Introduction

Diabetes in sub-Saharan Africa (SSA) is predicted to more than double by 2045, which is the highest increase worldwide [1]. Diabetic retinopathy is the commonest cause of blindness in people of working age, which has compounded consequences on the individual, family and wider society [2].

Typically, screening for diabetic retinopathy (DR) is performed annually and patients who need further management or an intervention are referred to the respective eye care centre. However, there are often limited screening and treatment programmes for DR particularly in SSA. As diabetes and its sequelae become more prevalent in SSA it is important these are developed.

Globally, 56% of the blind are female [3]. There are several reasons for this, not least that women live longer than men. Furthermore some blinding conditions are more likely to affect women than men, for example, trachoma and cataract [4, 5]. However, another key factor is that women do not access eye care services as often as men [6].

The Kilimanjaro Diabetic Programme (KDP) has been running since 2010 and screens eyes of persons with diabetes across the Kilimanjaro region of northern Tanzania in diabetic clinics. Of concern, only 42% of patients referred to the tertiary eye hospital (Kilimanjaro Christian Medical Centre) for treatment after screening attended [7]. It was noted that women were significantly less likely to attend a follow-up appointment at the tertiary eye hospital when referred. The aim of this study was to explore gender biases amongst persons registered with the KDP to help better understand why women access diabetic eye care services less than men.

Methods

This prospective study was carried out between September 2014 and February 2015. All patients registered with the Kilimanjaro Diabetic Programme (KDP) were considered eligible.

A questionnaire of closed questions and a separate semi-structured qualitative interview guide (Appendix 1) were developed by a team of eye care professionals involved in screening and management of patients with diabetes. The questionnaire and guide were piloted with 10 patients prior to the start of the study to ensure adequate understanding.

Quantitative component

307 patients were selected from the KDP database through random number generation to take part in the study.

Once the participants were identified, they were consented for participation in the study with an explanation of what this would involve. The questionnaire was then administered in Kiswahili, by a native speaker.

During the interviews, data were entered on paper forms and later entered into Microsoft Access (2007). STATA version 13 was used for statistical analysis.

Qualitative component

A total of 33 participants, selected through random number generation, underwent semi-structured qualitative face-to-face interviews from 7 diabetic clinics in the Kilimanjaro region. No participant took part in both the qualitative and quantitative components.

The interviews were conducted in Kiswahili and were recorded, after consent from the participants. They were later transcribed into Microsoft Word (2007) and translated into English by a bilingual speaker. NVIVO version 12 was used for data management.

Ethics approval

The study was approved by the Kilimanjaro Christian Medical College University ethics committee. Informed written consent was obtained from all participants.

Results

Quantitative component

A total of 307 persons were interviewed. 7 participants were excluded due to incomplete questionnaire forms. The remaining 300 participants had a median age of 61 years (IQR 55–69 years). 193 (64.3%) were female and 107 (35.7%) were male with a median age of 58 years (IQR 53–67 years) and 65 years (IQR 56–73 years) respectively (p = 0.002). The mean duration of diabetes was 6.76 years (SD 5.5 years), with no significant difference between men and women.

Of the included 300 participants, 252 persons (84%) had been screened for DR through the regional screening programme. Of the 252 screened persons, 75 persons (29.8%) were referred to KCMC and of the 75 persons referred 29 (38.7%) attended the referral appointment. 55.2% of men referred to the specialist eye clinic attended their follow-up appointment compared with 44.8% of women; however, this difference was not significant (p = 0.422).

Table 1 shows details of the education level, insurance status and the self-reported answer given when asked whether the participant would attend KCMC (the tertiary hospital) if referred, split by gender. Females were significantly less educated (p < 0.001) and self-reported as less likely to attend the tertiary hospital if referred. There was no significant difference in rates of health insurance between men and women.

Table 1 Socio-demographic details of participants and self-reported compliance with referral recommendation split by gender.
Full size table

Data were available on 290 participants concerning marital status with 10 preferring not to say. A significantly higher proportion of females were not married or widowed compared to men (Table 1).

There was a good level of understanding regarding diabetes, DR and the need for DR screening, with no significant difference between men and women. Of the 300 participants, 96% understood that it is important to attend hospital appointments for diabetes even if one is asymptomatic and, similarly, 96% knew that DR can cause blindness. 94% and 87% of participants respectively had heard of DR and understood the need for screening even if one has no visual complaints.

