[Skip to Content]
[Skip to Content Landing]
Table 1.  
Baseline Characteristics of Participantsa
Baseline Characteristics of Participantsa
Table 2.  
Associations Between Perceived Barriers to Diabetes Self-management and Presence and Severity of Diabetic Retinopathy (DR)
Associations Between Perceived Barriers to Diabetes Self-management and Presence and Severity of Diabetic Retinopathy (DR)
1.
Cheung  N, Mitchell  P, Wong  TY.  Diabetic retinopathy.  Lancet. 2010;376(9735):124-136.PubMedGoogle ScholarCrossref
2.
Holman  RR, Paul  SK, Bethel  MA, Matthews  DR, Neil  HA.  10-Year follow-up of intensive glucose control in type 2 diabetes.  N Engl J Med. 2008;359(15):1577-1589.PubMedGoogle ScholarCrossref
3.
Nathan  DM, Genuth  S, Lachin  J,  et al; Diabetes Control and Complications Trial Research Group.  The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus.  N Engl J Med. 1993;329(14):977-986.PubMedGoogle ScholarCrossref
4.
Genuth  S, Eastman  R, Kahn  R,  et al; American Diabetes Association.  Implications of the United Kingdom Prospective Diabetes Study.  Diabetes Care. 2003;26(suppl 1):s28-s32.PubMedGoogle ScholarCrossref
5.
Huang  OS, Lamoureux  EL, Tay  WT, Tai  ES, Wang  JJ, Wong  TY.  Glycemic and blood pressure control in an Asian Malay population with diabetes and diabetic retinopathy.  Arch Ophthalmol. 2010;128(9):1185-1190.PubMedGoogle ScholarCrossref
6.
Hiss  RG.  Barriers to care in non–insulin-dependent diabetes mellitus: the Michigan experience.  Ann Intern Med. 1996;124(1, pt 2):146-148.PubMedGoogle ScholarCrossref
7.
Corabian  P, Harstall  C.  Patient Diabetes Education in the Management of Adult Type 2 Diabetes. Edmonton, AB, Canada: Alberta Heritage Foundation for Medical Research; 2001.
8.
Munshi  MN, Segal  AR, Suhl  E,  et al.  Assessment of barriers to improve diabetes management in older adults: a randomized controlled study.  Diabetes Care. 2013;36(3):543-549.PubMedGoogle ScholarCrossref
9.
Nam  S, Chesla  C, Stotts  NA, Kroon  L, Janson  SL.  Barriers to diabetes management: patient and provider factors.  Diabetes Res Clin Pract. 2011;93(1):1-9.PubMedGoogle ScholarCrossref
10.
International Diabetes Federation.  IDF Diabetes Atlas. 7th ed. Brussels, Belgium: International Diabetes Federation; 2015.
11.
Yau  JW, Rogers  SL, Kawasaki  R,  et al; Meta-Analysis for Eye Disease (META-EYE) Study Group.  Global prevalence and major risk factors of diabetic retinopathy.  Diabetes Care. 2012;35(3):556-564.PubMedGoogle ScholarCrossref
12.
Man  RE, Sabanayagam  C, Chiang  PP,  et al.  Differential association of generalized and abdominal obesity with diabetic retinopathy in Asian patients with type 2 diabetes.  JAMA Ophthalmol. 2016;134(3):251-257.PubMedGoogle ScholarCrossref
13.
Brooke  P, Bullock  R.  Validation of a 6 item cognitive impairment test with a view to primary care usage.  Int J Geriatr Psychiatry. 1999;14(11):936-940.PubMedGoogle ScholarCrossref
14.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
15.
Willis  GB.  Cognitive Interviewing: A Tool for Improving Questionnaire Design. Los Angeles, CA: Sage Publications; 2004.
16.
Collins  D.  Pretesting survey instruments: an overview of cognitive methods.  Qual Life Res. 2003;12(3):229-238.PubMedGoogle ScholarCrossref
17.
Linacre  JM.  A User’s Guide to Winsteps: Rasch-Model Computer program. Chicago, IL: MESA press; 2002.
18.
Linacre  JM.  A User’s Guide to Winsteps/Ministeps Rasch-Model Programs. Chicago, IL: MESA press; 2005.
19.
Andrich  D.  A rating formulation for ordered response categories.  Psychometrica. 1978;43(4):561-573. doi:10.1007/BF02293814Google ScholarCrossref
20.
