peto

System for mass periodic infectious disease testing


License
GPL-3.0
Install
pip install peto==0.1.0

Documentation

Peto

Build Status Coverage Status

System to support mass periodic testing for infectious disease.

This system is motivated by the occurrence of the Cononavirus (COVID-19) pandemic.

Mass periodic testing is a response proposed by Julian Peto FRS which depends on dispatching and processing swabs, and recording test results. The requirements are described below.

This approach offers a way to end the lockdown that many of us are currently experiencing without returning to the "let rip" mass deaths scenario. Looking beyond the current crisis, there is a possibility that future epidemics could also be responded to with mass periodic testing.

Motivation

Coronavirus (COVID-19) is a novel human respiratory virus. The respiratory aspect means the virus is transmitted during the course of physical social interaction, which is unavoidable for many human activities on which society depends. Its novelty means there is no vaccine (and none expected for at least one year, perhaps two).

The virus causes no symptoms in some people. But of those people who do develop symptoms, some people will recover and some people will die. The dispositions to different outcomes are not very well understood, but age seems to be a strong factor, with older individuals being more at risk than younger ones. The outcome depends upon individual biological reaction to being infected, partly their immune system response but also other factors that are not fully understood. If the number of infected people reaches the point where more people develop severe symptoms than a health service has capacity to care for, the proportion of infected people who die (the case fatality rate) may increase.

Respiratory viruses are relatively easily transmitted from person to person, through exhalation of droplets containing the virus into air that is subsequently breathed in by another person, and through contact with contaminated surfaces and areas of the face such as eyes, mouth, and nose.

The average number of people that become infected by an infected person can vary. If the number of people who are infected by an infected person is greater than one, then over time the number of cases of tends to increase exponentially. However, if the number of people who are infected by an infected person is less than one, then over time the number of cases tends to decrease exponentially. This number is sometimes referred to as the "reproduction number" and varies according to the period of infectiousness, the degree of physical interaction, and the modes of transmission. Since the period of infectiousness is more or less a property of the the virus itself, attempts to reduce the reproduction number consider changing the degree of physical interaction ("stay at home") and the modes of transmission ("catch it, kill it, bin it", "wash your hands").

Because there is no vaccine, if personal hygiene measures (e.g. "wash your hands", or everybody wearing a mask) are insufficient to reduce the reproduction number below the critical level of one, the response to the existence of a novel respiratory virus becomes a choice regarding whether or not to implement quarantining of infected individuals. Quarantines prevent infected people from infecting other people by physical isolation for the period in which they are infectious ("stay at home").

Without implementing quarantine, most people will eventually become infected. This option may lead to "herd immunity" if the individuals that recover sustain resistance to reinfection. However this approach will lead to a large number deaths, and may take some time for all the individuals susceptible to developing severe symptoms to die. And any herd immunity may fade away, or be deprecated by mutation of the virus. This is the "let rip" or "mass deaths" strategy.

The alternatives to the "mass deaths" strategy are the various quarantine strategies. Infected individuals can be quarantined either by mass quarantine ("lockdown") or by selective quarantine. Selective quarantine can be implemented: either with "periodic mass testing" to detect infected individuals whether they have developed symptoms or not; or with "mass tracking" so that when a person develops symptoms and feels unwell they can be tested and their contacts can be traced and tested.

The "mass deaths" strategy causes severe detriment to well-being, through direct loss of life ("mass deaths", impacting on mental health through bereavement, whilst also impacting on economic activity through trauma and organisational disruption.

The "lockdown" strategy causes severe detriment to well-being, through direct damage to livelihoods and education, whilst also impacting on mental health through isolation ("quarantine for all").

The "mass tracking" strategy avoids mass deaths and severe damage to the economy, but involves enforced mass tracking of all individuals so that those who crossed the path of the infected individual can be identified, traced, and tested. By waiting until an infected individual develops symptoms, infected individuals have a greater duration of being infectious.

The "periodic testing" strategy also avoids mass deaths and severe damage to the economy, and it also avoids mass tracking. It involves repeat mass testing of a high proportion of individuals, so that infected individuals are detected regardless of whether or not they develop symptoms later. The period can be adjusted according to prevalence of infection, and would be expected to decline exponentially as the incidence of infection falls away.

