 |
|
Evaluating Interviewing Techniques for the Sexual Practices History
Use of Video Trigger Tapes to Assess Patient Comfort
Michael Floyd, EdD;
Forrest Lang, MD;
Kathleen L. B. Beine, MD;
Elizabeth McCord, MD
Arch Fam Med. 1999;8:218-223.
ABSTRACT
| |
Background Although physicians recognize the importance of assessing a patient's risk for human immunodeficiency virus (HIV) infection, many are reluctant to explore a person's sexual history, a principal determinant of this risk.
Objective To examine the feasibility of a research design that uses responses to viewing a videotaped interview as a proxy for how patients might feel if they were interviewed using a specific approach, and to determine comfort levels with a variety of interview approaches for conducting HIV risk assessment.
Methods Individuals responded to a videotape of several interviewing approaches for HIV risk assessment. Responses to the following aspects were collected: introduction of HIV risk assessment, assessment of patient comfort with the topic, techniques for collecting HIV risk and sexual information, and exploration of sensitive issues not previously identified. Participants expressed levels of comfort by means of a Likert scale to rate their comfort with each approach.
Results Participants expressed higher comfort levels with an introduction that used a ubiquity statement or lifestyle bridge question. Also, they expressed greater comfort when the interviewer addressed how they felt about responding to questions about their HIV risk. Participants reported highest levels of comfort with both patient-centered and closed-ended interviewing techniques. Women were less comfortable with an open-ended interviewing technique.
Conclusions Individuals have different comfort levels for approaches used to assess HIV risk. This trigger tape experimental design to assess comfort levels for interview techniques is feasible for exploring other aspects of physician-patient communication; additional validation studies are recommended.
INTRODUCTION
THERE IS GENERAL agreement on the need for human immunodeficiency virus (HIV) risk assessment in primary care,1-2 with a growing consensus of what content should be identified during such a screening.3-6 However, there has been less discussion regarding effectiveness of interview approaches for eliciting this information.7-10 Much of the research on physician-patient communication involves interactional analyses that separate verbal utterances into broad categories, eg, questioning, providing information, providing support, etc.10 Such studies do not provide information on particular interview techniques in specific interviewing situations. Little is known about comfort levels concerning specific interview techniques from the perspective of the patient as a consumer of medical services. As a sensitive communication situation, an interview assessing HIV risk is useful for studying the comfort levels of health care consumers with various interview approaches and, thus, may be an indicator of preferences patients may have for these approaches.
Although the most important factors in the acquired immunodeficiency syndrome (AIDS) epidemic are related to patients' lifestyles, sexual habits, and risks, physicians frequently do not ask patients for a sexual history.11-13 Proportions of patients who were asked or physicians who routinely requested this information ranged from 7% to about 50%.14-15 A study using standardized patients found that primary care physicians initiated HIV screening in 60% of interviews, but only 49% of risk behaviors were identified.16 There are many reasons physicians do not elicit a sexual history, including a physician's own discomfort,17-19 as well as perceptions of potential or real patient discomfort.11 There has been disagreement among communication experts over how best to elicit a sexual history in a way that is least traumatic for patients, yet efficiently obtains accurate information.20 However, there is a need for HIV risk information, as well as a demand from patients and managed care systems to address the issue.21-22 Knowledge about patients' comfort levels when being interviewed about this sensitive area should allow physicians greater comfort and proficiency in carrying out this task.
In personal discussions with other physicians who study physician-patient communication, we solicited recommendations on effective approaches to the HIV risk assessment and found no consensus. Considerable disagreement occurred around the following communications questions:
- What is an appropriate introduction into HIV risk assessment? One opinion was that any introduction to the subject would raise patient concern by implying that eliciting sexual information required special attention. Another suggestion was to introduce the topic within a discussion of prevention and health promotion. Others recommended a ubiquity statement to introduce the subject of HIV risk assessment, ie, a statement indicating that such questions are routinely asked of all patients.23
- Should the clinician explicitly assess the patient's feelings regarding responding to questions about sexual activity? About this question, some experts believed that, by raising the question of a person's comfort, one might cue the patient to feel uncomfortable about a topic that might otherwise flow naturally between clinician and patient. Others believed that exploring the comfort or discomfort of the patient is respectful and appropriate.
