This prospective cohort conducted in symptomatic and suspected patients of Covid-19 has shown that by combining serological and molecular tests, we are able to treat a larger number of patients by considering them to have covid-19; moreover, the risk of dying is better predicted by using the clinical classification in relation to the viral load of SARS-CoV2.
The study period coincides with the peak of the Covid-19 pandemic in the city of Kinshasa. Hospitals experienced a significant flow of patients and several cases of death were reported. In a complex scenario such as the ongoing pandemic, not only the diagnoses need to be timely and accurate, but laboratory testing needs also to provide information to avoid missing cases. By combining RT-PCR and serological tests, 43% of patients who had symptoms and or chest CT suggestive of Covid-19 were treated. With RT-PCR alone, we were going to limit ourselves to 31%. Although being a gold standard method for confirming the SARS-coV2 infection, the limitations of RT-PCR, both real-time and GeneXpert, are known. False-negative results have been reported at rates as high as 20 to 40% in cases for which both clinical symptoms and imaging evidence raised strong suspicions of disease (4, 16). False negatives may be caused by various factors, including the specimen source, timing of sampling, personnel operation, and the test kit quality (4, 16). Instead of waiting for a second RT-PCR result as some recommend (in the event of a first negative test) (17, 18), we believe that this strategy wastes the time to diagnose the disease (1 to 2 days or more), with a very high cost. Even considering the serological window which sometimes limits the value of serological tests, in our situation where patients consult the hospital several days after the onset of symptoms, serological assays can become helpful both to complete the epidemiological link when molecular diagnosis results are negative, and to alleviate the burden of laboratories implicated in molecular diagnosis. The time for seroconversion is about 10–14 days, but early seroconversion has also been documented at 3–5 days post infection (4–7).Table 2 shows that patients in whom the diagnosis of Covid-19 was confirmed were older, consulted later in the hospital and had a more serious clinical picture. The stigmatization of patients, the denial of the disease by a large part of the population and the habit of consulting late in the hospital after self-medication largely explain these results.
Contrary to the Covid-19 WHO clinical classification, the viral load result (expressed by CtE and CtN2 values) had no prognostic value to predict the mortality. Pujadas et al had showed an independent relationship between high viral load and mortality; so transforming qualitative testing into a quantitative measurement of viral load could assist clinicians in risk-stratifying patients and choosing among available therapies and trials (19). Our study may have the disadvantage of being monocentric with a small sample size. To date, the study of the viral load to predict the prognosis of Covid-19 has not been the subject of several publications. Cohort studies with large samples may resolve the issue. Nevertheless, independent of the viral load, several factors, in particular co-morbidities, explain the death in Covid-19 patients. In a recent systematic review and meta-analysis, Lu et al had showed that advanced age as well as comorbidities and laboratory indicators including lactate dehydrogenase, C-reactive protein, neutrophil, Blood urea nitrogen and albumin are correlated with Covid-19 mortality (20).
Among the risk factors studied, the clinical stages of the disease (severe and critical Covid-19) as well as the time between the first symptoms and admission to hospital were retained in the Cox model. Even if the risk linked to the stage of the disease can be explained, the probability of dying, which is multiplied by 26, shows how the management of severe forms of Covid-19 remains a major concern. In countries with low resources and where the health system is fragile, such as the DR Congo, efforts must focus on preventive measures that require compliance with barrier measures. The shorter hospital admission time for deceased patients reflects the fact that in our environment, individuals come to the hospital when the situation is serious. Likewise, it is possible that the lethal strains of the virus also have a shorter latency time; which may explain a faster hospital consultation. A limitation that must be recognized is the fact that in severe patients, the information was obtained by hetero-history. The exact date of onset of symptoms may be wrong. Although the patients who died were relatively older, this did not emerge as an independent risk factor in statistical analyzes. We have already mentioned the limit related to the size of the sample.