Nguyen  TH, Han  HR, Kim  MT, Chan  KS.  An introduction to item response theory for patient-reported outcome measurement.  Patient. 2014;7(1):23-35.PubMedGoogle ScholarCrossref
21.
Early Treatment Diabetic Retinopathy Study Research Group.  Grading diabetic retinopathy from stereoscopic color fundus photographs—an extension of the modified Airlie House classification. ETDRS report number 10.  Ophthalmology. 1991;98(5)(suppl):786-806.PubMedGoogle ScholarCrossref
22.
Gupta  P, Sidhartha  E, Tham  YC,  et al.  Determinants of macular thickness using spectral domain optical coherence tomography in healthy eyes: the Singapore Chinese Eye Study.  Invest Ophthalmol Vis Sci. 2013;54(13):7968-7976.PubMedGoogle ScholarCrossref
23.
Yeo  CP, Tan  CH, Jacob  E.  Haemoglobin A1c: evaluation of a new HbA1c point-of-care analyser Bio-Rad in2it in comparison with the DCA 2000 and central laboratory analysers.  Ann Clin Biochem. 2009;46(pt 5):373-376.PubMedGoogle ScholarCrossref
24.
Millán  J, Pintó  X, Muñoz  A,  et al.  Lipoprotein ratios: physiological significance and clinical usefulness in cardiovascular prevention.  Vasc Health Risk Manag. 2009;5:757-765.PubMedGoogle Scholar
25.
Norris  SL, Engelgau  MM, Narayan  KM.  Effectiveness of self-management training in type 2 diabetes: a systematic review of randomized controlled trials.  Diabetes Care. 2001;24(3):561-587.PubMedGoogle ScholarCrossref
26.
Sherifali  D, Bai  JW, Kenny  M, Warren  R, Ali  MU.  Diabetes self-management programmes in older adults: a systematic review and meta-analysis.  Diabet Med. 2015;32(11):1404-1414.PubMedGoogle ScholarCrossref
27.
Fisher  L, Gonzalez  JS, Polonsky  WH.  The confusing tale of depression and distress in patients with diabetes: a call for greater clarity and precision.  Diabet Med. 2014;31(7):764-772.PubMedGoogle ScholarCrossref
28.
Schillinger  D, Grumbach  K, Piette  J,  et al.  Association of health literacy with diabetes outcomes.  JAMA. 2002;288(4):475-482.PubMedGoogle ScholarCrossref
29.
Egede  LE, Ellis  C.  The effects of depression on diabetes knowledge, diabetes self-management, and perceived control in indigent patients with type 2 diabetes.  Diabetes Technol Ther. 2008;10(3):213-219.PubMedGoogle ScholarCrossref
30.
Lin  EH, Katon  W, Von Korff  M,  et al.  Relationship of depression and diabetes self-care, medication adherence, and preventive care.  Diabetes Care. 2004;27(9):2154-2160.PubMedGoogle ScholarCrossref
31.
Fenwick  EK, Pesudovs  K, Rees  G,  et al.  Republished article: the impact of diabetic retinopathy: understanding the patient’s perspective.  Postgrad Med J. 2012;88(1037):167-175.PubMedGoogle ScholarCrossref
32.
Chen  X, Lu  L.  Depression in diabetic retinopathy: a review and recommendation for psychiatric management.  Psychosomatics. 2016;57(5):465-471.PubMedGoogle ScholarCrossref
33.
Prieto  L, Alonso  J, Lamarca  R.  Classical Test Theory versus Rasch analysis for quality of life questionnaire reduction.  Health Qual Life Outcomes. 2003;1:27.PubMedGoogle ScholarCrossref
34.
Petrillo  J, Cano  SJ, McLeod  LD, Coon  CD.  Using classical test theory, item response theory, and Rasch measurement theory to evaluate patient-reported outcome measures: a comparison of worked examples.  Value Health. 2015;18(1):25-34.PubMedGoogle ScholarCrossref
35.
Simmons  D, Lillis  S, Swan  J, Haar  J.  Discordance in perceptions of barriers to diabetes care between patients and primary care and secondary care.  Diabetes Care. 2007;30(3):490-495.PubMedGoogle ScholarCrossref
36.
Chin  MH, Cook  S, Jin  L,  et al.  Barriers to providing diabetes care in community health centers.  Diabetes Care. 2001;24(2):268-274.PubMedGoogle ScholarCrossref
Original Investigation
December 2017

