PCR-test positive rate in Tokyo falls to a "non-emergency" level

On May 8, Tokyo’s Governor, Yuriko Koike surprised us by disclosing that the PCR test positive rate, the percentage of PCR test-takers who tested positive over the number of all test-takers, has actually been falling in the last three weeks. On May 11, Governor Koike announced that the average positive rate for the 7 days till May 10 was 5.9%, below the threshold typically thought to allow emergency measures to be lifted.  PCR-test positive rate in Tokyo falls to a ‘non-emergency’ level.

Chart 1: PCR test positive rate published by Tokyo Government

Source: Tokyo Metropolitan Government, JMA

Notes: 

Positive rate is the average of the last 7 days.

There are three categories of PCR test sites: (1) Tokyo Metropolitan Institute of Public Health (2) PCR test centers (3) Medical Institutions (tests for which medical insurance was applied). And the calculation of positive rates depending on the dates as follows.

Jan 24th to Apr 9th: Test results from (1) are included in the calculation for the positive rate but test results from (2) and (3) are not included.

From Apr 10th to May 6th: Results from (1) and (2) are included, but not from (3).   

From May 7th onward : All results are included in the calculation. 

MHLW was publishing misleading statistics

Until the news broke on May 8, the impression among the public was that the positive rate in Tokyo remained elevated in Tokyo, close to 40%. The high level of the positive rate was considered to be evidence that there are not enough PCR tests being conducted in Tokyo. Why did the public have such impression? It is because Ministry of Health, Labor and Welfare (MHLW) has been publishing such numbers. According to MHLW, the accumulative PCR positive rate for Tokyo as of May 7 was 37.0%.

So, which number is correct? Hands down, Tokyo Metropolitan government is the correct number to follow. For those who were reading the definition of the positive rate calculated by MHLW, it was clear MHLW’s data were faulty. The calculation done by MHLW has been based on “incomplete” data and the actual ratio should have been lower. What do we mean by “incomplete”?  MHLW calculated PCR test positive rate as “Number of those who tested positive” divided by “Number of people who took PCR test”. However, MHLW was using counting methods that were obviously flawed. For “Number of those who tested positive”, MHLW were including all patients whose PCR test result was positive, regardless of where they took their tests. But for “Number of people who took PCR test”, MHLW were only including those who took PCR tests at Tokyo Metropolitan Institute of Public Health. Those who took PCR tests at other facilities such as private medical facilities were not included.

It means that those who took the test in in private medical facilities and tested positive are included in the numerator, but not in the denominator. Obviously, this created an upward bias. When PCR tests started, all tests were conducted at public health centers, but as the number of tests conducted outside public health centers increased, the upward bias kept increasing.

Does it mean Tokyo Metropolitan Government could lift the emergency measures?

Unfortunately not. There are two other indicators often mentioned as conditions that need to be satisfied for municipalities to lift the emergency measures. One is the number of newly confirmed infection, especially those whose source of infections are unknown. As of May 7, the 7 days average of the confirmed cases are still elevated for Tokyo, although the average is likely to fall sharply as the daily cases between May 8 and May 10 dropped sharply.

Another condition that probably takes more time to improve would be the stringency of hospital beds. The number of patients still being hospitalized remains elevated for Tokyo, although there is a sign that the number has stopped rising anymore and perhaps started to decline.

 

How soon can the Tokyo government lift the emergency measures? It is still hard to say. As we look around in Tokyo, people seem to have gotten tired of staying home and there are increasingly more people on the street. While the presence of people on the street does not necessarily mean that it will start to spread again, but it really depends on how careful people are. Let’s see how quickly we see infection and the number of hospitalized to decline in the next one week and we could start to hope that the emergency measures would indeed to be lifted at the end of May.

Report written by:

Contributing Analyst 

Akira Morita

Email: akira.morita@japanmacroadvisors.com