JAMA & ARCHIVES
Arch Fam Med
SEARCH
GO TO ADVANCED SEARCH
HOME  PAST ISSUES  TOPIC COLLECTIONS  CME  PHYSICIAN JOBS  CONTACT US  HELP
Institution: CLOCKSS  | My Account | E-mail Alerts | Access Rights | Sign In
  Vol. 7 No. 1, January 1998 TABLE OF CONTENTS
  Archives
 • Online Features
  Original Contribution
 This Article
 •Abstract
 •PDF
 •Send to a friend
 • Save in My Folder
 •Save to citation manager
 •Permissions
 Citing Articles
 •Citation map
 •Citing articles on HighWire
 •Citing articles on Web of Science (29)
 •Contact me when this article is cited
 Related Content
 •Related article
 •Similar articles in this journal
 Topic Collections
 •Violence and Human Rights
 •Violence and Human Rights, Other
 •Alert me on articles by topic

Validating the Concept of Abuse

Women's Perceptions of Defining Behaviors and the Effects of Emotional Abuse on Health Indicators

Peggy J. Wagner, PhD; Patrick F. Mongan, MD

Arch Fam Med. 1998;7:25-29.

ABSTRACT



Objectives  To validate the construct of abuse in 2 ways: first, to examine female patients' perceptions of abusive behaviors that are typically used in standardized abuse scales; and second, to determine health status symptom and medical utilization differences between women who report emotional abuse and women who are not abused.

Design  Cross-sectional interviews and medical record reviews.

Main Outcome Measures  Modified directions to the Conflict Tactics Scale were used to identify women's perceptions of abusive behaviors. Personal history of abuse was determined by self-report. Health status was measured using the Medical Outcomes Study Short-Form Health Survey-36 and medical services utilization was determined from medical records. The Wahler Physical Symptom Inventory was used to measure symptom experience.

Setting  Patients were interviewed in either a rural primary care practice or an urban medical university practice.

Patients  Four hundred seven women older than 18 years were interviewed. Half were from an urban and half from a rural setting. Sixty-four percent of the sample was black.

Results  Women saw more behaviors as abusive than are typically identified by the Conflict Tactics Scale and abused women identified more abusive behaviors than nonabused women. Significant health status differences were found between women who reported emotional abuse with no concurrent physical or sexual abuse and nonabused women on 7 of the 8 dimensions of the Short-Form Health Survey health status scales and on 25% of measured symptoms.

Conclusions  These findings reflect the idea that women consider many behaviors to be abusive and that abused women perceive more behaviors as abusive than do nonabused women. Given that significant health status differences are shown between emotionally abused and nonabused women, emotional abuse can be viewed as a critical variable in patient health behavior.



INTRODUCTION


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

BOTH RESEARCH data and clinical evidence consistently report high rates of violence against women.1-13 However, in more informal settings, health care professionals often react to reports of family violence with disbelief. This may result from a lack of agreement with the definition of abuse or a belief that emotional abuse does not affect subsequent health status and health-related behaviors. The purpose of the present study is 2-fold: to examine women's definitions of abusive behavior, rather than using operational criteria established by professionals, and to validate the construct of emotional abuse by assessing its association with health status, symptoms, and utilization of medical services.

The rates of violence against women have been measured in a variety of ways and in multiple settings. More than 11% of emergency department visits reflect domestic violence problems, and 54% of the women visiting an emergency department report a lifetime prevalence of exposure to violence.1 One in 6 women report physical or sexual abuse during pregnancy.2 In the population they studied, Hamberger and colleagues3 found that nearly 25% of women in outpatient settings had reported being assaulted in the past year. Others have estimated that of the girls who are now 12 years old, 20% to 30% will suffer a violent sexual attack during their lifetime.12 Finally, overall estimates continue to suggest that the number of women experiencing domestic violence may be as high as 3 to 4 million per year.9

Coupled with the high prevalence is the consistent observation that providers are hesitant to address the issue of violence in their patients. Women patients in emergency departments, ambulatory primary care settings, and relevant specialty services such as obstetrics and gynecology are not queried about violence in their history.1, 5-6 The study by Friedman et al14 demonstrated that inquiries about physical abuse are not made in 67% of initial visits and sexual abuse is not investigated in 89% of cases. Even though physicians believe they could make a difference, they do not ask.14 Is the failure to inquire about abuse related to the questionable and elusive nature of the construct of abuse? Our study is an effort to approach the construct from the patient's point of view and to begin to understand the health-related implications of emotional abuse.


