Methicillin-resistant staphylococcus aureus nucleic acid amplification test (MRSA NAAT): Evaluation of predictive performance in skin and soft tissue infections

INTRODUCTION

Skin and soft tissue infections (SSTIs) are one of the most common infections in hospital practice, with common causative pathogens including Staphylococcus aureus and Streptococcus species. According to the 2014 Infectious Disease Society of America (IDSA) Clinical Practice Guidelines for the Diagnosis and Management of SSTIs, it is recommended that empiric antibiotic selection should include an agent effective against methicillin-resistant Staphylococcus aureus (MRSA) for moderate-to-severe purulent SSTIs, given the increasing prevalence of antibiotic resistance.^[1]

MRSA nucleic acid amplification test (NAAT), polymerase chain reaction (PCR), or MRSA nasal screening has been a valuable tool for de-escalating vancomycin and is a guideline-recommended tool when it comes to deciding and narrowing antibiotic coverage for community-acquired pneumonia, due to the average negative predictive value (NPV) of 95% per previous research demonstrating strong reliability.^[2] Utilization of MRSA nasal screening has extended out to other indications based on previous research, including for antibiotic management of SSTIs. New studies have shown pharmacist-driven protocols for discontinuation of vancomycin based on negative MRSA NAAT results significantly reduce the inappropriate prescribing of vancomycin, which in turn decreases the risk of toxicity and strain on hospital resources.^[3] Determining whether this practice could be expanded to all infection types would be beneficial for positive patient outcomes and cost savings.

Previous research shows MRSA nasal screening NPV for all SSTIs to be between 80% and 98.4%.^[4-6] Due to the variance in these results, there is some concern that NPV for SSTIs may be less than that of other infections. Due to the short turnaround time for MRSA NAAT results (about 24–72 h) compared to the extended amount of time required for a wound culture and susceptibilities to result (about 96–168 h),^[7] a negative MRSA NAAT result can lead to narrowing antibiotic selection before final culture results revealing MRSA growth, resulting in inadequate coverage of SSTI.

Based on the severity that infections can achieve when improperly covered, especially antibiotic-resistant infections, it is important to determine whether a popular tool, such as MRSA nasal screening, is reliable for guiding management.

MATERIALS AND METHODS

RESULTS

A total of 246 patients were reviewed for inclusion, with 84 patients being excluded, leaving a total of 162 patients included in this study. Patients were excluded for MRSA infection other than SSTI (n = 67), cultures/nasal screening laboratory collected 48+ h after vancomycin initiation (n = 7), and culture/tissues results collected over 48 h apart from each other (n = 10). Figure 1 demonstrates the distribution process of included and excluded patients, Table 2 outlines the complete list of baseline demographics, and Table 3 outlines culture collection details and infection characteristics. Most of our patients were white (75.3%) and of the male sex (59.3%). The ages of our patients were varied, with the majority being between ages 24 and 50 years old (41.4%). MRSA was the most common organism identified by culture (31.5%), and the most often used method of culture collection was a superficial wound culture (76.5%). The majority of infections were identified as purulent (82.7%), and cellulitis was the most common SSTI diagnosis (40.1%).

Close

Figure 1:

Flowchart Diagram of SSTIs pulled for review. SSTI: Skin and soft tissue infections, MRSA NAAT PCR: Methicillin-resistant Staphylococcus aureus nucleic acid amplification test–polymerase chain reaction. provides a full outline of the resultant values discussed above.

Export to PPT

Analysis of the 162 eligible patients included in this study revealed that 92 patients (56.8%) had a negative MRSA NAAT and a negative MRSA culture, 37 patients (22.8%) had a positive MRSA NAAT and a positive MRSA culture, 18 patients (11.1%) had a positive MRSA NAAT and a negative MRSA culture, and 15 patients (9.3%) had a negative MRSA NAAT and a positive MRSA culture. Table 4 provides a full outline of the resultant values discussed above.

Table 5 outlines the results of the calculated statistical testing. After completing statistical calculations, the MRSA NAAT NPV was determined to be 86.0% (95% confidence interval [CI], 77.6–91.7%). In addition, PPV was found to be 67.3% (95% CI, 53.2–79%), with a sensitivity of 71.1% (95% CI, 56.7–82.5%) and a specificity of 83.6% (95% CI, 75.1–89.8%). PLR and NLR were 4.35 (95% CI, 2.75–6.86) and 0.35 (95% CI, 0.22–0.53), respectively.

