Animal models have some serious flaws

Medical research, whether focused on pharmaceutical or herbal medicines, often makes use of animal models in the initial phases of study. For many herbal remedies, animal studies are the only data (other than, of course, thousands of years of traditional and experiential knowledge) available to the scientific community, and often these data are used to dismiss phytomedicines as ineffective. In an article in the British Journal of Medicine, researchers identify potential shortcomings of the animal model in drug trials. It comes down to two main conclusions: first, animal studies are often methodologically flawed (not sure why this is, perhaps people pay less attention if only little mice are involved). Second, the animal model has serious predictive deficits and very often has no relevance to actual human therapeutics.
We can speculate endlessly as to why this may be. An interesting thought comes from the European Journal of Cancer: perhaps mice that are genetically engineered to produce "research cancers" exhibit tumors that are substantially different in morphology and behavior from those that develop slowly and naturally in humans (or in pet dogs, as was the focus of this study). These differences may account for the discrepancy between the artificial models and actual human beings. But regardless of the "why", the bottom line remains a challenge to the established phase I, II, III drug trial model. It should also be food for thought for herbalists who are relying on animal models for therapeutic data. And perhaps, if this challenge remains relevant, the entire medical community can start to think about different research strategies that don't involve torturing and killing animals...