Inside “Pandora's Box”

Abstract

OBJECTIVE

To explore the attitudes and experiences of abused women to identify characteristics that helped or hindered abuse disclosure to clinicians and to determine how women viewed potential interventions to improve detection and treatment in a medical setting.

DESIGN

Focus group data conducted and analyzed with qualitative methodology.

SETTING

Three community-based mental health centers and one women's shelter.

PARTICIPANTS

Twenty-one women in group therapy for domestic violence.

MAIN RESULTS

Eighteen (86%) of the 21 women had seen their “regular doctor” in the prior year; only 1 in 3 had discussed the abuse with the clinician. The major discussion themes were medical problems that were exacerbated with abuse, lack of ability to access medical care due to abuser interference, emotional attitudes about abuse that acted as barriers to disclosure, clinician characteristics that helped or hindered disclosure, and treatment experiences and preferences. Women described how their medical problems began or worsened during the abusive period. one in three women described how abusers blocked them from receiving medical care. Women reported intense shame about the abuse and described their self-denial of abuse. Women stated they were inclined to discuss abuse if they felt the clinician was perceived to be caring, was easy to talk to, had a protective manner, or if the clinician offered a follow-up visit. There was no consistent clinician gender preference among the women. One in four women had received psychotropic medication for problems associated with abuse. Many feared addiction, or a loss of alertness, increasing their risk for more abuse.

CONCLUSIONS

Many abused women experience worsening health and seek medical care; most do not volunteer a history of violence even to their regular clinicians. Many of the barriers to disclosure of abuse could be overcome by a physician's knowledge of the link between abuse and medical illness, an understanding of the women's emotions about abuse, and her treatment preferences.

Physical or sexual abuse is experienced by 2 million to 4 million U.S. women each year.¹ Childhood and adult abuse are associated with increased numbers of physical symptoms, depression, anxiety, somatization, drug and alcohol abuse, and suicide attempts.^2–14 Not surprisingly women who have experienced violence have higher medical service utilization.^15–17 Studies indicate that as many as 44% of women presenting to primary care medical practices have been abused sometime in their lives.^{2, 12, 18–21}

Despite the high prevalence of abuse and the associated medical problems, most physicians do not routinely screen their patients for abuse.^{2, 22, 23} Although approximately two thirds of abused female patients have not discussed their abuse with a medical professional, one study found that patients welcome a physician asking about abuse.²⁴ Abused women are reported to have higher dissatisfaction with their regular physicians, feeling that their physicians do not listen, are difficult to talk to, or are not competent to treat illnesses.²⁵

A few studies have described barriers to clinician-patient discussion of abuse and are summarized in Table 1). Sugg and Inui described physician barriers to screening. Physicians feared opening “Pandora's box” and unleashing patient issues that the clinician had neither the time nor the expertise to address.²² Other barriers were fear of offending the patient, a feeling of powerlessness to help the women leave an abusive relationship, and disinclination to consider the possibility of abuse among women of higher socioeconomic backgrounds. In another survey Canadian physicians felt that they should screen patients for domestic violence, but few thought they could effectively treat abuse. Lack of clinician education on detection and treatment was found to be a major barrier to screening.²⁶

Table 1.

A Summary of Studies on Barriers to Physicians' and Patients' Discussion of Abuse

There is conflicting evidence as to whether physician gender is related to detection rates ^{26, 27}; it is also unclear if abused women have physician gender preferences. Prior professional or personal exposure either to violence or to educational programs on abuse has been associated with increased detection by physicians.^{22, 26, 27}

One study has focused on the abused woman's perspective of physicians and the health care system. In Rodriguez et al.'s focus group study in California, abused women talked about the silence surrounding the topic of abuse and what helped them “break the silence”.²⁸ The silence was described as a collusion between the abused women and other members of society: “The unspoken agreement between battered women and other members of society to not disclose or address the battering.” Women reported fearing physical retribution, feeling deeply ashamed, having an obligation to keep their families together, and lacking the economic resources to obtain medical help for their abuse. Many stated the California law mandating a police report for domestic violence might lead to added danger from the abuser. In this study, women identified a compassionate attitude, direct questioning, and appropriate referrals as physician characteristics and actions beneficial in helping them discuss the abuse.

