MRSA Nasal PCR Fell Short of Reducing Vancomycin Exposure in ICU Pneumonia
Fear of missing methicillin-resistant Staphylococcus aureus remains one of the most durable drivers of empiric vancomycin use in critically ill patients with suspected community-acquired pneumonia. In practice, that fear often persists long after microbiology suggests a narrower bacterial target. The STOP-Vanc trial tested whether a routine MRSA nares PCR strategy could change that behavior inside the intensive care unit, where diagnostic uncertainty and the pressure to treat early are both amplified.
The randomized design matters here because it separated the test’s laboratory performance from its bedside influence. A biomarker can have strong diagnostic characteristics and still fail to change prescribing if clinicians do not trust it, do not act on it quickly enough, or believe the consequences of undertreatment outweigh the benefit of de-escalation. This trial asked the more operational question: if ICU teams receive MRSA nares PCR testing after admission, does vancomycin come off sooner?
It did not.
High Negative Predictive Value, Limited Change at the Bedside
STOP-Vanc was a pragmatic, prospective, single-center, non-blinded randomized trial involving 277 adult ICU patients with suspected CAP. Participants were assigned 1:1 to usual care with or without the addition of MRSA nares PCR testing after ICU admission. The intervention arm produced a negative predictive value of 98.9%, a result that confirms why this assay has become so attractive in respiratory infection management. When the nasal PCR is negative, the likelihood of MRSA pneumonia is very low, and the test should, in theory, support earlier discontinuation of vancomycin.
Yet the primary endpoint moved only modestly. Vancomycin-free hours alive within the first 7 days were 105.7 in the control arm and 109.7 in the intervention arm, an adjusted difference of 4 hours with a 95% confidence interval from -9.5 to 18.2 and a P value of .458. The trial therefore did not show a statistically significant reduction in vancomycin exposure.
That gap between diagnostic certainty and prescribing behavior is clinically important. ICU clinicians often treat pneumonia before a microbiologic diagnosis is secure because the cost of missing MRSA can be immediate and severe. Once vancomycin is started, stopping it requires confidence not only in the test result but in the timing of the result, the severity of illness, the quality of the respiratory syndrome, and the perceived risk of de-escalation in a fragile patient. A highly specific stewardship tool can still fail if the workflow around it is not equally specific.
Why a Negative Test Did Not Move Prescribing Enough
The biology behind the assay is straightforward. MRSA nares PCR detects colonization patterns that correlate strongly with the absence of MRSA pneumonia, making a negative result useful for ruling out disease in many respiratory settings. In critically ill patients, however, clinical decision-making is not governed by microbiology alone. Ventilator dependence, shock, worsening oxygenation, and broad concern for severe bacterial pneumonia can make clinicians reluctant to narrow therapy even when the pretest probability of MRSA is low.
The ICU environment also compresses decision windows. Empiric therapy is often initiated rapidly, and de-escalation depends on active review of evolving data. If a nasopharyngeal PCR result arrives after the initial treatment plan has already hardened, the test may function more as reassurance than as an actionable trigger. Stewardship tools work best when they are embedded into a system that creates a specific next step, not merely an available piece of information.
This trial suggests that the barrier is not assay performance. It is behavior change.
Mortality Was Unchanged Despite Better Rule-Out Performance
The study also found no reduction in 30-day mortality. That result should not be read as a failure of the assay to identify MRSA absence; rather, it reflects the broader truth that shortening vancomycin exposure is only one part of pneumonia management in the ICU. Mortality in this population is shaped by respiratory failure severity, organ dysfunction, aspiration risk, pathogen burden, and comorbid illness. A de-escalation tool can be microbiologically sound and still have no measurable effect on survival if it does not alter the dominant drivers of outcome.
The absence of a mortality signal is also consistent with the role of MRSA nares PCR as a stewardship instrument rather than a therapeutic intervention. It helps clinicians avoid unnecessary anti-MRSA coverage, but it does not treat infection itself, and it cannot reverse the physiologic consequences of established critical illness. In that sense, the trial tested an information strategy, not a drug.
Stewardship Implications for ICU Pneumonia Pathways
The practical message is narrow but useful. If ICU teams want MRSA nares PCR to reduce vancomycin use in suspected CAP, the test likely needs to be paired with more active antimicrobial stewardship support. Education alone may not be enough. Automatic interpretation guidance, embedded order-set prompts, rapid-result workflows, and real-time pharmacist or stewardship review could all matter more than simply making the assay available.
For hospitals that already use vancomycin liberally in severe pneumonia, the study offers a cautionary lesson about relying on diagnostic tools without redesigning the clinical process around them. A test with a 98.9% negative predictive value can still leave prescribing patterns largely unchanged if clinicians continue to prioritize the fear of undertreatment over the probability of MRSA exclusion. The result is not a reason to abandon MRSA nares PCR. It is a reason to stop assuming that a good test automatically produces a better treatment decision.
In a single-center ICU population, STOP-Vanc showed that MRSA nares PCR is diagnostically reassuring but operationally insufficient on its own. Vancomycin use did not meaningfully fall, and 30-day mortality remained unchanged. The unanswered implementation question is how to turn a reliable rule-out assay into a prescribing change that happens fast enough to matter at the bedside.