Time to diagnosis of pediatric brain tumors: a report from the Pediatric Hematology and Oncology Center in Rabat, Morocco

Abstract

OBJECTIVE:

Delayed diagnosis of pediatric brain tumors is known to occur worldwide but is not well studied in developing countries. Here, we examined the extent of delayed pediatric brain tumor diagnoses in Rabat, Morocco and consider its potential causes and possible solutions.

METHODS:

We conducted a survey and interviews of the parents of children who were admitted to the Department of Hematology and Pediatric Oncology of Rabat Children’s Hospital from January 1, 2016 to June 30, 2016.

RESULTS:

The families of 27 patients (14 girls and 13 boys) participated in the survey and interview. The median patient age was 7 years (range, 1–15 years). The most common presenting symptoms were vomiting (n = 18) and headache (n = 17). The tumor locations were supratentorial in 13 cases and infratentorial in 14 cases. The median time to diagnosis was 2 months (range, 0.25–20 months). The longest times to diagnosis occurred in children older than 5 years and in patients with supratentorial tumors or low-grade glioma. We did not observe any differences in the time to diagnosis according to sex, socioeconomic status, or urban or rural origin.

CONCLUSIONS:

Delayed diagnosis of pediatric brain tumors is a universal problem, evidenced by many studies in different countries. We propose that a paradigm shift in medical curricula addressing the delayed diagnosis of pediatric brain tumors should occur in medical schools and clinical training programs.

Introduction

Brain tumors are the most common solid tumors that occur in children and are a contributor of cancer-related mortality and morbidity.^1-3 Management of pediatric brain tumors requires a multidisciplinary approach within pediatric oncology centers. However, relatively long delays in the diagnosis of brain tumors, compared with that of other childhood cancers, presents a unique challenge to their management.^4-6 Such delays in diagnoses often lead to urgent situations, high morbidity, cognitive impairment, and psychological sequelae.⁷

Delays in the diagnosis of pediatric cerebral tumors occur even in developed countries with abundant access to medical care.^8,9 In developing and middle-income countries, however, studies of the incidence and survival associated with pediatric cerebral tumors are limited, and those reporting long-term follow up are lacking.⁹ In Morocco, data describing the incidence and timing of diagnoses of pediatric brain tumors are particularly limited, and most reports are based on epidemiologic and histologic studies.¹⁰

The Pediatric Hematology and Oncology Center (PHOC) of the Rabat Children’s Hospital in Morocco is a major pediatric oncology center dedicated to the treatment of children under 15 years. Since its inception in 1980, the PHOC has had a low recruitment rate of children with brain tumors—approximately 5 of 300 new pediatric cancer cases per year. In 2010, the Moroccan Society of Hematology and Oncology created the Pediatric Brain Tumor (PBT) working group with the help of St. Jude Children’s Research Hospital and The Hospital for Sick Children (i.e., SickKids), which raised the recruitment rate by 5 to 6 fold. In 2015, the PBT working group treated 28 new cases. This rise resulted from a close collaboration with the Department of Neurosurgery at Rabat’s University Hospital and North American experts in the PBT working group.

With the increase in pediatric patients with brain tumors at the PHOC, it became apparent to PHOC health care providers that the causes of delayed diagnosis of brain tumors were different from those of other pediatric tumors. Therefore, we quantified the timing of diagnoses of brain tumors at the PHOC and identified the potential causes of delayed diagnoses and necessary interventions to improve symptom awareness and reduce delayed brain tumor diagnoses in children.

Methods

We conducted a survey and interview of the parents of patients with brain tumors who were treated at the PHOC on an inpatient or outpatient basis from January 1, 2016 to June 30, 2016. Some patient tumors were diagnosed in 2015. One or both parents participated in the survey and interview at the PHOC, which focused on the diagnosis experience for each eligible child, in addition to family socioeconomic status, parent education level, and demographic data. We included patients who were less than 15 years old and received care for primary brain tumors that were diagnosed on the basis of imaging and/or histologic confirmation. Participation occurred at presentation, during treatment, or at the end of therapy.