Many groups have recommended that clinicians routinely assess women for domestic violence. However, it has been difficult to design interventions and provide evidence that these interventions lead to improved outcomes. Abused women can provide valuable feedback regarding acceptability and potential effectiveness of proposed interventions.

For this study, our objectives were to explore abused women's experiences with, and perceptions of both clinicians and the health care system to identify characteristics that facilitated or acted as barriers to disclosure of abuse, investigate perceptions of the link between abuse and medical problems, and gain women's opinions of potential interventions to improve detection and management of abuse in the primary care setting.

METHODS

Recruitment

To identify abused women who would be willing to participate in focus group sessions, we contacted two sites providing counseling services to people experiencing abuse in Baltimore, Maryland. To be eligible, women had to be at least 18 years of age, English-speaking, and in group therapy for current or past domestic violence (either by self-referral or by court order). The regular therapists reviewed all focus group questions, explained the purpose of the study, and asked for participation the week before the focus groups were conducted. All participants who were approached agreed to participate. Three of our group sessions occurred in a community-based mental health center; the fourth occurred at a women's shelter for domestic violence. All sessions were audiotaped, the participants' real names were not used or revealed to the investigators, and the tapes and transcripts were reviewed only by the study team. Participants received a small stipend for their involvement. The Institutional Review Board committee at the shelter approved the study.

Study Design and Sessions

We selected a focus group methodology because we felt that women would feel more comfortable discussing this difficult topic in a familiar environment with group and counselor support than individually with a stranger. Focus groups have been shown to be extremely effective in providing in-depth information on the attitudes and feelings of participants about a particular issue.²⁹ Because we were interested in the range of opinions about our topics, we conducted different groups until there appeared to be little new information from the participants. We found that after four different focus groups new comments were limited. The focus group leader attempted to elicit opinions on all questions from all participants and encouraged different opinions.

The focus group leader was the principal investigator, a female physician with a substantial research background in domestic violence. The focus groups ranged in size from three to eight participants, and each lasted approximately 90 minutes. Each participant completed a short questionnaire with background information prior to the focus group session. The regular therapists were present during all the focus group sessions but did not participate.

Our three major focus group questions were (1) Tell us about an experience you had with a doctor or health care physician concerning violence—was it a good or bad experience and why?, (2) What made it easy or hard for you to discuss the violence with a doctor or other health care professional?, and (3) Is there any other information that you think that doctors should know when treating women who have experienced violence? The group leader also included probe questions based on the participant comments to the three major questions; these probes included questions about specific physician characteristics, medical system features (access, confidentiality, insurance issues), medical problems that triggered a physician visit, and treatment and screening preferences in a primary care setting.

Data Collection and Analysis

Audiotapes were transcribed verbatim, and the transcriptions were checked for accuracy by the study leader. The overall strategy was to have investigators independently review the transcript, meet regularly to review and discuss differences of opinion about transcript fragment length and other issues, and to develop themes driven by the participants' own words and phrases.

Two investigators (JM, RAY) independently reviewed the transcripts and separated the participants' statements into fragments with each fragment representing a discrete thought. They advanced through the transcripts a section at a time. They regularly met with a third investigator (DEF) to resolve any differences in fragment length and content and to develop and revise themes on the basis of participants' thoughts and comments. The investigators repeated this process until there was consensus on all fragment length and content, theme development, and the grouping of fragments into themes. A fourth investigator (MWJ) reviewed a portion of the fragments and helped to revise themes and regroup fragments.

The major themes are illustrated by quotations. We selected the quotations based on several criteria including how often comments were made about the particular theme, intensity of the emotion displayed by participants about the theme, and the ability of the quotation to provide new clarity for a theme.

RESULTS

Twenty-one women participated. Socioeconomic characteristics of the women are listed in Table 2. Most of the women were single, separated, or divorced, and there was a diversity of reported family incomes. The participants all had at least some high school education; many had some college or graduate school education. The racial breakdown was representative of Baltimore city with a larger percentage of African–American women. There were no Hispanic women in the groups.

