Introduction

Despite evidence supporting the effectiveness of diabetic retinopathy screening (DRS) in reducing the risk of sight loss, attendance for screening in particular demographic groups is often below recommended levels.1 Although adherence to screening recommendations is generally high in countries with universal health insurance coverage, there is considerable variability within population subgroups. For example, in Australia, DRS uptake in non-indigenous Australians was reported to be 53% compared with 78% of indigenous Australians.2 Only 38% of recent immigrants to Canada had at least one screening visit within 1 year of diagnosis compared with 51% of long-term residents.3

Understanding modifiable barriers and enablers to DRS is essential to develop tailored intervention strategies to improve screening uptake. There have been many studies internationally that have investigated the factors influencing DRS attendance.4 5 Barriers/enablers to attendance potentially operate at different levels, including the person with diabetes, the healthcare professional (HCP) or the healthcare system. Furthermore, factors influencing individual screening attendance are likely to differ according to the presence of variables that are known to impact on health equity, for example, type of diabetes, ethnicity or socioeconomic status.6–9 However, studies have often considered people with diabetes as a homogeneous group and relatively few studies have addressed barriers/enablers in particular population subgroups.

Another demographic group where adherence to DRS consistently falls below recommended levels is young adults (YAs) with diabetes aged under 35 years.10–13 Recent studies from the UK Diabetic Eye Screening Program (DESP) have shown that the time interval from registration with the screening program to DRS attendance is strongly age dependent.10 12 Time to first screening was significantly longer for the 18–34 year age group, with 70% meeting the requirement for routine annual screening and approximately 20% remaining unscreened 3 years after registration.12 In terms of ongoing attendance, it was reported that younger adults (<35 years) have a 75% reduction in the odds of attending annual DRS compared with those aged 60 years and over.12 Furthermore, younger adults are more likely to miss three successive DRS appointments and present with sight-threatening retinopathy.13

YAs are a particularly hard to reach group and there has been little previous research to understand the reasons for poor DRS attendance in YAs.14 A 2017 Australian study15 conducted semistructured interviews with YAs, n=10 aged 18–39 years and older adults, n=20 aged over 40 years with type 2 diabetes (T2D). This study used a behavioral science framework, the Theoretical Domains Framework (TDF),16 to explore the wide range of barriers and enablers to attendance. The TDF synthesizes constructs from 33 theories of behavior change into 14 domains, representing individual, sociocultural and environmental influences on behavior (eg, knowledge, emotions, social and professional identity, perceived consequences, intention, environmental context and resources). Although younger and older adults shared several screening behavior determinants, a number of TDF domains showed greater salience to YAs including misconceptions regarding diabetic retinopathy (Knowledge); social comparison with others; unrealistic optimism and perceived invulnerability (Beliefs about consequences); and lack of time and financial resources (Environmental context and resources).15 We have recently completed the National Institute for Health and Care Research (NIHR)-funded ‘Enabling diabetic RetinOpathy Screening: Mixed methods study of barriers and enablers to attendance (EROS study)’, which aimed to identify barriers and enablers to DRS attendance experienced by YAs with diabetes living in the UK. As part of this research, we conducted qualitative interviews with 29 YAs with type 1 diabetes (T1D) aged 18–34 years.17 We have also conducted a cross-sectional survey of HCPs working in the UK National Diabetic Eye Screening Program.18 We similarly applied the TDF to identify modifiable barriers and enablers to DRS attendance. In terms of the interviews, key influences fell within the TDF domains: Knowledge, for example, not understanding reasons for attending DRS or treatments available if diabetic retinopathy is detected; Social support, for example, lack of support following DRS results; Social role and identity, for example, not knowing other people their age with diabetes; feeling ‘isolated’ and being reluctant to disclose their diabetes; Environmental context and resources, for example, lack of appointment flexibility and options for rescheduling; and Emotion, for example, diabetes distress/burnout. Enablers included Social influences, for example, support of family/diabetes team; and Goals, for example, DRS regarded as ‘high priority’. Barriers/enablers were generally consistent across groups defined by patterns of attendance (regular attenders, occasional non-attenders, regular non-attenders).