The survey consisted of a 4-part form. The first section queried general information about the patient, family, and medical history. The second section inquired about patient experiences from the onset of the first signs/symptoms until brain tumor diagnosis. This included the number of consultations and the specialty of the health care providers who provided the consultations. The third section established the imaging modalities used in diagnosis. The fourth section ascertained patient treatment and care received at the PHOC.

Parent interviews were used to collect information describing brain tumor signs/symptoms, symptom duration, number of clinic visits, specialty of each clinician, and the potential causes of delayed diagnoses. The questions were asked in a simple Arabic dialect to ensure understanding of the questions by the parents. The interviewer then translated the responses into French, which is the official academic language of Morocco. Interviewers were fluent in dialectal Arabic and French and possessed a medical background. Before beginning the interview, the study and objectives were explained to the children, when age appropriate, and to the parents. Parents were asked to read a written consent form and provide oral approval. The time required for the interview and survey ranged from a minimum of 30 minutes to a maximum of 45 minutes.

Parent educational levels were defined as follows: illiterate and primary education only were scored as “low”, middle school and high school education levels were scored as “medium”, and college or university education levels were scored as “high”. Socioeconomic status was determined from investigations performed by the Department of Social Work at the PHOC. To assess the accessibility of patients to care, we recorded the duration of travel from patients’ homes to their nearest health care facilities and the duration of travel from patients’ homes to their nearest university hospitals.

Time of diagnosis was recorded as the date of first imaging by computed tomography or magnetic resonance imaging. The prediagnosis symptom interval (PSI) was defined as the interval between the onset of the first signs/symptoms and the date of diagnosis. The time between the onset of signs/symptoms and the first visit to a health care provider was termed parental duration (PD). The difference between the PSI and PD was termed doctor duration (DD). We used pathologic and radiologic reports and medical records to verify the dates of admission and confirm diagnoses.

Exclusion criteria included the onset of first symptoms outside of Morocco (n = 1, non-Moroccan patient referred for treatment), presence of more than 1 primary tumor (n = 1, simultaneous diagnosis of an abdominal mass and brain tumor), parent refusal to participate (n = 0), and brain tumor as a second cancer (n = 1, a patient in complete remission for mediastinal lymphoma referred for pineoblastoma). The data were compiled and analyzed with Microsoft Excel 2013. Quantitative measures were expressed as medians, and qualitative measures were expressed as percentages.

Results

Patient Characteristics

During the study period, 27 patients (14 girls and 13 boys) and their parents participated in the study. The median patient age at diagnosis was 7 years (range, 1–15 years). The patients were predominantly of urban origin (n = 17) and of low socioeconomic status (n = 15). The education levels of parents were predominantly low (n = 20). Eleven patients had more than 4 siblings.Two children were suspected to have type I neurofibromatosis. Most patients lived close to a health care facility. Fifteen patients lived within 15 minutes of a provincial hospital, and 13 patients lived within 1 hour of a university hospital (Table 1).

Table 1.

Patient Characteristics.

Characteristic	Number of patients
Sex
Female	15
Male	12
Age
≤ 1 year	2
1–5 years	10
> 5 years	15
Travel time to provincial hospital
< 15 minutes	15
30 minutes	7
45 minutes	2
60 minutes	2
120 minutes	1
Travel time to university hospital
< 60 minutes	13
60–180 minutes	11
>180 minutes	3
Origin
Urban	17
Rural	10
Socioeconomic status
Low	15
Medium	12
High	0
Parent education level
Low	14
Medium	7
High	6
Siblings
Only child	5
1 siblings	4
2 siblings	7
3 siblings	9
≥ 4 siblings	2

Characterization Of Signs And Symptoms

Each patient presented most often with several symptoms that either appeared successively or simultaneously and that worsened before imaging-confirmed diagnoses were made. The most common symptoms were vomiting (n = 18), headache (n = 17), oculo-visual disturbances (n = 11), motor deficits (n = 6), and endocrine dysfunction (n = 3) (Table 2). These signs/symptoms worsened in association with the appearance of intracranial hypertension and/or a focal sign in 7 patients and remained monosymptomatic in 1 patient (Table 2).

