Should Children Be in the Room When the Mother Is Screened for Partner Violence?

In 1992 the American Medical Association (AMA) and other professional organizations, including the American Academy of Family Practice (AAFP), began advocating the screening of adult women for intimate partner violence.¹ In 1998, the American Academy of Pediatrics (AAP) recommended screening for intimate partner violence and the abuse of women as part of anticipatory guidance at the well-child visit and whenever family violence is suspected.² That recommendation follows the AMA’s guideline, which recommends screening the adult victim alone without the partner or children present in the room¹Table 1. The feasibility of this recommendation, separating the mother and children, needs to be considered. Physicians are already challenged to find time to screen for the expected preventive issues. For example, a study examining family physicians’ practice patterns showed that less than 0.3% discussed violent injury prevention with patients.³ It may be unrealistic and logistically difficult for the physician to ask for privacy with the mother* for routine intimate partner violence screening. If this is the requirement, then screening may not occur. In addition, office staff may not be able to provide supervision for young children while mothers are being screened.

This raises the question: Who is the patient? It is not uncommon to excuse the child from the mother’s Papanicolaou test because the mother is the patient. Does it make sense to excuse children from their own well-child visit? It may be appropriate to have privacy with the mother for in-depth discussions, but routinely excusing children from their own office visit to screen the mother for partner violence is not practical. Similarly, chart documentation is complicated. Abuse of the mother does not belong in the child’s chart, since both parents or legal guardians have legal access.⁶ However, if the information is obtained during a well-child visit, where can it be documented?

Screening and documentation must be done in a confidential manner that ensures the mother’s safety. Failure to do this may have life-threatening consequences for her and the children. If a child shares information with the person committing the violence about “what Mommy discussed with the doctor today,” there may be retaliation toward the mother and the child. The same is true if the partner reads about his abusive behavior toward the mother in the child’s medical chart.

Are there ways of asking general screening questions in front of the children? What issues need to be clarified so intimate partner violence identification becomes routine and safe?

There is little in the literature about this subject. Qualitative research methods are useful for exploring issues that have gaps between knowledge and practice and for highlighting the areas that need more research. The purpose of this study was to weigh the benefits of routine screening for intimate partner violence against the risks to the mother’s safety that can occur when children are present during the screening. Our findings should help develop practical screening recommendations that preserve the safety of the mother and her children. This allows physicians to identify troubled families and link them to resources.

Methods

Experts from the Midwest participated in individual interviews and focus groups using open-ended questions. Sampling was done to the point of theoretical saturation (ie, no new information was being generated). Responses were coded and categorized into themes, and frequency counts were done. The Institutional Review Board at the University of Cincinnati approved the study protocol.

Measurement

Most physician education focuses on the adult victim of domestic violence and her presentation in the medical office;^1,7,8 our study focused on the child. The questions asked of the study participants Table 2 explored issues associated with screening for intimate partner violence in front of a child and the assessment of a child for the potential of actual abuse and the consequences of being a child witness.^2,8-11 This includes the potential consequences of involving child protective services.

Interview Subjects and Procedure

Twelve experts who work with children and their parents in a health/mental health office or domestic violence agency were interviewed. These experts included 2 child psychologists; 1 social worker; 1 domestic violence state coordinator, and 8 experienced physicians (5 pediatricians, 3 family physicians). The snowball technique was used for identifying experts, where one participant referred the researcher to other participants.¹² No one refused to participate in the study. Verbal consent was obtained and confidentiality was assured.

The average interview lasted 45 minutes and consisted of 5 open-ended questions. For the 2 long-distance participants, the interviews were conducted by phone. The participants set the pace of the interviews with their responses, and the researcher used close-ended questions to clarify any unclear answers. At the conclusion of each interview, the participant was asked for feedback. Suggestions were incorporated into subsequent interviews. Seven interviews were taped and transcribed; 5 were transcribed from the researcher’s notes because of technical problems.

Focus Group Subjects and Procedure

Three focus groups were convened involving a total of 17 experts. Domestic violence agency directors preferred a group format, and the director of social work suggested a focus group with her department Table 3.

Focus groups met once for approximately an hour and a half. The researcher used the same 5 open-ended questions as in the individual interviews to stimulate discussion. The researcher moderated the focus groups, encouraging all participants to share. The focus group sessions were transcribed using audiotapes and the researcher’s notes.

Data Analysis

Discussion content was reviewed and coded and categorized according to prominent themes. For example, opinions about screening for partner violence were divided into subcategories (eg, personal screening practices, use of general screening questions, use of questionnaires, age at which a child can stay with the parent during screening, age at which a child can be screened about the parents, and when to ask for privacy with the parent). The type of expert (physician, social worker, psychologist, domestic violence agency director) and method of interview (individual interviews or focus group) were noted for each subcategory. Frequency counts of prominent themes were used to characterize the data. Pertinent quotes were selected to illustrate the key issues.

Results

Demographics of Experts

Among those participants interviewed individually, the least experienced person had been in practice for 7 years; most had been in practice for more than 15 years. Eighty-three percent were white, and 83% were women. Of the physicians, 3 were family physicians, and 5 were pediatricians. Less is known about the length of practice of those participating in the focus groups, but all were currently employed in either a domestic violence agency or in social work. Of the focus group participants, 88% were women, and 88% were white.

Screening for Partner Violence with Children Present

All individuals and focus groups agreed that screening of the mother could be done with AMA-type questions (Table 1) in front of children aged younger than 2 or 3 years. A majority of the experts thought that general screening questions for domestic violence were appropriate in front of children of all ages. These screening questions were different from the questions recommended by the AMA. Suggested formats of the general screening questions were more general and family or child focused, such as: Everyone has conflicts; how do you resolve them? What happens in your house when people are angry? Has your child ever been exposed to anything that would make him nervous or upset? Does your child have nightmares as a result of family disruptions?

