Referral patterns for retinoblastoma patients in Ethiopia

Retinoblastoma (RB) is the most common intraocular malignancy of childhood, representing approximately 2.0% of all pediatric malignancies [1]. It is the most frequent intraocular malignancy of the eye in childhood occurring in early childhood; two-thirds are diagnosed before 2 years of age, and 95% before 5 years [2]. The incidence in various well-studied population groups around the world varies from 1 in 15,000 to 1 in 20,000 live births [3, 4], and is increasing in regions where the majority of patients survive and have affected children of their own [5].

The most common presentation of RB is leukocoria but if this early sign does not prompt the family to seek care, the tumor can continue to grow and spread relatively undetected until it may be too late for a cure, which can happen within months. In many low-income (LIC) and middle-income countries (MIC) presentation of RB is often late, and characterized by orbital involvement and metastasis [6‐8]. This aggressive tumor grows quickly, metastasizes early, and can be fatal; however, it is also curable if diagnosed and treated early. In general, survival from retinoblastoma in LIC, lower MIC, upper MIC, and high-income countries (HIC) is 30%, 60%, 75%, and 95%, respectively [6]. Yet even within these national income categories, there is wide variation between treating RB facilities; for example, single institution studies from LICs and upper MICs report survival of up to 73% [9] and 96% [10], respectively.

Likewise, the burden of disease is much greater in LICs and MICs, where an estimated 84% of all children with cancer in the world reside [11]. Survival and visual outcome in RB are dependent on the severity of disease at time of presentation, so early detection is paramount to survival; however, children in East Africa present an average of 9 to 11 months later than their counterparts in the US and Canada, leading to poor survival rates [4].

Ethiopia has one of the highest estimated burdens of RB in sub-Saharan Africa [12]. RB in Ethiopia is characterized by delayed presentation and advanced disease [8, 13]. Menelik II Hospital is Ethiopia’s most advanced RB treatment center. This study was conducted to understand the path from first onset of symptoms to arrival at Menelik II Hospital in order to address delays in accessing timely care.

This is the first study in relation to the referral pattern of RB patients in Ethiopia, examining timing of each step in seeking care. This study assessed the experience of patients from first noting signs of RB in their home, to arriving for treatment at Menelik II Hospital, Ethiopia’s most comprehensive RB treatment center. The study revealed the magnitude of lag time, specifically in the delay in treatment, as well as the relationship of caregiver knowledge, cost and travel distance to these delays. These findings have implications for developing health education campaigns and public awareness programs, and development of health care resources and counselling guidance for patient caregivers.

A prospective study from 11 RB centers showed that the lag time between first sign and treatment of RB was 5.4 times higher in LIC, 3.2 times higher in lower MIC and 1.6 times higher in upper MIC when compared to HIC [15]. The overall mean lag time observed in our study (14.31 months) was much longer than studied from UK (38 days) [14], North China (1 month) [16], Thailand (5 weeks) [17], Central and Southern China (54 days) [18], India (3 months) [19], Brazil (5.8 months) [20], and other sub-Saharan African countries like Kenya (6.8 months) [21] and Tanzania (10.6 months) [22], but shorter than a study from Mali (50 months) [23].

In comparison to other studies, the mean lag time 1 in our study population (7.97 months) was much longer than a report from China (8 days) [16] and UK (28 days) [14]. Participants in our study indicated that lack of knowledge was a reason for their late presentation, underscoring the importance of raising public awareness of the early signs of RB. This will have positive effects for pediatric eye conditions beyond RB, as leukocoria can also be indicative of Coats disease, persistent fetal vasculature, or familial exudative vitreoretinopathy, among others [24‐26].

Lag time 2, the delay from first presentation to diagnosis by the first care providers, was more prolonged in our study population (5.2 months) than in UK (2 days) [14] and China (3 days) [16]. A lag time of this magnitude even though patients have already presented to a health facilities possibility indicates poor awareness of RB among health care providers or potentially incorrect choice of initial health facility by patient caregivers. The wide range of lag time 2 (0‑38 months) we observed could be due to the difference in diagnosis capacity of the primary care providers for patients with RB.

Lag time 3 represented the time to treatment of RB once the diagnosis was made. Though its much shorter than lag time 1 and 2 of our study population, lag time 3 in our study participants (3.18 weeks) is longer than that reported in UK (6 days) [14]. Besides, there is wide range in delay of treatment initiation for RB patients (0–13 months). The reasons for the delay in treatment could be multiple, but lack of standard operating procedure and a clear guideline at Menelik II Hospital can be reasons.

Prolonged lag time before initiation of RB treatment can lead to advanced disease presentation and poor patient outcomes. A study of 4351 RB patients from 153 countries from different national income status revealed that patients from LIC had a larger proportion of patients with symptoms of advanced disease compared to patients from HIC [27]. A Brazilian study showed that a six-month delay in diagnosis was linked to an increased risk of extraocular disease [28]. Similarly, a study from India revealed that prolonged duration of symptoms > 6 months was predictive of high-risk histopathologic features of RB [29]. In our study, prolonged lag time of 12 months or more was significantly associated with death (Table 4), suggesting that efforts to reduce lag time are imperative to save lives from retinoblastoma.

The literature points to several interventions which may be successful in reducing lag time. With respect to lag time 1, in Honduras a relatively inexpensive awareness program in conjunction with a national vaccination campaign was shown effective in an early diagnosis of retinoblastoma; they observed an increase in the number of referrals and, more importantly, a significant decrease (from 73 to 35%) in the occurrence of extraocular disease [30]. In the 1990s, in Brazil, an education program comprising primarily magazines and news articles about RB leading to earlier diagnosis and a reduced (from 56 to 17%) frequency of extraocular disease [31].

Interventions at the level of the health system may also prove effective to improve the referral pathway and facilitate connection of families to appropriate care; this would have been useful for the 21 patients in our study who were not referred to Menelik II Hospital after presenting at a primary or secondary eye care centre. In the Ethiopian health care delivery system, health extension workers are the main grassroots health work forces for promotive, preventive and basic curative services focused on common diseases, and other community-based health care services [32]. However, a previous study revealed that their knowledge regarding common eye diseases is limited [33]. Enhancing the health extension training program to include instruction on common signs of eye cancer in children may improve awareness and facilitate timely and appropriate referral of children with RB. Additionally, the delay in presentation to health facilities and diagnoses may be alleviated by public education through mass media and further training of primary and secondary health workers. Integration of screening for RB and other common pediatric eye diseases at vaccination or under five clinics may be another way to improve early detection of the cancer within the heath system.

In Ethiopia ophthalmologists work at secondary or tertiary health centers, but in this study we observed that the majority of the patients were first seen at primary health facilities, increasing the lag time to achieve diagnosis, and exposing patients to unnecessary costs. We also noted that the secondary health centers in which patients presented to prior to arriving at Menelik II Hospital did not provide RB treatment, though theoretically they should be able to provide some aspects of care; this requires further study. The underrepresentation of patients and caregivers from the regions of Harar, Dire Dawa, Afar or Benishangul-Gumuz may be reflective of affected patients ending up at health facilities other than Menelik II Hospital. However, it is also possible that low capacity for eye care in Afar and Bunshangul-Gumuz regions may be missing RB patients in those regions, warranting further study.

Our results show that the RB patients in Ethiopia have long overall lag time from notice of first notice of sign to receiving treatment. Lack of RB awareness among caretakers, high cost and travel burden were significant barriers to receiving timely care. Delays in seeking care, diagnosis and on time initiation of treatment for patients with RB may be alleviated by public education and targeted capacity building in the Ethiopian health care system.