Sage Journals: Discover world-class research

Abstract

Background/Objectives:

Delays to diagnosis of multiple sclerosis (MS) are largely unknown in sub-Saharan Africa. This study utilizes a quantitative approach to determine factors independently associated with delays to diagnosis among a cohort of Zambian people with MS.

Methods:

This cross-sectional study enrolled people with a confirmed diagnosis of MS at a public outpatient neurology clinic in Lusaka, Zambia. Standardized surveys were administered. Diagnostic delay was categorized into patient delay, health-system delay, and all delay. A multivariable linear regression model was used to determine factors independently associated with time to diagnosis.

Results:

A total of 22 participants had a confirmed diagnosis of MS. Median all delay from symptom onset to MS diagnosis was 11.4 months (interquartile range (IQR) = 4–35.5), with health-system delay accounting for the majority. Asian race correlated with a decrease in all delay, while evaluation outside of Zambia was associated with a decrease in patient delay and birthplace outside of Zambia correlated with decreased health-system delay.

Conclusion:

Black African people born in Zambia and evaluated in Zambia experienced prolonged diagnostic delays when compared to non-black Zambians born and/or evaluated outside Zambia, likely due to a combination of health system and patient factors, which necessitate further study to shorten time to diagnosis.

Keywords

Multiple sclerosis epidemiology demyelination MRI quality of life relapsing/remitting

Introduction

Multiple sclerosis (MS) is one of the most common chronic neurological disorders in the world with the highest prevalence reported in North America, northern Europe, and southern Australia, but largely unknown in sub-Saharan Africa (SSA).^1,2 Because SSA has the lowest number of neurologists per capita in the world and limited availability of brain imaging,^3,4 a major ascertainment bias due to underdiagnosis is likely. A recent ecological study found that MS prevalence was lower in countries with lower healthcare expenditure per capita, reduced access to diagnostic technology, and lower number of neurologists per capita,⁵ suggesting prior reports of low MS prevalence in SSA may be based on absence of data rather than true absence of disease. Recent epidemiological studies from Sudan, Kenya, and South Africa all suggest the prevalence, or ascertainment, of MS is rising among Black African people.^6–8

Regional prevalence is associated with diagnostic delay of MS, which occurs due to a multitude of patient and health-system factors and can have important implications for prognosis, especially as delays in treatment initiation increase disability and even mortality.^9–12 The degree of diagnostic delay varies worldwide and, with increased awareness and modifications to diagnostic criteria, has improved in many regions over time.^13,14

Zambia is a country in southern Africa with a population of 20 million people and only 12 neurologists, less than 10 magnetic resonance imaging (MRI) units and often cost-prohibitive access to diagnostic testing. Since the formation of the country’s first neurology training program and expansion of specialized neurological care at the University Teaching Hospital (UTH) in the capital of Lusaka in 2018, approximately 30 patients have been diagnosed with MS. A qualitative assessment of 13 patients diagnosed with MS in Zambia found that most had long delays to diagnosis, visited multiple physicians, received misdiagnoses, underwent unnecessary tests, and had the socioeconomic means to pursue second opinions.^15,16 Based upon these findings, we hypothesized that higher education level and socioeconomic status would be associated with decreased delay. While this qualitative data explores the patient perspective of causes of delay, the specific factors that increase delay to diagnosis of MS in this population remain unknown.

This study aimed to utilize a quantitative approach to determine factors independently associated with patient delay, health-system delay, and all delay to diagnosis among a cohort of people with MS (PwMS) in Zambia.

Methods

This cross-sectional cohort study was conducted in PwMS who presented for clinical care at an outpatient neurology clinic at UTH, the national referral hospital in Lusaka, Zambia, from October 2019 to February 2022. Patients could present to the neurology clinic upon self-referral or referral from another provider. Adult patients aged 18 years or older with a confirmed diagnosis of MS based on 2017 McDonald’s criteria¹³ or clinically isolated syndrome (CIS) after evaluation by a neurologist at UTH were included in the study. Due to limitations in performing oligoclonal band testing at local laboratories, this criterion to fulfill dissemination in time was not utilized to aid with diagnosis; however, all patients underwent MRI. Patients with a suspected but unconfirmed diagnosis of MS due to incomplete evaluations were excluded (Figure 1). All participants in the study provided written consent.

Figure 1.

Flowchart of patient case identification and selection based on inclusion criteria.

Data were collected through standardized survey questionnaires developed by an MS specialist and administered by trained nurses and neurologists. Figure 2 summarizes the questionnaires including demographic variables (age, sex, self-reported race, marital status, birth province, highest education), clinical characteristics (date and symptoms of first attack), and diagnostic journey (date of first presentation to healthcare, number and type of prior doctors and facilities, and location of MS diagnosis). All data were de-identified and stored in a secure institutional REDCap database at Johns Hopkins University.^17,18

Figure 2.

