During the pandemic I wrote a lot, in terms of analysis and encouragement of public health measures. Some of my contributions were on this Blog, others were on Twitter. To this day, I think that the measures taken to manage the pandemic were for the most part necessary and successful, for example restricted activity, masking, and vaccine policy.
The COVID vaccines in particular have been one of the great achievements in the history of medicine.
But here are some ideas about ways I think it could have been done better: in sharing this I know I run the risk of dabbling into territory that many would consider outside my lane of expertise. That was a constant frustration for public health experts during the pandemic.
However, I was conscious all the way through the pandemic that we did not adequately use the most powerful tool available in science and medicine to evaluate the effectiveness and optimize the efficiency of an intervention: the Randomized Controlled Trial.
And also in dealing with people who were extremely resistant to adopting mandated restrictions or vaccines, there could have been a way to manage this situation that would have helped rapidly gather much better data about COVID itself, protect the population, while also appeasing people who did not want to get vaccinated or follow restrictions.
Randomized Controlled Studies (or "Randomized Controlled Trials -- RCTs")
There were many RCTs during the pandemic, but in my opinion there could have been much more done here, in almost every stage, and there could have been massive public investment to get this done, which subsequently could have saved billions of dollars of economic loss, in addition to saving lives.
For example, it was very clear from basic science knowledge that masking was valuable to reduce viral contagion. When some people raised the idea of doing more RCTs on masking, it was met with some resistance, as though we were wasting everyone's time in a dangerous way. Some used a comparison with doing an RCT of using parachutes when jumping out of an airplane -- obviously this would be recklessly inappropriate, and the entire placebo group would die! Masks, like parachutes, are obviously effective, and mask proponents made a reasonable case that randomizing people such that a placebo group would not get masks would be needlessly dangerous. But an RCT does not require that there be a "placebo" wing! It only requires that the study be randomized to compare one treatment with another. To follow the parachute analogy, it could be to randomize people jumping out of an airplane to receive one of two different types of parachutes, each of which an accepted standard; or for them to use two different timings for releasing parachutes, if each of these was also within an accepted standard.
In the case of masks, there could have been RCTs of using different types of masks (e.g. procedure vs. N95), different timings of masking (indoor only vs continuous outside the home), or different replacement times for masks (e.g. re-using N95 masks for days vs replacing them every use), or different N95 use details (e.g. receiving formal instruction on technique vs. not). And early in the pandemic there was an N95 shortage. One of the ways to deal with this could have been to randomly distribute the timing of the N95 supply, so that some entire communities would receive an adequate supply first. Then the entire community would use N95s while adjacent communities would temporarily make do with other types of masks. Then the disease prevalence rates and hospital admission rates could have been compared between adjacent communities. This type of design could have been tremendously valuable, since mask use has not only an individual benefit for infection control, but has a collective impact, akin mathematically to the effect of vaccines -- if everyone in the population has a modestly reduced probability of infection, then it could translate to a massive reduction in community prevalence. If such a study had shown reduced infection and hospitalization rates in the better masked areas, it could have propelled a much more urgent and timely effort to manufacture better masks for everyone, and in the medium term the whole community could have had better access to N95s, saving thousands of lives. But since such studies were lacking, there was enough doubt about mask effectiveness or effect size to delay the massive investment needed to increase mask production.
When RCTs are done, it does not settle questions once and for all: in good science, we are always repeating, tweaking, and refining. New RCTs would have to be done after the first ones, with different details being looked at, or simply for replication.
One type of mask use behaviour which should have been better guided by evidence, is the use of masks outdoors. I still see many people outside with their masks, or people wearing them in their cars on the way to work. It was pretty clear from the ventilation evidence that outdoor mask use was very likely unnecessary, unless one were in very close proximity to crowds, or doing a lot of talking up close. Perhaps masks would still be needed in playgrounds etc. but certainly not for walks at the beach alone or with just a few people close to you.
Similarly, RCTs could have been done on ventilation control in buildings. The basic science on ventilation was one of the most important and underappreciated areas of science during the pandemic. There was a wonderful group of engineers who had done great work in this area. Ventilation improvement was also a totally non-controversial intervention: regardless of one's views about masks or vaccines or restrictions etc., I think everybody would welcome the idea of having better fresh air inside our homes and workplaces. Ventilation improvements involved air filtration (such as with HEPA or MERV-13 HVAC filters) but also increased fresh air replacement rates. But the engineers again used the parachute analogy when there were challenges to do RCTs, arguing that their work was established basic science, which didn't need to be tested in an RCT. But once again, if RCTs had been done, of whole communities which made ventilation improvements, vs communities which did not, we could have much more quickly found a "signal" of improved infection control, and then made much more rapid investments in ventilation improvement technology for everyone.
In all of these studies, the data to gather should always have been not only rates of infection, but also most importantly rates of severe disease. Some interventions such as masks arguably could cause a reduction in infection rate, but perhaps in cases of people getting infected despite mask use, they would have inhaled a smaller inoculum, and possibly could subsequently have developed milder disease, since the immune system would have had a little bit more time to respond to the virus before getting overwhelmed. The question of whether inoculum size impacts disease severity is yet another one which I don't think is well-enough answered by the research.
The Covid Hotel
The "COVID hotel" idea was something I proposed early on as a thought experiment at the very least, and there was at least one other scientist in the US who shared this idea as well... but it was received very coldly by experts--when I gently suggested it I got the sense that they thought it was scandalously inappropriate or unethical. But this idea could have saved thousands of lives, and could have helped gather optimized, crystal-clear data about COVID in terms of the mechanism of transmission, the effectiveness of masks, the impact of ventilation, etc. This information could have been obtained within a few months, and then could have helped focus optimal interventions with much better clarity and urgency, and to mobilize public investments in such things as masks etc. much sooner.
