Development of a Comprehensive Survey to Assess Key Socioecological Determinants of Health

Abstract

Background:

While disparities in surgical outcomes are well-documented, our understanding of how socioecological factors drive these disparities remains limited. Comprehensive and efficient assessment tools are needed. This study’s objective was to develop and assess the acceptability and feasibility of a comprehensive tool evaluating socioecological determinants of health (SEDOH) in colorectal surgery patients.

Methods:

In the first phase, a comprehensive SEDOH assessment tool was developed. A review of validated socioecological health evaluation instruments was conducted, and a two-step modified Delphi method addressed length, clarity, appropriateness, and redundancy of each instrument. A comprehensive tool was then finalized. In the second phase, the tool was tested for acceptability and feasibility in adult colorectal patients using a theory-guided framework at three Alabama hospitals. Relationships between survey responses and measures of acceptability and feasibility were evaluated using results from initial pilot tests of the survey.

RESULTS:

In Phase 1, a modified-Delphi process led to the develop of a comprehensive tool which included 31 SEDOH (88 questions). Results of acceptability and feasibility were globally positive (>65%) for all domains. Overall, 83% of participants agreed that others would have no trouble completing the survey, 90.4% of respondents reported the survey was not burdensome, 97.6% of patients reported having enough time to complete the survey, and 80.9% agreed the survey was well-integrated into their appointment.

CONCLUSIONS:

An 88-item assessment tool measuring 31 socioecological determinants of health was developed with high acceptability and feasibility to colorectal surgery patients. This work aids in the development of research needed to understand and address surgical disparities.

INTRODUCTION:

Among surgical populations, it is increasingly clear that certain groups, such as Black and rural patients, suffer from worse surgical outcomes [1–5]. Major gaps exist in our understanding of relevant contextual factors contributing to these surgical disparities [6, 7] and the mechanisms by which these factors mediate adverse outcomes [8, 9]. An emerging body of research suggests that socioecological determinants of health (SEDOH), or the behavioral, social, and built environment factors which impact an individual’s health, have a significant impact on patients’ overall health [10–13] and surgical outcomes [8, 14, 15]. Countless instruments for measuring individual SEDOHs exist, but an instrument that measures a conceptually-guided constellation of key SEDOHs is needed to explore the complex interplay of these factors’ impact on surgical outcomes [16].

When examining mechanism(s) of surgical outcomes such as mortality, length of stay, readmissions, and complications, most studies focus on individual-level factors such as demographic and co-morbidity characteristics [17, 18]. However, knowledge gaps exist in understanding how factors from other socioecological levels of health (i.e. interpersonal, organizational, community, or policy levels) influence surgical outcomes [8, 9, 19]. Additionally, explorations of SEDOH’s impact on surgical outcomes utilize assessment tools which often only evaluate a single SEDOH. This approach creates a fragmented understanding of these critical factors’ true impact. As a result, SEDOH literature frequently calls for more comprehensive analyses of how SEDOH impact surgical experiences and outcomes [8, 9, 17, 19, 20]. To develop this understanding, practical evaluation tools are critically needed.

This study aims to develop and pilot test a comprehensive, practical tool designed to measure key SEDOHs at all socioecological levels. Delivery of such a tool would be a major step in assessing how individual, structural, and environmental factors impact surgical outcomes. Understanding these mechanism(s) of action will ultimately lead to development of comprehensive, multi-level interventions to address surgical disparities.

METHODS:

To develop this survey and subsequently test it for acceptability and feasibility, a stepwise process was followed which is detailed in Figure 1. The process had two phases: Survey Development and Acceptability and Feasibility Testing. Survey tool development was performed at UAB Hospital in Birmingham, AL. Testing was completed at three rural and urban centers across the state of Alabama. The state’s population is 68.9% White, 26.8% Black, 4.8% Hispanic or Latino, and 1.6% Asian [21]. 16.1% of Alabama lives below the poverty line [21]. IRB approval was obtained prior to beginning work to develop the survey.

FIGURE 1:

Methods Process Outline

1. Development:

The objective for this phase was to define a conceptual model and to create a comprehensive tool capable of measuring critical SEDOHs. Established survey development guidelines [22, 23] recommended a two-step process involving a Development Phase and a Testing Phase [22, 23]. A detailed overview of these stages is presented below.

