COVID-19 Mitigation: Are Testing And Contact Tracing Effective Enough?
The COVID-19 pandemic has revealed many loopholes in our understanding of viral respiratory disease mitigation. It has also shown what can only be termed as technocratic hubris, whereby measures that are not firmly rooted in science have been applied across the population without conducting the requisite cost-benefit analysis. Most importantly, these measures have been highly inequitable in their formulation, implementation, and impact and most of them correlate poorly with the downward trends in epidemic curves around the world. Instead, they appear to serve as symbolic gestures that give some semblance of agency where we have none.
The current COVID-19 containment measures are doubly problematic as an over emphasis on them has taken the limelight away from time tested pandemic mitigation measures that target the most vulnerable, economies and health systems. Along with prolonged lockdowns, the mass scale contact tracing established has been experimental, highly costly and ill-conceived.
Contact tracing and isolation, ‘Test and Track’ is increasingly implemented as a key strategy for controlling the spread of COVID-19. Many experts claim that a “consensus” exists, that contact tracing and testing are critical mitigation measures. Expert consensus is an important basis for health policy, but is usually relied on only when there is insufficient evidence base. Critically, this view is not consistent with prior WHO and CDC recommendations on contact tracing within a respiratory virus pandemic response. Evidence for these was systematically weighed. The WHO further noted: ‘contact tracing on a large scale can lead to ethical issues such as leakage of information, and inefficient usage of resources, including human resources’
Naturally, the question arises: what is the new evidence base (generated over the last few months) that has proven that contact tracing will mitigate this pandemic? This question is important to answer as we have years of knowledge and experience in this area that states that once community transmission of a respiratory virus has started, tracing becomes futile practice.
Out of the many narratives constructed around SARS-CoV2, a frequently stated one has been that a ‘novel’ virus confronts us. This notion of novelty has resulted in a widescale rejection of well-established principles within pandemic control. While genetically novel, it is no different from other coronaviruses, rhinoviruses, and the influenza virus in terms of mode of transmission. There has been no fundamental change in the evidence base that contact-tracing is only a feasible option in early outbreaks to slow or break transmission (e.g. New Zealand). It appears relatively ineffective in restricting COVID-19 once transmission has become largely widescale (e.g. UK, US). Now we know who is likely to experience adverse outcomes from COVID-19. Tracking and isolating healthy low-risk individuals takes considerable resources away from protecting and managing those with certain high risk.
A Problem with the Strategy
In the conventional sense, contact tracing is employed to use the web of contacts to reach the center, i.e., the index case, also known as backward tracing. For COVID-19, the process is reversed, and an outward (forward tracing) web of contacts is mapped, tested, and if found infected, is quarantined. When the epidemic is widespread, such mapping can quickly reveal that a high proportion of people in a locality could be potential contacts, granted the tracing is done diligently. In reality, it is also virtually impossible to account for all the cases – even in the most advanced economies with sophisticated healthcare systems and exceptional testing capacity. Seroprevalence studies have revealed that actual prevalence is far greater than the number of recorded acute cases. Additionally, it has been argued that even the current seroprevalence is an underestimation.
We cannot detect the majority of the infectious cases in the population to trace them and we cannot test all potentially infected contacts identified as such. So, what is the purpose of contact tracing? An argument has been made for focusing on hotspots i.e. identify pockets where the infection is particularly concentrated. This, however, is akin to trying to stop a wave by hurling pebbles at it since hotspots are partially a function of testing capacity which is limited, especially in developing countries. The determination of one hotspot versus another that is not on the radar then becomes arbitrary when surveillance cannot cater to the entire population. While resources are used where a cluster of positives are identified, other clusters go unnoticed, negating an overall impact as the virus continues to circulate.
A problem with the tool
Lack of population-wide testing is only a small part of a bigger problem. The entire programme of contact tracing, in most countries, hinges on the accuracy of the RT-PCR test, i.e., the determination of a ‘case’. The test has come under significant criticism for its ineffectiveness in both diagnostic and surveillance purposes. For one, the test has a high degree of false negatives ranging from 100 percent before the onset of symptoms to 33 percent. This implies that the test will miss out on a significant number of contacts. Ideally, the aim would be to find the contacts when infection is detectable but not significantly infectious. This would need multiple testings of many contacts to catch them just as they become infected.
False positive results also present a problem. As per CDC’s recommendation beyond the 10th day of symptom onset, it is advised that people should not get tested, and they can return to normal activity. This recommendation is in place because the test picks up inactivated RNA (remnants of the virus) and gives a positive result for up to 50 days even when the individual is no longer infectious or even infected.
