Background
Inner city populations are groups of marginalized individuals in urban settings who live with any combination of poverty, unstable housing, mental health issues, problematic substance use and involvement in survival sex or drug trade [
1,
2]. The term ‘inner city’ remains common in Canada; however, phrases used to describe inner city populations in other settings may include ‘marginalized populations’, ‘vulnerable populations’, or ‘urban underserved populations’. [
3] This is a difficult population to holistically define, in part due to the evolving nature of language and the social context in which terminology is used. Despite being a heterogeneous patient population, its members share similar unmet needs for care and past care experiences [
2,
3].
Inner city populations have a high burden of illness and mortality but less access to effective primary care [
4‐
6]. Providing health services to this population is often difficult because of mutual mistrust, population heterogeneity and the unique circumstances surrounding each individual. The provision of more holistic and evidence-based care to high-risk groups, which make up a higher proportion of frequent health-care users, is urgently needed [
5,
7].
Despite regular opportunity for interaction with this population, learners may feel uncomfortable or underprepared for professional interactions with individual patients and their unique context and health care needs [
8]. Learners may hold negative beliefs and/or attitudes which undermine the therapeutic relationship and may contribute to poor health outcomes in this group [
9,
10]. Over the course of health professions training, some learners develop progressively more negative attitudes towards, and greater reluctance to work with, specific marginalized populations [
1,
8]. Conversely, supported exposure to inner city patients and focused curricula can improve attitudes towards at-risk populations and increase the likelihood that learners will choose to work with these groups [
11‐
13].
Measuring attitudes is one component of evaluating the impact of curricular interventions. Attitudes can be thought of as ‘a relatively enduring organization of beliefs, feelings, and behavioural tendencies towards socially significant objects, groups, events or symbols’, although many definitions have been proposed [
14,
15]. In a commonly endorsed tripartite model, attitude is defined as a construct comprising affective, behavioural, and cognitive components [
16,
17]. A tool measuring learner attitudes towards members of the inner city would assist in evaluating curricula designed to improve those attitudes. While a literature review identified a number of published tools examining attitudes toward specific subpopulations (e.g. populations defined solely by homelessness or single health conditions like mental health issues, substance use or HIV positive status) [
18‐
26], no tool was sufficiently broad enough in context or language to apply to complex inner city populations that are encountered in generalist medical settings, such as in emergency or primary care [
27]. Attempting to apply existing tools would be insufficient to capture the experience of a healthcare professional caring for a patient who presents with multiple issues resulting from a complex interplay of health and social concerns, rather than an isolated single health condition or social problem. A generalist lens supports the delivery of comprehensive, high quality health care to complex populations; assessing competency in this area requires an equally comprehensive, generalist stance.
The objective of this study was to develop and provide validity evidence for a tool to measure health care learner attitudes towards inner city populations.
Discussion
The tool developed in this study, the Inner City Attitudinal Assessment Tool (ICAAT, Table
3), consists of 24 items within three conceptual themes (‘Affective’, ‘Behavioural’, and ‘Cognitive’). The validity evidence for the ICAAT was generated from a multidisciplinary, pre-clinical health care learner cohort. The factors identified in the ICAAT were noted to parallel previous research examining the construct of “Attitude”, notably the tripartite model [
14‐
17]. The ICAAT can contribute to health sciences education, by offering a literature- and expert-informed tool to measure health professions learners’ attitudes about providing care to members of inner city populations.
Table 3
Inner City Attitude Assessment Tool (ICAAT). Participants are instructed to indicate their level of agreement with each item using a six-point Likert-type scale (1–strongly disagree; 2 – disagree; 3 – somewhat disagree; 4 – somewhat agree; 5 – agree; 6–strongly agree). The items are meant to appear in a random format. The following preamble may appear with the items: ‘This tool assesses attitudes towards inner city populations. Your responses will remain anonymous. Please answer the following as honestly as possible.’
