In the absence of a 100% testing rate, every measure of COVID-19 deaths is going to be an estimate based on a large number of assumptions. As we learn more about the virus and its progression, these assumptions will change and death figures will get updated based on the new assumptions. As Dan Scally pointed out, this is not a China specific thing. Below, I use New York City as an example of the ambiguities in this kind of data. If you simply don't trust the Chinese government (not an unreasonable position), there's no way for us to prove whether they're lying or not, but the updated death tolls are to be expected and shouldn't be considered suspicious on their own. --- For example, on April 14th, New York City updated their estimated death figures by 3700 by including probable deaths that could not be confirmed due to a lack of available testing: > New York City, already a world epicenter of the coronavirus outbreak, sharply increased its death toll by more than 3,700 victims on Tuesday, after officials said they were now including people who had never tested positive for the virus but were presumed to have died of it. > The revised death toll renewed focus on shortcomings in testing that have hamstrung city and state officials since the beginning of the outbreak. A limited number of tests have been available, and until now, only deaths where a person had tested positive were officially counted among those killed by the virus in New York. >But for weeks, the Health Department also had been recording additional deaths tied to the virus, according to two people briefed on the matter. Those cases involved people who were presumed to have been infected because of their symptoms and medical history. > [N.Y.C. Death Toll Soars Past 10,000 in Revised Virus Count][1] There are also issues with deciding how to collect data for people who don't die in hospitals: >A staggering number of people are dying at home with presumed cases of coronavirus, and it does not appear that the state has a clear mechanism for factoring those victims into official death tallies. >In the last three days, 766 people were found dead in their homes, bringing the total for the first eight days of April to 1,891, according to the city’s medical examiner’s office. It’s likely that many have not been counted in the current tally. > [How Many People Have Actually Died From Coronavirus in New York?][2] Data inconsistencies can also come from differences in how deaths are reported between different health departments: >Mr. Cuomo said on Wednesday that the official death count numbers presented each day by the state are based on hospital data. Our most conservative understanding right now is that patients who have tested positive for the virus and die in hospitals are reflected in the state’s official death count. >The city has a different measure: Any patient who has had a positive coronavirus test and then later dies — whether at home or in a hospital — is being counted as a coronavirus death, said Dr. Oxiris Barbot, the commissioner of the city’s Department of Health. > [How Many People Have Actually Died From Coronavirus in New York?][2] Complications in sharing and reporting data, exacerbated by an emergency situation, can also cause inconsistencies: > On Tuesday, the city’s count of confirmed cases went up to 6,589. > The city and the state have at times differed in their counts of the dead in New York City. As of Monday, the state said that 7,349 had died of the virus in the city. City officials have complained that they are at the whim of the state, which has been slow to share the data it receives from hospitals and nursing homes. The state Health Department explained on its website that the discrepancy is caused by the city and state using “different data systems.” > [N.Y.C. Death Toll Soars Past 10,000 in Revised Virus Count][1] There are also questions about how to count "excess deaths", that is people who didn't die of the virus directly, but who wouldn't have died had it not occurred: >The outbreak is likely to have also led indirectly to a spike in deaths of New Yorkers who may never have been infected. >Three thousand more people died in New York City between March 11 and April 13 than would have been expected during the same time period in an ordinary year, Dr. Oxiris Barbot, the commissioner of the city Health Department, said in an interview. While these so-called excess deaths were not explicitly linked to the virus, they might not have happened had the outbreak not occurred, in part because it overwhelmed the normal health care system. > [N.Y.C. Death Toll Soars Past 10,000 in Revised Virus Count][1] [1]: https://www.nytimes.com/2020/04/14/nyregion/new-york-coronavirus-deaths.html [2]: https://www.nytimes.com/2020/04/10/nyregion/new-york-coronavirus-death-count.html