To interrupt onward COVID-19 transmission it is essential to swiftly identify cases with prompt identification of their exposed (high risk) contacts and the application of appropriate measures by cases and contacts. As the generation interval (the delay between an index case becoming infected and passing on the infection to a secondary case) of COVID-19 has been reported to be on average around 4-5 days [20, 21] the whole process of contact tracing should be as quick as possible. Additionally, it has been shown that cases become infectious before symptom onset, further reducing the time window to act if cases are identified after symptom onset [22–25].
The average delay between symptom onset and contact tracing initiation was around 5 days. This delay remained stable over the study period despite an increase in cases in March - April 2021. The longest delay is the time between the onset of symptoms and the consultation of a healthcare professional. The duration between the onset of symptoms and the testing by a healthcare professional is 3-4 days. This is slightly higher than the 2.8 days delay reported by France  but shorter than the 10.2 days by Brazil . However, data is not available for many countries and comparison is difficult due to differences in the organization of the health care system and the epidemical situation when these studies were carried out.
Modeling studies indicate 4–5 days between symptom onset of the index case and contact tracing initiation as an important tipping point for contact tracing effectiveness [28, 29]. Although efforts have been made to speed up the testing and tracing process to its maximum further improvements are essential, in particular, shortening the time between symptom onset and a positive test result. Frequent and clear communication towards the general population about the importance of rapid testing is needed to stimulate the public to get tested as soon as possible when COVID-19 symptoms occur and respect immediate isolation awaiting the test result. Also, the threshold for testing needs to be kept as low as possible. Early November 2021, an online self-assessment questionnaire was implemented in Belgium allowing persons with mild symptoms to be tested without consultation . This new tool could help to improve the speed of identification of cases and may also help to decrease the burden on traditional healthcare services by creating an alternative test circuit. Such solutions may need further exploration and assessment, not at least with variants having an even shorter incubation period  and considering the upcoming phase of transition towards more sustainable long-term prevention, control, and surveillance strategies for SARS-CoV-2.
More than 9 out of 10 index cases were successfully contacted and most were contacted on the day of diagnosis. Similar results were obtained for HRC: more than 9 out of 10 were contacted and 3 out of 4 were contacted the same day as the index case. This shows the high coverage and operational efficiency of the contact tracing system. In addition, field agents could be deployed to visit at home index cases that could not be contacted by the contact center (not included in the results). Although the study period included the third wave, the results remained stable despite a strong dependency on human resources to conduct interviews. The system however remains susceptible to facing high volumes of cases at a short time, especially when rapid increases happen (“waves”). This can overload the system, and calling of all cases and HRC will become unfeasible. Less in-depth, more automated contact tracing techniques were prepared (e.g. online self-reporting of HRC by index cases, inform HRC by SMS). The performance of such automated systems remains to be evaluated, but the lack of a human aspect might negatively affect contact tracing effectiveness.
Among the contacted index cases 72% reported symptoms. The proportion of asymptomatic cases (28%) is rather high compared to the range reported in other studies (17-25%) . However, our study lacks a follow-up of asymptomatic cases causing an overestimation of the asymptomatic fraction . Interestingly, the symptomatic proportion remained stable over the entire study period, despite different predominant lineages over time and the roll-out of the vaccination campaign in the general population in 2021. Cough, headache, and a runny nose were the symptoms most often reported. Anosmia was reported by 18% of the cases, which is at the lower end of the frequency reported in the literature .
Among the index cases that answered the questions on the Coronalert app, 28% reported that they had installed the Coronalert app, and 43% of the users reported that they had used the app to alert contacts after the positive test. These results are in line with the general statistics reported in Belgium on the app use. Digital proximity tracing via apps is a novel and promising measure to reduce the spread of COVID-19, with the potential to complement regular contact tracing and enhance contact tracing effectiveness [29, 35]. However, a majority of citizens need to be willing to install and use such an app to be effective. The proposed uptake threshold of 60% of the population remained unattainable in most countries [36–38]. Another limitation of the system is that the context (e.g. face-mask-wearing) of the possible exposure could not be taken into consideration in the system’s current form.
