[Skip to Content]
Access to paid content on this site is currently suspended due to excessive activity being detected from your IP address 23.23.54.109. Please contact the publisher to request reinstatement.
Sign In
Individual Sign In
Create an Account
Institutional Sign In
OpenAthens Shibboleth
[Skip to Content Landing]
Citations 0
Correction
March 18, 2009

Incorrect Data Analysis in: Circumcision Status and Risk of HIV and Sexually Transmitted Infections Among Men Who Have Sex With Men: A Meta-analysis

JAMA. 2009;301(11):1126-1129. doi:10.1001/jama.2009.92

Incorrect Data Analysis: In the Review entitled “Circumcision Status and Risk of HIV and Sexually Transmitted Infections Among Men Who Have Sex With Men: A Meta-analysis” published in the October 8, 2008, issue of JAMA (2008;300[14]:1674-1684), the authors have identified an error that occurred in the spreadsheet computations of the random-effects variance estimate. This had the effect of altering the random-effects summary odds ratios, some of the power estimates, and some interpretation of the odds ratios. Although the values have changed, the pattern of nonsignificant findings remains consistent with the originally published article. Based on the corrected reanalysis, the following changes should be made to the text.

In the Data Synthesis section of the Abstract on page 1674, the second sentence should have read as follows: “The odds of being HIV-positive were not significantly lower among MSM who were circumcised than uncircumcised (odds ratio, 0.95; 95% confidence interval, 0.81-1.11; number of independent effect sizes [k] = 15).” The fourth sentence should have read: “Among MSM who primarily engaged in insertive anal sex, the association between male circumcision and HIV was protective but not statistically significant (odds ratio, 0.71; 95% confidence interval, 0.22-2.28; k = 4).” The sixth sentence should have read: “Neither the association between male circumcision and other STIs (odds ratio, 1.06; 95% confidence interval, 0.97-1.15; k = 8) nor its relationship with study quality was statistically significant (β, 0.265; P = .47).”

In the Results section on page 1677, the third sentence of the fourth paragraph should have read: “The weighted overall effect size reflecting the association between circumcision and HIV infection among MSM was statistically nonsignificant (OR, 0.95; 95% CI, 0.81-1.11; k = 15).” The fifth sentence should have read: “There was moderate to high heterogeneity among the 15 findings (I2 = 61%), which warranted further examination via stratified analyses.” The sixth sentence should have read: “Most of the stratified analyses revealed associations that were not statistically significant (Table 2).”The fourth sentence of the fifth paragraph should have read: “The association between circumcision and HIV infection among the subset of higher-quality studies (determined by the Newcastle-Ottawa Scale) was protective, but statistically nonsignificant (OR, 0.84; 95% CI, 0.57-1.24; k = 5; I2 = 53%).” The first sentence of the sixth paragraph should have read: “A separate analysis (not shown in Table 2) of 4 findings from 3 studies29,33,34 reporting HIV infection and circumcision status for MSM who engaged exclusively or primarily in insertive anal intercourse (n = 2238) was protective, but not statistically significant (OR, 0.71; 95% CI, 0.22-2.28; k = 4).” The fourth sentence of the seventh paragraph should have read: “The weighted overall effect size for the association between circumcision status and STI among MSM was not statistically significant (OR, 1.06; 95% CI, 0.97-1.15; k = 8).”

In the Comment section on page 1679, the first sentence of the first paragraph should have read: “In this meta-analysis of 15 observational studies of the association of circumcision status and HIV infection among 53 567 MSM, the odds of being HIV positive among MSM who were circumcised vs among MSM who were uncircumcised were not statistically different.” On page 1680, the second sentence of the first paragraph should have read: “First, we found a statistically significant protective association for circumcision among MSM in studies conducted before the advent of HAART, but no significant association for studies conducted after HAART.” On page 1681, the fourth sentence of the second paragraph should have read: “The aggregated US studies, where circumcision prevalence is relatively high,62 had no significant reduction in odds of HIV infection among circumcised MSM.” The fifth sentence of the second paragraph should have read: “By contrast, the reduction in odds was higher in developing countries (49%) which have lower adult circumcision rates than the United States.”63 On page 1683, the first sentence of the first paragraph should have read: “We found no statistically significant association between circumcision and HIV infection in our meta-analysis of MSM observational studies, and no statistically significant association between circumcision status and STI.”

The corrected Figure 2, Figure 3, Table 2, and Table 3 appear here.

Figure 2. Overall Effect Size Estimates for Male Circumcision and HIV Infection Among Men Who Have Sex With Men (14 Studies; 15 Findings)
Figure 2. Overall Effect Size Estimates for Male Circumcision and HIV Infection Among Men Who Have Sex With Men (14 Studies; 15 Findings)

HIV denotes human immunodeficiency virus; CI, confidence interval. Odds ratios are from reconstructed 2 × 2 tables and may differ from those in original reports. Odds ratios of less than 1 indicate decreased odds of HIV infection among circumcised men who have sex with men.
aAdditional data were obtained from authors, coauthors, or principal investigators.
bDenotes black participants.
cDenotes Latino participants.

Figure 3. Overall Effect Size Estimates for Male Circumcision and STI (Other Than HIV) Among Men Who Have Sex With Men (7 Studies; 8 Findings)
Figure 3. Overall Effect Size Estimates for Male Circumcision and STI (Other Than HIV) Among Men Who Have Sex With Men (7 Studies; 8 Findings)

HIV denotes human immunodeficiency virus; CI, confidence interval; and STI, sexually transmitted infection.
Odds ratios are from reconstructed 2 × 2 tables and may differ from those in original reports. Odds ratios of less than 1 indicate decreased odds of STI among circumcised men who have sex with men.
aDenotes black participants.
bDenotes Latino participants.

Table 2. Odds of HIV Infection Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
Table 2. Odds of HIV Infection Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
Table 2. Odds of HIV Infection Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
Table 3. Odds of Sexually Transmitted Infections (Other Than HIV) Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
Table 3. Odds of Sexually Transmitted Infections (Other Than HIV) Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
Table 3. Odds of Sexually Transmitted Infections (Other Than HIV) Among Circumcised vs Uncircumcised Men Who Have Sex With Men by Study and Design Characteristics
×