Victims of domestic violence frequently seek medical care in emergency departments (EDs) and health care providers have the opportunity to detect DV and intervene. Unfortunately ED personnel often do not screen for DV due to time pressures and the sensitive nature of the problem. Novel methods are needed to improve detection and intervention in the ED. We developed and piloted a self-administered computer health risk assessment (ED Prevent) that is completed by patients before their ED visit. High-risk behaviors are provided to the ED physician. Our pilot demonstrated significant improvement in the chart documentation of DV by physicians in patients receiving ED Prevent (19 chart notes of DV) compared to controls (1 chart note).This study is a randomized control trial to test the effect of ED Prevent on the communication between physicians/nurses and patients about DV. 800 women patients presenting to two different EDs (a urban ED with largely African American patients and a suburban community ED with mainly Caucasian patients) will be randomized to either ED Prevent or Usual Care. All visits will be audio-taped. Primary outcomes for the study will be patient disclosure and physician/nurse discussion of DV based on the audio-tapes. Secondary outcomes will include patient knowledge of, or contact with, DV services, patient satisfaction with communication with ED staff, and chart documentation of DV. The effect of ED Prevent on rates of disclosure and discussion will be evaluated using logistic regression controlling for hospital site and patient demographics. In addition a qualitative analysis will explore the effect of ED Prevent on DV communication variables [such as the timing and initiation of discussions about DV and ED provider/patient affective responses] The study has the potential to provide a novel way to screen women in the ED relying on a computer instead of a busy physician and potentially shifting the role of the physician to discussion and referral, ultimately improving the likelihood that women will change the abusive situation. The study results may be generalizable to many other EDs with a variety of populations. Further the data will be available for future investigations of communication in the ED.

R01HS011096

2000-09-30

2003-08-31