Health care providers and patients agree that it is important to screen for intimate partner violence, particularly when they are informed of its association with numerous negative health outcomes.^1–3 However, screening occurs infrequently,^4–6 even though universal screening has been recommended for health care settings.^6–8 Survivors of partner violence often have more contact with health care providers than with social service professionals,⁹ and family planning service providers are the most likely ones to see women during the years that coincide with the highest risk of victimization.^10–11 Thus, family planning providers have a distinct opportunity to identify cases of, assess levels of risk for, and offer risk reduction and referral services related to partner violence; accordingly, a number of professional groups have recommended that screening and follow-up be part of routine practice at all reproductive health centers.^8,12–13 Nonetheless, providers’ perspectives on intimate partner violence policies and procedures in family planning centers have not been thoroughly assessed or documented.¹⁴

A number of factors may impact screening in family planning settings, from availability of empirically validated screening tools and institutional training to individuals’ attitudes and job responsibilities. Studies in other health care settings have indicated that providers’ screening behaviors are biased by lack of education and training; time constraints; perceptions of job responsibilities; and feelings of fear, discomfort, powerlessness and bias toward patients who are at risk.^15–16 Provider and patient beliefs about what constitutes a successful outcome of screening and fears about the possible results of a disclosure (e.g., retaliation by partner or police involvement) may also play a role.^17–19

Several studies have documented positive associations between provider training and subsequent adherence to screening protocols, knowledge of and communication skills for discussing intimate partner violence, and attitudes about the importance of screening.^4,20–22 A national cross-sectional study of general internists, family practitioners, obstetrician-gynecologists and emergency medicine physicians found that practitioners’ having had domestic violence training was associated with increased screening, particularly when training had occurred in the last 12 months.²¹ Provider training is associated with increased feelings of preparedness to inquire about and discuss partner violence with patients,²² but the most effective interventions have included provider education, the implementation of an institutional policy and protocol, and additional on-site resources, such as a victim advocate or social worker.^4,23–25

Two areas need further research. First, almost all studies have focused on one type of provider (mostly physicians).^4,9,26,27 However, job tasks, training levels and experience with patients differ by provider type, and these differences may lead to varied perceptions, attitudes and barriers. One study found that social workers perceived fewer barriers to screening than did family practitioners and obstetrician-gynecologists;²⁶ the researchers concluded that this disparity reflected greater attention to partner violence in social workers’ education. However, little else is known about differing job responsibilities within the same clinic setting, which could impact how health care workers perceive their role in screening or barriers (whether attitudinal or organizational) to doing so. Second, research has focused primarily on providers who were not trained in conducting screening or experienced in doing so within organizations with established institutional procedures.^2,4,18,28 A few studies have compared providers with and without training, but their findings have been equivocal. Some found that training was associated with increased confidence, comfort and preparedness;^22,23 others showed that training alone, without additional organizational policies and protocols in place that increase experiences with performing screening, had little relationship to screening practices.^4,29

Our study is intended to expand current knowledge by comparing licensed family planning service providers (advanced practice clinicians and social workers) and unlicensed ones (health care assistants) who work in a setting guided by an institutional policy and procedure for intimate partner violence screening. (Advanced practice clinicians are certified nurse-midwives, nurse practitioners and physician assistants.) We combined providers into these categories because we hypothesize that compared with unlicensed health care assistants, licensed professionals have more educational and professional training that enhances their understanding of how psychosocial factors impact health and their role in intervening. We report data from focus groups and surveys to describe perspectives on screening among providers at three large family planning centers in New York City.

METHODS

Study Sites and Their Screening Policy

The study was conducted in 2009 at three family planning centers in New York City (one each in Manhattan, Brooklyn and the Bronx). All sites operated under the same family planning organization, which had an intimate partner violence screening and response policy and protocol that had been in place for more than two years, as well as on-site social workers who provided additional assessment and counseling after a positive response to violence screening (a somewhat rare resource at family planning centers). These centers provide family planning services to more than 40,000 patients per year; patients are adolescents and adults from a wide variety of racial, ethnic and economic backgrounds.

