Interpreting results:
Though I have no medical background I find a few things quite interesting about these trial results. 1- At Day 15 of the study (the last day of treatment) there was 0% mortality in the 20 mg dose Ifenprodil treatment arm compared to a 3.3% mortality rate in the untreated control arm, p=0.18. (So this basically says 0% of patients on Ifenprodil died, but 3.3% on just standard of care did. This is literally a perfect score, but comes in at p=0.18 which is nowhere near statistical significance at p=0.05. Basically unless there were more patients in the trial or a whole bunch more people died in the trial it would never have been possible to do better than this as it was a perfect score!?) 2- Oxygenation (SpO2): Of patients with a low blood oxygen level (SpO2 <94%), 100% of patients in the 20 mg dose treatment arm returned to normal levels of oxygen at day 4 compared to day 9 for patients in the untreated arm (adjusted hazard ratio 1.91, 95% ci 0.97-3.77, p=0.061). power calculations project that 450 patients would be required to confirm a statistically significant result with this endpoint in a phase 3 trial. (once again, a perfect score of 100% of the patients returned to normal oxygen levels at day 4 as opposed to 9 days in the untreated arm. still only reaches p=0.061. so with another perfect score which is approaching statistical significance, it's once again mathematically impossible to achieve without more patients) 3 - time in icu: topline results for this endpoint indicate that there was also a strong trend to less time spent in the icu in the overall study by patients in the 20 mg dose arm, as compared to patients in the untreated arm (adjusted hazard ratio 10.45, ci 1.23-88.61, p=0.0315). however, the company cautions that additional, confounding variables were detected, and these numbers need to be confirmed with additional analysis, as well as power calculations conducted to project the required size for a phase 3 study. (here we surpassed statistical significance at p=0.0315 but i'm not entirely sure what "confounding variables detected" means...) maybe someone with a medical background could shine some light on my conclusions? 100%="" of="" patients="" in="" the="" 20="" mg="" dose="" treatment="" arm="" returned="" to="" normal="" levels="" of="" oxygen="" at="" day="" 4="" compared="" to="" day="" 9="" for="" patients="" in="" the="" untreated="" arm="" (adjusted="" hazard="" ratio="" 1.91,="" 95%="" ci="" 0.97-3.77,="" p="0.061)." power="" calculations="" project="" that="" 450="" patients="" would="" be="" required="" to="" confirm="" a="" statistically="" significant="" result="" with="" this="" endpoint="" in="" a="" phase="" 3="" trial.="" (once="" again,="" a="" perfect="" score="" of="" 100%="" of="" the="" patients="" returned="" to="" normal="" oxygen="" levels="" at="" day="" 4="" as="" opposed="" to="" 9="" days="" in="" the="" untreated="" arm.="" still="" only="" reaches="" p="0.061." so="" with="" another="" perfect="" score="" which="" is="" approaching="" statistical="" significance,="" it's="" once="" again="" mathematically="" impossible="" to="" achieve="" without="" more="" patients)="" 3="" -="" time="" in="" icu:="" topline="" results="" for="" this="" endpoint="" indicate="" that="" there="" was="" also="" a="" strong="" trend="" to="" less="" time="" spent="" in="" the="" icu="" in="" the="" overall="" study="" by="" patients="" in="" the="" 20="" mg="" dose="" arm,="" as="" compared="" to="" patients="" in="" the="" untreated="" arm="" (adjusted="" hazard="" ratio="" 10.45,="" ci="" 1.23-88.61,="" p="0.0315)." however,="" the="" company="" cautions="" that="" additional,="" confounding="" variables="" were="" detected,="" and="" these="" numbers="" need="" to="" be="" confirmed="" with="" additional="" analysis,="" as="" well="" as="" power="" calculations="" conducted="" to="" project="" the="" required="" size="" for="" a="" phase="" 3="" study.="" (here="" we="" surpassed="" statistical="" significance="" at="" p="0.0315" but="" i'm="" not="" entirely="" sure="" what="" "confounding="" variables="" detected"="" means...)="" maybe="" someone="" with="" a="" medical="" background="" could="" shine="" some="" light="" on="" my="">