Association Between Perceived Barriers to Diabetes Self-management and Diabetic Retinopathy in Asian Patients With Type 2 Diabetes

Author Affiliations
  • 1Singapore Eye Research Institute, Singapore National Eye Centre, Singapore
  • 2Duke-National University of Singapore Medical School, Singapore
  • 3Centre for Eye Research Australia, University of Melbourne, Victoria, Australia
  • 4Singapore National Eye Centre, Singapore
JAMA Ophthalmol. 2017;135(12):1387-1393. doi:10.1001/jamaophthalmol.2017.4888
Key Points

Question  What is the association between a patient’s perceived barriers to diabetes self-management and the severity spectrum of diabetic retinopathy in Asians with type 2 diabetes?

Findings  In this cross-sectional clinic-based study, greater perceived barriers to diabetes self-management were associated with higher odds of having any diabetic retinopathy, mild to moderate diabetic retinopathy, or severe diabetic retinopathy independent of hemoglobin A1c, blood pressure, and lipid levels.

Meaning  These data suggest that interventions to reduce patient-, practitioner-, and system-related barriers to diabetes care may help reduce the risk of diabetic retinopathy.

Abstract

Importance  A patient’s perceived barriers to diabetes self-management (DSM) may affect his or her risk of diabetic retinopathy (DR); however, few studies have examined this association.

Objective  To examine the association between perceived barriers to DSM and the severity spectrum of DR in Asian patients with type 2 diabetes.

Design, Setting, and Participants  A cross-sectional clinic-based study, the Singapore Diabetes Management Project, was conducted from December 28, 2010, to March 20, 2013, at the Singapore National Eye Centre, a tertiary eye care institute. After excluding patients with type 1 diabetes and ungradable fundus images, 361 participants were included in the analyses. Statistical analysis was conducted from July 20 to September 8, 2017.

Exposure  The degree of perceived barriers to DSM was assessed using a 23-item questionnaire comprising items about knowledge of DSM, access to care, and confidence in health care professionals. Rasch analysis was used to optimize the scale’s psychometric properties, with lower scores indicating a higher degree of self-perceived barriers.

Main Outcomes and Measures  Diabetic retinopathy was graded from 2-field retinal images into categories of no DR (Early Treatment Diabetic Retinopathy Study levels 10-15; n = 154), mild to moderate DR (levels 20-43; n = 112), and severe DR (levels ≥53 and/or presence of clinically significant macular edema; n = 95) using the modified Airlie House classification system of DR. Multinomial logistic regression models were used to assess the association between perceived barriers and severity of DR in the worse-affected eye.

Results  Among the 361 participants (105 women and 256 men; mean [SD] age, 57 [8] years), a greater magnitude of perceived barriers to DSM was independently associated with higher odds of having any DR (odds ratio, 1.32; 95% CI, 1.06-1.66), mild to moderate DR (odds ratio, 1.30; 95% CI, 1.01-1.68), and severe DR (odds ratio, 1.36; 95% CI, 1.03-1.79). This association was independent of diabetes control (hemoglobin A1c, blood pressure, and lipid levels), presenting visual acuity, and socioeconomic indicators.

Conclusions and Relevance  These results suggest that greater perceived barriers to DSM are independently associated with severity of DR. Although longitudinal data are needed, these findings suggest that evidence-based interventions to reduce patient-, practitioner-, and system-related barriers to diabetes care may help reduce the risk of DR.

Introduction

Diabetic retinopathy (DR), the most common visual complication of diabetes,1 can be delayed or prevented by optimal control of glucose, blood pressure (BP), and lipid levels.2-4 Optimal rates of diabetes control are, however, low in individuals with diabetes. For instance, in a population-based study, only 17.4% of ethnic Malays with DR met the recommended glycemic control threshold, and 10.3% met the BP control threshold.5 As shown previously, a large part of poor control of diabetes may be attributed to barriers to diabetes self-management (DSM),6 which is defined as a process in which the knowledge, skills, and abilities required for a patient to adequately manage his or her diabetes are facilitated.7

Previous studies have characterized many perceived barriers to DSM via qualitative methods, such as focus groups and semistructured interviews.6,8,9 However, it is difficult to quantify the independent effect of these perceived barriers on key clinical outcomes using qualitative data without accounting for confounding factors (eg, sex and duration of diabetes). To our knowledge, no study has evaluated the association between these perceived barriers and the severity of DR using a robust quantitative method. This information is particularly important for Asian populations, which have witnessed an exponential increase in the incidence rates of diabetes and DR as a result of rapid urbanization and an increasingly sedentary lifestyle combined with a Westernized type of diet.10,11 Data on the role of perceived barriers to DSM may potentially inform evidence-based interventions to prevent or delay the development and progression of DR.

In this study, we examined the association between perceived barriers to DSM (assessed using a questionnaire developed and validated in Asia) and the severity of DR in a multiethnic sample of Asian patients with type 2 diabetes.

Methods
Study Population

The Singapore Diabetes Management Project is a clinic-based cross-sectional study investigating the clinical, behavioral, and environmental barriers associated with optimal diabetes care in patients with diabetes with and without DR.12 We recruited 498 individuals with type 1 and 2 diabetes, 21 years of age or older, from the Singapore National Eye Centre from December 28, 2010, to March 20, 2013. All individuals were free of cognitive impairment (assessed using the 6-item Cognitive Impairment Test13), of sufficient hearing to be able to conduct normal conversations, and lived independently in the community (ie, not in assisted care facilities). Diabetes was defined as physician-diagnosed diabetes, with the information retrieved from participants’ case notes. For the current study, we included participants of Asian race/ethnicity (Chinese, Malay, and Indian) with diabetes who had available sociodemographic and clinical data (N = 361; comprising 261 Chinese, 31 Malay, and 69 Indian participants, with 127 excluded owing to missing data or diagnosis of type 1 diabetes). The study was approved by the Singapore Centralized Institutional Review Board (reference: 2010/470/A) and adhered to the tenets of the Declaration of Helsinki.14 All participants provided written informed consent.

