Violence: Recognition, Management and Prevention - Journal of ...

3 downloads 85 Views 71KB Size Report
James Anderson, PhD. ‡. *Section of Emergency Medicine, University of Nebraska Medical Center, Omaha, Nebraska; †University of Nebraska College of.
The Journal of Emergency Medicine, Vol. 23, No. 3, pp. 307–312, 2002 Copyright © 2002 Elsevier Science Inc. Printed in the USA. All rights reserved 0736-4679/02 $–see front matter

PII S0736-4679(02)00543-7

Violence: Recognition, Management and Prevention

PROBLEM GAMBLING IN THE PARTNER OF THE EMERGENCY DEPARTMENT PATIENT AS A RISK FACTOR FOR INTIMATE PARTNER VIOLENCE Robert L. Muelleman,

MD,*

Tami DenOtter, BS,† Michael C. Wadman, James Anderson, PhD‡

MD,*

T. Paul Tran,

MD,*

and

*Section of Emergency Medicine, University of Nebraska Medical Center, Omaha, Nebraska; †University of Nebraska College of Medicine, Omaha, Nebraska; and ‡Preventive and Societal Medicine, University of Nebraska Medical Center, Omaha, Nebraska Reprint Address: Dr. Robert L. Muelleman, University of Nebraska Medical Center, 981150 Nebraska Medical Center, Omaha, NE 68198-1150

e Abstract—It has been suggested that the increase in gambling activity nationally has resulted in an increase in intimate partner violence (IPV). There are apparently no studies that have assessed problem gambling as a risk factor for IPV. To determine if problem gambling in the partner is a risk factor for IPV, a cross-sectional study was conducted at a university-based Emergency Department (ED). All women aged 19 to 65 years who presented to the ED for treatment and were not decisionally incapacitated or acutely ill were eligible. Data were collected by a research assistant during 4 or 8-h blocks covering each day of the week over a 10-week period during the months of June through August 1999. There were 300 consecutive women approached, and 286 (95%) agreed to participate. Of the women who agreed to participate, 237 (83%) reported having an intimate partner in the last year, and 61 (25.7%) of these women were categorized as experiencing IPV. The odds ratio (OR) of experiencing IPV was the main outcome measure, estimated using standard logistic regression, given the presence of various personal and partner characteristics, including problem gambling in the partner. The results revealed that the relative odds were elevated for women whose partners were problem gamblers (adjusted OR: 10.5; 95% CI: 1.3– 82) or problem drinkers (adjusted OR: 6.1; 95% CI: 2.5–14). The presence of both problem gambling and problem drinking in the partner was associated with an even higher OR (adjusted OR: 50; 95% CI: 9 –280). Our study shows that problem gambling in the partner is associated with IPV. The causes of IPV are not fully known, but

the association of problem gambling in the partner with IPV could lead to new intervention strategies and Emergency Medicine research in the future. © 2002 Elsevier Science Inc. e Keywords— gambling; intimate partner violence; domestic violence; alcohol abuse

INTRODUCTION Intimate partner violence (IPV) is a complex societal and medical problem. The causes of IPV are not fully known, although some associated risk factors have been described (1). In recent years, with the increased activity in legalized gambling, anecdotal reports suggest that problem gambling may be associated with IPV. In 1996, the National Gambling Impact Study Commission was created by the 104th Congress “to conduct a comprehensive legal and factual study of the social and economic implications of gambling in the United States.” One recommendation of the commission was to direct the NIH to revise the special program announcements for research applications on pathologic gambling to include, among other things, the “effects on family members, such as divorce, spousal or child abuse, severe financial instability, and suicide” (2).

