Compliance with eye and foot preventive care in people with self-reported diabetes in Latin America and the Caribbean: Pooled, cross-sectional analysis of nine national surveys

Abstract

Aims:

To estimate the proportion of people with self-reported diabetes receiving eye and foot examinations in Latin America and the Caribbean (LAC).

Methods:

Cross-sectional analysis of national health surveys in nine countries. Adults aged 25–64 years with self-reported diabetes. We quantified the proportion who reported having an eye examination in the last two years or a foot examination in the last year. We fitted multilevel Poisson regressions to assess socio-demographic (age and sex) and clinical (oral hypoglycemic medication and insulin treatment) variables associated with having had examinations.

Results:

There were 7435 people with self-reported diabetes included in the analysis. In three countries (Chile [64%; 95% CI: 56%−71%], British Virgin Islands [58%; 95% CI: 51%−65%], and Brazil [54%; 95% CI: 50%−58%]), >50% of people with diabetes reported having had an eye examination in the last two years. Fewer participants (<50% across all countries) reported having had a foot examination in the last year, with Ecuador having the lowest proportion (12%; 95% CI: 8%−17%). Older people, and those taking oral medication or insulin, were more likely to have eye/foot examinations.

Conclusions:

The proportion of eye and foot examinations in people with self-reported diabetes across nine countries in LAC is low.

1. Introduction

Diabetes is highly prevalent worldwide and disproportionally affects low- and middle-income countries. [1,2] Latin America and the Caribbean (LAC) have one of the fastest-growing prevalence of diabetes worldwide; [3] however, there is very limited information about the quality of care that people with diabetes receive to prevent diabetes-related complications in LAC. [4]

While cardiovascular disease morbidity and mortality has significantly increased among people with diabetes in LAC, other debilitating and disabling microvascular complications can significantly impair their quality of life too. [5–9] Diabetic retinopathy [8,9] and peripheral neuropathy [5,7] if not detected and treated early can lead to blindness and limb amputations, and to adverse socio-economic consequences both for patients and health systems. Thus, in addition to adequate metabolic control, [10,11] eye and foot examinations are recommended for all patients with diabetes for early detection of diabetic retinopathy and peripheral neuropathy. [12]

Whether preventive eye and foot care recommendations are met by people with diabetes in LAC is largely unknown. This information is relevant for public health officials to understand the magnitude of the gap (i.e., people with diabetes not receiving eye and foot examinations), and to guide tailored policies and interventions. This information is also important to guide professional organizations (e.g., Latin American Association for Diabetes [13]) on the strength of preventive care recommendations and to support the design of country-specific pathways to optimize eye and foot preventive care using resources locally available.

To quantify the proportion of eye and foot examinations among adults with diabetes in LAC, we pooled data from nationally representative surveys from nine countries. We also explored socio-demographic and clinical factors associated with having had eye and foot examinations.

2. Methods

2.1. Study design and data sources

Pooled cross-sectional analysis of nine health surveys (Supplementary Table 1) in Argentina (Encuesta Nacional de Factores de Riesgo, 2018), Bahamas (WHO STEPS Survey, 2012), Bolivia (WHO STEPS Survey 2019), Brazil (Pesquisa Nacional de Saude, 2013), Chile (Encuesta Nacional de Salud, 2017), Ecuador (WHO STEPS Survey, 2018), Mexico (Encuesta Nacional de Salud y Nutricion, 2019), Uruguay (WHO STEPS Survey, 2014), and British Virgin Islands (WHO STEPS Survey, 2009). Each survey followed a multi-stage sampling design to collect health questionnaires and clinical measurements (e.g., anthropometrics and blood pressure) in a sample of the general population to make inferences at the national level.

We selected national surveys with information about self-reported diabetes status, and surveys with information about self-reported foot examination (in the last year) and eye examination (in the last two years) as shown in Supplementary Table 2. If there were more than one survey per country, we included the most recent survey only. We restricted the analysis to surveys in the public domain (i.e., open access data).