Table 2.

Patient Symptoms/Signs Reported During Consultations.

Symptoms/signs	Number of patients	Number of times reporteda
Neurologic
Motor deficit	6	8
Gait difficulty	5	6
Writing difficulty	2	4
Ataxia	3	7
Vertigo	4	4
Seizure	2	2
Loss of consciousness	1	1
Ophthalmic
Reduced visual acuity	1	4
Strabismus	8	8
Diplopia	1	1
Photophobia	1	1
General
Headache	17	29
Vomiting	18	39
Fever	4	8
Abdominal pain	1	1
Bone pain	2	2
Irritability	1	1
Endocrine
Weight gain	1	1
Early puberty	1	1
Acute dehydration	1	1

^aNumber of symptoms was determined from the number of times each symptom was reported during the 76 consultations.

Tumors Characteristics

Tumor locations were supratentorial in 13 patients and infratentorial in 14 patients (13 with posterior fossa and 1 with brain stem). The supratentorial tumors were located in the cerebral hemispheres and lateral ventricle (n = 3), suprasellar cistern (n = 4), optic pathway (n = 4), pineal gland (n = 1), and midbrain (n = 1). In total, 22 patients received biopsies, 18 occurring before admission to the PHOC and 4 after admission when surgery was possible. In 5 patients (4 with optic pathway and 1 with brain stem tumors), diagnoses were based on imaging only.

The tumors were classified according to the 2007 World Health Organization Classification of Tumours of the Central Nervous System.¹¹ The most frequently observed tumors were low-grade glioma (LGG) (n = 14), followed by medulloblastoma (n = 9), ependymoma (n = 2), pineoblastoma (n = 1), and atypical teratoidrhabdoid tumor (n = 1) (Table 3).

TABLE 3.

Consultation and diagnosis delays according to presenting signs/symptoms, diagnoses, and tumor site

Patient	NOC	PD (days)	PSI (months)	DD (days)	Time between diagnosis and referral to PHOC (weeks)	Presenting signs/symptoms	Diagnosis	Tumor site
1	4	7	0.3	3	0.3	Ophthalmologic	LGG/Neurofibromatosis	Optic pathway
2	4	2	2	58	1	Neurologic	LGG	Optic pathway
3	3	4	12	357	28	General	LGG	Posterior fossa
4	3	7	3	83	8	General	LGG	Right ventricle
5	2	15	1	15	40	Ophthalmologic	LGG	Midbrain
6	5	7	1	23	4	Neurologic	Medulloblastoma	Posterior fossa
7	2	7	0.3	4	10	General	Medulloblastoma	Posterior fossa
8	1	180	6	0	8	Neurologic	Ependymoma	Cerebral hemisphere
9	5	7	1	23	12	General	LGG	Posterior fossa
10	1	60	2	0	12	Neurologic	medulloblastoma	Posterior fossa
11	2	300	10	0	2	Endocrine	LGG/Neurofibromatosis	Optical pathway
12	2	5	1	25	8	General	Medulloblastoma	Posterior fossa
13	2	7	0.5	7	44	Neurologic	LGG	Suprasellar
14	3	7	3	83	24	General	LGG	Suprasellar
15	2	7	4	113	12	General	LGG	Suprasellar
16	2	7	2	53	20	General	Pineoblastoma	Pineal gland
17	1	7	0.5	7	0.7	General	Medulloblastoma	Posterior fossa
18	4	10	13	380	76	Ophthalmologic	LGG	Optic pathway
19	2	7	1	23	28	Ophthalmologic/Neurologic	Medulloblastoma	Posterior fossa
20	4	7	0.6	13	0.7	Neurologic	LGG	Brainstem
21	1	2	2	0	8	General	Ependymoma	Posterior fossa
22	5	7	4	112	4	General	Medulloblastoma	Posterior fossa
23	4	7	9	263	3	General	Medulloblastoma	Posterior fossa
24	2	4	8	236	8	General	Medulloblastoma	Posterior fossa
25	3	7	20	593	16	Ophthalmologic	LGG	Suprasellar
26	2	1	1	29	32	Neurologic	LGG	Posterior fossa
27	3	3	5	147	6	General	ATRT	Cerebral hemisphere

ATRT, atypical teratoid rhabdoid tumor; DD, doctor delay; LGG, low-grade glioma; NOC, number of consultations; PD, parent delay; PHOC, Pediatric Hematology and Oncology Center; PSI, prediagnosis symptom interval.