Violence: Recognition, Management and Prevention is coordinated by Ellen H. Taliaferro, General Hospital, San Francisco, California

RECEIVED: 18 September 2001; ACCEPTED: 4 December 2001 307

MD,

of San Francisco

308

R. L. Muelleman et al.

There is surprisingly little research information about the relationship between problem gambling and IPV despite surveys that suggest there may be such a relationship. A survey of 144 spouses of compulsive gamblers indicated that 50% were physically and verbally abused by their spouses and 12% had attempted suicide (3). In face-to-face interviews, 23% of pathologic gamblers admitted to “hitting or throwing things more than once at spouse or partner” (4). Another survey of 215 spouses of pathologic gamblers indicated that they often suffer from headaches, stomach problems, dizziness, and breathing difficulties, in addition to emotional problems of anger, depression, and isolation (5). Six of the 10 communities surveyed by others have reported an increase in domestic violence relative to the advent of casinos (6). Dual addiction to gambling and alcohol has been described (7–9). Although it has been shown that alcohol abuse in the partner is a significant risk factor for IPV, there are apparently no studies assessing problem gambling as a risk factor for IPV (1). The purpose of this study was to determine if problem gambling in the partner is a risk factor for IPV in patients presenting to an Emergency Department (ED).

worked random 4 or 8-h blocks covering every day of the week between the hours of 6:00 AM through 2:00 AM during the months of June through August, 1999. Data collection protocol. After triage, placement in a private room, and initial medical evaluation, any woman who met the study criteria was approached by the research assistant and asked if she would be willing to participate in a survey concerning women’s medical issues. If she was willing, all family members and friends were asked to leave the room. A scripted description of the study and assurance that her responses would remain confidential was read aloud before obtaining final verbal consent. If consent was obtained, a 45-item questionnaire was read to each study participant. The series of questions included: relationship with a male partner within the last year; episodes of intimate violence; age of patient and partner; race of patient and partner; education of patient and partner; employment status of partner; cohabitation with partner; gambling activities of partner; and alcohol use by partner. Each question was structured and answers were close-ended. Each woman identified as experiencing IPV was also asked if she thought there was an association between her partner’s gambling or alcohol habits and IPV. All women were offered information on resources in the community for IPV. Nebraska is not a mandatory reporting state.

MATERIALS AND METHODS Definitions The study was conducted at the University Hospital of Nebraska Health System, Omaha, NE. The hospital is an urban teaching hospital for the University of Nebraska Medical Center (UNMC). The ED has about 27,000 visits annually, with approximately 70% White, 25% African American, and 5% Hispanic and other patients. The Institutional Review Board (IRB) at UNMC granted approval for the study.

Legal Gambling Opportunities In Nebraska at the time of the study, the legal gambling options available were bingo, lottery, pari-mutuel betting, and Indian reservation casinos. Also at the time of the study, casino gambling was legal in Iowa. The closest casinos were within 10 miles of the study site. Study population. Inclusion criteria for the study were any women age 19 through 65 years who presented to the ED for treatment. Exclusion criteria were women who were decisionally incapacitated or incapacitated by illness, or who did not speak English. Data were collected by a single medical student research assistant who

INTIMATE PARTNER: Male with whom the patient considers having an intimate relationship within the past year. INTIMATE PARTNER VIOLENCE: Either physical injury inflicted purposely by the partner or excessive stress or fear related to threats or violent behavior of the intimate partner. IPV CASES: Women who: (1) presented to the ED with an acute injury caused by IPV; or (2) who had experienced IPV within 1 year with any partner; or (3) who had experienced IPV at any time with the current partner. NON-IPV CASES: Women: (1) with a current partner who had never experienced IPV; or (2) who had experienced intimate violence over 1 year ago, but who had a new partner without a history of domestic violence and to whom they referred when answering the questionnaire. Women without a partner in the past year were excluded from consideration in the analysis. PROBLEM GAMBLING: Problem gambling in the partner was determined by the use of the South Oaks Gambling Screen (SOGS). The SOGS is a 20-item questionnaire based on the DSM-III-R criteria for pathologic gambling (10). A score of five or more