Patients attending these centers routinely complete a written medical history form before they receive medical services. The form contains four empirically tested statements about partner violence in the past year:^1,30,31 "Things have been going well in my relationship"; "My partner threatened or frightened me"; "My partner forced me to have sex when I didn’t want to"; and "My partner hit, slapped or physically hurt me." Responses are answered on a five-point scale (1=never to 5=always). Two dichotomous questions assess lifetime prevalence of violence: "Have you ever been slapped, hit, or physically abused by a partner?" and "Has anyone ever raped you or forced you into a sexual act?" Depending on the type of service appointment a woman has, a licensed provider or health care assistant reviews this form and then conducts a secondary verbal screen. If violence is disclosed, the patient’s family planning needs are discussed in the context of partner violence and she is given the option to speak with a social worker at the center. Adolescents younger than 17 who report victimization of any kind are required to meet with a social worker for a fuller assessment. Any patient who speaks with a social worker receives a risk assessment, safety planning regarding sexual and reproductive health, and referrals for social services.

Women’s privacy and confidentiality regarding violence are maintained in the same way that all other health care information is protected. If a woman has someone accompany her to an appointment and cannot protect the privacy of her medical history form because she completes the form in the presence of that individual, her visit with the provider always includes a private secondary verbal screen for violence and other sexual and reproductive health concerns, without anyone else in the room.

Study Design

Our study used an integrated mixed-methods approach, a strategy for studying health services that can "help researchers to engage with the complexity of health, health care, and the environment in which studies take place."³² We designed our research questions and selected participants in such a way so as to collect qualitative and quantitative data concurrently (i.e., qualitative information did not inform later quantitative measurement, and neither method was dominant).^33–34 The same participants provided both types of data. We undertook this approach to compare and contrast data in a way that would highlight different facets of a phenomenon.³⁵

We invited family planning center staff whose duties included conducting screening for and follow-up of intimate partner violence to complete a brief, anonymous, self-administered survey at the end of routine, monthly administrative staff meetings. Recruitment and data collection took place at five staff meetings: one for advanced practice clinicians and one for social workers from all three centers combined, and one for health care assistants at each center. Physicians were not included because partner violence screening was not a standard part of their job. We asked a total of 75 staff to participate; 64 (85%) agreed.

After the same administrative meetings, we invited staff to participate in a 30-minute focus group discussion about the intimate partner violence policy and procedures. We informed staff that the discussion would not affect their job evaluations, as no information about the content of individual contributions or whether staff participated would be reported to their supervisors. All 75 providers agreed to participate. No supervisory staff were present at the discussions. Five focus groups were conducted: one with health care assistants at each center (participants numbered 12, 16 and 17), one with 25 advanced practice clinicians and one with five social workers. The principal investigator conducted the focus groups, which were audiotaped and included no identifying information about participants. Institutional review board approval was obtained for use of human subjects in this study; participants received no incentives other than snacks provided during the focus groups.

Measures

•Quantitative. The survey took approximately 10 minutes for staff to complete. Questions covered job characteristics (location, title, number of years working in family planning and receipt of training about intimate partner violence), as well as attitudes toward and perceived barriers to intimate partner violence screening. The items on attitudes and barriers, which were adapted from other studies in similar health care settings,^21,36–37 were used to create composite variables based on weighted means; such variables maintain the original measurement and are easier to interpret than summary scores.

Attitudes toward performing partner violence screening were measured with 14 items that reflected broad opinions about the usefulness of screening and providers’ frustration and comfort with it. The organization’s policy had been in place for several years, so we worded the questions in such a way that they would not reflect expected organizational views (i.e., positive views about screening). The questions were introduced by the statement "We acknowledge that you follow the organizational protocol on partner violence. We would like to know whether you agree or disagree that each statement below makes it more difficult to discuss partner violence with patients." Items included the following: "Asking patients about violence opens the door to time-consuming activities that aren’t part of my job"; "If the patient won’t leave the relationship, I should not have to spend my time talking to them about it"; "My patients’ relationship history is none of my business"; "I am uncomfortable discussing abuse with my patients"; "Asking patients about violence is frustrating because they don’t want to leave their partner"; and "I do not think my patients want me to ask them about it, if they haven’t told me themselves." All items were answered on five-point scales (1=strongly disagree, 5=strongly agree); on the composite variable (Cronbach’s alpha, 0.86), 5 reflected the most negative attitude, and 1 the most positive

Additionally, two items measured provider perceptions of the helpfulness of using a written screening form and conducting secondary verbal screening with patients. These were answered on five-point scales (1=not at all helpful, 5=very helpful). They were analyzed separately because they did not form good reliability with other attitudinal questions, probably because they asked about participants’ global perception of the helpfulness of screening for patients, rather than their view of conducting screening themselves.