1A: Conceptual Model Developed:

The goal of the developed survey was to be capable of assessing a broad range of factors established to be relevant to surgical outcomes (health literacy, physician trust, patient activation, and social support) which were present at multiple layers of an individual’s socioecological environment. A variety of previously validated evaluation tools such as the Protocol for Responding to & Assessing Patients’ Assets, Risks & Experiences (PRAPARE) and the PhenX ToolKit offer mechanisms to capture this data, but few existing tools utilized in surgical care possess the scope needed to capture sufficient detail about a patient’s socioecological environment [24]. To develop such a tool, a conceptual model (Figure 2) was developed with the guidance of the socioecological model and established sources such as the World Health Organization’s identified Social Determinants of Health (SDOH), the PhenX Toolkit, and the CDC’s Healthy People 2030 [16, 24]. Notably, the Policy/Society domain was not specifically included in this model though the Delphi panel noted that several factors such as Discrimination could have reasonably been placed there.

FIGURE 2:

Conceptual Model

1B: Resource Compilation:

At this stage, the Delphi gathered to discuss an initial selection of SEDOH domains to be included for consideration in the tool. The Delphi was composed of an expert panel (9 surgeons, 4 surgical residents, 2 social determinant of health experts, and 3 health disparities researchers) who were chosen for their clinical experience with the surgical patient population, knowledge of relevant health disparities, and interest in developing this intervention. Conversations at this stage centered on developing a comprehensive list of domains and potential instruments to be considered for inclusion in the finalized tool. The Delphi utilized details from the literature review and observations from previous clinical and research experiences to select SEDOH tools at this stage. To ensure coverage of general demographics and foundational SEDOH, the PhenX Toolkit’s “Social Determinants of Health: Core” was initially reviewed. This bank of questions covered the domains Access to Health Services, Annual Family Income, Biological Sex Assigned at Birth, Birthplace, Current Address, Current Age, Current Employment Status, Educational Attainment – Individual, English Proficiency, Ethnicity and Race, Food Insecurity, Gender Identity, Health Insurance Coverage, Health Literacy, Occupational Prestige, and Sexual Orientation [25]. Additionally, clinical insight was used to select PhenX tools assessing Access to Health Technology, Spirituality, Purpose in Life, Transportation, and Discrimination. Delphi members were able to reference examples and experiences wherein which these factors were relevant to perioperative care. Collectively, 20 domains were identified from the PhenX Toolkit. Additionally, four survey domains and corresponding assessment tools were prioritized for inclusion in the survey for either their relevance in the literature or clinical insight. These survey tools were the Patient Activation Measure (PAM^®), Wake Forest Physician Trust Scale (WFPTS), the BRIEF Health Literacy survey (BRIEF), and Social Support Survey (SSS) [26–29].

1C: Initial Survey Created:

Following literature review, 24 SEDOH were identified. Chosen domains at this stage were selected for their relevance to surgical disparities previously demonstrated in existing literature or based on clinical insight from the authors. Identified assessment tools were compiled and organized into an initial draft to be discussed at the first round of the Delphi process. Previously validated assessment tools were left in their original form without any modifications [20, 22, 30–34]. Nine questions were taken from previously validated tools and used to assess singular, simple domains such as Food Insecurity relating specifically to cost, Discrimination in the healthcare setting specifically, and aspects of home environment assessing issues such as mold, water leaks, and malfunctioning utilities [25, 35, 36]. In these instances, rigorous consideration was taken by the Delphi to ensure that the individual question was capable of fully assessing the domain in question independently (such as in the case of the domain “community involvement” which conducts a ten point evaluation of community involvement as a single question taken from the much lengthier World Value Survey) [37].

1D: Modified Delphi Round 1:

This study employed a two-step, modified-Delphi method to develop the survey tool. In the first round, the Delphi panel met to discuss and reach unanimous agreement about the inclusion of certain survey tools in the survey. Prior to the first meeting, the Delphi panel developed an initial draft of the survey which was then distributed individually to members ensuring that each was blinded to the responses of other voting members. Individual members of the Delphi panel voted to include or remove individual questions and left comments on questions which they felt needed additional discussion. Voting members then resubmitted their comments and votes to organizing members of the Delphi panel. Votes and critiques were consolidated and discussed at the first meeting of the Delphi panel, the results of which are described below. This first round focused on producing a robust, focused, and functional tool.

