Promotional literature

[Guest guest]

Respected members,

Appraisal

of Promotional literature:

Are

results shown Statistically significant or clinically significant?

Are

results shown are relative risk reduction or absolute risk reduction?

Statistical significance will be useful only if such a

significant difference

will translate into a meaningful benefit to the patient.

Confidence interval is a better statistical parameter. It gives

an idea about the precision of

treatment effects. For example, if the rate and the 95% CI of an event were estimated from 100

different samples

for the same population, then one can be 95% sure that

the true rate lies between the confidence intervals.

The outcome can be presented in terms of risk reduction

(absolute/relative), odds

ratio, number needed to treat. Relative

risk reduction (RRR) is a method commonly employed

to express results. It is the per cent reduction in events in

the treated group compared to

the control group event rate. The RRR is not a good way to compare outcomes. It amplifies small

differences and makes insignificant findings appear significant,

and it doesn’t reflect the baseline risk of the outcome event. A

better statistic is the Absolute Risk Reduction (ARR), which is

the difference in the outcome event rate between the control

group and the experimental treated group. The ARR does

not amplify small differences but shows the true difference

between the experimental and control interventions. For

example, if the outcome event rates in treatment groups A

(old drug) and B (new

drug) are 50% and 80% respectively, the RRR will be (80-50)/50×100 = 60%. This may be interpreted

as treatment B is 60% more effective than control, which is

incorrect. Let us look at the ARR, which is outcome rate with

B (80%) minus outcome rate with A (50%) = 30%. It can

be accurately stated that Treatment B is 30% more effective

than the control drug A. Inverse of the ARR gives the Number

needed to treat (NNT) i.e. how many patients must be treated

with the new treatment for one patient to benefit more than

if the patient had been treated with the standard treatment.

The NNT is a simple statistic that could be easily interpreted by

the physicians as well as patients. The NNT provides the likelihood

that the test or treatment will benefit an individual patient.

It also provides an impression of the baseline risk of the

adverse event, and the cost of treatment incurred to society. Thus,

it justifies the “reasonableness†of preference of the new treatment

over the existing one. Investigators may use other methods

to describe the size of a treatment effect, such as an odds

ratio. Point estimates of effect like odds ratio can often be

misleading unless accompanied by a measure of precision, such

as confidence intervals. Many claims about new drugs are based

on trials having surrogate end points, which are variables that

measure rare or distant outcomes of a new drug but are not

a measure of direct clinical benefit. They are commonly used

in clinical trials because the sample size, duration of the

trial and consecutively the cost of conduction of the trial are

reduced. Some studies may present their data based on surrogate

endpoints to show the superiority of the new drug if the

primary objectives of the study are not fulfilled. Physicians should

also be aware of the fact that graphical displays of the outcomes

in promotional literature often fail to convey the complexity

of the data and may distort the findings in favor of the product.Regards

Dr Sanjay Yallappa Choudhari

JR2,Dept of Pharmacology

GMC, Nagpur