Perhaps the single most serious surprise associated with COVID-19 is the apparent ability of asymptomatic people to infect others without even knowing they have the disease themselves. This has greatly complicated attempts to track and preempt the spread and rendered many standard control protocols less effective if not moot. An even more serious surprise may yet be in store, although it is simply too soon to tell, the time factor again coming into play. All future control and recovery scenarios have been built upon the assumption of at least temporary immunity developing for those who have had COVID-19 and, presumably, for those who would someday receive a vaccine against it. If this proves not to be true, and there are at least a few small disturbing indications that it may not be, then long-term caseloads, death rates and economic effects would be sustained at much higher levels than currently hoped. Notwithstanding all of the issues associated with the various COVID-19 statistics reported, simply ignoring the figures or dismissing them is not a viable or logical option. Few if any analyses of public policy decisions can be based on complete and perfectly reliable data. In a complex environment, decisions will always be based upon incomplete and potentially contradictory information. Fortunately, there are methods for dealing with uncertainty and bias in assessing the data available and making decisions based upon those assessments. Where measurement methods remain largely consistent, however flawed, trends can still be tracked and evaluated since the counting process has not changed. Where measurement methods change, presumably to improve accuracy or usefulness, counts and trends going forward can be much more useful and, over time, comparisons to prior conditions can be made once adjustments for differences in methodology can be inferred from the new experiences. Inevitably, at least some improvements in treatment protocols will develop, even without new medications or vaccines, indeed it is already happening, that may help reduce the severity and/or infectiousness of the disease. Even at present, however, the data with all of its flaws can be put to good use by experienced and capable analysts using effective modeling techniques. To do so, and more importantly to put modeling output to effective use in making public or personal policy decisions, it is important to understand the structure and purpose of the models themselves. The basic modeling technique for public health analysis of pandemic spread and appropriate mitigation is known as an SIR model, where the initials stand for Susceptible, Infected, Recovered. In a basic model, recovered also includes dead and the “R” can sometimes be termed resolved, meaning that one way or the other that person neither has the disease anymore nor can they infect anyone with it. In more complicated versions, a “D” is added to differentiate for death versus recovery and various other aspects are either added as additional variables or adjusted as inputs to reflect differing assumptions or public health actions.
∴ PROGNOSIS
25
Made with FlippingBook - Online catalogs