The Perceived Barriers to DSM Questionnaire

Even though a comprehensive search of the literature established the presence of a substantial amount of qualitative information on perceived barriers to DSM, it also revealed a lack of valid and reliable questionnaires to assess these perceived barriers.6,8,9 Consequently, based on these qualitative studies, we developed a 36-item questionnaire across 4 specific types of barriers (eTable 1 in the Supplement), namely, knowledge of diabetes and its effect (items 1-8), external systems (items 9-17), psychosocial (items 18-26), and psychological (items 27-36). Items were rated on one of six 4- or 5-point rating scales (eTable 1 in the Supplement). We conducted cognitive interviews with 10 participants with diabetes (5 with DR and 5 without) to ensure that the questionnaire was comprehensible and culturally relevant.15 In addition, after every 3 interviews, iterative revisions were made to the instrument to improve content validity, either via the addition of relevant items or improving the wording of existing items, based on the feedback received from the participants and after discussions among an expert panel (R.E.K.M., E.K.F., and E.L.L.). These revisions were then tested in the next group of participants.16 As a result of these interviews, 2 items (items 21 and 25) were added and the wording of 11 existing questions was amended.

Psychometric Assessment of the Perceived Barriers to DSM Questionnaire

Rasch analysis was used to assess the psychometric properties of the Perceived Barriers to DSM questionnaire using the Andrich rating scale model17 with Winsteps software, version 3.92.1.18,19 Rasch analysis is a form of item response theory in which the ordinal ratings of the questionnaire are transformed to estimates of interval measures (expressed in log of the odds units [logits]). The advantages of item response theory vs classical test theory has been well documented20: it provides a richer description of the performance of each item compared with classical test theory, which is useful during development of patient-reported outcome measures; provides greater detail on a measure’s precision, which may vary across different levels of the construct at the item or scale level, as compared with classical test theory, which uses only a single estimate (eg, Cronbach α); and scores estimated using item response theory methods are independent of item difficulty as opposed to observed scores using classical test theory.

During Rasch analysis, responses were recoded for each item so that higher scores indicated a higher degree of perceived barriers, and differential item functioning (a measure of whether an item measures an ability the same way across different groups) was assessed for age (median split, <55 vs ≥55 years) and sex. Initially, the questionnaire displayed poor fit to the Rasch model parameters (eTable 2 in the Supplement), with disordered thresholds (response categories are not endorsed consistently) for most of the rating scales (eFigure 1 in the Supplement), poor precision, evidence of multidimensionality (>1 dimension being assessed), and 4 misfitting items. To resolve these issues, we collapsed response categories from 5 to 4 to resolve disordered thresholds and iteratively removed 13 items (items 4, 18-20, 23, 26-28, 30-31, 33, 35, and 36). The final 23-item questionnaire displayed adequate psychometric properties, with ordered thresholds, optimal measurement precision, no misfitting items or differential item functioning, and, importantly, minimal evidence of multidimensionality (eTable 2 in the Supplement). As such, a composite interval-level person measure for each participant for the 23 items was exported and used for subsequent parametric testing.

Assessment of DR and Diabetic Macular Edema

Diabetic retinopathy was graded from 2-field fundus photographs (Canon CR6–45NM; Canon Inc) using the modified Airlie House classification system.21 Diabetic macular edema was defined by hard exudates in the presence of microaneurysms and blot hemorrhage within 1 disc diameter from the foveal center or the presence of focal photocoagulation scars in the macular area. Clinically significant macular edema was considered present when the macular edema was within 500 mm of the foveal center or if focal laser photocoagulation scars were present in the macular area. These retinal abnormalities were confirmed from central macular thickness (defined as the thickness of the central circular zone of 1 mm diameter) measurements on spectral-domain optical coherence tomography (Cirrus, version 3.0; Carl Zeiss Meditec) using the macular thickness cube scan protocol (512 × 128) and comparing it with normative population data22; only scans with signal strength of 6 or greater were included.

For analytical purposes, DR was categorized as no DR (Early Treatment Diabetic Retinopathy Study levels 10-15), mild to moderate DR (levels 20-43), and severe DR (levels ≥53 and/or presence of clinically significant macular edema). Data from the worst eye were used; if images from both eyes were ungradable, the participant was excluded from analyses.

Assessment of Covariates

Participants’ demographic and socioeconomic characteristics (eg, age, sex, income, and educational level), lifestyle factors (eg, smoking), and medical history (eg, duration of diabetes) were collected using a standardized interviewer-administered questionnaire in English, Mandarin, Malay, or Tamil. Ethnicities were defined by the Singapore census as indicated on the National Registration Identity Card. Clinical covariates were obtained via a standardized clinical examination. Presenting distance visual acuity was checked monocularly by a trained research officer with the participant wearing current refractive correction (if any) using a logMAR number chart at a distance of 4 m under standard lighting. Presenting visual impairment was defined as a logMAR score of greater than 0.3 in the better-seeing eye. Height was measured in centimeters using a wall-mounted measuring tape, and weight was measured in kilograms using a digital scale. Body mass index was calculated as weight in kilograms divided by the height in meters squared. Trained personnel performed systolic and diastolic BP measurements twice using a digital BP monitor (Dinamap Pro 100 V2; GE Healthcare), with the mean value of each parameter used in the analyses.