Problem Gambling in the Partner

points was classified as problem gambling. The SOGS is the only validated screening tool available to determine problem gambling. PROBLEM DRINKING: Problem drinking in the partner was determined by the use of the CAGE instrument (11). The CAGE questionnaire is a four-item validated questionnaire. A score of two or more points was classified as problem drinking. Data analysis and statistical methods. The primary goal of this study was to determine if a woman’s assessment of problem gambling in her partner was associated with the likelihood of her being a victim of IPV. We used logistic regression to model the probability that a woman was classified as a victim of IPV as a function of various characteristics of the partner (problem gambler or problem drinker status; employment status [full time versus not full time]) and of the woman (age [20 –29, 30 –39, 40 – 49, 50⫹]), race [White, African American, other], education [less than high school; high school; some college; college graduate]). Because the woman’s age, race or ethnicity, and educational level tended to be highly correlated with their partners, these partner factors were not considered in the logistic regression analysis. The model allowed for an assessment of the odds of being a victim of IPV, given certain characteristics of the women and their partners. The model assumes that the odds ratio (OR) associated with two factors (e.g., the woman’s age and education status) is the product of the OR associated with the individual factors. We used the model-based estimated OR as a magnitude of the association of various subject and partner characteristics and the self-report of having experienced IPV. Ninety-five percent confidence intervals for the true OR were computed based on the Wald tests. We used the Hosmer and Lemeshow test to assess model goodness-of-fit (12). SAS (SAS Institute, Cary, North Carolina) was used to conduct the statistical analyses.

RESULTS A total of 300 women were approached for participation, and 286 (95%) agreed to participate. Of the women who agreed to participate, 237 (83%) reported having an intimate partner in the last year, and 61 (26%) of these women were categorized as having experienced IPV. Table 1 shows the demographic characteristics of the female subjects. IPV cases were more likely to be younger, and have more education than the other women interviewed. Table 2 shows the demographic and behavior characteristics of the male partners. IPV case partners were younger and less likely to be fully employed. They

309 Table 1. Distribution of the Demographic Characteristics of the Female Subjects

Subject age (median, IQR) Subject Race or ethnicity White African American Hispanic Asian Other Subject education level Grade school Some high school High school graduate Some college College graduate

IPV Cases (n ⫽ 61)

Non-IPV Cases (n ⫽ 176)

29 (22–39.25)

34 (25–45)

59% 25% 7% 2% 8%

55% 35% 5% 2% 3%

2% 13% 36% 44% 5%

3% 14% 39% 27% 17%

were also more likely to have been classified as problem gamblers, problem drinkers, or both. Ten IPV case partners had problems with both gambling and alcohol. Table 3 presents the adjusted ORs, estimated using logistic regression, for IPV according to characteristics of the subjects and partners. Because of missing data, 78 of the 286 subjects were not included in the logistic regression analysis. The Hosmer-Lemeshow test suggested that the model provided a reasonable fit to the observed data. Compared to women who reported that their partner was neither a problem gambler or problem drinker, the odds of IPV were significantly elevated for women whose partners were problem gamblers but not problem drinkers (adjusted OR: 10.5, 95% confidence interval [CI]:1.4 –103) and also for women whose partners were problem drinkers but not problem gamblers (adjusted OR: 6.1; 95% CI:2.5–15). The presence of both problem gambling and problem drinking in the partner was associated with an even higher OR (adjusted OR: 50; 95% CI: 11–382). The adjusted OR for individuals whose partners were not working full-time was 2.2 (95% CI: 0.91–5.4). Neither the women’s age nor race appeared to be a predictor of risk. However, women with some college education (but not college graduates) appeared to be at increased risk (adjusted OR: 5.1, 95% CI: 1.5–20). Nine (64%) of the 14 women in the IPV case group with partners with problem gambling thought there was an association between their partners’ gambling habits and abuse; two (14%) were unsure. Twenty-six (74%) of the 35 women in the IPV case group with partners with alcohol abuse thought there was an association between the partners’ drinking habits and abuse; 1 (3%) was unsure. Of the 10 women in the IPV

310

R. L. Muelleman et al.

Table 2. Distribution of the Demographic Characteristics of the Female Subjects’ Partners

Partner’s Age (Median, IQR) Partner’s race or ethnicity White African American Hispanic Asian Other Partner’s education level Grade school Some high school High school Some college College graduate Partner’s employment status Full-time employed Part-time employed Intermittently employed Long-term unemployed Recently unemployed Partner’s alcohol abuse Positive Partner’s problem gambling Positive Partner’s alcohol abuse and problem gambling Positive

case group with partners with both problem gambling and alcohol abuse, 6 (60%) thought there was an association between the partners’ gambling habits and abuse, (10%) were unsure; and 9 (90%) thought there was an association between the partners’ drinking habits and abuse.