Summary of variables collected by standardized questionnaires (for additional details, please see Supplemental section including full questionnaires).

Demographic and clinical data are reported using descriptive statistics, including mean and standard deviation for normally distributed continuous variables, median and interquartile ranges for non-parametric continuous variables, and proportions for categorical variables.

Both mean and median delays are reported. Mean delay was calculated to allow direct comparison with other published studies of diagnostic delays in MS. However, because the diagnostic delay data did not follow a normal distribution and to mitigate the influence of significant outliers in the data set, median delay was preferentially utilized in analyses of delay sub-types. Delay was categorized into patient delay, health-system delay, and all delay. Patient delay was defined as the time between the date of symptom onset and the date of first presentation to healthcare. Health-system delay was defined as the time between date of first presentation to healthcare and date of diagnosis of MS. All delay, the sum of patient and health-system delay, was defined as the time between the date of symptom onset and the date of diagnosis of MS.

Association of factors with diagnostic delays was first analyzed with Wilcoxon Rank Sum or Kruskal–Wallis tests for non-parametric predictor variables, and Spearman’s correlation coefficients were used for parametric predictor variables.

Due to non-parametric distributions and outlier values, delay times were log transformed before input into regression models. Delays of zero months were replaced with values of 1 day (1/365 = 0.002739726) to avoid missing values. Variable selection for the regression models was based upon the literature of diagnostic delay studies in other MS populations and the assumed relationships of factors influencing the diagnostic pathway in Zambia (Figure 3).^19–23 Univariable linear regression models were used to identify variables for inclusion in a multivariable model. A priori, we planned to use a p-value threshold of <0.2 for inclusion of variables into all multivariable models. However, given the large number of variables that met this threshold for all delay and health-system delay models, in the context of a small sample size, we elected to use p < 0.05 for those models. We retained p < 0.2 as the threshold for the patient delay model. The inverse log of the beta coefficients of the linear regression models was calculated and is reported as they correspond to the percent change in delay for each variable. A two-sided p-value of less than 0.05 in the multivariable regression models was considered statistically significant. Fit statistics and variance inflation factors are reported in Supplementary Materials. Data analysis was conducted using Stata 17 SE (College Station, TX, USA).

Figure 3.

Directed acyclic graph of the relationships between selected variables influencing the MS diagnostic pathway in Zambia.

This study was approved by the ERES Converge Institutional Review Board, the Zambia National Health Research Authority, and the Johns Hopkins University Institutional Review Board.

Results

A total of 22 participants with a diagnosis of MS or CIS were included in this study. All patients underwent MRI of the brain. Most were female (n = 14, 64%), Black African (n = 20, 91%), and had at least a college or university level of education (n = 13, 59%) (Table 1). The mean age of symptom onset was 30.1 years (standard deviation (SD) = 9.5 years), while the mean age of diagnosis was 33.7 years (SD = 10.7 years), leading to a mean diagnostic delay of 3.7 years (SD = 8.6 years). Most participants had a diagnosis of relapsing-remitting MS (n = 19, 86.4%), with the remainder diagnosed with secondary progressive MS (n = 2, 9.1%) and CIS (n = 1, 4.5%).

Table 1.

Demographic and clinical characteristics of cohort (n = 22).

Demographics
Age at first symptom onset (years)
Mean ± SD	30.1 ± 9.5
Age at diagnosis (years)
Mean ± SD	33.7 ± 10.7
Sex (%)
Male	8 (36%)
Female	14 (64%)
Race/ethnicity (%)
Black African	20 (91%)
Asian	2 (9%)
Birth province (%)
Lusaka or Copperbelt	11 (50%)
Other Zambian province^a	7 (32%)
Outside Zambia	4 (18%)
Level of education (%)
Secondary school or less	9 (41%)
College/university or more	13 (59%)
Marital status (%)
Single or divorced	8 (36%)
Married	14 (64%)
Clinical characteristics
Initial MS symptoms included (%)
Vision problem	9 (41%)
Gait instability or imbalance	9 (41%)
Limb weakness	10 (45%)
Sensory changes	6 (27%)
Number of symptoms in first attack
Mean ± SD	2.1 ± 1.4
Number of symptoms within first year of symptom onset
Mean ± SD	1.6 ± 0.7
Type of multiple sclerosis (%)
Relapsing remitting	19 (86%)
Secondary progressive	2 (9%)
Clinically isolated syndrome	1 (5%)

Other Zambian provinces included Eastern, Western, Northern, Central, and Southern.

The overall median all delay from symptom onset to MS diagnosis was 11.4 months (interquartile range (IQR) = 4–35.5). The median patient delay from symptom onset to first presentation to care was less than 1 month (IQR = 0–3). The median health-system delay from first presentation to MS diagnosis was 8.8 months (IQR = 2.6–25.4). Of note, all participants reported seeking evaluation at a healthcare facility for their initial symptoms.

