Mathematical Modeling
Once efficacy is demonstrated by conducting numerous studies in different geographical areas, program planning and implementation must be examined. Owing to the high prevalence of anemia in most developing countries, such a program must be targeted to all children less than three years of age. For large scale community based intervention, individualized screening procedure for IDA is not possible because: (a) there is no single marker for IDA; and (b) multiple marker studies will be very expensive. Therefore all children will receive iron intervention and a single safe dose is desired that is effective for all at risk children under the age of 3 years. Such an ideal dose is challenging to compute, particularly because of the unique and complex metabolism of iron, and the changing needs of iron in rapidly growing children.
While dosing and frequency issues remain unresolved, the conventional approach to these challenges has been of making best ‘guesses’ about the appropriate dosing regimens that might work, preferably evidence-based, and proposing to conduct a randomized controlled trial to test the efficacy of those regimens. However, limitations to the application of this approach are:
1. RCTs are expensive and time consuming.
2. Statistical analyses with frequentist (classical) approach do not directly answer questions such as: (i) What is the probability that a 15mg or a 45mg dose will prevent anemia in Indian infants; (ii) Does an intermittent dosing regimen yield absorbed amounts of iron in a range of equivalence with a daily dosing regimen.
Dr Waseem Sharieff is exploring (as part of his PhD thesis) a simulation method for modeling clinical outcomes with an evidence-based approach. The proposed approach will incorporate evidence with Markov Chain and Monte Carlo methods to draw probabilistic inferences, thereby enabling the prediction of clinical outcomes in new settings and populations, while quantifying the uncertainty around estimates.
