Abstracts

Screening for partner violence does not improve women’s health

A three-arm trial tested approaches to addressing intimate partner violence against women who seek primary care for clinical complaints. Participants visited one of 10 urban primary care centres in Illinois, USA. Most were black (55%) or Latina (37%), with high school education or less (57%), and uninsured (57%).

In one group, 909 women were screened by means of a three-question, computer-delivered instrument, which was (in the case of a positive test) followed by a brief recording encouraging access to services and an information sheet on local resources available to women in abusive relationships, as well as other health and social resources. In the other groups 893 women received the information sheet without screening; 898 women had neither, and received only the list of general health and social resources available in the area.

One year later none of the measured outcomes differed between the groups. This included the physical and mental component of the quality of life; self-reported number of days unable to work in the past month, either at a job or in a household; use of a partner violence service or other health services, assessed by self-report and electronic health records, respectively; and recurrence of intimate partner violence, established in an interview asking specifically about 18 different forms of violence.

The negative results could be due to inadequate intensity of the interventions, as well as biases in the sample: nearly 1 in 5 of the invited women declined participation, 12% were lost to follow-up, and the study excluded all women who visited the clinic with a partner or older children.

Klevens J, et al. JAMA 2012;308:681-689.