Social Determinants of Health in Pediatric Ophthalmology Patients: Availability of Data in the Electronic Health Record and Association With Clinic Attendance

Abstract

Purpose: To characterize the availability of social determinants of health data in the electronic health record of pediatric ophthalmology patients and to examine the association of social determinants of health with attendance at scheduled operating room and clinic visits.

Methods: This was a retrospective cohort study of pediatric ophthalmology patients seen at The Hospital for Sick Children between June 1, 2018, and May 23, 2022. Data were collected on demographics, diagnosis, and management-plan. The χ² tests and multivariable regression were used to examine associations between social determinants of health and attendance at scheduled operating room and clinic visits.

Results: The cohort consisted of 26,102 study subjects with 31,288 unique eye-related diagnoses representing 57 unique ICD-10 codes. Availability of data in the electronic health record ranged from 100% for sex, age and postal code to 0.1% for ethnic group. Female sex (P = 0.004) and urbanicity (P = 0.05) were associated with higher operating room visit cancellations. Female sex (P = 0.002), age group 0–13 (P ≤ 0.001), low-medium neighborhood income quintile (P ≤ 0.001), residence of Northern Ontario (P ≤ 0.001), and urbanicity (P ≤ 0.001) were associated with higher clinic visit cancellations and no-shows.

Conclusions: At a major tertiary-care hospital in Canada, key social determinant data such as ethnicity are not consistently available in the electronic health record of pediatric ophthalmology patients. Female sex, younger age, and living in a rural area or neighborhood with low-medium income quintile may be predictors of missed visits and require further study.

Translational Relevance: This study highlights a need for improved documentation of social determinants of health variables in electronic health records.

Introduction

Social determinants of health (SDH) are the environmental conditions in which people are born, learn, live, work, play, age, and worship.^1–3 From a macro-perspective, SDH encompass five domains: economic stability, education access and quality, health care access and quality, neighborhood and built environment, and social and community context (Fig.).⁴ Economic stability is determined by one's employment, income, and wealth.^4–7 Education access and quality includes literacy, language, early childhood education, vocational training, and higher education.^4,7 Health care access and quality encompasses an individual's access to health coverage, health literacy, provider and pharmacy availability, access to linguistically and culturally appropriate care, and quality of care.^4,7 Neighborhood and built environment is described by an individual's housing infrastructure, transportation, health safety, walkability, access to healthy food, and geography.^4,7 Last, social and community context encompasses an individual's social integration, gender, ethnicity, family structure, community engagement, and exposure to violence/trauma.^4,7 These SDH domains intersect to influence the health and wellbeing of individuals, observed as differences in mortality, morbidity, life expectancy, health care expenditures, health outcomes, and functional limitations.⁸ 

Figure.

 

Social determinants of health encompass five interconnected domains: economic stability (purple), educational access and quality (yellow), health care access and quality (green), social and community context (pink), and neighborhood and built environment (blue). These domains intersect to influence health status. This study explored how specific social determinants of health variables within each domain (bullet points) influenced one aspect of health status, attendance to scheduled clinic and operating room appointments.

SDH such as poverty and low literacy are linked to poor health and well-being.⁹ Children, in a stage of rapid growth and development, are particularly vulnerable to the negative effects of SDH.^10,11 For example, children reared in poverty have been seen to have worse health outcomes than children from more privileged backgrounds, because they frequently lack basic necessities like access to food, transportation, and secure housing.¹² There is a growing body of evidence that SDH contribute to inequitable health care experiences in pediatric patients, specifically regarding patient access to and utilization of health care services.^5,11,13,14 Extensive literature emphasizes that adherence to the recommended treatment and follow-up schedules is essential for achieving high-quality health outcomes; poor attendance to scheduled medical visits can be a pervasive threat to health and wellbeing.¹⁵ 