IPV Cases (n ⫽ 61)

Non-IPV Cases (n ⫽ 176)

30 (23.75–41.5)

37 (26.25–49.0)

49% 34% 12% 0% 5%

49% 38% 6% 2% 5%

7% 25% 39% 16% 13%

3% 13% 38% 25% 21%

65% 10% 2% 22% 2%

82% 7% 2% 4% 5%

57%

14%

23%

3%

16%

2%

DISCUSSION Although previous surveys have alluded to the relationship between problem gambling in the partner and IPV, this is apparently the first study to demonstrate that relationship (3–5). It also confirms that problem drinking

Table 3. Adjusted Odds Ratio Estimates and 95% Confidence Intervals for Risk of Acute Injury Caused By Intimate Partner Violence for Women According to Partner and Subject Characteristics Characteristic Partner Not problem gambler, not problem drinker Problem gambler, not problem drinker Not problem gambler, problem drinker Problem gambler and problem drinker Employment full time Employment not full time Subject Age less than 30 Age 30–39 Age 40–49 Age 50⫹ Race: White Race: African American Race: Other Education: less than high school Education: high school Education: some college Education: college graduate

Adjusted OR

95% Confidence Interval

1.0 10.5 6.1 50.4 1.0 2.2

(reference group) (1.4, 103) (2.6, 15) (10.7, 382) (reference group) (0.91, 5.4)

1.0 0.48 0.85 0.59 1.0 0.48 1.8 1.0 1.7 5.1 1.2

(reference group) (0.13, 1.5) (0.25, 2.6) (0.11, 2.3) (reference group) (0.17, 1.2) (0.57, 5.5) (reference group) (0.51, 6.5) (1.5, 20) (0.20, 6.6)

Problem Gambling in the Partner

in the partner increases the risk of IPV. The presence of both problem gambling and problem drinking in the partner greatly amplifies the risk. We not only showed a statistical relationship, but when the women were asked, 64% who had partners with problem gambling and 60% who had partners with both problem gambling and problem drinking thought there was a relationship between gambling and partner violence. The results of this study demonstrate a correlation between problem gambling and problem drinking in the partner, with 71% of problem gamblers also classified as problem drinkers and 29% of problem drinkers classified as problem gamblers. In other settings, 47% of severe pathologic gamblers met the criteria for either alcohol or drug abuse at some point in their lives, and 33% of hospitalized substance abusers were also pathologic gamblers (7,8). The causes of IPV are unknown. Just as with alcohol abuse, it would be erroneous to conclude that problem gambling alone causes IPV, despite the strong association in our study. Although the precise relationship between problem gambling and IPV is not known, it is interesting that others have observed that IPV and addiction disorders do not merely co-exist—they actually share many features (13). These shared features include loss of control, continuation despite adverse consequences, preoccupation or obsession, tolerance and withdrawal, involvement of the entire family, and the use of the defenses of denial, minimization, and rationalization. The relationship between IPV and dual gambling and alcohol addiction, as seen in our study, may indicate the presence of some common underlying factors (9). One common factor suggested by others could be problems with impulse control. Alcoholic patients have been shown to have a higher rate of impulse– control disorders, with 24% exhibiting intermittent explosive disorder, and 9% with pathologic gambling (14). Substance abuse disorders, such as alcoholism, are linked to impulsivity and pathologic aggression by comorbidity data as well as neurobiologic evidence (15). Some studies examining the psychiatric comorbidity of pathologic gambling behavior also support its inclusion in the diagnostic category of impulse control disorders (16,17). The DSM-IV criteria for the diagnosis of intermittent explosive disorder state that the “essential feature. . . is the occurrence of discrete episodes of failure to resist aggressive impulses that result in serious assaultive acts.” These episodes are “preceded by a sense of tension or arousal and (are) followed immediately by a sense of relief” and the individual may then even “feel remorseful. . . about the aggressive behavior” (18). The preceding description bears close similarity to the ‘cycle of violence’ often used to characterize interpersonal violence episodes. Our study clearly

