Revisions in the case definitions in mainland China
February 21, 2020
When a new disease emerges and causes an outbreak or epidemic, one of the important early decisions is how to say whether a patient is a “case” or not. If we have a good case definition, counting cases and analysing their characteristics can tell us how fast the disease is spreading, what kind of people are at higher risk, and how effective are control measures in reducing spread.
In the COVID-19 outbreak there have been a number of revisions to the case definition in mainland China which have caused some confusion. Most recently, changes announced on the 12 February led to a sudden surge in case numbers in Hubei province, and a further change announced on 19 February led to a sudden drop in case numbers. Here, we discuss how cases are defined, why those changes might have been made, and what we can learn about the epidemic from the case numbers so far.
At the start of the outbreak in Wuhan, initial attention focused on patients with viral pneumonia and links to the Huanan Seafood Wholesale Market in Wuhan. In order to capture potential COVID-19 cases in other cities across China, if they were to occur, the first national case definition was issued on 15 January 2020 and in order to be a confirmed case, a patient had to meet these three conditions: (1) recent travel to/from Wuhan or an exposure to a wet market in Wuhan in the 14 days prior to illness onset; (2) clinical evidence of viral pneumonia; (3) laboratory identification of the virus in a patient sample plus virus sequencing to confirm that the virus matches SARS-CoV-2. This was a fairly strict definition. On January 27, recognising that some infections could be mild, the case definition was broadened to allow for milder infections without pneumonia.
On February 4, the fifth version of the case definitions allowed for a new type of case in Hubei province, called a “clinically confirmed” case. Under this new categorisation, a patient with radiographic evidence of pneumonia, and blood tests indicating that it was specifically viral pneumonia, doctors could confirm a case without requiring laboratory detection of the virus. Because hospital laboratories can only run a certain number of virus tests each day, if there are a large number of patients it is possible that there would be a limit to the number of positive cases that could be identified from among the patients in the hospital, and this would be called a “ceiling” effect. Virus tests are not only been used for patients in the hospital, but also for testing people in the community or in quarantine facilities who have mild symptoms but could be infected. Removing the need for virus confirmation in hospitalised patients could therefore free up laboratory capacity for some of these other tests, but avoid underreporting of the cases in hospitalised patients.
On February 12, Hubei province announced that it would group together the laboratory-confirmed cases and the clinically-confirmed cases and report the total as the confirmed cases for their province. This new approach led to a surge in the number of cases in Hubei Province on the day of introducing the new category. It is important to note that the increase of cases reported may not have indicated a significant change in the outbreak because the additional confirmed cases were not all new cases but some were clinically-confirmed cases from the preceding days or weeks who now became eligible for inclusion as confirmed cases.
However, on February 19, the National Health Commission announced that it had reversed the case definition in its 6th edition of the national treatment and diagnostic guidelines. In the new revision, clinically diagnosed cases are no longer counted as confirmed cases in Hubei province. This change brings Hubei back to the standardized case definition used across China. On February 20 and 21 the daily number of confirmed cases was much lower than on preceding days, perhaps partly because of reductions in the use of viral testing in some hospitalised patients.
The changes in case definitions, leading to surges and drops in reported case numbers, do make it difficult to interpret trends in case numbers. However, setting aside these fluctuations, there is a clear overall declining trend in the daily numbers of new cases in mainland China since early February, suggesting that the lockdowns of personal movement in major cities have been effective in curtailing the spread of infection.
Here are some news articles discussing this issue in English:
Here are some news articles discussing this issue in Chinese: