Former Health Minister Senaratne correctly diagnoses underestimation of Covid-19 cases but overstates extent

Fact Check

In his statement, former Health Minister Rajitha Senaratne made the overall claim that the number of people reported as having Covid-19 in Sri Lanka is a significant underestimate due to three reasons:

• Only tested and confirmed cases are reported, those who are infected but not tested are not counted.
• Only 15% of those infected develop symptoms of Covid-19, while the others can be carriers without showing symptoms.
• Sri Lanka has a low rate of testing at 100 per million people, in contrast to South Korea which has tested 8,000 per million.

This fact check evaluates the former minister’s overall claim by evaluating the specific reasons and facts provided in support of the overall claim.

The former minister is right in his first assertion. The statistic of confirmed cases significantly undercounts the number of infected cases, in line with the minister’s assertion. [For more details see https://ourworldindata.org/coronavirus-testing]

The second assertion states that only 15% develop symptoms, and 85% do not, implying that those who do not show syptoms may not be tested. Because of the preliminary and limited nature of the data available among the several studies done (see Annex 1 for a list of studies), it should not be possible to assert such a number conclusively. There is also a difference between not showing sypmtoms during the early stage of the infection and never showing symptoms through the full infection period; in other words, the difference between being pre-symtomatic and asymtomatic. Currently available studies show that many people who test positive don’t show symptoms (the estimates are in a wide range from below 20% to above 80%). The World Health Organization also suggests that 75% of such cases can go on to develop symptoms later. Therefore, the claim that only 15% show symptoms (which means 85% never show symptoms at all) could be considered an overstatement, on the information available at present.

The third assertion is on Sri Lanka’s relative rate of testing. The cumulative number of tests conducted per million people increases on a daily basis. The figures cited by the minister are correct as of the 31 March – when the relevant statistics for Sri Lanka and South Korea were 104 and 8,008 respectively. However, because the first local case in South Korea was on 22 January, and the first local case in Sri Lanka was on 09 March, this testing statistic, if evaluated against time passed since the first case, gives a lower relative difference of 159 for South Korea and 104 for Sri Lanka (on the 23^rd day).

Overall, the minister’s claim that the number infected in Sri Lanka would be much larger than the number of reported cases is supported by his reasoning – that the total number of those infected are not counted due to: not being detected through testing; not being discovered due to lack of symptoms; and limited testing. The specific numbers he ascribes to lack of symptoms and lack of testing can be seen as over-stating the applicable magnitudes, even while they remain applicable as reasons for the over-all claim.

Therefore, we classify the minister’s statement as PARTLY TRUE.

Appendix 1: Proportion of asymptomatic infections based on various studies

Exhibit 1: Covid-19 tests in Sri Lanka and South Korea