Sub-analysis of the married participants showed that females were less likely to make independent financial decisions compared to men (Table 2). Females were also significantly more likely to depend on their spouse when deciding if, and when, they, or family members, should attend hospital appointments compared to men (Table 2). There was no significant difference between education level and whether individuals made financial decisions. However, the more educated were more likely (p = 0.016) to decide if family members should attend hospital.

Table 2 Intra-family and financial decision-making practices split by gender amongst married participants (n = 229).
Full size table

On multivariate analysis, men were significantly more likely to make both financial decisions in the household and to decide if, and when, family members should attend hospital compared to women. This was independent of age, education level and whether the participants were from an urban or rural area (Table 3).

Table 3 Univariate and multivariate logistic regression analysis assessing the association between financial decision making and intra-family decision making and socio-demographic factors amongst married participants (n = 229).
Full size table

Qualitative component

Thematic analysis was used to interpret the qualitative data. The results from the qualitative interviews are presented in two sections. Firstly, the perceived need for DR screening and themes relating to the poor follow-up rates seen within the Kilimanjaro Diabetic Programme. Secondly, the results focus on the family dynamics between husband and wife in regards financial decision making.

Perceived need for diabetic retinopathy screening

Most interviewees had a good understanding of the potential for diabetes to affect the eyes and lead to a reduction in, or loss of, vision, with no difference between men and women. This is consistent with our results from the quantitative section of the study that also showed no significant gender difference in the understanding of the threat to vision from diabetes and the need for regular DR screening.

However, despite a good level of understanding of the risk to vision from diabetes, the rates of follow-up for those referred after DR screening remain poor. The interviews identified several themes potentially explaining this, including: cost of treatment and transport to and from the hospital, fear of treatment and a cultural lack of understanding and appreciation of chronic disease.

The majority of participants interviewed (n = 30) were aware that diabetes could affect vision and that screening was important in the absence of symptoms:

  • “Definitely! Because they say diabetes runs slowly to the eyes so it’s important to check frequently.”

When asked why a lot of patients do not comply with follow-up recommendations, financial concerns were most commonly cited:

  • “…it could be due to the cost, because KCMC is very far. No-one does not want treatment Doctor. The problem is money!”

The costs incurred included not only the cost of treatment, but the cost of transport, accommodation and the loss of income through missed work, which was particularly cited as a factor for men. An escort, which is often required, doubles the cost of travel and accommodation. In addition to their transport and accommodation costs, some escorts were reported as asking for payment to accompany the patient:

  • “…you must find someone to escort, get treatment pay for transport for two people and others will say pay me if you want me to escort you.”

Several patients commented on the cultural relationship with disease and a limited appreciation of the chronicity of diabetes. There was, amongst some participants, a lack of understanding the concept of being “sick” without any symptoms:

  • “Most Africans don’t have the habit of checking our health status until we get sick.”

Family dynamics with regard to financial decision making in households

Men were cited as acting as head of the household (n = 14) and making financial decisions most commonly; thereby determining access to healthcare:

  • “Of course father is the final decision maker! Because he is the head of the family. And all family members depend to him!”

There were some participants who stated that, even when the male household member is absent, a female household member is unable to make decisions and must wait until the man household member returns:

  • “…no one, even a wife can make decision in my absence. Without my permission she cannot decide!”

However, a relatively large proportion of respondents (n = 11) stated that men and women make financial decisions together. This arrangement, whereby decisions are made collectively as a couple, seemed to be more common in educated families:

  • “Both mother and father. My husband understands he was a medical assistant…for those who are not educated, the husband is deciding and sometimes those not educated take money and use for alcohol.”

In those participants who reported female household members making decisions, the purchase of alcohol by their husbands was cited as a reason why women made financial decisions:

  • “Father if given to make decision may use in bad manner like buying alcohol. But mother use for the benefit of the family like buying food and important things needed at home.”

Discussion

Men, in many areas of low- and middle-income countries, are twice as likely as women to access eye care services [8]. This study is the first to report and investigate reasons for gender differences in compliance with follow-up recommendations after screening for diabetic retinopathy (DR) in Africa.