In the current study, we used the results of the previous interview study17 to further explore the research problem. Themes from the qualitative data informed the design of an online survey. The purpose of the survey was threefold: (1) to assess the generalizability of the perceived barriers and enablers in a larger and more diverse sample of YAs with regard to particular demographic characteristics (eg, age, employment, gender, ethnicity, educational level); (2) to enable a qualitative investigation of differences in perceived barriers and enablers between YAs with T1D and T2D; and (3) to investigate differences in those that attend DRS regularly versus those who did not. The survey also allowed us to triangulate findings from qualitative and quantitative methods to gain a more complete picture of the factors that influence screening uptake in YAs.19

Research design and methodsDesign

A cross-sectional web-based survey.

Participants and recruitment strategy

Eligible participants included YAs aged 18–34 years with diabetes. Previous studies have shown that people in this age group are least likely to attend DRS and have high rates of referable retinopathy.10 12 13 As this was a descriptive survey, we did not have a predefined target sample size in mind and aimed to maximize response rate from as many YAs as possible. Two recruitment strategies were used:

A text (SMS) message with the link to the survey was sent to all YAs with T1D and T2D aged 18–34 on the register of a large DESP in London whose mobile number was available. Screening providers often use this mobile phone strategy to request feedback from patients about the care they receive.

The survey was also promoted via the web pages of the Juvenile Diabetes Research Foundation UK (JDRF UK) and Diabetes UK and further supported using the Facebook and Twitter accounts of these organizations (see online supplemental file S1 for examples of promotional material).

Materials: questionnaire

The full survey is available in online supplemental file S2. In brief, the survey was developed based on guidance for conducting surveys using the TDF20 and the findings of our previous interview study with YAs in the UK.17 The survey was fully anonymous and divided into three sections:

Section 1—Participant demographics: age, gender, ethnicity, geographical location, highest level of education; type and duration of diabetes; screening appointments missed in the last 3 years (either forgotten and rescheduled or deliberately not attended) (12 questions).

Section 2—Perceived influences on DRS attendance: Participants were presented with 30 belief statements representing barriers and enablers to DRS attendance. These statements were developed based on the inductively generated themes based on frequency and elaboration from our semistructured interview study with YAs in the UK17 (eg, the theme ‘Diabetic retinopathy is a concern’ was reflected in the belief statement ‘I worry about diabetic retinopathy’)). To ensure theoretical coverage and that the wide range of potential influences were considered, belief statements covered 13 of the 14 TDF domains (our earlier qualitative study17 did not identify themes for the domain Optimism). Participants rated their agreement with each statement using a 5-point Likert scale (strongly agree→strongly disagree).

Section 3: Free text question, ‘Please describe any other factors which influence your attendance at diabetic eye screening which we have not covered’.

To assess participant burden, clarity of questions and face validity, a draft questionnaire was sent to the project Patient and Public Involvement (PPI) panel consisting of four YAs with diabetes, who were asked to comment on the following:

How long the survey takes to complete.

If the survey items make sense, are appropriate and relevant.

If there are any survey items which are ambiguous or unclear.

A final version based on the feedback was uploaded and pretested for technical quality prior to distribution.

Procedure

The survey took place in June 2021. The questionnaire was hosted online using Qualtrics survey software (https://www.qualtrics.com/uk/core-xm/survey-software/). The survey was fully anonymous, and participants consented to participate in the survey by completing a brief consent form on the survey home page. Respondents were offered an incentive in the form of the chance to win a £20 Love2shop voucher (we offered twenty £20 vouchers).