Nonfasting venous blood samples were also collected to assess hemoglobin A1c (HbA1c), serum creatinine, serum total cholesterol, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein cholesterol, and triglycerides levels. All samples were analyzed at the Singapore General Hospital Hematology Laboratory. Hemoglobin A1c was assessed via immunoassay conducted using the Roche Cobas c501 (Roche Diagnostics) calibrated according to the standards set by the National Institute of Standards and Technology.23 Likewise, serum total cholesterol, HDL cholesterol, low-density lipoprotein cholesterol, and triglycerides were assessed via spectrophotometry conducted using the Beckman Coulter Unicel DxC 800 (Beckman Coulter Inc). Total to HDL cholesterol ratio was used in analyses, as this measure has greater predictive value than isolated parameters used independently.24

Statistical Analysis

Statistical analysis was conducted from July 20 to September 8, 2017. All analyses were performed using intercooled Stata, version 14.1 for Windows (StataCorp). Participant characteristics of persons with and without DR were first compared using the χ2 statistic for proportions for categorical variables and a t test for means or Mann-Whitney test for medians for continuous variables. The degree of perceived barriers experienced by participants was analyzed continuously and categorically in tertiles. We categorized the perceived barriers score into tertiles ad hoc, as it allowed us to maintain a reasonable number of participants within each tertile when analyzing the association between perceived barriers and severity of DR. Multivariable logistic regression models were then used to determine the independent associations of the degree of perceived barriers to DSM on DR presence, adjusted for participant characteristics found to be significantly different between those with and without DR plus variables previously found to be associated with DR in people with diabetes, including age, sex, race/ethnicity, educational level (≤6 vs >6 years), monthly income (<SGD2000 vs ≥SGD2000 [US $1466.17]), smoking status (current vs nonsmoker or ex-smoker), HbA1c level, duration of diabetes, body mass index, presenting visual acuity in the better eye, systolic BP, total to HDL cholesterol ratio, insulin use, and antihypertensive medication use. Finally, we used multinomial logistic regression adjusted for the above confounders to determine the independent effect of the magnitude of perceived barriers to DSM with the severity of DR. In addition, ordinal logistic regression models were used to estimate the overall trend of the categorical exposures (in tertiles) with the severity of DR. P < .05 (2-sided) was considered significant.

Results

The mean (SD) age of the sample was 57 (8) years, and 105 participants (29.1%) were female. The mean (SD) score for perceived barriers to DSM was 1.63 (1.07) logits (range, –0.51 to 6.81), with higher scores indicating greater perceived barriers (eFigure 2 in the Supplement). A total of 207 participants (57.3%) had DR in at least 1 eye; of those, 112 had mild to moderate DR and 95 had severe DR, while 154 participants did not have DR. Table 1 summarizes the participants’ clinical and demographic characteristics stratified by DR status. Patients with DR were more likely than those without DR to have a lower mean (SD) body mass index (26.0 [4.0] vs 26.9 [4.2]; P = .04), have a higher mean (SD) HbA1c level (8.14% [1.68%] vs 7.50% [1.42%] [to convert to proportion of total hemoglobin, multiply by 0.01]; P < .001), have a higher mean (SD) systolic BP (137.7 [18.2] vs 134.0 [15.5] mm Hg; P = .04), have a longer mean (SD) duration of diabetes (16.2 [9.3] vs 10.1 [8.1] years; P < .001), were more likely to use insulin (43 [20.8%] vs 8 [5.2%]; P < .001), and experienced a greater degree of perceived barriers (mean [SD], –1.52 [1.09] vs –1.78 [1.02] logits) (Table 1). No differences were found in the perceived barriers score and severity of DR between those who were included compared with those who were excluded in analyses (eTable 3 in the Supplement).

In age- and sex-adjusted models, a higher perceived barrier to DSM was associated with higher odds of having any DR (per logit decrease: odds ratio [OR], 1.26; 95% CI, 1.03-1.54) (Table 2, model 1). This association remained even after multivariable adjustments (Table 2, model 2). In addition, participants with the greatest degree of perceived barriers (ie, tertile 3) were more likely to have DR compared with those in tertile 1 (OR, 2.11; 95% CI, 1.25-3.57; P = .006 for trend), and these associations remained after further adjustments. Subgroup analyses for age (<65 vs ≥65 years), sex (male vs female), or race/ethnicity (Chinese, Malay, or Indian) found no between-group differences.

Table 2 also shows the associations between the magnitude of perceived barriers to DSM with DR severity. We found that, in age- and sex-adjusted models, while there was an increase in the odds of having mild to moderate DR with a greater magnitude of perceived barriers (per logit decrease: OR, 1.23; 95% CI, 0.97-1.56; model 1), this association increased in magnitude after multivariable adjustments (OR, 1.30; 95% CI, 1.01-1.68; model 2). We also demonstrated that, after multivariable adjustments, participants with the highest tertile of perceived barriers to DSM were more than 2.5 times more likely to have mild to moderate DR (OR, 2.52; 95% CI, 1.30-4.89; P = .006 for trend; model 2) than those with the lowest tertile of perceived barriers.

