Health advocacy among resident physicians

Background: Health advocacy is increasingly recognized as an integral part of a physician’s professional role. This survey examines advocacy attitudes, competencies, and participation among psychiatry residents. Methods: Psychiatry residents from one program were surveyed using a 13-item, self-report questionnaire. Fisher’s exact tests were used to examine the association between advocacy attitudes and competencies and past, current, and planned future participation in advocacy. Results: Thirty-two residents responded to the survey. Nineteen percent of respondents endorsed being able to describe what a health advocate does, 91% endorsed that advocacy extends beyond the needs of the individual patient, 44% endorsed that being an advocate is an important part of their professional identity, and 26% endorsed being confident in their ability to advocate. Endorsing advocacy as an important part of one’s professional identity was associated with higher endorsement of future participation in advocacy (38% vs. 9%, p < 0.001). Discussion: Perceived competency in advocacy may lag behind positive attitudes towards advocacy, suggesting a knowledge gap among residents that warrants further curriculum development. Feeling that advocacy was an important part of one’s professional identity was associated with future plans to participate in advocacy, suggesting that fostering positive attitudes to advocacy may help produce future advocates.


Introduction
Health advocacy, defined as healthcare professionals' responsible use of "their expertise and influence to as they work with communities or patient populations to improve health" (Frank et al., 2015, p. 22), is increasingly recognized an integral part of a physician's professional role and a core competency in residency training. The American Medical Association's Declaration of Professional Responsibility has outlined health advocacy as one of the nine basic tenets of physicianhood (American Medical Association, 2001), and the Royal College of Physicians and Surgeons of Canada has officially adopted a physician competency framework that includes "health advocate" as Vance M, Bui E, Shtasel D, Borba C MedEdPublish https://doi.org/10.15694/mep.2019.000029.1 Page | 2 one of seven essential physician roles to be emphasized in medical specialty training (Frank et al., 2015).
With the increasingly acknowledged importance of teaching health advocacy, researchers in medical education have started to examine advocacy attitudes, competencies, and participation in resident physicians (Verma et al., 2005;Leveridge et al., 2007;Stafford et al., 2010;Bachofer et al., 2011;Chamberlain et al., 2013;Long et al., 2014;Shepard et al., 2015;DeCesare and Jackson, 2016). However, there is, to our knowledge, no published data to date examining these attributes in psychiatric residents. This represents a gap in the current state of medical education knowledge and practice, especially given the significant societal and economic toll of mental illness worldwide. These factors speak to an ongoing need to train more psychiatrists and other healthcare professionals in health advocacy. The current study seeks to evaluate health advocacy attitudes, competencies, and participation in psychiatric resident physicians.

Participants and procedures
Participants were resident physicians from one US adult psychiatry residency program. There were 16 residents in the PGY-1, PGY-3, and PGY-4 classes and 17 residents in the PGY-2 class, comprising a total of 65 residents in the entire program who were eligible for the study. They were recruited through residency-wide emails, inviting them to respond to a confidential survey on psychiatry residents' attitudes towards and competencies in health advocacy. Information on each resident's postgraduate year of training was recorded, but in order to preserve confidentiality for the participants, who were providing quality improvement feedback to their residency program, no further demographic information was obtained. This project was undertaken as a quality improvement initiative within a hospital-based residency program, and, as such, the Institutional Review Board declined to review it as per hospital policy.

Measures
The Resident Physician Health Advocacy Questionnaire (R-PHAQ) is a 13-item, self-report measure designed to assess health advocacy attitudes and competencies in resident physicians. It was adapted from the Health Advocacy Questionnaire, a self-report tool for internists developed and validated by Stafford et al. (2010). Items 1-12 of the R-PHAQ are displayed in Table 1. The first 8 items are statements scored on a 5-point Likert scale (ranging from 1=strongly disagree to 5=strongly agree). Thematically, items 1-5 were intended to assess health advocacy competencies, and items 6-8 were intended to assess health advocacy attitudes. Items 9, 11, and 12 are multiplechoice items with "yes/no/unsure or undecided" responses assessing actual or planned participation in advocacy across time. Item 10 elaborates on when previous participation in advocacy occurred and asks respondents to "select all that apply." Finally, Item 13 is a free-text area for additional comments (not shown in Table 1). Before it was administered to residents, this survey was validated for content by an expert in medical education and an expert in health advocacy, who recommended no additional changes.

Data analysis
Survey responses on the Likert scale were dichotomized as "disagree" (1 through 3) and "agree" (4 and 5) for purposes of statistical analysis, since not all items followed a normal distribution. Percentages were then calculated for these dichotomized responses. For items 9, 11, and 12, responses were also dichotomized into "yes" (yes) and "no" (no/unsure or undecided). Fisher's exact tests were used to examine the association between advocacy attitudes and competencies (survey items 1-8) and previous, current, and planned future participation in advocacy. To assess whether dichotomization of the Likert scale responses into "disagree" (1 through 3) and "agree" (4 and 5), instead of separating 3 as a neutral value, had an impact on results, Fisher's exact tests were also performed on responses dichotomized as "disagree" (1 and 2) and "agree" (3 through 5). Since this alternate method of dichotomization did not meaningfully impact the results, the original method of dichotomization was retained in the final analysis. The level of statistical significance was initially set to 0.05 (two-tailed) for all analyses. After a Bonferroni correction was made for the presence of multiple comparisons in a small sample, the final level of statistical significance was set to 0.002 (0.05/24 total hypotheses tested). Analyses were conducted using Stata/IC 14.2 statistical software (Stata, College Station, TX).

