Electrodiagnostic referrals and neuromuscular disease pattern in East Africa: Experience from a tertiary hospital in Ethiopia.

Objective: We present a retrospective cross-sectional review of the electrodiagnostic (EDX) referral and diagnostic patterns in patients with suspected neuromuscular conditions at a tertiary hospital in Ethiopia.
Methods: Between 2016 and 2019, 313 patients were evaluated at the EDX lab in Tikur Anbessa Specialised Hospital, Addis Ababa, Ethiopia. In our patients, nerve conduction study and when appropriate needle electromyography was done. Demographic and clinical data were extracted from a digital registry.
Results: Polyneuropathy (19.8%), carpal tunnel syndrome (12.1%), and lumbosacral radiculopathy (9.9 %) were the top three reasons for EDX referral in Ethiopia. Among them, polyneuropathy was the most frequent electrodiagnosis, where diffuse axonal and demyelinating subtypes accounted for 54% and 18.8%, respectively. Guillain-Barre syndrome was suspected in 18 patients and 15 had EDX confirmed motor axon polyneuropathy while three patients had demyelinating variant. Although a quarter (26.2%) of the referrals had a normal EDX, abnormal test results were significantly associated with weakness (95% CI: 3.29-7.04, p < 0.001), bound to wheelchair (95% CI: 1.86-2.87, p = 0.01) and having a specific diagnosis at time of referral (95% CI: 2.53-4.68, p = 0.007). Conclusions: Diffuse and entrapment neuropathies were the main reasons for electrodiagnosis test referrals in Ethiopia. Also, motor axonal variant was the most common type of inflammatory polyneuropathy diagnosed with EDX. Significance: Proper patient evaluation and documentation significantly improves the diagnostic yield and cost-effectiveness of EDX testing in a resource-poor countries like Ethiopia. This might be achieved through educating medical students, residents, and other care providers on the basics of EDX and its indications for correct use in the clinical care.

Introduction

Electrodiagnostic (EDX) studies, including nerve conduction studies (NCS) and needle electromyography (EMG), remain the most effective diagnostic tools in the evaluation of patients with suspected neuromuscular disorders (NMD). NCS and needle EMG studies are complementary and usually performed together, playing a vital role in assessing peripheral nervous system (Podnar, 2005).

In standard clinical practice, EDX study do not replace a detailed neurologic examination, but rather are performed as an additional tool of a complete clinical assessment, to assist with localization and establishing a working diagnosis (Fuller, 2005). EDX might also offer an insight into treatment selection and prognosis (Gutierrez and Sumner, 2014).

In Ethiopia, EDX services are only available sporadically with limited patient access, as the services are largely concentrated in a small number of private centres in the capital, Addis Ababa. Additionally, each test costs between $60 and $90 from an out-of-pocket payment, which is quite expensive for an ordinary Ethiopian citizen. Since 2018, the Department of Neurology at Addis Ababa University revitalized the EDX service at the largest public referral hospital in Ethiopia, Tikur Anbessa Specialized hospital (TASH), with subsidized cost and free of charge for patients covered by the Community-based Health Insurance. Despite the growing burden of neuromuscular disorders, there is a dearth of published data on the epidemiology of neuromuscular disorders in Ethiopia. To the best of the authors knowledge, this is the first study to report on the NCS/EMG profile of Ethiopian patients. This study aims to review the clinical indications, and electrodiagnostic profiles among patients with neuromuscular disorders referred to the electrophysiology laboratory in TASH, Addis Ababa, Ethiopia. We believe this study lays the ground to facilitate further research and contribute to the current understanding of neuromuscular disorders in the region.

Methods and materials

Study area

The study was conducted at the electrophysiology laboratory of Tikur Anbessa Specialized Hospital, the largest tertiary level referral and teaching hospital, under the College of Health Sciences of Addis Ababa University (CHS-AAU) in Addis Ababa, Ethiopia. The hospital serves a large population of patients referred from every corner of the country for outpatient and inpatient care in multidisciplinary specialties (neurology, internal medicine, emergency, pediatrics, neurosurgery, and orthopedics). TASH is the only neurology training center in Ethiopia, and it has been providing comprehensive outpatient, in-patient, and emergency neurological services since 2012. In May 2016, under the leadership of the Department of Neurology, the EDX lab was renovated to provide NCS, EMG, Electroencephalogram (EEG), and Evoked Potentials (EP) services. It also serves as EDX referral center for Zewditu Memorial Hospital (ZMH), a CHS-AAU affiliated neurology training center under the Addis Ababa City Administration. Although the department lacks a neuromuscular or neurophysiology sub-specialist, the nine neurologists at the department who co-authored this paper run the lab and performed all the EDX tests. The EDX lab also functions as one of the training sites for neurology residents.