Several experts warned about being sensitive to cultural differences. One participant said, “Some homes may be more emotive or physical. Don’t focus only on hitting. Sometimes the controlling behaviors are worse, such as limiting access to the car, money, or food.”

Those who advocated general screening questions said: “Err on the side of getting more information. We have been too conservative.” “The children know what is going on anyway. The victim often denies the child knows. It is important for the kids to know that there are other ways to live and that help is available.”

All individuals and focus groups agreed that further questioning of a mother who gives an equivocal or a positive response and the sharing of resources and crisis numbers should be done in private. A minority of the experts (advocates from one domestic violence agency and one nonphysician expert) discouraged general screening for domestic violence in front of children older than 2 or 3 years. In fact, although the 2 focus groups with domestic violence advocates both started by supporting private screening, they reached different conclusions. One group reached the consensus that general questions were suitable and perhaps therapeutic for children. The other focus group of domestic violence advocates remained concerned about screening in front of the children. The concerns about screening in front of older children included: the possibility of violence if the child tattles on the mother’s conversation with the physician, and further traumatization of the child by hearing the mother recount the abusive events or by hearing her deny the violence at home, thereby continuing the family secret.

“If the child is aligned with the batterer, the child could go home and say: ’Mom told what you do when you are mad at the doctor’s office today.’”

“The child may understand that Mom is breaking the family secret and be concerned for Mom’s welfare.”

“If the victim lies, and victims often do, the child will hear the mother lie, reinforcing the continued secret.”

Written questionnaires were consistently mentioned by participants as an expedited method for screening in a busy practice. Questionnaires covering developmental and preventive issues are often handed out as part of the well-child visit, and additional questions could be included. By distributing the written questionnaire in advance, victims or potential victims of domestic violence can be identified by their responses; clinic staff can then arrange for follow-up questioning in private. This eliminates the awkwardness of face-to-face questioning followed by escorting the children out of the room if the parent gives a positive response. One participant said, “Questionnaires might help a victim organize her thinking for later discussion.”

The main concerns raised about questionnaires were confidentiality and the impact of concerns about confidentiality on reliability. Also, the questionnaires cannot be used if the violent partner brings the child in for the visit. It was suggested that it would be better if the questionnaires were not part of the medical record.

“If the questionnaire done at the well-child visit becomes part of the permanent chart, the father/perpetrator has access to the chart. This might be dangerous for the victim. Questionnaires could be filled out and then discarded.”

“She might be honest if she knows that it is not being filed in the chart.”

Physician Knowledge and Screening Practices

Only one of the participating physicians was screening for partner violence at the well-child visit. That physician was asking the mother if it was permissible to ask questions about violence in front of her children. At the time, the physician had not experienced any negative effects of this practice and had not considered the specific risk of the child’s telling anyone or of the child hearing the mother’s denial. Another physician was planning protocols for her hospital’s ambulatory clinic to begin screening for intimate partner violence. All of the family physicians screened some adult women but felt that they should screen more consistently. All individuals and focus groups stressed the value of physicians sharing resources and crisis numbers with victims.

The signs and symptoms of the children who witness violence between adults in their home^13-22 were not well described by the physicians. These signs and symptoms are important, because recognizing them in the office setting may trigger the physician to screen for violence when universal screening is not being done.

The Links Between Child Abuse and Domestic Violence

Consistent with the literature, nonphysician experts and focus groups thought it was imperative to screen for intimate partner violence when child abuse was suspected and to screen for child abuse when intimate partner violence was identified.^9-11 An expert stated, “Looking only at domestic violence or only at child abuse is dealing with just half of the family.”

Only 3 of the 9 physicians had screened for child abuse when intimate partner violence was identified, despite the fact that the pediatricians were more comfortable screening for child abuse than for intimate partner violence.

Participants suggested the following child abuse screening questions when intimate partner violence has been identified: When abuse occurs, are the children hurt? Have you been afraid of your child’s being hurt?

All individuals and focus groups supported the mandatory reporting responsibility of child abuse and neglect even when partner violence has been identified. One member of a focus group suggested: “Encourage the mother to make the report about child abuse to Child Protection as a way to help empower her. If she is unwilling, the report must be made, but do it with her knowledge.”

Discussion

Our study demonstrated that current practice is based on clinical experience, and there is no clear consensus on how to screen mothers for intimate partner violence.

Feasibility in the Busy Office

Experts did not debate the value of screening women for domestic violence, but they discussed the best way to do such screening. Current practice is based on clinical experience; evidence-based guidelines will need to be developed on the basis of research that identifies the most appropriate screening questions to ask in front of the children and at the well-child visit. This research will need to focus on the most effective method (ie, interview vs questionnaire) and the specific questions to be asked.

The Perspective of the Patient, Mother, or Child

Whether to document a mother’s positive response to the screening in her child’s chart is still in question because the person committing the violence, if a legal guardian, has access to the child’s chart. Chart confidentiality issues need further deliberation by medical legal experts. Should the questionnaire become part of the permanent record if it contains questions about intimate partner violence? Is documentation about such violence appropriate in the child’s chart?