In the field of pediatric ophthalmology, effective eye care delivery depends on understanding and addressing the social needs of patients.^11,16–18 Social determinants of health significantly influence pediatric eye care, because children from disadvantaged backgrounds often face barriers such as limited access to transportation, and financial constraints, in accessing timely and appropriate vision screenings and treatment, leading to higher rates of undiagnosed and untreated eye conditions.^11,17,19 In Canada, children from lower socioeconomic neighborhoods are less likely to undergo routine comprehensive eye exams.^19,20 Additionally, recent studies have shown that access to care and adherence to clinical recommendations for several ophthalmic conditions are associated with various socioeconomic or racial and ethnic factors.^17,21 

Despite the growing importance of SDH in pediatric ophthalmology, there is limited information on the specific impact of the SDH health outcomes in the pediatric health care setting. One concern is that despite the ability to document social determinants in electronic health records (EHRs), compliance is low in recording this data comprehensively.^16–18,22 The purpose of this study was to quantify the availability of SDH data of pediatric ophthalmology patients in a tertiary hospital setting and to identify the association between SDH and patient attendance to scheduled medical visits. The information obtained from this study will help identify the associations between SDH and health outcomes in pediatric ophthalmology and guide future efforts to provide comprehensive support to critical segments of the patient population. 

Material and Methods

Study Design and Objectives

This was a retrospective cohort study of pediatric ophthalmology patients seen at The Hospital for Sick Children (SickKids) between June 1, 2018, and May 23, 2022. This study aimed to determine (i) the availability of SDH data in the electronic medical record of ophthalmology patients managed at SickKids, and (ii) the association of SDH and attendance at clinic and operating room visits overall, and before and after the COVID-19 pandemic. This study was approved by the SickKids Research Ethics Board (REB No: 1000079687) and was conducted in accordance with the guidelines set forth by the Declaration of Helsinki. 

Study Subject Eligibility

Study subjects were included if they resided in Ontario and had at least one scheduled visit (operating room or clinic) within the SickKids Department of Ophthalmology and Vision Sciences (DOVS) during the observation period. 

Data Collection

Data abstracted from the electronic medical record (Epic) included demographics (sex, birthdate, ethnicity, preferred language, need for interpreter, postal code, public insurance status, guardians, legal guardian status, and guardian relationship to study subject), diagnosis (ICD-10 code) and management plan (date of visit, type of visit [operating room or clinic], visit status [completed, rescheduled, canceled, no show], and reasons for canceled or rescheduled visits).²³ 

Data Processing

The neighborhood income quintile for each study subject was deduced using their postal code and the Postal Code Conversion File available via the University of Toronto's CHASS Canadian Census Analyzer through SAS. Neighborhood income was determined using 2016 Census Profile data.^24,25 Urbanicity and geographic region of residence were determined using the forward sortation area of postal code, as defined by the Canada Post Corporation.²⁶ Age at each visit and at the end of study period were calculated for each study subject using date of birth and date of visit or study end date, respectively. Specific reasons for canceled or rescheduled visits were classified as either patient-initiated or provider-initiated. 

Data Analysis

Summary statistics were used to calculate availability of SDH variables and characterize the study cohort; these SDH variables covered the domains of economic stability (neighborhood income quintile), health care access and quality (public insurance coverage), neighborhood and built environment (geographic region in Ontario, urbanicity), and social and community context (age, sex, ethnic group, preferred language, need for interpreter, and guardians) (Fig.). Statistical analyses were carried out using SPSS. A χ² test was used to test for univariate associations between SDH variables with sufficient coverage (age, sex, income quintile, region, and urbanicity) and health outcomes (attendance at clinic or operating room visits). Multivariable regression was used to examine associations between SDH and outcomes. The government of Ontario declared a state of emergency because of the COVID-19 pandemic on March 17, 2020; we use “pre-pandemic” and “post-pandemic” to describe the periods before and after this date.²⁷ 

Results

Availability of SDH Data in the EHR

The study cohort included 26,102 unique study subjects (Table 1). Data on sex, birthdate, and postal code were recorded for 100.0% of study subjects, whereas public insurance status was recorded for 99.1%, preferred language for 85.2%, and ethnic group for 0.1% (Table 1). Postal codes were available to determine geographic region of residence and urban or rural residence status for 100% of study subjects. Neighborhood income quintile could be determined for 99.2% of participants (Table 1). 

