Doxycycline use to reduce death in DCS cases

[tarponchik]

Might be more reason to use doxycycline as an antimalarial in areas where it still works for that purpose. Just in case it has other side benefits to reducing malaria risk...

If you try this at home keep in mind that doxy severely increases your sensitivity to UV light. They even tried it as psoriasis treatment.

 

[loosenit2]

Okay great now it's possible an antibiotic can reduce DCS mortality rate when getting to a chamber is delayed. Wouldn't it be a better option to perform in-water recompression at that point?

I think the origin of study was to apply to submariners rather than divers, but there has been a lot of historical cross over in studies involving divers and submariners.

 

[drjtprice]

Apparently it was for both divers and submariners - NMRC Researchers think a Common Antibiotic can Save the Lives of Divers Suffering from Decompression Sickness

duplicates link above, but this is the one that discusses Navy divers as well

 

[Duke Dive Medicine]

I think this is meant to encourage further research. An effect that isn't statistically significant will (or should) not change practice.

Best regards,
DDM

 

[Storker]

Usually it means it did not meet the standard of p<.05

I believe that what @uncfnp meant was "how the heck can they give a conclusion that isn't statistically significant?"

That was my reaction too, BTW. If it isn't statistically significant (typically p<0.05 or whatever you decide is the criterion before you do the study), the proper conclusion is "inconclusive" (which is a fancy word meaning "we can't really say").

 

[drjtprice]

The p value is not the be all and end all of interpretation. Even if it met the p<.05 there is still a 5/100 probability of achieving the result by chance alone. Some scientists have been caught out by 'p milking', to try and meet a threshold that is typically used but by no means certain. That is why I would want to see the numbers. As Duke Dive Medicine says - it is an indicator as to whether more research 'should' be done. So, was it .06?, .10? or .5 (clearly a level for no further research). In terms of health, I think I might want higher than .05 depending on treatment risks versus non-risks. The question of sample size remains. While this plays a role in how easy/hard it is to meet the p threshold it plays other roles as well (statistical power for example). So, yes, you decide the criterion before the study but failing to reach the value does not mean you throw in the towel. It may mean it is time to gather more data, or it may mean give up. Again, with some physics discoveries the threshold is much higher and the scientists will say they think they are on the right track but have not reached the .01, or .001 level. It is an indicator of sources of error (just as it is significant does not mean it might not have occurred by chance).

 

[Storker]

failing to reach the value does not mean you throw in the towel. It may mean it is time to gather more data, or it may mean give up.

I agree. However, what it doesn't mean is that you should say "we got this result but it isn't statistically significant"

 

[Valyngar]

30mg/kg is also massive dosage. For 75kg person it would be 2250mg, about 10 times normal.

 

[afdgf]

This was obviously a preliminary study that was reported as a poster at a conference so it had not been in a peer-reviewed journal. What the poster show is Kaplan-Meier survival curves. If you look at the curves you will see that there are three lines. All lines start at 100% because in survival analysis all subjects are alive at time zero.The curves then show the percent surviving at each time point. You will see that the three curves are different. The one on top show higher survival than the one on the bottom. Presumably the one on the top is the group that was given the doxicycline. At this point in the research, it is just descriptive which is why they are saying that it is not significant. These kinds of preliminary findings are used to justify funding for further research which would then use a sufficient sample size and a long-enough followup to adequately test the hypothesis. This poster is just a preliminary study that is trying to say that the preliminary data suggest that further research may be promising.

 

[afdgf]

You read the curves left to right.