A plan for a test strategy | Analysis – analysis
India has been stuck for a month and finally there is a plethora of test kits, although their quality is apparently a real concern. Locks are the best time to test as movement and new contacts are minimal. While it may be necessary to adapt testing strategies to the local context of states, are there common principles that should govern when, where, how, and to whom to evaluate?
Since mid-March, testing has increased 30-fold, to around 30,000 tests daily. But even if a million additional tests were done, it would equal one test per 1,000, which would keep India a low-test country. Therefore, even without quality problems, efficient testing strategies are needed to quickly learn and achieve the two key goals of testing: first, find infected people, even if they are very few and asymptotic, and treat them and prevent transmission; and second, generating data to implement a smart containment strategy, reducing the need for wholesale locks.
Should states start testing at hot spots? If people test positive and are asymptomatic, they will be isolated and hospitalized if symptoms develop. Ideally, tests should change the action, but, at a critical point, they are quarantined anyway.
But, access point tests can be informative. If the positivity is less than expected, or grouped, it provides information about the transmission; if it is higher, it indicates an increase in the demand for hospitalization. This information is best obtained by antibody testing, if accurate, but learning is only possible with adequate randomization, that is, testing according to a predetermined statistical plan. Within the access point, further testing further from the nucleus, where less infection is expected, may indicate the extent of the spread. Importantly, it can also help decide when an area is no longer an access point.
But focusing only on the current hot spots cannot find infected people in areas where the virus has not yet been reported. Can tests prevent future hot spots by locating asymptomatic infections?
The current testing strategy recommended by the Indian Council of Medical Research (ICMR) only allows evaluation of asymptomatic individuals if they are direct, high-risk contacts of a confirmed case. But, ICMR also says that about 69% of confirmed patients with coronavirus disease (Covid-19) are asymptomatic. Considering that surveillance of patients with severe acute respiratory infections (SARI) indicated a positivity of more than 2%, we can miss many infected people. If even 0.1% is Covid positive and asymptomatic, 20,000 infected people could be infecting others in Delhi alone, as the blockage is alleviated.
By going beyond ICMR recommendations and analyzing a sample of the population, you will likely find few people infected. One of us calculated that if 0.1% of 10 million are infected and 1,000 tests are randomized, the chance of not even finding a positive case is more than 40%! Therefore, an efficient test design should maximize the chances of finding an infection, especially those most vulnerable to the disease.
This can be done in three ways.
First, assess asymptomatic people who may become “super spreaders,” that is, those who are susceptible to infection, but who frequently interact with others, including during blockades, such as health and police workers, but also civic workers. in essential services and street vendors. Many of these groups were found to be infected after they became symptomatic. These groups can be evaluated at work, using reverse transcription polymerase chain reaction (RT-PCR) methods, with pooled samples for those working together. If necessary, their contacts can be traced back to where they live. Some states already plan to do this. Such potential super spreaders can also form a sentinel network that is checked weekly for symptoms.
Second, stratify the risk of the areas, that is, demarcate the areas with an initially high risk of expected transmission and / or high vulnerability, such as dense and narrow settlements, with a greater number of older people and, this is very Important, choose people, within the areas, in an explicit statistically structured random form and test them using RT-PCR methods. The chances of finding infected people can be improved by using local information first: people who report symptoms of an influenza-like illness (ILI) and then randomly draw from the area’s voter list.
Along with data on occupation, age, sex and any recent morbidity and / or disease, basic information on the intensity of contact with others should be collected from all examinees, even at critical points. Despite the increase in evidence, the percentage of positive tests has remained stable at around 4.5%: every 100 people evaluated led to five new patients with Covid-19. Is this because high-risk contacts are selected or because the prevalence is worryingly high? If it is, hopefully, the first, it means that the contact tracing processes are not standardized in practice, since the positivity of the test varies considerably between states.
Third, states must anonymize and publish this data, along with the results of associated tests. Randomisation would allow high-quality, real-time analysis of potential determinants of transmission, even in locations not directly tested. Information, this is important, from the initial stages should be used to iteratively improve sampling designs after each round of testing. This can be analyzed internally, but public supply will exponentially increase the speed and quality of the analysis, allowing for a better understanding of the disease and improving the policy response.
This suggested ICMR-plus strategy evaluates more asymptomatic people and uses structured randomization, to enable active learning by governments. Minimizes the risk of transmission of super spreaders, maximizes the possibility of detecting hidden infections and configures the real-time strategy for future tests, to optimize scarce medical and test resources. Importantly, you can also calibrate containment strategies. States should consider using these principles when embarking on their test trips, which unfortunately may take longer than expected.
Jishnu Das teaches at Georgetown University and Neelanjan Sircar at Ashoka University. This article was written by Partha Mukhopadhyay. All authors are affiliated with the Policy Research Center.
The opinions expressed are personal.