{"id":100848,"date":"2020-04-06T16:58:34","date_gmt":"2020-04-06T23:58:34","guid":{"rendered":"http:\/\/www.lankaweb.com\/news\/items\/?p=100848"},"modified":"2020-04-06T17:00:56","modified_gmt":"2020-04-07T00:00:56","slug":"covid-19-how-much-testing-do-we-need","status":"publish","type":"post","link":"https:\/\/www.lankaweb.com\/news\/items\/2020\/04\/06\/covid-19-how-much-testing-do-we-need\/","title":{"rendered":"Covid-19,   how much testing do we need?"},"content":{"rendered":"<h2><span style=\"color: #0000ff;\"><em>By&nbsp; Chandre Dharmawardana.<\/em><\/span><\/h2>\n\n\n<p>A number of distinguished medical\nprofessionals&nbsp; led by Dr. Ruvaiz Haniffa\n,&nbsp; Head, Dept. of Family Medicine,&nbsp; University of Colombo, published an appeal\n(see&nbsp; the Island , 1<sup>st<\/sup> April,\nDaily Mirror&nbsp; and other newspapers)\nemphasizing the&nbsp; Need for rapid\nexpansion in testing for COVID-19 in Sri Lanka\u201d.<\/p>\n\n\n\n<p>While creating a data base is essential for\nconstructing epidemiological models, another very effective approach uses&nbsp; information-based intervention. Detectives\nexamining social media or gossip learn about&nbsp;\nCovid-19 contacts, religious gatherings etc., and aggressively follow\nevery lead.&nbsp; Sri Lanka is currently in\nstage 3A (WHO classification)&nbsp; where\nclusters are being identified. &nbsp;&nbsp;To avoid\ndeteriorating to stage 3B,&nbsp; authorities\nhave to&nbsp;&nbsp; trace contacts&nbsp; and&nbsp;\nisolate cases. This&nbsp; is&nbsp; an intelligence-driven attacking approach\u201d\n(IDAA). The military&nbsp; in Sri Lanka&nbsp; was thus able to identify over 20,000 people\nand enforce&nbsp; self-quarantining. Direct\nsampling would not have found these people, or if found, it would be too late.<\/p>\n\n\n\n<p>So we need BOTH the IDAA and direct TESTING\nof optimally selected samples.<\/p>\n\n\n\n<p>Even people who seem perfectly healthy may\nbe carries of the virus. Testing suspected cases\u201d may not be enough. But no\ncountry can test&nbsp; even a significant\nfraction of the population? But what is that significant fraction? How can we\nget away with testing even smaller fractions? How do we sample populations? <\/p>\n\n\n\n<p>Test kits for the new virus are in&nbsp; short supply and now reserved for priority\ncases. Leading companies are scrambling to provide test kits to countries&nbsp; buckling&nbsp;\nunder the virus. There are also cheap wonder test kits\u201d that quick-buck\ninventors and alternative-medicine Gurus&nbsp;\nhave put out, claiming fast test results. These should be avoided and\nonly&nbsp; kits using proven methods&nbsp; must be used. <\/p>\n\n\n\n<p>&nbsp;According to Health Canada guide lines, a&nbsp; person with a confirmed&nbsp; virus infection\u201d must have positive nucleic\nacid amplification tests on at least two specific genome targets or a single\npositive target with nucleic acid sequencing.<\/p>\n\n\n\n<p>Many Western governments (e.g., France) had\ncomprehensive plans for dealing with Pandemics, put in place after the SAARs\nepidemic. But the rise of neo-liberals and Human-Rights lobbies with their\nbelief in minimal government intervention and devolution (e.g., in Spain)&nbsp; led to the disbanding of all such programs,\nleaving matters&nbsp; to the private sector\u201d\nand to local authorities\u201d.&nbsp; USA&nbsp; is still in that mode and lagging behind&nbsp; in fighting Covid-19. Today, many governments\nare scrambling to re-assemble those safety nets at great cost.&nbsp; <\/p>\n\n\n\n<p>The article by Dr. Haniffa and his learned\ncolleagues suggests the use of existing laboratories in the country to do the\ntesting. Even if we had 25 such labs, and even if each lab did several hundred\ntests a day, it is clear that even if test kits are available,&nbsp; the time delay in getting&nbsp; useful answers, coordinating timing&nbsp; and data processing&nbsp; a huge number of samples reliably under\nemergency conditions is impossible. We need good data in real time. A further\nsix months delay for&nbsp; peer review can be\navoided&nbsp; if the program is under&nbsp; medical and scientific experts. <\/p>\n\n\n\n<p>So, testing boils down to the\u00a0 crucial question of what is the smallest\u00a0 sample\u00a0 needed\u00a0 to\u00a0 get at a reliable answer\u201d? The medical professionals had not addressed this most crucial question, but their involvement is needed. Otherwise we can only address the basic statistical question of how big a sample should be,\u00a0 without additional wisdom.<\/p>\n\n\n\n<figure class=\"wp-block-image\"><img loading=\"lazy\" decoding=\"async\" width=\"299\" height=\"280\" src=\"https:\/\/www.lankaweb.com\/news\/items\/wp-content\/uploads\/2020\/04\/image-2.png\" alt=\"\" class=\"wp-image-100851\"\/><\/figure>\n\n\n\n<p><strong>Ramanayake&#8217;s Audio files.<\/strong><\/p>\n\n\n\n<p>Take a simpler problem. For instance, actor\nturned MP&nbsp; Ramanayake&#8217;s now notorious\ntelephone recordings run into, say 10,000&nbsp;\naudio files. It takes too long to review ALL the tapes. What&nbsp; is the minimum number (sample size) of\nfiles&nbsp; that we should listen to, to have\na very good chance of picking up the most juicy one?