Table 1.

 

Coverage and Characterization of Patient Social Determinant of Health Data

 

Study Cohort Characteristics

The study population was 52.5% male and 47.5% female (Table 1). Study subjects most commonly were aged five to 13 years at the end of study period (44.6%), resided in urban areas (93.0%), or lived in Central Ontario (48.9%) (Table 1). The most common preferred language was English (75.9%) (Table 1) and the most common non-English language was Chinese-Mandarin (1.6%) (Supplementary Table S1). Most study subjects (99.1%) were registered with the Ontario Health Insurance Program (Table 1). Most study subjects had two guardians recorded in their medical record (80.1%, Table 1), most commonly a mother and a father (78.5%, Supplementary Table S2). For study subjects with only one guardian recorded (19%, Table 1), it was most commonly a mother (80.0%, Supplementary Table S2). 

Association Between Social Determinants and Operating Room Visit Attendance

Females were more likely to reschedule or cancel operating room visits in comparison to males (P = 0.004) (Table 2). Individuals from rural neighborhoods were more likely to reschedule visits in comparison to individuals from urban neighborhoods (P = 0.006) (Table 2). When controlling for other SDH variables, sex and urbanicity remained significant. Age, neighborhood income quintile, and region of residence were not associated with operating room attendance. Rescheduled operating room visits were most commonly provider-initiated (Supplementary Table S3), but there was no difference observed for canceled operating room visits. There were no associations between SDH and patient-initiated cancellations or rescheduling (Supplementary Table S4). 

Table 2.

 

Association Between Social Determinants of Health and Operating Room Visit Attendance

 

When looking at rescheduled operating room visits pre- and post-pandemic, there was no significant impact of any SDH variable (Table 3). However, living in Metropolitan Toronto or Northern Ontario was associated with a lower likelihood of completing an operating room visit as scheduled post-pandemic (P = 0.02) (Table 4). 

Table 3.

 

Impact of COVID-19 and Social Determinants of Health on Rescheduled Operating Room Visits

 

Table 4.

 

Impact of COVID-19 and Social Determinants of Health on Completed Operating Room Visits

 

Association Between Social Determinants and Clinic Visit Attendance

Clinic visit attendance was associated with study subject's sex (P = 0.002), age (P ≤ 0.001), neighborhood income quintiles (P ≤ 0.001), geographical region of residence (P ≤ 0.001), and urban vs rural residence (P ≤ 0.001) (Table 5). Females were more likely to cancel clinic visits (P = 0.002), whereas males were more likely not to show for clinic visits (P = 0.002). Individuals aged zero to 13 were more likely to cancel and no-show to clinic visits (P ≤ 0.001). Study subjects living in medium neighborhood income quintile were more likely to cancel clinic visits, and individuals from the low-medium neighborhood income quintile were more likely to no-show to clinic visits (P ≤ 0.001). Living in Eastern, Southwestern, and Northern Ontario was associated with a higher likelihood of cancelling clinic visits and living in Metropolitan Toronto was associated with a higher likelihood to no-show to clinic visits (P ≤ 0.001). Residing in a rural area was associated with a higher likelihood to cancel and no-show to clinic visits (P ≤ 0.001) (Table 5). When controlling for other SDH variables, sex, age, income quintile, region of residence, and urbanicity remained significant. Canceled clinic visits were most commonly patient-initiated (60.8%) (Supplementary Table S5). Being aged zero to 13 (P = 0.02), living in a neighborhood with medium-high quintile (P ≤ 0.001), living in Central Ontario, Eastern Ontario, Southwestern Ontario, or Northern Ontario (P ≤ 0.001), and living in a rural area (P ≤ 0.001) were factors associated with patient-initiated clinic visit cancellations (Supplementary Table S6). 