Study design and study period

This retrospective, cross-sectional study was conducted by reviewing the electronic records of consecutive patients referred for the first time to the EDX laboratory at TASH between 2016 and 2019. Patients who had a repeat test and incomplete data were excluded.

Data collection

An excel sheet was prepared to extract all the necessary information from a structured referral form completed by the treating physician and approved by the attending neurologist at the time of EDX referral. We collected the socio-demographic data (age, sex, referral unit, and mobility), comorbidities, working diagnosis of the referring physician, and indication for EDX referral. Also, the final diagnosis from the electrodiagnostic evaluation was documented.

One Neuropack X1 MEB-2300 (Nihon Kohden) NCS/EMG machine was used for the EDX assessment and different combinations of sensory and motor conduction studies, including F-wave and H reflex, were performed in common nerves in the limbs. For suspected cases of neuromuscular junction disorders, repetitive nerve stimulation (RNS) was done in addition to routine NCS. Needle EMG study using concentric needle was done in selected patients when the examining neurologist considered this test essential to exclude a competing diagnosis or confirm the suspected diagnosis based on his or her clinical judgments.

The World Federation of Neurology (WFN) Classification of Neuromuscular Diseases (Swash and Schwartz, 1997) was used to categories the referral diagnosis and the electrodiagnosis in a standardized form. Accordingly, neuromuscular disorders were classified as; 1) Spinal Muscular Atrophy (SMA) and other disorders of motor neurons, 2) Disorders of motor nerve roots, 3) Disorders of the peripheral nerves, 4) Disorders of neuromuscular transmission and 5) Disorders of muscle.

In the absence of local normative data, all test results were interpreted using normative data from standard published references (Preston and Shapiro, 2013) or in cases where side-to-side asymmetries found, intra-personal comparisons were taken. All test results and technical limitations were interpreted in the context of referral diagnosis, examiners' clinical judgment, and any additional investigation results to reach the final diagnosis.

Statistical analysis

Cleaned and completed data from the excel sheet was exported to the Statistical Package for Social Science (SPSS, IBM, Chicago, IL, USA) Version 25 for data analysis. Descriptive statistics were summarized using mean, standard deviation (SD), and range for continuous data. Frequencies and proportions were utilized for categorical data. Pearson’s Chi-square or Fisher’s exact tests were used to evaluate the association between the electrodiagnosis and independent variables. Variables with p values < 0.05 were considered statistically significant.

Ethical approval and consent to participate

The study protocol obtained ethical approval from the Institutional Review Board (IRB) of the College of Health Sciences in Addis Ababa University (Protocol number: 003/20/Neuro). Waiver of consent was received due to the nature of the study.

Results

Demographic data and referral pattern

During the three- and a half-year period, 313 patients with presumed neuromuscular disorders were referred to the EDX lab in TASH for comprehensive EDX evaluation. Among them, the majority (93.9%) were referred by neurology, pediatric neurology, orthopedics, internal and emergency medicine physicians at TASH while the rest 6.1% were from the neurology outpatient clinic in ZMH. Males accounted for 57.5% of the referrals. Most of the electrophysiology tests (84%) were done for adult patients aged between 18 and 59 years, while children under 18 years represented only 7.3% of the referrals. The mean age (±SD) of participants was 39.3 (±15.6) years (Ranges: 2–85 years).

During the EDX lab visit, 90.4% (n = 283) of the participants were self-ambulating. Fifty-one (16.3%) patients had a documented comorbidities where diabetes and HIV infection were the most frequent ones at 31.4% (n = 16) and 25.5% (n = 13), respectively. Motor weakness (40.2%) was the main reason for referral followed by sensory complaints (29.7%) and trauma/iatrogenic complications (10.5%). Demographic, clinical, and referral characteristics of patients referred for EDX tests are summarized in Table 1.

Table 1

Demographic, comorbidity, and referral characteristics of patients evaluated at the EDX lab in TASH.

