The COVID‐19 epidemic in Madagascar: clinical description and laboratory results of the first wave, march‐september 2020

Abstract Background Following the first detection of SARS‐CoV‐2 in passengers arriving from Europe on 19 March 2020, Madagascar took several mitigation measures to limit the spread of the virus in the country. Methods Nasopharyngeal and/or oropharyngeal swabs were collected from travellers to Madagascar, suspected SARS‐CoV‐2 cases and contact of confirmed cases. Swabs were tested at the national reference laboratory using real‐time RT‐PCR. Data collected from patients were entered in an electronic database for subsequent statistical analysis. All distribution of laboratory‐confirmed cases were mapped, and six genomes of viruses were fully sequenced. Results Overall, 26,415 individuals were tested for SARS‐CoV‐2 between 18 March and 18 September 2020, of whom 21.0% (5,553/26,145) returned positive. Among laboratory‐confirmed SARS‐CoV‐2–positive patients, the median age was 39 years (IQR: 28‐52), and 56.6% (3,311/5,553) were asymptomatic at the time of sampling. The probability of testing positive increased with age with the highest adjusted odds ratio of 2.2 [95% CI: 1.9‐2.5] for individuals aged 49 years and more. Viral strains sequenced belong to clades 19A, 20A and 20B indicative of several independent introduction of viruses. Conclusions Our study describes the first wave of the COVID‐19 in Madagascar. Despite early strategies in place Madagascar could not avoid the introduction and spread of the virus. More studies are needed to estimate the true burden of disease and make public health recommendations for a better preparation to another wave.


| INTRODUC TI ON
In December 2019, a new coronavirus later named SARS-CoV-2 emerged in the city of Wuhan (province of Hubei), China, causing deadly pneumonia. 1,2 Since then, this virus has spread worldwide and the World Health Organizations (WHO) declared coronavirus disease 2019 , the disease resulting from SARS-CoV-2 infection, a global pandemic on 11 March 2020. 3 Despite many efforts from countries to contain the spread at the national level, the epidemic is still ongoing in many countries, including those in Africa, although the African epidemic has been somewhat blunted in comparison with European countries and other territories. 4,5 At the time of submission (31 December 2020), COVID-19 has resulted in more than 83 million cases and 1,8 million deaths worldwide. 5 In Africa, the number of cases (2,76) and deaths (65 468) represents a small fraction of the global data. With the exception of anosmia and ageusia in some patients, COVID-19 is non-specific and similar to many other respiratory viruses. 6,7 Therefore, laboratory confirmation is required to positively identify a case.
Madagascar is a large island located in the South-West of the Indian Ocean with an estimated population of about 27 million, most of whom (65%) inhabit rural areas. 8 International connection through air-traffic remains limited with fewer than 50 international flights per week and around 500 000 passengers annually. 8 In order to mitigate the introduction of SARS-CoV-2 to Madagascar from patients arriving from affected countries, the Institut Pasteur de Madagascar established a real-time RT-PCR detection platform in country as early as 29 January 2020, thanks to technical support from the Hong Kong University-Pasteur Research Pole. 9 Following an increasing number of cases in Europe and Asia, one of the regions with high volume of travellers, the Malagasy Government screened all incoming international travellers from 12 to 20 March 2020 and eventually decided to close the country to all air-traffic on 20 March 2020. After the detection of the first SARS-CoV-2 case in Madagascar from an incoming traveller on 19 March 2020, other nonpharmaceutical interventions were adopted, including curfew, stay-athome order, closure of non-essential businesses and social distancing in order to prevent or limit the spread of the virus in the country.
The objective of the current study was to describe the epidemiology of the first epidemic wave of SARS-CoV-2 affecting Madagascar from 18 March to 18 September 2020, and in particular the proportion of asymptomatic positive cases since the national strategy was to test both travellers and contacts regardless of the presence of symptoms ate the time of sampling. tions. 14 Patients that presented with solely anosmia and/or ageusia were also considered as COVID-19 suspected cases. From each suspected case, demographic and clinical information was collected.

| Study subject and specimen collection
-We also received a high number of specimens from public clinics that were opened during the epidemic offering free sampling and screening test. Convenient specimens were also received from public and private institutions.

| Viral detection
Nasopharyngeal and/or oropharyngeal swabs were taken and were placed into viral transport media and transported at 4°C to the Virology Unit (National Influenza Centre) at the Institut Pasteur de Madagascar (IPM). Specimens were stored at 4°C before nucleic acid extraction and real-time RT-PCR processing. Due to the scarcity of reagents available, specimens were tested using different methods upon availability of reagents. Overall, five real-time RT-PCR

| Full Genome sequencing and genomic analysis
Methods for generating full genome sequences from SRAS-CoV-2 strains circulating in Madagascar and subsequent genomic analysis are detailed in Supplementary file.

| Data management and analyses
The data included in the record form accompanying the biological samples were collected and managed using REDCap electronic data capture tools hosted at IPM. 18,19 REDCap (Research Electronic Data Capture) is a secure, web-based software platform designed to support data capture for research studies, providing 1) an intuitive interface for validated data capture; 2) audit trails for tracking data manipulation and export procedures; 3) automated export procedures for seamless data downloads to common statistical packages; and 4) procedures for data integration and interoperability with external sources. In our analyses, all continuous variables are expressed as median with interquartile range (IQR); categorical variables are presented as percentage, subject to a chi-squared test. All statistical analysis was performed in R 20 and at individual level, and P-value < .05 was considered statistically significant.
We carried out a mapping of the geographical distribution of confirmed cases according to the health district where the sample collection originated from.