- Which questioning style is most effective in assessing HIV risk? Communications colleagues frequently recommend the use of open-ended questions to collect information. However, most published guidelines addressing HIV risk assessment identify specific information and list a series of closed-ended questions to obtain this information.10 Still others advocate collecting HIV risk assessment information with a patient-centered approach that emphasizes eliciting the patient's own ideas and concerns in the clinical dialogue.24 But what does it mean to be patient-centered while conducting HIV risk assessment? How might patients react to efforts to explore their ideas and concerns on a topic that may make them feel uncomfortable?
Much of the disagreement about how to conduct the HIV risk assessment rests on potential patient reactions to different interview techniques. Consequently, we evaluated how individuals respond to different options, comparing introductions to the interview, the issue of addressing the patient's feelings, and 3 questioning styles: patient-centered, open-ended, and closed-ended approaches.
A second purpose of this pilot study was to assess the feasibility of using trigger tapes to evaluate different interview techniques. We hoped this new research method could provide objective data on the desirability of using one communication approach over another. A review of the literature yielded only one study using a similar approach. The study by Gerbert et al25 of patients' responses to an interactive multimedia HIV risk assessment demonstrated a higher disclosure of risky behaviors when the interview was conducted in a "warm" communication style.
PARTICIPANTS, MATERIALS, AND METHODS
PARTICIPANTS
Fifty-seven individuals participated in this pilot study, including volunteers from a senior citizens' center (n=10), an undergraduate health education class (n=32), graduate students in health education (n=10), and a group of volunteer undergraduate college students (n=5). Their ages ranged from 20 to 75 years. Nineteen participants were men and 37 were women (this information was missing for 1). Of the 56 participants who reported their race, 51 (91%) were white, 3 (5%) were black, 1 was Hispanic, and 1 was Asian. All but 1 participant had attended college, with 29 participants (52%) reporting some college and 26 (46%) college graduates.
MATERIALS
Materials included a demographic data form, an interview comfort rating form, and videotaped vignettes. The following information was requested on the demographic data form: age, sex, race, educational attainment, a self-report of strength of religious faith, religious preference, date of last sexual contact, level of concern regarding issues related to sexual activity, and level of comfort answering physicians' questions regarding sexuality.
The Interview Comfort Scale was a self-report measure that used the following categories: (1) very uncomfortable, (2) somewhat uncomfortable, (3) somewhat comfortable, and (4) very comfortable. Participants also wrote comments on the form regarding interview techniques.
Videotaped examples of a variety of interview approaches were developed, evaluated by an independent panel of consultants in the clinical interviewing field, and determined to represent 1 of several recommended interview techniques. On the videotape, a male physician demonstrated the different techniques for each of 5 areas selected for study within the HIV risk assessment interview: introduction to sexual history; assessing comfort; style of collecting information: sexual practices; style of collecting information: identification of risk for AIDS; and exploring sensitive issues not previously identified. The sequence of presentation of the different interview approaches was varied to balance the potential effect of primacy or recency (Table 1).
|
|
|
|
Videotaped Interview Techniques for HIV Risk Assessment*
|
|
|
PROCEDURE
After the purpose of the study was explained, questions about the study were answered, informed consents to participate were obtained, demographic data forms were completed, and participants were provided the Interview Comfort Scale. The videotaped vignettes were presented to groups of participants as sequenced in Table 1. Within each HIV risk assessment area, all of the alternative techniques were shown initially to participants in quick succession. Participants then viewed the videotape a second time; the videotape was paused after each technique, and participants assessed comfort with the technique by means of the Interview Comfort Scale. This process was repeated until all 5 study areas were evaluated. After completing the videotape, participants were provided an opportunity to write additional comments and explore further questions during a group discussion.