PATIENTS AND METHODS


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

DESIGN

A cross-sectional interview combined with retrospective medical record review using a case-control approach was conducted.

SUBJECTS

Four hundred seven women, all of whom were older than 18 years, were interviewed. Approximately half came from a rural health clinic that is part of a resident rotation and half from an urban clinic in a university practice setting. About 36% of the sample was white and 64% was black.

INTERVIEWERS

Six women, all white, were trained as interviewers: 2 were medical students, 2 were masters'-level psychologists, and 2 held bachelor's degrees. One-day training included presentation of standard interview techniques, interview rehearsal under observation, and procedural details. In addition, all interviewers were accompanied by a trainer during their first day at each of the 2 test sites, observed for 1 interview, and advised on specifics throughout the first day.

DATA COLLECTION PROCEDURES

One interviewer was in a clinic at a time. Interviewers would approach a woman, tell her the intent of the interview, and if she agreed to participate they would interview her prior to her medical appointment. Women were excluded owing to age younger than 18 years, extreme illness, and the presence of children if no other caretaker was available. While each woman was being interviewed, no other woman was approached. Approximately 86% of women approached agreed to participate. Others did not, for a variety of reasons including lack of time (5%), not interested (7%), and not wanting to discuss the issue (3%).

INSTRUMENTATION

The primary measure of abuse was self-report (have you ever been physically, sexually, or emotionally abused in your lifetime). In addition, the Abuse Risk Inventory for Women15 and Conflict Tactics Scale (CTS)16 were administered, although not used as the abuse criteria in the current results.

Symptoms were measured using the Wahler Physical Symptom Inventory,17 which includes 42 specific physical symptoms such as headaches, stomach trouble, and weight gain. Medical services utilization, quantified by number of visits and number of prescriptions, was extracted from medical records for the past 5 years in the clinic in which the participants were seen. Health status was measured using the Short Form Health Survey-36 (SF-36) from the Medical Outcomes Study, a well-validated outcome scale resulting in 8 health dimensions: physical role functioning, emotional role functioning, physical functioning, social functioning, bodily pain, mental health, vitality, and general health perception.18

Women's perceptions of abuse were measured by modifying the instructions to the CTS, an instrument that has been well tested and frequently used in the literature16 as an abuse assessment scale. Standard instructions state that "Couples often have problems. Here are some things that couples sometimes do when problems arise. Have you ever done any of these?" Women are then "stepped" through a hierarchy of increasingly violent behaviors beginning with "discuss the issue calmly" and ending with "using a knife or gun." In our study, directions were modified so that women were asked 2 questions about each of the items in the hierarchical behavioral list of the CTS: whether such a behavior had ever happened to them, and if they viewed that particular behavior as abusive. Women were randomly assigned to 1 of 2 orders of presentation of the modified CTS, with approximately half receiving each question first. No differences in responses were obtained based on order of presentation of the items.

STATISTICAL ANALYSES

Women's perceptions of abuse were compared using {chi}2 analyses. The differences between the emotionally abused and nonabused groups of women were tested using analysis of variance (ANOVA) with the measures of health status, symptoms, and medical services utilization. Differences in medical services utilization data were tested using nontransformed and log-transformed data to normalize the distributions.


RESULTS


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

CLASSIFICATION OF ABUSE

Women were classified into 1 of 4 discrete categories of abuse: not abused, emotional abuse, physical abuse, and sexual abuse. These results have been presented in detail elsewhere.19 Briefly, these categories were derived from women's responses to the open-ended questions "Have you ever been physically, sexually, or emotionally abused?" To be placed in the nonabused group, women had to answer no to all 3 questions (n=140 [33.9%]).

