current thinking is "uh...we don't know for sure"
Much of the information is increasing and changing (and is of course expected as patient data increases, but is naturally frustrating from a policy-making/recommendations point of view!)
One example is defining common symptoms for this novel coronavirus in order to make all the recommendations that have been rolling out over the last weeks. Ideally, there's one or a cluster of specific things that
universally and
specifically identify the disease (medical term for that is "pathognomonic," for your future scrabble games!) but that's unlikely to be the case for a virus like this, so instead the most common signs (currently, dry cough and fever) have to be used as a threshhold for possible infection and implementations of precautions/quarantine/treatment.
Some reports now are highlighting gastrointestinal components of disease (nausea/vomiting/diarrhea) and implications for fecal/oral transmission (another route aside from aerosol/droplet exposure). Different populations may also exhibit different symptoms (i.e. populations with greater respiratory comorbidities due to a larger population of smokers), which can complicate making "universal" recommendations.
Another example that has gained significant traction today due to new WHO recommendations regards limiting use of NSAIDs (like ibuprofen or aspirin) in place of acetaminophen (Tylenol
- which is also an anti-pyretic (fever-lowering drug) and analgesic (pain drug), but different mechanism of action) due to observations of more significant disease morbidity in patients that had taken NSAIDs. I first saw this reported several days ago by the French minister of health, and both now and then information is still lacking as to whether the process is causal or correlative, that is:
a. if you are infected with coronavirus and take NSAIDs to manage your symptoms, the drug actively makes your disease worse (some people have postulated mechanisms for this, but as yet haven't had time to test them!)
b. if you are infected with coronavirus and have more severe symptoms of disease because the disease itself is bad, you are more inclined to take a "stronger" medication than Tylenol, and so those with pre-existing worse disease are attempting to use a higher tier of drugs, and thus use of NSAIDs might actually be a predictor of worse disease symptoms/morbidity due to perceived necessity of use, rather than actually cause the greater morbidity.
(But to be safe, if it's the former, best to use Tylenol to manage symptoms of fever right now, if you're able to!)
Sorry, this got wordier than I had intended, but I'm having slight deja vu for this situation, as we experienced it on a MUCH smaller scale several years ago in the equine veterinary world - interestingly, also with a (horse-specific) coronavirus. Historically, equine coronavirus was something we tested for, but rarely found causing disease in adult animals, until roughly 5 years ago, at which point we started seeing it with increasing frequency as a cause of contagious, acute, moderate to severe gastrointestinal disease in horses (they don't thus far exhibit any respiratory component, nor does that strain infect humans, thank goodness!)
/epidemiology rambling