Out of a population of 333 patients hospitalized at the Epidemic Treatment Center (CTE) for COVID-19, 203 subjects had diabetes mellitus, i.e. a diabetes prevalence of 60.96%. The average age of our study population was 60.18 ± 12 years with extremes of 30 years and 100 years. The predominance was female with a sex ratio of 0.70.

Diabetes was first discovered in 32 patients (15.76%) and old in 171 patients (84.23%). Only one patient had known type 1 diabetes. Mean blood glucose on admission was 3.07 ± 1.4 g/l with extremes of 0.94 and 6 g/l. Diabetes was poorly balanced in 71.43% of cases (145 patients including a simple hyperglycemic imbalance in 61.08% (n = 124), ketosis in 5.41% (11 cases), acid-acid decompensation ketosis in 4.92% (10 cases).

Cardiovascular risk factors associated with diabetes were found in 76.35% (155). These were arterial hypertension (80 cases, 39.40%), age (121 cases, 59.60%), obesity (20 cases, 8.85%), heart disease (6 cases, 2.95%), obliterating arteriopathy of the lower limbs (3 cases, 1.47%), dyslipidemia, chronic kidney disease, smoking in 1 case each. Two cases of cerebrovascular accidents were noted (0.98%). Table 1 shows the distribution according to epidemiological profile (Table 1).

Clinically, it was mainly asthenia in 90.14%, dyspnea in 78.81%, cough in 52.21%, fever (≥38˚C) in 43.84%, and among them 6.4% had chills, chest pain in 26.60%. So-called specific signs such as ageusia, anosmia or odynophagia were found in 1.97%, 5.91%, 5.41% respectively. Oxygen saturation was <90% in 35.46% (72 cases). The main clinical signs associated with COVID-19 are shown in Table 2.

Biologically, the RT-PCR was performed in all patients was positive in 86.21% (175 cases). The D-dimer level, achieved in 151 patient cases, was greater than or equal to 1000 ng/l in 71.52% (108 cases). C-reactive protein was ≥100 mg/l in 32.01% (65 cases). The other biological abnormalities were anemia in 10.34% (21 cases), hepatic cytolysis in 14.28% (29 cases), hypokalemia in 27.09% (55 cases), kidney failurein 10.34% (21 cases) and electrocardiographic abnormalities in 17.24% of cases (35 cases).

Chest computed tomography, performed in all patients, was pathological in 94.08% (191 cases). Lung involvement was less than 25% (weak to moderate) in 55.17% of cases (112 cases), between 25% and 75% lung involvement (significant to severe) in 21.67% of cases (44 cases). Achievement was greater than 75% (critical lesions) in 17.24% (35 cases). A CT lesion of pulmonary embolism was found in 1.97% (4 cases). Table 3 shows the distribution of additional signs in our patients.

The antidiabetic treatment was noted in 141 patients (69.46%) at the entrance to the CTE, and the rest had no drug treatment. Among those who had an anti-diabetic drug treatment, we found: monotherapy in 53.19% (including insulin therapy in 52.21%), dual therapy in 4.43%, and triple therapy in 3.94%. The oral antidiabetics were a biguanide (16.31%), a sulfonylurea (0.71%), a dipeptide peptidase 4 inhibitor (7.09%). Insulin therapy was used in 85.82% (121 cases).

The symptomatic treatment of COVID-19 included: anticoagulants in 99.00%, analgesics mainly paracetamol in 98.95%, corticosteroids in 92.11%, antibiotics in 100%, zinc and hydroxychloroquine in respectively 86.53% and 21.29% of cases. Oxygen therapy was systematic for all patients with a saturation of less than 95%. Among the antibiotics prescribed, they were macrolides in 89.11% (180 cases), beta-lactams in 193 cases (95.07%) (Including a 3rd generation cephalosporin in 52.48% (106 cases) and an association of amoxicillin and clavulanic acid in 43.07% (87 cases), aminoglycosides were used in 11.35%.

The mean length of hospital stay was 6.27 ± 4.5 days. It was <7 days in 29.08% and ≥7 days in 70.92%. The evolution was favorable in 85.71% of patients (174 cases). A transfer to an intensive care unit was noted in 69 patients (33.99%). Death was observed in 14.36% of patients (29 cases). Death in diabetic patients with COVID-19 was statistically correlated with age ≥ 60 years (OR = 3.013 [1.266 - 7.173], p = 0.005), short duration of hospitalization (OR = 3.154 [1.149 - 8.663], p = 0.009), C-reactive protein > 100 mg/l (OR = 6.370 [2.704 - 15.006], p = 0.000), lung saturation < 90% (OR = 6.331 [2.633 - 15.222], p = 0.000), lung CT lesions ≥ 50% (OR = 7.855 [3.367 - 18.322], p = 0.000) (Table 3).

The hospital prevalence of diabetics infected with SARS Cov-2 was 60.96%. This was higher than the data in the literature which varied from 5.3% to 33.9% in China, the United States and Africa [7] [10] [11] [12] [13]. In France, during the first half of 2020, 19,315 people treated pharmacologically for diabetes were hospitalized for COVID-19 in France, i.e. approximately 20% of all people hospitalized for COVID-19 during the period [14].

This prevalence difference could be explained by a concentration bias insofar as the Abass Ndao hospital is a reference center in the management of diabetes. In our study, new cases of diabetes accounted for 15.8%. This incidence remains similar to that reported by a meta-analysis by Sathish et al. [15] where a result of 14.4% was found.