A significant unadjusted association was found between birth province and health-system delay (p = 0.03), with a shorter health-system delay for those who were born outside Zambia (Table 2). The presence of sensory changes during the initial attack was associated with a decrease in all delay (p = 0.01) and health-system delay (p = 0.006). When evaluating the diagnostic journey, visit to a specialist surgeon significantly increased both all delay (p = 0.01) and health-system delay (p = 0.003). Symptom onset before 2019, prior to the establishment of specialty neurology care in Zambia, significantly increased both all delay (p = 0.01) and health-system delay (p = 0.003). Among continuous factors, the age at time of diagnosis was positively and significantly correlated with all delay (p < 0.001) and health-system delay (p < 0.001) (Table 3).

Table 2.

Association of categorical characteristics and diagnostic delay.

Characteristic	n	All delay (months)		Patient delay (months)		Health-system delay (months)
Characteristic	n	Median (IQR)	P-Npar	Median (IQR)	P-Npar	Median (IQR)	P-Npar
Demographics
Sex
Male	8	9.8 (2.5, 24.9)	0.47	2 (0, 6.6)	0.08	2.6 (0.5, 23.4)	0.23
Female	14	11.4 (4.0, 40.6)		0 (0, 0)		11.4 (3.3, 35.5)
Race/ethnicity
Black African	20	14.3 (4.0, 38.1)	0.06	0 (0, 3)	0.79	11.4 (2.8, 30.4)	0.22
Asian	2	2 (0, 4)		0 (0, 0)		2 (0, 4)
Birth province
Lusaka or Copperbelt	11	24.3 (4–40.6)	0.08	0 (0–3)	0.61	21.3 (4–35.5)	0.03
Other Zambian province^a	7	15.5 (4–48.7)		0 (0–3)		13.2 (3–48.7)
Outside Zambia	4	2.5 (0.5–4.7)		1.5 (0.5–2.5)		0.5 (0–2.2)
Level of education
Secondary school or less	9	5.3 (4–13.2)	0.23	0 (0–1)	0.49	4 (1–13.2)	0.28
College/university or more	13	24.3 (4.0–40.6)		0 (0–3)		21.3 (3–25.4)
Marital Status
Single or divorced	8	11.4 (3.3–33)	0.99	0 (0–2)	1	11.4 (1.8–23.3)	0.85
Married	14	11.8 (4–35.5)		0 (0–3)		6 (3–35.5)
Clinical characteristics
Initial MS symptoms
Vision problem	9	8.1 (4.0–24.3)	0.4	0 (0–0)	0.17	8.1 (2.6–24.3)	0.54
No vision problem	13	15.5 (4–40.6)		0 (0–3.1)		13.2 (3.0–35.5)
Gait instability/imbalance	9	15.5 (4.0–40.6)	0.59	1 (0–3.1)	0.12	8.1 (1.3–21.3)	0.86
No gait issue	13	9.6 (4–25.4)		0 (0–0)		9.6 (3.1–25.4)
Weakness	10	6.0 (4–24.4)	0.35	0 (0–3)	0.94	6 (1–21.3)	0.41
No weakness	12	19.9 (4.7–44.7)		0 (0–6)		15.4 (2.9–37.0)
Sensory changes	6	3.5 (2.6–4.0)	0.01	0.5 (0–3)	0.75	1.8 (1.0–3.0)	0.006
No sensory changes	16	24.4 (6.7–44.7)		0 (0–2.6)		21.3 (4–42.1)
Other symptoms^b	8	20.4 (2.5–66.4)	0.68	2 (0–12.2)	0.05	11.3 (0.5–58.8)	0.58
No other symptoms	14	8.8 (4–24.4)		0 (0–0)		8.8 (3.3–24.3)
Diagnostic journey
Prior doctors visited
Internist/general practitioner	16	14.3 (4.7–30.4)	0.32	0 (0–2.5)	0.93	11.4 (3.2–30.4)	0.3
No internist/GP	6	4 (2.6–40.6)		0 (0–3)		3.3 (1.0–21.3)
Ophthalmologist	3	25.4 (24.3–483.4)	0.11	0 (0–0)	0.47	25.4 (24.3–483.4)	0.053
No ophthalmologist	19	8.1 (4–35.5)		0 (0–3)		4 (1.3–21.3)
Neurologist	7	35.5 (1–117.7)	0.64	0 (0–1)	0.53	35.5 (0–107.5)	0.42
No neurologist	15	9.6 (4.0–24.4)		0 (0–3)		8.1 (2.6–21.3)
Specialist surgeon^c	5	35.5 (25.4–92.2)	0.01	0 (0–0)	0.15	35.5 (25.4–92.2)	0.003
No surgeon	17	5.3 (4–15.5)		0 (0–3)		4 (1.3–13.2)
Prior facilities visited
Public or government only	9	5.3 (4–13.2)	0.34	0 (0–2.1)	0.7	4 (3–13.2)	0.2
Private only	9	15.5 (4–24.4)		0 (0–3.1)		9.6 (1.3–21.3)
Both public and private	4	76.6 (19.0–300.5)		0 (0–5.1)		72 (19–295)
Country of diagnosis
Zambia	18	15.4 (4–35.5)	0.12	0 (0–3)	0.23	13.2 (3–25.4)	0.26
Other^d	4	4 (1–8)		0 (0–0)		4 (0–8)
Year of first symptoms
Before 2019	13	24.4 (9.6–48.7)	0.01	0 (0–0)	0.3	21.3 (9.6–48.7)	0.003
2019 and afterwards	9	4 (3–5.3)		1 (0–3)		2.6 (1.0–3.3)
Year of first presentation
Before 2019	10	18.8 (8.1–92.2)	0.21	0 (0–0)	0.25	17.3 (8.1–92.2)	0.07
2019 and afterwards	12	4.7 (3.5–24.8)		0.5 (0–3)		3.2 (1.2–22.8)
Year of diagnosis
Before 2019	7	9.6 (4–48.7)	0.9	0 (0–3.1)	0.82	9.6 (4–48.7)	0.57
2019 and afterwards	15	13.2 (4–35.5)		0 (0–3)		4 (1.3–25.4)