Table 2.

Sociodemographic Characteristics of Participants (n= 21)

Background Questionnaire

Results of the background questionnaire are presented in Table 3. Most of the women had a regular physician and had had a visit within the year; only approximately one in three had discussed the abuse with the physician. Most said that they would answer a written questionnaire about abuse in a doctor's office if they felt that their privacy was protected. One in four had received medication for depression, anxiety, or sleep problems associated with the abuse; nearly all the women who took the medicine stated that it helped.

Table 3.

Background Questionnaire of Participants

Focus Group Themes

In the initial discussion, women described a wide range of positive and negative health care experiences. Ninety-eight percent of all the comments from the focus groups could be classified into three broad categories: (1) physical or mental problems that started or were exacerbated during the abusive relationship, (2) barriers or facilitators to disclosure and treatment, and (3) interventions or office aids that were potentially beneficial or detrimental (Table 4

Table 4.

Major Themes of Focus Groups Exploring Abused Women's Perceptions of and Experiences with Clinicians and Health Services

Medical Problems Exacerbated by Abuse.

Women described many medical and emotional symptoms that began or worsened with abuse which caused them to seek medical care. The physical problems included worsening asthma or hypertension, headaches, eye pain, chest pain, stomach pains, back problems, vaginal bleeding, massive weight gain or loss, insomnia, depression, and anxiety. Though most of the women did not recognize an association between abuse and their physical symptoms at the time that they first presented to their physicians, about one third of the women said that the medical problems improved when they either left the relationship or talked with a health professional about the abuse. For example:

I had asthma for years, but I noticed that during the [abusive] time period, I had a really hard time breathing, and now it's minimal.

I was always sick. People used to call me a hypochondriac because I never told them what was going on.

Some of the women did note that the “stress” of the abuse affected their health adversely. These women would tell their doctors that they were under stress but would not specifically identify their abuse as the source of the stress unless the doctor inquired further or asked directly. Women noted that they wished their doctors would probe more. For example:

I had some [vaginal] bleeding … I tried to explain to the doctor that I was having a lot of stress, and I was kind of hoping he would ask more questions.

Abillity to Access Medical Care.

The major barrier to access was the abuser preventing his partner from receiving medical care. One third stated that an abuser had specifically prevented them from receiving emergency or routine medical care or from being alone with a clinician. For example:

My doctor, I could always talk to her, but I didn't know that she had been calling my house, sending letters, making appointments, because my mail was being intervened. The phone calls were being screened.

If you have a husband hovering over a wife, doting on her … there should be signals going off in the doctors' minds, ‘Maybe we should get this person alone a bit.’

Other expected barriers to access were only infrequently mentioned. Only a few women noted financial constraints or difficulty leaving work to receive health care. Some women were critical of HMOs because of the lack of continuity with a regular doctor; however, others mentioned that the low copayments in HMOs make medical care possible for them.

Attitudes About Abuse That Affected Disclosure.

Participants' feelings about abuse were the most common barriers to disclosure. These feelings included intense shame; denial that abuse was occurring even when it was severe; the fear of the reaction of friends, family, or medical professionals if the abuse was revealed; fear of the consequences to their children; lack of readiness to change the relationship with the abuser; and fear of the abuser's reaction to disclosure. Some felt they had to be “tough” and endure the abuse or were too “proud” to discuss the abuse; others felt “stupid” for not leaving the relationship.

The shame surrounding abuse and the denial that abuse was occurring were the two most frequent and intense emotions. One woman stated:

I think that going to a hospital for domestic violence is like going to the sexually transmitted disease clinic … you feel like the doctors look at you like you're dirty or you weren't protecting yourself.

Interestingly, women of both the highest and lowest family income groups related inferior social status to abuse. They did not view themselves as being part of a low social status, and this perception contributed to their denial of abuse. For example:

He tried to choke me … even when I came to the women's group here, I still didn't think I was abused. I felt like it was someone else's problem, those people, you know, downtown.

When I first came here, I thought it would be a group of low-lifes … You know, people that break beer bottles, scream and argue on Saturday nights.

