Determinants of delayed diagnosis among pediatric cancer patients from Ayder Comprehensive Specialized Hospital, Mekelle, Northern Ethiopia

According to the 2016 estimates of the International Agency for Research on Cancer (IARC), 300,000 new cases of cancer in children less than 19 years appear in the world every year. Of these, 80% live in low and middle-income countries (https://​www.​acco.​org/​global-childhood-cancer-statistics). Cancer is an important cause of childhood mortality, with an estimated 80,000 cancer-related deaths per year worldwide with developing countries taking the largest share (https://​www.​acco.​org/​global-childhood-cancer-statistics). In LMIC advanced disease at presentation and delayed diagnosis of cancer in children is common, however there are limited studies regarding the determinants of delayed diagnosis [1‐4].

The determinants of delayed diagnosis in developing countries varied across studies. Some of the factors associated with delay includes tumor type, health insurance and health care system. Studies from Western Kenya and Uganda found longest delay among Burkitt lymphoma [5], retinoblastoma and Hodgkin lymphoma from Nigeria [3, 6] and retinoblastoma patients from Kenya had longest median patient delay as well as physician delay [1]. South African study demonstrated considerable delay in diagnosing childhood cancer due mostly to a physician delay [7]. Absences of health insurance in Kenya was associated with delay and abandonment of treatment [1]. Dang-Tan et al. published a review of 23 studies that explore diagnosis delays in childhood cancer and revealed that the determinants of delay in developed countries also varied across studies [8].

Early diagnosis is fundamental for cancer management because it allows treatment of early stagedisease, which results in better prognoses [9]. The delays in cancer diagnosis may be related to the parents, the health personnel or the health system. Failure to recognize cancer symptoms may result in patient delay according to national survey from NHS patients [10]. Health care providers are expected to make the diagnosis of cancer as early as possible, but most cancer symptoms are vague and nonspecific making it difficult to detect early [11]. Knowing the different levels of delay is the first step to improve the care and treatment options given for the patients. However, most of the studies addressing delay were done in high-income countries [12‐15].

In Ethiopia, there are approximately 6000 estimates of new cases of pediatric cancer each year, according to INCTR-USA hospital twinning initiative with Tikur Anbessa Specialized Hospital. Most children present late, with advanced disease [16].

There are no comparable studies available in Ethiopia about the determinants of diagnosis delay. But, we have observed that a lot of patients visit our center late and many days had elapsed even after visiting our center until the diagnosis of cancer. Once the diagnosis was made, treatment delay was not an issue; most patients started treatment with in few days of diagnosis. Hence, understanding the potential factors influencing the delayed diagnosis of cancer is needed to fill in the existing gap in knowledge and develop practical strategies to address the delay in cancer diagnosis among the pediatric population in Ethiopia.

The aim of this study was to identify determinants of delayed diagnosis of pediatric cancer patients at Ayder Comprehensive Specialized Hospital. Earlier presentation leads to early diagnosis of cancer which is a fundamental goal in oncology because it allows an opportunity for timely treatment. Consequently, the prognosis may improve, and a cure can be attained with minimal side effects. In the current study, the age of the child, the area of residence, parent’s level of education, family income and health insurance were all found to be statistically significant determinants of delayed diagnosis of pediatric cancer patients.

Delay in diagnosis among pediatric cancer patients was common in our settings, with much of the delay occurring prior to the first health encounter. About 71 (69.6%) had significant total diag nosis delay i.e. > 170 days. The median patient delay was 50 days, the median physician delay was 32 days and the median diagnosis delay (i.e. patient delay + physician) was 90 days. The Kenyan and Nigerian total delays were 102 and 110 days respectively which was longer than this study [1, 2]. This could be explained by the inclusion of treatment delay in the studies from Kenya and Nigeria. Studies from Egypt and South-Africa reported similar findings with a median total delay of 47 and 34 days, respectively. The short diagnosis delay reported from South-Africa could be a reflection of the superior infrastructure which was similar to the delays found in high-income countries, such as Canada [4, 7, 13]. The main reasons for the delayed presentation were use of alternative medicine in 47(46.1%). Alternative medicines sought by parents included: holy water 25 (24.5%) cases, praying ceremonies 4 (3.9%) cases, visiting herbalist 12 (11.7%), use of over the counter drugs 3 (2.9%) and others. A study done in Kenya reported that more than half of patents have used alternative medicine prior to admission [1]. The use of holly water was unique in our setting which was sought by significant number of patients (i.e. 25). There are many mentions of holy water in Ethiopian church literature [17]. Despite holy water is considered a cure to a variety of illness in Ethiopia, literature is scarce. This could be due to the highly spiritual community [18].

