– The news was abrupt and, to some, surprising: Overnight, a Chinese province near Russia had cut its count of confirmed coronavirus cases by more than a dozen.

The revision stemmed from what appeared to be a bureaucratic decision, buried in a series of dense documents from the national government. Health officials said that they would reclassify patients who had tested positive for the new coronavirus but did not have symptoms, and take them out of the total count of confirmed cases.

The documents offered little detail or explanation, and skepticism was immediate. A Hong Kong newspaper called the decision a “disguise.” World Health Organization officials seemed caught off guard when asked about the move at a news conference this week.

The change in counting cases is only one factor that has made it difficult for experts to determine the true scale of the epidemic. In fact, the shifting landscape of how infections are defined and confirmed has led to significant variations in the estimates for the extent of outbreak.

Early Thursday, provincial officials in Hubei province, the center of the outbreak, announced that nearly 15,000 new cases and 242 new deaths were recorded in a single day, largely because the authorities expanded their diagnostic tools for counting new infections.

Until now, only infections confirmed by specialized testing kits were considered accurate. But those kits have been in such short supply — and so many sick people have gone untested — that the authorities in Hubei province have started counting patients whose illness have been screened and identified by doctors.

The result was a sudden — and large — spike in the overall tally for the coronavirus: more than 1,300 people killed and well over 50,000 infected.

The surge in cases in Hubei underscored how elusive the exact scale of the epidemic is.

The change in how cases are counted reflects a two-headed problem in the global fight against the disease. On the one hand, health officials need to stay flexible in dealing with new outbreaks.

One the other hand, mistrust of the Chinese government — especially when it comes to being transparent about the threat and extent of the virus — remains pervasive.

“It’s pretty clear that there is an issue with trust about whatever the Chinese government comes out with at the moment,” said Kerry Brown, a former diplomat and director of the Lau China Institute at King’s College, London.

“That may be terribly unfair,” Brown said. But, he added, “to redefine things — even legitimately — at a moment like this is always going to be a presentational challenge, because people are going to be very sensitive, and they’re going to suspect there’s another agenda.”

The new numbers out of Hubei came only a day after China reported that new infections had hit the lowest point in a single day since Jan. 30. Experts cautioned then that it was premature to draw any conclusions from the drop.

The shifting case counts are not the only example of conflicting or spotty information. Researchers have given differing estimates on when the outbreak might peak, ranging from a date already past to several months in the future. The Chinese authorities have closely guarded the demographic details about the fatalities, creating uncertainty about who is most susceptible.