background [11]. We also used Spearman's rank correlation to explore the relations among neurocognition, social cognition, and symptoms.…”