Similarly, for severe DR, a greater degree of perceived barriers was associated with greater odds of having severe DR both continuously (per logit decrease: OR, 1.29; 95% CI, 1.00-1.66; model 1) and categorically (OR, 2.13; 95% CI, 1.13-4.03; model 1) compared with the lowest tertile of perceived barriers. These associations persisted after multivariable adjustments (model 2).

Discussion

In this study, we demonstrated that a greater level of perceived barriers to DSM in Asian patients with type 2 diabetes was associated with higher odds of having any DR, mild to moderate DR, and severe DR. More important, these findings were independent of the parameters of diabetes control (HbA1c, BP, and lipid levels), suggesting that, aside from pharmaceutical management, measures to address these perceived barriers, including counseling, goal setting, and logistical support for access to health care, may help to prevent or delay the risks of DR.

Our data suggest that there is a linear association between the magnitude of perceived barriers to DSM and the presence and severity of DR. At first glance, these results appear to corroborate studies showing a strong correlation between a greater awareness of and proficiency in DSM activities and systemic metabolic parameters25,26 since it is widely accepted that optimal control of these parameters (HbA1c level, 6.5%-7% low-density lipoprotein cholesterol level <100.4 mg/dL [to convert to millimoles per liter, multiply by 0.0259]; HDL cholesterol level >50.2 mg/dL [to convert to millimoles per liter, multiply by 0.0259]; and BP ≤130/80 mm Hg) form the cornerstone of good management of DR.2,3 However, we found that the association between barriers to DSM and DR persisted even after accounting for these systemic variables, suggesting the involvement of factors other than classic poor control of metabolic parameters.

The mechanisms underpinning the association between perceived barriers to DSM activities and DR were not studied in this work, although previous research has found consistent correlations between poor DSM behavior and possible contributory factors including diabetes-related distress,27 health literacy,28 and depression,29,30 all of which have been demonstrated to be associated with a higher likelihood of more severe DR.28,31,32 As we did not assess participants’ mental health or health literacy status, we were unable to determine the role of these factors in the observed association between barriers to DSM and DR. Further cohort studies are needed to better understand the mechanisms underlying the above association.

As a valid questionnaire on perceived barriers to DSM is not currently available, we developed a new one and psychometrically tested it using Rasch analysis. Initially, the 36-item questionnaire had substantial psychometric issues even after cognitive testing; however, using an iterative, Rasch-guided approach involving collapsing poorly used response categories and deleting misfitting items, we produced a valid 23-item scale to measure our desired construct. Our study hence highlights the usefulness of using modern psychometric methods in the development and validation of questionnaires.33,34

Strengths and Limitations

Strengths of this study include a study sample characterized using a standardized clinical testing protocol, a large sample of persons with varying severity of DR, the objective categorization of DR using fundus and spectral-domain optical coherence tomography images, a comprehensive assessment of DR confounders, and a robust and psychometrically validated questionnaire on barriers to DSM. Limitations include the cross-sectional nature of this study, which limited causal inferences, as well as the clinical nature of the study population, which limited the generalizability of the study. However, these results are still helpful to inform novel targeted interventions for individuals with DR to address any perceived barriers to DSM, with the aim of slowing the progression of the disease. Moreover, a substantial proportion of our study population were men (256 [70.9%]), which may have affected responses to the questionnaire. However, Rasch analysis revealed that sex did not significantly affect the way that participants responded to the items within the questionnaire. In addition, we quantified the magnitude of perceived barriers only from the perspective of patients with diabetes. It has been established that perceived barriers to DSM on the part of health care professionals plays an equally important role in poor management of diabetes.9,35,36 Further studies on the effect of the magnitude of perceived barriers to DSM on DR from the perspective of the health care professional may further illuminate the issue. In addition, we conducted only 3 cognitive interviews to improve the clarity and comprehensiveness of our instrument, which may have affected content validity. As such, more comprehensive foundational work, together with the use of multidimensional item response theory models during psychometric assessment to cope with the potentially multidimensional constructs, may be needed to develop new item content and refine existing item content. Last, we did not evaluate the facilitators of DSM in our study, which would have given us greater insight into the steps needed to overcome these perceived barriers. Future qualitative and quantitative studies should be performed in order to inform intervention strategies to address perceived barriers to DSM.

Conclusions

We found a systematic linear association between the magnitude of perceived barriers to DSM and the severity of DR in Asian patients with diabetes. Future studies to optimize understanding of the mechanisms contributing to the observed association are needed. In addition, community and clinical health efforts to overcome patients’ perceived barriers to DSM, including diabetes support groups and evaluation of individual patient perception toward diabetes with the goal of making realistic and specific recommendations for self-care activities, are warranted to combat the increasing incidence of diabetes and DR in Asia.