Patient Diagnosis Experiences

Parents sought medical consultations within 7 days after the onset of the first symptom in 15 cases. The median consultation time was 7 days (range, 1–300 days). Most diagnoses were made after several consultations with different physicians. In total, 27 patients received consultations 76 times before diagnosis confirmations were made, with a mean of 2.74 consultations per patient (range, 1–5 consultations per patient). Approximately half of all patients received at least 3 consultations (13 patients) before diagnosis. Four patients were provided a diagnosis at the first consultation. Nine patients received a diagnosis at the second consultation. Five patients received a diagnosis at the third consultation. Six patients received a diagnosis at the fourth consultation, and 3 patients received a diagnosis at the fifth consultation.

General practitioners provided most consultations (n = 33), followed by pediatricians (n = 27), ophthalmologists (n = 12), neurologists (n = 2), and neurosurgeons (n = 2). The median PSI was 2 months (range, 0.25–20 months), with a median DD of 25 days (range, 0–593 days) (Table 3).

In this study, the PSI varied according to 3 main factors; increased age, tumor type and tumor location(Figure 1), The shortest PSIs (< 1 month) were observed in patients younger than 5 years, with infratentorial tumors and medulloblastomas (median, 1 month). The longest PSIs were observed in children over 5 years and in those with supratentorial tumors, especially LGG (median, 3 months; range, 0.3–20). Single-child families had reduced PSIs (median, 1 month; range, 0.3–2), whereas sex, socioeconomic status, and urban or rural origin did not affect PSI length (Table 4).

Fig. 1.

Prediagnosis symptom interval

Table 4.

Diagnosis Delay According to Patient and Tumor Characteristics.

Characteristic	Number of patients	Median PSI (months)	Range (months)
Sex
Female	15	2	0.25–20
Male	12	2	0.30–13
Age
≤ 1 year	2	1 and 2	–
1–5 years	10	1	0.30–10
> 5 years	15	4	0.25–20
Origin
Urban	13	3	0.25–20
Rural	14	1	0.50–10
Socioeconomic level
Low	15	2	0.50–13
Medium	12	2	0.25–20
Tumor site
Supratentorial	13	3	0.30–20
Infratentorial	14	1	0.25–12
Histologic tumor type
LGG	14	3	0.30–20
Medulloblastoma	9	1	0.25–9
Ependymoma	2	2 and 6	–
Other	2	2 and 5	–

Abbreviation: LGG, low-grade glioma.

60% of the patients with PSI longer than 4 months had general symptoms (figure 2c).Headache, vomiting, and strabismus were the primary initial symptoms for patients with PSIs longer than 4 months. At the time of diagnosis, these symptoms had considerably worsened, and all but 1 child were older than 5 years (Table 5).

Fig. 2.

General symptoms

Table 5.

Signs/Symptoms Initially Present and at the Time of Diagnosis for Patients with PSIs Longer than 4 Months.

Patient	Age	PSI	Initial signs/symptoms	Presentation at diagnosis
3	6	12	Headache, vomiting, photophobia	Worsening initial signs/symptoms, papilledema (ICHT)
8	3	6	Parents didn’t consult	Headaches and motor deficit
11	5	10	Parents didn’t consult	Precocious puberty (pubic hair and elevated scrotum volume)
15	6	4	Headache, vomiting, photophobia	ICHT, dehydration
18	7	13	Strabismus	ICHT
22	10	4	Vomiting, fever, abdominal pain	Pyrexial meningeal syndrome
23	15	9	Headache	ICHT, gait trouble, irritability
24	10	8	Strabismus	Strabismus, vertigo, ataxia
25	9	20	Strabismus	Strabismus
27	15	5	Headache, vomiting	Worsening of initial signs/symptoms

Abbreviations: ICHT, intracranial hypertension; PSI, prediagnosis symptom interval.

