Wuhan Coronavirus: Case classification standards change causing reported-cases spike
A key issue in any observation-based data is changes to the criteria for observation. That's coming to play in the 2019-nCoV outbreak numbers today, with Reuters and other sources reporting "Coronavirus death toll leaps in China's Hubei province":
[T]he 2,015 new confirmed cases reported in mainland China on Wednesday were dwarfed by the 14,840 new cases reported in Hubei alone on Thursday, after provincial officials started using computerized tomography (CT) scans to look for infections.
Hubei had previously only allowed infections to be confirmed by RNA tests, which can take days to process and delay treatment. RNA, or ribonucleic acid, carries genetic information allowing for identification of organisms like viruses.
The current reports stand at 60,330 confirmed cases and 1,369 official deaths, according to current data at Wikipedia. I'll note that fifteen days ago based on then-current trends, projection without epidemiological containment was for 200,000 cases, 10,000 deaths. Thankfully, even with a broader inclusion criteria, we're well below those values. Containment does largely appear to be working, and the epidemic may be further self limiting. (More below.) My existing projections are based on the earlier criteria and should not be applied to the broader clinical-diagnosis numbers.
Numbers reported depend on observational criteria. In general, my recommendation is to look to the case count as a trend indicator, across successive periods with consistent criteria, and for deaths as a far stronger magnitude indicator.
That is: we know that the cases were underreported. That's been widely criticised, and I suspect you'll see reports now of a sudden explosion in cases as Another Dire Sign. It's ... more complicated than that. Case reports to date have largely been a function of:
- Limited numbers of virus test kits sensitive to the 2019-nCoV specific RNA (viruses don't have DNA, so this is the equivalent).
- Testing of severe cases only. For non-severe cases (the majority), home isolation and care are largely sufficient.
- Accepting most cases as "suspected" rather than confirmed.
The new numbers should be compared against earlier "suspected" cases, not "confirmed" cases. What are now being included are clinically diagnosed cases, based not on a definitive RNA test, but on symptoms as presented at a clinic, largely CT scans of lungs, among other measurements. This is actually very common in medical practice, and many conditions are diagnosed based on symptoms rather than a definitive test, often because symtoms are far more accessible (a case of availability heuristic), faster, cheaper, and in virtually all cases, sufficient. It's nice to have a definitive diagnosis, but not necessary. My understanding is that test kits have often been reserved for medical personnel themselves, many of whom have been infected with 2019-nCoV.
As before, epidemiological and public health responses are most important. If you are in an outbreak area (China, generally), rather than rushing out to buy masks (of marginal use), make sure you have and are using cleaning supplies, and minimise hand-to-face transmission. Wash surfaces, door handles, light switches, bannisters, and other contacted surfaces. Disinfect (spray bleach) around bath and toilet facilities. Use alcohol-based hand-sanitisers. (Again, antibacterial treatments do not work for viruses, that's wasted money.) Wash your hands frequently. Avoid crowded public spaces. Wipe down shopping trolly handles and other surfaces if possible. Be aware of what you're touching (elevator controls, any public touch-screen devices, keyboards, etc.). If you are infected and must go out, wear a mask but primarily to protect others --- you're limiting spread of droplets from your mouth and nose. (It's possible that applying table salt to the masks may increase their anti-viral properties, possibly by spraying with a saline solution.)
Expect to see travel bans and restrictions in place both within and transiting China for another few weeks, possibly months.
The real risk with 2019-nCoV is that it escapes containment and becomes a recurring annual disease much as the current common cold and influenza, though with far greater impacts. Given its high mortality rates, this could have severe impacts worldwide (deaths in the 10s to 100s of millions), though it's likely that this would eventually moderate. That's the scenario people are hoping to avoid.
Risks remain high, the containment trends still look very positive. Coronavirus fares poorly at higher temperatures and humidities -- odds of it spreading especially within the subtropical and tropical regions, and particularly in summer months, are fairly low. This means that China and neighbouring countries (again, notably North Korea with limited public health and public information resources and practices), Europe, and North America are probably most at risk. Community transmission elsewhere has remained low (cruise ships being a notable exception), and deaths outside China are also extremely rare: 2 (Hong Kong and Japan) despite 9,525 cases outside of China.
Fingers crossed.
The other notable news is that the Chinese Communist Party chiefs of Wuhan City and Hubei Province have both been sacked for their mishandling and information suppression of the 2019-nCoV initial outbreak. Again, China and the world could have had an additional 4-6 weeks' prior notice of the disease had initial reports not been covered up, before the mass internal migration of Chinese New Year celebrations. That was extreme misgovernance, and I expect to see more heads fall, possibly higher up, as well as further political reforms toward free speech within China.