STATISTICAL ANALYSIS
A Friedman analysis of variance (ANOVA) was used to measure the strength of associations between comfort levels expressed for the various interview techniques. Where the strength of these associations was significant at  <.05, a post hoc Wilcoxon matched-pairs signed rank test was used to differentiate among groups. Kruskal-Wallis ANOVAs, Mann-Whitney test, and Friedman ANOVAs were used to evaluate the associations among and within demographic groups. In all cases where multiple intergroup comparisons were made after a Friedman or Kruskal-Wallis ANOVA, an adjustment (with Bonferroni P value correction) was made to the resulting P values to compensate for the added probability that we would erroneously find a statistically significant result purely by chance.
RESULTS
As seen in Figure 1, participants expressed different comfort levels for certain techniques in HIV risk assessment. With introduction to sexual history, participants expressed higher comfort levels with the ubiquity statement and lifestyle bridge over the direct statement (P<.001). An example of this is illustrated by the written comment, " . . . a good intro made me feel more at ease; helps to know that everyone is asked the same questions." However, differences between the ubiquity statement and the lifestyle bridge were not significant (P=.10).
|
|
|
|
Figure 1. Mean comfort levels for interview methods (N=57). AIDS indicates acquired immunodeficiency syndrome. The comfort level was graded on a 4-point Likert scale, where 1 indicates very uncomfortable; 4, very comfortable.
|
|
|
In assessing feelings, an interview technique that addressed patients' feelings before beginning the sexual history was rated more comfortable than one that did not (P<.001). Written comments such as "Feelings can affect the truth about activities, so they are important to discuss first" reveal the value attached to this issue.
Participants were significantly more comfortable with the patient-centered interview technique than the open-ended technique for both styles of information collection: sexual practices (Wilcoxon matched pairs post hoc, P<.001) and identification of risk for AIDS. Additionally, while participants were more comfortable with the closed-ended technique than the open-ended technique in both of these areas, differences were statistically significant only within the sexual practices area (P<.001) and not in identification of risk for AIDS. There was no statistically significant difference between patient-centered and closed-ended techniques for either of these 2 areas. Participants' comments recognized the benefits and limitations of these different techniques. For example, one participant wrote that while the open-ended technique "gives the chance to tell [the sexual history] exactly, others might accidentally leave out something that might be crucial." Another concern was raised about the closed-ended technique: "[It] may exclude some things the doctor could need to know by only targeting [a very specific piece of information]."
The last area, exploring sensitive issues not previously identified, included a question about anal intercourse. An open-ended interview approach was rated as more comfortable than the closed-ended direct question (P<.001). One participant stated, "I think it is rude to assume that ‘most' people have had anal intercourse. I certainly have not and would be extremely offended if someone asked me if I had. Once you offend your client, you can forget open communication thereafter." Other participants indicated the term was "too forward" or "too blunt."
SEX DIFFERENCES
Significant differences in comfort levels based on consumer sex were found in introduction to the sexual history (Figure 2). Both men and women preferred lifestyle and ubiquity approaches over a direct statement (Wilcoxon matched pairs post hoc, P<.01). However, men's comfort levels were significantly higher than women's levels when the direct statement method was used (Mann-Whitney, P=.009).
|
|
|
|
Figure 2. Mean comfort levels for interview methods based on sex (N=56), graded on a 4-point Likert scale, where 1 indicates very uncomfortable; 4, very comfortable.
|
|
|
Significant difference based on sex was also found in style of information: sexual practices, where women reported significantly greater comfort with patient-centered and closed-ended techniques than with the open-ended technique (Friedman ANOVA, P<.001; Wilcoxon matched pairs post hoc, P<.001). Men expressed higher comfort ratings with the patient-centered and open-ended approaches than with the closed-ended approach (Friedman ANOVA, P<.04). Comfort ratings for men were significantly higher than those for women with the open-ended technique (Mann-Whitney, P=.004).
