Effect of screening for partner violence on women's quality of life: a randomized controlled trial

Although partner violence screening has been endorsed by many health organizations, there is insufficient evidence that it has beneficial health outcomes.

To determine the effect of computerized screening for partner violence plus provision of a partner violence resource list vs provision of a partner violence list only on women's health in primary care settings, compared with a control group.

A 3-group blinded randomized controlled trial at 10 primary health care centers in Cook County, Illinois. Participants were enrolled from May 2009-April 2010 and reinterviewed 1 year (range, 48-56 weeks) later. Participants were English- or Spanish-speaking women meeting specific inclusion criteria and seeking clinical services at study sites. Of 3537 women approached, 2727 were eligible, 2708 were randomized (99%), and 2364 (87%) were recontacted 1 year later. Mean age of participants was 39 years. Participants were predominantly non-Latina African American (55%) or Latina (37%), had a high school education or less (57%), and were uninsured (57%).

Randomization into 3 intervention groups: (1) partner violence screen (using the Partner Violence Screen instrument) plus a list of local partner violence resources if screening was positive (n = 909); (2) partner violence resource list only without screen (n = 893); and (3) no-screen, no-partner violence list control group (n=898).

Quality of life (QOL, physical and mental health components) was the primary outcome, measured on the 12-item Short Form (scale range 0-100, mean of 50 for US population).

At 1-year follow-up, there were no significant differences in the QOL physical health component between the screen plus partner violence resource list group (n = 801; mean score, 46.8; 95% CI, 46.1-47.4), the partner violence resource list only group (n = 772; mean score, 46.4; 95% CI, 45.8-47.1), and the control group (n = 791; mean score, 47.2; 95% CI, 46.5-47.8), or in the mental health component (screen plus partner violence resource list group [mean score, 48.3; 95% CI, 47.5-49.1], the partner violence resource list only group [mean score, 48.0; 95% CI, 47.2-48.9], and the control group [mean score, 47.8; 95% CI, 47.0-48.6]). There were also no differences between groups in days unable to work or complete housework; number of hospitalizations, emergency department, or ambulatory care visits; proportion who contacted a partner violence agency; or recurrence of partner violence.

Among women receiving care in primary care clinics, providing a partner violence resource list with or without screening did not result in improved health.

clinicaltrials.gov Identifier: NCT00526994.