Job-related resource barriers to screening were measured with three items: "There is not enough time to identify and refer patients for partner violence in addition to attending to other health concerns,""There is a lack of adequate training" and "There is a lack of resources for community referrals." All items were answered on five-point scales (1=strongly disagree, 5=strongly agree); the composite variable had moderate internal reliability (Cronbach’s alpha, 0.72).

Eleven statements assessed how prepared providers felt to perform different aspects of screening and respond to patient disclosures. We introduced these items by telling participants that their answers would help us plan ongoing professional development. Items included "asking directly about an observed physical injury,""referring the patient to a social worker,""conducting a risk assessment with a patient,""documenting a statement from a patient who discloses abuse,""bringing up the issue when the patient returns for another visit,""calling the domestic violence hotline with a patient" and "informing the patient she is not to blame." These items were scored on five-point scales (1=not at all prepared, 5=very prepared); the aggregated preparedness variable had good internal reliability (Cronbach’s alpha, 0.94).

•Qualitative. One discussion guide was used to facilitate all five focus groups. The discussions started with a review of the official policies and procedures for screening, as well as the screening questions on the medical history form. Staff were asked about their experiences with patient disclosures of partner violence; about barriers to screening (with probes about time constraints, personal comfort, and cultural and age differences related to discussions of partner violence); and about the usefulness of, gaps in and needed adaptations to the screening policy and procedures. The discussion ended with needs and desires for future training or professional development in the area of intimate partner violence and sexual and reproductive health.

Analysis

Chi-square analyses were performed to test for differences between licensed and unlicensed health care staff in job characteristics. Pearson two-tailed correlations among training, attitudes, barriers and preparedness were assessed to identify bivariate associations. Multiple analysis of variance was used to test mean differences between licensed and unlicensed providers simultaneously across five linearly related dependent variables: attitudes about conducting screening, perceptions of the helpfulness of written and verbal screening, perceived resource barriers and preparedness. Multiple analysis of variance provides an advantage over separate t tests because it accounts for the correlations between the dependent variables and decreases type 1 error.³⁸ We then compared these results with qualitative data from the focus groups.

We manually doubled-coded all qualitative data to increase reliability of thematic identification using a grounded theory approach to develop theoretically derived categories and themes emerging from the participants as outlined in Strauss and Corbin.³² We created theoretical categories to complement the quantitative data on barriers, attitudes and preparedness (level 1 coding), and then identified themes through the constant comparative method (level 2 coding). Finally, we classified the themes into broader categories (level 3 coding).

RESULTS

Sample Characteristics

Of the 64 family planning center staff who completed surveys, 56% were unlicensed health care assistants and 44% licensed practitioners. In both groups, about half of participants worked at the Manhattan center, and the rest were about equally divided between staff from the Bronx and Brooklyn. Fifty-three percent of unlicensed providers had been working in family planning care for less than five years, 25% for 5–10 years and 22% for longer; among licensed providers, the proportions were 29%, 32% and 39%, respectively. A significantly greater proportion of licensed professionals than of unlicensed practitioners had received formal training on intimate partner violence (54% vs. 46%).

Quantitative Findings

In general, staff attitudes and barriers were moderately intercorrelated (Table 1). Number of years working in family planning is positively associated with having received training on partner violence, but is not associated with any other variables; thus, it appears that years worked may not, in and of itself, create positive attitudes toward screening. However, receipt of training is positively (although marginally) associated with perceived helpfulness of written and verbal screening. It is also associated with having less-negative attitudes about screening, perceiving relatively few barriers to screening and feeling prepared to discuss partner violence with clients. Further, the more prepared providers feel, the less likely they are to have negative attitudes toward screening and to perceive barriers.

Results of the multiple analysis of variance (Table 2) revealed no differences between licensed and unlicensed practitioners in perceptions of the helpfulness of written and verbal screening; both groups rated these services as helpful to very helpful (means, 4.5–4.7 on the scale of 1–5). Unlicensed health care staff had more negative attitudes than did licensed professionals; however, in both groups, levels of negative attitudes were fairly low (2.6 and 1.9, respectively) and varied little. Licensed professionals reported marginally lower levels of resource-related barriers than did unlicensed staff (2.7 vs. 3.1) and rated themselves as more prepared to conduct screening and follow-up than did unlicensed providers (3.5 vs. 2.7).

