Over the weekend, there was news that gave us hope that we might be able to escape from the current coronavirus pandemic. Clinical trial (Phase 1) of an oral drug being developed by Pfizer to treat the new coronavirus 2/3 (interim analysis of phase trial) to reduce the risk of severe disease. 89% Results were reported to reduce
In this article, we will discuss this oral drug, Paxlobid. (PAXLOVID) However, the risk of worsening 89% We will tell you what kind of results were used to calculate the reduction. At the same time, risk 89 Since the % reduction is an estimate obtained from a limited number of patients, we will calculate and discuss the confidence interval for the estimate.
Pfizer homepage According to a news article published in ( placebo ) As a result of allocation to the group administered with 28 We are comparing the proportion of patients admitted within days.
Paxlobid administration: 389 philtrum 3 People hospitalized (no deaths)
Placebo administration: 385 philtrum 27 person hospitalized (then 7 name died)
■Comparison of hospitalization risks
In order to arrive at the risk reduction figure mentioned in the title, we need to calculate the relative risk. To do this, create the data in JMP as follows. Column ``Group'' indicates the drug administration group, column ``Hopitalized'' indicates whether hospitalization resulted (Yes, No), and column ``N'' corresponds to each value of ``Group'' and ``Hospitalized.'' Enter the number of people.
After this, we usually use the "Fit Y by X" platform to find it, but here we will do the same thing using the "Categorical" platform.
Select Analyze > Consumer Research > Categorical from the menu bar and specify the columns as follows: The key is to specify "N" as [Frequency].
Once specified, select the option Relative Risk. You will then be asked for the target level and target sample level, so select "Yes" and "PAXLOVID" respectively. That is, we are specifying that we are comparing the hospitalization rate for the paxlobid group to the hospitalization rate for the placebo group.
The report displays a crosstabulation table and a graph (share chart) showing the proportions of each group. As shown in the cross-tabulation, the hospitalization rate for each treatment group is 0.77% (= 3 / 389) for the paxlobid group and 7.01% ( = 27 / 385) for the placebo group.
At the bottom of the report, you will see a relative risk report. In this example, the “paxlobid administration group” ” The relative risk for hospitalization for the ``placebo group'' is calculated as follows:
relative risk = Percentage of hospitalizations in the paxlobid group ÷ Percentage of hospitalizations in the placebo group
= (3 / 389) ÷ (27 / 385)
= 0.11
That is, the relative risk of administering paxlobide versus administering a placebo is 0.11(11%) It shows that it decreases to Considering how much the risk of hospitalization is reduced from this, 89% ( 100% - 11% ) is the number reported on the news.
■Confidence interval for relative risk
The relative risk value calculated in the previous section ( = 0.11) This is just an estimate based on the limited number of patients included in the study. Therefore, it is better to consider not only the estimated value but also its confidence interval.
Let's look at the relative risk report again. 95% confidence limits are shown.
From this, the relative risk (= 0.11) against 95% The confidence interval is approximately 0.03 ~ 0.36 You can see that it is a surprisingly wide area. Therefore, to put it in rough terms, the reduction in the risk of hospitalization is 64% ~ 97% You will have to think broadly. 64% Even if I take it lightly, I feel that it is quite effective.
Naturally, the larger the number of samples, the more confident we are in the estimate, and therefore the narrower the confidence interval. Hypothetically, if both the number of patients administered and the number of hospitalized patients in this clinical trial are Ten Suppose it was twice as much. That is, the paxlobid-treated group 3,890 philtrum 30 Human hospitalization, placebo administration group 3,850 philtrum 27 When a person is hospitalized, the relative risk at this time is the same as before. 0.11 but, 95% The confidence interval is 0.08 ~ 0.16 is. The confidence interval is much narrower than the actual example.
The results of this interim analysis have shown sufficient efficacy and safety, and it appears that an application will be filed for approval in the future. It seems that both the medical institution and the medical institution can enjoy great benefits. By then, we may be able to live a life where we don't have to wear masks in public, like before the coronavirus.
