Debating Ideas is a new section that aims to reflect the values and editorial ethos of the African Arguments book series, publishing engaged, often radical, scholarship, original and activist writing from within the African continent and beyond. It will offer debates and engagements, contexts and controversies, and reviews and responses flowing from the African Arguments books.
Contact tracing by public health officials in SA
In the absence of a vaccine, the main tool for control of the current pandemic of Covid-19 (caused by the SARS Cov-2 virus) is human behavioural change. Social scientists are not fully agreed on what determines behavioural change, but there is a broad consensus that individual agency is influenced by social factors. It matters what your family, friends and neighbours think.
So why haven’t social factors been more thoroughly scrutinised in the huge upsurge of scientific effort to combat Covid-19?
One answer to this question concerns the sociology of science itself. Some decades back the anthropologist Mary Douglas shone a torchlight on the social organization of institutions, including the institutions of science, in How Institutions Think (1986). The book focused attention on the need to understand how collective rules imposed their own constraints on sharing of information, shaping who knew what, and what information was considered more or less salient.
At a loss to cope with a new and lethal virus, many governments have argued that they were “following the science”. But that begs the important questions asked by Douglas. How is the science that they are following socially constituted to offer advice? Which of many groups of scientists are invited to the table and why?
The selection of epidemic advice is far from being a random sample of all secure empirical knowledge on the topic.
In Britain, a committee called SAGE (Science Advisory Group on Emergencies) is the main conduit for scientific advice to the government on Covid-19. At first its membership was secret. More recently, names have been published, and it appears that there is a strong representation of biomedical-sciences and epidemiological modellers.
Social scientists are not unrepresented, but it seems that those who seek to modify behaviour (often experimentally) are more numerous than those whose main emphasis is on understanding behaviour through social observation.
Observational information on social behaviour was pushed further down the ladder by a decision made by the British government, early in the pandemic, to abandon contact tracing. It seems to have been thought that once the spread of Covid-19 became general in the community, following up and isolating the contacts of known cases was too difficult, and no longer of much value. The effect was to cut off a flow of potentially important information on who was becoming infected and why.
Reliance was placed on lockdown and social distancing. Neither are very satisfactory ways of imposing social control over infection. Lockdown is too blunt a tool. It cuts down too many activities, and the loss of these activities has damaging effects that eventually cause as much suffering or loss of life as the control measure itself. Social distancing is too imprecise – it is not specific in distinguishing between physical distancing and control over frequency and types of social contacts.
Local Market in Arusha- Tanzania
It would be better if we knew more exactly the contribution of various kinds of social and economic activities to the actual infection risk. For this we need more precise data, and experience with Ebola in West Africa shows that contact tracing was essential to gathering these data.
In the Ebola epidemic in Sierra Leone in 2014 contact tracing was introduced early. Initially, it was done largely by communities themselves, under their chiefs and other local authorities, as part of a civil defence response addressing movement of strangers, or cases of hidden sickness. The disease was strange and devastating and communities quickly decided they needed to protect themselves by finding where the cases were.
Super-spreader events were quickly recognised. This terrible new vomiting fever occurred in clusters of patients treated by well-known herbalists who had also sickened and died, or among afflicted mourners who had attended funerals of persons killed by the disease. Subsequently, it began to be understood that the disease was a family sickness. Those who cared for the sick – partners, or children caring for parents – were prominent among those who were next to show the symptoms.
Ebola Check points and information centres in Sierra Leone
Mende-speakers in southern and eastern Sierra Leone then found a word for Ebola that exactly summed up their observations – bonda wotei (“family turn back”).
The international community stepped in to scale up a national Ebola response. It organised and properly funded contact tracing. Many young people with basic education found employment and a way of contributing to ending the epidemic by signing on for work following infection chains to their roots. This activity was vital in identifying persons at risk and quarantining them before they infected others.
It is worth adding that this surge in contract tracing also hugely improved the accuracy of disease modelling. In August 2014 it was being forecast that by March of the following year there might be as many as 1.4 million cases of Ebola in the Upper West African region. Once the family clustering effect was properly modelled the predictions tumbled. The eventual total number of cases was somewhere below 30,000.
In a recent paper, Susannah Mayhew, Gelejimah Mokuwa, Ahmed Vandi and I have looked again at some of these Ebola infection pathways (Richards et al. 2020).
We begin by reviewing several large-scale numerical studies undertaken using data on laboratory-confirmed cases. Some of these studies speculate about the part played by trade, ethnicity or cultural factors in spreading infection – but none is able to discern evidence for these speculations in the numerical data. The summary is always that analysis of Ebola numbers reveals heterogeneities – a fine word but not much of an explanation!
Our own paper takes a different angle. We revert to a contact tracing approach. Starting with the index case in each of a series of infection chains we trace how subsequent spread took place – who was involved, and with what outcome – using qualitative social science methods of in-depth interviews and careful contextualization.
This allowed us to glimpse underlying explanatory factors – the desperate search for local cures as information spread via international radio stations about a disease with no cure, the dogged determination of families not to abandon the sick and dying, their equally moving if foolhardy persistence in ensuring their loved ones were decently buried, and the accidents and confusions with testing that allowed some people to believe they were free of infection when they were not.
Researchers in our group have also been able to match information about changing attitudes to the disease over time with improvements in disease response (Mokuwa and Maat 2020). The introduction of “respectful burial” improved compliance with regulations on safe burial. Localised Ebola care centres helped convince communities that testing and treatment were worthwhile options.
So why hasn’t the same approach been enthusiastically embraced to deal with Covid-19?
Part of the answer is undoubtedly that the two diseases are different. Covid-19 spreads more rapidly to larger numbers of people, and some of those infected show no symptoms, or infect others before they know they are sick. This makes contact tracing more difficult, not least because it requires rapid and accurate testing.
Where the approach has been tried, however, the benefits are immediately obvious. A new study from Germany looking at patterns of Covid-19 infection based on community-wide testing came up with two surprising facts (Streeck et al. 2020). Much of the infection had been driven by street celebrations associated with Carnival. On the other hand, levels of intra-household infection (the big problem with Ebola) were lower than expected.
This is but one study, even though based on a large community sample, and not yet peer-reviewed, but it suggests a different approach. Contact tracing and testing should be able not only to isolate infectious people in the population but also to throw light on the relative risks associated with different kinds of social interactions.
Even if not directly applicable to Africa, these results provide encouragement for those arguing that the trace-and-test approach is what is needed, to provide a more fine-tuned precision to lockdown and distancing initiatives.
For Africa, a question concerns whether there is enough laboratory capacity to support the contact tracing approach. Almost certainly the answer for many countries is “no”. The symptoms of Covid-19 are various and only a test can identify the disease definitively. But to reject the approach on the grounds of limited laboratory capacity may miss the point.
Cases occur in clusters. At a time of global pandemic clusters of respiratory sickness are likely to include Covid-19 cases, even if some are suffering from other respiratory diseases (available testing could be used to confirm that at least some cases were Covid-19 and not something else entirely – this approach is currently being tried in Ghana).
Contact tracing data on why and where people thought they had contracted their sickness would still be valuable as a way of gaining a sense of which activities are the new super-spreader events – perhaps crowded urban markets, for example, rather than funerals.
At the moment there is no specific information on the hierarchy of risks, and everything is threatened with stand-still. A strongly implemented contact tracing approach designed to throw light on causation and infection risk would show where to concentrate response efforts.
Scientific advisory committees in Africa currently appear to reproduce some of the institutional biases of former colonial powers or prominent aid donors, especially in regard to the weighting given to the bio-medical versus observational social sciences. A certain rigidity of response is thereby apparent. Africa needs to cut its own path. The observational social sciences have so far been underutilised and need to be heard more in formulating national strategy. Even more to the point, Ebola taught that the people themselves are the best observers of the social realities behind epidemic disease. Contact tracing should be organised as mass observation, and through that the people will be given a voice to speak about the pandemic and how to resist it.
References
Mokuwa, E. and Maat, H., 2020,‘Rural populations exposed to Ebola Virus Disease respond positively to localised case handling: evidence from Sierra Leone’, PLoS Negl Trop Dis14(1): e0007666. https://doi.org/10.1371/journal.pntd.0007666.
Richards, P., Mokuwa, G., Vandi, A. and Mayhew, S.G., 2020, ‘Re-analysing Ebola spread in Sierra Leone: the importance of local social dynamics’ (pre-print).
Streeck, H. et al., 2020, ‘Infection fatality rate of SARS-CoV-2 infection in a German community with a super-spreading event’ (pre-print).