Factor 1 – Affective | 1. I feel uneasy when interacting with patients from the inner city. 2. I feel uncomfortable when I talk to a patient from the inner city about their social circumstances. 3. I feel uneasy when I am in a room alone with someone from the inner city. 4. I avoid contact with people from the inner city when I am outside of a health care setting. 5. I avoid contact with people from the inner city when I am in a health care setting. 6. I find it difficult to work with patients from the inner city because I have no way of relating to them. 7. I find it difficult to view things from the perspective of a patient from the inner city. 8. I am reluctant to talk to patients from the inner city about their social circumstances. 9. People from the inner city are disruptive to health care staff and other patients. 10. People from the inner city do not adequately value their own health. 11. People from the inner city overuse the health system and waste health care dollars. |
Factor 2 – Behavioural | 1. I feel I know enough about the health issues of inner city populations to provide care to a patient from the inner city. 2. I feel that I know enough about the social determinants of health to provide care to a patient from the inner city. 3. I feel capable of establishing a good working rapport with patients from the inner city. 4. I feel capable of communicating effectively with a patient from the inner city. 5. I feel capable of facilitating trust with a patient from the inner city in a professional setting. |
Factor 3 – Cognitive | 1. It is my professional responsibility to provide care to underserved populations. 2. My profession should be involved in providing care to underserved populations. 3. Professionals in my discipline should address social determinants of health (such as unstable housing) when interacting with patients. 4. Professionals in my discipline should advocate for the health of inner city populations. 5. Professionals in my discipline should adapt how care is provided in order to meet the needs of patients from the inner city. 6. A person from the inner city deserves hospital space and resources as much as any other patient. 7. It is worth my time to provide care to someone from the inner city. 8. Providing care to inner city populations is pointless. |
Inner city populations have disproportionate health care needs for their population size [
4‐
6], yet are also a group that can be challenging to treat for a variety of reasons [
5,
7]. The current model for health care is not effective for inner city populations whether looked at from a patient outcomes perspective, or a system cost perspective. A shift in training models is urgently needed. The common pattern of health care use by people within an inner city context is that of higher acuity on presentation, higher medical complexity, lower preventative health uptake, and higher health costs [
46,
47]. The likelihood of poor outcomes for these patients is exacerbated when they are seen by staff who may have little to no training in managing health inequities resulting from social determinants of health [
48]. Increasing evidence confirms the failure of crisis-oriented care delivery and the value of social determinants of health and relationship-centred care [
49,
50], highlighting the need for effective training programs and curricula to expose health professions trainees to these contextual influences on health.
A shift in training is needed to ensure that inner city populations get appropriate care, and to improve health care outcomes for this population. Unfortunately, health care educators and role models may sometimes hold negative attitudes towards patients in crisis who are experiencing the impact of adverse social circumstances, past trauma, and untreated medical conditions; these negative attitudes may be modeled for learners, whose attitudes are also seen to worsen over the course of training [
8,
51‐
53]. Any training program designed to address attitudinal competencies must be accompanied by ongoing program evaluation to monitor whether the programs are having the intended effect of positive changes in attitudes. The ICAAT is a tool that could assist in the evaluation of these programs through measuring the self-reported attitudes of the learners being trained. This could take the form of a before-and-after analysis of an educational intervention, or a pooled comparison of groups.
In recent years, health sciences education programs have been shifting to competency-based approaches. Several competency frameworks have been developed in response to this shift, such as the CanMEDS roles in Canada [
54], the Accreditation Council of Graduate Medical Education competencies in the United States [
55], and the Royal College of General Practitioners competencies in the United Kingdom [
56]. These frameworks all emphasize the need for health professional learners to demonstrate competencies beyond medical knowledge. However, addressing non-clinical aspects of health care is relatively more difficult than addressing knowledge-based domains [
57]. The ICAAT is structured to address attitudes toward inner city populations, and can potentially be used in learner self-assessment.
A strength of the ICAAT is that it was designed for use with more than one health care discipline, reflecting the multidisciplinary environment that characterizes our current health system; as such, the development process included input from representatives of different disciplines. The research team included medical and nursing professionals; the expert panel included representatives of medicine, nursing, social work, and community members with lived inner city experience; and the readability and pilot testing involved medicine and nursing students. By having more than one discipline represented in the development of the tool, there is greater likelihood that the final tool is useable in cohorts of learners in several health care disciplines.
Through use of items from a short form of the Marlow-Crowne social desirability scale [
38], tendency of responses primarily due to social desirability bias was evaluated. Results from factor analysis yielded evidence against social desirability bias in the responses. This is a reassuring finding regarding the extent to which learners might provide answers to appear socially appropriate as opposed to answers representing their true attitudes. However, these results are taken in the context of anonymity of the respondents.