Qualitative Findings

•Attitudes. In the focus groups, four themes emerged related to attitudes about partner violence screening (Table 3). The first two were expressed only by unlicensed assistants: Screening is important, but should take a backseat to immediate family planning concerns; and screening for current violence is more important than screening for past violence. Reflecting the first theme, one p-articipant commented, "Violence screening is okay if there’s time, but it’s more important to ask about [STDs] and other medical problems during their visit. This is what they came here for." The second theme seemed to be driven by a lack of knowledge of potential long-term effects of partner violence. For example, one staff member said, "Why should we ask clients about something that’s already over? We can’t do anything about it, and they usually tell us, ‘It’s over; I don’t want to talk about it.’"

The third theme related to attitudes emerged among both licensed and unlicensed staff. Providers expressed frustration over clients’ unwillingness to utilize referrals to the agency social worker or other services after disclosing violence, as well as a sense that the assistance they are able to give is inadequate. For example, one licensed professional related, "I can talk to them about their health care [after a disclosure], but they need to see someone who can help them more than that. When they refuse to see a social worker, it’s frustrating because they need more help than I can give them." Another licensed provider said, "I offer [to set them up with] social work services, but when they refuse, I just want to shake them, because I can’t help them." And as providers discussed their sense that their responses to violence disclosure are inadequate to "fix" the problem, a health care assistant said, "We can ask lots of questions, but what matters is what we are going to do about it."

The fourth theme in the attitudes category was also discussed by both provider types, who related that verbal screening and follow-up are more appropriate for licensed professionals than for unlicensed providers to do. Licensed providers considered screening and assessment of related health care needs part of their job responsibility, and unlicensed staff reported ambivalence about whether they had the time, skills or duty to provide these services.

•Resource barriers. In relation to resource barriers, two themes emerged. First, both licensed and unlicensed providers stated that they do not have enough time to effectively address all aspects of patients’ needs. Yet, participants had differing views on the importance of screening for partner violence, regardless of the time it takes. Some providers felt that screening is just as important to spend time on as other aspects of health care, while others thought that it should be done only if time permits after other needs are addressed. An unlicensed staff member said, "We are not in a room long enough to build rapport." A licensed professional related, "I ask clients, ‘How’s your relationship going?’ If the answer is ‘fine,’ I move on, because I don’t have time to ask anything else."

Second, providers expressed frustration that they do not always have immediate access to social work services within and outside of the organization. One licensed professional remarked, "We need to know there is immediate access to a social worker in case of an [emotional] crisis [that might result from a disclosure]." In addition, some participants voiced concern over not having enough written materials about community services to give to clients. A licensed professional stated, "Referrals need to be in the form of packets and palm cards, and need to be replenished in all clinic offices so that all staff have easy access to give them to clients." At the same time, social workers expressed frustration that some providers do not routinely refer clients who have reported partner violence. The focus group discussions revealed some lack of clarity about what providers should say to a client when making a referral, and which situations were appropriate for a referral and which could be handled in a discussion of family planning needs. One social worker stated, "Sometimes I hear about clients who reported serious violence, but I did not speak with them, while others are referred to me when they only wanted information about birth control options that the health care provider could have given them."

•Preparedness. Many providers, especially unlicensed staff, felt they do not adequately understand the connection between partner violence and reproductive health outcomes, and cannot explain to patients the importance of screening. Some staff, mostly unlicensed assistants, said they do not know what to say if a woman discloses violence or how to comfortably offer a referral to an agency social worker. One health care assistant said, "I want to know exactly what we are allowed, and not allowed, to say to a client about [partner violence]." Licensed professionals also related some lack of preparation in these areas; for example, a clinician related, "I need some helpful scripts or specific sentences to say to patients, because I don’t know what to say when they tell me they are being abused." Another clinician commented, "I don’t want to scare or upset a client by asking her about abuse." Licensed professionals also commented on their lack of knowledge about how to document details of discussions about partner violence (e.g., risk assessment and safety planning) and any related symptoms or injuries. They were uncertain about how this type of documentation can be used to help the client at future health care visits or for legal purposes, if the client ever needs documentation of injuries.

