A research group led by Associate Professor Ryosuke Omori of Hokkaido University and Assistant Professor Kenji Mizumoto of Kyoto University found that changes in the number of reports of new coronavirus infections in Japan in the early stages of the epidemic were generally observed under the epidemic. It was discovered that the fluctuation did not apply to the curve, but was a linear fluctuation.
In order to understand the epidemic situation of the new coronavirus infection, the time series of the number of positive test reports (number of reports) of patients who tested positive by PCR is published in many countries.Generally, in the early stage of an infectious disease epidemic, the time variation of the number of reports is approximated by a curve (exponential function).This time, we compared using the daily report data of the number of reports of new coronavirus infections in Japan and Italy published by John Hopkins University.
The report count data in Italy was more applicable to the exponential function over the entire period, and it was highly possible that the report count data reflected the epidemic situation.However, in Japan, a straight line is more applicable up to a certain point, which means that a certain number of new positives will be reported over time, and other than the interpretation that "an epidemic situation in which one infected person can be transmitted to only one person", the test is conducted. It can be interpreted that the number is limited, the number of new positives per day is constant, and the epidemic situation is not captured.In fact, the exponential function model became more applicable to the report number data this time after the date of the significant increase in the number of tests, and it is considered that the report number data after that date began to reflect the infectious disease epidemic.
Time-series data on the number of reports is important for grasping the epidemic situation, but caution is required in its analysis because there may be bias due to inspection policies and plans.Elimination of the bias requires analysis using multiple data, including time-varying data on the number of reports on the day of onset, time-varying data on the number of hospitalizations, serious injuries, and deaths.
Paper information:[International Journal of Infectious Diseases] Changes in testing rates could mask the novel coronavirus disease (COVID-19) growth rate (PDF)