Age of the children was a significant predictor of delayed diagnosis. Fajardo-Gutierrez et al. found that the risk of increased delay for children between the ages of 10 years and 14 years is 1.8 times that of infants younger than 1 year of age (OR = 1.8(95% CI, 1.4–2.3) [17]. Another Kenyan study showed that the longest total delay was for the age group 6–10 years though it was not significantly longer when compared to the other age groups [1]. It was also reported that diagnosis delay was shortest for children aged 0–2 years despite no significant differences in histopathology, grade or location of tumors [18]. Those younger children had the shortest delay which could be because the common tumors in this age group are aggressive and may lead to a faster presentation and earlier diagnosis.

This study did not find a significant difference in the diagnosis delays between the two genders which is consistent with many studies. The sex of the child didn’t influence diagnosis delay in different studies [3, 4, 7, 15]. This may imply that there are no differences in health-seeking behavior regardless of the gender of the child.

The fact that children are usually under the care of their parents, parental characteristics and behaviors are also important factors in recognizing symptoms and signs of cancer. Thus, socioeconomic status and parental education were also identified as determinants of delayed diagnosis in this study. Parents who never attended school were more likely to be diagnosed late as compared to those who attended primary school and above school. In the Argentinean study, patients whose parents had an elementary education or lower had a greater risk of longer patient delay [18]. This was also observed in Mexico, where children whose parents had the lowest level of education had longer delays in diagnosis than children with parents with the highest level of education (OR = 1.4 for fathers and 1.5 for mothers) [17]. The rural residency was also associated with a significant delay in diagnosis which was consistent with Nigerian study where patients with longer travel distance and rural residence was associated with a delayed presentation [2]. This could be due to long travel distance and lack of transportations.

Parents with monthly income less than 1000 ETB were around 6 times more delayed in this study. This was consistent with a study done in Egypt where lower socioeconomic status and delayed diagnosis showed statistically significant correlations (P < 0.001) [4].

Having health insurance resulted in significantly shorter diagnosis delay compared to those without. This is consistent with a study done in Kenya where having insurance reduced the risk of abandonment [19]. Another study done in Kenya showed that health-insurance at diagnosis resulted in a significantly less patient delay (P = 0.049) [1].

Pediatric cancer patients who used holy water prior to first health encounter were 3 times more likely to get delayed and those who think cancer is incurable were 3 times delayed than those who think that cancer is curable. The use of holly water was unique in our setting which was sought by significant number of patients (i.e. 25). There are many mentions of holy water in Ethiopian church literature [19]. Despite holy water is considered a cure to a variety of illness in Ethiopia, literature is scarce. This could be due to the highly spiritual community.. Other factors which were found to be determinants in different studies including the type of cancer, parental occupation, and type of referring facility were not associated with delayed diagnosis in this study [1, 14, 20]. Therefore, some associated factors that were found to be determinants of delayed diagnosis in one study may not necessarily be determinants in another study supporting the argument that possible determinants of delay differ across the geographical location because of socioeconomic differences, beliefs and health-seeking habits of the specific community and other factors.

This study had certain limitations, including its relatively small sample size and recall bias. Recall bias was an issue when asking about the beginning of symptoms, but we attempted to achieve accuracy by crosschecking with other family members, reviewing dates of referral letters and initial written suspected diagnoses.

Delay in diagnosis among pediatric cancer patients was common in our settings, with much of the delay occurring prior to the first health encounter. The median patient delay was 50 days, the median physician delay was 32 days and the median diagnosis delay was 90 days. About 71 patients had significant diagnosis delay. These findings are more or less similar to other studies in LMIC. Delayed diagnosis of childhood cancer was most influenced by the child’s age, residence, family’s socioeconomic status, and parental education, absence of health insurance, visit of holy water and caregiver’s perception on curability of cancer.

Much of the delay occurred prior to the first health encounter. Thus, every effort should be made to promote public and parental awareness of childhood cancer and promoting health insurance to those with low socioeconomic status. We also recommend that training on childhood cancer be incorporated into the curricula of medical training institutions.