Middle East African Journal of Ophthalmology

: 2014  |  Volume : 21  |  Issue : 4  |  Page : 321--327

Retinoblastoma referral pattern in kenya

Joseph M Nyamori1, Kahaki Kimani1, Margaret W Njuguna1, Helen Dimaras2,  
1 The Department of Ophthalmology, University of Nairobi, Kenya
2 The Department of Ophthalmology and Vision Sciences, University of Toronto; Division of Vision Science, Toronto Western Research Institute; Sick Kids Research Institute, The Hospital for Sick Children, Toronto, Canada

Correspondence Address:
Joseph M Nyamori
The Department of Ophthalmology, University of Nairobi


Purpose: Kenya is a large country with a widely dispersed population. As retinoblastoma requires specialized treatment, we determined the referral pattern for patients with retinoblastoma in Kenya to facilitate the formulation of a national policy. Materials and Methods: A retrospective study was performed for retinoblastoma patients who presented from January 1, 2006 to December 31, 2007. Data were collected on the referral process from presenting health facility to the hospital where patient was treated. Data were also collected on the time interval when the first symptoms were noticed to the time of presentation at a health facility (lag time). For cases that could be traced to a referral hospital, the time delay due to referral (referral lag time) was recorded. Results: There were 206 patients diagnosed with retinoblastoma in 51 Kenyan and 2 foreign healthcare facilities, and they received final treatment at a Kenyan hospital. Mean lag time was 6.8 months (۰.45). Of all patients, 18% (38/206) were treated at the hospital where they first presented and 82% (168/206) were referred elsewhere. Of those referred, 35% (58/168) were lost to follow-up. The mean referral lag time was 1.7 months (۬.5). Conclusions: A significant proportion of cases presented late, and either delayed seeking further treatment or were lost after initial referral. We recommend the implementation of a national strategy that emphasizes early detection, documentation and follow up of retinoblastoma patients.

How to cite this article:
Nyamori JM, Kimani K, Njuguna MW, Dimaras H. Retinoblastoma referral pattern in kenya .Middle East Afr J Ophthalmol 2014;21:321-327

How to cite this URL:
Nyamori JM, Kimani K, Njuguna MW, Dimaras H. Retinoblastoma referral pattern in kenya . Middle East Afr J Ophthalmol [serial online] 2014 [cited 2023 Feb 6 ];21:321-327
Available from: http://www.meajo.org/text.asp?2014/21/4/321/142270

Full Text


Retinoblastoma is the most common pediatric eye cancer worldwide. [1] In Kenya, it ranks tenth among all cancers. [1] Cure is possible with early detection, prompt referral, timely treatment and follow up.

In the Kenyan referral system, children with suspected retinoblastoma rarely present to a healthcare facility where treatment can be provided, thus most patients are referred to another center. This referral may cause delays in treatment, or failure of patients to present to the referral center. We performed a retrospective study to document the referral process and associated treatment delays for children with retinoblastoma in Kenya.


0Selection of the study population

This retrospective study was conducted as a part of a larger study that estimated the incidence of retinoblastoma in Kenya [2] with approval from the Kenyatta National Hospital ethics board and the Kenyan Ministry of Health. Data were collected for retinoblastoma cases observed over a 2-year period between January 1, 2006 and December 31, 2007. After confirmation of true diagnosis of retinoblastoma (excluding those with histology inconsistent with retinoblastoma or lost/missing files) and identification of cross-referred cases to avoid duplicate counting, data were collected on: Age and gender, laterality, family history, presenting signs and symptoms, health facility/site of presentation, health facility/site referred to, time from first noticing signs of retinoblastoma to clinical diagnosis, time from clinical diagnosis to presentation at a referral health facility (if no presentation at referral facility, the case was considered lost), name of health facility where treatment was performed, and the type of treatment received.

Cases that were lost after referral were analyzed separately using the limited details recorded in the registers at the facility where they initially presented.

