Richard Jefferys understandably assumed that simple percentages based on raw numbers would be a guide to the direction and limits of the results. While this may be the case for a properly controlled clinical trial, the Lancet study was not a clinical trial. It was an analysis of data from multiple cohort studies in 12 different centers. For such data to be meaningful, sophisticated techniques and adjustments are needed for comparability. Otherwise, the results would be swamped and distorted by arbitrary differences in the size of particular cohorts and the particular biases in their data.

“The adjusted hazard ratios only come into play for the comparison with the reference year that the researchers picked (1998)…”

The above statement by Mr. Jefferys is also in error. Since the adjusted hazard ratios for different years were expressed in terms of the results for 1998 (which equalled “1″), they can be compared proportionally with each other, as is true of any values divided by the same constant ie coding by division (see the early section on coding in any basic textbook on statistics).

Elsewhere, it’s been asserted by Mr. Jeffreys that none of the reported changes over the years of the study were significant. This is also incorrect. The increase in the risk of AIDS events (”including AIDS-related deaths”) from 1998 compared with 2002-2003 went from 1.00 to 1.35 (95% confidence interval: 1.06-1.71). Because the confidence interval starts higher than the 1.00 of 1998, the increase is statistically significant (p

Oh oh, reading this over, Richard, this is all wrong, so we hope that our response “No argument here” can be rescinded. Let’s repeat. Essentially what you have said and continue to say based on the raw percentages is misplaced, because it is the adjusted hazard ratios you have to go by, since you don’t know how each of these groups year by year were made up by other smaller groups with differing characteristics. Maybe one group was 900, and were followed for two months, and no one died. Maybe another group of 100 was followed for the year, and ten died. For the total of 1000, one would not say Aha! 1 per cent died. Nor that 11 out of 1000 was the number to use for comparison. That is what ‘adjusted’ means and represents: the effort we have to make to straighten out the raw data.

Regrettably, you seem to continue to labor under what Walter Kerr called a “delusion of adequacy” in regard to your primitive analysis. Ask a statistician. Ask your colleague in HIV apologia, Chris Noble. Where is Chris Noble, by the way? Have you noticed that he has kept out of it? Why would that be, do you wonder? Because even he cannot rescue you from your error, we would say.

Bottom line, HAART stinks by all indications, and its atrocious effects are reflected by the Lancet study in three respects – 1) a statistically insignificant but visible indication of a sharp rise of 29% in overall mortality when it was introduced, when one would have expected it to go down if HAART did any immediate good, 2) a good indication that that deterioration was maintained, and 3) a statistically significant rise in AIDS symptoms and mortality. The stuff ain’t good for you, Richard, face it. In fact we are told that you have already published your own skepticism in the form of support for immunologically helpful solutions.

And by the way, with such brilliant results in curbing HIV load, and no improvement whatsoever shown in one or even two years in the condition of patients, doesn’t that tell you something about the relevance of HIV to the disease. Or should we spell that i-r-r-e-l-e-v-a-n-c-e?

Link

I also discovered that électrophiles as NO+ was capable of accelerating the opening of the acetalic ring of lamivudine, to release glycolic aldehyde , probably used by the aldehyde réductase to reduce nitroantibiotics, isoxazoles ( bactrim )… (Carey and Sundberg).

Finally, it is possible that lamivudine exerts its antioxidizing action only in an already very “oxidized” cell, where from its weak pharmacological toxicity. The haserd makes sometimes well things.

Wilhem, do you really believe that the risk of disseminated MAC or CMV retinitis is no different in someone with >500 CD4 T cell count versus someone with
No, I don’t really believe that. But my thoughts on the subject run a little deeper. I’m getting the impression that CD4 T-cell counts in healthy subjects run all over the map, both individually and on a time basis. This would be easy to check, by doing counts on a large cohort of healthy subjects.
But I’m afraid that won’t be done for two reasons: 1. Nobody is making a buck off that kind of study. 2. The old lawyer’s credo: “If you think you won’t like the answer, don’t ask the question.”
If CD4 counts vary in a natural way, a low count could mean nothing in terms of susceptibility to infectious disease. But if the immune system is already in bad shape, there is, of course, a structural reason for low T-cell counts.
Anyhow, I see no reason why, if a subject’s CD4 count comes out low, we should destroy his immune system further by administering toxic drugs. Doesn’t it make more sense to build up immunity again? (For example, with lemons, garlic, and olive oil, as Tine van der Maas does in South Africa).
Practicing physicians still assume (correctly) that a high count of white blood cells means there’s an infection going on. Why does HIV/AIDS science put everything on its ear by saying that a high count of a certain sub-class of these cells is suddenly a sign of good health, while a low (resting) count is bad news that requires drastic measures?

John, do you have any thoughts on the mutations in hepatitis B reverse transcriptase that are associated with 3TC resistance?

Well, I’m not John, but I would be very interested too in his views on this subject.
My question: What has a vague virus such as Hepatitis B (Hey! I’m NOT saying it doesn’t exist, OK? Hep-C is the nonexisting one) to do with reverse transcriptase? Is Hep-B now a retrovirus, all of a sudden? What if all these reverse transcriptases were endogenous after all?