This software is designed to support periodic mass testing, by recording testing results and establishing the test status of individuals.

The approach was conceived by Julian Peto FRS.

Proposal

https://www.bmj.com/content/368/bmj.m1163

In Editor’s Choice of 19 March Godlee mentions the urgent need for increased capacity to test frontline healthcare workers serologically to verify their immunity to the covid-19 virus.[1] Even more urgent is capacity for weekly viral detection in the whole UK population. This, together with intensive contact tracing, could enable the country to resume normal life immediately. The virus could only survive in those who are untested, and contact tracing would often lead to them. Within the tested population anyone infected would be detected within about a week (0 to 7 days plus sample transport and testing) of becoming infectious.

Centrally organised facilities with the capacity to test the entire UK population weekly (in 6 days at 10 million tests per day) can be made available much more quickly and cheaply than a vaccine, probably within weeks. This heroic but straightforward national effort would involve a crash programme to enlist all existing PCR (polymerase chain reaction) facilities, acquire or manufacture the PCR reagents, and agree protocols including a laptop program for barcode reading in smaller laboratories. The US Food and Drug Administration (FDA) has just authorised a test kit for detecting the covid-19 virus that can be run on machines used in the NHS for HPV screening. Only laboratories that do PCR routinely would participate, subject to central quality control and at cost price. The Wellcome Sanger Institute, UK Biocentre, and smaller academic laboratories, together with all commercial facilities, should have enough machines or can get more immediately from the manufacturers. The 24-hour extra staffing to run their machines continuously would be bioscience students, graduates, and postgraduates familiar with PCR who already work in or near the laboratory. Processing capacity equivalent to 4000 Roche COBAS 8800 systems is needed, and the UK may already have both the machines and the trained staff in post or immediately available.

All patients registered with a GP would be sent a test kit (a swab for throat and nasal self-sample, and a transport tube labelled with their name, NHS number, and a barcode). Homeless people and other disadvantaged groups would be served by charities already in contact with them. The Post Office, Amazon, and other companies already have the capacity to collect swabs from everyone with an address. Swabs might go to central facilities for preparation and arraying before dispatch to local laboratories for PCR.

Everyone should be tested weekly. All households and care homes would return self-taken swabs from all residents together. In most homes all residents would test negative and they could resume normal life immediately. An identification card certifying date and result of latest test (positive, negative, negative contact of a positive case) might be useful for policing arrangements. By the time the first test is done there may be more than a million infected people who must be treated or remain quarantined at home or in care until all residents at the address test negative. That unavoidable crisis for the NHS would be ameliorated by earlier diagnosis and treatment, and hence reduced pressure on intensive care, and by having all staff as well as patients tested regularly. Contacts of positive people who test negative could choose continued home quarantine or, at little extra risk, choose to join a group of up to 10 test-negative contacts (usually with other family members). Subsequent weekly national testing, together with behavioural changes and efficient contact tracing, would find progressively fewer infections and might soon be extended to a month. This emergency system would only be needed for about 2 months but could be rapidly reintroduced to control any future epidemic caused by a new virus.

Requirements overview

The IT for this is easy. The national master file (separately for England, Wales, Scotland and N Ireland) is everyone's name, date of birth, NHS no. and (if recorded) tel and email for those registered at each household address, based on current GP practice records. That is created by downloading centrally from the servers of the companies (EMIS and TPP) who provide almost all GP databases. Barcoded sample tubes with preprinted name and date of birth are delivered and collected weekly from each household and distributed to labs for testing. A direct access facility for authorised people to submit changes of address is needed. Each testing lab creates a new Excel file of samples received and test results for each run of 96 or 48 samples (depending on PCR machine capacity) and uploads it after each run. The national master file is thus always up to date and labs retain complete records of their work. All scheduling of sample deliveries and collections, together with household status on all residents (all negative, all negative or untested, any positive) and hence quarantine status is also on the file, together with free fields for other info. A facility for adding people with no NHS number or address for samples distributed by outreach workers is also needed, plus a mobile app for them to enter the tube ID no. and add name, DoB and any further info on such samples. Results would go back to the outreach worker if no address is recorded.

It would be a mistake to have more complex IT. It can all be managed through one simple master file by giving NHS, police and public health full access to the master database and limited access for other users.