Shifting Standards

via @goldencaskcap

An Apple Daily article dated Feb 10th argued that China’s CDC was seeking to downplay the number of confirmed coronavirus pneumonia cases by revising the criteria used for designating ‘confirmed’ cases to exclude patients with no symptoms from the tally, seemingly against WHO guidelines.

The claim was based on a reading of the fourth edition of China’s official treatment guidelines and an article from February 9^th reporting that Heilongjiang province had excluded 13 cases of asymptomatic patients who tested positive for coronavirus from confirmed cases because they were asymptomatic.

To understand if the story is correct, we need examine several points one by one:

Do China’s Definitions Contradict the WHO Guidelines?

The WHO guidelines define “confirmed cases” as “[a] person with laboratory confirmation of 2019-nCoV infection, irrespective of clinical signs and symptoms”, which is why China’s exclusion of asymptomatic cases was viewed as problematic.

The guidelines also considers ‘suspect cases’ including several types of patient with respiratory tract symptoms and a history of contact with China or the virus. Another group, ‘Probable cases”, are in turn defined as “a suspect case for whom testing for 2019-nCoV is inconclusive or is tested positive using a pan-coronavirus assay and without laboratory evidence of other respiratory pathogens.”

If you read carefully, you’ll notice that two different diagnostic assays are mentioned in the definition of probable cases: one specifically for 2019- nCoV and the other for pan-coronavirus which includes testing for other varieties of coronaviruses.The WHO guidelines address asymptomatic patients testing positive for 2019-nCoV, but because probable cases are all suspect cases and thus show symptoms, the WHO guidelines do not address specifically address asymptomatic carriers that are tested positive on assays other than the 2019-nCoV PCR.

China’s 4^th Edition guidelines defines confirmed cases as clinically diagnosed cases or suspected cases with either a positive covid19 rt-PCR test or gene sequencing.

（三）确诊病例。 临床诊断病例或疑似病例，具备以下病原学证据之一者： 1.呼吸道标本或血液标本实时荧光 RT-PCR 检测新型冠状病 毒核酸阳性； 2.呼吸道标本或血液标本病毒基因测序，与已知的新型冠状 病毒高度同源。

Asymptomatic patients are defined as those without clinical symptoms but with positive “new coronavirus virology test results” ( （四）无症状感染者。 无临床症状，呼吸道等标本新型冠状病毒病原学检测阳性 者。).  While the tests used for confirmed cases are clearly defined, the wording for for asymptomatic patients is rather nebulous.  Does it include cases tested with the pan-coronavirus assay?  It is technically not in conflict with WHO guidelines to separately classify asymptomatic cases testing positive for pancoranavirus.

Was this even a change in definition?

As early as the third edition of China’s Guidance, published on Jan 22nd asymptomatic patients were already separated into a separate category than confirmed patients. If all provinces had followed this guidance, then confirmed cases should not have included these asymptomatic patients from at least that date. So there is a possibility that Heilongjiang’s correction may have been due to having not followed the national guideline and having to correct its classification, rather than a nationwide change in definition.

Are they doing this to make numbers look lower? 

The quick answer is yes as they have reclassified some cases from confirmed to suspect. However, they have disclosed that number reclassified each day and it’s at most 119, or less than 0.3% of total cases. See relevant disclosures here: [1][2][3][4]

The second part of the answer is that other recent diagnostic changes have actually broadened the definition for confirmed cases. Most notably classifying symptomatic patients as confirmed cases based only on positive CT imaging scans without requiring PCR confirmation. This has led to a huge UPTICK in number of confirmed cases allowing earlier isolation of more cases, which far outweighs any minor adjustments that might have been caused by the asymptomatic case definition change has caused.