Nexus - 1404 - New Times Magazine-pages

Page Content (OCR)

suggesting that herpes zoster be included in active surveillance beginning with the new project cycle starting in January 2000. The CDC then wrote a letter for publication to Vaccine, additionally expressing dissatisfaction with the first three manuscripts but providing criticisms with respect to only one of those manuscripts.'' I happened to be perusing the publicly available online list of manuscripts to be published in Vaccine when I located the CDC's letter and wrote to the editor of Vaccine, stating that it was customary to allow the author an opportunity to provide a related rebuttal. The editor agreed that this was ethical procedure and within several days I submitted a point-by-point rebuttal to each of the specious arguments that the CDC had presented in its letter. My attorney communicated with the CDC legal department, which agreed that my characterisation as a "data manager with no input into the studies" was untrue; and so a correction was made prior to publication of the CDC letter, which appeared with my rebuttal immediately following."* The historical fact was that in addition to developing the entire project's database and overseeing data entry, I implemented numerous statistical programs (including capture-recapture with 95 per cent goodness-of-fit-based confidence intervals, or Cls, as described by Dr E. B. Hook and Dr R. R. Regal), analysed and summarised household contact information, studied breakthrough from those currently presented in the CDC article that I was compelled to publish a fifth manuscript in the peer-reviewed International Journal of Toxicology (July-August 2005).'° This manuscript demonstrated how important trends in the data had been masked in the CDC's article and other presentations concerning the incidence of herpes zoster. wo-p-- ere ee when I located the CDC's letter and wrote to the editor of PART Ill: What the Public Was Not Supposed to Know: Vaccine, stating that it was customary to allow the author an — Chickenpox Vaccine Effectiveness Declines opportunity to provide a related rebuttal. The editor agreed that The CDC's contagiousness article states: "...finding no trend, this was ethical procedure and within several days I submitted a | we conducted subsequent analyses for the 5-year period point-by-point rebuttal to each of the specious arguments that the [1997-2001]". Yet, interesting trends became apparent when CDC had presented in its letter. My attorney communicated with using the exact same database; the vaccine effectiveness was the CDC legal department, which agreed that my characterisation stratified or computed by each year. as a "data manager with no input into the studies" was untrue; and With reference to Table 1, notice that the vaccine efficacy so a correction was made prior to publication of the CDC letter, increased to a high of 95.7 per cent in 1999, and then decreased to which appeared with my rebuttal immediately following." 73.9 per cent and 58.4 per cent in 2001 and 2002 respectively. The historical fact was that in addition to developing the entire The initial increase in vaccine effectiveness or efficacy project's database and overseeing data entry, I implemented demonstrates the "honeymoon" effect, whereby there is a numerous statistical programs (including capture-recapture with relatively brief period following the introduction of the 95 per cent goodness-of-fit-based confidence intervals, or Cls, as chickenpox vaccine during which time vaccinated children described by Dr E. B. Hook and Dr R. R. Regal), analysed and receive additional immune boosting by virtue of their contact with summarised household contact information, studied breakthrough children having natural chickenpox. During 1999 and thereafter, the number of children reported as having natural Table 1: Vaccine efficiency (efficacy) by year, 1997 to 2002, in| chickenpox dramatically declined (along with the households with contacts less than 20 years old and verified cases of | boosting effect), and this contributed to the chickenpox reported in 5-18 year olds, Antelope Valley, California. satically significant downward trend in vaccine . . a effectiveness. Year Vaccine Efficacy Verified Cases There is a legitimate concern that vaccinated (95% confidence interval) of Chickenpox children will be left unprotected and may prove to 1995 —_— 1,290 be susceptible to chickenpox as adults, when 99° 86.7 (75.0- 92.9) Gee chickenpox can be a more serious illness if 1 (75.0-92. ; contracted then. 1998 93.7 (83.2-97.7) 92 omneeeewNN sobe ae ae ae VASP/CDC Reports are Misleading . . . Civen et al.” report an incidence of shingles of 2001 73.9 (57.9-83.8) 442 . 