Back to top
Article Information

Corresponding Author: Ecosse L. Lamoureux, PhD, Singapore Eye Research Institute, 20 College Rd, The Academia, Discovery Tower Level 6, Singapore 169856 (ecosse.lamoureux@seri.com.sg).

Accepted for Publication: September 23, 2017.

Published Online: November 16, 2017. doi:10.1001/jamaophthalmol.2017.4888

Author Contributions: Dr Man and Mr Gan had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.

Study concept and design: Wong, Tan, Lamoureux.

Acquisition, analysis, or interpretation of data: Man, Fenwick, Gan, Sabanayagam, Gupta, Aravindhan, Tan, Lamoureux.

Drafting of the manuscript: Fenwick, Gan, Lamoureux.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Man, Fenwick, Gan, Aravindhan, Lamoureux.

Obtained funding: Tan, Lamoureux.

Administrative, technical, or material support: Sabanayagam, Gupta, Aravindhan, Tan, Lamoureux.

Study supervision: Wong, Lamoureux.

Conflict of Interest Disclosures: All authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest and none were reported.

Funding/Support: The Singapore Diabetes Management Project was funded by the National Medical Research Council (Singapore) as part of its Centre/Programmatic Project Grant. Dr Fenwick is funded by Early Career Fellowship 1072987 from the Australian National Health and Medical Research Council.

Role of the Funder/Sponsor: The funding sources had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

