In response to the rapid global spread of COVID-19 and in presence of local secondary infections, the understanding of the transmission dynamics of the virus is crucial to guide effectively containment measures and interrupt the spread of the virus. In this context, this is the first study documenting transmission chains and clusters of SARS-CoV-2 in Tunisia. This investigation showed how contact tracing and early investigation, upon a case of COVID-19 was confirmed, was crucial to document locally transmission and identify transmission chains. Using these methods, most possible sources of exposure could be identified and most clusters could be documented. This surveillance method is indicated to identify cases in the community and this will be missed if we focus only on symptomatic person and it is the most effective way to enable containment measures in order to stop super spreading events and transmission of the virus (10) .
Analysis of transmission chains at early stage of the epidemic before implementation of control measures highlighted the sustained transmission of the SARS-CoV-2 in Tunisia. In fact, 47% of secondary cases were detected during the first month of the spread the virus in Tunisia (March). This sustained transmission was demonstrated also in a recent publication (13) which concluded to a temporal reproduction number R0 of 3.18, 95% CI [2.73–3.69].
The evolution of secondary cases showed two peaks: the first one occurring on April 2 was probably related to the lack of compliance to at home quarantine among travelers coming from high risk countries. Whereas the second peak, of April 16 could be the consequence of the delayed effect of non-compliance with general confinement measures related to mass gatherings as weddings, meetings, funerals.
The “Superspreaders” identified in this investigation have significantly participated in the spread of the disease. In fact these 9 “Superspreaders” had transmitted the infection to a total of 95 close contact people. Similar events were associated with outbreaks of SARS and MERS (11, 14, 15) and reported recently with SARS-CoV-2 (16).
Concerning the evolution of the index cases, we have noted that imported cases contribute to secondary disease transmission in Tunisia at the early phase of the epidemic. This finding support the impact of government measures put in place such as travel restrictions since March 20 and mandatory quarantine of repatriated population since March 22.
As a consequence of the mounting number of COVD-19 cases in Tunisia and the sustained transmission, the government implemented many control measures from 18 March: curfew, lockdown, travel restrictions, social distancing measures including restriction of public transportation, cancelling of social and mass gathering events, school and university closure, and some professional activities as well as promotion of preventive measures among general population (barrier measures, physical distance, hand washing) and among health workers (Personal Protective Equipment and care hygiene).
In this study the four largest clusters showing a local transmission were analyzed. Transmission of infection in these clusters was probably attributed to close and prolonged contacts with symptomatic case and was occasioned by family gatherings (clusters 1, 2 and 3), a wedding ceremony (cluster1), a conference and meetings (cluster 4). Direct physical contact and handshaking was reported in family gatherings and meetings. In the wedding event (cluster 1), and one of family gatherings (cluster 1) sharing of meals during a ceremonial dinner was also reported.
These findings are similar to those shown in other published case series analyzing transmission chains. In fact, transmission by close contact is known as the most common transmission mode particularly when contact is prolonged (17). Whereas the possibility of indirect transmission can’t be excluded since surface contamination and indirect transmission (via fomites and shared food) has been documented in some studies and this evidence emphasizes the importance to sensitize community to adopt strict personal and hand hygiene as a key preventive measure (18).
We recognized that this study had some limitations and some results should be interpreted carefully. First, as symptom onset dates and close contacts were self-reported, there could have been reporting information bias.
To mitigate these biases we collected information immediately after case confirmation. However, biases concerning the identification of close contacts could not be completely eliminated since this identification is mainly based on the statements of patients and contacts, nevertheless we explained the exact meaning of a “close contact” to all interviewees and we collected all the exposure history in order to define accurately all close contacts.
Second limitation concerned the use of only RT-PCR testing in active case-finding among close contacts. This method is known to offer a rapid diagnostic solution, but it can only detect SARS-CoV-2 during the period of viral shedding which still not certain and RT-PCR is so limited by its ability to detect convalescent cases of COVID-19 (19). The first preliminary analysis (20) of SARS-CoV-2 IgM and IgG indicated that the antibody response in COVID-19 patients is similar to seroconversion kinetics indicated with Middle East respiratory syndrome (3 weeks after symptoms started) (21) and with severe acute respiratory syndrome (93% of patients seroconverted at an average of 20 days from symptom onset) (22), if not earlier than, these times.
Serological testing were demonstrated in a recent study (17), to have a crucial role in identifying convalescent cases or people with milder disease who might have been missed by RT-PCR and helped so to establish connections between COVID-19 clusters.. This will be useful for epidemiological investigation by identifying more accurately infected people in clusters and track transmission dynamics which would better inform disease control policies for more or less containment efforts. It is also crucial to determine more accurately the number of infected patients because this will influence all other epidemiological estimations: attack rates, case fatality rate …
Some difficulties were noted during this first step of outbreak investigation aimed to identify transmission chains of COVD-19. Difficulties concerned mainly the use of traditional epidemiologic methods to establish links between cases and clusters, which were sources of delays and inaccuracy to obtain information from cases and contacts so that the implementation of tracking device that will help to achieve contact tracing in a precise and valid manner is necessary in the next steps of the management of the epidemic.
The other perspectives in the next phases of epidemic management allowing to improve the identification of links between cases and between clusters concern adoption of additional laboratory techniques, such as serological tests and phylogenetic analysis, as well as the acquisition and activation of applications allowing accurate detection of close contacts and quarantine violators, enabling identification new close contacts in a timely manner.