An Exploration of Factors Promoting Patient Participation in Primary Care Medical Interviews

Abstract

Street's (2003) ecological model of communication in medical encounters was used to select and examine factors that potentially promote or retard patient participation. Patient participation was defined as information seeking and provision, assertive utterances, and emotional expressions. Patient participation discourse scores were used as the dependent variable in a multilevel regression analysis with 19 predictor variables representing cultural, organizational, and interpersonal factors of the ecological model. The analysis revealed eight significant predictors of patient participation. The results were discussed with respect to other research using the ecological model and their implications for continued study of factors that promote or retard patient participation.

ACKNOWLEDGMENTS

Funding for this study was provided by grant R03 HS90110-01T from the former Agency for Health Care Policy and Research (now the Agency for Healthcare Research and Quality) and a grant from the American Academy of Family Physicians Foundation. An earlier version of this paper was presented at the annual meeting of the National Communication Association, 2009.

Notes

¹See Cegala et al. (2007) for a literature-based justification for these components and sample discourse segments illustrating them.

²This data set was gathered initially to assess the effects of a patient communication skills intervention (see Cegala, McClure, et al., 2000).

³The number of information provision units per patient was much larger than the number of discourse units comprising the other components of patient participation. Accordingly, the information provision units across several topics were first averaged, and then the result was summed with the frequencies of the remaining three components of patient participation.

⁴There are advantages to using multilevel modeling over traditional regression analysis even when there are small differences between level-2 units (Hayes, 2006; Kreft & de Leeuw, 1998). Additionally, although there was a nonsignificant physician effect for the complete regression model, individual analyses on predictor variables indicated that several had a significant physician effect.

⁵Other differences were found regarding Asian physicians, but they are less relevant here due to the lack of an Asian physician sample.

⁶It should be noted that several studies show a positive relationship between physician patient-centeredness and patient participation, including the present study.