Statistics and population details

The data were analyzed using the Statistical Package for Social Scientists (SPSS) version 13. Odds ratio (OR) with 95% confidence intervals (CI) were calculated. A Yates corrected P value less than 5% was considered statistically significant. The independent samples t-test was used to compare equality of means.

In the period reviewed, no adverse events or changes in health policies that affected record-keeping were reported. This study assumed that all patients with retinoblastoma presented to health facilities where eye care personnel diagnosed, registered and referred them appropriately.


0Demographics of study population

After identifying cross-referrals, there were 206 unique cases of clinically diagnosed retinoblastoma who were presented in 51 Kenyan and 2 foreign healthcare facilities and received final treatment in a Kenyan hospital within the time period studied [Table 1]. Of the total patients, 18% (38/206) were treated at the hospital where they first presented (originating center) while 82% (168/206) were referred to another facility [Table 1]. Of the referred cases, 65% (110/168) were traced to the point of final treatment (final treatment center), while 35% (58/168) were lost after referral [Table 2]. This left 148 traceable patients; 16 patients were excluded (final diagnosis inconsistent with retinoblastoma or missing records [2] ) for a total of 132 evaluable retinoblastoma cases [Figure 1].{Table 1}{Table 2}{Figure 1}

The cases first presented in healthcare institutions from all Kenyan provinces, plus Tanzania (Mwanza) and Uganda (Tororo). Rift Valley and Central provinces observed the most cases at presentation (24 and 20%, respectively) while North Eastern and Coast provinces reported the least (3 and 5%, respectively) [Table 1] and [Figure 2]a. For the patients that presented at the foreign centers, two were of Kenyan origin. We were unable to determine if other Kenyan patients presented to foreign centers and received treatment or were lost, as our study focused on records available for children that received final treatment at Kenyan healthcare facilities.{Figure 2}

Lost cases after initial referral to a treatment centre

We examined the records of referred cases (n = 168) to define any characteristics that might predict lost versus traced cases. While in every province lost cases were observed, the proportion of lost cases varied by province; Nyanza province lost the most (52%, 15/29 cases) while Nairobi lost the least (27%, 4/15 cases) [Table 2] and [Figure 2]a. The lost cases originated from 22 of the 51 Kenyan healthcare facilities where cases initially presented [Table 2].

Lag time

The delay from noticing the first sign of retinoblastoma to presentation at clinic (lag time) was not recorded in 5% (7/132) of confirmed cases; therefore our analysis covers the remaining 95% (n = 125) [Figure 1]. The mean lag time was 6.8 ± 6.45 months (range, 0.25 to 30 months) [Table 3]. Lag time was longer in patients who presented with metastasis (8.3 ± 6.9 months) than those who did not (6.6 ± 6.4 months), however this was not significantly different (P = 0.3). The difference in lag time between bilateral and unilateral patients was not significant (6.4 months ± 6.4 vs. 6. 9 months ± 6.5, P = 0.7). There was a statistically significant difference in lag time among the province of presentation (P = 0.04); the highest mean lag time was in Coast province (13.8 ± 10.5 months). The reason(s) for the lag time was not recorded in most cases. However, the longest lag time was observed for a child from a nomadic community, where leukocoria remained unattended until the child presented with proptosis.{Table 3}

Referral lag time

Of the referred and traced cases that were confirmed retinoblastoma (n = 106), we could not determine the referral delay for 29% (31/106) of cases due to incomplete records. For the remaining 71% (n = 75), the mean delay from the date of first referral to the date of presentation at the final treatment center was 1.7 ± 2.5 months [Table 4]. The reasons for the delay were not documented. Referral lag time was longer for cases with metastases (2.5 ± 2.7 months) than those without (1.5 ± 2.4 months) but the difference was not significant (P = 0.2). When referred patients were analyzed by province of residence, the longest mean referral delay was noted in Rift Valley province (2.8 ± 3.8 months) [Table 4], however this difference was not significantly different from the lag times noted in the other provinces (P = 0.7).{Table 4}