Upon admission to the PHOC, 5 patients did not have any clinical findings during their examinations, although the remaining patients exhibited varying neurologic, ophthalmologic, or general signs/symptoms of brain tumors (Table 6). All patients were referred by neurosurgeons for further treatment. Four patients did not receive surgical resections or ventricular shunt before referral to the PHOC (Table 6). Of the remaining 23 patients, 16 received ventricular shunt and surgical resection, 5 received only ventricular shunt, and 2 received surgery alone. The median time between the imaging date confirming diagnosis and ventricular shunt was 1 week (range, 0–20 weeks). The median time between the imaging date and surgical resection was 1 month (range, 0.5–12 months). In 9 patients, surgical resection did not occur until longer than 2 months after the imaging date. The median time between the imaging date and admission to the PHOC was 3 months (range, 0.07–19 months) (Table 6).

Table 6.

Time Between Diagnosis and Surgery or Ventricular Shunt and Clinical Findings at Initial Examination upon Admission to the PHOC.

Patient	Time from imaging to PHOC admission	Time from imaging to surgery	Time from imaging to VS	Clinical findings at examination
1	2 days	NP	NP	Bilateral mydriasis, reduced photomotor reflexes, reduced visual acuity
2	1 week	NP	NP	Hemiparesia
3	7 months	NP	1 months	Inversion of patellar reflexes
4	2 months	NP	2 months	Absence of ankle jerk reflexes, speech disorder
5	10 months	9 months	1 week	Hemiparesia, ataxia, strabismus
6	1 months	3 weeks	1 week	No findings
7	2.5 months	2 months	1 months	No findings
8	2 months	2 months	NP	Motor deficit, gait disturbance, strabismus
9	3 months	4 months	2 days	Cerebellar syndrome, strabismus
10	3 months	2.5 months	1.5 months	Gait disturbance, motor deficit
11	2 weeks	NP	NP	Early puberty
12	2 months	1 month	1 week	Gait disturbance
13	11 months	12 months	3 months	No findings
14	6 months	NP	2 days	Paresthesia
15	3 months	2 months	2 weeks	Hemiparesia, facial asymmetry
16	5 months	NP	0 days	Hemiplegia, sphincter weakness
17	5 days	NP	14 days	Ataxia, gait disturbance
18	19 months	20 days	1 week	Blindness
19	7 months	5 months	5 months	Strabismus, nystagmus
20	5 days	NP	NP	Extrapyramidal syndrome
21	2 months	27 days	13 days	Positive Romberg sign
22	1 months	16 days	2 days	Facial palsy, strabismus, cerebellar syndrome, positive Babinski sign
23	3 weeks	18 days	0 day	Nystagmus, cerebellar syndrome
24	2 months	2 weeks	4 days	Facial hemiparesia, strabismus, nystagmus, balance problems
25	4 months	2 weeks	NP	Hemiplegia, third nerve palsy, amyotrophia of the right hand
26	8 months	7 months	1 month	No findings
27	1.5 months	1.5 months	NP	No findings

Abbreviations : NP, not performed; PHOC, Pediatric Hematology Oncology Center of Rabat Children’s Hospital; VS, ventricular shunt.

Discussion

Delays in diagnoses pose particular challenges for treating brain tumors in children.^6,12,13 Although an increasing prevalence of computed tomography and magnetic resonance imaging scanners has reduced DD length in Morocco, the diagnosis of brain tumors often occurs after the onset of intracranial hypertension. Furthermore, compared with the PSIs of other childhood malignant neoplasms, considerable delays in diagnosis of pediatric brain tumors have not improved for several decades.^4,12,14,15

