From the Research
The negative predictive value of a negative MRSA culture from the nose, axilla, and groin for MRSA infection is approximately 98.5%. This means that if a patient has negative surveillance cultures from these three sites, there is a 98.5% chance that they truly do not have an MRSA infection, as supported by a recent study 1. The high negative predictive value depends on proper collection technique, appropriate culture methods, and the absence of recent antimicrobial therapy that could suppress bacterial growth.
Key Points to Consider
- The nose, axilla, and groin are the most common sites of MRSA colonization, and sampling all three sites significantly increases the sensitivity of detection compared to sampling just one site.
- A negative MRSA culture from these sites can be used to rule out MRSA infection with a high degree of confidence, but it is not perfect and additional testing may be warranted in high-risk patients or those with strong clinical suspicion for MRSA infection despite negative surveillance cultures.
- More sensitive molecular testing methods like PCR may be useful in certain cases, especially when there is a high clinical suspicion for MRSA infection despite negative surveillance cultures.
Evidence Summary
The study by 1 found that MRSA PCR nasal swabs had a negative predictive value of 98.5% for any MRSA infections, supporting the use of negative MRSA cultures from the nose, axilla, and groin to rule out MRSA infection. Other studies, such as 2, 3, 4, and 5, also provide evidence on the sensitivity and specificity of MRSA screening methods, but the study by 1 provides the most recent and highest quality evidence on this topic.
Clinical Implications
In clinical practice, a negative MRSA culture from the nose, axilla, and groin can be used to inform decisions about empiric antibiotic therapy and infection control measures. However, it is essential to consider the patient's individual risk factors and clinical presentation when interpreting the results of MRSA surveillance cultures.