[An academic exercise and part of a larger project on authenticity in writing, 11/08.]
[THIS PAPER MAY BE USED IN WHOLE OR IN PART FOR ILLUSTRATION OR INSTRUCTIONAL PURPOSES WITH OR WITHOUT PERMISSION.]
Abstract
A recent Associated Press health news report discusses a new study of a vaccine that already prevents cervical cancer in women. It is called Gardasil, and it has now proven successful in preventing genital warts in men. Here the news report is summarized and is assessed as concluding appropriately. The research study itself is sketched sufficiently to conclude that it relies on standard research design and statistical methods, but there is not enough detail to identify these methods or assess the appropriateness of the study’s findings.
Health News Analysis, An Example
Mike Stobbe (2008), in an Associated Press news report printed in the online version of the Charlotte Observer, discusses a new study of a vaccine that already prevents cervical cancer in women. Gardasil has also proven successful in preventing genital warts in men. The vaccine "targets the two types of HPV, or human papillomavirus, believed to be responsible for about 70 percent of cervical cancer cases, and two other types that cause most genital warts."
Stobbe reports that the study involved "about 4,000 males ages 16 to 26 in nearly 20 countries. Results showed the vaccine was 90 percent effective in preventing genital warts."
Stobbe also notes that research on Gardasil is continuing, but that to date there is no evidence it prevents "penile cancer or other HPV-associated cancers in men. There also is no evidence it prevents men from spreading HPV to women." Right now, Gardasil as it may be used in the U.S. for males appears to prevent an unwanted but benign condition.
Stobbe's report is informative in the style of news reporting. The reader can get additional details once the main points have been covered at the beginning. Following the V structure for news writing allows for a "stay-tuned" kind of conclusion. Stobbe quotes an HPV expert at the U.S. Centers for Disease Control and Prevention as saying that policy makers, presumably the ones charged with passing judgment on Gardasil's availability to males in the United States, "will be looking at a variety of issues."
In spite of the limitations of news reporting and any subtexts one can glean from the article, Stobbe carefully sketches what has and has not happened yet in Gardasil's use and scrutiny for extended uses.
Among the research subjects for the study, a group was given a fake vaccine; and in ninety percent of the vaccinated subjects, Gardasil prevented genital warts. From this we can determine that the "new study" was of the experimental, empirical design type and blind, or double-blind, standard for drug companies doing clinical trials. In addition, quantitative analysis of factors such as three administrations of the drug to 4,000 participants over six months in twenty countries with examinations at various points to determine effectiveness, and so forth, clearly point to a clinical trial and the classic drug company research design.
Without having the specific procedures of the design, however, or the study itself, it is difficult to tell which statistical methods would have been applied other than standard ones for this kind of study. Randomization of the test population, careful data collection from experimental and control groups, pre- and post-assessment for evidence of the target disease would all need standardization over a specific time with subsequent proper analysis and summary.
The difficulty in identifying which statistical methods increases given the rigor needed to conduct a double-blind study the size of this one. The study's population and the different countries and therefore different languages involved seem daunting. It is likely that the study design and statistical procedures would have had to account for errors and inaccuracies.
However, an international drug company such as Merck might well have been able to carry out such a detailed and comprehensive study. It would seem so, for Stobbe reports that experts have found the reported results promising. Presumably they would be able to pass this preliminary judgment based on the study or Merck's reputation, or? (Well, we need not surmise further.)
The above is clearly about the study itself; however, the rest of Stobbe's news report suggests other data sources and statistical procedures. For example, several assertions in the news report suggest a simple-counts methodology was used to set the Gardasil study in context. Stobbe notes that "HPV causes at least 20,000 cases of cancer in the U.S. each year," and "about 40 other countries have approved the vaccine for males." How these numbers were derived cannot be determined by the information given, but how many countries have already approved the vaccine is a simple show of hands. Stobbe's statement of "at least 20,000 cases of cancer" is a conservative one based on other studies and data not specified, but not difficult at least to find. The Centers for Disease Control and Prevention would have or have access to these. Merck would also have considered these in their development of Gardasil or other drugs.
A noteworthy event that happened prior to this news article's appearance is that experts have weighed in on HPV's role in causing cervical cancer. It appears that women who have HPV is positively associated with the likelihood of their getting cervical cancer. Although again the precise measures that were used to weigh in on the side of "causes" of cervical cancer are not given, we can assume a role for inferential statistics, or perhaps meta-analyses of relevant cervical cancer studies.
Finally with regard to the Gardasil study, it relies upon other quantitative studies. For example, here is a likely research question the Merck study had to answer with statistical evidence: At what age should males be vaccinated with the drug? In answering this question, Merck and its research staff, or contractor?, had to address: Given the formula for Gardasil, at what age or above is it safe to administer?
The finding of the Gardasil study as reported by Stobbe is that "the vaccine was 90 percent effective in preventing genital warts, with only 15 cases of persistent infection in the vaccinated group, compared to 101 cases in a group that was given a fake vaccine."
It is not possible to assess whether the finding is appropriate. Not enough of the study is given to see how Merck and researchers got from A to B to C. As suggested above, the study was large and complex, and thus challenges to producing quality research would have been great, but not insurmountable. Lack of information or, better, access to the study itself would be most useful in determining whether the finding of the study is appropriate.
The conclusion of Stobbe's health news report is "we will see." This conclusion is appropriate. Stobbe has provided enough in the way of background information and questions yet to be answered to warrant this assessment. The take-up rate for girls and women in the U.S. for the vaccine (one in four) sounds a cautionary note for the commercial success of the product for males, especially if the warts are benign. In addition, "the vaccine's effect on precancerous lesions," when and if that is established as in the desired direction, might be sufficient reason not only for Merck to push forward its commercial interests in Gardasil, but also for experts and medical professionals to support approval of the drug for males--in the interest of public health.
Reference
Stobbe, M. (2008, November 13). Study: HPV vaccine prevents genital warts in males. Charlotte Observer. Retrieved November 22, 2008 from http://www.charlotteobserver.com/nation/story/337661.html