Analysis

After the closure of the survey, all data were imported into an Excel spreadsheet. Data were summarized using descriptive statistics (percentages (n)). For the responses to the 5-point scale in Section 2, scores for ‘Strongly agree’ and ‘Somewhat agree’ were combined into an overall mean and percentage agreement score.

Statistical analysis for comparison between participants with T1D and T2D and pattern of attendance (regular attenders vs occasional non-attenders) was carried out using MedCalc statistical software V.18 (MedCalc Software, Ostend, Belgium; http://www.medcalc.org) in the form of χ2 tests to determine differences in endorsed barriers and enablers. Occasional non-attenders were grouped as those who had either (1) unintentionally forgotten/missed previous appointments and rescheduled, or (2) actively chosen to not attend on at least one occasion.

A single investigator (JGL) coded each line of responses to the free-text question by adding a descriptive label. Codes representing similar thematic topics were then grouped together to identify common themes and concepts. A second investigator (FL) reviewed the thematic groupings and any discrepancies were resolved by discussion.

ResultsRespondent demographic characteristics

One hundred and two responses were received. Detailed respondent demographic characteristics are presented separately for respondents with T1D and T2D in table 1 (two participants reported having an ‘other’ type of diabetes (eg, maturity onset diabetes of the young). Almost all respondents (98.5%) were located in England. The greatest proportion of respondents were female (59.8%), had T1D (65.7%), were aged 30–34 (52.9%), identified as white British (57.8%), employed (76.4%) and educated to degree level or higher (60.8%).

Table 1

Characteristics of persons with type 1 and type 2 diabetes

Most respondents were regular attenders for eye screening, with 76.5% having not missed a DRS appointment in the last 3 years.

Barriers and enablers

The results of agreement/disagreement with belief statements related to barriers/enablers to DRS are presented in table 2. In this well-engaged population, the most salient TDF domain for DRS attendance was Goals, with 93% of respondents agreeing that DRS was a high priority in terms of diabetes management, with a clear intention to attend future eye screening appointments (TDF domain, Intention) (97%). The reason for attending was understood by almost all respondents, with 98% being aware that DRS can help detect eye problems early (‘Knowledge’). Conversely, 84% did not know what treatments were available if retinopathy was detected (Knowledge). Importantly, only 52% felt that diabetes education and training covered eye screening in detail (Skills).

Table 2

Mean scores and percentage agreement with belief statements representing barriers and enablers to diabetic retinopathy screening (DRS)

In terms of the screening process itself, overall, 67.4% indicated that they would like to be offered more options for appointment days/times when booking DRS appointments (Environmental context/resources) and approximately one-third of respondents had difficulties getting time off work/study to attend DRS appointments (Environmental context/resources). Scheduling was compounded by retinopathy screening appointments being at a different time to other diabetes appointments for most respondents (Environmental context/resources) (75.2%). Comparison between regular attenders and occasional non-attenders (table 2) found that non-attenders were more likely to report more difficulty taking time off to attend appointments (59.1% vs 23.4%, p=0.002) and felt that appointments took up too much of the day (50.0% vs 23.1%, p=0.015).

Additional negative aspects of the screening process were acknowledged, for example, being upset by seeing older people with worse complications in the waiting room (Emotion) (40%) and the adverse effects of the dilating eye drops (Beliefs about consequences) (64%). Most respondents were worried about diabetic retinopathy (74.3%), reported anxiety when receiving screening results (Emotion) (63%) and would like more support and information after getting their results (Social influences) (66%).

Enablers of attendance included an awareness that DRS allows early detection of eye problems (Beliefs about consequences) (98%), feeling reassured by attending DRS (Emotion (75%) and feeling comfortable disclosing diabetes to others (Social identity) (71.6%).

