Intention to Treat

 

Epidemiological studies are usually analysed on the basis of intention to treat because this reduces confounding due to non-random dropout from the study. The alternative approach is known as on treatment analysis.

 

Pros of intention to treat analysis:

• reduces confounding due to non-random drop out or non-compliance with the intervention studied
• more pragmatic for answering real world questions because it is the effect of offering the treatment that is analysed as opposed to the effect of taking it

 

Cons of intention to treat analysis:

• reduces statistical power and may thus fail to demonstrate a real effect (wash-in periods are sometimes used to minimise power loss by ensuring tolerability of the intervention and thus reducing non-compliance)