The in-vitro discovery of penicillin mold 3 for bacteria cytotoxicity has saved billions of lives and is an excellent example that justifies the consideration for upward positioning of in-vitro investigations in the evidence hierarchy above expert opinions, ideas, and editorials.4 This agrees with Dr Kokich 2 that the lowest quality of evidence consists of expert opinion, and hence the necessary first building blocks of in-vitro and animal studies should hold a higher position than expert opinions in the evidence pyramid4.

Actual clinical application is not represented at all in the older evidence pyramid diagram used in medicine.4 Thereby, we propose an “evidence iceberg” as a more realistic and complete representation of the hierarchy of investigations to now include the clinical application in orthodontics. Studies of highest relevance correspond to the largest and widest portion of the knowledge base, such as the systematic review (SR) with meta-analysis cross-comparing data, and the randomized controlled trial (RCT) (Fig). In contrast, the pyramid assigns the smallest areas to SRs and RCTs and the widest area to in-vitro studies. The evidence iceberg demonstrates a wide evidence-base of SRs and RCTs as a foundation, keeping afloat our viable clinical applications.

As a final consideration, it is recognized from the application of thalidomide in rats that false-positives can have disastrous effects in humans. However, similar to the conditional usefulness of in-vitro studies, animal studies such as those conducted by Banting et al 5 in beagle dogs for the discovery of insulin and rabbit investigations for Branemark's titanium implants 6 have proven to be highly useful in humans. The results of primate monkeys, with 90% of human proteins, treated with fixed functional dentofacial orthopedic appliances showed histologically significant glenoid fossa and condylar modifications.7 Both of these findings were similarly verified by Ruf and Pancherz 8 using MRI studies in humans. Controlled animal and in-vitro biomaterial studies in orthodontics with the caveat “results in humans do not always have a 1:1 relationship” deserves some greater consideration and attention, above expert opinion, in the evidence iceberg due to their merits and notably similar successes in human applications.9 I thank James Poulos, Ohio State University, for assistance with editing.