2.2. Study population

We studied men and women aged between 25 and 64 years. This age bracket was available in all surveys. The study population included only people who reported having been diagnosed with diabetes (i.e., self-reported diabetes). The study population was restricted to such group because people with diabetes diagnosis would be expected to have regular eye and foot examinations as part of their usual care. Of note, the questionnaires did not specify type 1 or type 2 diabetes, therefore throughout the manuscript we refer to diabetes in general rather than type 1 or type 2. For the purpose of this analysis, arguably, the distinction between diabetes types is not essential because all people with diabetes should receive regular eye and foot examinations.

2.3. Variables

Consistent with international [12] and LAC [13] clinical guidelines, information about eye and foot examinations was dichotomized as yes (within the last two years for eye and within the last year for foot) versus no (all other possible time frames). Information about eye and foot examinations was elicited with questionnaires. The exact questions together with how these questions were operationalized for the analysis are shown in Supplementary Table 2. In seven surveys, information about eye examination was asked about the last two years; however, in two surveys (Argentina and Mexico) they asked about the last year.

For descriptive purposes we stratified the analysis by sex (men and women), age, and body mass index (BMI; kg/m²). Four age (25–34, 35–44, 45–54, and 55–64 years) and four BMI (<18.5 kg/m², 18.5–24.9 kg/m², 25.0–29.9 kg/m², and 30+ kg/m²) strata were included in the analysis. We also included hypertension status (no versus yes) based on self-reported diagnosis and the average of the first two blood pressure measurements (systolic blood pressure/diastolic blood pressure >=130/80 mmHg, consistent with stage 1 hypertension [14]). We averaged the first two measurements -and not the second and third measurements- because some surveys did not collect three blood pressure measurements. We included variables for diabetes oral medication (no/yes) and insulin treatment (no/yes). Finally, we also included data on whether the participants had close relatives (parents and siblings) with diabetes (no/yes; Supplementary Table 3), and about diabetes duration (<5 and 5+ years since self-reported diagnosis; Supplementary Table 4).

2.4. Statistical analysis

The statistical analyses were conducted with R (version 4.1.2) and the regression models were conducted with Stata (version 17, College Station, Texas 77845, USA). Analysis codes are available as Supplementary Materials.

First, to examine preventive care practices at the country level, we computed the proportion of eye and foot examinations, including a sex- and age-stratified analysis. All estimates (both sex- and age-stratified) accounted for the sampling design of each survey. The srvyr library in R was used for the analysis of survey data. Because we did not aim to claim differences between countries or groups, rather describe the landscape of eye and foot examination in adults with self-reported diabetes in LAC by socio-demographic variables, we did not perform statistical tests across groups.

Second, to identify associated factors with having had eye and foot examinations, we constructed a multilevel model of the Poisson family with log link, using country as a random intercept, including robust standard errors, and the covariance matrix was set as unstructured. We developed two models, one for each outcome (eye and foot examination). We presented adjusted results by sex, age in categories, BMI in categories, hypertension status, oral hypoglycemic medication, and insulin treatment. Exploratorily, we also adjusted the regression models with having a direct family member with diabetes and with years since diabetes diagnosis. These results were considered exploratorily because a subset of surveys had information about diabetes in close relatives and diabetes duration.

The regression results are presented as Prevalence Ratios (PR) with 95% Confidence Intervals (95% CI) and associated p-values. A p-value <0.05 was considered statistically significant to claim that a given variable was associated with having eye or foot examinations independent of the other predictors in the regression model.

2.5. Ethics

We analyzed deidentified open-access data. No human subjects were directly involved in this work. We deemed this work as of minimal risk and did not seek approval by an Ethics Committee.

2.6. Role of the funding source

There was no specific funding for this work. The opinions in this document belong to the authors alone, and do not necessarily reflect the opinions of the institutions to which the authors belong. RMC-L and WCG-V had access to the data and vouch for the accuracy of the results. RMC-L is the guarantor and had final decision to submit for publication.

