This Article  • References  • Full text PDF  • Send to a friend  • Save in My Folder  • Save to citation manager  • Permissions  Citing Articles  • Contact me when this article is cited  Related Content  • Similar articles in JAMA  Social Bookmarking   What's this?

Ronald K. St. John, MD, MPH; Ping Yan, PhD
Laboratory Centre for Disease Control Ottawa, Ontario

JAMA. 1993;270(13):1544.

	Since this article does not have an abstract, we have provided the first 150 words of the full text PDF and any section headings.

To the Editor.

—Von Reyn et al¹ have underestimated the upper limit of the 95% CI (0 to 4.37/100 000 patients). The conclusions of von Reyn et al minimize the risk. Nevertheless, the finding of 0% prevalence among patients recalled and tested in their article and in the companion articles^2,3 is reassuring.

Based an the binomial model, when testing a sample representing 50.7% of the study population (1174 tested of 2317 potentially exposed patients), the chance that there is one positive individual who is not included in the sample is high (50.2%). This probability is precisely calculated by the following: (1-1/n)^0.507n, where n is the population size (2317).

If one HIV-infected patient was missed, then the sero-prevalence in this population would be 4.3 per 10 000 patients (1/2317), or 10-fold higher than the 95% upper CI of von Reyn et al. At the critical 0.05 confidence level, . . . [Full Text PDF of this Article]

CiteULike    Connotea    Del.icio.us    Digg    Reddit    Technorati    Twitter     What's this?