311

demonstrates a relationship between problem gambling, described by DSM-IV as an impulse control disorder, and IPV, suggesting that in some relationships, impulsecontrol could be one common pathologic thread of these diseases. Another common factor could be the presence of expectation disease (19). Expectation disease has been described as occurring when expectations disable rather than enable; dominate instead of motivate; when efforts to attain increased levels of control result in loss of control. It is characterized by recurring episodes of loss of control resulting from a consistent pattern of disorders of control. Disorders of control include: unmet or excessive need for control, impaired recognition of controllability, malattribution of control, control dissimulation, and fear of loss of control. There is overlap between the control disorders, and the greater the degree of overlap, the more likely loss of control will occur. The manifestation of loss of control results in impatience, jealousy, envy, and patterns of failed relationships as well as avoidance behavior such as alcoholism, violence, nicotine addiction, eating disorders, problem gambling, or sex addiction. This description, tying together issues of control, violence, problem gambling, and problem drinking, again suggests the possibility of a common pathologic thread. Many patients are in the ED as a result of the consequences of substance abuse or violence. Some EDs have developed programs to address these issues (20,21). If in fact many of these problems are not only inter-related, but may be the result of a common pathology, then ED based intervention programs may need to be broader in their approach. The finding that the risk of IPV increased if the partner was not fully employed is not new (1). Possibly the financial stress increases the risk that a man will physically abuse his partner. The additional financial stress brought on by problem gambling could well add to that risk, although the relationship is probably far more complex. Others have found that women with more than a high school education appeared to be at higher risk for injury than women who were high school graduates (1). Our study further delineated this group by finding that women who completed some college but did not graduate were at high risk. This relationship needs further investigation. One limitation of the study is the possibility of selection bias. This was limited by the use of the same eligibility criteria and exclusion for both groups. Also, by including women who were not acutely injured in the IPV case group, it made it less likely to know who was going to be an IPV case until the questionnaire was administered. The high degree of participation and the

312

R. L. Muelleman et al.

similarity of demographics between groups probably limited selection bias. Another limitation is the possibility of interviewer bias. The interview was structured so that the IPV questions were asked before the gambling or alcohol use questions. This may have influenced the approach to the gambling or alcohol use queries. The structured nature of the questions may have limited this bias. Another limitation is the possibility of misclassification. Some of the variables could have been subject to recall bias, including whether there had been abuse in the past year. Because the CAGE and SOGS questions were answered by the patient and not her partner, and have not been validated for use in collateral contacts, she may have over or underestimated the extent of alcohol use or gambling behavior of her partner. The 3% problem gambling rate in the non-IPV case group was similar to a reported 3.3% prevalence rate reported in the Midwest (22). There may be limitations on the use of the SOGS to determine problem gambling. The SOGS was closely based on the then new inclusion of the diagnosis of problem gambling in the 1980 DSM-III (23). Since then, SOGS has been used in population based research in over 45 jurisdictions around the world. The diagnostic criteria for problem gambling in the DSM-IV developed in 1994 are different from the DSM III. Although the SOGS was the most used validated tool at the time of the study, future tools certainly will be based on the DSM-IV criteria (24). Because we examined potential risk factors among women who sought emergency care in one Midwest urban university ED, the risk factors identified may be different from those in the general population. This possibility makes it uncertain whether our findings are generalizable to other settings in or outside of the healthcare system. CONCLUSION Our study identified the association between IPV and problem gambling in the partners of patients presenting to an urban ED. The causes of IPV are not fully known, but the association of problem gambling in the partner with IPV could lead to new intervention strategies in the future for both victims of IPV presenting to the ED and their partners, who may present on separate occasions. The results of this study indicate than when IPV is suspected or identified, a discussion about the partner’s substance abuse or gambling habits may be useful in determining what sort of resources are needed for the patient and her partner. Also, if the NIH revises its

special program announcements for research on pathologic gambling, more funding will be available to study the effects of pathologic gambling not only on IPV, but also on child abuse and suicide.