Other Zambian provinces included Eastern, Western, Northern, Central, and Southern.

Other initial MS symptoms included bowel/bladder symptoms, facial weakness, fatigue, headache, muscle spasms, and tremor.

Other specialist surgeons visited included neurosurgeons, orthopedic surgeons, and otolaryngologists.

Other countries of diagnosis included South Africa, the United States, and India.

Italicized values represent those meeting p-value cutoff to be included in the multivariable analysis.

Bolded values indicate the values meeting a p-value threshold of <0.05.

Table 3.

Association of continuous characteristics and diagnostic delay.

	All delay		Patient delay		Health-system delay
	Corr coefficient	95% CI	Corr coefficient	95% CI	Corr coefficient	95% CI
Age at first symptom onset	-0.29	(-0.64, 0.15)	0.05	(-0.38, 0.46)	-0.3	(-0.64, 0.14)
Age at diagnosis	0.53	(0.14, 0.78)	0.02	(-0.41, 0.44)	0.53	(0.14, 0.78)
Number of attacks within first year of symptom onset	-0.25	(-0.61, 0.19)	-0.14	(-0.53, 0.3)	-0.24	(-0.6, 0.2)
Number of symptoms in first attack	-0.25	(-0.61, 0.19)	0.10	(-0.33, 0.5)	-0.25	(0.61, 0.19)

Bolded values indicate the correlation coefficients meeting a p-value threshold of < 0.05.

The results of the final multivariable regression model elicited several factors that were independently associated with a decrease in delay, as outlined in Table 4. Patient race, if Asian, correlated with a decrease in all delay (p = 0.01), when compared to Black African participants. Evaluation outside of Zambia was significantly associated with a decrease in patient delay (p = 0.02), and birthplace outside Zambia correlated with a decrease in health-system delay (p = 0.01). In this model, the number of symptoms during the first attack and Asian race also trended toward a significant association with health-system delay.

Table 4.

Multivariable linear regression analysis of log of diagnostic delay.