Percentage agreement between YAs with T1D and T2D were broadly similar (table 2). Statistically significant differences were found for only three statements. Persons with T1D were more likely to feel overwhelmed by their diabetes (76.9% vs 56.7%, p=0.046) (Emotion). YAs with T2D were less comfortable in disclosing their diabetes to others (77.6% vs 57.6%, p=0.039) (Social identity) and were more likely have developed strategies to help them to remember to attend their DRS appointments (63.6% vs 37.9%, p=0.019) (Behavioral regulation).

Other factors influencing attendance at DRS

Free text responses were received from 30 respondents. These covered the following areas: impact of the COVID-19 pandemic on scheduling DRS appointments (‘Last appointment later due to COVID-19’); fear of vision loss (‘My mother in law lost her eye due to diabetic (retinopathy)’); appointment inflexibility (‘They can also be inflexible-the nearest one to me only does Tuesday mornings’); impact of eye drops and transport issues (‘They can be difficult to get to, especially as you cannot drive-in one occasion I had to get three buses which took nearly 2 hours-the return trip with dilated pupils wasn’t fun’), interactions with screening staff and issues with receiving screening results (‘I always receive letters that are extremely distressing and usually on a weekend when I cannot call anyone’). The complete set of free text comments can be found in the online supplemental material.

DiscussionSummary of findings

This study aimed to build on the findings of our earlier qualitative interview study to identify perceived barriers and enablers to DRS in YAs in the UK.17 The results broadly confirmed our qualitative interview findings in YAs with T1D and converged with the findings from a qualitative study of YAs with T2D in Australia.15

Based on the level of agreement with each belief statement, the most salient TDF domains associated with DRS included ‘Social influences’, ‘Intentions’, ‘Emotion’, ‘Environmental context/resources’, ‘Knowledge’, ‘Skills’, and ‘Goals’. However, the level of agreement with belief statements differed between respondents with T1D and T2D and between regular attenders and occasional non-attenders. This emphasizes the importance of understanding barriers/enablers for specific population subgroups.

Overall, survey respondents represented a well-engaged population, with approximately 78% reporting that they had not missed any screening appointments in the last 3 years. This may explain the high level of understanding of the purpose of DRS, the high priority (Goals) given to this particular aspect of diabetes care and the strong intention to attend further screening appointments. However, despite a large majority understanding the purpose of DRS and the need for regular screening, there were specific knowledge gaps such as an awareness of the treatments available should sight-threatening retinopathy be detected.

Competing time demands and practical issues with making appointments have been previously shown to be important barriers to DRS attendance.4 9 15 17 This was particularly pertinent among respondents who reported occasionally missing DRS appointments. Occasional non-attenders were more likely to agree with the belief statements: ‘Diabetic eye screening appointments take up too much time in my day’ and ‘I am unable to get time off work/study to attend diabetic eye screening appointments’. Most survey respondents agreed that their DRS appointments were at a different time and place to their other diabetes appointments. Possible solutions could include evening and weekend appointments coupled with a flexible online booking system to allow them to schedule appointments more easily. Furthermore, the use of an integrated diabetes care in the form of a ‘one-stop shop’ clinic that combines several processes of diabetes care, including DRS, which would reduce the number of appointments. Although there is currently no high quality evidence from the UK that integrated diabetes clinics improves DRS uptake specifically in YA, ‘collaborative case management’, which coordinates processes of diabetes care, has been shown to improve diabetic retinopathy outcomes in clinical trials of a general adult population with diabetes.1

The eye drops used to temporarily dilate the pupils were perceived to be unpleasant by the majority of respondents. Free text comments also alluded to the impact of the eye drops on attendance due to not being able to drive to and from the screening venue and having to rely on family members or use public transport (‘With the drops you can’t drive but it’s also hard to then see where the train is. The drops knock me out for the rest of the day and really affect work and everything’). Previous studies have also reported on barriers relating to dilating eye drops.4 9 The National Screening Committee in the UK currently recommends pupil dilation (mydriasis) for all attendees for DRS based on the ease of organization and improvement in image quality; however, there is evidence that using mydriasis only when clinically necessary can be effective for DRS.21