Baseline demographic and clinical patterns		Frequency	Percentage (%)
Age (years)	<18	23	7.3
	18–59	263	84.0
	≥60	27	8.6
Sex	Male	180	57.5
	Female	133	42.5
Referral unit
	TASH	294	93.9
	ZMH	19	6.1
Mobility	Walking	283	90.4
	Wheelchair	30	9.6
Comorbidity	Diabetes mellitus	16	31.4
	HIV infection	13	25.5
	History of malignancy	10	19.6
	Chemotherapy	8	15.7
	Others	12	23.5
	Not recorded	262	83.7
Reason for referral	Motor weakness	126	40.2
	Sensory abnormality	93	29.7
	Post-trauma/iatrogenic complications	33	10.5
	Lower limb radiculopathy	29	9.3
	Upper limb radiculopathy	15	4.8
	Others	17	5.4

EDX: Electrodiagnosis; TASH: Tikur Anbessa Specialized Hospital; ZMH: Zewditu Memorial Hospital.

Working diagnosis and electrodiagnosis for referred patients

More than half of the patients referred with sensory complaint or pain had a normal EDX result while two-third of the cases with focal or diffuse motor weakness showed an abnormal EDX (Fig. 1). The commonest indications for EDX referral were disorders of the peripheral nerves (38.9%). Among them, (26.2%) had a clinical diagnosis of polyneuropathies followed by mononeuropathies (19.1%) at different anatomic sites. The second most common (18.8%) reason for EDX referral were disorders of the nerve roots including lumbosacral (9.9%) and cervical (4.2%) radiculopathy and brachial plexopathy (4.8%). Motor neuron disease was suspected in 7.7% of the patients. The least frequent causes of referrals were disorders of the muscle (5.4%) and neuromuscular transmission (2.2%). A quarter (26.8%) of patients were referred without a clear working diagnosis, but neurologic symptoms such as pain, numbness, and weakness were documented in the request form.

Fig. 1

The frequency of electrodiagnosis test outcomes based on main referral complaint.

NCS was done for all patients and when appropriate, needle EMG was done in 33.8% of the participants. EDX abnormalities were documented in (73.2%) of the participants with variable degrees of severity. The most frequent (27.1%) electrodiagnosis were polyneuropathies with further categorization into axonal (11.8%) and demyelinating (5.1%) type based on an established criterion (5). Acute and chronic inflammatory polyneuropathies accounted for 5.8% and 1.3% of the referrals, respectively. In both conditions, the clinical and electrodiagnosis showed 100% congruency. Among the Guillain-Barre syndrome (GBS) cases, 15/18 (83.3%) met the EDX criteria for acute motor axonal polyneuropathy (AMAN) variant, while only 3 (1%) patients showed acute inflammatory demyelinating polyneuropathy (AIDP) features on their NCS. One (0.3%) case was consistent with multifocal motor neuropathy with conduction block (MMNCB).

A quarter (24.6%) of EDX referrals were for mononeuropathies. Among them, the majority 30 (9.6%) had carpal tunnel syndrome (CTS) with variable severity (mild 23.3%, moderate 26.6%, and severe 50%) and anatomic locations (unilateral: 60% and bilateral: 40%). Other less frequent focal neuropathies include peroneal neuropathy (4.8%), multifocal mononeuropathies (2.2%), and ulnar neuropathy at the elbow (1.9%).

NCS/EMG evaluation confirmed the diagnosis of amyotrophic lateral sclerosis (ALS) in all 24 (7.7%) patients suspected of motor neuron disease with 100% congruence. Five had definite ALS, 13 probable and 6 possible ALS diagnosis based on the revised El Escorial criteria (Brooks et al., 2000). Seven patients suspected with myoneural junction disorders tested by NCS, EMG and RNS and four had the typical EDX features of myasthenia gravis (MG) (decremental response on RNS) but the rest three had normal studies.

Although 17 (5.4%) of the referrals were made for suspected muscle disorders, NCS/EMG studies identified thirteen inflammatory and twelve non-inflammatory myopathies based on the presence or absence of abnormal spontaneous activity on EMG. Besides, two (0.6%) patients had features of myotonic dystrophy. A quarter (26.8%) of the referrals had normal EDX tests. Fig. 2 summarizes the referral (suspected) diagnosis and electrodiagnosis categorised according to the WFN Classification of Neuromuscular Diseases.