3. Results

3.1. Study population

There were 7435 adults aged between 25 and 64 years with self-reported diabetes in the analytic sample (Supplementary Table 5). The analytic sample was largest in Argentina (2503) and the smallest in Bahamas (116). The mean age of the study populations was similar across surveys, ranging from 47.5 (standard deviation (SD): 11.2) years in Argentina to 51.7 (SD: 9.3) years in Brazil.

3.2. Eye examination

Overall (Table 1), over one in two adults with self-reported diabetes in Chile (64% [95% CI: 56%−71%]), British Virgin Islands (58% [95% CI: 52%−65%]), and Brazil (55% [95% CI: 51%−59%]), reported having had an eye examination in the previous two years. The same profile was observed in the sex-specific results (Fig. 1), though men in British Virgin Islands (48% [95% CI: 43%−54%]) were slightly behind the 50% threshold. On the other hand, fewer than 30% of adults with self-reported diabetes in Bolivia (22% [95% CI: 16%−29%]), Ecuador (22% [95% CI: 16%−29%]), and Mexico (15% [95% CI: 12%−18%]) reported having had an eye examination; readers are kindly reminded that questionnaires in Mexico and Argentina referred to the past year (Supplementary Table 2). The same pattern was largely observed in the sex-stratified results. In sex-age-specific results, in Uruguay, Ecuador, Chile (men), and Bahamas (women), we observed a consistent age pattern whereby older age groups had higher proportion of eye examination (Supplementary Figure 1 and Supplementary Table 7).

Table 1.

Proportion (%) of eye and foot examination in adults (25–64 years) with self-reported diabetes by country in Latin America and the Caribbean.

Country, year	Proportion	Lower 95% Confidence Interval	Upper 95% Confidence Interval
Eye examination
Argentina, 2018	39.62	32.81	46.84
Bahamas 2012	48.21	37.10	59.50
Bolivia, 2019	21.46	15.57	28.81
Brazil, 2013	54.89	50.88	58.85
Chile, 2017	64.24	56.39	71.40
Ecuador, 2018	22.04	16.43	28.91
Mexico, 2019	14.68	11.90	17.99
Uruguay, 2014	42.17	35.03	49.66
Br. Virgin Islands, 2009	58.34	51.51	64.86
Foot examination
Argentina, 2018	26.80	20.80	33.79
Bahamas 2012	36.80	21.74	54.96
Bolivia, 2019	17.61	12.24	24.68
Brazil, 2013	31.66	28.17	35.37
Chile, 2017	40.55	32.94	48.64
Ecuador, 2018	12.81	8.88	18.14
Mexico, 2019	22.89	19.73	26.39
Uruguay, 2014	20.97	15.89	27.17
Br. Virgin Islands, 2009	28.01	24.30	32.05

These proportions account for the sampling design of each survey. The results show the proportion and the 95% confidence interval.

Fig. 1. Proportion (%) of adults with self-reported diabetes who have had eye examination in the last two years by sex.

These proportions account for the sampling design of each survey. The results show the proportions and the 95% confidence interval. Results also shown in Supplementary Table 6.

Older age was associated with higher prevalence of eye examination; for example, in comparison to the youngest age group (25–34 years), the oldest age group (55–64 years) had 37% (PR: 1.37 [95% CI: 1.05–1.78]) higher chance of having had an eye examination (Table 2). People receiving insulin were 58% (PR: 1.58 [95% CI: 1.28–1.95]) more likely to have had an eye examination, and people on oral hypoglycemic medication were 47% (PR: 1.47 [95% CI: 1.09–1.99]) more likely too.

Table 2.

Associated factors with eye examination in the last two years.