From the identified collection of SEDOH screening tools, Delphi members filtered questions and instruments to reduce redundancy, reduce time of administration, and ensure that only the most well-suited assessment tools and questions were selected for future drafts. Additionally, based on the recommendations of SEDOH experts who referenced the widely recognized Social interventions Research & Evaluation Network (SIREN) resource library and PRAPARE, 8 new domains were suggested for inclusion in the survey: Military Status, Refugee Status, Recent Incarceration, Material Security, Living Environment, Safety (Current Residence), Housing, Social Cohesion & Informal Social Control, Community Involvement, Ease of Scheduling an Appointment, Stress, Access to Healthy Foods, Physical Activity - Neighborhood Environment, and Neighborhood Safety [38]. Corresponding surveys for these new SEDOH were identified. Concise sources such as the PRAPARE and shortened, efficient questionnaires such the Perceived Stress Scale (PSS-4) were prioritized for inclusion [39].

The Delphi reviewed the domains and questions, voted, and then met collectively via online video communication to unanimously reach consensus with a selection of domains and questions to be refined further at subsequent meetings. The first meeting concluded with a robust selection of 31 SEDOH from 12 unique survey tools.

1E: Modified Delphi Round 2:

In the second round, Delphi members further refined the survey while maintaining a thorough and reliable assessment of all identified domains. To accomplish this, each question was considered with respect to the total length of the survey, readability, and redundancy.

Following a discussion of each of the eligible question’s appropriateness in the larger survey, consensus was reached about the survey’s final contents, and a coherent and logical ordering of questions was reached. The final ordering and organization of the SEDOH-88 is shown in Table 1.

Table 1:

SEDOH-88 Domains and Sources

Domains	Socioecological Level	# Items	Source
Income (%FPL)	Individual	1	PhenX
Sexual Orientation, Gender Identity	Individual	2	PhenX
English Proficiency	Individual	1	PhenX
Current Employment	Individual	1	PhenX
Educational Attainment	Individual	1	PhenX
Military Status	Individual	1	PRAPARE
Refugee Status	Individual	1	PRAPARE
Incarceration--Past Year	Individual	1	PRAPARE
Health Insurance	Organizational	1	PRAPARE
Transportation	Community	1	PRAPARE
Food Insecurity – Cost Prohibitive	Individual	1	PhenX
Delayed Medical Care for Cost	Individual	1	PRAPARE
Access to Health Technology	Individual	2	PRAPARE
Material Security	Individual	1	PRAPARE
Living Environment	Community	1	AAFP
Safety (Current Residence)	Community	1	PRAPARE
Housing (Type of Residence, Fear of Losing)	Individual	2	PRAPARE
Health Services (Last Accessed, Usual Places, Most Often)	Individual	3	PhenX
Ease to Schedule an Appointment	Organizational	1	UAB*
Health Literacy	Individual	4	BRIEF
Patient Activation	Individual	13	PAM
Trust	Interpersonal	7	WFPTS
Social Support	Interpersonal	8	SSS
Spirituality and Religion	Individual	3	PhenX (2), UAB* (1)
Discrimination - Lifetime	Individual	8	PhenX (7), CHIS (1)
Community Involvement	Community	1	World Value Survey
Stress	Individual	4	PSS-4
Access to Healthy Foods	Community	3	PhenX
Community Cohesion and Informal Social Control	Community	10	PhenX
Neighborhood Recreation Infrastructure	Community	2	PhenX
Neighborhood Safety	Community	1	PhenX
Total		88

^*These questions were developed by a group of SEDOH experts at UAB Hospital and then subsequently approved with consensus from the Delphi.

Acronyms: American Academy of Family Physicians Social Needs Screening Tool (AAFP)[53], Protocol for Responding to & Assessing Patients’ Assets, Risks & Experiences (PRAPARE). PhenX (The PhenX Toolkit). CHIS (California Health Interview Survey). BRIEF (BRIEF Health Literacy)[50]. PAM (Patient Activation Measure)[29]. WFPTS (Wake Forest Physician Trust Scale)[28]. SSS (Social Support Scale)[26]. PSS-4 (Perceived Stress Scale)[39].

