Chronic kidney disease (CKD) has become a public health problem worldwide, particularly in Sub-Saharan Africa, due to its increasing frequency [1] . In recent decades, the number of people on dialysis has been estimated at 2.6 million worldwide, and statistics suggest that the number will double by 2030 [2] .

The frequency and lethality of emergency hemodialysis (EH) vary widely from country to country. According to data from an American meta-analysis, undocumented immigrants with end-stage renal disease treated with emergency hemodialysis have a higher 3-year mortality rate than patients who received standard hemodialysis [3] . In Madagascar, the frequency of EH was 41.93% at Joseph Raseta Befelatanana Hospital, with a mortality rate of 23.08% [4] .

In Mali’s capital, out of 62 EH performed at the Point G University Hospital in 2020, the case fatality rate was 69.35% [5] . The low socio-economic level in our context has a major impact on the regular follow-up of nephrology patients, who generally reach the advanced stages of CKD complications [6] [7] . As a result, hemodialysis (HD) carried out on an emergency basis, i.e., immediately or within the first 48 hours, in the face of a life-threatening situation, is responsible for a high morbi-mortality rate, with a deterioration in quality of life, longer hospital stays and increased cost of care [8] [9] .

To date, no data are available on EH in a regional health facility in Mali, hence the initiative to carry out this study, which aims to describe the epidemioclinical profile of emergency hemodialysis patients at the Fousseyni Daou Hospital in Kayes, Mali.

Data were collected from patients’ individual medical records (consultation register, hospitalization record, dialysis record). Using the medical records, data were collected on an individual survey form including:

- Socio-demographic data included age, gender, socio-economic level (group I = senior government and/or private sector executives and import-export traders, group II = government and/or private sector employees and medium-sized traders, group III = manual workers, peasants, retail traders, and casual urban workers, provenances), insurance coverage or not. Group III represented patients with low socio-economic status.

- The clinical data sought were gastrointestinal symptoms (hiccups, anorexia, nausea, vomiting, gastrointestinal haemorrhage, constipation, diarrhoea), cardiovascular symptoms (hypertension, congestive heart failure, pericarditis, acute pulmonary oedema, cardiac arrhythmia, cardiomyopathy, accelerated atherosclerosis), respiratory symptoms (Kusmaul respiration, bronchopneumopathy, pleurisy), urinary symptoms (increased or decreased frequency of micturition, nocturia), skin symptoms (palpebral edema, facial puffiness, peripheral edema, pallor, pruritus, scratch lesions), neuromuscular symptoms (headache, sleep disorders, confusion, temporo-spatial disorientation, muscle weakness, myoclonus, cramps, convulsions, coma) and symptoms of mineral-bone disorders (bone pain, bone fractures).

- The risk factors investigated were classic (tobacco, alcohol, drugs, obesity, sedentary lifestyle, hypertension, diabetes, dyslipidemia) and specific to CKD (fluid retention, phosphocalcic disorders, anemia, left ventricular hypertrophy, homocysteinemia, arteriovenous fistula).

- Biological tests included blood tests (hemogram, urea, creatinemia, uric acid, ionogram, phosphocalcic balance, martial balance, lipid balance, vitamin balance, albuminemia, protidemia, transaminases and infectious diseases), urine tests (urine cytobacteriological study, 24-hour proteinuria, urine ionogram). The standards used were those of the Kayes region laboratories. In the absence of gasometry, severe metabolic acidosis was clinically evoked by the presence of Küssmaul respiration with free lungs on auscultation.

- Renal ultrasound was used to assess size, cortico-medullary differentiation and the presence or absence of kidney dilatation.

- Signs of acute pulmonary oedema (APO), pleurisy, cardiomegaly and pulmonary infection were detected by chest radiography.

- Electrocardiogram (ECG) and echocardiography were used to detect various cardiac abnormalities.

- The fundus was examined for signs of hypertensive and/or diabetic retinopathy.

- Data on vascular access included type of approach and time between catheter insertion and removal.

- The absence of renal biopsy, the etiology of renal failure was based on available clinical and paraclinical arguments.

- Patients’ progress within 30 days of the first dialysis session was assessed. We chose this duration because of the higher mortality during the first month of dialysis [4] [9] . It included lethality, total recovery of renal function in HD, partial recovery of renal function in HD and maintenance in chronic HD.

Mean age: 44.63 ± 18.02 years, extremes: 3 and 88 years.

During the study period, out of 178 hemodialysis patients, 110 underwent dialysis in an emergency setting, representing a prevalence of 61.79%.