References
1.
Cheung  N, Mitchell  P, Wong  TY.  Diabetic retinopathy.  Lancet. 2010;376(9735):124-136.PubMedGoogle ScholarCrossref
2.
Holman  RR, Paul  SK, Bethel  MA, Matthews  DR, Neil  HA.  10-Year follow-up of intensive glucose control in type 2 diabetes.  N Engl J Med. 2008;359(15):1577-1589.PubMedGoogle ScholarCrossref
3.
Nathan  DM, Genuth  S, Lachin  J,  et al; Diabetes Control and Complications Trial Research Group.  The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus.  N Engl J Med. 1993;329(14):977-986.PubMedGoogle ScholarCrossref
4.
Genuth  S, Eastman  R, Kahn  R,  et al; American Diabetes Association.  Implications of the United Kingdom Prospective Diabetes Study.  Diabetes Care. 2003;26(suppl 1):s28-s32.PubMedGoogle ScholarCrossref
5.
Huang  OS, Lamoureux  EL, Tay  WT, Tai  ES, Wang  JJ, Wong  TY.  Glycemic and blood pressure control in an Asian Malay population with diabetes and diabetic retinopathy.  Arch Ophthalmol. 2010;128(9):1185-1190.PubMedGoogle ScholarCrossref
6.
Hiss  RG.  Barriers to care in non–insulin-dependent diabetes mellitus: the Michigan experience.  Ann Intern Med. 1996;124(1, pt 2):146-148.PubMedGoogle ScholarCrossref
7.
Corabian  P, Harstall  C.  Patient Diabetes Education in the Management of Adult Type 2 Diabetes. Edmonton, AB, Canada: Alberta Heritage Foundation for Medical Research; 2001.
8.
Munshi  MN, Segal  AR, Suhl  E,  et al.  Assessment of barriers to improve diabetes management in older adults: a randomized controlled study.  Diabetes Care. 2013;36(3):543-549.PubMedGoogle ScholarCrossref
9.
Nam  S, Chesla  C, Stotts  NA, Kroon  L, Janson  SL.  Barriers to diabetes management: patient and provider factors.  Diabetes Res Clin Pract. 2011;93(1):1-9.PubMedGoogle ScholarCrossref
10.
International Diabetes Federation.  IDF Diabetes Atlas. 7th ed. Brussels, Belgium: International Diabetes Federation; 2015.
11.
Yau  JW, Rogers  SL, Kawasaki  R,  et al; Meta-Analysis for Eye Disease (META-EYE) Study Group.  Global prevalence and major risk factors of diabetic retinopathy.  Diabetes Care. 2012;35(3):556-564.PubMedGoogle ScholarCrossref
12.
Man  RE, Sabanayagam  C, Chiang  PP,  et al.  Differential association of generalized and abdominal obesity with diabetic retinopathy in Asian patients with type 2 diabetes.  JAMA Ophthalmol. 2016;134(3):251-257.PubMedGoogle ScholarCrossref
13.
Brooke  P, Bullock  R.  Validation of a 6 item cognitive impairment test with a view to primary care usage.  Int J Geriatr Psychiatry. 1999;14(11):936-940.PubMedGoogle ScholarCrossref
14.
World Medical Association.  World Medical Association Declaration of Helsinki: ethical principles for medical research involving human subjects.  JAMA. 2013;310(20):2191-2194.PubMedGoogle ScholarCrossref
15.
Willis  GB.  Cognitive Interviewing: A Tool for Improving Questionnaire Design. Los Angeles, CA: Sage Publications; 2004.
16.
Collins  D.  Pretesting survey instruments: an overview of cognitive methods.  Qual Life Res. 2003;12(3):229-238.PubMedGoogle ScholarCrossref
17.
Linacre  JM.  A User’s Guide to Winsteps: Rasch-Model Computer program. Chicago, IL: MESA press; 2002.
18.
Linacre  JM.  A User’s Guide to Winsteps/Ministeps Rasch-Model Programs. Chicago, IL: MESA press; 2005.
19.
Andrich  D.  A rating formulation for ordered response categories.  Psychometrica. 1978;43(4):561-573. doi:10.1007/BF02293814Google ScholarCrossref
20.
Nguyen  TH, Han  HR, Kim  MT, Chan  KS.  An introduction to item response theory for patient-reported outcome measurement.  Patient. 2014;7(1):23-35.PubMedGoogle ScholarCrossref
21.
Early Treatment Diabetic Retinopathy Study Research Group.  Grading diabetic retinopathy from stereoscopic color fundus photographs—an extension of the modified Airlie House classification. ETDRS report number 10.  Ophthalmology. 1991;98(5)(suppl):786-806.PubMedGoogle ScholarCrossref
22.
Gupta  P, Sidhartha  E, Tham  YC,  et al.  Determinants of macular thickness using spectral domain optical coherence tomography in healthy eyes: the Singapore Chinese Eye Study.  Invest Ophthalmol Vis Sci. 2013;54(13):7968-7976.PubMedGoogle ScholarCrossref
23.
Yeo  CP, Tan  CH, Jacob  E.  Haemoglobin A1c: evaluation of a new HbA1c point-of-care analyser Bio-Rad in2it in comparison with the DCA 2000 and central laboratory analysers.  Ann Clin Biochem. 2009;46(pt 5):373-376.PubMedGoogle ScholarCrossref
24.
Millán  J, Pintó  X, Muñoz  A,  et al.  Lipoprotein ratios: physiological significance and clinical usefulness in cardiovascular prevention.  Vasc Health Risk Manag. 2009;5:757-765.PubMedGoogle Scholar
25.
Norris  SL, Engelgau  MM, Narayan  KM.  Effectiveness of self-management training in type 2 diabetes: a systematic review of randomized controlled trials.  Diabetes Care. 2001;24(3):561-587.PubMedGoogle ScholarCrossref
26.
Sherifali  D, Bai  JW, Kenny  M, Warren  R, Ali  MU.  Diabetes self-management programmes in older adults: a systematic review and meta-analysis.  Diabet Med. 2015;32(11):1404-1414.PubMedGoogle ScholarCrossref
27.
Fisher  L, Gonzalez  JS, Polonsky  WH.  The confusing tale of depression and distress in patients with diabetes: a call for greater clarity and precision.  Diabet Med. 2014;31(7):764-772.PubMedGoogle ScholarCrossref
28.
Schillinger  D, Grumbach  K, Piette  J,  et al.  Association of health literacy with diabetes outcomes.  JAMA. 2002;288(4):475-482.PubMedGoogle ScholarCrossref
29.
Egede  LE, Ellis  C.  The effects of depression on diabetes knowledge, diabetes self-management, and perceived control in indigent patients with type 2 diabetes.  Diabetes Technol Ther. 2008;10(3):213-219.PubMedGoogle ScholarCrossref
30.
Lin  EH, Katon  W, Von Korff  M,  et al.  Relationship of depression and diabetes self-care, medication adherence, and preventive care.  Diabetes Care. 2004;27(9):2154-2160.PubMedGoogle ScholarCrossref
31.
Fenwick  EK, Pesudovs  K, Rees  G,  et al.  Republished article: the impact of diabetic retinopathy: understanding the patient’s perspective.  Postgrad Med J. 2012;88(1037):167-175.PubMedGoogle ScholarCrossref
32.
Chen  X, Lu  L.  Depression in diabetic retinopathy: a review and recommendation for psychiatric management.  Psychosomatics. 2016;57(5):465-471.PubMedGoogle ScholarCrossref
33.
Prieto  L, Alonso  J, Lamarca  R.  Classical Test Theory versus Rasch analysis for quality of life questionnaire reduction.  Health Qual Life Outcomes. 2003;1:27.PubMedGoogle ScholarCrossref
34.
Petrillo  J, Cano  SJ, McLeod  LD, Coon  CD.  Using classical test theory, item response theory, and Rasch measurement theory to evaluate patient-reported outcome measures: a comparison of worked examples.  Value Health. 2015;18(1):25-34.PubMedGoogle ScholarCrossref
35.
Simmons  D, Lillis  S, Swan  J, Haar  J.  Discordance in perceptions of barriers to diabetes care between patients and primary care and secondary care.  Diabetes Care. 2007;30(3):490-495.PubMedGoogle ScholarCrossref
36.
Chin  MH, Cook  S, Jin  L,  et al.  Barriers to providing diabetes care in community health centers.  Diabetes Care. 2001;24(2):268-274.PubMedGoogle ScholarCrossref
×