INTERVENTION TO IMPROVE PHYSICIAN DOCUMENTATION AND KNOWLEDGE OF CHILD SEXUAL ABUSE - A RANDOMIZED, CONTROLLED TRIAL

Objective. To determine if written feedback improves the chart documen tation and knowledge of physicians doing evaluations for child sexual abuse and to learn what other factors are associated with better docum entation and knowledge. Design. Randomized, controlled trial. Setting. A statewide network of physicians performing child abuse evaluations. Participants. All physicians who performed evaluations for sexual abus e during 1991 to 1992. One hundred forty-seven physicians were randomi zed to control (n = 75) and intervention (n = 72) groups, 122 (83%) re mained at follow-up, and 87 of the 122 (71%) had done evaluations for child sexual abuse. Interventions. Tailored written feedback based on chart reviews and relevant articles were sent to a randomly selected o ne-half of the physicians during a 9-month period. Main Outcome Measur es. The quality of documentation and physician knowledge before and af ter the intervention. Results. Documentation by chart review of up to five randomly chosen records per physician (preintervention, n = 552; postintervention, n = 259) by reviewers blinded to intervention status and physician knowledge was assessed by survey (78% completion). Chan ge in documentation and knowledge for physicians in the intervention g roup was not statistically significant compared with the control group . The risk ratio for a mean overall history rating of excellent/good w as 0.89 (0.63, 1.25) and for a mean overall physical examination ratin g of excellent/good was 1.03 (0.73, 1.45). Both groups improved signif icantly during the time period. The largest improvements in the time p eriod were in documenting the history of where abuse occurred, in the physical examination position, hymenal description, penile findings, a nd knowing that chlamydia infection should be assessed by culture. A s tructured medical record, female physicians, and credits in continuing medical education were associated with better documentation. Conclusi ons. Tailored feedback to the physician with directed educational mate rials did not seem to improve most aspects of documentation and knowle dge of child sexual abuse, although notable improvement was seen durin g the time period studied. This study suggests that chart audits may n ot be the best use of resources for trying to improve physician behavi or; credits in continuing medical education and use of structured reco rds may be more likely to be beneficial.