Relevance of the number of positive bottles in determining clinical significance of coagulase-negative staphylococci in blood cultures

Coagulase-negative staphylococci (CNS) are the most commonly isolated conta minants from blood cultures, yet they frequently cause true infections. Det ermining the clinical significance of CNS is difficult, and clinicians ofte n consider the number of positive bottles within a set of blood culture bot tles in their assessment. Therefore, in three separate studies, we counted the number of positive bottles within blood culture sets comprising two, th ree, or four bottles in order to predict whether or not CNS were clinically significant isolates (CSI) in adult patients with suspected sepsis. Each c ulture was evaluated by independent, published clinical criteria to determi ne its clinical importance. Of 486 positive sets that included two adequate ly filled bottles, 127 (26%) CNS were CSI, 329 (67%) were contaminants, and 30 (6%) were indeterminate as a cause of sepsis. Among CSI, 39 and 61% wer e isolated from one and two bottles, respectively. The positive predictive value for sepsis was 18% when one bottle was positive and 37% when both bot tles were positive. Of 235 positive sets that included three adequately fil led bottles, 81 (34%) were CSI, 109 (46%) were contaminants, and 45 (19%) w ere indeterminate as a cause of sepsis. Of CSI, 43, 38, and 19% were found in one, two, and three bottles, respectively. The positive predictive value for sepsis was 28, 52, and 30% when one, two and three bottles were positi ve. Of 303 positive blood culture sets that included four adequately filled bottles, 64 (21%) were considered CSI, 197 (65%) were contaminants, and 42 (14%) were indeterminate as a cause of sepsis. Of CSI, 27, 28, 19, and 27% were found in one, two, three, and four bottles, respectively. The positiv e predictive value for sepsis was 11, 30, 34, and 37% when one, two, three, and four bottles were positive. We conclude that the number of culture bot tles positive in a given culture set cannot reliably predict the clinical s ignificance of the CNS isolated and, therefore, should not be used as a cri terion for determining whether or not an isolate represents true infection or contamination.