Short Term Predictive Validity of a Clinical Performance Examination for Senior Medical Students

Abstract

Objective: Korean medical licensing exam for medical doctors will newly include SP-based CPX(clinical performance examination) and model-based OSCE(objective structured clinical examination) in 2009. A valid SP(standardized patient)-based exam is required to have high correlation with another future evaluations after the exam. This study was conducted to investigate the validity of exit CPX to predict the competency thereafter.

Method: Both eighty eight senior medical students in 2007 and 94 senior medical students who took the pilot skills assessment in 2008 were study subjects. The CPX scores during the middle of the fourth year of medical school were compared with the results of written preliminary comprehensive clinical exam, pilot clinical skill assessment for medical licensing exam, and written medical licensing examination. Pearson`s correlation coefficient between scores of two different exams was calculated.

Results or Conclusions: CPX scores were moderately correlated with the previous GPA grades(r=0.37, p<0.0001) and OSCE(r=0.49, p<0.0001) taken at the beginning of the 4th year of medical school. CPX scores were moderately correlated with those of the preliminary comprehensive written clinical exam which was run to prepare senior students for the written medical licensing exam(r=0.35, p<0.0001). CPX scores were not correlated with those of the traditional written Korean Medical Licensure Exam (KMLE)(r=0.14, p=0.20). CPX scores were highly correlated with the pilot clinical skill assessment for medical license exam(r=0.62, p<0.0001). CPX scores did not show any significant prediction for the traditional KMLE results compared with highly significant prediction of preliminary comprehensive exam. In conclusion, CPX measure a domain different from the traditional written cognitive test. Newly developed skills test may be correlated with the exit CPX medical school.