The phenomenon of new-onset diabetes has already been observed with other viral infections and acute illnesses. The precise mechanisms of new-onset diabetes in people with COVID-19 are not fully understood. But it is likely that a number of interrelated processes are involved, including previously undiagnosed diabetes, stress hyperglycemia, steroid-induced hyperglycemia, and the direct or indirect effect of the cytokine storm occasioned by the COVID-19 on pancreatic cells [16].

The sex ratio in our series was in discordance with the Asian, European and African series [10] [11] [15] [17] [18] [19] [20]. In the West, people with diabetes hospitalized for COVID-19 were mainly men (60.3%). On average, they were subjects in their sixties and over, as evidenced by our series and those of Western [9] [14] and African [10] [11] literature. As already reported in the international literature, the main comorbidity in diabetics infected with COVID-19 was arterial hypertension [5] [6] [9] [10] [11] [17] - [22].

Indeed, advanced age leads to an increase in vascular resistance due to the phenomenon of aging of arterial walls. This could be explained by the fact of the advanced age of the patients and the predominance of arterial hypertension in diabetic subjects. However, Chinese data have highlighted the effect of other morbidities such as smoking, chronic obstructive pulmonary disease, cancer, renal failure and coronary artery disease [17].

The clinical signs were represented in our study by a major triad: asthenia (90.14%), dyspnea (78.81%), and cough (52.21%). This same predominant triad was already reported in Guinea [10] in diabetic patients. On the other hand, in the French CORONADO study [8] [9], fever was at the forefront of the clinical manifestations in diabetic patients with an HbA1c < 7.7%. The low representativeness of fever in our cohort was explained by the fact that our population consisted of poorly balanced diabetics where fever is generally absent in an infectious context.

Oxygen saturation was low (<90%) in 27.94% of cases. In an Algerian study [21], a greater frequency of oxygen saturation disorders was reported in 52% of cases. It was correlated with higher mortality. Indeed, the mortality rate in this patient group was 30.53% in our study. The presence of specific signs such as anosmia and ageusia respectively in 5.91% and 1.97% in diabetics would seem to be less than in non-diabetics. These results were different from those reported in Europe where the prevalence of anosmia and ageusia was 14% in the Coronado study [8] [9].

Mean blood sugar was 2.02 ± 4.27 g/l. This poor glycemic balance, although less frequent, was observed in Asia [17] [23] and in Africa [10]. Diabetics are at increased risk of infection due to abnormal humoral and cellular immunity deficiencies. Although the mechanisms that link diabetes and immune disorders are probably multiple and complex, several studies point to the bad process exchange between diabetes and COVID-19 [9] [24]. In addition, the high frequency of severe cases of COVID-19 could also be justified by the high prevalence of type diabetes in the elderly. The elderly are very prone to cardio-metabolic pathologies due to senescence phenomena, which could themselves explain the unfavorable evolution of COVID-19 in diabetic patients [25].

In our study, the RT-PCR test was positive in 86.21%. These results were consistent with those of a study, the CORONADO study, where RT-PCR was positive in more than 85% of cases [9]. False negatives could be explained by an insufficient amount of viral material in the sample, a laboratory error or restrictions on the transport of samples. This justifies that certain studies recommended not to take the results of PCR tests as the only indicator for the diagnosis, treatment and isolation of COVID-19 patients.

The frequency of biological abnormalities reported in this study was similar to that observed in the literature [10] [11] [26] [27]. In an Asian study, diabetic patients infected with COVID-19 had a higher risk of inflammatory responses with elevated serum levels of inflammatory biomarkers such as C-reactive protein, D-dimers. A Chinese study [18] drew attention to three variables that would raise fears of a negative development. These were advanced age, SOFA score and D-dimer level > 1000 ng/ml. However, in our series, only age and C-reactive protein > 100 mg/l were significantly associated with mortality.

Computed tomography occupies a prominent place in the initial diagnosis, especially in the event of negative RT-PCR and for the evaluation of pulmonary lesions [28]. The severity of lung lesions in our study was correlated with literature data [8] [9] [29].

In a Tunisian series [22], moderate to extensive lesions were found in 25% each and severe ones in 36.5%. The most characteristic scan abnormalities of COVID-19 pneumonia were ground glass, multifocal, bilateral areas [30]. The low prevalence of pulmonary embolism was also observed in Tunisia [22] where it was 1.2%.

The therapeutic strategy is different depending on whether you are in Africa, Asia or Europe. In our study, the management protocol for diabetic patients infected with COVID-19 was that of the Senegal national program [31]. It was consistent with most COVID-19 management protocols in most African countries [10] [11] [20].

We recorded 14.36% of deaths. Our results were consistent with those of the literature where mortality varies between 5.4% and 35%. This mortality rate, strongly influenced by the diagnostic strategy and the type of population, was all the higher in populations with comorbidities. Reported mortality among diabetics was 7.3% in China [15] [17] [18], 35% in Italy in subjects of a very high average age [12], 24.2% in France [14], in a population hospitalized in intensive care units and 5.4% in Mali [10].

Death was statistically associated in diabetics with advanced age, severe lung damage, increased C-reactive protein and saturation greater than 90%. In the CORONADO study [8] [9], the severity of COVID-19, and in particular the risk of death, appears to be lower in type 1 diabetic patients compared to type 2 (5.4% versus 10.6%, respectively). On the other hand, the risk of death is identical in the oldest (≥75 years), suggesting that the prognosis of COVID-19 is essentially linked to age rather than to the type of diabetes and associated comorbidities.

In the literature, mortality was statistically associated with advanced age, severe lung disease, elevated C-reactive protein [32], and saturation less than 90%. A multivariate analysis [33] revealed that the risk of death was increased by the following factors: advanced age, multivisceral impairment score, plasma levels of D-dimer > 1000 ng/ml on admission.