Read on Twitter
Thread:
Irregular heart beats (PVCs) after heart attack were associated with death but the common sense approach of suppressing them increased the number of deaths/arrests from 26 to 63 or by a factor of 2.4. (1/n) https://www.nejm.org/doi/full/10.1056/NEJM199103213241201
Irregular heart beats (PVCs) after heart attack were associated with death but the common sense approach of suppressing them increased the number of deaths/arrests from 26 to 63 or by a factor of 2.4. (1/n) https://www.nejm.org/doi/full/10.1056/NEJM199103213241201
If 5% of symptomatic cases with #coronavirus die, imagine if we have a "common sense" approach that has not been rigorously evaluated in a randomized trial that behaves like anti arrhythmias in the CAST Trial. The morality could be increased by 2.4 fold to 12%.
(2/n)
(2/n)
The public health implications just in the US of this are staggering:
Instead of 2,000,000 people dying we could have 4,800,000
Instead of 200,000 dying, we could have 480,000
Instead of 60,000 dying we could have 144,000
All the while thinking we were actually saving lives!
Instead of 2,000,000 people dying we could have 4,800,000
Instead of 200,000 dying, we could have 480,000
Instead of 60,000 dying we could have 144,000
All the while thinking we were actually saving lives!
Historically, over the past 30 years, patients have agreed to be randomized to placebo vs active treatment in scenarios with as high if not higher mortality as #coronavirus: heart attack and heart failure for instance. These patients were also acutely and critically ill.
There are 2,000 patients dying DAILY in the US alone. Worldwide the number is much higher. This provides enormous statistical power (i.e. ability to have statistical certainty) We could have quickly answered the question(s) about the safety and effectiveness of drugs.
Scientists have put forth "promising therapies" to be tested. We MUST become better organized to conduct randomized controlled trials of these therapies. To date this has been an organizational / operational shortcoming, not a scientific shortcoming.
We know MUCH MORE about how to do randomized trials than we do about #coronavirus !
We should leverage our operational knowledge to speed the execution of randomized trial in this area.
We should plan ahead now for a cohesive national approach to answering these questions.
We should leverage our operational knowledge to speed the execution of randomized trial in this area.
We should plan ahead now for a cohesive national approach to answering these questions.
We should have the following:
1. Optimally, there should be central randomization for the country.
2. Optimally, there should be a shared standard of care control group.
3. There should be standardized electronic case report forms for consistency as a minimum.
1. Optimally, there should be central randomization for the country.
2. Optimally, there should be a shared standard of care control group.
3. There should be standardized electronic case report forms for consistency as a minimum.
4. We need consistent definitions of what constitutes a #coronavirus death. If someone dies at home that should probably count for instance. If a #coronavirus patient has a cardiac arrest, that should also count as a #coronavirus death in most cases.
5. We should have a centralized IRB that approves the a consistent and uniform trial protocol for all centers in the US and this could be reactivated at anytime in the future. This would obviate the delays at local IRBs.
6. We should have a centralized data safety monitoring board to frequently monitor the safety of all the different therapies and if a single therapy shows overwhelming benefit based upon pre-specified stopping rules, that therapy could be introduced into practice.
7. If we were organized centrally, we could also test promising combinations of therapies rather than single therapies.
8. The FDA should be deeply embedded in the process to allow evaluation of results in real time and speed approval.
8. The FDA should be deeply embedded in the process to allow evaluation of results in real time and speed approval.
9. Trials at single centers in hundreds of patients are not likely to be statistically powered to answer the question definitively.
10. Even if +, they may be a false +!
10. Central organization would dramatically increase the sample size and offer more definitive conclusions.
10. Even if +, they may be a false +!
10. Central organization would dramatically increase the sample size and offer more definitive conclusions.
11. Central organization would also increase the sample size the answer if drugs are effective & safe in different subgroups of patients.
12. This is no time to argue over which academic group gets credit. Papers could be authored by a large collaborative to give everyone credit
12. This is no time to argue over which academic group gets credit. Papers could be authored by a large collaborative to give everyone credit
13. We should also capitilize on wearable technology to track heart rate and fever after discharge, track patient reported outcomes, and alert family members to recurrent symptoms.
14. We need to leverage claims databases to efficiently collect clinical outcomes at scale.
14. We need to leverage claims databases to efficiently collect clinical outcomes at scale.
15. Clearly we need to identify the most sensitive and specific serologic tests that provide the greatest reliability.
16. A centralized clinical event committee could determine who did and did not have events on a consistent basis.
16. A centralized clinical event committee could determine who did and did not have events on a consistent basis.
#coronavirus is NOT just a scientific challenge.
It is just as much if not more of an organizational, collaborative and operational challenge.
We must organize and collaborate as a team to succeed.
It is just as much if not more of an organizational, collaborative and operational challenge.
We must organize and collaborate as a team to succeed.
