Our study of 385 children with SARS-CoV-2 related disease in hospitalised children aged 0–19 years in Addis Ababa, found high levels of severity and long length of stay. Almost 10% of children died. MISC cases were almost 5% of the cohort. Children with pre-existing conditions and comorbidities had the highest risk of severe disease. No association was found with malnutrition.
Clinical features of SARS CoV 2 related disease (fever, LRTI, GIT, rash, cutaneous inflammatory signs, coagulopathy, hyposmia) were similar to other studies globally and in Africa.[4, 13] Severity was also similar to studies from Africa and in Ethiopia.[9–11, 13] Duration of symptoms prior to presentation was 4 days indicating some delay in presentation. The mortality rate of 9.4% is similar to other African studies,[9–11, 13] and also highlights the importance of SARS CoV 2 disease for children and the need for high quality care of children in future pandemics.
The proportion of children with SARS CoV2 infection who had MISC (4%) was similar to other studies.[4, 5, 12, 13, 20–22] No children were diagnosed with MISC In 2020 indicating a delay in recognition of this severe disease in the hospitals. No children were aged > 15 years which requires further investigation as MISC Is well known to be an important problem in this age group.[13, 22] 30% of children with MISC died and the length of hospital stay was over 1 month (33 days) indicating the severity of MISC disease in our study population. The duration of symptoms prior to presentation 10 days which indicates a concerning delay in presentation which may have contributed to the poor outcomes in the MISC children.
In the children with MISC all had GIT features, most had hypotension (79%), and half had myocardial dysfunction (50%). These findings are similar to other studies.[4, 5] However, only five children had rash or bilateral non-purulent conjunctivitis or mucocutaneous inflammation signs (oral, hands or feet) (5/19, 26.3%) which requires further investigation.
Twenty percent of children with SARS CoV 2 disease and 70% of children with MISC received corticosteroids (either intravenous or inhaled). After three large MISC treatment studies were completed in 2021,[12, 20, 21] WHO MISC guidelines for use of steroids in all children with MISC were introduced in Nov 2021.[2] No IVIG or immunomodulators were used in study area, however there are no currently guidelines that recommend their use in children with SARS CoV2 related disease.[2] All children with hypoxaemia received oxygen and all children with severe hypotension received vasopressors in our study, indicating that quality of care was well maintained in the hospitals.
Children with pre-existing illness or comorbidities had 40% greater risk of severe disease than children without these conditions, findings that were similar to other studies.[4, 13] Our apriori hypothesis was that children with both under and over nutrition would also have a greater risk. This effect was seen in our univariable but not in our multivariable models, however we are aware that our sample size was limited. Pooling data with the companion studies is likely to provide more robust estimates. Children with obesity are well known to have increased risk of severity of SARS CoV 2 related disease and MISC. [4, 5, 13]To our knowledge to date there are no other studies that have examined the effect of undernutrition on risk of severity in SARS CoV 2 related disease. No other factors such as age or male sex were associated with severity in our study but these analyses were also limited by sample size.
We also noted that severe disease was more prevalent in the second year of the pandemic. However, we consider that this was not due to increased pathogenicity of the virus but that it was due to a change in care seeking and admission practices observed throughout Addis Ababa. In the early phases of the pandemic children with mild disease were admitted to hospital as it was unclear how their condition would evolve. By the second year only children with more severe disease who required usual hospital therapies such as oxygen and IV fluids.
Our study had some limitations. In our hospitals tests such as echocardiograms and coagulation tests such as D dimers were not done which reduced our ability to diagnose MISC disease. However basic tests such as ECGs, cardiac enzymes, INR and APTT and platelet counts were available. Hospital physicians also received careful training in diagnosis and management which assisted our case ascertainment rates. Our data was collected by busy hospital clinicians. However, all physicians were well trained and there were few missing data especially on mortality and severity variables.
We intentionally used a broad definition of SARS CoV2 disease which included contact or exposure to a COVID-19 positive person as well as positive pathogen testing. This was because access to testing was poor in our hospitals especially in the first phases of the pandemic and we did not want to restrict our study only to infants who had access to pathogen testing. Our results were similar when children who had negative testing were excluded (data not shown). IgM and IgG serology may indicate past infection or vaccine use.[2, 23] However no children had received vaccines in our study. Results were also similar when children who had positive serology were excluded (data not shown).
Our hospitals were government public hospitals, and we consider that they were representative of general care of children in Addis Ababa during the pandemic. Sociodemographic characteristics (i.e., age, education, underlying illness, malnutrition) distribution was typical of Addis Ababa general hospital care and also children admitted to hospital during the pandemic. The pandemic waves seen in our study were also similar to the Ethiopian data reported by WHO.[3, 24] Other strengths were our use of standard WHO case definitions for both severe disease and MISC.[1, 15, 16] and our a standard protocol harmonised with three other study sites.[14] We consider that our study is the largest published to date of severe SARS CoV-2 disease in low income African countries. We also plan to pool our data with the other sites in South Africa, India and Pakistan to increase the precision and power of study estimates.