Fig. 2

The frequency (%) of referral diagnosis and electrodiagnosis categorized according to the WFN Classification of Neuromuscular Diseases.

Association between electrodiagnosis finding and patient variables

We hypothesized that patients who had a specific diagnosis at the time of referral and those with motor weakness will have a specific final EDX. Accordingly, we performed Pearson’s Chi-square or Fisher exact test to assess the correlation between abnormal electrodiagnosis findings and different clinical and demographic variables such as gender, age group, mobility, presenting symptom, and referral diagnosis. Patients presented with motor weakness (OR: 4.81, 95% CI: 3.29–7.04, p < 0.001), those who were wheelchair bounded at presentation (OR: 2.32, 95% CI: 1.86–2.87, p = 0.01) were significantly associated with abnormal EDX test results. Besides, patients referred with a completed forms mentioning a clear EDX indications and specific referral diagnosis were associated with congruent electrodiagnosis (OR: 3.44, 95% CI: 2.53–4.68, p = 0.007) as compared to those with incomplete form or without diagnosis. Gender and younger age group did not show statistically significant agreement with EDX test outcome (Table 2).

Table 2

The association between electrodiagnosis and patient profiles among patients referred to the EDX lab in TASH.

Variables	Final electrodiagnosis		χ2 test	OR [95% CI]	p-value
	Normal 83 (26.5%)	Abnormal 230 (73.5%)
Sex
Male	44	136	0.93	2.41 [1.66–3.5]	0.34
Female	39	94	1
Age category
< 18 years	5	18	1
≥ 18 years	78	212	0.29	0.75 [0.27-2.1]	0.59
Mobility
Wheelchair	2	28	1
Walking	81	202	6.71	2.32 [1.86–2.87]	0.01*
Presenting symptom
Sensory abnormality	51	76	1
Motor weakness	32	154	20.41	4.81 [3.29–7.04]	<0.001*
Referral diagnosis
No specific diagnosis	31	51	1
Specific diagnosis	52	179	7.27	3.44 [2.53–4.68]	0.007*

EDX: Electrodiagnosis; TASH: Tikur Anbessa Specialized Hospital. *Statistically significant; χ2 test: Chi-square test.

Discussion

In Ethiopia, due to the lack of widely available and affordable EDX service, patients presenting with neuromuscular complaints often remain underreported, misdiagnosed or untreated. In 2016, the department of neurology revitalized the electrodiagnostic laboratory in TASH. Since then, we have noticed a growing number of patients referred for EDX evaluation. To the best of our knowledge, this is the first study to critically review the referral pattern and EDX profile of patients with neuromuscular symptoms who were evaluated in the electrophysiology laboratory of one of the largest public hospitals in Ethiopia, TASH, over a three-year period.

We found that polyneuropathy was the most common (26.2%) reason for EDX referral, and it was also the most frequently (27.1%) documented electrodiagnostic abnormality. Similar results were reported from tertiary neurology referral centres in Serbia (Nikolic et al., 2016) and Zambia (Kvalsund et al., 2019), where polyneuropathies were the most common suspected and confirmed diagnosis among patients referred for EDX testing. However, other studies from Siena (Cocito et al., 2006, Mondelli et al., 2014, Podnar, 2005), Italy showed that carpal tunnel syndrome and upper/lower limb radiculopathies were the frequent cause of referral as well as EDX findings Also, a report from tertiary centre in the USA (Cho et al., 2004) reported that only 13% of the referral cases had a diagnosis of polyneuropathy. This gap might be explained by a shortage of skilled professionals and poorly equipped facilities noted at the primary care level in Ethiopia and other developing nations, leading to delay in diagnosis and referral of chronic conditions like polyneuropathies to tertiary centers. The opposite is true in the developed countries like the USA, where a higher percentage of such patients likely obtain proper and timely evaluations at the community hospital and only complicated cases are sent to the higher level of care.

