Is there another way to take account of noncompliance in randomized controlled trials?

The article discusses non-compliance in the design and conduct of randomized controlled trials (RCTs). Noncompliance can cause problems at the analysis stage. The ideas behind choosing complier average causal effect (CACE) to analyze trials involving noncompliance is presented. The only disadvantage of CACE approach is that it can produce wider confidence intervals than the ITT and per-protocol analyzes. The ITT is suggested as primary and the CACE as a secondary approach to analysis.