“Between me and the computer”: Increased detection of intimate partner violence using a computer questionnaire

Abstract

Study objective: The emergency department is a problem-focused environment in which routine screening for intimate partner violence (IPV) is difficult. We hypothesized that screening for IPV during computer-based health-risk assessment would be acceptable to patients and improve detection. Methods: We performed a descriptive study of IPV data collected during a controlled trial of computer-based health promotion in an urban hospital ED. Patients received computer-generated health advice, and physicians received patient risk summaries. Outcomes were patient disclosure and physician documentation of IPV and associated risks. Results: Two hundred forty-eight patients (69% female, 90% black, mean age 39 years) participated in a clinical trial of computer-based health promotion in the ED. Of 170 women, 53 (33%) disclosed emotional abuse, and 25 (15%) disclosed physical abuse. Of 78 men, 22 (29%) disclosed emotional abuse, and 5 (6%) disclosed physical abuse. Patients were also willing to self-report a history or concern of hurting someone close to them. This was true for 21 (14%) women and 15 (22%) men. Controlling for demographic factors, disclosures of victimization and perpetration were associated with multiple psychosocial risks. Computer screening resulted in chart documentation in 19 of 83 potential cases of IPV compared with 1 case documented in the group that received usual care. Conclusion: Providing an opportunity for patients to confidentially self-disclose IPV has the potential to supplement current screening efforts and to allow providers to focus on assessment, counseling, and referral for those at risk. However, further measures will be needed to ensure that information gathered through computer screening is adequately addressed during the acute care or follow-up visit. [Ann Emerg Med. 2002;40:476-484.]

Introduction

Intimate partner violence (IPV), defined as a pattern of coercion, physical abuse, or threat of violence in an intimate relationship, remains a major source of morbidity and mortality worldwide.¹ In the United States, the prevalence of physical abuse in a current relationship is approximately 8.4% for cohabiting women aged 18 to 65 years, 92% of whom have never told a health care provider.² For this reason, there are multiple recommendations that all health care providers routinely screen for abuse.3, 4, 5 These recommendations are based on the burden of suffering and evidence that victims of abuse are overrepresented in health care settings.⁶ Although ED studies with dedicated screeners report prevalence rates in the 25% to 35% range,7, 8 detection rates rapidly decrease when screening is left to busy physicians and nurses.9, 10 Most emergency departments have protocols for routine IPV screening, but the current system often fails to identify battered women. This is due, in large part, to provider time constraints and reluctance to initiate discussions about partner violence.¹¹ Primary care settings, with increasingly abbreviated scheduled appointments, face similar obstacles.12, 13 Nonetheless, patients expect physicians to inquire and will usually disclose abuse if directly questioned.¹⁴

Qualitative reports from women who have escaped abuse indicate that even brief discussions with physicians are therapeutic when conversations are conducted in a concerned, nonjudgmental way.¹⁵ Effective communication with a health care provider has been linked to improved health outcomes in a number of areas.16, 17 For domestic violence, physician-nurse communication skills might also be a key predictor of patient outcome. Rodriguez et al¹⁸ found that women who had experienced abuse favored direct inquiry and identified provider reluctance to inquire about abuse as a significant barrier to disclosure. They reported that when a provider both acknowledged the abuse and validated the patient's self-worth, it had a powerful effect on her perception of the situation and, in some cases, was a turning point in the process of extrication from the abusive relationship. However, the provider has to be able to recognize cues to abuse before this communication can take place.

The vast majority of patients experiencing IPV in their lives will present for non-injury-related complaints and will only be identified as victims of abuse through routine screening.19, 20 Although the ED patient population is associated with high prevalence rates of IPV,²¹ the ED setting presents formidable barriers to routine screening.22, 23 Recently, there has been increasing recognition of the value of EDs as potential sites for injury surveillance and violence prevention.24, 25, 26 Practical methods for conducting IPV screening are needed to realize this potential.

Survey research data suggest that interactive computer-based screening can achieve higher rates of disclosure of sensitive issues than personal interviews.²⁷ In the health care setting, self-administered computer surveys have the potential to provide a relatively low-cost and staff-free method to identify serious health risks. Patients find computer-based health-risk appraisal methods acceptable. They might even be more likely to answer sensitive questions truthfully.²⁸ A systematic review of the literature found that computer-based, clinical decision support systems enhance preventive care.²⁹ However, experience is limited regarding integration of computer-based screening technology into clinical practice. The current study is drawn from a larger clinical trial of computer-based screening published in Annals of Emergency Medicine.³⁰ The purpose of this analysis was to evaluate the feasibility and utility of using computers to screen for IPV. We hypothesized that screening for risk of partner abuse during computer-based health-risk assessment would be acceptable to patients and improve detection of IPV.