The concepts behind limitations of intention to treat (ITT) analysis are interesting but at the same time too complex for an average practising clinician. As rightly pointed out by Altman,[1] trial investigators often do not describe how they analyzed data on patients who discontinued the study. It may also be argued that current researchers are of the opinion that their research observations are close to perfect since they are supported by advanced statistical methods. In contrast, researchers before the 1950s considered all possible alternative conclusions for their study, and I contend that clinicians were more comfortable reading these older yet more transparent research articles than the current sophisticated yet confusing ones. The worrying part for the clinician are contradictory observations on a similar topic of interest by studies done within a span of 5 years.As a clinician I look at the intention to treat analysis in a much simpler way. Let’s presume that in a randomized controlled trial with 2000 participants in each arm, 200 patients (10%) are lost to follow up in the intervention arm and 20 (2%) are lost to follow-up in the control arm, and that the authors conclude that the intervention resulted in favorable outcome based on ITT analysis with statistically acceptable modifications. The question from the clinician will be “Does this conclusion hold good if the unfavorable outcome occurred to all the 200 patients who were lost to follow-up in the intervention group?”I do not think that this worst scenario is considered in the ITT analysis. This idea may not look logical. But we should remember that the observation made on 2000 patients is going to be applied to entire global populations. Ethically this worst possible scenario should be considered in ITT analysis. If the outcome is in favor of the intervention despite this worst scenario, then probably the conclusion of the study is reasonable. If not, then the observations lead only to doubtful conclusions.Discussion should consider various possible outcomes to patients lost to follow-up and researchers should be encouraged to raise doubts about their observations in the discussion section of the paper so that the scientific community gets maximum benefit from the study. We usually know our mistakes before others identify them! It will be nice if CONSORT drives this initiative.