The concept of social accountability consists in medical schools’ obligation to direct their activities towards addressing the priority health concerns of the communities or regions they serve as identified jointly with governments, health care organizations, health professionals and the public (Boelen and Heck, 1995). On the social obligation gradient, social accountability goes beyond simple awareness of health inequities and the role of social determinants of health (social responsibility) or undertaking activities to address those health inequities (social responsiveness) (Boelen and Woollard, 2011; Boelen et al., 2016). In addition to these features, social accountability is defined by engagement with the communities served by the medical school (Boelen et al., 2016). Canadian faculties of medicine were early adopters (Health Canada, 2001; Association of Faculties of Medicine of Canada, 2010; Busing et al., 2015) and report regularly on the alignment of their activities in this area, for example, in the literature (Jarvis-Selinger et al., 2008; Meili et al., 2011; Ross et al., 2014; Duke and Brunger, 2015; Cobb et al., 2018; Strasser et al., 2018), and in their accreditation responses. However, it is not clear to what extent communities have been able to identify their needs in those activities, or how medical schools have engaged with communities to address these needs, even in those that emphasize the importance of such engagement (Strasser et al., 2018).

The Faculty of Medicine at the University of Ottawa (uOttawa FoM) has expanded its social accountability mandate to enhance the quality, equity, relevance and effectiveness of health care by integrating stakeholder and community engagement into medical education, clinical service, and health research (University of Ottawa Faculty of Medicine, 2020). The present study is part of this new commitment, with the purpose of improving organizational learning and performance. We conducted an internal environmental scan of documentary and other data sources such as websites and social media, to address the following questions: Where is social accountability informing uOttawa FoM activities? How is it doing so?

For our environmental scan, we adapted a previously-published seven-step environmental scan protocol (Wilburn, Vanderpool and Knight, 2016). The Ottawa Health Science Network Research Ethics Board and the Bruyère Continuing Care Research Ethics Board approved this scan as a quality improvement project.

Data Sources

For quantitative analysis, an extensive list of potential data sources was created based on the THEnet Evaluation Framework for Social Accountability in Health Professional Education and mapped to the uOttawa FoM official website (The Training for Health Equity Network, 2011; Larkins et al., 2013). We used a snowball process to iteratively update our list as new sources were identified. We submitted data requests to specific faculty departments and programs to fill in data gaps. We targeted searches for origins of data sources with a large number of items, such as PubMed, granting agencies, and social media, built around the word “community” (see Supplementary File 1). We used a look-back window of five years starting June 1, 2015 for all data sources and included items in both English and French.

For qualitative analysis, we solicited brief descriptions of projects involving community and stakeholder engagement to gain an additional perspective of social accountability within the FoM. We requested that each department and office circulate a request to their respective Faculty members, and invited submission of descriptions in a Faculty-wide announcement. When potential projects were suggested by faculty members other than the project lead, we proactively sought descriptions from the leads themselves.

Data Extraction

Data collection and extraction took place between June 3 and July 27, 2020. We used a pre-specified form (Supplementary File 2) to extract information on each item, highlighting social accountability factors such as the values addressed, populations represented, and determination of category on the social obligation gradient (Table 1). Two investigators (SL, GZ) piloted the form for a subset of items in each type of data source in duplicate to ensure reproducibility and adapted our form accordingly, creating working definitions along the way. Brief descriptions were collected free-form on a separate document, along with the contact information and the name of the project. We excluded data sources with no relevance to social accountability.

Analysis and Synthesis

We used descriptive statistics to analyze characteristics of the items across the fields of the data extraction form. Fields were not mutually exclusive.

For qualitative analysis, four team members (SL, GZ, NK, MF) independently coded each brief description without a preconceived framework to identify initial codes, comparing each segment of data with other segments for similarities and differences. Team members met to discuss the codes and their relevance in relation to the research questions and established initial overarching themes and a corresponding coding framework. The final framework was reviewed and approved by the review team.