Physicians’ Skill and Knowledge

Physicians in this study lacked knowledge about how a child who witnesses violence between adults in the home presents to the medical office. There is a growing body of literature about the adverse effects on children of hearing and seeing violence between adults. Studies estimate that 11% to 20% of children witness violence between their parents or their mother and her intimate partner.¹³ Children who hear and see this violence display a variety of behavioral, physical, cognitive, and emotional problems and symptoms.^13-22 This is important knowledge for those clinicians who see pediatric patients in an environment where universal screening is not feasible. Current domestic violence screening protocols and guidelines should be expanded to include guidance about the signs and symptoms of children of all ages who witness violence between the adults in their home.

Child Abuse and Mandatory Reporting

Only 3 of the 9 physicians in this study were aware of the link between child abuse and intimate partner violence. Physician education should include more information about this relationship so that documented child abuse will be followed up with questions about intimate partner violence, and intimate partner violence protocols will include screening questions for child abuse. However, this raises other difficulties, since child abuse requires mandatory reporting to authorities, usually child protection agencies, in all states. Bringing outside agencies into a volatile situation can be difficult when there is a risk of retaliation from the person committing the violence against the mother. Therefore, there should be procedures to manage child abuse reporting in situations of intimate partner violence.

Limitations

The limitations of our study include the small, regional sample. Results may not be generalizable because of the mandatory report requirement for intimate partner violence in some states.* The snowball technique we employed is widely used in the social sciences,¹² but it may lead to persons who have common interests and experiences. However, in our study, a variety of opinions and experiences were identified. This was an exploratory study that mixed 2 different qualitative methods, in-depth individual interviews, and focus groups. These 2 processes may not lead to equivalent results as indicated by the fact that only one focus group changed it’s viewpoint on private screening as a part of the process. However, there were no obvious systematic differences between the responses from the 2 different settings.

Directions For Further Research

Our study examined important issues on a topic rarely raised in discussions about partner violence: the needs of the child. Exploring how to screen for partner violence with children present revealed that clinicians’ opinions are based on individual experience, and thus there are disagreements and differences in their procedures. More research is needed to develop evidence-based guidelines regarding intimate partner violence screening and intervention when children are involved. This research should be multidisciplinary including, at a minimum, lawyers, child psychologists, social workers, and physicians. In addition, the inclusion of the mother’s perspective will add to the understanding of the problem. Further research should explore: (1) Methodology — It will be important to examine the advantages and disadvantages of face-to-face interviewing and questionnaires; (2) Content — General screening questions for intimate partner violence that can be used with the children present or on a well-child examination questionnaire should be developed and evaluated. Interpretative guidelines that are culturally sensitive and consider the potential of a highly emotive but not abusive family should be included; and (3) Documentation — It is necessary to establish legal guidelines that clarify what kind of documentation or notation about the mother’s abuse by her adult partner is appropriate in the child’s chart.

Conclusions

The development of evidence-based guidelines for screening women for intimate partner violence in clinical practice is critical. Failure to screen because of feasibility issues in a busy practice means that the family continues to live with violence and abuse without intervention. Screening in a manner that puts the victim at further risk for violence (eg, screening in front of a child who is aligned with the abuser) risks retaliation for both the mother and her children. Screening in a manner that allows the child to hear details about the violence may further victimize the child and support his already inappropriate role in the family. Therefore, despite the clear need for future research, we make the following recommendations for current practice:

1. The AMA-formatted questions should be used in the presence of children younger than 2 or 3 years.
2. Screening of a mother in front of older children should be done only with her prior permission. For example, clinicians can say: “I have some routine questions about violence in the home; may I ask them in front of Suzy?”
3. Physicians should educate patients about intimate partner violence and distribute resource materials and crisis phone numbers. This can be done routinely even when screening is not possible; offices should have pamphlets and posters in the examination room or bathroom.
4. When either child abuse or intimate partner violence is identified, the other should be screened for and considered.

Acknowledgments

Thanks to Jennifer L. Gossett and Susan L. Rosenthal for their assistance, and to the participants in our interviews and focus groups.

In 1992 the American Medical Association (AMA) and other professional organizations, including the American Academy of Family Practice (AAFP), began advocating the screening of adult women for intimate partner violence.¹ In 1998, the American Academy of Pediatrics (AAP) recommended screening for intimate partner violence and the abuse of women as part of anticipatory guidance at the well-child visit and whenever family violence is suspected.² That recommendation follows the AMA’s guideline, which recommends screening the adult victim alone without the partner or children present in the room¹Table 1. The feasibility of this recommendation, separating the mother and children, needs to be considered. Physicians are already challenged to find time to screen for the expected preventive issues. For example, a study examining family physicians’ practice patterns showed that less than 0.3% discussed violent injury prevention with patients.³ It may be unrealistic and logistically difficult for the physician to ask for privacy with the mother* for routine intimate partner violence screening. If this is the requirement, then screening may not occur. In addition, office staff may not be able to provide supervision for young children while mothers are being screened.

This raises the question: Who is the patient? It is not uncommon to excuse the child from the mother’s Papanicolaou test because the mother is the patient. Does it make sense to excuse children from their own well-child visit? It may be appropriate to have privacy with the mother for in-depth discussions, but routinely excusing children from their own office visit to screen the mother for partner violence is not practical. Similarly, chart documentation is complicated. Abuse of the mother does not belong in the child’s chart, since both parents or legal guardians have legal access.⁶ However, if the information is obtained during a well-child visit, where can it be documented?

Screening and documentation must be done in a confidential manner that ensures the mother’s safety. Failure to do this may have life-threatening consequences for her and the children. If a child shares information with the person committing the violence about “what Mommy discussed with the doctor today,” there may be retaliation toward the mother and the child. The same is true if the partner reads about his abusive behavior toward the mother in the child’s medical chart.