Women feared the negative perceptions of family, friends, and their physicians if they disclosed the abuse. They also feared the effect of disclosure on their children, as exemplified by such statements as:

I felt like if I shared it, the doctor would make fun of it

and

They want to check your children for sexual abuse … I don't want my children up on nobody's table at the age they are with their legs spread.

Several women felt that they did not mention the abuse because they were not at the stage to leave. Some of the women described a pivotal moment for seeking help, usually precipitated by fear for their own safety, fear of losing their children, or fear of killing the abuser:

I thought they were just disagreements and didn't realize what they were building up into … when he threatened me “I'm going to take your body out in the woods where nobody will ever find you.”

I know when I picked up the screwdriver and was ready to put it through his chest, I knew I had to get out of there because I was ready to kill him.

About one third of the women described fear of either repercussions from the abuser or losing the abusers' love or financial support:

I didn't want them to know what was going on. Then, by me supposedly loving him so much, I didn't want to get him in trouble.

I'll get beat up more. I mean, God forbid they're gonna give him something to be angry about. He's angry about nothing.

Clinician Characteristics That Helped or Hindered Disclosure

The most frequently mentioned barriers were fear that the clinician would look down on them or would blame them for the abuse. The abused women also noted that many clinicians were uncaring, appeared uncomfortable with the topic, were not listening, were too busy or rushed, or were only interested in money. One said:

Sometimes I feel like I should recite a nursery rhyme to see if they're listening.

Some feared that the doctor would disclose the information to other family members; others felt that treatment for abuse was not in the “domain” of a primary care doctor.

Women consistently mentioned that they were more inclined to discuss the abuse if they perceived the clinician to be caring, easy to talk to, protective, or if the clinician offered follow-up.

A few women noted that physicians may not be aware that their actions impact on the abused women's decision to disclose. One commented:

… And just to realize that every chance when they [the doctors] have a victim in their care, that they strongly affect the person later on. Any little comment that they [the doctors] make can affect what the victims' next steps are with their abuse.

The majority of patients felt that physician gender was not an important clinician characteristics in the decision to disclose. Those who had a preference were evenly divided as to whether they preferred a female or male physician. The following were typical comments:

Females are actually more critical than males.

In my experiences, I haven't had many sensitive men in my life [and would not feel comfortable talking to a male doctor] although he was a kind doctor.

Sex [of the doctor] doesn't matter … they're not the ones who hurt me.

Screening and Treatment

Approximately half of the women who commented stated that brochures and domestic violence posters in an office helped them talk about their abuse. Though several women expressed concerns about a loss of confidentiality and the fact that many people have access to the medical record, most said that they would answer a questionnaire about abuse truthfully if administered in a private area. One said:

If I had seen something in his [the doctor's] office, about it [abuse], I would have definitely felt more comfortable bringing it up.

One in four women described receiving medication for depression, anxiety, or sleep disturbances for the emotional problems associated with abuse in the written questionnaire, and nearly all stated that it helped. However, in the focus group sessions, many women were critical of drug therapy for the emotional disturbances associated with abuse. Many feared addiction to the prescribed medications. As one woman said:

You have to very careful [about taking medication] because a lot of times if you follow the doctor's advice, you can become addicted.

Others felt labeled as “crazy” for taking medications. Some felt “brushed off” when they were given medication without counseling. One complained:

But he [the psychiatrist] did not know me from Adam, and within 15 minutes he was writing out prescriptions.

Some women feared a loss of control or alertness when taking medications, as exemplified by this woman's statement:

I'm still living with my husband, and I'm afraid that if I take any kind of medication, that this man is definitely going to have control all over again. So I'm scared, I refused to take anything … I know if I take any kind of medication, I won't be as strong as I am now …

All women who commented felt that referral to women's groups or to other agencies for help was beneficial. However, the women in our groups consistently described negative experiences with psychologists or psychiatrists. For example, one woman described a therapist suggesting that she fought with her husband because she was suffering from premenstrual syndrome. She stated that the the therapist said:

Better start tracking your [menstrual] cycle … you may find that your fights go with your cycle …

DISCUSSION

This study of abused women and health experiences in Baltimore verified some results of former studies and added some new insights into this difficult-to-research area. Despite the fact that the majority of these abused women had regular physicians, only one in three had discussed the abuse with her physician; this low rate of disclosure has been sited in other studies.² As noted in prior studies, the abused women were experiencing a wide variety of health problems.^{2––5, 7, 8, 10, 11, 13, 14} In our focus groups, shame and denial were the most frequent themes mentioned as barriers to discussion about abuse; in some cases the intensity of the shame was so severe that women feared the reaction of their regular physicians.

