Table 2. Results of six studies evaluating effectiveness of problem based learning in continuing medical education Study Participants Study design Educational intervention No of participants Outcome variables (PBL v control) PBL Control PBL Control Participants' knowledge Participants' performance Participants' satisfaction Patients' health Follow up Chan et al 199915 General practitioners RCT Email PBL Internet resources  8 11 No effect NA NA NA NA Heale et al 198816 General practitioners RCT Small group PBL Large group discussion, didactic lecture  22*  39* No effect No effect Higher NA No effect Doucet et al 199817 General practitioners CT PBL Lecture format 34 29 Positive Positive Higher NA NA Benjamin et al 199918 Staff of outpatient clinic CT PBL None ~50§ ~50§ NA NA NA Positive Positive¶ Moran et al 199619 General practitioners CT Learner centred small group None  5 10 NA Positive** NA NA Positive¶ ** Premi et al 199420 General practitioners CT Practice based small group None 78 46 Positive** Positive High score NA NA RCT=Randomised controlled trial. CT=Controlled trial. PBL=Problem based learning. NA=Not assessed. * No information on number who completed knowledge test at end of trial. P value not given in article. Only participants' knowledge tested on follow up. § Information from author, not given in article. ¶ Only participants' performance tested on follow up. ** PBL group's post-intervention scores compared with own pre-intervention scores. 75/78 participants reported 127 practice changes as result of PBL. Mean (SD) score of 5.8 (0.9) on 7 point scale for PBL group.