Are there ways of asking general screening questions in front of the children? What issues need to be clarified so intimate partner violence identification becomes routine and safe?

There is little in the literature about this subject. Qualitative research methods are useful for exploring issues that have gaps between knowledge and practice and for highlighting the areas that need more research. The purpose of this study was to weigh the benefits of routine screening for intimate partner violence against the risks to the mother’s safety that can occur when children are present during the screening. Our findings should help develop practical screening recommendations that preserve the safety of the mother and her children. This allows physicians to identify troubled families and link them to resources.

Methods

Experts from the Midwest participated in individual interviews and focus groups using open-ended questions. Sampling was done to the point of theoretical saturation (ie, no new information was being generated). Responses were coded and categorized into themes, and frequency counts were done. The Institutional Review Board at the University of Cincinnati approved the study protocol.

Measurement

Most physician education focuses on the adult victim of domestic violence and her presentation in the medical office;^1,7,8 our study focused on the child. The questions asked of the study participants Table 2 explored issues associated with screening for intimate partner violence in front of a child and the assessment of a child for the potential of actual abuse and the consequences of being a child witness.^2,8-11 This includes the potential consequences of involving child protective services.

Interview Subjects and Procedure

Twelve experts who work with children and their parents in a health/mental health office or domestic violence agency were interviewed. These experts included 2 child psychologists; 1 social worker; 1 domestic violence state coordinator, and 8 experienced physicians (5 pediatricians, 3 family physicians). The snowball technique was used for identifying experts, where one participant referred the researcher to other participants.¹² No one refused to participate in the study. Verbal consent was obtained and confidentiality was assured.

The average interview lasted 45 minutes and consisted of 5 open-ended questions. For the 2 long-distance participants, the interviews were conducted by phone. The participants set the pace of the interviews with their responses, and the researcher used close-ended questions to clarify any unclear answers. At the conclusion of each interview, the participant was asked for feedback. Suggestions were incorporated into subsequent interviews. Seven interviews were taped and transcribed; 5 were transcribed from the researcher’s notes because of technical problems.

Focus Group Subjects and Procedure

Three focus groups were convened involving a total of 17 experts. Domestic violence agency directors preferred a group format, and the director of social work suggested a focus group with her department Table 3.

Focus groups met once for approximately an hour and a half. The researcher used the same 5 open-ended questions as in the individual interviews to stimulate discussion. The researcher moderated the focus groups, encouraging all participants to share. The focus group sessions were transcribed using audiotapes and the researcher’s notes.

Data Analysis

Discussion content was reviewed and coded and categorized according to prominent themes. For example, opinions about screening for partner violence were divided into subcategories (eg, personal screening practices, use of general screening questions, use of questionnaires, age at which a child can stay with the parent during screening, age at which a child can be screened about the parents, and when to ask for privacy with the parent). The type of expert (physician, social worker, psychologist, domestic violence agency director) and method of interview (individual interviews or focus group) were noted for each subcategory. Frequency counts of prominent themes were used to characterize the data. Pertinent quotes were selected to illustrate the key issues.

Results

Demographics of Experts

Among those participants interviewed individually, the least experienced person had been in practice for 7 years; most had been in practice for more than 15 years. Eighty-three percent were white, and 83% were women. Of the physicians, 3 were family physicians, and 5 were pediatricians. Less is known about the length of practice of those participating in the focus groups, but all were currently employed in either a domestic violence agency or in social work. Of the focus group participants, 88% were women, and 88% were white.

Screening for Partner Violence with Children Present

All individuals and focus groups agreed that screening of the mother could be done with AMA-type questions (Table 1) in front of children aged younger than 2 or 3 years. A majority of the experts thought that general screening questions for domestic violence were appropriate in front of children of all ages. These screening questions were different from the questions recommended by the AMA. Suggested formats of the general screening questions were more general and family or child focused, such as: Everyone has conflicts; how do you resolve them? What happens in your house when people are angry? Has your child ever been exposed to anything that would make him nervous or upset? Does your child have nightmares as a result of family disruptions?

Several experts warned about being sensitive to cultural differences. One participant said, “Some homes may be more emotive or physical. Don’t focus only on hitting. Sometimes the controlling behaviors are worse, such as limiting access to the car, money, or food.”

Those who advocated general screening questions said: “Err on the side of getting more information. We have been too conservative.” “The children know what is going on anyway. The victim often denies the child knows. It is important for the kids to know that there are other ways to live and that help is available.”

All individuals and focus groups agreed that further questioning of a mother who gives an equivocal or a positive response and the sharing of resources and crisis numbers should be done in private. A minority of the experts (advocates from one domestic violence agency and one nonphysician expert) discouraged general screening for domestic violence in front of children older than 2 or 3 years. In fact, although the 2 focus groups with domestic violence advocates both started by supporting private screening, they reached different conclusions. One group reached the consensus that general questions were suitable and perhaps therapeutic for children. The other focus group of domestic violence advocates remained concerned about screening in front of the children. The concerns about screening in front of older children included: the possibility of violence if the child tattles on the mother’s conversation with the physician, and further traumatization of the child by hearing the mother recount the abusive events or by hearing her deny the violence at home, thereby continuing the family secret.

“If the child is aligned with the batterer, the child could go home and say: ’Mom told what you do when you are mad at the doctor’s office today.’”

“The child may understand that Mom is breaking the family secret and be concerned for Mom’s welfare.”

“If the victim lies, and victims often do, the child will hear the mother lie, reinforcing the continued secret.”