Final treatment centers

The 132 evaluable patients were treated at 10 of the nation's treatment facilities [Table 5] and [Figure 2]b. Three of these final treatment centers did not receive referrals, but represent treatment of patients that primarily presented there (all in Rift Valley); the remainder (7/10) received both referrals and primary presentations. The majority of patients (64%) were treated at Kenyatta National Hospital (Nairobi), with no patients treated in North Eastern or Eastern Province [Table 5] and [Figure 2]b.{Table 5}

Treatment and outcomes

Of the all evaluable cases, 95% (126/132) were treated primarily by enucleation in one (108/132) or both (18/132) eyes, while 5% (6/132) declined any form of treatment. Of the 53% (70/132) of total cases that received chemotherapy, 2% (3/132) were palliative. We assumed that the majority of treatments with chemotherapy were to treat presumed risk of extraocular metastasis. Only 5% (6/132) received focal therapy, indicating they were amenable to eye salvage. Radiotherapy was administered to 12% (16/132). Focal and radiotherapy treatments were available only at one teaching and referral hospital, while chemotherapy was only available at the two teaching and referral hospitals. At the time of data collection 16% (21/132) of cases had died at hospital within an average of 7.2 months from diagnosis.


To the best of our knowledge, this is the first study of the Kenyan medical referral system in relation to the childhood eye cancer retinoblastoma. Our observations suggest striking effects on the retinoblastoma experience in Kenya: (1) A lengthy lag time results in diagnosis delay; (2) frequent referrals and the resultant lengthy referral lag time delays treatment; and (3) a significant proportion of referred cases are lost from the system, and presumably die.

The mean lag time observed in our study (6.75 months) was much longer than published figures from the UK (0.5 months) [3] , Belgium (3.2 months) [4] or the USA (1.5 and 2.25 months for unilateral and bilateral cases, respectively). [5] We speculate that this lag time significantly contributes to the high mortality observed among retinoblastoma patients in Kenya; however we were unable to assess outcomes for the cases we present in this study. Prompt treatment is essential for the cure of retinoblastoma, especially when the extent of tumor in the eye necessitates enucleation; all cases in our study required enucleation. Delay of enucleation of over 3 months in such eyes (IIRC [6] Group E) even when chemotherapy is used instead, has been shown to increase the risk of death. [7] Thus, we can only surmise that the longer mean lag time we observed suggests that Kenyan patients are already at a high risk of mortality by the time of first presentation. The range of lag times (1-30 months) we observed suggests that early diagnosis and prompt examination in clinic are possible in Kenya, but not common. This emphasizes the need for public education on the presenting signs of retinoblastoma, particularly in marginalized communities, where access to health information and services might be limited. This is consistent with a previous prospective study of 52 retinoblastoma patients from Kenyatta National Hospital that found the main reason for delay in the first presentation was ignorance of presenting signs of retinoblastoma rather than financial or other constraints. [8]

Most cases of retinoblastoma are unilateral, and enucleation is most often indicated for the sole affected eye. [9],[10] Follow-up by histopathological evaluation determines the need for adjuvant chemotherapy, when features indicating the risk for metastasis are observed. Bilateral disease is more complex, and may be treated by enucleation (even of both eyes, if the cancer poses a threat to the child's life, as for our 18/34 bilateral cases), or a combination of chemotherapy and focal therapy, if eye salvage is possible (only 6 our cases fit this category). [9],[10] Radiation is best avoided as primary eye salvage therapy, since it has been shown to increase the risk of second primary cancers. [11] Metastasis is difficult to cure. [12]