This categorization was designed to reflect a hierarchy of abuse with sexual abuse as the most severe type of abuse, followed by physical and emotional abuse. In both the physical and sexual abuse categories, other concurrent lifetime abuses may have occurred (physical and emotional). These categories reflect the increased psychiatric pathological characteristics often evidenced by victims of sexual abuse vs victims of physical and emotional abuse. In addition, the stereotypical perception of severity of type of abuse corresponds to these mutually exclusive categories.

The emotional abuse category includes women who answered yes to the emotional abuse question and no to both the sexual and physical abuse questions. The intent was to identify a group of women who reported emotional abuse but no concurrent physical or sexual abuse (n=80 [20%]). The physical abuse category includes those women who reported physical abuse, did not report sexual abuse, but may have reported emotional abuse. This group thus represents women who were physically abused or physically and emotionally abused (n=90 [22%]). The sexual abuse category includes women who had any combination of abuse that included sexual abuse (n=97 [24%]).

WOMEN'S PERCEPTION OF ABUSE

Table 1 presents the percentage of women from the entire sample who consider each behavior to be abusive. A large percentage of women consider several of the less physical behaviors to be abusive; eg, 67% of the overall sample considered insults or swearing to be abusive and 75% considered saying something to spite the other person as abusive. Lines have traditionally been drawn between minor and severe violence between the behavior of "slapped the other one" and "kicked, bit, or hit with fist." This break does not occur in the ratings of abuse given by this sample of women.


View this table:
[in this window]
[in a new window]
Percentage of Women Characterizing Behaviors on the CTS* as Abusive


Differences in the definition of abuse between women who self-report abuse and women who do not are presented in Table 1. Differences observed are in a consistent direction, with a greater percentage of abused than nonabused women labeling behaviors as abusive. Significant differences using {chi}2 analysis include the following: insulting or swearing, sulking and refusing to talk about it, saying something to spite the other person, threatening to hit or throw something, and throwing something at the other. In all cases, the behavior is more likely to be labeled as abusive by women who self-report abuse vs women who report no abuse.

No differences were found among the categories of abuse in terms of what they perceived as abusive behavior. That is, emotionally, physically, and sexually abused women were similar in their perceptions.

EMOTIONAL ABUSE AND HEALTH STATUS INDICATORS

Emotionally abused and nonabused women were compared in 3 areas: symptoms, medical services utilization, and health status. Symptom differences were compared using the Wahler Physical Symptom Inventory.17 Emotionally abused women reported significantly greater symptom experience for about 25% of the symptoms, with virtually all symptoms being at greater levels even when not statistically different. Significant differences included difficulty sleeping (P<.05), difficulty losing weight (P<.05), excessive perspiration (P<.05), feeling tired (P<.05), muscular tension (P<.0l), poor health in general (P<.05), feeling hot or cold regardless of weather (P<.05), arm or leg aches or pains (P<.05), shakiness (P<.05), and swelling of arms, hands, legs, or feet (P<.01). Overall symptom score for the Wahler Physical Symptom Inventory was greater in the emotionally abused group than in the nonabused group (P>.01).

In terms of medical services utilization, the emotionally abused groups had more medical visits than the nonabused group—11.46 during that time period vs 10.43 (P<.56). Similarly, emotionally abused women received more prescriptions—11.25 vs 8.96 (P<.19).

Health status differences were tested using the 8 subscales of the SF-36. Emotionally abused women scored as significantly less functional than nonabused women on 7 of the 8 subscales: physical role functioning (P<.06), emotional role functioning (P<.001), sexual functioning (P<.001), bodily pain (P<.01), mental health (P<.001), vitality (P<.02) and general health perceptions (P<.04) (Figure 1).



View larger version (10K):
[in this window]
[in a new window]
Health status score differences between nonabused and emotionally abused women.



COMMENT


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

The results of this study indicate that when women are asked what they think is abusive, they consider more behaviors to be abusive than are typically expressed in the literature. Prevalence rates for abuse would be even higher using the criteria suggested by the women in this sample. Thus, dismissal of the high prevalence of violence in women's lives based on the idea of faulty and overinclusive measurement may be erroneous; that is, from the perspective of the woman patient, even more "acceptable" behaviors are abusive. The overall perceptions of all the women in our study indicate that the line of demarcation of abuse should be even more stringent. In addition, we found that the experience of abuse in one's life relates to more restrictive definitions of appropriate behaviors.