Written questionnaires were consistently mentioned by participants as an expedited method for screening in a busy practice. Questionnaires covering developmental and preventive issues are often handed out as part of the well-child visit, and additional questions could be included. By distributing the written questionnaire in advance, victims or potential victims of domestic violence can be identified by their responses; clinic staff can then arrange for follow-up questioning in private. This eliminates the awkwardness of face-to-face questioning followed by escorting the children out of the room if the parent gives a positive response. One participant said, “Questionnaires might help a victim organize her thinking for later discussion.”

The main concerns raised about questionnaires were confidentiality and the impact of concerns about confidentiality on reliability. Also, the questionnaires cannot be used if the violent partner brings the child in for the visit. It was suggested that it would be better if the questionnaires were not part of the medical record.

“If the questionnaire done at the well-child visit becomes part of the permanent chart, the father/perpetrator has access to the chart. This might be dangerous for the victim. Questionnaires could be filled out and then discarded.”

“She might be honest if she knows that it is not being filed in the chart.”

Physician Knowledge and Screening Practices

Only one of the participating physicians was screening for partner violence at the well-child visit. That physician was asking the mother if it was permissible to ask questions about violence in front of her children. At the time, the physician had not experienced any negative effects of this practice and had not considered the specific risk of the child’s telling anyone or of the child hearing the mother’s denial. Another physician was planning protocols for her hospital’s ambulatory clinic to begin screening for intimate partner violence. All of the family physicians screened some adult women but felt that they should screen more consistently. All individuals and focus groups stressed the value of physicians sharing resources and crisis numbers with victims.

The signs and symptoms of the children who witness violence between adults in their home^13-22 were not well described by the physicians. These signs and symptoms are important, because recognizing them in the office setting may trigger the physician to screen for violence when universal screening is not being done.

The Links Between Child Abuse and Domestic Violence

Consistent with the literature, nonphysician experts and focus groups thought it was imperative to screen for intimate partner violence when child abuse was suspected and to screen for child abuse when intimate partner violence was identified.^9-11 An expert stated, “Looking only at domestic violence or only at child abuse is dealing with just half of the family.”

Only 3 of the 9 physicians had screened for child abuse when intimate partner violence was identified, despite the fact that the pediatricians were more comfortable screening for child abuse than for intimate partner violence.

Participants suggested the following child abuse screening questions when intimate partner violence has been identified: When abuse occurs, are the children hurt? Have you been afraid of your child’s being hurt?

All individuals and focus groups supported the mandatory reporting responsibility of child abuse and neglect even when partner violence has been identified. One member of a focus group suggested: “Encourage the mother to make the report about child abuse to Child Protection as a way to help empower her. If she is unwilling, the report must be made, but do it with her knowledge.”

Discussion

Our study demonstrated that current practice is based on clinical experience, and there is no clear consensus on how to screen mothers for intimate partner violence.

Feasibility in the Busy Office

Experts did not debate the value of screening women for domestic violence, but they discussed the best way to do such screening. Current practice is based on clinical experience; evidence-based guidelines will need to be developed on the basis of research that identifies the most appropriate screening questions to ask in front of the children and at the well-child visit. This research will need to focus on the most effective method (ie, interview vs questionnaire) and the specific questions to be asked.

The Perspective of the Patient, Mother, or Child

Whether to document a mother’s positive response to the screening in her child’s chart is still in question because the person committing the violence, if a legal guardian, has access to the child’s chart. Chart confidentiality issues need further deliberation by medical legal experts. Should the questionnaire become part of the permanent record if it contains questions about intimate partner violence? Is documentation about such violence appropriate in the child’s chart?

Physicians’ Skill and Knowledge

Physicians in this study lacked knowledge about how a child who witnesses violence between adults in the home presents to the medical office. There is a growing body of literature about the adverse effects on children of hearing and seeing violence between adults. Studies estimate that 11% to 20% of children witness violence between their parents or their mother and her intimate partner.¹³ Children who hear and see this violence display a variety of behavioral, physical, cognitive, and emotional problems and symptoms.^13-22 This is important knowledge for those clinicians who see pediatric patients in an environment where universal screening is not feasible. Current domestic violence screening protocols and guidelines should be expanded to include guidance about the signs and symptoms of children of all ages who witness violence between the adults in their home.

Child Abuse and Mandatory Reporting

Only 3 of the 9 physicians in this study were aware of the link between child abuse and intimate partner violence. Physician education should include more information about this relationship so that documented child abuse will be followed up with questions about intimate partner violence, and intimate partner violence protocols will include screening questions for child abuse. However, this raises other difficulties, since child abuse requires mandatory reporting to authorities, usually child protection agencies, in all states. Bringing outside agencies into a volatile situation can be difficult when there is a risk of retaliation from the person committing the violence against the mother. Therefore, there should be procedures to manage child abuse reporting in situations of intimate partner violence.

Limitations

The limitations of our study include the small, regional sample. Results may not be generalizable because of the mandatory report requirement for intimate partner violence in some states.* The snowball technique we employed is widely used in the social sciences,¹² but it may lead to persons who have common interests and experiences. However, in our study, a variety of opinions and experiences were identified. This was an exploratory study that mixed 2 different qualitative methods, in-depth individual interviews, and focus groups. These 2 processes may not lead to equivalent results as indicated by the fact that only one focus group changed it’s viewpoint on private screening as a part of the process. However, there were no obvious systematic differences between the responses from the 2 different settings.