We conducted a comprehensive environmental scan of the University of Ottawa Faculty of Medicine. There is little consensus in the literature as to design, analysis, and reporting of such scans (Charlton et al., 2019). Therefore, we drew on previously published methods to design our processes, and on existing frameworks of social accountability for our data extraction and triangulation of data sources. Half of the items our scan found were education-related, with the majority taking place in UGME, which was the only origin of data sources with specific committees explicitly dedicated to social accountability. However, overall, only a small percentage (17.1%) of the items overall were socially accountable, suggesting that many FoM activities still need to embrace key traits  of social accountability.

Socially accountable activities were more likely to address specific populations/communities than were FoM items overall. We found that “partnership” was the value most often exemplified by socially accountable items, highlighting their extensive engagement with communities and stakeholders. These findings were also reflected in the qualitative analysis of the brief descriptions, which described a diverse range of projects, spanning Faculty domains, populations/communities, and the social obligation gradient. Many of these projects also met social accountability criteria such as a focus on priority health needs and populations affected by health inequities.

Although there is extensive literature on social accountability and medical schools, only a small number of studies have actually assessed social accountability across all activities of a medical school, and fewer still have relied primarily on the medical school’s documentation in their analyses. Hosny, Ghaly and Boelen combined document review with interviews to assess the level of social accountability at the Faculty of Medicine, Suez Canal University using Boelen and Woollard’s Conceptualisation/Production/Usability model (Boelen and Woollard, 2009), and found that, although it didn’t meet the criteria in some areas, its “history of community engagement” meant that it would be able to “fulfill a socially accountable mandate in the near future” (p. S53) (Hosny, Ghaly and Boelen, 2015). Elsanousi et al. (2016) used the WHO social accountability grid (Boelen and Heck, 1995) to assess faculty documents, internal and external reports, and some programme data from the University of Gezira faculty of medicine, a member of THEnet. Social accountability was less prominent in the research domain, but, overall, the authors concluded that the faculty “is socially responsible and responsive and is progressing successfully towards being fully socially accountable” (p. 265) (Elsanousi et al., 2016). A third study also combined document analysis with interviews, assessed against the WHO social accountability grid, to assess progress towards social accountability at the College of Medicine, Qassim University (COMQU). These authors found that COMQU “needs to have more interaction with stakeholders” to realize “the full spectrum of social accountability”. Although these studies reinforce the usefulness of document analysis in assessing social accountability, none involved the granularity of the THEnet evaluation framework from which we drew. The level of detail provided by that framework is useful for medical schools to more accurately assess their social accountability strengths and weaknesses.

An advantage of the environmental scan approach we used is that it provides an overview across both categories and domains, as we have defined them, and thus has the potential to indicate relevant socially accountable activities that might otherwise be missed. For example, we found evidence of social accountability in the domains of Admissions and Curriculum across three categories (Faculty Administration, Departments and Offices, and Education), although we might have expected to find them only in the Education category. Our use of the social obligation gradient in our evaluation framework has allowed us to map out the FoM’s progress towards full social accountability and, in particular, to identify areas and activities that are in the vanguard.

Our study has some limitations. Our efforts to fill gaps in the data through direct email contact with departments, offices or project personnel did not always result in a timely response. As grants and publications are collated differently by different research institutes, we were unable to collect these data across the entire Faculty, and we were also only able to search the databases of the three major funding agencies, rather than those more likely to fund community and stakeholder engagement activities. While we anticipate our broad request for projects related to social accountability filled in some of these potential gaps, it is still likely that we missed some items and activities occurring in the Faculty.

This study demonstrates the value of the environmental scan approach in assessing the current state of social accountability at a medical school. In the case of the University of Ottawa Faculty of Medicine, some activities clearly meet the criteria for social accountability, whereas other domains could benefit from a greater emphasis on engagement and identification of priority health needs. We anticipate our methods may be useful to other Faculties of Medicine seeking to comprehensively strengthen their own social accountability mandates.