The mean age of patients was 44.63 ± 18.02 years, with extremes of 3 and 88 years. Females accounted for 54.5% (60 cases). Hypercreatininemia was the main reason for hospitalization, accounting for 98.2% (108 cases), with a mean admission value of 1846.2 µmol/l and extremes of 461 and 3300 µmol/l. One patient out of two came from the emergency department of the facility. 86.4% of patients were uninsured and of low socio-economic status ( Table 1 ).

A history of hypertension was found in 71.8% (79 cases), with irregular follow-up in 94.9% of patients and a mean duration of hypertension of 4.89 ± 4.65 years. Patients were known diabetics in 10.9% (12 cases). Functional signs were dominated by physical asthenia, vomiting, anorexia and nausea in 92.7%, 83.6%, 80.9% and 67.3% respectively. At physical assessment, 56.4% of patients were suffering from fluid retention, compared with 16.4% from dehydration. Twenty-four patients (21.8%) had pericardial friction on cardiac auscultation ( Table 2 ).

The blood pressure profile according to the WHO classification revealed grade 3 hypertension in 26.4% of patients, and Grades 2 and 1 in 30.9% and 20% of cases respectively.

Eighty-nine patients (80.9%) were in chronic end-stage renal failure versus 21 cases (19.1%) of severe acute renal failure, KDIGO stage 3. Patients had a mean hemoglobin level of 7.08 ± 2.01 g/dl, including 57.3% (63 patients) with a hemoglobin level below 8 g/dl ( Table 3 ). Blood ionograms showed hypocalcemia, hyponatremia and hyperkalemia in 39.1%, 30% and 62.7% of cases respectively.

CKD: chronic kidney disease; AKI: acute kidney injury.

Hyperparathyroidism and hypovitaminosis D were found in 64.9% and 85.1% of cases (N = 74). Proteinuria was minimal in 91% of cases. Urinary tract infection was found in 57.2% of patients (N = 92), with Escherichia coli predominating in 45.3% of cases.

Among hypertensive patients who had a fundus examination (N = 70), 42.9% (30 cases) had a normal fundus. Hypertensive retinopathy was stage 2 in 31.4% of cases, stage 3 and 1 in 12.9% and 11.1% respectively. Diabetic retinopathy was noted in 4 cases (3.6%). Left ventricular hypertrophy was found on ECG (N = 67) in 34 patients (50.7%).

The initial nephropathy of CKD patients (N = 89) was, in order of frequency, malignant nephroangiosclerosis (malignant hypertension) 38 cases (42.7%), chronic glomerulonephritis 21 cases (23.6%), chronic interstitial nephritis 17 cases (19.1%), benign nephroangiosclerosis 7 cases (7.9%), diabetic nephropathy 4 cases (4.5%), and undetermined nephropathy 2 cases (2.2%). The etiologies of organic AKI cases (N = 21) included 13 cases of acute tubular necrosis (62%), 4 cases of acute interstitial nephritis (19%), 2 cases of acute rhabdomyolysis (9.6%), 1 case of acute obstructive uropathy (2.2%) and 1 case of hemolytic uremic syndrome (2.2%) ( Table 4 ).

Urgent indications for hemodialysis were uremic syndrome with encephalopathy or coma (70%), anuria lasting more than 48 hours with severe renal failure (11.8%), acute lung edema (7.3%), uremic pericarditis (6.4%) and severe hyperkalemia (4.5%). The right femoral catheter was the most commonly used vascular approach (87.3%). Native AVF was used in 3 patients (2.7%) ( Table 5 ).

The overall outcome was favorable in 71 patients (64.5%), compared with 24 cases of death (21.8%) and 15 cases (13.6%) of abandonment of care against medical advice. Deaths occurred in a clinical context of deep coma (6 cases), septic shock (5 cases), ischemic stroke (5 cases), hemorrhagic stroke (2 cases), PAO (3 cases), pulmonary embolism (1 case) and 2 unspecified cases. Progression on dialysis was marked by total recovery of renal function in 16 patients (14.5%) out of 21 in acute renal failure versus 4 cases (3.6%) of partial recovery, 47 patients (42.7%) were maintained on chronic HD and 6 patients (5.5%) transferred to another center ( Table 6 ).

There was a statistically positive relationship between uraemic coma and the occurrence of patient death [Chi-square = 13.750 ddl = 1 P = 0.0002] ( Table 7 ).

During the study period, of 178 hemodialysis patients, 110 were dialyzed in an emergency setting, a prevalence of 61.79%. In the literature, the prevalence of EH varies between 40% and 60% [9] [13] . This variation in frequency is explained by the absence of a clear consensus on the variability of the definition of EH, which changes from one study to another.

This high proportion of EH in sub-Saharan Africa bears witness to the delay in consultation and management of pathologies that can lead to renal failure, but also to the increasing incidence of these pathologies [14] [15] .