World Health Organization (WHO) recognises regular and reliable follow-up of DR patients and timely treatment as key components in the effective management of the condition [9]. This is a particular challenge in low-income countries, such as Tanzania. There was a good level of understanding regarding the potential for sight loss from DR and on the importance of screening, yet follow-up remains poor. Financial challenges, fear of treatment and a lack of understanding of disease chronicity were cited as reasons for this. Similar themes pose challenges in the management of other chronic ocular conditions in SSA, such as glaucoma [10]. The reasons for poor follow-up of persons with DR are not well studied; however, are likely to be multi-factorial and not solely a consequence of financial challenges [11].

As with other eye health conditions, data from the Kilimanjaro Diabetic Programme (KDP) demonstrates that women are significantly less likely to comply with referral recommendations post DR screening than men [7]. As the number of persons with diabetes increases substantially over the next two decades in Africa [12], an understanding of why women access tertiary eye care services less than men, in relation to DR, and measures to address this will enable more effective screening and treatment facilities to be developed; potentially reducing visual impairment and blindness. For example, it is estimated that if women were to receive cataract surgery at the same rate as men, blindness and visual impairment from cataract in low and middle income countries could be reduced by up to 11% [8].

The results from this study suggest that women are significantly less likely than men to have financial independence and women more commonly rely on their spouse to make financial decisions within the household. The is irrespective of age and education. Moreover, women are significantly less likely than men to decide if household members should attend hospital, again irrespective of age and education. This provides insight into the social and family dynamics that may explain why women are less likely than men to access eye care services.

In African societies, it is typical for financial decisions to be made by male family members; women often do not have financial independence to make decisions regarding their own healthcare [13]. As seen in the KDP, this may contribute to women’s reduced engagement with eye care services. This is likely to be particularly relevant for eye care services where the disease process is not always obviously apparent, such as DR.

There are no previous data on gender discrepancies within DR screening and treatment programmes in Africa. However, there are three strategies that have been shown to improve cataract surgical coverage for women in low- and middle-income countries: the use of outreach services, educating family members and using women to reach women [14, 15].

Outreach services could, for example, include the use of portable laser equipment to enable the treatment of DR at outreach clinics. This would help both the general low rates of follow-up, as seen in the KDP, as well as increase access to DR treatment services for women.

It is important to educate whole family groups on the detrimental effects of diabetes on eye health. As diabetic eye disease is not always obviously apparent and can be advanced and remain asymptomatic, it is important that male family members understand and appreciate the importance of compliance with screening and referral. This will help empower women to engage with eye care services.

The use of female healthcare workers to engage with family groups and women has been shown to be potentially beneficial in Egypt and Pakistan [14, 15]. In Egypt, for example, female health visitors were employed to educated family groups and particularly women on eye health. The intervention significantly improved blindness and visual impairment in women and reduced the prevalence of cataract and trachomatous trichiasis in women as well as improving engagement for men [14].

It is important that interventions such as those outlined above are audited as DR screening programmes are developed; data should be split by gender when published. This will allow an understanding of the challenges faced and analysis of the effectiveness of interventions on engaging women with eye care services.

In conclusion, women represent 56% of global blindness [3]. In low-income countries, there is reduced engagement with eye care services in women compared to men. As DR screening programmes are developed in low-income countries, strategies, such as the use of outreach screening combined with treatment services and using women to educate women and family groups, should be built into screening programmes, with audit data split by gender. This could help to alleviate the growing burden of blindness and visual impairment from DR caused by the under-utilisation of eye care services by women.

Summary

What is known about this topic

  • Women make up 56% of the global blind.

  • In many low- and middle-income countries men are twice as likely as women to access eye care services.

  • Diabetic retinopathy is an increasingly important cause of vision loss in low- and middle-income countries; however, follow-up within existing screening services is poor, particularly for women.

What this study adds

  • The study highlights cost, fear of treatment and a lack of appreciation of disease chronicity as important reasons for the low rates of follow-up for those referred after screening for diabetic retinopathy.

  • The study is the first to explore why women are less likely than men to access diabetic retinopathy treatment services after screening in Africa.

  • Women were noted to have less financial independence than men and were less able to make independent decisions relating to access to healthcare than men in this region of Tanzania.