Also of importance is the finding that the health status of women who report only emotional abuse is lower on 7 of the 8 subscales of the SF-36. Although statistical differences in terms of the frequency of medical services utilization between these groups are not observed, the functional differences are large. Emotional abuse produces differences in the health status of a woman. When a woman defines herself as a victim of abuse, even "just" emotional abuse, her health status and her experience of symptoms seems to be modified. Recognition that emotional abuse has substantial psychological effects has long-term health care implications.

Although the study is limited in terms of the representativeness of the sample, the use of only white interviewers, and the collection of medical services utilization data in a retrospective manner, it suggests that nonabused women and prior abuse victims are substantially different in their cognitive perceptions of abuse, and that emotional abuse has substantial health effects.

The clinical implications are consistent with prior research in this area. We need to ask if, as providers, we have hesitated and doubted the data. And if so, why has such hesitation occurred? Our own personal histories with abuse may limit our willingness to be involved. Furthermore, violence is not an easy issue for anyone. Faced with a problem of this magnitude, the health implications of violence become a huge burden for health care providers. Perhaps at some level we maintain a vague, illusionary hope that the data is not accurate, and thus the specter of violence fades.

Continued verification and exploration of the cognitive constructs of abuse of all types by both providers and patients is critical to the clarification and understanding of the clinical impact of family violence. Meanwhile, vague complaints and diverse symptom experiences continue, medical services utilization increases, and pain and lack of vitality remain in these women's lives. Abuse is a real problem with tremendous effects on the healthy functioning of women, which subsequently affects the cost of health care. Women who seek medical care for a variety of somatic problems should be queried about emotional abuse as a life event. Symptoms should be recognized, inquiries initiated, and referrals for treatment made. Emotionally abused women may appear to function well, but the effect on health status and health behaviors may be substantial and long lasting.


AUTHOR INFORMATION


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

Accepted for publication December 20, 1996.

This study was supported by a grant from the American Academy of Family Physicians Foundation, Kansas City, Mo.

An earlier version of this article was presented at the meeting of the North American Primary Care Research Group, Houston, Tex, November 9, 1995, and received the Blue Ribbon Award for Best Paper.

Corresponding author: Peggy J. Wagner, PhD, Director of Research, Department of Family Medicine/HB-3040, Medical College of Georgia, Augusta, GA 30912.

From the Department of Family Medicine, Medical College of Georgia, Augusta.


REFERENCES


 Jump to Section
 •Top
 •Introduction
 •Patients and methods
 •Results
 •Comment
 •Author information
 •References