Emergency hemodialysis patients were young, with a mean age of 44.63 ± 18.02 years, compared with 36.82 years at the Point G hospital reported by Traoré AK et al. in 2020 [5] . Our results are similar to data recently reported from Tunisia and Senegal, with mean ages of 58 years (±4) and 46.39 years (±17.13) respectively [16] [17] . The difference with Western countries is explained by the ageing of the population and the greater accessibility of care for the elderly in industrialized countries. Numerous studies show the predominance of men in chronic kidney disease [13] [18] . The predominance of women in this study could be explained by the high level of male immigration to the Kayes region.

The predominance of CKD, similar to other studies, and the absence of a permanent vascular approach may be explained by the late referral of patients with CKD to nephrologists, coupled with possible denial of management or abandonment of follow-up in our low socio-economic context [6] [7] [17] . On the whole, before their first nephrological consultation in Kayes, patients had already transited through one or two health care districts, where creatinine dosage is often not given priority over general pathologies such as hypertension and diabetes [19] . Indeed, early referral to nephrologists of patients in stage 3/4 of CKD reduces the prevalence of EH and lowers the mortality rate after initiation of dialysis [20] [21] .

This early recourse prepares them for dialysis by sparing the venous capital and creating permanent vascular access, notably the arteriovenous fistula.

Hypertension and diabetes were the main pathologies reported. In our study, they were encountered in 71.8% and 10.9% of cases respectively. In the studies by Traoré AK et al. and Brown et al. emergency dialysis patients were hypertensive in 77.4% and 93% of cases respectively [5] [22] . According to the literature, 28% to 46% of patients were diabetic [23] . These data further confirm the epidemiological transition through the emergence of chronic non-communicable pathologies.

Our study also revealed the predominance of malignant hypertension among the etiologies of CKD. These results reflected the lack of proper follow-up of hypertensive patients. In the earlier Malian study by Traoré AK et al. in 2020, chronic glomerulonephritis was the main cause of CKD, followed by hypertension [5] . According to data from several studies, hypertension remains the most frequent etiology of chronic kidney disease in Kayes [19] [24] .

On admission, the mean hemoglobin level was 7.08 g/dl, compared with between 9.5 and 10.6 g/dl in studies [4] [18] [24] .

In EH indications, the majority of patients were in the poor general condition of uremic syndrome with encephalopathy or coma, in contrast to other studies where hyperkalemia and clinical metabolic acidosis represented the main indications [4] [25] . This difference may be explained by late referral on the one hand, and the low socio-economic status of patients on the other.

Our short-term case fatality rate was 21.8% versus 69.35% and 23.08% reported by Traoré AK et al. [5] and Randrianarisoa RMF et al. [4] . These rates are high compared with the Senegalese study with a rate of 17.8% [17] . The difference could be explained by the fact that 4.6% of their patients had permanent vascular access, whereas 97.3% of patients in our study were dialyzed using a catheter. The use of central venous catheters entails infectious and cardiovascular risks. Observations have shown that the survival of patients dialyzed in a planned manner using an AVF is better than that of patients dialyzed urgently on a catheter [20] [21] .

Some potential factors linked to the risk of death were evaluated in our patients at the time of dialysis initiation. In a multivariate analysis using a logistic regression model, uraemic coma increased the risk of death by a factor of 10, and by a factor of 6 for serum azotemia > 30 mmol/l ( Table 8 ). In Burkina Faso in 2021, Ilboudo CS et al. [26] . also noted that an altered state of consciousness before the hemodialysis session was associated with the risk of death (P = 0.02).

Aside from the complications of end-stage CKD, patients’ low socio-economic status was an important factor in increasing their risk of mortality. This finding is supported by data from an American meta-analysis among undocumented immigrants showing that 3-year mortality was higher with hemodialysis versus standard hemodialysis [3] .

Our study, although innovative in a region of Mali, had certain shortcomings. Our study and follow-up periods were relatively short compared with studies in the literature. This may underestimate the prevalence of EH and the mortality rate. Further work is needed to determine the long-term fate of emergency hemodialysis patients. The predominance of CKD and the absence of pre-existing permanent vascular access demonstrate our difficulty in following good medical practice recommendations. More frequent monitoring of glomerular filtration rate is needed to enable earlier recognition of rapidly progressing renal disease.

Apart from the difficulty of following recommendations, the management of CKD in Mali faces other problems. Kidney transplants can only be performed abroad. Although hemodialysis is free of charge, complementary examinations and parallel care are the patient’s responsibility, and are often not honored for financial reasons. Other problems arise from the lack of equipment in hospitals. The majority of care services for chronic kidney disease are concentrated in the capital, penalizing the majority of the Malian population.