REPORTED LEVELS OF PARTICIPANT COMFORT DISCUSSING SEXUALITY
Participants were evaluated on their a priori level of comfort with discussion of sexual matters, as indicated on the demographic data form. Forty-five participants reported feeling either very comfortable or somewhat comfortable about answering physicians' questions regarding sexuality on the demographic data form. Eleven participants were uncomfortable, reporting feeling either somewhat uncomfortable or very uncomfortable. However, during the videotape assessment, both comfortable and uncomfortable groups preferred the same interview techniques for each HIV risk assessment area.
OTHER DEMOGRAPHIC VARIABLES
Statistical analysis of other demographic variables in relation to comfort ratings was not done because of the small sample size and composition.
COMMENT
Typically, efforts to develop more effective methods of medical interviewing have been based on expert opinions and on communication principles derived from pooled interactive analysis data.26 Rarely have patient preferences been solicited or evaluated. Our study demonstrated that individuals expressed significantly different levels of comfort for interview techniques used to obtain their sexual history in all 5 content areas of the HIV risk assessment. This study points to the need for clinicians to acquire an array of interviewing techniques and use them at appropriate points in the interview. A knowledge of patient comfort should prove valuable in helping clinicians pursue HIV risk assessment with confidence and effectiveness.
When introducing the topic of sexual history during an interview, some authors have found no difference in disclosure of risky behaviors when using a "preamble."25 However, in our study, participants expressed greater comfort with some form of transition. While there was a trend favoring the ubiquity statement over the lifestyle bridge, this difference was not found to be statistically significant. Based on these findings, clinicians should introduce the topic before asking the sexual history and consider using a ubiquity statement rather than simply initiating questioning or announcing a shift in the focus of the interview.
There are numerous benefits in addressing patients' feelings when exploring sexual questions. Such an inquiry allows an opportunity for the patient to verbally express reservations. There are times when the level of discomfort needs to be addressed before proceeding, for example, when a patient expresses discomfort nonverbally. Otherwise, proceeding without addressing major reservations may jeopardize the validity of the data as well as the relationship.
Generalizations based on this pilot study are limited because of sampling procedures and sample size resulting in unintended selection bias. A much larger population is needed to identify potential differences among various subgroups. Several different methods were used to recruit study participants, including flyers announcing the project to a random sample of university students along with a monetary incentive. Despite these efforts, less than 1% of those recruited responded. Similarly, more than 100 senior center adults were invited, but only 10 (each with some college education) volunteered. There was a nearly 2:1 female to male ratio; consequently, the identified differences between sexes may have been underestimated. Future studies should use a random sampling of a more representative population of clinic patients, including individuals who are both at high and at low risk for HIV exposure.
Other possible challenges to the study's validity include the asking of respondents to suspend disbelief. The vignettes are brief segments presented by an interviewer unknown to the participant, as opposed to their own caregiver in the context of health maintenance or a health-related problem. An existing clinical relationship may entail other factors that could enhance or diminish emotional response, comfort, and preferences. Participants appeared to be at relatively low risk for HIV. It is not known if participants' self-assessment of personal risk for HIV would change their responses to the portrayed interviewing techniques.
With these limitations in mind, several findings become apparent. Patient comfort with a transition into the interview and with having feelings addressed have already been mentioned. A further important observation is the uniformly high level of relative comfort for patient-centered communications, among both men and women. Patient-centered communication begins with an attempt to understand what the patient knows, believes, feels, or values, and then engages in a dialogue providing a context for any patient-specific information that may follow and may diminish potential tensions for both parties.24 This approach may encourage broad, rich responses and may begin the groundwork for developing and negotiating a treatment plan that empowers the patient as an active participant.
Another finding in this study was that women felt more comfortable with patient-centered or closed-ended questions. In contrast, men expressed greater comfort with open-ended and patient-centered questions. Several factors may have contributed to these sex-related differences. First, the trigger tape depicted a male interviewer. For some women, the sex of the interviewer may have raised fears of sexual intrusiveness. Alternatively, men may culturally be more comfortable in discussing their sexual history. We are in the process of exploring whether the sex of the interviewer affects the comfort level of male and female patients.
