From the National Development and Research Institutes, Inc., New York, New York
Correspondence: Samuel R. Friedman, National Development and Research Institutes, Inc., Two World Trade Center, 16th Floor, New York, NY 10048.
To the Editor:
An article by Giuliani et al1 about hepatitis C transmission appeared in the same issue of Sexually Transmitted diseases (October, Vol. 24, No. 9, 1997) as two articles that discuss validity problems in ascertaining condom use.2,3 Given this context, it is striking that Giuliani et al nowhere question the validity of self‐reports of clients with sexually transmitted disease (STD) that they do not inject drugs‐a behavior that is probably stigmatized far more than unprotected sex.
The authors say little about how they determine whether participants are injecting drug users (IDUs) at intake or follow‐up, just "Information on behavioral risk factors was collected by a standard interview conducted during the HIV pretest counseling session. The interview included questions on demographic characteristics, age at first sexual intercourse, contraceptive and condom use, sexual practices in the past 6 months, and histories of STDs and of injecting drug use" (p. 534). In the United States, several reports suggest that similar questions asked of AIDS patients often lead to under‐disclosure of injection drug use4–6; and Turner et al7 report that the use of audio‐CASI (computer‐assisted self‐interviewing) techniques elicit approximately four times the number of disclosures of other highly stigmatized behaviors (such as having anal intercourse and not having had any sex partners since one was 18 years of age) as do standard interviews.
Why does this matter in studies of hepatitis C transmission? Giuliani et al found that 12 of 709 participants self‐report IDU; and of these, 5 became hepatitis C positive. Among the other 697 participants, 10 became infected. But what if there was underreporting of IDU? If half the IDUs hid their behavior, then we might expect (ceteris paribus) that about 5 of them were among the 10 "non‐IDU" seroconverters; and if two thirds of IDUs failed to disclose this behavior, then the true number of IDUs in the sample would be 36 (of 709, or 5.1%), and they might account for all 15 of the hepatitis C virus (HCV) seroconverters.
It should be added that, since IDUs are probably more likely to be HIV‐positive than other participants ‐ and since they may even be more immunosuppressed than other HIV‐positive participants (depending on number of years of infection) ‐ their findings that HIV is strongly related to HCV seroconversion, and that lower CD4 counts are associated with HCV seroconversion, could simply mean that the observed associations between HIV (and CD4) result from these measures serving as proxies for unreported injection drug use.
1. It is very difficult to guard against the problem of nondisclosure of stigmatized behaviors. Researchers should be aware of this limitation and should discuss it appropriately.
2. Researchers in the field of STD need to be aware that other behaviors, such as injection drug use, may be at least as stigmatized as unsafe sex.
3. Simple sensitivity analyses of possible effects of nondisclosure on findings may be a useful tool. In this case, such an analysis suggests the possibility that the results and discussion of the article by Giuliani et al may (or may not) be reporting as sexual transmission what may be largely or entirely parenteral transmission.
1. Giuliani M, Caprilli F, Gentili G, et al. Incidence and determinants of hepatitis C virus infection among individuals at risk of sexually transmitted diseases attending a human immunodeficiency virus type 1 testing program. Sex Transm Dis 1997; 24:533-537.
2. Turner CF, Miller HG. Zenilman's anomaly reconsidered: fallible reports, Ceteris Paribus,
and other hypotheses. Sex Transm Dis 1997; 24:522-527.
3. Shew ML, Remafedi GJ, Bearinger LH, et al. The validity of self-reported condom use among adolescents. Sex Transm Dis 1997; 24:503-510.
4. Serraino D, Franceschi S, Del Maso L, et al. The classification of AIDS cases: concordance between two AIDS surveillance systems in Italy. Am J Publ Health 1995; 85:1112-1114.
5. Nwanyanwu O, Conti L, Ciesielski C, et al. Increasing frequency of heterosexually transmitted AIDS in southern Florida: Artifact or reality. Am J Public Health 1993; 83:571-573.
6. Murphy JT, Mueller GE, Whitman S. Redefining the growth of the heterosexual HIV/AIDS epidemic in Chicago. J Acquir Immune Defic Syndr Hum Retrovirol 1997; 16:122-126.
7. Turner CF, Forsyth BH, O'Reilly J, et al. Automated self-interviewing and the survey measurement of sensitive behaviors. Computer-assisted survey information collection. New York: Wiley (in press).