This final version of the survey will be referred to as the SEDOH-88 from this point forward. 2: Acceptability and Feasibility Testing: The objective for this phase was to test the SEDOH-88 for acceptability and feasibility in a cohort of colorectal surgery patients.

2A). Post-Survey Development:

A holistic, previously validated post-survey to assess acceptability and feasibility of surveys was not identified in literature reviews [40–42]. Thus, to evaluate the newly developed tool’s acceptability and feasibility, this study utilized established theoretical frameworks detailed by Sekhon and Bowen to create a theory guided post-survey [40, 42]. The concept of acceptability was assessed in this study by considering its substituent elements: Burden, Ethicality, Intervention Coherence, Affective Attitude, Opportunity Costs, Self-Efficacy, and Perceived Effectiveness [40]. Similarly, this study assessed the following elements of feasibility: acceptability, demand, implementation, practicality, adaptation, integration, and expansion [42]. Limited Efficacy testing was initially performed, but, due to small sample size and limitations in data analysis, ultimately not utilized as a component of acceptability and feasibility testing. SEDOH-88 questions, listed in Table 2 and Table 3, were developed to assess each of the listed domains of acceptability and feasibility [40, 42]. All SEDOH-88 questions were assessed using a five-item Likert Scale except for Burden which used a four-item Likert scale. Assessing concepts like Demand and Adaptation required more nuanced approaches which are described below. The post-survey was administered immediately after the patient completed the 88-item survey.

Table 2:

Questions to Evaluate Acceptability

Elements of Acceptability	Post-SEDOH-88 Questions (Patient)
Burden (Number of Questions)	Thinking about the number of questions on this survey, did you feel that the survey was:   a) Unacceptable   b) Very burdensome   c) Somewhat burdensome   d) A little burdensome   e) Not at all burdensome
Ethicality (Appropriateness of Q’s)	Do you think it is appropriate to be asked about your social and economic needs at this clinic?   a) Very inappropriate   b) Somewhat inappropriate   c) Neither appropriate nor inappropriate   d) Somewhat appropriate   e) Very appropriate
Intervention Coherence (Face Validity)	How much do you agree with the following statement? I feel that my doctors are able to provide better care for me when they know this information about me.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree
Affective Attitude (Feelings; strong positive to strong negative)	How does it make you feel that your healthcare team wants to collect this information?   a) Strongly Positive Feelings   b) Positive Feelings   c) Neutral   d) Negative Feelings   e) Strong Negative Feelings
Opportunity Costs (Time Spent)	How much do you agree with the following statement? I feel that providing this information to my healthcare team was worth my time.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree
Self-Efficacy (Coherence)	How much do you agree with the following statement? The questions in this survey were asked in a way that I could understand.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree
Perceived Effectiveness	How much do you agree with the following statement? This survey captured details about my life that are important for my health.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree

Table 3:

Questions to Evaluate Feasibility

	Post-SEDOH-88 Questions (Patient)
Implementation	Thinking about taking this survey on the iPad, how easy was it for you to complete?   a) Extremely easy   b) Somewhat easy   c) Neutral   d) Somewhat difficult   e) Extremely difficult
Practicality	How much do you agree with the following statement? I had enough time to complete the full length of the survey during today’s visit.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree
Integration	How much do you agree with the following statement? I feel that this survey was smoothly integrated into today’s visit.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree
Expansion	How much do you agree with the following statement? All in all, I feel that most patients would have no trouble completing this survey.   a) Strongly disagree   b) Disagree   c) Neutral   d) Agree   e) Strongly Agree

^*The domains Adaptation and Demand Testing were evaluated using alternative methods. See Methods section 2A.

Assessing Demand:

Demand was evaluated by asking participants to choose which themes from the SEDOH-88, if any, were relevant to their health outcomes. This question was designed in accordance with the idea that demand could “be assessed by gathering data on estimated use or by actually documenting the use of selected intervention activities in a defined intervention population or setting” [42]. From the patient perspective, the benefit of completing the research tool and, thus, the best estimate of a patient’s desire to “use” the tool, is the potential for improved services and health outcomes. By asking patients to identify factors they believe to be associated with their health, data collected from this question characterized a patient’s demand to be surveyed on factors like patient-physician trust or health literacy. A free response option was offered with this question in which patients were invited to input additional factors they believed to be relevant to their health.

