gms | German Medical Science

Introduction: Since October 2005 the German Medical Faculties select 50% of their students in an in-house selection process, the remaining university places being centrally allocated mostly on the basis of marks in the university entrance diploma and the time on a waiting list. According to federal and land law regulations the in-house selection criteria must also grant a leading role of the marks in the university entrance diploma supplemented by a set of additional criteria. The present study analyses the predictive value of school leaving examination grades in two student cohorts starting undergraduate medical training at Heidelberg Medical School in 2005 and 2006 with respect to the endpoints “performance in medical examinations”, “delay of study progress” and “drop out rate”. Furthermore, the predictive value of the university entrance diploma is compared with the criterion “time on a waiting list” and the in-house criteria for admission with respect to the same endpoints. In-house admission criteria at Heidelberg Medical School are prices in educational contests, previous training in paramedical professions and voluntary social work.

Methods: During the application process the overall grade in the university entrance diploma, subject specific recent school marks, data on success in educational competitions, previous paramedical employment and engagement in voluntary social work are documented. During training the performance in medical examinations of our students are documented and compared to their individual application data.

Results: In the first year of undergraduate training the correlation coefficient between the performance in medical education and the overall marks in the university entrance diploma and the grades in mathematics reached 0.51 and 0.55, respectively, but decreased to 0.15 and 0.08 during clinical training. Already in the first year of training the variability of performance in medical examinations was too high across all levels of results in the final school examination and individual subjects for those being an adequate criterion for individual selection decisions. Furthermore, the drop out rate was similar among all students with university entrance diploma grades between 1.0 and 2.7. The performance in medical examinations during the first two years of those students being selected from a waiting list were lower on average than that of the best school alumni, but the difference disappeared during clinical training. The prognostic value of all other criteria analysed was weak already during the first two years of preclinical education.

Conclusions: Grades in university entrance diplomas and specific school subjects bear a low predictive value for study success in the first two years of medical education and none at all for the clinical training. To date, other reliable admission criteria are missing. New criteria have to be sought and must come under scrutiny.