Results/Analysis
The R-PHAQ questionnaire was completed by 32 out of 65 residents in the four classes, for a response rate of 49%. There were 6 (19%) respondents from the PGY-1 class, 9 (28%) from the PGY-2 and PGY-4 classes, and 8 (25%) from the PGY-3 class. Table 1 provides the characteristics of respondents, including postgraduate year of training, n (%) of respondents who expressed agreement with the Likert-scale items; n (%) of respondents who answered "yes" to questions about past, current, or planned future participation in advocacy; and when past participation in advocacy (if any) occurred. legend: ^ numbers do not total to 100%, because respondents were instructed to "select all that apply" In the statistical analysis examining the associations between advocacy attitudes and competencies and previous, current, and planned future participation in advocacy, endorsement of advocacy as an important part of one's professional identity had a significant association with planned future participation in advocacy (38% vs. 9%, p < 0.001). No other associations reached statistical significance.

Discussion
This cross-sectional survey in one adult psychiatry residency training program is, to our knowledge, the first study to assess health advocacy attitudes, competencies, and participation among psychiatric residents. We found that a large minority of resident participants (44%) believed that advocacy was an important part of their professional identity, and that the vast majority (91%) agreed that advocacy extends beyond the needs of individual patients. In contrast, fewer residents (26%) expressed confidence in their ability to advocate, and fewer still (19%) endorsed that they could describe the health advocate role. Specifically, residents' perceived competency to participate in advocacy declined as distance from the perspective of the individual patient increased (e.g., 59% stating that they knew how to advocate for individual patients, versus 19% stating that they knew how to advocate for changes in health policy). This suggests a gap between positive attitudes towards advocacy among psychiatric residents and their level of confidence and training to actually perform advocacy-related activities, which is in accordance with previous surveys of residents in other specialties (Leveridge et al., 2007;Stafford et al., 2010) as well as a previous survey of residents and fellows across specialties at two US medical centers (Long et al., 2014). Given the traditional focus in medical training programs on individual patient care rather than on systems and policy issues, this gap in knowledge is not surprising.
Data for participation in health advocacy showed the highest level of participation during medical school and the lowest level during residency (13 versus 3 positive responses). Although it is difficult to draw firm conclusions from this small sample size, it is nevertheless notable that there appears to be a drop-off in advocacy participation between medical school and residency. This finding is consistent with those from a prior study (Stafford et al., 2010) and cannot be fully accounted for, in our study, by a simple difference in the length of time spent in medical school versus in residency training (medical school and psychiatric residency are both four-year programs). Other reasons can be posited for this phenomenon: in Stafford et al.'s survey residents' free-text comments revealed that stress, the need for rest, and insufficient time were common barriers to advocacy participation (Stafford et al., 2010). Based on this information, it is possible that allotting more time during residency specifically for structured advocacy training and participation may help to increase residents' advocacy behaviors.
One significant association emerged between advocacy attitudes and competencies and participation in advocacy. Endorsing the importance of advocacy to one's professional identity was significantly associated with planned future participation in advocacy, suggesting that curricula aimed at fostering positive attitudes towards and increasing the perceived importance of advocacy may increase future advocacy participation.
This study had limitations. The R-PHAQ was adapted from a previously published instrument for ease of use in a project aimed at quality improvement. Although it underwent content validation by an expert in medical education and an expert in health advocacy, it did not undergo other rigorous psychometric testing before being administered. This may limit the ability to compare between these results and other results in the literature. In addition, demographic data other than postgraduate year of training were not collected for the study, in order to preserve the confidentiality of survey participants who were providing feedback to the program. The lack of additional demographic information constrains our ability to determine if survey non-respondents differed significantly from survey respondents. Likert scale responses were dichotomized for statistical analysis due to their non-normal distributions, and dichotomization reduced our study's statistical power. Finally, the survey was administered at only one psychiatry residency program and had a small sample size, which limit the survey's generalizability across psychiatry residency programs and across residency programs in general.

Conclusion
Nevertheless, several notable findings emerged from our survey that accord with the existing literature. These include that 1) training and competency in advocacy lags behind positive attitudes towards and professional identification with advocacy; 2) participation in advocacy drops off during residency; and 3) the perceived importance of advocacy is strongly associated with planned future participation in advocacy. Taken together, these data suggest that residents are interested in learning more about and participating in advocacy, and that curricula geared towards fostering a positive perception of advocacy may help in producing future advocates.

Take Home Messages
In one psychiatry residency program, knowledge of health advocacy lagged behind positive attitudes to health advocacy. Feeling that advocacy was an important part of one's professional identity was significantly associated with future plans to participate in advocacy. This suggests that curricula geared towards fostering positive attitudes to advocacy may help in producing future advocates.