Referrals were common in our study population, indicating an inability of the center of first presentation to adequately treat the patients. In the Kenyan health care system, ophthalmologists are mainly found at secondary and tertiary level hospitals, whereas most patients initially present to a primary healthcare facility. In our study of retinoblastoma patients, however, patients initially presented even at secondary (district) centers with trained ophthalmologists and still required referral to a tertiary center [Table 2]. All of the patients in our study required enucleation of one or both eyes as primary treatment, which for trained ophthalmologists is a relatively simple procedure. Performed promptly on eyes with tumors dangerously close to becoming extraocular, enucleation can prevent metastasis, for which treatment and cure requires much more intensive and costly treatment. The significant number of referrals in our population suggests two possibilities: (1) The presenting centers could not perform enucleation and/or corresponding pathological examination of the eye, and chose to refer; and/or (2) the patients presented with advanced disease requiring not only enucleation and pathology, but also adjuvant chemotherapy, which was not possible at the presenting centers. A combination of both these factors may have resulted in the large number of referrals we observed. Building the capacity for presenting secondary centers that are equipped with ophthalmologists to perform enucleation, as well as access to a quick and reliable histology service, may reduce the number of referrals required, and reduce strain on teaching and referral hospitals. Indeed, this is the goal of the Kenyan National Retinoblastoma strategy, [13],[14] which has strengthen a network of ophthalmologists and other caregivers nationwide, to begin to triage retinoblastoma cases, so that simple cases can be managed locally, and only complex cases be referred to tertiary centers.

The aim of reducing referrals is to combat the high number of lost patients and subsequent delayed treatment, both of which we observed in our study. We studied the lost cases in an attempt to find any features that might predict the risk of loss to follow-up. The largest proportion (25.9%) of cases was lost after referral in Nyanza province. Geographical distances from the patient's home to the clinic may account for lost cases, as travel may be financially impossible for many families. In addition, patients at the border towns of Busia and Moyale may have preferentially sought treatment at nearby facilities in Uganda (Tororo) and/or Ethiopia, however, this does not account for the lost cases in the facilities further inland. Central province also lost a significant number of cases (19%) even though it is the nearest to Kenyatta National Hospital in Nairobi, which received the most referrals, suggesting distance alone does not explain all the lost cases. Anecdotal evidence indicates that when families return home after obtaining a referral, they may be convinced by other family members or community members to seek alternative forms of treatment, like a traditional healer. A better understanding by parents of the risks involved in delaying treatment may improve compliance and speedier travel to referral centers. Also, when coupled with our previous recommendation to perform more enucleations at capable presenting centers, referral may not be necessary for as many cases.

The mean delay caused by referral was nearly 2 months [Table 4], indicating patients did not seek further treatment promptly. To a growing retinoblastoma tumor, 2 months without intervention is more than enough time to put a child's life at risk. Many of the patients who did not initially present with metastasis may have developed it in this time, or even extension of the tumor into the choroid and/or optic nerve, all of which may indicate the need for adjuvant chemotherapy. Indeed, a large number of patients underwent chemotherapy (but did not require focal therapy) suggesting most children were treated for either metastatic disease, or perceived risk of metastasis. At the time of data collection, we observed a 16% mortality rate, but the large number of patients undergoing adjuvant chemotherapy suggests many more children eventually died of their disease. This is unacceptable for a disease that is treatable when first diagnosed as intraocular and treated promptly. The delay caused by referral could have turned a tumor that is curable into one that is not. We could not determine the reasons for the referral delay for the families, however a previous study suggested mismanagement at the first center followed by ignorance of the severity of disease by medical personnel and families, distance and financial constraints. [8]

We recommend initiatives to assist families to seek prompt medical attention for retinoblastoma. Together with increasing public awareness of retinoblastoma to promote early diagnosis, improving the capacity of presenting centers where ophthalmologists are available to perform simple enucleation and access effective histopathology services could reduce referrals and have dramatic improvements on patient survival. Subsequently in 2008, a multidisciplinary group initiated the Kenyan National Retinoblastoma Strategy to strengthen the referral system, improve treatment and raise awareness. [13],[14] Key studies over the next few years will reveal the results of this strategy on outcomes for children with retinoblastoma.

This study provided key insight on the referral and treatment pattern of retinoblastoma; however it was not without its limitations. At most government health facilities, patients retained their outpatient records and could not be traced because the telephone and address details were not routinely recorded in clinical registers at these health facilities. Therefore, it was difficult to track cases that were lost after referral. Different record keeping systems at the various health facilities may have influenced our retrieval rate. We are working on a prospective registration of cases in a national database intended to improve follow-up and provide more accurate data for eye health planning in Kenya.