Accepted communications recommendations encourage the use of open-ended questions, since closed-ended questions may restrict the range of response. However, in this study women expressed greater comfort levels with closed-ended questions exploring the sexual history. This supports Williams'10 observation that patients often do not volunteer sexual information when asked general, open-ended questions, thus necessitating beginning with direct questions. This is in contrast to the findings that men in our study preferred open-ended and patient-centered questions.
Given women's greater comfort for closed-ended questions, why did both male and female participants rate comfort levels high with the open-ended technique in the area that asked for a history of anal intercourse? The answer seems to be linked with the highly charged term anal intercourse, which appeared offensive. These findings suggest that certain words or terms carry negatively charged emotional responses that can interfere with the interview. It should be noted that the open-ended approach to questioning for extravaginal sexual practices asks, "Other than vaginal intercourse, what other forms of sexual activities have you been involved with?" Reliance on this approach to collect valid data assumes that the physician will be comfortable with the response "no other types." In some situations, however, the clinician may still feel the need to explicitly ask about specific sensitive subjects, such as anal intercourse. Conversely, this study assumes that reported levels of comfort are related to preferred interviewing techniques, but this may not always be the case. For example, someone may be more uncomfortable being asked directly about anal sex but at the same time prefer that it be asked directly rather than indirectly.
Because this was a pilot study, we believe this research approach would be strengthened through additional studies validating this method by comparing patient responses to video review with their responses to the same approaches in a live interview. Another limitation of this study is that research participants were presented with only one interviewer, a man. Future studies should use a variety of interviewers, carefully matched in their presentations in all ways except for their sex, race, age, or other meaningful variables. Also, the issue of confidentiality was not addressed in this study; it should be evaluated.
It is time, in research of physician-patient communication, that we listen directly to our patients for guidance in how best to communicate with them. This research method may be helpful in defining certain aspects of the acceptability of various communication techniques and phrasing options. It may also be useful in researching general strategies for clinical interviewing as well as approaches to specific sensitive medical or social topics. Other questions that may be addressed through future research using this method include the use of silence, preference for delivering "bad news," discussing advance directives, and screening for alcoholism, substance abuse, and domestic violence.
AUTHOR INFORMATION
Accepted for publication April 30, 1998.
We acknowledge Patricia Buck, MA, who provided editorial assistance; Carol Abell, BS, who produced the video stimulus material and assisted with data entry and analysis; and Tony Narkawicz, MS, who assisted with the statistical analyses.
Corresponding author: Michael Floyd, EdD, Johnson City Family Medicine Associates, 917 W Walnut St, Johnson City, TN 37614 (e-mail: floyd{at}etsu.edu).
From the Department of Family Medicine, James H. Quillen College of Medicine, East Tennessee State University, Johnson City.
REFERENCES
| |
1. Health and Public Policy Committee, American College of Physicians, Infectious Diseases Society of America. Acquired immunodeficiency syndrome. Ann Intern Med. 1986;104:575-581.
2. Jewell ME, Jewell GS. How to assess the risk of HIV exposure. Am Fam Phys. 1989;40:153-161.
ISI
| PUBMED
3. Antoniskis D, Sattler FR, Leedom JM. Importance of assessing risk behavior for AIDS: why and how to obtain a relevant history. Postgrad Med. 1988;3:138-152.
4. Gabel LL, Pearsol JA. Taking an effective sexual and drug history: a first step in HIV/AIDS prevention. J Fam Pract. 1993;37:185-187.
PUBMED
5. American Medical Association. HIV Early Intervention: Physician Guidelines. 2nd ed. Chicago, Ill: American Medical Association; 1994.
6. Hearst N. AIDS risk assessment in primary care. J Am Board Fam Pract. 1994;7:44-48.
7. Locke SE, Kowaloff HB, Hoff RG, et al. Computer-based interview for screening blood donors for risk of HIV transmission. JAMA. 1992;268:1301-1305.