This study highlighted the fact that emergency dialysis patients were unprepared for dialysis and arrived in a serious condition. Mortality was high in the first month, and prognosis was influenced by their condition on admission. Efforts must be made to increase the proportion of patients undergoing dialysis in a planned manner, with mass education, promotion of geographical accessibility of hemodialysis and insurance coverage for the proper management of patients with kidney disease.

Emergency Hemodialysis, Kayes, Mali.

Survey form:

Full name: ___________________________

Socio-demographic data:

Age: /##/

Sex: /#/

Service of origin: /#/ 1 = Emergency 2 = City clinic 3 = Other: ________

MDC: /#/ 1 = Hypercreat 2 = Renal insufficiency on echo 3 = Other: ________

ATCD:

HTA: /#/ 1 = Yes 2 = No

Diabetes: /#/ 1 = Yes 2 = No

Symptoms:

Asthenia: /#/ 1 = Yes 2 = No

Headache: /#/ 1 = Yes 2 = no

Tinnitus: /#/ 1 = Yes 2 = no

Phosphene: /#/ 1 = Yes 2 = no

Vertigo: /#/ 1 = Yes 2 = no

Exertional dyspnea: /#/ 1 = Yes 2 = no

Anorexia: /#/ 1 = Yes 2 = no

Nausea: /#/ 1 = Yes 2 = no

Vomiting: /#/ 1 = Yes 2 = no

Clinical parameters on admission:

PAD/###/NOT/###/

If hypertension WHO grade: ____________

Mucocutaneous pallor: /#/ 1 = Yes 2 = no

Heart murmurs: /#/ 1 = Yes 2 = No

Crackles: /#/ 1 = Yes 2 = No

Pericardial friction: /#/ 1 = Yes 2 = No

IMO: /#/ 1 = Yes 2 = No

Ascites: /#/ 1 = Yes 2 = No

Dehydration folds: /#/ 1 = Yes 2 = No

Uremic frostbite : /#/ 1 = Yes 2 = No

Uremic encephalopathy: /#/ 1 = Yes 2 = No

Uremic comma: /#/ 1 = Yes 2 = No

Diuresis: /#/ 1 = oliguria less than 500 ml/24 H 2 = diuresis anuria less than 100 ml/24 H 3 = Preserved diuresis

Complementary tests on admission:

Hb: /#/ 1 = high 2 = normal 3 = diminished Val ___________

Créat: /#/ 1 = high 2 = normal 3 = diminished Val _________

Urea: /#/ 1 = high 2 = normal 3 = diminished Val _________

Uric acid: /#/ 1 = high 2 = normal 3 = diminished Val _________

Natraemia: /#/ 1 = high 2 = normal 3 = diminished Val ___________

Kalemia: /#/ 1 = high 2 = normal 3 = diminished Val _________

Calcemia: /#/ 1 = high 2 = normal 3 = diminished Val _________

Phosphorus: /#/ 1 = high 2 = normal 3 = diminished Val _________

VitD: /#/ 1 = high 2 = normal 3 = diminished Val _________

24 H proteinuria (g/24 H): /#/ 1 = minimal (less than 1 g/24 H) 2 = medium (1 − 3 g/24 H) 3 = massive (more than 3 g/24 H)

Cytobacteriogical study of urine:

Leukocyturia (10000/ml): /#/ 1 = yes 2 = no

Hematuria (10/mm 3): /#/ 1 = yes 2 = no

Urinary tract infection: /#/ 1 = yes 2 = no If yes germ ____________

Renal ultrasound:

Kidney size: /#/ 1 = normal (100 - 130) 2 = diminished (inf 100) 3 = increased (sup 130)

Differentiation: /#/ 1 = good 2 = poor

Other to specify: __________________

Fundus: /#/ 1 = normal 2 = RH stage1 3 = RH stage2 4 = RH stage3 5 = RD

LVH (ECG): /#/ 1 = Yes 2 = No

Type of IR: /#/ 1 = IRA 2 = IRC

CKD Initial kidney disease: /#/

AKI Initial kidney disease: /#/

Indication for dialysis: /#/ 1 = Uremic syndrome 2 = Pericardial friction 3 = Anuria longer than 48 H 4 = Hyperkalemia 5 = OAP 6 = Metabolic acidosis 7 = Refractory fluid retention 8 = Uremic encephalopathy 9 = Uremic coma

Vascular approach: /#/ 1 = KTFD 2 = KTFG 3 = KJD 4 = KTJD

General evolution: /#/ 1 = favorable 2 = deceased 3 = stopped against medical advice

Progress on dialysis: /#/ 1 = Chronic dialysis 2 = HD withdrawal 3 = Temporary withdrawal