Abuse Of Pregnant Women -- Can Physicians Make A Difference?

Deborah S. Lyon, M.D. and Marghani M. Reever, LCSW, ACSW
Deborah S. Lyon, M.D. is Associate Professor, Department of Obstetrics and Gynecology
at the University of Florida Health Science Center / Jacksonville.
Marghani M. Reever, LCSW, ACSW, is with the Mental Health Resource Center, Jacksonville.

Introduction

Pregnant women are particularly vulnerable to abuse, and by some estimates suffer partner abuse at rates higher than the background population.¹ Ironically, pregnancy is also a time in most women's lives when they interact most frequently and intensively with the medical profession, and during which they are most likely to modify unhealthy behaviors. Thus physicians may have an optimal opportunity to intervene during this time. In addition, abuse poses particular risks in pregnancy which may be mitigated if the risk is identified and an effective treatment plan initiated. Barriers to effective screening and treatment are profound, but not insurmountable. This article outlines the epidemiology and pathology of the problem, as well as offering options for screening and intervention.

Reports of domestic violence during pregnancy place prevalence at 0.9 -- 20.1 percent.² The extremely large range reflects the difficulty of research in this area; screening tools are not standardized, and various cultural and community factors significantly affect responses. Lower estimates are based on self-reports, and higher ones on in-office personal interviews. Each method is likely to have reporting bias. There is not a well-defined lower limit for violence, so that patient and interviewer interpretation of terms like "coercion" may vastly affect reporting. Review of the available literature, accounting for methodology, suggests a prevalence of 4-8 percent, or 156,000 - 332,000 women annually.³

While traditionally viewed as a problem of low-income, poorly educated women, recent media coverage has highlighted the ubiquitous occurrence of domestic violence across ethnic and socioeconomic boundaries.⁴ There is no satisfactory profile of the at-risk woman which would allow for selective screening, nor is there a single personality type or educational deficit which, if addressed, would solve the problem. Identifying victims is made more difficult by the frequent collusion of the victim with the perpetrator in suppressing necessary information. Only a well-designed universal screening mechanism can identify the at-risk population, and only a comprehensive plan of services and supports can reduce the sequelae of domestic violence.

The impact of domestic violence on women's health is the subject of other material in this journal. This article will focus on perinatal consequences. The literature suggests increased rates of preterm labor and chorioamnionitis,¹ low birthweight⁵, anemia, infections, first- and second-trimester bleeding, and fetal distress.⁶ This literature is somewhat inconsistent,³ but suggestive in the aggregate of both direct and indirect effects on the fetus of abuse to the mother. More consistent are findings of associated tobacco, alcohol and drug use, late prenatal care, poor maternal weight gain, depression, and suicide attempts.^5,6 These are confounders which might explain the variable results of attempts to evaluate the impact of domestic violence on perinatal outcome. It is not clear whether there is a cause-and-effect relationship between the abuse and the other destructive behaviors, versus each being symptomatic of an underlying common trigger. Nevertheless, whether viewed as a separate risk factor or simply a marker, presence of a domestic violence history may identify patients at increased likelihood of untoward perinatal outcomes.

Several tools have been developed to help identify at-risk patients. The simplest of these is a 3-question screen (see Table 1)⁷. Several more detailed surveys have been used, and some authors recommend more open-ended questions such as "How do you and your partner resolve disagreements?" The literature shows lower rates of detection, however, when the tool used is a self-administered questionnaire compared to one administered by a trained clinician in a direct interview.^2,4,7 Since there is no "gold standard" for identifying the true prevalence of domestic violence in a given population, it is difficult to assess the sensitivity of any given screening tool. Nonetheless, the literature is virtually unanimous in declaring virtually any formal, universally applied screening mechanism to be more effective than risk-based or incident-based questioning.

Contrary to modern dogma that "they don't want help," several interventions have been undertaken for victims of domestic violence with proven success.^8,9 Unfortunately, these have been quite time-intensive, and the published work comes from large indigent care clinics. There is no available literature on interventions which might be practical in the time-constrained, low volume private office setting. There is suggestion, however, that even the initial screening, along with a pocket reference card containing a list of emergency numbers, might reduce abuse prevalence.

Many clinicians fail to apply universal domestic violence screening based on one of three major barriers. The first is time-intensiveness. Because of the above-cited need for direct patient interview, universal screening would add time to the patient encounter. Positive answers require further time. Many practices are designed in such a way that it is difficult to incorporate this step into the flow of patient care. The second major barrier is the perception that intervention is unlikely to change patient behavior. The highly publicized tendency of abused women to return to the environment of abuse discourages many clinicians from investing what they perceive as wasted time. Screening for diabetes makes sense to us because we clearly understand what to do with a positive result and how that will help the patient; screening for abuse does not have the same immediate rewards. The third, and closely related, barrier to universal screening is clinician lack of understanding regarding what resources are available for intervention and how to best apply these resources for any individual patient. Few private clinicians have the resources or inclination to launch the kind of intensive education and intervention programs described in the literature, and no practical alternative has been proposed. Together, these three barriers are quite formidable. Lack of resources, lack of proven interventions, and lack of faith that intervention will be of benefit make for a strong demotivation to investigate for abuse.