Table 5.

 

Association Between Social Determinants of Health and Clinic Visit Attendance

 

Post-pandemic, study subjects older than 13 (P = 0.05) and living in Metropolitan Toronto or Central Ontario (P = 0.002) were likelier to no show at clinic visits (Table 6). Similarly, study subjects older than 13 (P = 0.05) or living in Central or Northern Ontario (P = 0.03) were more likely to cancel clinic visits (Table 7). Conversely, study subjects aged zero to 13 (P < 0.001), living in a medium-high neighborhood income quintile (P = 0.002), or living in Central Ontario (P < 0.001) were more likely to complete clinic visits as scheduled (Table 8). 

Table 6.

 

Impact of COVID-19 and Social Determinants of Health on Clinic Visit No Shows

 

Table 7.

 

Impact of COVID-19 and Social Determinants of Health on Clinic Visit Cancellations

 

Table 8.

 

Impact of COVID-19 and Social Determinants of Health on Clinic Visit Completions

 

Discussion

Inequities in eye care and disparities in visual health are significantly influenced by SDH.^11,28 Health care professionals are increasingly being called upon to address SDH in clinical practice, for example, by adopting a biopsychosocial approach to care; this rests on the systematic and comprehensive collection of SDH variables to inform such practice.²⁹ Our study revealed that in a Canadian tertiary pediatric ophthalmology health care setting, documentation of SDH in the electronic health record was limited, with very low rates of documentation for social variables such as ethnicity, preferred language, use of language interpreter and legal guardian status, despite having dedicated fields in the EHR platform for this information. Ethnicity, primary language, and family structure are important influences of health outcomes, yet we were unable to perform analyses with them because of the low availability of data. Most of the analysis in this study was therefore based on sex, age, and SDH that could be deduced from residential postal code. 

Our study examined attendance or rescheduling of operating room or clinic visits as a measure of adherence to care, and assessed how SDH might influence these outcomes. When comparing the association between SDH and operating room visits and clinic visits we found that fewer SDH disrupt operating room visit attendance. Because of the complicated and sensitive nature of the surgical care, operating room visits in pediatric ophthalmology are frequently more severe and carefully adhered to than routine clinic visits.³⁰ We noted that female sex and urbanicity were associated with cancellations and rescheduled operating room visits. Females were more likely to reschedule or cancel operating room visits than males, suggesting that they may require additional attention or supports to ensure adherence to planned care. Other studies have shown that surgical visit cancellation occurs because of the menstrual cycle of females;³⁰ however, it is unclear if that is the case here, because older female age did not appear to be a factor in our cohort. Still, it may be important for health care providers to be aware of the potential impact of sex on adherence to scheduled visits. Living in a rural area was associated with rescheduled operating room visits in our study; this is consistent with other studies indicating that children in rural or remote areas face barriers such as long distances or lack of transportation and accommodation options when seeking care.³¹ Patients from rural areas may face challenges finding suitable accommodation, which can be particularly problematic for when surgeries and related follow up require an overnight stay.³² In addition, rural residents often have responsibilities related to farming, family, or community commitments that make it harder to adhere to scheduled surgery dates.³³ No differences in cancellations and rescheduling of operating room visits were seen in the pre- and post-pandemic phases of our analysis. The absence of disparities in cancellations and rescheduling of operating room visits between the pre- and post-pandemic periods could potentially be attributed to the critical nature of these procedures; it is worth noting that during the pandemic hospitals in Ontario were advised to carry out essential surgical care and postpone only nonessential care.³⁴ 

