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

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

Objective: 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.

Design, Setting, And Participants: 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%).

Intervention: 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).

Main Outcome Measures: 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).

Results: 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.

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

Trial Registration: clinicaltrials.gov Identifier: NCT00526994.