•Recommendations. Both types of providers recommended changes for future practice. All staff discussed the need for more educational materials about the connection between relationship factors and sexual and reproductive health to be made available to clients in the waiting area. They felt this would "prime" clients about why they are asked about violence on the written medical history form and verbally by a provider. All groups discussed a desire for more available referral materials, such as discreet brochures and small cards, to give to clients. Participants recommended that written and verbal screening be conducted with every client; both provider types believed that the agency protocol should change to specify that licensed providers conduct the verbal screening and unlicensed staff just review the medical history form to make sure that the patient has completed it. They requested that the organizational policy and protocol be changed to reflect this. For example, one health care assistant said, "Only one person should do the verbal screen so [patients] don’t have to be asked several times." And a licensed practitioner stated, "We need a new procedure for who does the verbal screen. It should be the clinician."

DISCUSSION

Overall, providers expressed positive attitudes about screening for intimate partner violence and considered it an important part of comprehensive health care. Negative attitudes were more nuanced and were mostly related to provider perceptions that clients were not responsive enough to advice or referrals that providers offered after a disclosure of partner violence. This finding is consistent with the expectation, commonly documented in the literature on intimate partner violence, that an individual who discloses an experience of violence will, or should, take actions that the listener deems appropriate.^4,17,20 Health care providers who are accustomed to administering specified and immediate treatment may find the inability to remedy partner violence particularly frustrating, as partner violence often occurs over a longer term than health problems, and women may have less control over it than they do over health issues.³⁹ Our focus group participants also expressed that partner violence screening was more appropriate for licensed providers to conduct than for unlicensed providers. It is unclear to what extent this is a function of educational and professional training or job demands.

Underlying some negative attitudes may also be a lack of knowledge about the association between intimate partner violence—past and present—and sexual and reproductive health. Adequate training about the relationship of violence to sexual behaviors and contraceptive use may help improve providers’ attitudes toward screening. Including both the short-term and the long-term effects of violence could increase their understanding of the importance of screening for both women’s present and past experiences, and emphasizing the relevance of violence to family planning counseling may strengthen the importance of this issue for both licensed and unlicensed providers. Moreover, providers need help developing their skills for responding to disclosures, including training on what to assess, how to counsel about the impact of violence on sexual and reproductive health, and how to make a referral.

Strengths and Limitations

To our knowledge, this is the first mixed-methods study of family planning providers’ perspectives on intimate partner violence screening. Concurrent validity between qualitative and quantitative measures was good; the themes that emerged in the focus groups were very consistent with survey responses. However, the qualitative data illuminated some issues not captured in the surveys—for example, providers’ need for role clarification. Findings support those of prior research in other health care settings regarding the association between training and attitudes about screening,^4,20 and participants themselves offered some recommendations for future intimate partner violence screening practice.

The varying sizes of the focus groups may have affected the diversity of opinions. Additionally, participants all worked for the same organization, and the discussions were held during working hours; as a result, social desirability may have played a greater role than it would have in another setting. However, we did not find more negative attitudes or greater perceived barriers reported on the anonymous surveys than in the focus groups. Furthermore, although participants within groups were not in complete agreement, opinions were more similar than different, and themes reflected ideas expressed by several participants.

Focus groups were brief, and having enough time to discuss a full range of issues was a concern. However, discussions were not cut short, and all groups answered each question on the focus group guide. The same participants had completed a survey prior to the focus group discussion, so having already expressed their opinion in that format, they may have been less compelled to do so in a group discussion; alternatively, the survey may have primed them for the questions in the focus group.

Recommendations

We recommend training all family planning service providers every few years on screening for and following up on disclosures of intimate partner violence. Training should include a review of the latest research findings about the potential effects of partner violence on a variety of reproductive health outcomes, and should make providers aware that survivors may need time and ongoing, consistent support, information and resources before they can take steps to address the problem. Furthermore, to help providers enhance their communication skills, training might include scripts and role-playing scenarios related to asking about partner violence and responding to disclosures. It also could include clarification of job responsibilities and of how and when to make a referral, rather than addressing clients’ needs through health care counseling alone. Follow-up protocols and risk reduction services—such as counseling about birth control and periodic testing for STDs, including HIV—may be necessary for clinical settings.