Here, we found the median time to diagnosis was 2 months, ranging from 1 week to 20 months, which is consistent with other studies. In a cohort study of 200 patients with brain tumors diagnosed between 1988 and 2001 in the United Kingdom, the median PSI was 2.5 months,⁹ whereas the median PSI of a cohort of 252 children with brain tumors diagnosed between 1980 and 1999 in Switzerland was 60 days.¹² Overall, we found that the longest PSIs (> 3 months) occurred in patients over 5 years old or in those with LGGs and supratentorial tumors. The shortest PSIs (< 1 month) occurred in patients younger than 5 years old and in those with medulloblastomas and infratentorial tumors. With exception of single-child families, PSI length was not changed by any demographic factors, including sex, socioeconomic status, parental education level, and urban or rural origin, which was also observed in other studies.^5,12,16

Despite advances in medical imaging and improved access to care in developing countries, long PSIs for children with brain tumors are still prevalent. Many studies have attempted to identify why cerebral tumors remain undetected for longer times.^15,17,20 The causes appear multifactorial and linked to interdependent variables (eg, age, symptoms, parents, physicians, histologic tumor type and grade, tumor localization, and health care systems).^{4-6,15,16,18,21,22} Previous studies that defined PSI length according to histologic tumor type, localization, and grade reported similar finding to ours. Specifically, long median PSIs are associated with LGG, supratentorial localization, and older ages of children and are even longer in children with convulsions (> 10 months).¹⁷ Tatencloux et al. reported a median diagnosis time of 4 months for low-grade brain tumors and 1 month for high-grade cerebral tumors, although this study contained a limited number of patients.²³ Advanced medulloblastomas are more associated with shorter PSIs (< 7.4 weeks) than are localized medulloblastomas (22.5 weeks).²⁴ The PSIs of metastatic brain cancer are also considerably shorter,¹⁹ suggesting that tumor biology is an important determinant for delayed diagnoses and survival probability.

Similar with that of other previous studies, we found marked variability in the clinical presentation of brain tumors, consisting primarily of general, nonspecific signs/symptoms. A clear predominance of vomiting and headache, followed by other neurologic signs/symptoms that generally occur successively, have been reported by several studies.^{12,15,16,18,19,25} The heterogeneity and predominance of nonspecific symptoms (eg, headache and vomiting) most likely play an essential role in the delayed diagnoses of pediatric brain tumors because they are often attributed to other self-limiting illnesses.^9-15 Although behavioral disorders and school problems are less frequently observed, these signs/symptoms alone can also predominate the clinical picture for a long time, delaying diagnoses.^9,15,21,25 Convulsions and endocrine disturbances are also known to be associated with diagnosis delays.¹⁹

Because parents and physicians are generally more attentive to nonspecific signs/symptoms that occur in toddlers and infants, reduced PSI lengths (< 1 month) are more likely to occur in younger children. In addition, the behavior of brain tumors in young children is often aggressive, with multiple symptoms that develop in a relatively short period, leading to noticeable developmental delays.¹² Furthermore, brain tumors are frequently localized to the posterior cranial fossa in infants, permitting rapid diagnoses.^12-18 In older children, long PSIs may be associated with an adaptation of the cerebral parenchyma to the development of the tumor mass.^4,18,24 In addition, pediatric cerebral imaging often requires anesthesia, which is not available in many centers in developing countries.^9-18

Although most families in our study had low socioeconomic status and low educational levels, we found that most sought consultations within 1 week after the onset of the first symptom, contrary to other studies reporting that a combination of disadvantaged social backgrounds and low parent education levels contributes to longer PSIs.^4,5,21 We also found that the PD was relatively shorter than those reported by other studies (7 versus 14 days).¹² We observed that in 24 cases, parents took their children to physicians within 1 month after the first sign/symptom appeared. However, these families sought multiple consultations at public and private institutions, without follow-up by the same doctor. This has also been reported by others.^6,15-19 In such cases, the primary reason for delayed diagnosis are incorrect diagnoses, such as tonsillitis, gastroenteritis, sinusitis, meningitis, or refractive eye disorders. Headache is particularly confounded with migraine or tension headaches, resulting in incorrect and subsequently delayed diagnoses. Therefore, headache is specifically recommended for careful assessment, especially when associated with seizures or ocular disorders.^25-27 The specialty of the first consulting physician, accessibility to care, type of social protection, and organization of the health care system are also associated with long PSIs.⁴