Directions For Further Research

Our study examined important issues on a topic rarely raised in discussions about partner violence: the needs of the child. Exploring how to screen for partner violence with children present revealed that clinicians’ opinions are based on individual experience, and thus there are disagreements and differences in their procedures. More research is needed to develop evidence-based guidelines regarding intimate partner violence screening and intervention when children are involved. This research should be multidisciplinary including, at a minimum, lawyers, child psychologists, social workers, and physicians. In addition, the inclusion of the mother’s perspective will add to the understanding of the problem. Further research should explore: (1) Methodology — It will be important to examine the advantages and disadvantages of face-to-face interviewing and questionnaires; (2) Content — General screening questions for intimate partner violence that can be used with the children present or on a well-child examination questionnaire should be developed and evaluated. Interpretative guidelines that are culturally sensitive and consider the potential of a highly emotive but not abusive family should be included; and (3) Documentation — It is necessary to establish legal guidelines that clarify what kind of documentation or notation about the mother’s abuse by her adult partner is appropriate in the child’s chart.

Conclusions

The development of evidence-based guidelines for screening women for intimate partner violence in clinical practice is critical. Failure to screen because of feasibility issues in a busy practice means that the family continues to live with violence and abuse without intervention. Screening in a manner that puts the victim at further risk for violence (eg, screening in front of a child who is aligned with the abuser) risks retaliation for both the mother and her children. Screening in a manner that allows the child to hear details about the violence may further victimize the child and support his already inappropriate role in the family. Therefore, despite the clear need for future research, we make the following recommendations for current practice:

1. The AMA-formatted questions should be used in the presence of children younger than 2 or 3 years.
2. Screening of a mother in front of older children should be done only with her prior permission. For example, clinicians can say: “I have some routine questions about violence in the home; may I ask them in front of Suzy?”
3. Physicians should educate patients about intimate partner violence and distribute resource materials and crisis phone numbers. This can be done routinely even when screening is not possible; offices should have pamphlets and posters in the examination room or bathroom.
4. When either child abuse or intimate partner violence is identified, the other should be screened for and considered.

Acknowledgments

Thanks to Jennifer L. Gossett and Susan L. Rosenthal for their assistance, and to the participants in our interviews and focus groups.

In 1992 the American Medical Association (AMA) and other professional organizations, including the American Academy of Family Practice (AAFP), began advocating the screening of adult women for intimate partner violence.¹ In 1998, the American Academy of Pediatrics (AAP) recommended screening for intimate partner violence and the abuse of women as part of anticipatory guidance at the well-child visit and whenever family violence is suspected.² That recommendation follows the AMA’s guideline, which recommends screening the adult victim alone without the partner or children present in the room¹Table 1. The feasibility of this recommendation, separating the mother and children, needs to be considered. Physicians are already challenged to find time to screen for the expected preventive issues. For example, a study examining family physicians’ practice patterns showed that less than 0.3% discussed violent injury prevention with patients.³ It may be unrealistic and logistically difficult for the physician to ask for privacy with the mother* for routine intimate partner violence screening. If this is the requirement, then screening may not occur. In addition, office staff may not be able to provide supervision for young children while mothers are being screened.

This raises the question: Who is the patient? It is not uncommon to excuse the child from the mother’s Papanicolaou test because the mother is the patient. Does it make sense to excuse children from their own well-child visit? It may be appropriate to have privacy with the mother for in-depth discussions, but routinely excusing children from their own office visit to screen the mother for partner violence is not practical. Similarly, chart documentation is complicated. Abuse of the mother does not belong in the child’s chart, since both parents or legal guardians have legal access.⁶ However, if the information is obtained during a well-child visit, where can it be documented?

Screening and documentation must be done in a confidential manner that ensures the mother’s safety. Failure to do this may have life-threatening consequences for her and the children. If a child shares information with the person committing the violence about “what Mommy discussed with the doctor today,” there may be retaliation toward the mother and the child. The same is true if the partner reads about his abusive behavior toward the mother in the child’s medical chart.

Are there ways of asking general screening questions in front of the children? What issues need to be clarified so intimate partner violence identification becomes routine and safe?

There is little in the literature about this subject. Qualitative research methods are useful for exploring issues that have gaps between knowledge and practice and for highlighting the areas that need more research. The purpose of this study was to weigh the benefits of routine screening for intimate partner violence against the risks to the mother’s safety that can occur when children are present during the screening. Our findings should help develop practical screening recommendations that preserve the safety of the mother and her children. This allows physicians to identify troubled families and link them to resources.

Methods

Experts from the Midwest participated in individual interviews and focus groups using open-ended questions. Sampling was done to the point of theoretical saturation (ie, no new information was being generated). Responses were coded and categorized into themes, and frequency counts were done. The Institutional Review Board at the University of Cincinnati approved the study protocol.

Measurement

Most physician education focuses on the adult victim of domestic violence and her presentation in the medical office;^1,7,8 our study focused on the child. The questions asked of the study participants Table 2 explored issues associated with screening for intimate partner violence in front of a child and the assessment of a child for the potential of actual abuse and the consequences of being a child witness.^2,8-11 This includes the potential consequences of involving child protective services.

Interview Subjects and Procedure

Twelve experts who work with children and their parents in a health/mental health office or domestic violence agency were interviewed. These experts included 2 child psychologists; 1 social worker; 1 domestic violence state coordinator, and 8 experienced physicians (5 pediatricians, 3 family physicians). The snowball technique was used for identifying experts, where one participant referred the researcher to other participants.¹² No one refused to participate in the study. Verbal consent was obtained and confidentiality was assured.

The average interview lasted 45 minutes and consisted of 5 open-ended questions. For the 2 long-distance participants, the interviews were conducted by phone. The participants set the pace of the interviews with their responses, and the researcher used close-ended questions to clarify any unclear answers. At the conclusion of each interview, the participant was asked for feedback. Suggestions were incorporated into subsequent interviews. Seven interviews were taped and transcribed; 5 were transcribed from the researcher’s notes because of technical problems.

