We thank Dr O'Farrell for his comments  that essentially address two points: whether HIV incidence in men with smaller foreskin surfaces areas (<75th percentile)  are lower than those observed among circumcized intervention arm men in our trial , and whether screening and selectively targeting men with larger foreskins might be of utility in service programs. We will address these two issues.
It is important to note that the study of HIV acquisition associated with foreskin surface area was observational, and included HIV infections among men prior to their enrollment into the randomized trial . Thus, it is difficult to directly compare the incidence in these men to the incidence observed during the trial. Furthermore, in Table 1 of our study , we found that foreskin surface area increased with age and, in Table 2, we found that HIV incidence was higher among older men. The latter observation of increasing incidence with age was also reported in our trial Table 4 . Thus, confounding by age could affect direct comparisons between the observational and trial data. To address the issue raised by Dr O'Farrell, we would really need to analyze the trial data to assess the efficacy of circumcision by strata of foreskin surface area, and these analyses have not been done. Therefore, it would seem unwise to base policy or program procedures on a possibly confounded comparison between different studies.
Dr O'Farrell rightly draws attention to his studies on foreskin length  that we should have cited in our study, and implies that this metric could be a surrogate for surface area, and could be measured in programs to prioritize surgeries for men with longer foreskins. We found that both foreskin length and circumference were related to surface area, and assessed the associations between HIV acquisition with either foreskin length or circumference. The incidence rate ratio of HIV acquisition associated with foreskin length in centimeters was 1.16 [95% confidence interval (CI) = 0.94–1.42, P = 0.16] and not statistically significant, whereas the association with foreskin circumference was 1.19 (95% CI = 1.03–1.38) that was statistically significant (P = 0.02). Thus, measuring foreskin length alone may not be predictive of HIV risk and measurement of circumference in vivo would be difficult. Therefore, we do not think that measurement of foreskin length could be used to prioritize circumcision for men with longer foreskins. Moreover, our experience with over 10 000 surgeries suggests that preoperative measurement of foreskin length would add to the burden of service providers, and could lead to discriminative practices disadvantaging younger men. In conclusion, we would recommend providing a timely service to all men requesting circumcision as an optimum strategy.
1. O'Farrell N. Foreskin surface area and HIV acquisition. AIDS 2010; 24:786–787.
2. Kigozi G, Wawer M, Ssettuba A, Kagaayi J, Nalugoda F, Watya S, et al
. Foreskin surface area and HIV acquisition in Rakai, Uganda (size matters). AIDS 2009; 23:2209–2213.
3. Gray RH, Kigozi G, Serwadda D, Makumbi F, Watya S, Nalugoda F, et al
. Male circumcision for HIV prevention in men in Rakai, Uganda: a randomised trial. Lancet 2007; 369:657–666.
4. O'Farrell N, Chung CK, Weiss HA. Foreskin length in uncircumcised men is associated with subpreputial wetness. Int J STD AIDS 2008; 19:821–823.