Our study highlights the need for increased awareness of retinoblastoma among the public as well as health care providers to reduce lag time, and enhanced resources in presenting treatment centers to reduce referrals for cases that can be treated with enucleation alone. These two key actions could result in dramatic increase in overall survival of patients with retinoblastoma in Kenya.


The Division of Ophthalmic Services, the Kenya Ministry of Health staff at all participating institutions, Kenyatta National Hospital ethics board and registry staff; Moi Teaching and Referral Hospital - Drs. Odede, Wanjala, Rono; PCEA Kikuyu Eye Unit - Drs. Mundia, Kabiru; Lions Sight First Eye Hospital - Dr. Khan; Mombasa Lighthouse - Dr. Matende; Sabatia Eye Hospital, AIC Litein and AIC Tenwek Mission hospitals; and the Kenya Medical Research Institute - Dr. Mutuma and Ms. Rugut. We also thank the members of the Kenyan National Retinoblastoma Strategy group for their discussions and input.


1Mutuma G, Korir R. Cancer Incidence report 2000-2002. Kenya Med Res Inst 2006;3:26.
2Nyamori JM, Kimani K, Njuguna MW, Dimaras H. The incidence and distribution of retinoblastoma in Kenya. Br J Ophthalmol 2012;96:141-3.
3Goddard AG, Kingston JE, Hungerford JL. Delay in diagnosis of retinoblastoma: Risk factors and treatment outcome. Br J Ophthalmol 1999;83:1320-3.
4Wirix M, Parys-Vanginderdeuren R, Casteels I, Uyttebrouck A. Delayed diagnosis of retinoblastoma. Bull Soc Belge Ophtalmol 2000;278:37-41.
5Butros LJ, Abramson DH, Dunkel IJ. Delayed diagnosis of retinoblastoma: Analysis of degree, cause, and potential consequences. Pediatrics 2002;109:E45.
6Murphree A. Intraocular retinoblastoma: The case for a new group classification. In: Singh A, editor. Ophthalmic Oncology, Ophthalmology Clinics of North America. Philadelphia: Elsevier Saunders; 2005. p. 41-53.
7Zhao J, Dimaras H, Massey C, Xu X, Huang D, Li B, et al. Pre-enucleation chemotherapy for eyes severely affected by retinoblastoma masks risk of tumor extension and increases death from metastasis. J Clin Oncol 2011;29:845-51.
8Kahaki K, Ilako D, Kollman M. A review of retinoblastoma presentation, diagnosis and management at the Kenyatta National Hospital. Ophthalmology. Nairobi: University of Nairobi; 2000.
9National Retinoblastoma Strategy Canadian Guidelines for Care: Strategie therapeutique du retinoblastome guide clinique canadien. Can J Ophthalmol 2009;44 Suppl 2:S1-88.
10Dimaras H, Kimani K, Dimba EA, Gronsdahl P, White A, Chan HS, et al. Retinoblastoma. Lancet 2012;379:1436-46.
11Kleinerman RA, Tucker MA, Tarone RE, Abramson DH, Seddon JM, Stovall M, et al. Risk of new cancers after radiotherapy in long-term survivors of retinoblastoma: An extended follow-up. J Clin Oncol 2005;23:2272-9.
12Dimaras H, Héon E, Budning A, Doyle JJ, Halliday W, Gallie BL, et al. Retinoblastoma CSF metastasis cured by multimodality chemotherapy without radiation. Ophthalmic Genet 2009;30:121-6.
13Dimaras H, White A, Gallie BL. The Kenyan National Retinoblastoma Strategy: Building local capacity in the diagnosis and management of pediatric eye cancer in Kenya. Ophthalmol Rounds 2008;6:1-6.
14Kimani K, Ouma B, Gallie BL, White A, Dimaras H, Livingstone M. and the Kenyan National Retinoblastoma Strategy Group. Rati's Challege: A Vision for Africa. Report from the First Kenyan National Retinoblastoma Strategy Meeting, Daisy's Eye Cancer Fund, 2009. p. 39.