Clinic visits that were not completed as originally scheduled were categorized as cancellations (advance notice provided) and no-shows (no advance notice provided). Our data indicated that urbanicity and ages between zero to 13 were predictors of both cancellations and no-shows. Cancellations were additionally associated with female sex, medium neighborhood income quintile, and living in Eastern, Southwestern, or Northern Ontario; in contrast, no-shows were associated with male sex, low-medium neighborhood income quintile, and living in Metropolitan Toronto. Although our study did not tease out causative factors for these associations, it is possible that younger age cohorts are more prone to or deal with unexpected emergencies or illnesses that often arise in young children.³⁵ It is also possible that young children's fears about medical procedures may lead to increased cancellations or no-shows.³⁵ Individuals from lower neighborhood income quintiles may face barriers to attending scheduled clinic visits (e.g., unpredictable working hours/shift work, unreliable transport options, etc.), often arising at the last minute and therefore resulting in no-shows. 

Our study results, though limited, are a starting point for addressing the SDH in pediatric ophthalmology. Experts agree that improved data sets on SDH is a necessary step to improve eye health equity.^11,17 A framework has been proposed for ophthalmologists to address social determinants of vision health in their practice, consisting of five steps: awareness (screening for social needs), assistance (putting patients in touch with social care resources), adjustment (adjusting clinical care to take into account social needs), alignment (appreciating social assets and working with community organizations), and advocacy (promoting policy).¹¹ The first step, systematic screening for social needs, is inextricably linked to systematic collection of data, which we found lacking in our setup. Although streamlined intake forms and online patient portals that allow data collection directly from patients may improve workflow and overall quality of care, patients may have privacy concerns or not have enough time to complete intake forms in clinical settings. In addition, marginalized patient groups who have been hurt by research and medical communities in the past may distrust health facilities and be reluctant to contribute data on ethnicity or race without knowing why the data are being collected and how the data will be used to improve their health.^30,31 Indeed, individuals may feel that revealing such sensitive data may put them at higher risk of receiving poorer quality of health care.³⁶ Health care professionals and researchers must address these concerns in light of historical wrongs, ongoing systemic racial discrimination, and societal inequities.³⁷ 

The main limitations of our study were that it was retrospective in nature and limited to data captured in discrete fields within the EHR. The low availability of SDH data may be reflective of poor practice by health care teams to collect this data or reluctance of patients to provide this sensitive data; our study was unable to ascertain the underlying reasons. It is possible that better-quality data could have been collected directly from the patient in a prospective study or that additional SDH could have been extracted by analyzing the free-text narrative fields of the EHR (although this was infeasible for a cohort of ∼26,000 study subjects). We also note the possibility that there may have been inconsistencies in visit documentation (i.e., a cancellation might have been rescheduled but not documented as such) or reasons for cancellation (i.e., reasons may not have been documented accurately or at all) because the data was manually recorded by hospital staff. Last, we recognize that data deduced from postal code data may not be reflective of individual patients socioeconomic situation. 

Understanding how SDH affect eye health outcomes is of growing interest. However, our study indicated that coverage of SDH within EHRs was highly variable and that there was a significant quantity of missing data. Despite this, our study suggests that SDH including sex, age, neighborhood income quintile, geographic region of residence, and urban versus rural living status warrant further investigation to deduce their influence on attendance to scheduled operating room and clinic visits, as well as to their impact on overall health outcomes. Promoting access to quality eye care for all children, regardless of their socioeconomic status or other factors, is an important step towards achieving child health equity. Future studies should focus on the construction of relevant datasets through which social determinants of health can be quantified and measured. 

Acknowledgments

Omer Jamal was supported by a University of Toronto Vision Science Research Program (VSRP) scholarship and by an Edwin S.H. Leong Centre for Healthy Children MSc Restracomp Scholarship. The sponsors had no role in the design or conduct of this research. 

Disclosure: O. Jamal, None; A. Mallipatna, None; S.W. Hwang, None; H. Dimaras, None