	PR (95% CI)	p-value
Sex
Men	Reference
Women	0.98 (0.92–1.06)	0.690
Age groups (years)
25–34	Reference
35–44	1.08 (0.91–1.28)	0.386
45–54	1.32 (1.09–1.60)	0.005
55–64	1.37 (1.05–1.78)	0.022
BMI groups (kg/m2)
<18.5	Reference
18.5–24.9	1.09 (0.68–1.75)	0.707
25.0–29.9	1.24 (0.81–1.89)	0.318
30.0+	1.20 (0.78–1.85)	0.398
Hypertension
No	Reference
Yes	1.06 (0.94–1.19)	0.333
Insulin
No	Reference
Yes	1.58 (1.28–1.95)	<0.001
Oral medication
No	Reference
Yes	1.47 (1.09–1.99)	0.013

BMI: body mass index (kg/m2); PR: prevalence ratio; 95% CI: 95% confidence interval. 5640 observations were included in this multilevel Poisson regression with log link where the country was the random intercept.

3.3. Foot examination

Overall, Chile (41% [95% CI: 33%−49%]) was the country closest to having 50% of their adult population with self-reported diabetes who have received a foot examination in the last year. Bahamas (37% [95% CI: 22%−55%]) and Brazil (32% [28%−35%]) ranked second and third, respectively. In the sex-specific results (Fig. 2), no country was above the 50% threshold for men, whilst for women Chile (44% [95% CI: 35%−54%]) and Bahamas (37% [95% CI: 15%−67%]) were closest to such threshold. On the other hand, in Bolivia (18% [95% CI: 12%−25%]) and Ecuador (13% [95% CI: 9%−18%]) fewer than one in five of adults with self-reported diabetes reported having had a foot examination, with similar sex-specific results. In the sex-age-specific results, in Mexico (women), Ecuador, Chile, and British Virgin Islands (men), the proportion of foot examination in people with self-reported diabetes increased with age (Supplementary Figure 2 and Supplementary Table 9).

Fig. 2. Proportion (%) of adults with self-reported diabetes who have had foot examination in the last year by sex.

These proportions account for the sampling design of each survey. The results show the proportions and the 95% confidence interval. Results also shown in Supplementary Table 8.

Older age was associated with higher prevalence of foot examination (Table 3). Women were 12% less likely to have had a foot examination in the last year (PR: 0.88; 95% CI: 0.78–0.99). People receiving insulin were 106% (PR: 2.06 [95% CI: 1.80–2.35]) more likely to have had a foot examination, and people on oral hypoglycemic medication were 91% (PR: 1.91 [95% CI: 1.36–2.67]) more likely too.

Table 3.

Associated factors with foot examination in the last year.

	PR (95% CI)	p-value
Sex
Men	Reference
Women	0.88 (0.78–0.99)	0.043
Age groups (years)
25–34	Reference
35–44	1.11 (0.94–1.31)	0.224
45–54	1.41 (1.14–1.75)	0.002
55–64	1.54 (1.19–2.01)	0.001
BMI groups (kg/m2)
<18.5	Reference
18.5–24.9	1.21 (0.83–1.75)	0.318
25.0–29.9	1.18 (0.86–1.62)	0.315
30.0+	1.15 (0.87–1.53)	0.338
Hypertension
No	Reference
Yes	1.05 (0.98–1.13)	0.173
Insulin
No	Reference
Yes	2.06 (1.80–2.35)	<0.001
Oral medication
No	Reference
Yes	1.91 (1.36–2.67)	<0.001

BMI: body mass index (kg/m2); PR: prevalence ratio; 95% CI: 95% confidence interval. 5640 observations were included in this multilevel Poisson regression with log link where the country was the random intercept.

3.4. Exploratory analysis

In the regression models which also included diabetes in close relatives and diabetes duration, for eye examination, age, insulin treatment and oral medication were still strongly associated, and people with five or more years since diabetes diagnosis were 28% (PR: 1.28 [95% CI: 1.16–1.41]) more likely to have had an eye examination (Supplementary Table 10). For foot examination, age, insulin and oral medication were still strongly associated; similarly, people with close relatives with diabetes were 9% (PR: 1.09 [95% CI: 1.00–1.17]) more likely to have had a foot examination while people with five or more years since diabetes diagnosis were 14% more likely (PR: 1.14 [95% CI: 1.12–1.17]) (Supplementary Table 11).