	10^β (95% CI)	% Change in delay (95% CI)	p
All delay ^a
Race/ethnicity
Black African	Ref.
Asian	0.00 (0, 0.12)	-100% (-100%, -88%)	0.01
Birth Province
Lusaka or Copperbelt	Ref.
Other province	9.55 (0.08, 1072)	855% (-92%, 107,100%)	0.33
Outside Zambia	0.02 (0.0001, 2.0)	-98% (-99.99%, 100%)	0.09
Age at diagnosis	1.02 (0.83, 1.26)	2% (-17%, 26%)	0.85
Number of symptoms in first attack	0.65 (0.14, 2.88)	-35% (-86%, 188%)	0.54
Prior doctors visited
No specialist surgeon	Ref.
Specialist surgeon	19.50 (0.20, 1950)	1850% (-80%, 194,900%)	0.19
Diagnosed outside of Zambia	0.32 (0.003, 3631)	-68% (-99.7%, 363,000%)	0.62
Patient delay ^a
Sex
Male	Ref.
Female	0.05 (0, 93.3)	-95% (-100%, 9230%)	0.4
Initial MS symptoms
No sensory symptoms	Ref.
Sensory symptoms present	0.58 (0, 173,780)	-42% (-100%, 17,377,900%)	0.92
No other symptoms	Ref.
Other symptoms present	93.33 (0.02, 537,032)	9233% (-98%, 53,703,100%)	0.28
Number of symptoms in first attack	2.88 (0.06, 144)	188% (-94%, 14,300%)	0.58
Prior doctors visited
No ophthalmologist	Ref.
Ophthalmologist	0.42 (0, 1,737,800)	-58% (-100%, 173,779,900%)	0.9
No specialist surgeon	Ref.
Specialist surgeon	0.0001 (0, 66.1)	-99.97% (-100%, 6510%)	0.16
Diagnosed outside of Zambia	0.0001 (0, 0.25)	-99.99% (-100%, -75%)	0.02
First presentation to care
Before 2019	Ref.
2019 or afterwards	3.24 (0.001, 8912)	224% (-99.9%, 891,100%)	0.75
Health-system delay ^a
Race/ethnicity
Black African	Ref.
Asian	0.001 (0, 1.26)	-99.9% (-100%, 26%)	0.056
Birth Province
Lusaka or Copperbelt	Ref.
Other province	1.35 (0.007, 257)	35% (-99.3%, 25,600%)	0.9
Outside Zambia	0.0003 (0, 0.10)	-99.97% (-100%, -90%)	0.01
Age at diagnosis	1.32 (0.79, 1.29)	32% (-21%, 29%)	0.92
Number of symptoms in first attack	0.15 (0.02, 1.10)	-85% (-98%, 10%)	0.06
Initial MS symptoms
No sensory symptoms	Ref.
Sensory symptoms present	0.48 (0.001, 398)	-52% (-99.9%, 39,700%)	0.82
Prior doctors visited
No specialist surgeon	Ref.
Specialist surgeon	9.55 (0.06, 1514)	855% (-94%, 151,300%)	0.35
Diagnosed outside of Zambia	0.40 (0.002, 93.3)	-60% (-99.8%, 9230%)	0.72

Variables were included in the all delay and health-system delay models if p < 0.05 in univariable analyses and in the patient delay model if p < 0.20 in the univariable analyses.

Bolded values indicate the values meeting a p-value threshold of <0.05.

Italicized values represent those trending towards significance with a p-value less than 0.1.

Discussion

Delay to diagnosis of MS varies widely worldwide. Contemporary cohort studies report a median diagnostic delay of 1.6 years in Italy,²³ and mean diagnostic delay of 0.9 years in Germany,²⁴ 1.1 years in the United States,²⁵ and 0.6 years in Iran.²² From other countries, however, even recent cohort studies suggest longer delays. A Columbian study found a mean delay of 3.07 years,²⁰ and cohort studies from Kenya and Nigeria have found that mean delays to diagnosis were approximately 4 years.^7,26 These differences are likely due to a host of factors, including limitations in resources, lack of referral pathways and/or access to specialized neurological care, and limited regional awareness of MS, and highlight the need for further characterization of the factors that influence diagnostic delay in different countries.²⁷ Prominent neurophobia among African medical students likely contributes to few neurologists and limited regional awareness about neurological conditions.²⁸

In this study, the mean delay to diagnosis of MS in Zambia was 3.6 years, which is comparable to contemporary cohorts from other countries with reported low MS prevalence.^7,20,26 Prominent findings were that race, birthplace, and country of evaluation most significantly impacted diagnostic delay, even after controlling for other factors.

When all delay was examined in a multivariable regression model, only race was significant. However, when this delay was further categorized into patient delay and health-system delay, evaluation outside of Zambia and birthplace outside of Zambia were significantly associated with patient and health-system delay, respectively. Race also trended toward significance in health-system delay. Of note, associations between race and time to diagnosis have not been previously reported in other cohorts.^29,30 One explanation is that those of Asian race or born outside Zambia are more likely to seek care outside Zambia in regions with greater MS awareness and, therefore, have a shorter delay to diagnosis. Alternatively, local physicians may be more likely to consider an MS diagnosis and/or referral to neurology when the person is not Black African, resulting in a shorter diagnostic delay. As such, long-held beliefs that MS is rare in SSA may prevent clinicians from even considering an MS diagnosis in a Black African person and further perpetuate under-ascertainment of MS in the region.^31,32 Both explanations support the underlying conclusion that Black African people born in Zambia and evaluated in Zambia experience prolonged diagnostic delays when compared to non-black Zambians born and/or evaluated outside Zambia.

Besides race and birthplace, another demographic variable that correlated with delay in unadjusted analyses was age at diagnosis, with older patients experiencing a prolonged health-system delay. This is consistent with findings from other cohort studies in Italy, Germany, and Iran^22–24 and may be explained by provider attribution of MS symptoms to an alternate diagnosis that is more common among older patients, especially cerebrovascular disease. However, age at diagnosis did not reach significance in a multivariable model, suggesting that the univariable association is more likely due to the prolonged diagnostic journey of patients who grew older by the time of diagnosis or a survivorship bias.