| Patient Consent Statement
All data used by this study were from state-wide surveillance of a notifiable disease and were de-identified. March 2020. Thereafter, several imported cases from passengers were detected. The first laboratory-confirmed cases without a travel history, therefore considered to be community transmission, were detected on 25 March 2020. Although some cases are still being detected in December 2020, our study focuses on the first six months, or the "first wave," of the pandemic in Madagascar (ie from 18 March to 18 September 2020).

| Clinical symptoms of patients
We found that 75.2% (19,864/26,415) of patients tested declared no symptom at the time of sampling. The proportion of asymptomatic individuals was 56.6% (3,311/5,553) among laboratory-confirmed cases ( Table 1). The most common symptoms of illness onset among confirmed cases were cough (27.2%), fever (18.7%), weakness (14.7%), runny nose (13.3%) and headache (13.1%) (Table S2;   Table 2). In multiple logistic regression, age and the five most common symptoms observed in confirmed cases were associated with SARS-CoV-2 positivity. The probability of having a positive RT-PCR increased with age ( Figure S1). Compared to individuals less than 16 years, individuals aged 16 and above had higher probability to have a positive RT-PCR. The   Although cases were still detected at the end of our study, the positivity rate was below 10% by end of August and the number of daily cases was below 10 by week 38 (14-20 September 2020).

| Monitoring of COVID-19 through the Influenza Surveillance System
A proportion of the overall specimens received during the COVID-19 epidemic were acquired through the extension of the ISS to include SARS-CoV-2. Although our SARI surveillance system only detected a few COVID-19 cases (with very few samples received from May to July and only two SARI-derived SARS-CoV-2-confirmed cases in August and September), the ILI system sourced a substantial number of COVID-19positive samples ( Figure 4). Overall, among ILI suspected cases, 35.0% (205/584) of them were found positive for SARS-CoV-2. The peak positivity rate reached 69.2% (164/237) in July and decreased thereafter.

| Genetic characteristics of the newly introduced SARS-CoV-2 virus in Madagascar
The entire genomes of the 10th and 19th cases of SARS-CoV-2 detected in Madagascar (from 20 and 22 March 2020) were obtained on   (Table S3). In addition to these common mutations, both early sequences also exhibited C-to-T muta-  (Table   S3). In addition, they also show a C-to-T mutation at position 19 524 and a non-synonymous G-to-A mutation at position 1268 (Orf1a-Nsp2: D335N).   September during the dry and cold season in the highlands. [23][24][25] Further sequencing of SARS-CoV-2 isolates will be critical to "tracing" the spread of these two different outbreaks.

| D ISCUSS I ON
Overall, the total number of laboratory-confirmed cases of COVID-19 in Madagascar as of 20 September 2020 (16,020, a third of which were detected in part with this study) remained low per inhabitant, when compared to Europe and the Americas. 26 Within Africa, Madagascar is among the ten countries reporting the highest number of cases of COVID-19 but is still reporting far fewer cases than the northern African countries, as well as South Africa. 26,27 Several reasons could explain this result. First, almost 65% of Madagascar's inhabitants live in rural settings, 8 and the population is, on average, very young (median age = 20.3 years). In our study, SARS-CoV-2 infected patients aged less than 20 years represented only 9.4% of all positive cases. This particularity may have limited the spread of COVID-19 as suggested by the modelling study conducted by Diop et al. 28 Secondly, it is possible that the total number of confirmed cases of COVID-19 in Madagascar is underestimated and/or underreported due to several factors, including (i) the testing capacity of labs that could not exceed 1,000 tests/day, (ii) insufficient staff to conduct efficient contact tracing and (iii) behavioral resistance to healthcare seeking in the population. Limited healthcare seeking behaviour often presents challenges to efforts to estimate the burden of diseases in sub-Saharan and other low-income countries. 29,30 Resistance to seeking health care can have many drivers, but recent studies have shown a reduction in patient presentation in clinics or hospitals during the COVID-19 pandemic and associated lockdown. [31][32][33] An ongoing serological survey among blood donors in Madagascar should be able to address the true burden of COVID-19.
For future monitoring of SARS-CoV-2 circulation, WHO has recently recommended that countries extend the Influenza

F I G U R E 4 Monthly laboratory results from the Severe Acute Surveillance Infection (SARI) and Influenza-Like Illness (ILI) surveillance in
Madagascar from February to September 2020. Each bar represents the total number of negative cases (grey) and SARS-CoV-2-positive cases (red). Numbers above bars indicate the number of positives. The dark blue line represents the positivity rate Surveillance System (ISS) to include  In Madagascar, an effective ISS has been in place for decades and was used effectively to detect and monitor the last pandemic virus A/H1N1pdm09 in Madagascar. 11,12 Although the ISS was disrupted during the first few weeks of the COVID-19 epidemic, due to a lack of personal protective equipment for clinicians and their excessive workload, it was rapidly reinstated and has been used thereafter for effective monitoring of SARS-CoV-2 circulation in the Madagascar community. and in contrast to previously described patterns of respiratory virus circulation in Madagascar. 24,25,39,40 Regarding clinical signs, although symptoms of COVID-19 are considered to be non-specific, the five most common clinical manifestations (fever, cough, weakness, headache and runny nose) were significantly associated with SARS-CoV-2 infection in our study. Ongoing studies (seroprevalence surveys, first few hundred cases and contact analyses, and genomic epidemiology) will support efforts to estimate the burden of disease, various epidemiological parameters (eg R0, clinical attack rate, immune response…), and underreporting of cases and inform public health strategies critical to avoiding or reducing the impact of subsequent waves of infection on the health systems and the economy of a country with limited resources.

ACK N OWLED G EM ENTS
We Biohub, or the Innovative Genomics Institute.

CO N FLI C T O F I NTE R E S T
All authors declare that they have no commercial or other associations that may pose a conflict of interest.