Does Ascertainment Bias Affect Reports on the Incidence of Multidrug-Resistant, Community-Associated Methicillin-Resistant Staphylococcus aureus Infection?

IN RESPONSE:

We thank Dr. Katz and colleagues for their concern regarding ascertainment bias of multidrug-resistant USA300 cases in our population-based survey. We reason that ascertainment bias would similarly affect the incidence estimates of multidrug-resistant and non–multidrug-resistant USA300 infections because they are indistinguishable to patients and clinicians. When compared with our previously published spatial distribution of multidrug-resistant USA300 infections according to San Francisco ZIP codes, non–multidrug-resistant USA300 infections had a markedly different distribution, not clustering around the 8 ZIP codes with higher proportions of male same-sex couples (see Figure from our article). In fact, the incidence of non–multidrug-resistant USA300 infections was 339 cases per 100 000 persons in ZIP codes with higher proportions of male same-sex couples, compared with 192 cases per 100 000 persons in the other 18 ZIP codes (relative risk, 1.8 [95% CI, 1.6 to 1.9]). In contrast, 59 cases of multidrug-resistant USA300 infections per 100 000 persons occurred in the 8 ZIP codes with a higher proportion of male same-sex couples compared with 4 cases per 100 000 persons in the other 18 ZIP codes (relative risk, 16.1 [CI, 9.8 to 26.5]).

The Castro District (ZIP code 94114) had a disproportionate burden of 170 cases of multidrug-resistant USA300 infection per 100 000 persons when compared with the citywide incidence of 26 cases per 100 000 persons. A correspondingly disproportionate burden of non–multidrug-resistant USA300 infections was not observed when comparing the incidence of non–multidrug-resistant USA300 infection in the Castro District of 571 cases per 100 000 persons with the citywide incidence of 251 cases per 100 000 persons (Figure). Taken together, ascertainment bias is unlikely to explain the clustering of multidrug-resistant USA300 infections in ZIP codes with higher proportions of male same-sex couples. More important, our cross-sectional studies from San Francisco and Boston also indicated that men who have sex with men have an increased risk for infection with multidrug-resistant USA300.