Limitations
Some may find the term ‘inner city’ to be inadequate to define the population, or even pejorative in its meaning. Language evolves as the conceptual understanding of phrases changes, as does the social context in which they are used. Recognizing this challenge, the chosen phrasing is common and remains acceptable within the current Canadian context. Indeed, a number of Canadian health care services and related academic initiatives specifically reference inner city populations within their mandate. The term ‘inner city’ appeared to be sufficiently understood by the expert panel and students involved in readability testing, and the ICAAT concept was warmly received. However, this term might not be equally understood or accepted in all contexts in which the ICAAT might be employed. Where the term ‘inner city’ is less common, the ICAAT may perform differently, and educators might consider gathering validity evidence, with or without modified terminology, prior to widespread use in such settings.
It is beyond the scope of the ICAAT to assist with broader case finding for marginalizing circumstances. Inner city residence does not necessarily imply marginalization, nor are marginalizing conditions constrained to inner city locations. However, marginalizing conditions can congregate in specific urban settings, and students working in these communities should be mindful of this prevalence and poised to provide appropriate care. Moreover, exposure to inner city learning experiences, with the ICAAT as a means of self-reflection, will better equip students to address marginalization in a variety of settings [
58].
Educators might worry that the use of instruments assessing attitudes will introduce or solidify implicit learner biases against inner city populations. Although we did not perform longitudinal assessment when gathering our validity evidence to refute this possibility, we did not elicit any concerns about a potential Hawthorne effect from participating student cohorts, medical and nursing school personnel, or Delphi panelists. If paired with targeted curriculum that acknowledges and challenges implicit biases, the ICAAT is unlikely to override that learning experience or have a net negative effect.
Although the connection is expected conceptually, a direct connection between attitudes addressed in the ICAAT with real-life practice and patient outcomes has not been made. We would expect that students who score higher on measures of positive attitudes towards inner city patients would have greater therapeutic alliance and presumably better patient outcomes, but one of the limitations of educational assessment tools is the difficulty in making explicit links to changes in future practice and patient outcomes. It has previously been shown that greater scores on measures of empathy, for example, can be linked to patient outcomes [
59]. Notably, several items in the ICAAT address empathy, particularly in the first factor of ‘Affective’. Future research to examine the connection between attitudes measured by the ICAAT and real-world practice is warranted.
An important next step in tool development would be confirmatory factor analysis (CFA) in a larger representative cohort. CFA was not pursued because of the feasibility limitations of recruiting a large enough cohort. In the form described here, the ICAAT can be considered at an intermediate point on a continuum of tool development.
The modified Delphi process described herein was not a strict Delphi process in that it did not seek uniform consensus, but rather incorporated feedback from the experts. True consensus would not be feasible because panel members varied considerably with respect to specific expertise and discipline.
Future work
The ICAAT will require testing with new cohorts to gather additional validity evidence. This includes testing with more experienced clinicians, such as medical residents and practicing nurses, as well as with other health care-oriented disciplines, such as social workers and pharmacists. Further, knowledge translation activities are underway to encourage uptake of the ICAAT in undergraduate clinical settings; this ‘real-world’ use of the ICAAT will provide additional feedback and the opportunity to refine the tool as needed.
Acknowledgments
The authors wish to thank Dr. Monica Jepsen for her work on the scoping review; Ms. Jessica Muller for her representation of the nursing discipline on the study team; Dr. Allison Kirkham for her involvement in study design, data collection, and interpretation; panel members Dr. Rabia Ahmed, Dr. Vanessa Brcic, Ms. Rosemary Fayant, Dr. Stephen Hwang, Dr. Louanne Keenan, Dr. Thomas Kerr, Dr. Ryan Meili, and Ms. Jean Repchuk; Dr. David Buck and LC See for permission, on behalf of their project teams, to adapt items from their homelessness and HIV/AIDS attitudinal instruments for use in the tool described herein; and the undergraduate health sciences education programs from the University of Alberta Faculty of Medicine & Dentistry, the University of Alberta Faculty of Nursing, and the University of Calgary Cumming School of Medicine, who facilitated validation with an undergraduate learner cohort.
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.