Friday, January 31, 2014 has more clinical trial data than Pubmed

A recent PLOS publication examined clinical trial results available at and clinical trial data published in journals. Their results demonstrated that there is more trial data available at than what is available in the published literature. Data that was significantly more reported included efficacy, adverse events and serious adverse events. Is anyone using this as a resource for clinical trial data?


Wednesday, January 22, 2014

Conflict of Interest in Systematic Reviews

There is a recent systematic review of systematic reviews in PLOS Medicine examining the effect of industry funding on conclusions. The authors found that industry sponsored systematic reviews are 5 times more likely to support a conclusion of no positive association between sugar-sweetened beverages and obesity or weight gain as compared to studies that did not report industry funding. The authors do not give suggestions for the mechanism by which bias may have entered the systematic reviews other than the heterogeneity of literature selection. Examine systematic reviews carefully!


Sunday, January 19, 2014

Curling Science - Sweeping

I often ponder the physics of curling, and have previously posted on the drag effect here on the bayblab. These musings have yet to improve my game but as the winter olympics approach and as my beer league curling team suffers embarrassing defeats from inebriated senior citizens, I thought I would post some information I found on sweeping.

Sweeping is done by one or more players on the team whose rock is being shot. The players sweep the ice, using specific curling brooms, directly in the path of the travelling rock. Sweeping reduces the deceleration and the degree to which the path of the rock curls.
A common belief is that sweeping melts a thin layer of ice ahead of the rock and that the resulting thin layer of water decreases friction between the rock and the ice. An investigation into ice temperatures resulting from sweeping demonstrated that sweeping raised the ice temperature by 1.5C and achieved a maximum ice temperature of -1C. It was also found that raising the temperature of the ice was crucial for the effectiveness of sweeping but the temperature measurements, as determined by infrared camera, suggested that sweeping doesn't melt the ice. I do not understand the exact methodology or much about infrared cameras but I remain skeptical that a sufficiently thin layer of ice temperature can be measured in this way. It is known that the melting point of ice is decreased under pressure, such as the pressure experienced by ice under the weight of the rock. Therefore I don't know how conclusive it is that the effectiveness of sweeping is not due to contributing to the creation of a thin layer of lubricating water. Nonetheless, it is the momentarily increased ice temperature created by sweeping immediately ahead of the moving rock that makes it effective.
Not surprisingly, the mechanism by which sweeping increases the temperature of the ice is friction. The evolution of materials used for curling brooms reflects this fact, with the latest fabric broom heads enabling a dense contact surface area with the ice. The latest innovation in sweeping brooms is the incorporation of a heat reflective material behind the fabric broom head [link to patent (pdf)]. This reflects some of friction heat back towards the ice to increase the efficiency of sweeping.
None of the above information is going to help one's curling performance, however I did find a review that does have some practical applications for curlers. The first thing that I learned from this review is that it is actually against the rules to not sweep across the entire width of the contact point of the rock with the ice. 'Corner sweeping', as it is known, is when the rock is swept only on one side. By reducing the friction asymmetrically as the rock travels, the amount of curling can be influenced. While 'corner sweeping' is illegal, it is perfectly legal to sweep from either side of the rock. The review points out that more sweeping force is produced on the side closer to the sweeper. Therefore changing the side on which the rock is swept, or having the stronger sweeper on a particular side, might be advantageous for some shots.
Another big question is 'hurry or hard' when sweeping. The review finds that sweeping fast for fast moving rocks and hard for slow moving rocks produces the most friction heat beneath the moving rock. Sweeping fast for fast moving rocks makes sense as the speed of the rock may prevent multiple sweeping passes over the surface area under the rock. When the rock is moving slower it is easy to do multiple passes so it is best to push down hard when sweeping to maximize the heat produced by friction.
In reality none of this information is going to make me a better curler. However, at least the old boys I play with can't say that my poor performance is due to ignorance.