Furthermore, false positive results occur intrinsically due to the nature of the test, increasing with the number of ‘cycles’ conducted. While it has been noted that live (infectious virus) is rarely detected after more than 25 cycles, 35 to 40 cycles are frequently being used, ensuring that many results will falsely indicate an infectious case. The rigorous tools needed to determine that a positive result really is someone infected and an infected case is truly infectious are missing from the global mitigation strategies. Thus, this crude measure of the pandemic, i.e., ‘cases’, is a poor proxy for the actual sick or infectious population. Failure to distinguish true from false positive results has contributed to a greater stress for the public and inappropriate policies by governments.
The criticism against population-wide testing does not imply that testing is useless. It has significant use in diagnosis if complemented with clinical analysis. Furthermore, testing can play an important role in protecting the elderly in nursing homes and similar facilities, by ensuring those caring for them to be virus-free.
This unprecedented mass testing program imparts a huge cost on health systems, either increasing debt or taking away funding from other services. The growing burden of healthcare would take a greater toll on health budgets not only in high income but especially the low income countries. Let us take the example of a single week in Pakistan.
From 22nd November to 28th November, approximately 20,000 cases were identified. The number of tests done over the week was about 290,000 tests. The price per test is around $45. It implies that the testing done roughly cost the country $1.3 million. An average household in Pakistan has 7 members in it and a conservative estimate would suggest around 12 contacts per week. So, on top of this we add the cost of testing the contacts. It is important to note that this is just the first level of contact tracing, and we are not looking at contacts of contacts. These costs can substantially increase if ‘cases’ start to increase, whether due to increased baseline testing or due to seasonality. For a sense of the project’s scale, at its peak, Pakistan saw around 40,000 cases per week, and if contact tracing were to be conducted at that point, we would be looking at a weekly cost of $34 million on testing and tracing at the first level of contact tracing. All of these costs pale in comparison to the actual scale of the problem reflected in seroprevalence studies. The entire exercise of contact tracing is rendered moot when confronted with the spread of infection’s real scale.
Another argument has been made that a traditional form of contact tracing and isolation can be done; testing kept to a minimum while contacts are asked to isolate. This would imply that irrespective of whether people have the infection or not, they will be asked to quarantine. Furthermore, all of these measures are based on a test that cannot tell if the index case was infected or infectious. Both from a social and economic perspective, this policy is costly and bound to backfire. This strategy again fails to address the critical issue of the immense gulf between detected ‘cases’ and the true prevalence of SARS-CoV 2.
Contact tracing is a fruitless endeavor unless used during the pandemic’s early phase, and even then, it has a time-bound efficacy. This phase of intervention is so early that it implies that the global pandemic never resulted in a regional/local epidemic. In other words, it may be suitable to stem out a localized outbreak. In this early stage, it is a useful tool only if aggressively implemented. This was done in Australia and New Zealand. However, such contact tracing will make sense if you envision an indefinite relationship with the virus that entails closing borders intermittently and quarantining all those traveling into the country (or till a viable and safe vaccine is available) often at the peril of ignoring other public health and social imperatives. Such a regimen will have to be continued until the virus is expunged from the rest of the world. That is not anticipated to happen as the virus will most likely become endemic, while the vaccines under development will take years to roll out and, as per the trials, do not include reducing transmission as one of their end goals.
Any contact tracing strategy raises issues concerning individual liberty and compliance. Most people are not comfortable revealing all their activities and the locations they have visited. There is also a potential breach of medical confidentiality. This is exacerbated by the conceptualization of contagion in the popular imagination. A more immediate rights issue comes from being quarantined, the right to earn a living – which is particularly acute in low-income families. Further human rights issues emerge if proscriptive or coercive means are used to force people to be tested. Given the test’s low efficacy and multiple lines of evidence suggesting that asymptomatic cases appear not to play a larger role in spreading the virus, on what ethical grounds are we asking these individuals to isolate themselves and potentially lose out on income that is key to their sustenance?
Even with costly population-wide screening, contact tracing that relies on RT-PCR will miss out on a substantial number of asymptomatic cases, whilst falsely identifying many people as infected and infectious. We are not in the early phase of this pandemic. Once community transmission has taken root, going after the virus through contact-tracing is akin to an unfit person running a hundred-metre race against an elite sprinter. The unfit person will eventually catch up but not because of his speed; the race would have been over.
The resources spent on contact tracing, particularly in developing countries with fragile healthcare systems, would be better employed in behavioral change communication and bolstering the health system’s capacity. It is thereby imperative to isolate truly symptomatic and highly infectious cases to protect the relatively vulnerable from infection.
Muhammad Usman Khan is a public health researcher and David Bell is a Public Health consultant.