FREE FULL TEXT
8. Fleisher JM, Minkoff HL, Senie RT, Endias RE. Assessing prior history of sexually transmitted disease [letter]. JAMA. 1991;266:1646.
FREE FULL TEXT
9. Kripke CC, Vaias L, Elliott A. The importance of taking a sensitive sexual history [Clinical Guidelines]. JAMA. 1994;271:713.
FULL TEXT
| PUBMED
10. Williams S. The sexual history. In: Lipkin M, Putnam SM, Lazarre A, eds. The Medical Interview: Clinical Care, Education, and Research. New York, NY: Springer-Verlag NY Inc; 1995:235-250.
11. Merrill JM, Laux LF, Thornby JL. Why doctors have difficulty with sex histories. South Med J. 1990;83:613-617.
FULL TEXT
|
ISI
| PUBMED
12. Ross PE, Landis SE. Development and evaluation of a sexual history-taking curriculum for first- and second-year family practice residents. Fam Med. 1994;26:293-298.
PUBMED
13. Lewis CE. Sexual practices: are physicians addressing the issues? J Gen Intern Med. 1990;5(5 suppl):S78-S81.
14. McCance KL, Moser R Jr, Smith KR. A survey of physicians' knowledge and application of AIDS prevention capabilities. Am J Prev Med. 1991;7:141-145.
ISI
| PUBMED
15. Gerbert B, Bleecker T, Bernzweig J. Is anybody talking to physicians about acquired immunodeficiency syndrome and sex? a national survey of patients. Arch Fam Med. 1993;2:45-51.
FREE FULL TEXT
16. Wenrich MD, Curtis JR, Carline JD, Paauw DS, Ramsey PG. HIV risk screening in the primary care setting. J Gen Intern Med. 1997;12:107-113.
FULL TEXT
|
ISI
| PUBMED
17. Vollmer SA, Wells KB. How comfortable do first-year medical students expect to be when taking sexual histories? Med Educ. 1988;22:418-425.
ISI
| PUBMED
18. Stevenson RWD, Szasz G, Maurice WL, Miles JE. How to become comfortable talking about sex to your patients. CMAJ. 1983;128:797-800.
ABSTRACT
19. Darr MS, Sinclair AE. HIV disease: educational needs of Oregon family physicians. Fam Med. 1992;24:21-23.
PUBMED
20. Gabel LL, Crane R, Ostrow DC. HIV-related disease: family physicians' multiple opportunities for preventive intervention. J Am Board Fam Pract. 1994;7:218-224.
21. Waitzkin H. Doctor-patient communication: clinical implications of social science research. JAMA. 1984;252:2441-2446.
FREE FULL TEXT
22. Ende J, Rockwell S, Glasgow M. The sexual history in general medicine practice. Arch Intern Med. 1984;144:558-561.
FREE FULL TEXT
23. Johnson MC. Taking sexual histories: now it's essential. Iowa Med. 1988;78:118-120.
PUBMED
24. Stewart M, Brown JB, Weston WW, McWhinney IR, McWilliam CL, Freeman TR. Patient-Centered Medicine: Transforming the Clinical Method. Thousands Oaks, Calif: Sage Publishers; 1995.
25. Gerbert B, Johnston K, Bleeker T, et al. HIV risk assessment: a video doctor seeks patient disclosure. MD Comput. 1997;14:288-294.
PUBMED
26. Bryson-Brockmann WA, Fischbein D. Demonstrating the effectiveness of videotape feedback for teaching interviewing skills: a multiple-baseline, single-participant experimental design. Teach Learn Med. 1995;7:149-154.
THIS ARTICLE HAS BEEN CITED BY OTHER ARTICLES
Patients Prefer the Method of "Tell Back- Collaborative Inquiry" to Assess Understanding of Medical Information
Kemp et al.
J Am Board Fam Med 2008;21:24-30.
ABSTRACT
| FULL TEXT
Use of Videotaped Interventions in Research
Steinke
West J Nurs Res 2001;23:627-643.
ABSTRACT
|
