Here's a good summary that provides some useful insights:
1) There is precious little peer-reviewed clinical trial data. Much of the political and regulatory support is based on patient testimonials and the luster of tax revenue.
2) Dosing is highly variable and dependent on a mix of over a hundred active ingredients, some of which are intentionally manipulated to develop different plant strains.
3) A marijuana pill has been approved by the FDA, but typically goes unmentioned by advocates. Small wonder, since smoking weed allows the user to not only titrate any medical effects, but the euphoria that goes along with them.
4) Absent clinical trial data, short and long term harms are also largely unknown. There are worrisome reports of structural brain changes, decline in IQ, mental illness and respiratory disease. Legalization would further increase the public's perception of safety.
5) FDA involvement is minimal. If contamination occurs (pesticides, herbicides or fungal infestation), there is little hope of a recall.
The authors conclude with the usual academic call for more research. The Population Health Blog wholeheartedly agrees.
The PHB also predicts the population health vendors and their outcomes registries may become an important factor in better understanding the role of medical marijuana in the management of chronic illness. In the meantime, an evidence-based approach would suggest that until we have better data, informed skepticism should prevail in the course of patient coaching and decision-making.
Image from Wikipedia