REFERENCES 1. Kyriacou DN, Anglin D, Taliaferro E, et al. Risk factors for injury to women from domestic violence. N Engl J Med 1999;341: 1892– 8. 2. The National Gambling Impact Study Commission, Executive Summary, June 1999. 3. Lorenz VC, Shuttlesworth DE. The impact of pathologic gambling on the spouse of the gambler. J Comm Psych 1983;11:67–75. 4. Bland RC, Newman SC, Orn H, Stebelski G. Epidemiology of pathologic gambling in Edmonton. Can J Psychiatry 38:108 –12. 5. Lorenz VC, Yaffee RA. Pathological gambling & psychosomatic, emotional and mental difficulties as reported by the spouse. J Gambling Behav 1988;4:13–26. 6. National Opinion Research Center at the University of Chicago. Gambling impact and behavior study: report to the national gambling impact study commission. 1999:78. 7. Ramirez LF, McCormick RA, Russo AM, Taber JI. Patterns of substance abuse in pathologic gamblers undergoing treatment. Addict Behav 1983;8:425– 8. 8. Daghestani AN, Elenz E, Crayton JW. Pathological Gambling in hospitalized substance abusing veterans. J Clin Psychiatry 1996; 57:360 –3. 9. Murray JB. Review of research on pathological gambling. Psychological Reports 1993;72:791– 810. 10. Lesieur H, Blume S. The South Oaks Gambling Screen (SOGS): a new instrument for the identification of pathologic gamblers Am J Psychiatry 1987;144:1184 – 8. 11. Ewing JA. Detecting alcoholism. The CAGE questionnaire. JAMA 1984;252:1905–7. 12. Hosmer DW, Jr, Lemeshow S. Applied logistic regression. John Wiley and Sons, New York, 1989. 13. Irons R, Schneider JP. When is domestic violence a hidden face of addiction? J Psychoactive Drugs 1997;29:337– 44. 14. Lesieur HR, Blume SB, Zoppa RM. Alcoholism, drug abuse, and gambling. Alcohol Clin Exp Res. 1986;10:33– 8. 15. Brady KT, Myrick H, McElroy S. The relationship between substance use disorders, impulse control disorders, and pathological aggression. Am J Addict 1998;7:221–230. 16. Blaszcynski A. Pathological gambling and obsessive— compulsive spectrum disorders. Psychol Rep 1999;84:107–13. 17. Steel Z, Blaszcynski A. Impulsivity, personality disorders and pathological gambling severity. Addiction 1998, 93:895–905. 18. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 4th edn. Washington, DC: Author; 1994 19. Frank SH. Expectation disease: a model for understanding stress, control, and dependent behavior. Family Practice 1993;10:23–33. 20. Muelleman RL, Feighny KM. Effects of an emergency department-based advocacy program for battered women on community resource utilization. Ann Emerg Med 1999;33:62– 6. 21. Bernstein E, Bernstein J, Levenson S. Project ASSERT: an EDbased intervention to increase access to primary care, preventive services, and the substance abuse treatment system. Ann Emerg Med 1997;30:181–9. 22. Volbert RA. Prevalence studies of problem gambling in the United States. J Gambling Studies 1996;12:111–28. 23. American Psychiatric Association. Diagnostic and statistical manual of mental disorders, 3rd edn. Washington, DC: Author; 1980. 24. National Opinion Research Center at the University of Chicago. Gambling Impact and behavior study: Report to the national gambling impact study commission. 1999:13–25.