Randomization in a hospital-based study.

Sir,

I went through the article entitled “A comparative study of itraconazole in various dose schedules in the treatment of pulmonary aspergilloma in treated patients of pulmonary tuberculosis” published in “Lung India” (2015; 32:342-6) with interest.[1] The authors deserve appreciation for their effort. However, I have a concern regarding this study. The authors state that the study patients randomly received itraconazole either in a fixed dose schedule of 200 mg (group I) 200 mg twice daily (group II) or a variable dose schedule (group III) for 12 months. The authors further state that 60 patients were enrolled, 20 in each group, and that there were no intergroup differences with regard to age, sex, body weight, smoking status, alcohol intake, symptoms, potassium hydroxide (KOH) mount, fungal culture, pattern of radiological lesions, or antiaspergillus antibodies (anti-Asp-Ab) titers.

The authors have actually conducted a clinical trial and not a randomized trial. Conducting a randomized trial in a clinical setting is difficult. Furthermore, there is practical difficulty in conducting randomization on a sample size such as the one used by the authors. Small study samples limit the use of randomization. Randomization in its simplest form is like the tossing of a coin; therefore, the chance of being equally distributed in two groups in a small sample is rare. Now, imagine tossing a coin 60 times. How much probability is there for heads or tails? In the rarest of situations, it will be 30:30. As is apparent, the authors have chosen a deterministic pattern for choosing study participants, which is nonrandom. This pattern is further highlighted in the statement of the authors. The authors state that there are no intergroup differences in the three groups chosen for the purpose of this study. By all probabilities, this is impossible in a random selection of study participants.

To make the point clearer, there are two processes involved in randomizing patients to different interventions. The first is choosing a randomization procedure to generate an unpredictable sequence of allocations. This may be a simple random assignment of patients to any of the groups at equal probabilities; it may be “restricted,” or may be “adaptive.” A second and more practical issue is allocation concealment, which refers to the stringent precautions taken to ensure that the group assignment of patients is not revealed prior to definitively allocating them to their respective groups. Both of these procedures are effective in generating patient distribution capable of yielding unbiased results.

The authors seem to have chosen nonrandom “systematic” methods of group assignment such as alternating subjects between one group and the other. “limitless contamination possibilities” and can cause a breach of allocation concealment.[2]