Here's the idea: if people refused to be vaccinated, or insisted on having unrestricted freedoms, instead of punishing them using the justice system, they could instead opt to check into a "covid hotel" in which they would choose, with informed consent about risks, to be deliberately infected with COVID under controlled conditions, with optimal medical support available. Then they would stay in the hotel for a few weeks under quarantine until they were no longer infectious. Upon checking out, they would have a much lower risk of spreading COVID--the risk would be comparable to a person who had been vaccinated. In this environment, there could be meticulously controlled experiments to determine if COVID could be transmitted through an airborne route (perhaps all the time, perhaps only in some cases of "superspreaders" etc.), or through a surface contamination route (after all this time, it is not crystal clear that surface contamination was ever a major route of spread). And there could have been masking studies in this environment to determine if masks (including styles of mask usage and mask type such as N95 vs procedure masks, as well as the proportion of people wearing masks, etc.) reduced the likelihood of contagion, or reduced the ensuing severity of disease (since the masks even if they didn't prevent infection might at least reduce the inoculum size). Similarly there could have been meticulous ventilation control studies, to see if improved ventilation reduced contagion.
In this environment, participants could even be offered to choose modalities of treatment of their choice, delivered by their practitioner of choice. They could try the "alternative treatments" in vogue if they wished, or opt for standard medical care. This way, there could have been much more rapid evidence to establish the impact of these alternative treatments (all of these alternative remedies have been utterly disproven, but this could have happened much more quickly and persuasively in the "COVID hotel" environment).
Some of the benefits of this idea would have been much, much better quality data about mask effectiveness, mode of contagion, effectiveness of ventilation improvement, etc. And there would have been much less spreading of COVID to vulnerable people by people who refused to adhere to public health guidelines. And there would have been much less upset from people who wanted more freedoms. In fact these people, instead of being vilified, could have felt like true heroes, even from a scientific point of view. The cost of this, of course, would have been that people who chose the "COVID hotel" route would have been much more likely to die, or to have severe long-term consequences of COVID. But this would have been their choice, and if they didn't check into the COVID hotel, they would have subjected themselves to the same risk in the community, with less medical support and therefore an even higher likelihood of medical harm, and all the while they would have spread COVID to many more people, without contributing anything useful to the world's knowledge about the disease. There are many other examples in life of people who are willing do risky activities, following informed consent: for example, joining the military, the fire department, or doing risky sports such as hang gliding.
Animal Studies
There were animal studies during COVID. It's a sensitive topic, since it is important to respect the rights of animals. But COVID affected the animal world as well, and the research about contagion would have led to benefit for not only human populations but animals as well. One very particular type of animal study that was never done well enough was to use an animal model to demonstrate spreading modality. For example, the ventilation outflow from hospital rooms with human COVID patients could have been pumped into an animal enclosure of susceptible animals. If these animals developed COVID it would have been tremendously strong evidence for airborne transmission in humans. If, in a follow-up experiment, the same ventilation pipe passed through a HEPA filter first, and then into the animal enclosure, and if these animals did not contract COVID, it would have been incredibly powerful evidence that a simple filtration technique could prevent contagion. If animals were simply allowed to visit hospital rooms where COVID patients had spent a few days, but who had left, and where the air in these rooms had been replaced using ventilation, then it could have helped determine if surface contamination unequivocally could cause COVID spreading. It is quite possible that surface spreading was never a major problem, while airborne spreading was a huge problem, hence efforts would have been directed towards ventilation rather than as much surface cleaning. But we would have needed the research to prove this.
Vaccine / Restriction Timing
Restrictions were deployed in the pandemic quite wisely, particularly with a view to prevent the nightmare of ICU and hospitalization overflow. For some individuals, going beyond mandates, they voluntarily maintained restrictions for months or years following vaccination. One interesting study issue could have been to randomize people to maintain strict restrictions after vaccination for a long period of time, vs. ending restrictions for those people starting about 4 weeks after each vaccination. This would have caused the unrestricted individuals to have greater exposure to ambient circulating COVID strains, but this would have occurred in the context of good immunity. As the vaccine strains kept changing, the vaccinated people would continue having new exposures with new strains, and especially as 3-6 months passed after their vaccines, they most likely would have had some mostly mild cases of COVID along the way. But I wonder if this process would have in the long term led to improved, robust immunity to multiple strains, with the same or lower long-term health risk, while also improving community freedoms, compared to the situation of maintaining continuous long-term restricted behaviour. In a sense, this idea would suggest that the vaccine and annual boosters would be the primary preventative defense, but then exposures to the ambient COVID strains in the community would subsequently act as "boosters" for previously vaccinated people, and in the long term (measured over 3-5 years or more) lead to equivalent or better health outcomes, with fewer restrictions needed. Conversely, the studies might instead show that maintaining more restrictions over the longer term would have led to better long-term outcomes. We can't know for sure, since the studies were never done.
Unfortunately, in the aftermath of the pandemic, there has been increased polarization in the world about public health measures of all types. We are seeing decreased rates of vaccination against other diseases, and we are seeing a return of various diseases which had previously been nearly eradicated, such as measles. Just as with Covid, most people who get these diseases will recover ok, but there will be needless cases of severe disease and death, including among young children. I hope that the field of public health can work hard on the sociopolitical aspects of their profession as well as the epidemiological parts. But I also wish that some of the best scientific tools, such as RCTs, could be done much more quickly and on a much larger scale than what we saw during the worst years of COVID.
No comments:
Post a Comment