The barriers to disclosure of abuse described by these abused women are remarkably similar to the physician barriers to discussion described by Sugg and Inui (Table 1).²² Some women noted that physicians lacked the time to address their problems, but other women noted that educational brochures, referrals, and brief supportive comments were helpful. Women in our focus groups also felt that abuse was related to an inferior social class; this belief was part of their basis for denying their abuse. Women noted that the physicians seemed “uncomfortable” when the patient admitted she was abused. Some women affirmed that they chose not to volunteer a history of abuse if they weren't at a stage to address the problem, though others noted that the physician's counsel caused them to realize the magnitude of the problem and to seek help or leave the relationship.

We noted some themes that were different or new from those described by Rodriguez et al.²⁸ Lack of financial resources to see a doctor was not mentioned as a significant barrier even by most of the women with the lowest incomes. The most common barrier discussed was the abuser's blocking visits or interfering with privacy during medical appointments. We found reporting abuse to the police was not a major concern of women in our study, and only one women reported lying about the cause of her injuries when directly questioned. These differences in themes may be due to the difference in ethnic backgrounds of the women in Rodriguez's study, which included Hispanic and Asian Americans, lack of medical insurance coverage in their groups, and the presence of a mandatory domestic violence reporting law in California that is not present in Maryland. In addition, though the groups were similar in the percentage that were single, separated, or divorced, the California women may have been in a different stage of readiness to change their abusive relationships or at a different level of perceived threat from the abuser.

Because of our focus group approach, we were able to further explore certain themes. For example, even though women had worsening health with abuse and sought medical care for their problems, they frequently did not seem to perceive the link between the two and so did not volunteer a history of abuse. Some women hinted to their physicians about a stressful personal problem in the hope the physicians would “invite” them to give more details; this “invitational disclosure,” in which the discloser provides sufficient cues that invite the respondent to notice and perhaps ask more questions, has been described before in a small group of patients with stigmatizing conditions.³⁰ Physician gender was not a consistently important determinant in the decision to disclose; a supportive, nonjudgmental clinician attitude was consistently mentioned as much more important.

In an effort to understand treatment preferences, we explored the women's perceptions of and experiences with several therapies. We found previously described barriers to mental health care such as stigma and addiction.^{29, 31} The fear of addiction was so great that some women doubted specific physician reassurances about prescribed medications. At least one unique fear was described: women worried that the use of psychotropic medications might cause decreased alertness leading to slower reaction time in the event of an escalating argument.

Our study has several strengths. We obtained comprehensive, in-depth information about participants' opinions; use of the background questionnaire and the focus groups was particularly valuable to shed more light on this complex problem. For example, the background questionnaire suggested that although practically all women who received psychotropic medications benefitted from them, the focus group discussion revealed many women had concerns about prescription medications. We sampled a diverse sample of women with different incomes and educational backgrounds; all participated actively in the discussions.

Conversely, focus groups have limited generalizability because of the small numbers of patients involved. In addition, the fact that all of the women were in group therapy might introduce a bias in favor of group therapy and another bias in that these women may be in a different therapeutic phase because they are receiving help. More focus group studies or larger quantitative studies need to be performed to confirm the findings of this study.

This article supports and broadens our understanding of abused women and their perceptions of, and experiences with, clinicians and the health care system. Physicians may be less likely to fear the contents of the “Pandora's box” of domestic violence if they have a better understanding of what's inside. Many of the barriers to discussion of abuse could be overcome by a physician's understanding of the emotions surrounding abuse and the unique treatment concerns of abused women.

SGIM Website

Please visit the Society of General Internal Medicine on their World-Wide Website. SGIM is located at

http://www.sgim.org