The lack of significant association of education level with delay in any analysis was notable, especially in a predominantly highly educated cohort, and supports the viewpoint that education among black Zambian people does not necessarily shorten diagnostic delay. This contrasts with findings from studies in Iran and Italy, where lower education levels prolonged delay.^22,23 However, education levels of this cohort were much higher than average education levels in the Zambian population. This may have led to an undercoverage bias due to the absence of data from those with lower educational attainment. Furthermore, low levels of MS awareness across the Zambian population may diminish the association between higher education and shorter diagnostic delay, as even those with higher education may not have awareness of MS.

In terms of notable clinical characteristics that were associated with delay in univariable analyses, the presence of sensory symptoms correlated with a shorter all delay and health-system delay, a finding which was not noted in other symptom categories, including vision, gait, or motor symptoms. There are mixed reports from different studies, with certain cohorts, especially in Portugal and Canada, reporting that isolated vision or motor symptoms were associated with prolonged delays, possibly due to these patients being initially referred to a medical or surgical specialist, while sensory symptoms resulted in a shorter referral delay directly to neurology.^19,21 In this cohort, being seen by a surgeon was associated with a significantly longer health-system delay in univariable analyses.

In contrast, studies from Iran and Italy reported that motor and vision symptoms at onset were associated with shorter delays, presumably due to the functionally impairing nature of these symptoms resulting in increased care seeking by patients or shorter referral delay to neurology.^22,23 In our cohort, due to most patients reporting multiple symptoms at onset of MS and sensory symptoms being a commonly co-occurring symptom, it is possible that the presence of sensory symptoms correlated with those who had more symptoms at onset. Indeed, in the multivariable analysis, it was not the presence of sensory symptoms but rather the number of symptoms during the first attack that trended toward significance in decreasing health-system delay. In addition, this supports the possible conclusion that patients presenting with multiple symptoms, especially sensory, were more likely to be referred to neurology sooner, whereas those presenting with isolated motor or vision symptoms, which may have better established referral pathways to specialists such as ophthalmologists or neurosurgeons/orthopedic surgeons, experienced longer delays. Longer delays in those referred to surgeons were also supported by qualitative work in Zambia and led to a targeted educational course for these specialists.^16,33

Within the median all delay of 11.4 months, much of this delay (8.8 months) occurred after first presentation to healthcare and was categorized broadly as health-system delay. This finding points to the role of country-specific healthcare system factors in delaying MS diagnosis, which may include lack of provider knowledge of MS, delays in appropriate referrals to neurology, and limited access to resources to aid with an expedited diagnostic evaluation, including both human resources (e.g. neurologists) and physical resources (e.g. MRI machines). When the diagnostic journey was examined further, visitation to a private versus public healthcare facility did not play a role, suggesting that despite the private healthcare sector’s lower barrier to access diagnostic tests, under-recognition of MS symptoms, lack of neurologists in local private practice, and low rates of neurology referral by private healthcare providers cannot be overlooked. In terms of prior provider subtype, consultation with a specialist surgeon associated with increased health-system delay in the univariable analysis, possibly due to a lack of knowledge of MS among local surgical practitioners or a higher threshold for specialist referrals compared to general practitioners. However, this association was not seen in the multivariable model. Furthermore, there was a reduction in all and health-system diagnostic delay for patients with symptom onset after the establishment of a neurology program in Zambia in 2018, underscoring the importance of access to neurology specialty care in making a timely diagnosis of MS.

There were a few notable limitations to this exploratory study, particularly as a single-center cohort study in a resource-limited setting. As such, we recognize the risk of overfitting the multivariable regression models and not detecting true associations due to the small sample size. However, we believe this unique cohort represents an important and under-investigated population, and these data may be useful in hypothesis generation for future, larger cohort studies, especially regional collaborations with pooled MS cases across sub-Saharan Africa. Still, they should be interpreted with caution considering these limitations. To ensure rigorous and accurate data, participants who were unable to complete diagnostic evaluation due to socioeconomic or geographic factors were excluded, as denoted by 59% of this cohort being college-educated, which is not reflective of Zambia’s overall population. As such, these data are likely not representative of the overall population of people with MS in Zambia, but they call attention to the continued persistence of underdiagnosis in the most socioeconomically vulnerable segments of the population. In addition, of those who were able to seek neurologist evaluation and afford diagnostic testing, potential participants with clinical and radiographic presentations atypical of MS were excluded due to local systemic limitations in obtaining supporting data to definitively rule out alternative diagnoses. Lack of an electronic medical record meant we relied on participant-provided historical data, raising a possibility of recall bias, and these could not be secondarily verified. While household income and occupation were included on questionnaires, participant refusal to disclose financial information was high, so these data were excluded, and education level was instead used as a proxy for these variables. In addition, the delay imposed by waiting for diagnostic tests such as MRI was not measured due to lack of validated dates, and this may have falsely inflated the health-system delay. Because this study was only conducted in Zambia among participants with MS, the results are highly contextualized and not generalizable to other populations. Regardless of these limitations, the strengths of this study include its conduct in an understudied and underrepresented region and population that, despite local resource limitations, still had access to fully trained neurologists who conducted the clinical evaluations and confirmed diagnoses.

