print PRINT


Prefectures use different methods to calculate positive test rate for COVID-19

  • May 24, 2020
  • , Asahi , p. 2
  • JMH Translation

By Chiaki Hagiwara, Yusuke Morishita, and Shohei Sasagawa


There is no standardized method in Japan for calculating the “positive test rate,” a statistic that has attracted attention on account the new coronavirus outbreak. The positive test rate is the proportion of people who tested positive for the coronavirus among those who underwent PCR tests to check for infection. Tests performed by private hospitals are not included in some prefectures, and cases in which the same person was tested multiple times are handled differently depending on the prefecture. The government’s experts panel views this situation as problematic. There are calls to standardize the calculation method.


Calculation of the positive test rate:

Positive test rate = Number of positive test results / Number of people tested


Although the positive test rate is considered to be an important indicator for determining the infection situation, there are differences in how it is calculated. The difference centers around two points: 1) Inclusion of tests conducted by private hospitals: 2) Inclusion of confirmation tests performed at the time of hospital discharge.


There are disparities among the four prefectures under the state of emergency as of May 24 and the thee Kansai prefectures in which the state of emergency was lifted on May 21.


The Tokyo Metropolitan Government (TMG) initially included only tests performed at public centers. The total number of tests did not include tests performed by private hospitals and therefore TMG did not release Tokyo’s positive test rate. Later, TMG changed its methodology to include cases from private hospitals in the total number of cases, and released the positive test rate for the first time on May 8. Excluding confirmation tests conducted at the time of discharge, the positive test rate was 1.3% between May 16 and May 22.


Kanagawa and Hyogo still do not include tests performed by private hospitals. Chiba does not include tests performed by private hospitals because the data it receives from a private testing organization includes confirmation tests. Chiba plans to change its calculation method to include tests by private hospitals, but excluding confirmation tests.


Osaka and Kyoto include tests conducted by private hospitals. Saitama initially only counted tests performed by private hospitals and 13 prefecture-operated public health centers. Starting May 15, Saitama will include tests by four public health centers operated by ordinance-designated cities and core cities in its calculation of the prefecture’s positive test rate.


Hokkaido does not perform PCR tests at private institutions. None of the eight prefectures mentioned include confirmation tests in the calculation of the positive test rate. According to the Ministry of Health, Labour and Welfare (MHLW), nearly 20 prefectures include confirmation tests in their calculations of the positive test rate.


Although there are no standards, Shigeru Omi, vice chair of the government’s expert panel has pointed out that “basing the calculation only on tests by public centers makes the denominator smaller. Inclusion of the tests by private hospitals makes the denominator more accurate.” Omi also said with respect to the confirmation tests, that “since hospitalized patients are tested multiple times, double counting them will make the denominator too large.” Omi’s position is that tests from private hospitals should be included in the calculation, but that confirmation tests should not be included because they are unrelated to the positive test rate.


The amount of variation due to calculation method has not been publicized. Different prefectures have different calculation methods. The MHLW does not have an accurate account of Japan’s positive test rate. The government’s expert panel raised this issue, stating that ““it is important to be able to compare the situation across different prefectures” in its analysis and proposal on coronavirus countermeasures.


Some people point out that it is not enough to standardize the calculation method for the positive test rate. University of Tokyo Graduate School of Public Policy project professor Isao Kamae (health care policy) says that there is a “statistical bias in the positive rate calculated from groups which were tested because they are suspected of infection.” Kamae says that a “reliable figure cannot be calculated because the number of tests is small.”


The MHLW approved a coronavirus antigen test kit on May 13. This test is less accurate than a PCR test, but results can be obtained in about 30 minutes, so is expected that testing will be expanded. In June 2020, antibody tests to determine whether a person has been previously infected will be performed on about 10,000 people. Kamae says that expansion of testing with antigen tests and antibody tests in addition to PCR tests will lead to estimates for the spread of community infection. “The government needs to act quickly to establish rules for calculation and survey systems,” says Kamae.

  • Ambassador
  • Ukraine
  • COVID-19
  • Trending Japan