Assessing Adaptation:

To assess the SEDOH-88’s adaptability in a diverse patient population, we evaluated differences in answers to post-survey and SEDOH-88 questions between differing participant cohorts (age, race, gender).

2B: Recruitment and Administration

Inclusion criteria required the patient to be ≥18 years of age, English speaking, able to give consent and undergoing or having recently undergone (within last 7 days) colorectal surgery for colorectal disease (diverticular disease, colorectal cancer, or inflammatory bowel disease). Eligible patients were identified and approached following screening of EMR records for eligibility criteria. Colorectal surgery patients were chosen as the primary study group due to the presence of significant disparities in this population, the longitudinal nature of care of many colorectal diseases, and the high frequency of colorectal procedures even in rural communities [8, 18]. Patients were recruited from both inpatient and outpatient settings in the authors institution (a large academic medical center in an urban area) as well as two smaller, rural, community hospitals affiliated with the authors institution (Whitfield Regional Medical Center in Demopolis, AL and Regional Medical Center of Central Alabama in Greenville, AL). A small incentive gift card ($20) was offered for the completion of the the 88-item tool and subsequent post-survey to encourage participation in the study and to compensate participants for their time. Patients were administered the SEDOH-88 and the post-survey assessing acceptability and feasibility on iPad’s using RedCap.

2C. Statistical Analysis

Univariable analyses were conducted on post-survey data to evaluate the acceptability and feasibility of the newly created SEDOH-88. Descriptive statistics also sought to characterize meaningful relationships between answers to the post-survey and the SEDOH-88.

Univariate analyses were conducted on four- or five-item Likert-scale questions to assess acceptability and feasibility. To account for the small number of low scores in the post-survey, scores were grouped by combining the frequencies of lower-end (1’s, 2’s, and 3’s) and higher- end (4’s and 5’s) Likert scale responses. Scores for each question were calculated by adding the frequency of 4’s and 5’s (and for the four-item Likert Scale, the frequency of 4’s) and dividing it by the total number of participants.

Two SEDOH-88 domains (Adaptability and Demand) were not assessed with a Likert-scale-based responses. Adaptability was assessed using spearman rank coefficients between participant demographics (age, race, sex, and institution) and post-survey responses. Demand was assessed using univariate analysis of post-survey responses to the post-survey question evaluating demand.

Spearman rank coefficients were used to evaluate the strength and direction of relationships between acceptability and feasibility metrics measured in the post-survey in relation to domains assessed in the SEDOH-88. As an example of this, health literacy status (a SEDOH-88 domain) was compared to perceived burden (a post-survey domain) to reveal if low or high health literacy status may correlate to low or high reported burden of completing the SEDOH-88. Correlations were categorized as strong positive, medium positive, medium negative, or strong negative based on computed ρ (−1 to 1). Significance was assessed at the 0.05 alpha level, and the relevance to acceptability and feasibility was analyzed.

RESULTS:

1: SEDOH-88 Development:

1B-1E: A conceptual model was developed to guide development (Figure 1). Literature review of relevant existing SEDOH, SDOH, and demographic tools identified >30 individual instruments which collectively contained >300 items. An initial draft was created which contained 24 domains (217 items). Following the first round of the modified Delphi method, the tool contained 31 domains (>100 items), and, after the second and final round of the Delphi, the tool contained 31 domains (88 items). The final product tested domains at four levels of the socioecological model (Individual, Interpersonal, Organizational, and Community). For the domains “Ease to Schedule an Appointment” and “Spirituality/Religion”, a concise assessment tool was not identified in the literature search. To address this, the Delphi method panel reached consensus on a customized Likert Scale based question to assess these domains (listed as “UAB” in Table 1). A complete list of domains and sources is available in Table 1.

2: Acceptability and Feasibility Testing:

2A: Post-Survey Development:

The finalized post-survey was a short, 12-item tool capable of assessing nine fundamental constituent elements of acceptability and feasibility.

2B: Recruitment and Administration:

44 participants were approached in person using iPads and asked to complete both the SEDOH-88 and post-survey. This number was based on recommendations from recognized survey development guidelines [22, 33]. 41 participants completed the SEDOH-88 and post-survey in its entirety, and there were no partially completed surveys. Demographic characteristics taken from EMR information of participants is shown in Table 4. Estimated time for completion of the SEDOH-88 was 12–15 minutes.

Table 4:

Demographic Data

Demographics	Overall (n=41)	Demopolis (n=15)	Greenville (n=3)	UAB (n=23)
Age, mean (STD)	68.2 (8.1)	70.2 (3.7)	72.6 (7.5)	54 (13.6)
Gender, n (%)
Male	18 (43.9)	5 (33.3)	0 (0)	13 (56.5)
Female	23 (56.1)	10 (66.7)	3 (100)	10 (43.5)
Race (n, %)
White	18 (43.9	2 (13.3)	0 (0)	16 (70.0)
Black	21 (51.2)	12 (80.0)	3 (100)	6 (26.0)
Asian	2 (4.8)	1 (6.6)	0 (0)	1 (4.3)

Demographic makeup of SEDOH-88 participants recruited from three locations across the state of Alabama taken from EMR data.

2C: Post-Survey Results:

Post-survey results were positive (indicating that for each question >65% of participants selected answers coded as either a four or five based on Likert scale data) for all assessed domains based on Likert Scale data (Figure 3). Overall, 83% of participants agreed that other patients would have no trouble completing the SEDOH-88, 90.4% of respondents reported that the SEDOH-88 was not at all burdensome, 97.6% of patients reported they had enough time to complete the SEDOH-88, and 80.9% agreed that it was well-integrated into their appointment. Of note, high health literacy scores (r(39=0.48, p=0.002) and higher levels of education (r(39)=0.51, p=0.0007) were positively correlated with higher reported self-efficacy.

FIGURE 3:

Acceptability and Feasibility of SEDOH-88

Assessment of Demand:

Demand was evaluated for each individually assessed domain.

Patients were given a complete list of SEDOH-88 domains and asked to select which of them they believed to be relevant to their health. The percentage of participants selecting a certain domain was the outcome for the assessment of Demand. The highest scoring domains were Patient-Physician Trust (21%), Healthcare Access (24%), Health Maintenance (Patient Activation) (21%), and Health Literacy (21%). The lowest scoring domains were Gender Identity/Sexuality (5%), Religious Faith/Purpose in Life (5%), and General Demographics, Education, and Income (5%).

Assessment of Adaptation:

Correlations between age and post-survey responses were successfully evaluated using the methods described above. Higher age was positively correlated with higher reported ethicality (r(39)=0.37, p=0.017) and affective attitude (r(39)=0.37, p=0.018). Higher age was mildly negatively correlated to self-efficacy (r(39)=−0.38, p=0.014) and the perception that employment status would have an impact on a participants health outcomes (r(39)=−0.31, p=0.046).

DISCUSSION:

This study developed and assessed for the acceptability and feasibility of a comprehensive tool measuring key SEDOHs in colorectal surgery patients. Developed using a modified-Delphi method with key participant engagement, this 88-item tool (SEDOH-88) is a collection of validated items that captures a diverse representation of SEDOHs. The SEDOH-88 was demonstrated to be clinically acceptable and feasible. Colorectal surgery patients reported low levels of reported burden, positive affective attitude towards the SEDOH-88’s content and expressed that completing the SEDOH-88 was worth their time. Together, these results indicate that the SEDOH-88 could be a useful tool to assess SEDOHs within surgical populations and, subsequently, assist in the creation of interventions aimed at addressing health disparities. Further work with validity and reliability testing will be critical to ensure accurate implementation of this tool in broader populations as well.

