A direct observation study of provider patient communication about contraception: How does race affect communication, and how does communication affect contraception use?
Awarded 2008
Large Research Grants
Christine Dehlendorf, MD, MAS
University of California, San Francisco

Background: Women in the US have extraordinarily high rates of unintended pregnancy, with poor and minority women disproportionately experiencing this adverse reproductive health outcome. This high rate of unintended pregnancies is caused in part by the under-use and misuse of effective contraceptive methods. As in the US all non-barrier methods of contraception require consultation with a health care provider, clinicians who provide family planning care have an opportunity to positively impact contraceptive use during contraceptive counseling. Little research has investigated this interaction, however. In addition, this interaction has not been investigated from the perspective of health care disparities. Research in other areas indicates that there are differences in the quality of communication between patients and providers, which may contribute to differences in health outcomes. As contraception is a preference-sensitive decision and raises issues of sexuality, we hypothesize that in this context, variations in provider-patient communication by race/ethnicity are pronounced. Methods: We have initiated a cohort study to investigate provider-patient communication about contraception in order to (1) determine if providers discuss contraception with their patients differentially depending on the patients’ race/ethnicity, and (2) determine which aspects of communication impact use of contraception. Patients are recruited at the time of a visit in which they wish to start a contraceptive method. They complete a pre- and post-visit survey, and their visit is audiotaped. We then perform telephone follow-up at 3- and 6-months to determine contraceptive continuation and adherence. The audiotaped visits are analyzed using a modified Four Habits Coding Scale (a measure of interpersonal quality of care) a facilitation score, as a measure of shared decision making, and a content checklist. In order to investigate differences in contraceptive counseling by race/ethnicity, we will determine if these measures differ by patient race/ethnicity. In addition, we will test whether elements of communication, including shared decision making and quality of communication, are associated with improved contraceptive use (method choice, method continuation at 3 and 6 months, and method adherence). We will also analyze the audio recordings for specific counseling behaviors to determine if these factors are associated with improved contraceptive use. Results: To date, 285 patients, out of planned 414, have been recruited at six clinical sites in the San Francisco Bay Area, and we have achieved a 90% follow-up rate at 6 months . We have recruited 114 White patients, 94 African-American patients and 77 Latina patients. As we have received other funding to expand the study to an additional site, due to difficulty recruiting from our original sites, we have not yet begun the formal data analysis phase of the study. Demographics of our patient participants include an average age of 26 years, with 75% having an income of less than 200% of the Federal Poverty Level. 90% of participants had completed high school. Oral contraceptive pills is the most commonly selected method followed by the contraceptive ring, levonorgestrel IUC, contraceptive injection and the copper IUD. With respect to differences by race/ethnicity, a significant difference in communication between Black and White patients was identified using the Four Habits Coding Scale, with Blacks receiving lower quality of interpersonal communication. No other differences in communication were identified in this preliminary analysis, including no difference in likelihood of provider self-disclosure of contraceptive use and no difference in the likelihood of the clinician mentioning specific contraceptive methods. Analyses of predictors of contraceptive continuation found no association between the Four Habits Coding Scale or the facilitation checklist and contraceptive continuation. Overall, 55% of women are continuing their method at 6 months, with the only significant difference by method being that those choosing the ring are less likely to continue (31%). Final analyses will include the association of additional provider behaviors with patient race/ethnicity, correlation between the patient’s assessment of the quality of care with patient race/ethnicity and with contraceptive continuation and adherence, and the association between contraceptive method use and specific counseling behaviors. In these analyses, we will utilize multivariate hierarchal models to control for clustering by provider and clinic and confounding by patient variables. Discussion: While we have experienced significant challenges with recruitment due to a variety of barriers, we have completed 70% of our desired sample size and have achieved a higher than expected follow-up rate. As departmental funding is available to complete recruitment at an additional site, we anticipate having the necessary power to address our two aims following completion of data collection. In addition, during the course of this study we realized the depth of information available in the audiorecorded data, and plan to initiate qualitative analyses of the audiotaped data in order to further investigate counseling about specific contraceptive methods and about specific populations, such as adolescents and older women. The data generated by this study has the potential to expand our understanding of contraceptive counseling and to inform counseling interventions designed to improve contraceptive use.