40 per 100,000 among children in Antelope 2002" 58.4 (13.7-79.9) Figure n/a Valley, where vaccine coverage exceeds 80 per * Efficacy based on verified cases reported January through June (half-year) only. cent. These authors also report incidence of shingles of 45 per 100,000 among individuals varicella, lesion severity, vaccine efficacy, seasonal incidence aged 10 to 19 years, a group which is largely unvaccinated but has patterns, outbreaks among schools, and much more. I also had a previous experience of natural chickenpox. Notice there is participated in project presentations and on-site seminars at the hardly a difference in incidence rates between these groups CDC, and provided suggestions regarding methodology. I wrote despite differences in their vaccination status. the preliminary justification for adding herpes zoster to the active There are two fundamental problems with the manner in which surveillance and contributed much more than act as a "data these figures were derived. First, Civen et al. assume that 100 per manager with no input into the studies". cent of shingles cases are voluntarily reported to VASP. This I considered I had certainly fulfilled all my ethical obligations assumption is rarely true in any study. Because there were two and responsibilities. But then the CDC published a new article on reporting sources (schools and healthcare providers), a statistical contagiousness of varicella in household settings.'* This research technique known as "capture-recapture" can be used to quantify article stated that the vaccine was highly effective and that the the percentage of underreporting, which was estimated to be about efficacy (effectiveness) of the varicella vaccine did not 50 per cent. Without adjusting for underreporting, incidence rates significantly change over the period 1997 to 2001. reported by the VASP merely reflect the case ascertainment and Irecalled having collected these data and transmitted them to cannot be compared to other studies. the CDC to permit independent analysis. Early in 2002, I had Second, Civen et al. present what is called an average (or mean) performed a data analysis of vaccine efficacy and placed my of a bimodal incidence rate. To illustrate why this is statistically report in a notebook containing other manuscripts awaiting invalid, consider that when the same data are stratified by review. vaccination status, this yields an estimated rate of 22 cases of Just like the other manuscripts discussing shingles, this shingles per 100,000 among vaccinees and 223 per 100,000 among manuscript was not reviewed and the summary of vaccine children with a previous history of natural chickenpox. Combined, efficacy by year was also deleted from the annual report without the weighted average of these two very different incidence rates is explanation. The results and trends that I had previously derived 52 per 100,000; however, this rate is not representative of the rate and confirmed with a CDC data manager were so very different in either of the two diverse groups. Year Vaccine Efficacy (95% confidence interval) 1995 1996 1997 1998 1999 2000 2001 2002* 86.7 (75.0-92.9) 93.7 (83.2-97.7) 95.7 (82.7-98.9) 85.5 (73.9-92.0) 73.9 (57.9-83.8) 58.4 (13.7-79.9) varicella, lesion severity, vaccine efficacy, seasonal incidence patterns, outbreaks among schools, and much more. I also participated in project presentations and on-site seminars at the CDC, and provided suggestions regarding methodology. I wrote the preliminary justification for adding herpes zoster to the active surveillance and contributed much more than act as a "data manager with no input into the studies". I considered I had certainly fulfilled all my ethical obligations and responsibilities. But then the CDC published a new article on contagiousness of varicella in household settings.'* This research article stated that the vaccine was highly effective and that the efficacy (effectiveness) of the varicella vaccine did not significantly change over the period 1997 to 2001. Irecalled having collected these data and transmitted them to the CDC to permit independent analysis. Early in 2002, I had performed a data analysis of vaccine efficacy and placed my report in a notebook containing other manuscripts awaiting review. Just like the other manuscripts discussing shingles, this manuscript was not reviewed and the summary of vaccine efficacy by year was also deleted from the annual report without explanation. The results and trends that I had previously derived and confirmed with a CDC data manager were so very different 22 = NEXUS JUNE — JULY 2007 www.nexusmagazi ne.com