This study elucidated several factors associated with delays to diagnosis of MS in Zambia. While a nationwide educational lecture series targeted for local providers was recently launched to advance knowledge of MS in Zambia, a more comprehensive study to understand gaps in provider knowledge, appraisal of MS symptoms, and referral patterns is needed. Furthermore, patient-specific factors, including perceptions and health literacy of MS symptoms, care-seeking thresholds, and selection of specialist and facility type, need to be more fully understood, especially among black Zambian people of varied socioeconomic backgrounds; while these factors have been qualitatively studied to understand delays in presentation for meningitis in Zambia, this remains unknown for neurologic symptoms of MS.³⁴ In addition, given the costs associated with delayed diagnosis, the role of economic factors remains largely understudied in this population. Further studies are imperative to improve health equity for other similar populations in SSA, as shortening diagnostic delays should lead to earlier access to treatment and less accrual of disability.

Supplemental Material

sj-docx-1-msj-10.1177_13524585251344832 – Supplemental material for Factors associated with delay to diagnosis of multiple sclerosis in Zambia

Supplemental material, sj-docx-1-msj-10.1177_13524585251344832 for Factors associated with delay to diagnosis of multiple sclerosis in Zambia by Malya Sahu, Mashina Chomba, Dominique Mortel, Sarah Braun, Lorraine Chishimba, Frighton Mutete, Naluca Mwendaweli, Coolwe Namangala, Stanley Zimba and Deanna Saylor in Multiple Sclerosis Journal

Supplemental Material

sj-pdf-2-msj-10.1177_13524585251344832 – Supplemental material for Factors associated with delay to diagnosis of multiple sclerosis in Zambia

Supplemental material, sj-pdf-2-msj-10.1177_13524585251344832 for Factors associated with delay to diagnosis of multiple sclerosis in Zambia by Malya Sahu, Mashina Chomba, Dominique Mortel, Sarah Braun, Lorraine Chishimba, Frighton Mutete, Naluca Mwendaweli, Coolwe Namangala, Stanley Zimba and Deanna Saylor in Multiple Sclerosis Journal

Footnotes

Data availability statement

Anonymized data not published within this article will be made available by request from any qualified investigator that follows Zambian regulations regarding data sharing.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This study received funding from a National Multiple Sclerosis Society Pilot Grant (PP-1807-32127), NIH Fogarty International Center (K01TW011771, D43TW009340), Multiple Sclerosis International Federation Du Pre Award, and American Academy of Neurology/American Brain Foundation Neurodisparities Clinical Research Training Scholarship.

ORCID iDs

Malya Sahu

Mashina Chomba

Sarah Braun

Frighton Mutete

Naluca Mwendaweli

Deanna Saylor

Supplemental material

Supplemental material for this article is available online.

References

Makhani

Morrow

Fisk

, et al. MS incidence and prevalence in Africa, Asia, Australia and New Zealand: A systematic review. Mult Scler Relat Disord 2014; 3(1): 48–60.

Browne

Chandraratna

Angood

, et al. Atlas of multiple sclerosis 2013: A growing global problem with widespread inequity. Neurology 2014; 83: 1022–1024.

Bower

Zenebe

. Neurologic services in the nations of Africa. Neurology 2005; 64: 412–415.

Aarli

Diop

Lochmuller

. Neurology in sub-Saharan Africa: A challenge for World Federation of neurology. Neurology 2007; 69: 1715–1718.

Hwang

Garcia-Dominguez

Fitzgerald

, et al. Association of multiple sclerosis prevalence with sociodemographic, health systems, and lifestyle factors on a national and regional level. Neurology 2022; 99: e1813–e1823.

Idris

Sokrab

Ibrahim

, et al. Multiple sclerosis in Sudan: A prospective study of clinical presentation and outcome. Mult Scler 2009; 15(12): 1537–1538.

Jamal

Shah

Mativo

, et al. Multiple sclerosis in Kenya: Demographic and clinical characteristics of a registry cohort. Mult Scler J Exp Transl Clin 2021; 7(2): 20552173211022782.

Bhigjee

Moodley

Ramkissoon

. Multiple sclerosis in KwaZulu Natal, South Africa: An epidemiological and clinical study. Mult Scler 2007; 13(9): 1095–1099.

Kavaliunas

Manouchehrinia

Stawiarz

, et al. Importance of early treatment initiation in the clinical course of multiple sclerosis. Mult Scler 2017; 23(9): 1233–1240.

10.

Goodin

Reder

Ebers

, et al. Survival in MS: A randomized cohort study 21 years after the start of the pivotal IFNbeta-1b trial. Neurology 2012; 78: 1315–1322.