Fear of diabetic retinopathy was identified as a cause for concern, with a high level of agreement that screening attendance provided reassurance. However, there was a particular anxiety associated with receiving screening results and a desire for more support and information on receiving results. The previous literature identified that while, for some, the fear of losing vision is a strong incentive to attend DRS, for others, the fear of a diagnosis of diabetic retinopathy may act as a barrier.17 Interventions to address this could include training suitably qualified screeners to give immediate feedback on the results of eye screening or the provision of more support after receiving results. This strategy is potentially acceptable to implement in practice, with a recent cross-sectional survey of HCPs working in the UK National Diabetic Eye Screening Program highlighting that screening providers would like to be more involved in discussing screening results with YAs and promoting diabetes self-management.18

A recommendation by a HCP has been shown to be an important enabler for DRS uptake and receiving a recommendation from a healthcare provider to attend screening is associated with improved attendance.4 15 However, survey respondents reported that members of the diabetes team did not always check their DRS attendance record or encourage them to attend. This is clearly a missed opportunity to improve screening uptake in this population. Furthermore, the lack of integration of DRS with other diabetes services creates a further missed opportunity for screening providers to discuss general diabetes self-management, due to a lack of information on how attendees are managing their diabetes.18

T2D is increasingly prevalent in YAs. This trend is particularly pronounced in South Asian ethnic groups.22 23 Previous research has established that YAs with T2D are at higher risk of developing diabetic retinopathy,24 25 face unique barriers to diabetes self-management and have specific unmet psychosocial needs.26 Although all participants in our earlier interview study17 had T1D, 32% of respondents in the current study were YAs with T2D. The levels of agreement between these subgroups were very similar, with significant differences in agreement found for only three belief statements (‘I sometimes feel overwhelmed due to my diabetes’; ‘I feel comfortable disclosing to others that I have diabetes’; ‘I have developed strategies to help me remember to attend my diabetic eye screening appointments’). Respondents with T1D were more likely to feel overwhelmed by their diabetes. YAs with T2D were less comfortable in disclosing their diabetes to others. Research in Australia suggests that young people with T2D are sensitive to stigmatizing attitudes.27 Although there was no difference between respondents with T1D and T2D in terms of likelihood of forgetting to attend DRS appointments, respondents with T2D were more likely to have developed strategies to help them to remember to attend.

Strengths and limitations of the current study

One of the strengths of the current study is that it addresses an important evidence gap. Although there are many studies that have reported modifiable barrier/enablers to DRS,2 3 the majority of these studies tend to treat people with diabetes as a homogeneous group, and therefore, it is not possible to identify determinants of DRS uptake from the perspective of particular population subgroups. Relatively few studies have reported barriers from the perspective of YAs, who are at a high risk of developing sight-threatening retinopathy.12 24 25

Another strength of our approach is the use of a theory-informed methodology to identify barriers and enablers.14 We used the TDF to guide data collection, which provides a basis for generating future behavior change strategies that can be tailored to YA to address barriers or enhance facilitators.

Although this hypothesis-generating study was limited in terms of its small sample size, we received responses from a demographically diverse sample of YA, including 33% of responses from YA with T2D. The results confirmed many of barriers and enablers identified in previous qualitative interview studies15 17 and suggest that the determinants of screening attendance are broadly similar for YAs with T1D and T2D.

The main limitation was the difficulty experienced in recruiting non-attenders. Despite using a variety of recruitment strategies (see online supplemental material), only 20% of survey respondents had missed a DRS appointment in the last 3 years and nearly all of these had unintentionally missed the appointment (ie, they forget or were unable to attend). We were only able to recruit two participants who had made a deliberate decision not to attend DRS, which could impact on the generalizability of the findings to repeat non-attenders. Considering the study aimed to identify barriers and enablers of DRS attendance among YAs in the UK, a further limitation is that 98.5% of respondents were from England.