Focus Group Subjects and Procedure

Three focus groups were convened involving a total of 17 experts. Domestic violence agency directors preferred a group format, and the director of social work suggested a focus group with her department Table 3.

Focus groups met once for approximately an hour and a half. The researcher used the same 5 open-ended questions as in the individual interviews to stimulate discussion. The researcher moderated the focus groups, encouraging all participants to share. The focus group sessions were transcribed using audiotapes and the researcher’s notes.

Data Analysis

Discussion content was reviewed and coded and categorized according to prominent themes. For example, opinions about screening for partner violence were divided into subcategories (eg, personal screening practices, use of general screening questions, use of questionnaires, age at which a child can stay with the parent during screening, age at which a child can be screened about the parents, and when to ask for privacy with the parent). The type of expert (physician, social worker, psychologist, domestic violence agency director) and method of interview (individual interviews or focus group) were noted for each subcategory. Frequency counts of prominent themes were used to characterize the data. Pertinent quotes were selected to illustrate the key issues.

Results

Demographics of Experts

Among those participants interviewed individually, the least experienced person had been in practice for 7 years; most had been in practice for more than 15 years. Eighty-three percent were white, and 83% were women. Of the physicians, 3 were family physicians, and 5 were pediatricians. Less is known about the length of practice of those participating in the focus groups, but all were currently employed in either a domestic violence agency or in social work. Of the focus group participants, 88% were women, and 88% were white.

Screening for Partner Violence with Children Present

All individuals and focus groups agreed that screening of the mother could be done with AMA-type questions (Table 1) in front of children aged younger than 2 or 3 years. A majority of the experts thought that general screening questions for domestic violence were appropriate in front of children of all ages. These screening questions were different from the questions recommended by the AMA. Suggested formats of the general screening questions were more general and family or child focused, such as: Everyone has conflicts; how do you resolve them? What happens in your house when people are angry? Has your child ever been exposed to anything that would make him nervous or upset? Does your child have nightmares as a result of family disruptions?

Several experts warned about being sensitive to cultural differences. One participant said, “Some homes may be more emotive or physical. Don’t focus only on hitting. Sometimes the controlling behaviors are worse, such as limiting access to the car, money, or food.”

Those who advocated general screening questions said: “Err on the side of getting more information. We have been too conservative.” “The children know what is going on anyway. The victim often denies the child knows. It is important for the kids to know that there are other ways to live and that help is available.”

All individuals and focus groups agreed that further questioning of a mother who gives an equivocal or a positive response and the sharing of resources and crisis numbers should be done in private. A minority of the experts (advocates from one domestic violence agency and one nonphysician expert) discouraged general screening for domestic violence in front of children older than 2 or 3 years. In fact, although the 2 focus groups with domestic violence advocates both started by supporting private screening, they reached different conclusions. One group reached the consensus that general questions were suitable and perhaps therapeutic for children. The other focus group of domestic violence advocates remained concerned about screening in front of the children. The concerns about screening in front of older children included: the possibility of violence if the child tattles on the mother’s conversation with the physician, and further traumatization of the child by hearing the mother recount the abusive events or by hearing her deny the violence at home, thereby continuing the family secret.

“If the child is aligned with the batterer, the child could go home and say: ’Mom told what you do when you are mad at the doctor’s office today.’”

“The child may understand that Mom is breaking the family secret and be concerned for Mom’s welfare.”

“If the victim lies, and victims often do, the child will hear the mother lie, reinforcing the continued secret.”

Written questionnaires were consistently mentioned by participants as an expedited method for screening in a busy practice. Questionnaires covering developmental and preventive issues are often handed out as part of the well-child visit, and additional questions could be included. By distributing the written questionnaire in advance, victims or potential victims of domestic violence can be identified by their responses; clinic staff can then arrange for follow-up questioning in private. This eliminates the awkwardness of face-to-face questioning followed by escorting the children out of the room if the parent gives a positive response. One participant said, “Questionnaires might help a victim organize her thinking for later discussion.”

The main concerns raised about questionnaires were confidentiality and the impact of concerns about confidentiality on reliability. Also, the questionnaires cannot be used if the violent partner brings the child in for the visit. It was suggested that it would be better if the questionnaires were not part of the medical record.

“If the questionnaire done at the well-child visit becomes part of the permanent chart, the father/perpetrator has access to the chart. This might be dangerous for the victim. Questionnaires could be filled out and then discarded.”

“She might be honest if she knows that it is not being filed in the chart.”

Physician Knowledge and Screening Practices

Only one of the participating physicians was screening for partner violence at the well-child visit. That physician was asking the mother if it was permissible to ask questions about violence in front of her children. At the time, the physician had not experienced any negative effects of this practice and had not considered the specific risk of the child’s telling anyone or of the child hearing the mother’s denial. Another physician was planning protocols for her hospital’s ambulatory clinic to begin screening for intimate partner violence. All of the family physicians screened some adult women but felt that they should screen more consistently. All individuals and focus groups stressed the value of physicians sharing resources and crisis numbers with victims.

The signs and symptoms of the children who witness violence between adults in their home^13-22 were not well described by the physicians. These signs and symptoms are important, because recognizing them in the office setting may trigger the physician to screen for violence when universal screening is not being done.

The Links Between Child Abuse and Domestic Violence

Consistent with the literature, nonphysician experts and focus groups thought it was imperative to screen for intimate partner violence when child abuse was suspected and to screen for child abuse when intimate partner violence was identified.^9-11 An expert stated, “Looking only at domestic violence or only at child abuse is dealing with just half of the family.”