4. Discussion

4.1. Main findings

This analysis of pooled data from nine national health surveys in LAC shows an alarming low proportion of eye and foot preventive care among people with self-reported diabetes. Only three countries (Chile, Brazil and British Virgin Islands) were close to or over 50% of people with self-reported diabetes who had an eye examination in the past two years. The proportion of foot examination was lower than 50% in all countries. In six countries (Argentina, Bolivia, Ecuador, Mexico, Uruguay, and British Virgin Islands) three or fewer than three out of ten adults with self-reported diabetes had a foot exam in the prior year.

These results show that in LAC, care to prevent major diabetes-related complications (i.e., retinopathy and peripheral neuropathy) warrants further investment, resources and standard management to secure more adults with diabetes receive eye and foot examinations. Our results also call for innovative research to identify patients who must not miss eye or foot examinations in the context of limited resources where not all patients may have access to such care. Interventions and health systems research are needed to secure opportunities for patients with diabetes to receive the evidence-based care they need to prevent diabetes-related complications.

4.2. Public health relevance

A previous report from Brazil using the same national survey we analyzed is consistent with our estimates for foot examination; [15] however, they reported lower proportion of eye examination probably because they used a different time frame (in the last year rather than in the last two years as herein reported). [15] Of note, the time frame we chose for eye and foot examinations is consistent with international [12] and LAC [13] clinical guidelines. There is a lack of equivalent information based on large or national population-based surveys from other countries in LAC; however, there is relatively more evidence about the prevalence and incidence of the complications. [5–9] High-quality evidence about the complications is also needed in LAC. These early and intermediate milestones can be intervened or inform monitoring and benchmark thresholds.

The 2019 [13] guidelines from the Latin American Association for Diabetes (ALAD for its name in Spanish) proposed the National Committee for Quality Assurance (NCQA) metrics to assess the quality of diabetes care. For example, they aim for 60% of patients with an eye examination and 80% with a foot examination. Acknowledging that these metrics are for healthcare facilities or units providing care for people with diabetes and our estimates are at the general population level, for several countries there is a large gap between these targets and the observed frequencies herein reported. Whether these or other similar thresholds should be used to monitor the progression towards having more people with diabetes with eye and foot examinations deserves careful consideration and analysis. On the other hand, specific metrics for eye and foot examinations can be proposed for LAC, or globally, to be monitored at the population level. As it has been the case recently with the Global Diabetes Compact by the World Health Organization with new monitoring metrics (e.g., 60% of people with diabetes aged 40+ receiving statins) [16] based on empirical data. [17] The implementation of the Global Diabetes Compact strategies in the Americas [18] may boost the number of people with diabetes receiving eye and foot examinations in LAC, and our results can be used as baseline to monitor the progression.

5. Strengths and limitations

We analyzed nine national health surveys across LAC, including large countries such as Brazil and Mexico. In addition, the number of people with diabetes in Mexico (~12 million in 2019 [19]) and Brazil (~13 million in 2019 [19]) is among the highest worldwide, [1] making our results relevant to inform the global profile of eye and foot examinations. We analyzed individual data following a consistent protocol, and we conducted an individual-level meta-analysis which provides more flexibility than study-level meta-analyses. Our results accounted for the sampling design of each survey, making our estimates informative -if not representative- of the underlying population with self-reported diabetes in each country. Finally, the time frame we used for eye and foot examination is consistent with clinical guidelines.