1. Abbott J, Johnson R, Kozoil-McLain J, Lowenstein SR. Domestic violence against women: incidence and prevalence in an emergency department population. JAMA. 1995;274:1763-1767. FREE FULL TEXT
2. McFarlane J, Parker B, Soeken K, Bullock L. Assessing for abuse during pregnancy: severity and frequency of injuries and associated entry into prenatal care. JAMA. 1992;267:3176-3178. FREE FULL TEXT
3. Hamberger LK, Saunders DG, Hovey M. Prevalence of domestic violence in community practice and rate of physician inquiry. Fam Med. 1992;24:283-287. PUBMED
4. Burge SK. Violence against women as a health care issue. Fam Med. 1989;21:368-373. PUBMED
5. Jones RF. Domestic violence: let our voices be heard. Obstet Gynecol. 1993;81:1-4. WEB OF SCIENCE | PUBMED
6. Saunders D, Hamberger LK, Hovey M. Indicators of woman abuse based on a chart review at a family practice center. Arch Fam Med. 1993;2:537-543. FREE FULL TEXT
7. Goodman L, Koss MP, Russo NF. Violence against women, physical and mental health effects, I: research findings. Appl Prev Psychol. 1993;2:79-89. FULL TEXT
8. Walker EA, Torkelson N, Katon WJ, Koss MP. The prevalence rate of sexual trauma in a primary care clinic. J Am Board Fam Pract. 1993;6:465-471.
9. Sugg N, Inui T. Primary care physicians' response to domestic violence: opening Pandora's box. JAMA. 1992;267:3157-3160. FREE FULL TEXT
10. McCauley J, Kern DE, Kolodner K, et al. The battering syndrome: prevalence and clinical characteristics of domestic violence in primary care internal medicine patients. Ann Intern Med. 1995;123:737-803. FREE FULL TEXT
11. Delahunta EA. Hidden trauma: the mostly missed diagnosis of domestic violence. Am J Emerg Med. 1995;13:74-76. FULL TEXT | WEB OF SCIENCE | PUBMED
12. Johnson AG. On the prevalence of rape in the United States. Signs: J Women Culture Soc. 1980;6:136. FULL TEXT
13. Berenson AB, San Miguel VV, Wilkinson GS. Prevalence of physical and sexual assault in pregnant adolescents. J Adolesc Health. 1992;13:466-469. FULL TEXT | WEB OF SCIENCE | PUBMED
14. Friedman LS, Samet JH, Roberts MS, Hudlin M, Hans P. Inquiry about victimization experiences: a survey of patient preferences and physician practices. Arch Intern Med. 1992;152:1186-1190. FREE FULL TEXT
15. Yegides BL. Abuse Risk Inventory for Women. Palto Alto, Calif: Consulting Psychologists Press; 1989.
16. Straus MA. Measuring intrafamily conflict and violence: the Conflict Tactics (CT) Scales. Marriage Fam. 1979;41:75-88. FULL TEXT
17. Wahler HJ. Wahler Physical Symptom Inventory. Los Angeles, Calif: Western Psychologist Services; 1993.
18. Ware JD, Sherbourne C. The MOS 36-Item Short-Form Health Survey (SF-36), I: conceptual framework and item selection. Med Care. 1992;30:473-483. WEB OF SCIENCE | PUBMED
19. Wagner PJ, Mongan P, Hamrick D, Hendrick KL. Experience of abuse in primary care patients: racial and rural differences. Arch Fam Med. 1995;4:956-962. FREE FULL TEXT

RELATED ARTICLE

How Do You Define Abuse?
Sandra K. Burge
Arch Fam Med. 1998;7(1):31-32.
FULL TEXT  


THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES

Is It Abuse? Deaf Female Undergraduates' Labeling of Partner Violence
Anderson and Kobek Pezzarossi
J Deaf Stud Deaf Educ 2011;0:enr048v1-enr048.
ABSTRACT | FULL TEXT  

Psychosocial Effects of Physical and Verbal Abuse in Postmenopausal Women
Mouton et al.
Ann Fam Med 2010;8:206-213.
ABSTRACT | FULL TEXT  

Computer-Assisted Screening for Intimate Partner Violence and Control: A Randomized Trial
Ahmad et al.
ANN INTERN MED 2009;151:93-102.
ABSTRACT | FULL TEXT  

Screening for Intimate Partner Violence in Medical Settings
Phelan
Trauma Violence Abuse 2007;8:199-213.
ABSTRACT  

Attitudes and Beliefs About Domestic Violence: Results of a Public Opinion Survey: I. Definitions of Domestic Violence, Criminal Domestic Violence, and Prevalence
Carlson and Worden
J Interpers Violence 2005;20:1197-1218.
ABSTRACT  

Issues in Estimating the Prevalence of Intimate Partner Violence: Assessing the Impact of Abuse Status on Participation Bias
WalterMaurer et al.
J Interpers Violence 2003;18:959-974.
ABSTRACT  

Identifying domestic violence: cross sectional study in primary care
Richardson et al.
BMJ 2002;324:274-274.
ABSTRACT | FULL TEXT  

Physical Health Consequences of Physical and Psychological Intimate Partner Violence
Coker et al.
Arch Fam Med 2000;9:451-457.
ABSTRACT | FULL TEXT  

How Do You Define Abuse?
Burge
Arch Fam Med 1998;7:31-32.
FULL TEXT  




HOME | CURRENT ISSUE | PAST ISSUES | TOPIC COLLECTIONS | CME | PHYSICIAN JOBS | HELP
CONDITIONS OF USE | PRIVACY POLICY | CONTACT US | SITE MAP
 
© 1998 American Medical Association. All Rights Reserved.

DCSIMG