Only 3 of the 9 physicians had screened for child abuse when intimate partner violence was identified, despite the fact that the pediatricians were more comfortable screening for child abuse than for intimate partner violence.

Participants suggested the following child abuse screening questions when intimate partner violence has been identified: When abuse occurs, are the children hurt? Have you been afraid of your child’s being hurt?

All individuals and focus groups supported the mandatory reporting responsibility of child abuse and neglect even when partner violence has been identified. One member of a focus group suggested: “Encourage the mother to make the report about child abuse to Child Protection as a way to help empower her. If she is unwilling, the report must be made, but do it with her knowledge.”

Discussion

Our study demonstrated that current practice is based on clinical experience, and there is no clear consensus on how to screen mothers for intimate partner violence.

Feasibility in the Busy Office

Experts did not debate the value of screening women for domestic violence, but they discussed the best way to do such screening. Current practice is based on clinical experience; evidence-based guidelines will need to be developed on the basis of research that identifies the most appropriate screening questions to ask in front of the children and at the well-child visit. This research will need to focus on the most effective method (ie, interview vs questionnaire) and the specific questions to be asked.

The Perspective of the Patient, Mother, or Child

Whether to document a mother’s positive response to the screening in her child’s chart is still in question because the person committing the violence, if a legal guardian, has access to the child’s chart. Chart confidentiality issues need further deliberation by medical legal experts. Should the questionnaire become part of the permanent record if it contains questions about intimate partner violence? Is documentation about such violence appropriate in the child’s chart?

Physicians’ Skill and Knowledge

Physicians in this study lacked knowledge about how a child who witnesses violence between adults in the home presents to the medical office. There is a growing body of literature about the adverse effects on children of hearing and seeing violence between adults. Studies estimate that 11% to 20% of children witness violence between their parents or their mother and her intimate partner.¹³ Children who hear and see this violence display a variety of behavioral, physical, cognitive, and emotional problems and symptoms.^13-22 This is important knowledge for those clinicians who see pediatric patients in an environment where universal screening is not feasible. Current domestic violence screening protocols and guidelines should be expanded to include guidance about the signs and symptoms of children of all ages who witness violence between the adults in their home.

Child Abuse and Mandatory Reporting

Only 3 of the 9 physicians in this study were aware of the link between child abuse and intimate partner violence. Physician education should include more information about this relationship so that documented child abuse will be followed up with questions about intimate partner violence, and intimate partner violence protocols will include screening questions for child abuse. However, this raises other difficulties, since child abuse requires mandatory reporting to authorities, usually child protection agencies, in all states. Bringing outside agencies into a volatile situation can be difficult when there is a risk of retaliation from the person committing the violence against the mother. Therefore, there should be procedures to manage child abuse reporting in situations of intimate partner violence.

Limitations

The limitations of our study include the small, regional sample. Results may not be generalizable because of the mandatory report requirement for intimate partner violence in some states.* The snowball technique we employed is widely used in the social sciences,¹² but it may lead to persons who have common interests and experiences. However, in our study, a variety of opinions and experiences were identified. This was an exploratory study that mixed 2 different qualitative methods, in-depth individual interviews, and focus groups. These 2 processes may not lead to equivalent results as indicated by the fact that only one focus group changed it’s viewpoint on private screening as a part of the process. However, there were no obvious systematic differences between the responses from the 2 different settings.

Directions For Further Research

Our study examined important issues on a topic rarely raised in discussions about partner violence: the needs of the child. Exploring how to screen for partner violence with children present revealed that clinicians’ opinions are based on individual experience, and thus there are disagreements and differences in their procedures. More research is needed to develop evidence-based guidelines regarding intimate partner violence screening and intervention when children are involved. This research should be multidisciplinary including, at a minimum, lawyers, child psychologists, social workers, and physicians. In addition, the inclusion of the mother’s perspective will add to the understanding of the problem. Further research should explore: (1) Methodology — It will be important to examine the advantages and disadvantages of face-to-face interviewing and questionnaires; (2) Content — General screening questions for intimate partner violence that can be used with the children present or on a well-child examination questionnaire should be developed and evaluated. Interpretative guidelines that are culturally sensitive and consider the potential of a highly emotive but not abusive family should be included; and (3) Documentation — It is necessary to establish legal guidelines that clarify what kind of documentation or notation about the mother’s abuse by her adult partner is appropriate in the child’s chart.

Conclusions

The development of evidence-based guidelines for screening women for intimate partner violence in clinical practice is critical. Failure to screen because of feasibility issues in a busy practice means that the family continues to live with violence and abuse without intervention. Screening in a manner that puts the victim at further risk for violence (eg, screening in front of a child who is aligned with the abuser) risks retaliation for both the mother and her children. Screening in a manner that allows the child to hear details about the violence may further victimize the child and support his already inappropriate role in the family. Therefore, despite the clear need for future research, we make the following recommendations for current practice:

1. The AMA-formatted questions should be used in the presence of children younger than 2 or 3 years.
2. Screening of a mother in front of older children should be done only with her prior permission. For example, clinicians can say: “I have some routine questions about violence in the home; may I ask them in front of Suzy?”
3. Physicians should educate patients about intimate partner violence and distribute resource materials and crisis phone numbers. This can be done routinely even when screening is not possible; offices should have pamphlets and posters in the examination room or bathroom.
4. When either child abuse or intimate partner violence is identified, the other should be screened for and considered.

Acknowledgments

Thanks to Jennifer L. Gossett and Susan L. Rosenthal for their assistance, and to the participants in our interviews and focus groups.