Notwithstanding, there are limitations we ought to acknowledge for clarity and thorough interpretation of our results. First, although we included nine countries, there are more countries in LAC in need of these estimates. We analyzed all national health surveys available in the public domain. Where possible, local researchers and authorities should analyze their country-specific data to complement our findings. Second, except for the STEPS surveys, the other national health surveys herein analyzed did not follow the same protocols and procedures; nevertheless, their sampling frame targeted a nationally representative sample and the questionnaires they used were similar. Readers are advised to be careful if they intend to pool the results into one metric, and they should consider differences in sampling, in survey techniques, and when the surveys were conducted. Third, although the questions (Supplementary Table 2) about eye and foot examinations were part of the diabetes questionnaire or asked in the context of other diabetes-related questions, the wording of these questions was not specifically and unequivocally about eye or foot examinations for the prevention of diabetic retinopathy or peripheral neuropathy. Given the context in which these questions were applied we argue that they referred almost exclusively about diabetes. Nonetheless, if some participants answered these questions thinking outside the diabetes context, our results would be overestimated. Then, our findings would suggest there is a potentially lower proportion of eye and foot examination for the prevention of retinopathy and peripheral neuropathy in people with diabetes in LAC. Notably, report and desirability bias should be considered when interpreting the results though it is not possible to ascertain whether these biases would have moved the results upwards or downwards. For example, report bias (assuming the participants forgot they had their eye/foot checked) could have led to smaller proportions; conversely, desirability bias (assuming the participants said they had their eye/foot checked) could have led to higher proportions. Fourth, even though we included large national health surveys, the effective analytical sample was <10,000 adults. This, because we only included people with self-reported diabetes, and diabetes awareness is still poor in LAC [20] (hence the small number of people with self-reported diabetes). In this line, potential misclassification in self-reported diabetes should also be considered. For example, people could have answered positively to the question if they have been diagnosed with diabetes when they were told they had high blood glucose or pre-diabetes. Fifth, as reported in the methods section, the questionnaires did not dissect between type 1 and type 2 diabetes and we have, therefore, used the generic term diabetes throughout the manuscript. It may be the case that people with type 1 diabetes are more closely monitored than those with type 2 diabetes, hence the type of diabetes could have influenced our findings. This hypothesis can be reflected in the fact that the likelihood (prevalence ratio) of having had eye examination was higher for people receiving insulin (likely type 1 diabetes) versus people receiving oral medications (Tables 2 and 3). Moreover, the proportion of cases with type 1 diabetes in the 25–64 age group should be low with limited impact on the results. Nevertheless, in this case and from a public health perspective, the distinction between type 1 and type 2 is not utterly relevant. Our findings alert of a small proportion of people with diabetes receiving eye and foot examination, and policies, interventions, as well as guidelines should be strengthened to improve access to this preventive care regardless of whether patients have type 1 or type 2 diabetes. Sixth, even though the estimates accounted for the sampling design of each survey, given the small analytical sample size for each country, the results may be informative at the general population level, not necessarily representative of the whole underlying national population. Our results warrant further verification with larger samples, ideally at the general population level or with large registry-based platforms of people with diabetes. Seventh, according to data availability, we harmonized the variables about eye and foot examinations to be consistent across surveys. However, for two surveys regarding eye examination, this was not possible (i.e., the time frame in two surveys -Argentina and Mexicowas within the last year whereas for other countries the time frame was within the last two years). Thus, due to these potential differences we did not make strong comparisons between countries and focused on describing the results for each country. Regional organizations could clearly define a suggested time frame for eye and foot examinations, so that researchers and future national health surveys include questions with the same time frame for all countries.

6. Conclusions

The proportion of eye and foot preventive care among people with self-reported diabetes across nine countries in LAC is poor. In the three best performing countries, ~50% of patients with self-reported diabetes have received an eye examination in the last two years; however, foot examinations were consistently low in all countries. Eye and foot examinations are instrumental for the prevention or early diagnosis of diabetic retinopathy and peripheral neuropathy. Because recommendations for eye and foot examinations are clearly outlined in diabetes guidelines, opportunities and resources need to be secured to improve the uptake and compliance with these strategies to prevent diabetes-related complications.

Appendix A. Supporting information

Supplementary data associated with this article can be found in the online version at doi:10.1016/j.pcd.2024.03.001.