Despite these barriers, there is reason to believe that universal screening for abuse in the pregnant population is worthwhile.⁷ First, there is strong support for the notion that pregnancy itself is an ideal time to address destructive behaviors. Literature on smoking cessation shows dramatic successes during pregnancy, albeit with some postpartum recidivism. Other issues, such as use of vehicle seat belts, alcohol consumption, and illicit drug use, have been shown to have somewhat improved intervention success in pregnant compared to non-pregnant patients. Abuse is perhaps not completely comparable to these behaviors, and indeed may more difficult to deal with in light of the greater need for financial and emotional support experienced by most pregnant women. Nevertheless, the models mentioned above give us some reason to hope for greater intervention successes during pregnancy.

Secondly, women typically spend more time engaged with the medical profession during pregnancy than at any other time in their lives. This allows multiple opportunities for identification, education, and intervention. It is reasonable to believe that even low-level interventions such as a poster advertising a women's shelter might be of more benefit given frequent exposure.

The third reason for optimism rests in the wider availability of resources offered to pregnant as opposed to non-pregnant patients. Virtually all pregnant women are eligible for Medicaid, and many qualify for other programs as well. This may allow a greater degree of financial freedom as well as more opportunities for counseling, legal assistance, and public housing priority.

Given the constraints on the time and talents of most medical professionals, the kind of intensive evaluation and therapy described in the literature is unlikely to become standard practice. Although perhaps less effective, the modest approach we advocate is practical and likely to be an improvement over no screening at all.

First, we suggest an assessment tool such as that in Table 1 be administered to all pregnant patients. If the practice has a nurse take the initial OB history, he or she can incorporate this into that history, in the same way we have now incorporated questions regarding HIV risk. Ideally, this should be done with the partner absent. If the partner is present for the history, the nurse might insist that office policy dictates privacy for the examination, and then perform the abuse assessment in the exam room. If there is no person-to-person process in place in the practice (for instance, if patients fill out their own histories prior to the initial exam) then the questionnaire may be self-administered.

Second, for all positive responders, some effort must be made to determine immediate safety. All clinicians should have telephone numbers of emergency shelters and be able to advise patients of these options.

Third, for all patients, whether or not they responded positively to the abuse assessment, a pocket reference card should be given with the numbers of medical, legal, shelter, counseling and other resources in the local area. For clinicians in the Jacksonville area this card may be generated by reproducing Tables 2 and 3. This card should be given along with any other printed materials and samples given to the patient at her first obstetrical visit. The person who normally gives these materials to patients should point this card out and explain that it is given to all pregnant patients. By identifying the resource card, patients are given the opportunity to hide or remove it if they feel its discovery by the abuser might have violent repercussions. On the other hand, if it is presented as simply a routine part of the prenatal education package, its presence is likely to arouse less notice than if only selected patients are provided the card.

Finally, clinicians should remember the associated problems of tobacco, drug, and alcohol use, missed appointments, depression and suicide attempts. These problems should be addressed individually and regardless of whether the patient accepts abuse intervention. In addition, special attention should be given to adequate fetal growth and maternal nutrition, and any complaints of contractions, bleeding, vaginal discharge or fever should be thoroughly evaluated. These patients should be considered at high risk of untoward pregnancy outcome, and may merit antepartum fetal testing.

Pregnancy represents a time of particular risk and vulnerability for the abused woman. It may also represent an ideal opportunity for successful intervention. Although optimal diagnosis and intervention may not be achievable in every practice, even small efforts at screening and education can be surprisingly effective in improving patient safety. Given the high prevalence of abuse compared to many other conditions for which we routinely screen, and its attendant obstetrical hazards, universal screening is the only rational option for lowering the morbidity associated with domestic violence.

1. Berenson AB, Wiemann CM, Wilkinson GS, et al. Perinatal morbidity associated with violence experienced by pregnant women. Am J Obstet Gynecol. 1994;170:1760-9.
2. Gazmararian JA, Lazorick S, Spitz AM, et al. Prevalence of violence against pregnant women. JAMA. 1996;275:1915-20.
3. Petersen R, Gazmararian JA, Spitz AM, et al. Violence and adverse pregnancy outcomes: A review of the literature and directions for future research. Am J Prev Med. 1997;13:366-73.
4. Poirier L. The importance of screening for domestic violence in all women. The Nurse Practitioner. 1997;22:105-22.
5. Curry MA, Perrin N, Wall E. Effects of abuse on maternal complications and birth weight in adult and adolescent women. Obstet Gynecol. 1998;92:530-4.
6. McFarlane J, Wiist W, Watson M. Predicting physical abuse against pregnant hispanic women. Am J Prev Med. 1998;15:134-8.
7. Mayer L, Liebschutz J. Domestic violence in the pregnant patient: Obstetric and behavioral interventions. Obstet Gynecol Survey. 1998;53:627-35.
8. McFarlane J, Parker B, Soeken K, et al. Safety behaviors of abused women after an intervention during pregnancy. JOGNN. 1998;27:64-9.
9. Parker B, McFarlane J, Soeken K, et al. Testing an intervention to prevent further abuse to pregnant women. Res Nurs Health. 1999;22:59-66.