Intention to treat analysis
- "Once randomised, always analysed"
- All enrolled patients have to be a part of the final analysis
- This preserves the bias-protective effect of randomisation
- Contrasts with per-protocol analysis and as-treated analysis:
- Per-protocol analysis: only includes patients who adhered to the protocol; subject to selection bias. May be appropriate for analysis of adverse effects in drug trials.
- As-treated analysis: groups patients according to the treatment they received, rather than the treatment they were allocated to receive. Subject to confounding.
- Modified intention-to-treat analysis: allows the exclusion of patients which never started treatment for whatever reason. Very subjective, and open to selection bias.
- A more reliable estimate of treatment effectiveness
- Prevents bias
- Minimises Type 1 errors (false positives)
- Supported by the CONSORT statement
- When intention-to-treat analysis agrees with per-protocol analysis, it increases the validity of the study
- Treatment effect is diluted (ends up underestimated)
- ITT is inaccurate unless there are negligible protocol violations
- ITT alone is inappropriate for non-inferiority trials