Among the index cases, only 49% reported HRC. This proportion might signal underreporting, but it should be taken into account that it is impacted by the index cases in collectivities (e.g. school, kindergarten, nursing home) and companies for which local contact tracing is carried out for their contacts within that collectivity, which are not reported to the central contact center. Among the index cases that reported high-risk contacts, the mean number of HRC was around 2.7 and this number remained stable between January and September 2021. This appears to be low, especially taking into account the lifting of most nonpharmaceutical interventions impacting social contacts since June 2021 . On the other hand, several measures that could reduce the number of HRCs remained mandatory, e.g. mask-wearing in public transport and shops. The number of identified contacts in other countries varies widely with only 1.15 in the US, 1.4 in the UK, and more than 17 in Taiwan [36, 40, 41]. Differences in nonpharmaceutical interventions, the definition of an HRC, the inclusion of low-risk contacts in the tracing system, coverage of the system, and compliance of the population all impact these numbers. It is essential that index cases report all their contacts to maximize the effect of contact tracing activities. Sensitization of the general population would be needed to improve the awareness about the importance of reporting all contacts and to find the willingness and trust to do so.
Our results support that contact tracing is a useful targeted public health tool that effectively results in finding cases and presumably in an earlier phase than compared to a system that would only rely on symptom-based testing and case finding. Overall the proportion of COVID-19 cases that were previously identified as HRC was steady around 24%. This proportion may be an underestimation, partly due to some HRC not reporting their NRN (14%) as well as due to contact tracing in companies and collectivities not included in the central tracing system. The latter may also have contributed to a possible overestimation of household transmission. Transmission at home was most often reported as the suspected place of infection by the index cases, 2 out of 3 reported HRC where household members and the SAR among household HRC was with 34% double as high as among non-household HRC.
Testing and quarantine of HRC are crucial measures to reduce the risk of transmission and to identify newly exposed HRC whenever an HRC becomes an index case. Overall 89% of the HRC got tested at least once. We reported a high overall SAR of 27%, which is partly thanks to the double testing strategy of the HRC, as many cases were only identified during their second test. On the other hand, not all HRC were reported, leading to an overrepresentation of household members, who have a higher positivity rate. Nevertheless, even with a SAR of 16% among non-household HRC, our results suggest that contact tracing is an effective way of case finding as the reported HRC (with NRN) represented only 6% of all tests in Belgium but 24% of all new cases.
Data of the adult population between January and September 2021 also illustrates that vaccinated and unvaccinated people complied with the testing strategy as respectively 92% vs 85% were tested at least once. All HRC received an officially recognized quarantine certificate for the quarantine period, with information about the mandatory and recommended measures but the follow-up of these measures is more difficult to evaluate.
In Belgium, the first vaccines were delivered on 28 December 2020. The campaign continued, in different phases. By the end of September 2021, 73% of the general population was fully vaccinated . In contrast, only 36% of index cases were fully vaccinated at that time. The over-representation of unvaccinated persons among index cases is in line with the previously described protection of vaccination against infection among the Belgium population . The proportion of HRC fully vaccinated was also lower than the proportion of fully vaccinated persons in the general population. This difference may be explained by the fact that the largest proportion of the unvaccinated population by the end of the study period were children and adolescents, who in general had a higher incidence when vaccination increased in the general adult population, causing a higher number of high-risk contacts in unvaccinated, compared to the oldest age groups who have the highest vaccination coverages. Furthermore, there might be clustering due to household members or friends who may tend to have a similar vaccination status. SAR among unvaccinated HRC are two times higher compared to vaccinated HRC, illustrating the protection of vaccination against infection during the first nine months of 2021.