Preregistration house officers in the Thames regions: changes in quality of training after four years.

OBJECTIVE To measure changes in the training and workload of preregistration house officers over four years. DESIGN Postal questionnaire. SETTING The Thames health regions. PARTICIPANTS 1049 preregistration house officers. RESULTS Response rate was 69% (725 replies). The proportion of house officers officially on duty > 83 hours a week fell from at least 42% to 21%, and the proportion officially on duty < or = 72 hours rose from no more than 9% to 40%. Adequate guidance in breaking bad news increased from 25% to 46% (p < 0.0001; 95% confidence interval for difference, 16.2% to 25.8%) and guidance in pain control increased from 36% to 46% (p < 0.01; 5.0% to 15.0%). The number of house officers attending an induction course increased from 61% to 94% (p < 0.001; 28.9% to 37.1%). There was no change in the proportion unable to attend formal educational sessions because of clinical commitments or in levels of satisfaction with consultants' educational supervision. The median number of inpatients under house officers' care fell from 20 to 17, but the numbers of patients clerked in an average week showed little change. House officers were less satisfied with the clinical experience their post provided (proportion dissatisfied rose from 30% to 39%; p < 0.01; 4.2% to 13.8%) and less enthusiastic about recommending their post to a friend (proportion neutral or not recommending rose from 30% to 42%, p < 0.0001; 7.9% to 16.9%). CONCLUSION Despite progress in reducing hours of duty and providing induction courses, the training that hospitals and consultants provide for house officers is still unsatisfactory and inconsistent with the General Medical Council's recommendations.