Posts by James Green
Last ←Newer Page 1 2 3 4 5 Older→ First
-
It was an interesting exercise, but I think I could do with putting some time into some other things : )
My final (hopefully) comment is that I think there is much greater complexity than meets the eye. Prediction is one of the most fraught jobs of a statistician/epidemiologist/public healther. Also, WRT warnings about side-effects, there is some good evidence that if you provide "accurate" information about side effects, people tend to overestimate the risk. -
Y'know how you think people were stopping dying anyway, have a look at these figures (B strain deaths)
1996 7
1997 16
1998 17
1999 11
2000 12
2001 18
2002 5
2003 7
In 1999 and 2000, you could also have been forgiven for thinking the problem was going away. There are some big relative percentage drops, and some huge gains. While it is possible that the 2002/2003 was the end of the epidemic, it also could have jumped up to 16 (like it did from 96 to 97). Having had the vaccine there is no way of knowing. -
In 2004, 4 people (across all age groups) died from the epidemic strain; in 2005, 6 people died fromthe epidemic strain. That's an absolute increase of 50%.
I can't see the figures for the epidemic strain but for all B types it was 5 and 7. However, if you do division to get your percentage increase, that's the relative increase, not absolute. And much as I feel for the families of the 2 people, you'd need to see an increase from 4 to around 8-10 for it to be a statistically significant change.
In 2006, there were 161 cases (the offical report into the disease has yet to be published) of all strains of he disease. That constitutes an epidemic, maybe not to you or I, but according to the MOH. According to the ministry, only 3 cases per 100,000 are required for an epdiemic.
Uh yeah. And in 2005 there were 229, with an incidence of 6.1 per 100,000, so I'd guestimate that the incidence for '06 was down to around 4, and with the trend evident in my graph, it looks like the epidemic will be over, er, next year.
Between 2001 and rollout, total deaths for all age groups declined naturally by about 70 percent while case numbers dropped by about 50 percent.
I assume you're talking about only the specific strain, and that you conviently chose 2001 because it makes the figures look better?
Alas, I was awaiting your analysis of the MeNZB vaccine. Who does that affect? Well, it has been given to more than one miilion NZers but will affect less than 100 of them.
Uh no. There are hundreds not hospitalised. Or does it only affect you if it kills you?
I conclude that since you apparently believe that the numbers don't stack up for the funding of Herceptin
I'm confident that the cost per life saved comfortably exceeds a million dollars, and that there are quite severe side effects for others.
the numbers nonethless look a helluva lot better than they do for the MeNZB vaccine.
The numbers for MeNZB are a bit harder to calculate, but it would appear that if account for both deaths and hospitalisations, MeNZB is probably cheaper than Herceptin. Essentially, breast cancer statistics require a lot less extrapolation that menningoccal B, so there are some things that are unknowable. We'll never know how the disease rates would be if there had been no vaccine, and trend prediction is a difficult and inexact science. I think the call to do the vacccination was a big one, and we'll never really know for sure if it was the right call.
-
not commendable
That related to side-effects. Not deaths. Way to misquote me.
bad science
Uh, where did anyone say 2 actual deaths and MoH 7? And speaking of bad science, no response to my critique of your own bad 'science'? Happy to discuss Herceptin, but if your argument falls over there as well, will you change the topic again?
-
Incidentally, that list of grusome side-effects seems about average. Here's one (at semi-random*) on an ibuprofen (ie Nurofen) formulation for children
http://www.medsafe.govt.nz/profs/Datasheet/f/Fenpaedsusp.htm
-
I'm in the wrong demographic to know anything about what MoH said beforehand, however, 2 things strike me. 1) The sheer number of vaccinations mean that even very uncommon side effects will occur (in a quite noticeable number). 2) Given the misinformation routinely produced by anti-vaccine lobbies, I could see the temptation to err in the other direction by the group trying to get people to accept vaccination. Not commendable, but I can see the temptation.
-
Flash as an ad format lets you do more interesting things than just wink and blink at people - we could run an RSS feed straight into an ad if someone wanted it - but people don't tend to.
Oh, I absolutely appreciate that (and that PA is more broadband). Actually, I'm not sure that I do have PA adblocked at home a) because I mostly read it at work, and b) because the ads aren't annoying slow to load. nzherald.co.nz on the other hand. That is excruciating on dial-up with no ad-blocking.
-
With respect to the ethics of using adblock. If your site has flash ads, and I'm at home browsing it with dial-up, they will be blocked. At work, it's not an issue.
-
I think there are legitimate questions about the way the MeNZB rollout was handled, but these aren't those.
There is definitely some interesting stuff in there, but packaging it in an explicable fashion for public consumption would be pretty difficult. There's an enormous number of different considerations in there.
I'm pretty much ready to give up on this. However, I've prepared a graph of my own to compare with the Ron Law/Barb Sumner one. As you can see here, if you're anti-MeNZb, you'd do well to concentrate on the fatality rate. The number of cases and the number of deaths have both declined a great deal, but if you combine them mathematically, and ignore the concerns I raised about both it's reliability and validity, it's the only way you can spin the figures to look bad.
-
"the number of deaths has increased"
Incorrect. The number of deaths for 04 and 06 are lower than every year since '96; and 05 is lower than all pre-vaccination years except 03.
"Figures for 2006 show that incidence of the disease continues to be at epidemic levels, depsite the stated goal of the MOH to wipe it out."
The incidence in 2003 (pre) was 14.5/100,000. In 2005, it was 6.4. The number of cases in 06 was lower than 05, and continues to trend down. That looks a lot like it's being wiped out.
The number of women who die from breast cancer far exceed the number of people who die from meningococcal disease."
While this is true, it couldn't be less relevant. Herceptin only treats some types of breast cancer, and only helps a small portion of those (approx. 19). Inconveninently, 18 people died from MeB in 2001.
You're wrong. It is an absolute AND relative increase.
Uh, no. You're wrong. The increase they refer to is from 2.4% (2003) to 6.1%. The relative increase is 6.1/2.4=2.54 (254% of original, or a 154% increase). The absolute increase is 6.1-2.4, which is 3.7%.