Our study addresses issues raised in SEDOH literature, most notably from Britt et al. who calls on, “[p]atients, surgeons, researchers, measure developers, hospital administrators, and policymakers… [to] unite to specify, test, and implement new surgical access measures.” [43]. The need for such collaboration stems from a limited understanding of surgical disparities driven by a previously “myopic focus on process and outcome measurement” in SEDOH research thus far [40]. Further, there exists a need for the development of tools that will elucidate contextual socioecological factors potentially driving surgical disparities [40]. The assessment tool described in this paper will be well suited to assist in defining and characterizing a broad array of SEDOH which impact surgical outcomes. Additionally, while specifically designed to assess surgical populations, this comprehensive tool introduces a path to merge data into ongoing, multidisciplinary efforts to characterize and foster health equity across medical, institutional, and social contexts such as the ongoing Gravity Project [44]. Our project is a relevant adjunct to SEDOH data collection efforts while being specifically designed to understand opportunities relevant to surgical populations. Additionally, the SEDOH-88 is a potential adjunct for providers conducting CMS annual wellness visits, which mandate providers collect information regarding, “[p]sychosocial risks including, but not limited to, depression, life satisfaction, stress, anger, loneliness or social isolation, pain, and fatigue” in addition to traditional health metrics. [45]. In short, the SEDOH-88 is an efficient, consolidated assessment tool with high clinical acceptability and feasibility in the surgical population. The SEDOH-88 provides a mechanism for surgeons, clinicians, and researchers to gather comprehensive socioecological and psychosocial data points which are being recognized and prioritized by a variety of industry leaders and reimbursement agencies.

The socioecological perspective provides needed context to surgical disparities research [24]. Current literature frequently focuses on individual-level factors (Health Literacy, Patient-Physician Trust) which limits the development of interventions capable of engaging factors at multiple socioecological levels [17, 46, 47]. With a multilevel approach, interactions between multiple levels of an individual’s environment can be assessed, and a more detailed understanding of adverse and protective contributing factors can be obtained [24]. For example, the protective influence of employment (an individual level factor) is better understood with knowledge of organizational (e.g., health insurance provided by their employer) and interpersonal factors (e.g the social support obtained from co-workers) as these may also mediate the impact of employment on surgical care and outcomes. The use of a multilevel perspective allows a path for characterizing relationships between contributing factors accurately and comprehensively which yields data capable of informing complex and durable interventions.

Research detailing assessments of acceptability and feasibility for survey-based screening tools are rare, and even fewer studies assess the specific needs of the surgical population [31, 40–42, 47–49]). Existing studies in this space detail either the creation of new tools [23, 29, 50] or processes for testing acceptability and feasibility of pre-existing tools [32, 41, 48, 51]. Thus, the methods and results described in this study may be particularly valuable for researchers seeking a clear and relatively simple path to create and test tools used to study SEDOH in surgery.

This study had several important limitations. First, to develop a concise and clinically feasible tool, individual items were taken from larger previously validated tools such as the World Value Survey. This limits the scope of these larger assessment tools to the output of the selected item [52]. This was remedied by restricting the use of individual items from larger surveys, using exact language from previously validated instruments, and seeking consensus in the Delphi panel when proceeding with these changes. Second, insufficient sample size limited the ability of this study to evaluate observed differences in responses with respect to gender, race, and type of visit. However, rates of acceptability and feasibility were globally positive, and the researchers are currently working to increase the sample size to obtain these measurements. Third, insufficient sample size also prevented the statistical evaluations that would be needed to establish the post-survey’s reliability and validity. As stated above, to address this, items were taken from previously validated sources in their original form. Additionally, no conglomerate score for the SEDOH-88 was produced. Rather, the SEDOH-88 is a collection of smaller tools which produce data corresponding to the constituent surveys. Due to limitations in RedCap, our study was not able to quantify the time needed to complete the SEDOH-88. However, given that participants reported that they felt they had enough time to complete the tool, the length and time needed was deemed acceptable. Finally, while our studied population was socioeconomically diverse, pilot testing of the SEDOH-88 included only colorectal surgery patients which limits the generalizability of these results. Future studies should assess the feasibility of this intervention in other surgical populations.

CONCLUSION:

Using a two-step design, the SEDOH-88 was successfully constructed and deemed acceptable and feasible by a cohort of colorectal surgery patients. This tool may assist researchers attempting to link SEDOHs to surgical outcomes and, ultimately, contribute to the development of multilevel interventions addressing surgical disparities.

Abbreviations:

SEDOH

Socioecological Determinant of Health

SEDOHs

Socioecological Determinants of Health

SEDOH-88

Our 88 item, comprehensive tool assessing 31 unique socioecological domains