Appropriateness of Using Daily Positive Cases to Predict the Trend of Covid-19

Recently, we get a daily release of data on the development of new cases of COVID-19. This is good, because it shows the transparency of the government in handling this case. For some people, this situation is more or less like a football league fan waiting every day for his favorite club's score and standings. If the total of new cases are small, psychologically, the community will be happy, and hope this continue until the outbreak cease to zero. But on the contrary, if the total of new cases are high, concerns arise that the outbreak will last long. If the total of new cases is consistently high, some people even will assume that it is, probably, a sign that the second wave has entered, which can be more dangerous than the first wave. But is that right? Let's learn from the West Nusa Tenggara Province (WNT), by looking at the data below, obtained from the official release of the WNT Covid-19 Task Force.

Graph-1 showed the trend of total daily positive cases in WNT from March 24, 2020 (confirmation of the first case) to June 6, 2020. Initially, cases were fluctuating but tended to rise until reached its peaked on April 23, 2020. Then there was a decline to May 18, 2020, which was soon followed by relatively high eruptions of daily positive cases, which was even higher than the cases on April 23. Does Graph-1 showing the actual or real daily new cases? Does it indicate if WNT has entered the second wave? To answer this, let's look at the other graphs.

Positive cases presented in Graph-1 were obtained from RT-PCR tests conducted by several laboratories in WNT. The total of PCR tests that have been carried out in WNT from 6 April 2020 - 6 June 2020 are 7266 tests. Let's focus on data on the number of PCR tests per day for the past 2 weeks and the number of positive cases found (Graph-2). At first glance, it appears that the more test conducted (red bars), the more the number of new positive cases (blue bars). This was even clearer when we saw the Graph-3. In this last chart, it appeared that the number of tests was positively correlated with the number of daily new positive COVID-19 cases. The correlation coefficient (r) was 0.91, indicates a very strong association, and there was no better explanation to explain this association. In plain language, the ups and downs of positive cases reported in the release over the past 15 days, depending on the number of PCR tests conducted. If there were many tests, there would be many positive ones, and vice versa. Thus, it would be very misleading if the trend of the daily cases was used to predict the development of this outbreak in WNT.

Based on simple calculations, the number of RT-PCR tests carried out to find one positive case is between 4-15 tests, with an average of 9 ± 4 (SD).

In the future, the things that should be done, perhaps, is to increase the number of PCR test per day, to make it into e.g. 500 test per day. This is important to ensure the status of all high-risk groups such as patients under surveillance, person in monitoring, asymptomatic carrier, and the traveller without symptom. If the total number of samples examined per day is fluctuating and relatively small, the data will be less meaningful, and our responses (tracing and isolating the positive cases) would be far behind the speed of spread of this virus. Theoretically, the sooner all risk groups are tested, the faster the isolation action can be taken (for those who are positive), so that the spread of disease can be minimized (WA).