Read on Twitter
A dive into'Patanjali Clinical Study'
According to registration on CTRI, Initial sample size was planned for 60 in each group (placebo and Coronil).Primary Endpoint was Virological "Cure" at 14 days.(PCR negative)for mild/moderate severity COVID Patients
http://ctri.nic.in/Clinicaltrials/pmaindet2.php?trialid=43900&EncHid=&userName=COVID
According to registration on CTRI, Initial sample size was planned for 60 in each group (placebo and Coronil).Primary Endpoint was Virological "Cure" at 14 days.(PCR negative)for mild/moderate severity COVID Patients
http://ctri.nic.in/Clinicaltrials/pmaindet2.php?trialid=43900&EncHid=&userName=COVID
However when trial results were launched n Coronil Market blitz rolled out. It turned out only ASymptomatic and mild Patients recruited. And no moderate ones . (45 Coronil,50 Placebo). Also they stopped trial at 7 days only. https://theprint.in/india/just-45-healthy-cases-interim-results-no-peer-review-%e2%81%a0-how-patanjali-found-covid-cure/447540/?amp&__twitter_impression=true
Can a trial stop midway. Yes for efficacy ,if interim looks were planned (Both Remedesivir and Dexamethasone were stopped midway for efficacy) and p values adjusting for multiple comparisons. But unlikely that it was done here. An unprofessional look.
So why did they stop?. Yes 100% virological cure at 7 days only.
In Coronil Group 69%(31) and 100%(45) were PCR negative at day 3,7 respectively
In Placebo 50%(25) and 65%(33)respectively.
And Baba went to town with "cure" story.
In Coronil Group 69%(31) and 100%(45) were PCR negative at day 3,7 respectively
In Placebo 50%(25) and 65%(33)respectively.
And Baba went to town with "cure" story.
Was it appropriate? No given that PCR fluctuates at borderline values,they should have documented one more negative IMO as was done in Recent RCT and followed up till 14 days.(remember India had 2 negative PCR criteria even for Discharge till recently) https://www.bmj.com/content/369/bmj.m1849
Is Primary Outcome Significant?
I calculated on basis of the Reported values inprint n PressRlz
Hazard rate : Rate of change of conversion is
2.36 (confidence interval 1.48-3.76),p=0.0002(log rank),
Effect size is yuuge..Almost 236% faster cure rate not seen commonly with drugs.
I calculated on basis of the Reported values inprint n PressRlz
Hazard rate : Rate of change of conversion is
2.36 (confidence interval 1.48-3.76),p=0.0002(log rank),
Effect size is yuuge..Almost 236% faster cure rate not seen commonly with drugs.
Taken into perspective it is almost 2 fold larger than time to clinical Recovery for Remedesivir (which incidentally didn't have any virological cure effect in other RCT) and dexamethasone,both of which have been labelled as reasonably successful drugs.
The NIMS group also reports modest reduction in CRP(p<0.03) and IL-6(?remarkable decrease). In my view these biomarkers are often done and reasonable. https://twitter.com/vasudha_ET/status/1275456401431646217?s=20
So why am I Skeptical of their results.
The Rate of cure is pretty high in both Treatment and Placebo group. Check out cure rate in Hydroxychloroquine trial by Tang et Al .Pretty low at 7 days. That despite using a lot of off-label Antivirals in Standard of Care Group.
The Rate of cure is pretty high in both Treatment and Placebo group. Check out cure rate in Hydroxychloroquine trial by Tang et Al .Pretty low at 7 days. That despite using a lot of off-label Antivirals in Standard of Care Group.
Even in observational study by Raoult,cure rate at 7 days was 70%% at 7 days and only 12.5% in Control group.
In fact Tang et Al had to change their end point to 28 days to account for negative conversion in all patients.
Even in India median Recovery rate almost lagged 2 weeks behind active cases a till negative PCR criteria was there.(when it was changed to symptom based Discharge). https://twitter.com/anupampom/status/1270615638248091648?s=20
So why was the cure rate (Negative PCR) very high in both Coronil and Placebo group. (100% and 65% at 7 days)
I suspect it had to do with some PCR interfering chemical in Coronil group(Anu tail) and placebo(nasal drip).
I suspect it had to do with some PCR interfering chemical in Coronil group(Anu tail) and placebo(nasal drip).
Even if data is published, I will be skeptical . Due to huge effect size , mismatch with historical data and questionable conduct of trial group with marketing blitz.
I have an open mind on this will be more happy if these results are reproduced in other centres or by more experienced trial group.
I believe we need an oral drug even in mild cases(which has a high recovery rate) because we need to cut down on transmission as well to others. This is one of key principles of public health. Unfortunately Coronil is far from being that drug.
Is it possible that they took Patients already diagnosed a week back with very mild/no symptoms [since they already dropped Moderate severity Patients] ,that might have contributed to faster recovery in both groups.
Yes.
Hence imp to document
A)2 PCR negative (to limit variability)
B) Diagnosis date
C) Get trial results replicated at other places https://twitter.com/shananalla/status/1276037879261589505?s=20
Hence imp to document
A)2 PCR negative (to limit variability)
B) Diagnosis date
C) Get trial results replicated at other places https://twitter.com/shananalla/status/1276037879261589505?s=20