11.

Kaisey

Solomon

. Multiple Sclerosis Diagnostic Delay and Misdiagnosis. Neurol Clin 2024; 42: 1–13.

12.

Uher

Adzima

Srpova

, et al. Diagnostic delay of multiple sclerosis: Prevalence, determinants and consequences. Mult Scler 2023; 29(11–12): 1437–1451.

13.

Thompson

Banwell

Barkhof

, et al. Diagnosis of multiple sclerosis: 2017 revisions of the McDonald criteria. Lancet Neurol 2018; 17: 162–173.

14.

Solomon

Weinstein

Shinohara

, et al. Diagnostic delay and misdiagnosis of symptoms reported by patients with multiple sclerosis participating in a research registry. Mult Scler J Exp Transl Clin 2025; 11(2): 20552173251333390.

15.

Chomba

Mortel

Saylor

. Impact of delayed diagnosis on Zambians living with multiple sclerosis (P4-8.009). Neurology 2023; 100: 4565.

16.

Chomba

MMD

Saylor

. Pathways to diagnosis of multiple sclerosis in Zambia. In: American Neurological Association Annual Meeting, Chicago, IL, 22–25 October 2022.

17.

Harris

Taylor

Minor

, et al. The REDCap consortium: Building an international community of software platform partners. J Biomed Inform 2019; 95: 103208.

18.

Harris

Taylor

Thielke

, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform 2009; 42(2): 377–381.

19.

Aires

Barros

Machado

, et al. Diagnostic delay of multiple sclerosis in a Portuguese population. Acta Med Port 2019; 32: 289–294.

20.

Cárdenas-Robledo

Lopez-Reyes

Arenas-Vargas

, et al. Delayed diagnosis of multiple sclerosis in a low prevalence country. Neurol Res 2021; 43(7): 521–527.

21.

Kingwell

Leung

Roger

, et al. Factors associated with delay to medical recognition in two Canadian multiple sclerosis cohorts. J Neurol Sci 2010; 292: 57–62.

22.

Mobasheri

Jaberi

Hasanzadeh

, et al. Multiple sclerosis diagnosis delay and its associated factors among Iranian patients. Clin Neurol Neurosurg 2020; 199: 106278.

23.

Patti

Chisari

Arena

, et al. Factors driving delayed time to multiple sclerosis diagnosis: Results from a population-based study. Mult Scler Relat Disord 2022; 57: 103361.

24.

Blaschke

Ellenberger

Flachenecker

, et al. Time to diagnosis in multiple sclerosis: Epidemiological data from the German multiple sclerosis registry. Mult Scler 2022; 28(6): 865–871.

25.

Jakimovski

Kavak

Zakalik

, et al. Improvement in time to multiple sclerosis diagnosis: 25-year retrospective analysis from New York State MS Consortium (NYSMSC). Mult Scler 2023; 29(6): 753–756.

26.

Okubadejo

Ojo

Lawal

, et al. Unveiling multiple sclerosis in Nigeria: The conundrum of diagnosis and access to disease modifying therapies (P5.152). Neurology 2014; 82: P5152.

27.

Solomon

Marrie

Viswanathan

, et al. Global barriers to the diagnosis of multiple sclerosis. Neurology 2023; 101: e624–e635.

28.

McDonough

Chishimba

Chomba

, et al. Neurophobia in Africa: Survey responses from fifteen African countries. J Neurol Sci 2022; 434: 120161.

29.

Langer-Gould

Brara

Beaber

, et al. Incidence of multiple sclerosis in multiple racial and ethnic groups. Neurology 2013; 80: 1734–1739.

30.

Cree

Reich

Khan

, et al. Modification of multiple sclerosis phenotypes by African ancestry at HLA. Arch Neurol 2009; 66(2): 226–233.

31.

Dean

Bhigjee

Bill

, et al. Multiple sclerosis in black South Africans and Zimbabweans. J Neurol Neurosurg Psychiatry 1994; 57(9): 1064–1069.

32.

Dabilgou

AANC

Teinture / Kambou

Kyelem

JMA

, et al. Multiple sclerosis in West Africa, about a case confirmed at Ouagadougou, Burkina Faso. J Mult Scler 2017; 4: 2.

33.

Duodu

Simpson

Gouider

, et al. Effectiveness, relevance, and feasibility of an online course on multiple sclerosis for African healthcare workers (P4.009). In: Merino

(ed.), American Academy of Neurology. San Diego, CA: Neurology, 2025.

34.

Elafros

Bwalya

Muchanga

, et al. A qualitative study of factors resulting in care delays for adults with meningitis in Zambia. Trans R Soc Trop Med Hyg 2022; 116: 1138–1144.

Supplementary Material

Please find the following supplemental material available below.

For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.

For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.

0.00 MB

0.02 MB

0.10 MB