&nbsp;&nbsp; Elementary mathematics shows that the sample\nshould be a fraction 1\/e of the total, where e\u201d&nbsp; is the number 2.7182 which is the base of\nNapier&#8217;s system of natural logarithms. So, it is probably enough to listen to\napproximately 10,000 divided by 2.7, i.e., about 3704 audio files to pick up\nthe most interesting one!&nbsp; As the\nattached graph shows, increasing the sample size even to 50%&nbsp; does not strongly&nbsp; increase the chances of catching the\nincriminating case significantly.<\/p>\n\n\n\n<p>The same sampling technique can be used if\nyou have to select the best candidate from 1000 applicants to a job. Your best\ncandidate is most likely to be within the first 370 that you interview! <\/p>\n\n\n\n<p>The audio tapes, or candidates for a\nprescribed job, presumably all form a set of similar objects or individuals.\nBut human populations are not uniform. Even the individuals in the population\nof each district are still too dis-similar to form a statistically uniform\nsample\u201d. It is here that judicious planning is needed. Judging that the\nCovid-19 is most likely to be found in urban populations than in rural\npopulations (say), one may take samples from each city population in a given\ndistrict. One doesn&#8217;t even have to take a census of each city to know their\nproportions \u2013 they follow a mathematical law known as&nbsp; Zipf&#8217;s&nbsp;\nlaw. In any case, once the uniform demographic to be tested\u201d (having a\nnormal distribution) is&nbsp; identified, one\ncan use the inverse of Napier&#8217;s number&nbsp;\nand so take about 37% of each demographic to make&nbsp; the test samples for each district. <\/p>\n\n\n\n<p>Of course, a good team of statisticians can\nwork out a better statistical models&nbsp; to\nminimize the sample size,&nbsp; improve&nbsp; the&nbsp;\ngathering of data and deal with non-normal, multi-modal and other&nbsp; distributions.&nbsp; The data provides an empirical basis for&nbsp; constructing a trustworthy epidemiological\nmodel&nbsp; for&nbsp; Covid-19&nbsp;\naffecting the country. However, one really needs a time dependent model\ntaking account of the amount of virus (viral load ) absorbed by the patient\nduring the process of infection. The interaction between the virus and the\nimmune system is a race in time.&nbsp; If you\nget a large dose, you get higher viremia, more dissemination, higher infection,\nand worse disease.<\/p>\n\n\n\n<p>So, the suggestions by the health\nprofessionals must be taken seriously, but using a good sampling scheme rapidly\ndeployed to make the testing affordable and meaningful. This will also be an\nimportant research contribution to the epidemiology of the disease.&nbsp;&nbsp; No country should rely on the Covid-19\ngrowth curves of other countries. For instance, the rate of growth of Covid-19\nin a country with a large demographic of young people will be quite different\nfrom that of an European country with a large demographic of old over-fed\npeople. Furthermore, once a patient is identified as having the virus, there is\nstill no clear way of predicting if the disease will become acute. The British\ngovernment under Mr. Boris Johnson used a limited epidemiological model\nconstructed by Imperial College, London, and is said to have paid the price in\nmoney and lives on being misled on how Covid-19 will develop in the UK.<\/p>\n\n\n\n<p>Sri Lanka should also attempt making&nbsp; its own vaccines. Time-tested techniques for\nmaking SARS-CoV-2 vaccines are by inactivating whole virus particles with\nformaldehyde and adding a booster like alum, and doing clinical trials.<\/p>\n\n\n\n<p>All this can be derailed by a politically\npowerful Natha Deiyyo devotee\u201d&nbsp; who may\ncome forward with a quick revelation of what the status of the&nbsp; epidemic is, after claiming that Western\nScience is a Patta Pal Boru\u201d fabrication, as happened with the investigation\nof the Chronic Kidney Disease Endemic affecting the North Central Province of\nSri Lanka. Similarly, die-hard Muslim groups have already risen asking for\nMulsim burials (bathing the corpse etc.), showing the ever persistent power of\nfaith over reason.&nbsp; <\/p>\n\n\n\n<p>By&nbsp;\nChandre Dharmawardana.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>By&nbsp; Chandre Dharmawardana. A number of distinguished medical professionals&nbsp; led by Dr. Ruvaiz Haniffa ,&nbsp; Head, Dept. of Family Medicine,&nbsp; University of Colombo, published an appeal (see&nbsp; the Island , 1st April, Daily Mirror&nbsp; and other newspapers) emphasizing the&nbsp; Need for rapid expansion in testing for COVID-19 in Sri Lanka\u201d. While creating a data base [&hellip;]<\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"closed","sticky":true,"template":"","format":"standard","meta":{"footnotes":""},"categories":[122],"tags":[],"class_list":["post-100848","post","type-post","status-publish","format-standard","hentry","category-health"],"_links":{"self":[{"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/posts\/100848","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/comments?post=100848"}],"version-history":[{"count":0,"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/posts\/100848\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/media?parent=100848"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/categories?post=100848"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.lankaweb.com\/news\/items\/wp-json\/wp\/v2\/tags?post=100848"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}