In our study, GBS was clinically considered in eighteen (5.8%) patients and their EDX evaluation confirmed the diagnosis in all of them. Additionally, EDX studies allowed us to classify the syndrome into their different subtypes. One of the unique finding of this study was that the motor axon variant (AMAN) of GBS was the predominant electrophysiologic subtypes identified, accounting for (n = 15, 83.3%) of the GBS referrals. Only three patients were diagnosed with the demyelinating (AIDP) variant. In line with this finding, Tian J. et al (Tian et al., 2019) from Hebei province in northern China assessed the electrophysiologic subtypes of GBS patients and they found that most patients (55.8%) had AMAN subtype while 21.2% had AIDP subtype of GBS. They also reported that the AMAN subtype was significantly associated with a history of antecedent diarrhea, longer duration of hospitalization, and slightly slower recovery than AIDP cases. On the other hand, a multicentre study from Saudi Arabia (Alanazy et al., 2021) reported that AIDP was the commonest variant (37.7%) followed by AMAN and AMSAN subtypes in 29.5% and 19.2%, respectively. The international Guillain-Barré Syndrome Outcome Study, a multi-national cohort study in different regions of the world, also reported that demyelinating subtypes of GBS were the most prevalent EDX variant detected in all regions (Europe/Americas: 55%; Asia: 45% and Bangladesh: 40%) while the axonal subtype was predominant (36%) in Bangladesh (Doets et al., 2018). These variations might be explained by the variable environmental and genetic risk factors found in different geographic regions. Furthermore, the high prevalence of axonal variants in our study can be explained by the study setting where patients having severe weakness from axonal involvement are likely to get referred to a tertiary centre for EDX evaluation unlike the milder forms noted with AIDP variants.

Twenty-four (7.7%) patients with suspected MND underwent NCS/EMG evaluation. All cases fulfilled the revised El Escorial diagnostic criteria for ALS diagnosis. Using the same diagnostic criteria, a Zambian study reported a relatively higher frequency (11%) of ALS cases among the 12% adults referred for MND evaluation (Kvalsund et al., 2019). In contrary, Nikolic and his colleagues (Nikolic et al., 2016) confirmed the diagnosis of MND in less than half of the patients who arrived to their EDX lab for suspected MND. Similarly, P.B. Adebayo and his colleagues from Ibadan, Nigeria evaluated three suspected cases of MND and only one of them met the diagnostic criteria (Adebayo et al., 2018). One possible explanation for 100% congruency between clinical and electrodiagnosis of MND patients in our lab is that most of our patients see a neurologist or referred for EDX evaluation once the disease advances and patients developed a severe weakness. The reason behind these diagnostic delays needs to be investigated in another study.

A report from a Zambian EMG lab identified 22% of the patients referred for the NCS/EMG study had comorbid HIV infection while a history of diabetes was documented in 3% of their cohort (Kvalsund et al., 2019). In our study, the opposite is true as diabetes was documented in 31.3% (n = 16) of patients and HIV positive cases accounted for 25.5% ((n = 13 of the documented comorbidities. This reflects different frequency of HIV infection in both countries.

In our lab, almost three-quarter (73.5%) of the EDX evaluations had an abnormal test result. Other prior studies reported similar diagnostic frequencies ranging from 37% (Haig et al., 1999) to 77% (Kvalsund et al., 2019) of the referred cases. Prior studies reported that electrodiagnostic studies allowed them to make a new diagnosis at a variable frequency ranging from 13 to 43% (Cho et al., 2004) of the cases. In our study, we identified that a quarter (26.5%) of the EDX evaluations had normal test results and the major contributing factor was that an equal proportion (n = 82, 26.2%) of the referrals lack a specific clinical diagnosis. This reflects the challenge we faced in our daily practice where providers are highly dependent on diagnostic tests and procedures, giving little attention to the history and physical examination. Our data indicates that patients referred with a specific clinical diagnosis, those having motor deficits and those bound to a wheelchair at the time of referral were significantly associated with the abnormal electrophysiology test result.

Some of the limitations of our study includes a retrospective study design, incomplete documentation, lack of normative data to our study population, and a single centre study serving mostly an adult population might limit the generalizability of our study finding. The main strength of this study is that it adds to the dearth of knowledge in the electrophysiology literatures and shine a light on the electrophysiology profile of different neuromuscular disorders in Ethiopia and the eastern Africa.

In conclusion, electrophysiologic studies have a tremendous advantage in confirming the clinical diagnosis and excluding competing differential diagnosis among patients with neuromuscular disorders. Polyneuropathy, carpal tunnel syndrome, and lumbosacral radiculopathy were the most common reasons for EDX referral in Ethiopia. Having a specific working diagnosis improves the diagnostic yield of EDX study. Our study indicates that the EDX service is an integral part of patient care. Future prospective case-control studies are recommended to establish a normative data for our setting and additionally generate a local epidemiological knowledge.

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.