In several recent studies, it has become evident that parents often detect certain disorders better than physicians do.^15,19,22 This is clearly shown by the high numbers of prediagnosis consultations and the variety of misdiagnoses given to the parents of children with brain tumors. This can lead to conflicts between parents and health care providers, further exacerbating delayed diagnoses. Poor communication between physicians and parents may also be associated with delayed diagnoses, but this requires further assessment. In Morocco, as in most developing countries, access to care is a major obstacle to early care, even for patients living near health care facilities. All patients in our study were referred to the PHOC approximately 2 months after receiving neurosurgery. According to the parents of these patients, a slow administrative process and frequently delayed appointments were a contributing factor to delayed diagnoses.

The association between delayed diagnosis of childhood brain tumors and survival has been studied only in monocentric studies, without consideration of confounding factors.⁵ The long-term survival rate without recurrence for all histologic tumor types remains low when diagnoses are delayed.¹⁰ Children with short PSIs are more likely to have smaller brain tumors that are easier to resect, which in a variety of childhood brain tumors improves prognosis.^28-30 However, the association between delayed diagnosis and mortality is less clear, as PSIs are also relatively short for aggressive tumors.^5,24 For most childhood brain tumors, long PSIs are not associated with an unfavorable outcome and may paradoxically be favorable in some cases.^5,24,31 In a study of children with medulloblastoma in France, long PSIs were associated with increased survival, but this association was not present after adjusting for confounding factors (ie, age, presence of metastases, histologic tumor type, complete character of tumoral excision, and tumor volume).²¹ Long-term survival of gliomas, which constitute the majority of childhood brain tumors (40%–50%), is associated with patient age, tumor subtype, and the extent of resection.³² As the PSIs of some aggressive brain tumors may not influence survival because of their rapid progression, the effect of delayed diagnoses on survival should be analyzed separately for distinct tumor types. The association between delayed diagnoses and sequelae is also unclear because data are very limited and confounding factors are numerous (eg, neurosurgery and radiotherapy).

Early diagnosis remains a high priority because it minimizes progressive neurologic morbidity, prevents acute symptoms and sequelae, allows extensive tumor resection, reduces parent psychological distress, and reduces health care costs. Therefore, we propose the following potential solutions to reduce PSI lengths in developing countries such as Morocco: (1) adapt clinical guides for diagnoses that optimize the approach physicians use to reach diagnoses, (2) introduce flexible health care systems, and (3) improve health care infrastructures. Such initiatives have successfully reduced the PSIs of pediatric brain tumors in other countries. This is particularly evident in the United Kingdom, which launched a project titled HeadSmart that provided guidelines for assessment and radiologic screening of brain tumors in children.^7,33 This approach is highly likely to be successful in other countries with similar centralized health care systems. However, in Morocco and other developing countries without centralized health care systems, improved training in medical and nursing schools and in residency/fellowship training programs is also needed. As we and others have shown, delayed diagnoses of pediatric brain tumors in Morocco and other developing countries most often results from misdiagnoses and a failure to provide necessary assessments by health care providers. A particular emphasis should be placed on pediatric symptoms, which are nonspecific and often mimic other benign illnesses, rather than on the symptoms observed in adults.

In conclusion, we found that delayed diagnoses of brain tumors are prevalent for Moroccan children. The longest PSIs occurred in older children and in those with LGG and supratentorial tumors. Our survey indicates that a lack of symptom awareness, a complex health care system, an absence of patient referral and follow-up, and inadequate access to specialized radiologic examinations are contributing factors to these delayed diagnoses. To attenuate PSIs, physicians should consider brain tumors among differential diagnoses in children presenting with nonspecific neurologic signs/symptoms. Because our study was limited by a small sample size and a reliance on anecdotal accounts of the prediagnosis clinical experiences of patients, further large cohort studies are needed to confirm our findings.

WHAT’S NEW:

Delayed diagnosis of pediatric brain tumors is exacerbated by older age, supratentorial tumor location, and low-grade glioma. A lack of association between general symptoms, such as nausea and headaches, with brain tumors is a major contributing factor.