Patient-centred communication and provider avoidance: Does body mass index modify the relationship?

Data source

Data come from the Health Information National Trends Survey (HINTS) 3. Data for HINTS 3 were collected between January 2008 and April 2008 by computer-assisted telephone interview and by mail. The final sample for HINTS 3 (n = 7,674) is representative of the population of the USA aged 18 years and over when data are weighted. The survey’s methods are described in detail elsewhere (Cantor et al., 2009). We used HINTS 3 rather than more recent iterations of HINTS because it included our dependent variable of interest (i.e. provider avoidance), as well as measures of patient-centred communication, BMI and covariates.

Study variables

Provider avoidance was measured using the following question: ‘Some people avoid visiting their doctor even when they suspect they should. Would you say this is true for you, or not true for you?’ Response categories included (1) ‘true’ or (2) ‘not true’.

The measure of patient-centred communication is based on the six core functions of patient-centred communication identified by Epstein and Street (2007). The patient-centred communication score is an average of responses to six items, and this measure has been used in previous research (e.g. Blanch-Hartigan et al., 2015; Underhill and Kiviniemi, 2012). Participants were asked, within the past 12 months, how often did doctors, nurses or other healthcare professionals (1) ‘give you the chance to ask all the health-related questions that you had?’, (2) ‘give the attention you needed to your feelings and emotions?’, (3) ‘involve you in decisions about your healthcare as much as you wanted?’, (4) ‘make sure you understood the things you needed to do to take care of your health?’ and (5) ‘help you deal with feelings of uncertainty about your health or healthcare?’ Participants were then asked how often, over the past 12 months, they felt they could rely on doctors, nurses or other health professionals to take care of their healthcare needs. Response categories for all six items were (1) ‘never’, (2) ‘sometimes’, (3) ‘usually’ and (4) ‘always’. Respondents who were missing responses to three or more of the individual questions were treated as missing for the scale score. The scale had good reliability (Cronbach’s α = .89), consistent with previous research (e.g. Blanch-Hartigan et al., 2015; Finney Rutten et al., 2015; Underhill and Kiviniemi, 2012). Higher scale score indicated better patient-centred communication.

BMI was calculated from self-reported height and weight. We then classified respondents into standard BMI categories (Expert Panel on the Identification, Evaluation, and Treatment of Overweight in Adults, 1998) as follows: (1) <18.5 as ‘underweight’, (2) 18.5–24.9 as ‘normal weight’, (3) 25.0–29.9 as ‘overweight’ and (4) >30.0 and above as ‘obese’. Due to a small sample size for the underweight category (n = 130), participants in that group were excluded from analysis.

Potential covariates included age, sex, race/ethnicity, education level, working status, marital status, perceived health status, health insurance status, having a regular healthcare provider and frequency of provider visits in the last year.

Data analysis

All analyses were conducted in Stata version 12.1. Data were weighted using sampling weights provided by HINTS. For all variables, ‘don’t know’ and ‘refused’ responses were coded as missing.

We first examined bivariate associations. We estimated the weighted means of patient-centred communication and tested for differences between BMI categories. We used cross-tabulations to assess the statistical significance of associations between BMI categories and all other variables. To assess effect modification of BMI on the relationship between patient-centred communication and provider avoidance, we estimated multiple logistic regression models stratified by BMI category.

Limitations

This study also has limitations that may have affected the results. Specifically, it has a cross-sectional design, relies on self-report data and may be confounded by unmeasured variables. For example, the measure of patient-centred communication asks respondents to reflect on the prior 12 months, the measure of provider avoidance does not pertain to a specific period and the measure of BMI is based on self-reported height and weight at the time of data collection. Because people tend to underreport their weight (Kuczmarksi et al., 2001), some overweight and obese respondents may be misclassified in lower weight categories, which would lessen the observed effect of BMI on the communication–avoidance relationship. Studies that can provide better evidence of the temporal sequence of the variables and include the collection of weight and height data by research staff may produce different results. In addition, due to the small size of obese subgroups, we were not able to stratify by obesity categories (e.g. Obese I, II and III).

Another limitation of the measurement is that the questions used to measure patient-centred communication and provider avoidance did not pertain to specific providers or types of visits; furthermore, the measure of provider avoidance was binary and broad. Certain healthcare services or types of procedures may be more threatening or uncomfortable for overweight or obese patients, especially if excess body weight is relevant to the service or procedure. For example, previous studies linking weight status with avoidance of care have focused on breast and cervical cancer screenings (Amy et al., 2006, Wee et al., 2000). Future studies that use more specific measures are needed to understand the relationships between these variables in specific healthcare contexts and better inform interventions. Finally, although HINTS data, when weighted, are representative of the US population aged 18 years and over, our findings may not be generalisable to persons who have not recently visited their provider. Any participants who did not visit the doctor in the previous 12 months were excluded from our models because they were not asked to respond to the patient–provider communication questions.

Our findings highlight the importance of patient-centred communication in patients’ healthcare experiences, regardless of patient characteristics such as BMI. For all BMI subgroups, those who reported greater patient-centred communication had lower odds of provider avoidance. These findings add to the body of evidence indicating that patient-centred communication may have important implications for use of healthcare services (Faith et al., 2015; Kannan and Veazie, 2014; Moore et al., 2004), as well as outcomes related to patient satisfaction (Rathert et al., 2013) and patient-reported well-being and quality of life (Epstein and Street, 2007; Robinson et al., 2013). In conclusion, our results suggest that patient-centred communication and healthcare seeking are linked for diverse groups, even those who may experience weight stigma and discrimination in healthcare. Further research is needed to better understand the role of patients’ weight in their healthcare experiences, the quality of patient–provider interactions and healthcare use.