Correspondence analysis to evaluate the transmission of Staphylococcus aureus strains in two New York State maximum-security prisons.

Prisons/jails are thought to amplify the transmission of Staphylococcus aureus (SA) particularly methicillin-resistant SA infection and colonisation. Two independently pooled cross-sectional samples of detainees being admitted or discharged from two New York State maximum-security prisons were used to explore this concept. Private interviews of participants were conducted, during which the anterior nares and oropharynx were sampled and assessed for SA colonisation. Log-binomial regression and correspondence analysis (CA) were used to evaluate the prevalence of colonisation at entry as compared with discharge. Approximately 51% of admitted (N = 404) and 41% of discharged (N = 439) female detainees were colonised with SA. Among males, 59% of those admitted (N = 427) and 49% of those discharged (N = 393) were colonised. Females had a statistically significant higher prevalence (1·26: P = 0·003) whereas males showed no significant difference (1·06; P = 0·003) in SA prevalence between entry and discharge. CA demonstrated that some strains, such as spa types t571 and t002, might have an affinity for certain mucosal sites. Contrary to our hypothesis, the prison setting did not amplify SA transmission, and CA proved to be a useful tool in describing the population structure of strains according to time and/or mucosal site.