DISCUSSION

Our primary objective was to assess the NPV of MRSA NAAT or nasal screening for patients with SSTIs. The calculated NPV of 86% of our study is within the values of 80–98.4% that have been demonstrated by previous well-powered studies assessing a similar outcome.^[4-6] Clay et al. retrospectively reviewed the MRSA PCR predictive value in 3595 patients diagnosed with SSTI and found a sensitivity of 97.6% (97.5–98.5), a specificity of 94.9% (94.3–95.7), a PPV of 92.3% (91.4–93.2), and an NPV of 98.4% (98.0–98.8). This study did have a MRSA prevalence of 38.9% (1398 patients), which is similar to the current study; however, the incidence of substance use disorder, a factor that increases the risk of MRSA infection, was documented as 23.1% (830 patients) and associated with a higher NPV of 98.6%.^[5] This factor was not addressed in the current study, but may have been beneficial to review to determine aspects of past medical history that influence MRSA PCR predictive value. The NPV discovered in this study is numerically less than the MRSA NAAT NPV than the previous study and those that have proven to be reliable enough to be included in guideline recommendations, such as pneumonia, which has an NPV of ≥95% per previous studies.^[7,8] Conversely, a retrospective study of 57 patients, specifically with a diagnosis of DFI and a MRSA prevalence of 29.8%, found relatively poor performance of MRSA nasal screening with a sensitivity of 41%, specificity of 90%, PPV of 67%, and NPV of 80%.^[4] Our incidence of DFI was 10 patients (6.2%); however, this previous data suggests that subgroup analysis of these patients would be beneficial with a larger sample size. Overall, with our data providing an NPV on the lower end of previous data, additional judgment and analysis of other patient symptoms or characteristics are necessary to be taken into consideration to avoid premature discontinuation of antibiotic therapy, such as infection type or significant past medical history. Vancomycin is a common agent selected for empiric therapy to manage moderate-to-severe, purulent SSTIs, though it comes with significant risks, including nephrotoxicity, ototoxicity, infusion reactions, and increased incidence of Clostridioides difficile infection. Because a negative test result would likely still be effective for helping to make an ultimate decision to de-escalate, this test may be used as an adjunctive tool to aid in discontinuation of vancomycin in patients at high risk of adverse drug reactions.

The PPV of 67.3% calculated for this study was higher than anticipated, as this has been reported at a lower value of 30–50% in cases of pneumonia.^[8,9] However, the value was similar to a previous study regarding MRSA NAAT utility in SSTI, which reported a PPV of 70%.^[10] Regardless, this value is still within a range to allow for significant error if using it to determine true culture results, and positive MRSA NAAT results are likely to be unreliable. The values of sensitivity and specificity closely mirror the results of NPV and PPV. PLR of 4.35, which, while >1 is <10, indicates that there may be a slight influence on increasing post-test probability of a patient having a MRSA infection after testing positive, though likely not extremely significant. An NLR is 0.35, which while <1 is still >0.1. This indicates that there is a small potential of decreasing the post-test probability of having a MRSA infection after a negative test.

This study does have some limitations. First, our sample size was small, with only 162 patients included. Given the retrospective nature of data collection, the implementation of appropriate collection techniques of both MRSA NAAT PCRs and wound cultures could not be verified, which could result in contamination. In addition, of the methods used to collect wound cultures, the most common collection method was superficial wound cultures, with a reported 124 patients (76.5%) included having their cultures identified using this method. Multiple reports have identified that using superficial cultures rather than deep wound cultures results in decreased reliability of identifying the correct infectious pathogen and likely increases the probability of contamination.^[11,12] Organism growth results were also not consistent with reported bacteria from both past results and previous reports from our local antimicrobial stewardship team.^[13] MRSA was the most common pathogen included in this study, with 51 cultures of the 162 total assessed (31.5%) compared to methicillin-sensitive Staphylococcus aureus (MSSA) growth in 38 cultures (23.5%). Figure 2 demonstrates a brief report of the Providence – Oregon Region incidence of gram-positive pathogens in 2023. According to these reports, the incidence of MRSA is numerically less than MSSA, at 22.5% versus 43%, respectively. The high prevalence of MRSA in this study had the potential to influence results, given there was an increased opportunity for incongruent results between MRSA NAAT and wound cultures. This study did not address the severity of infection and/or blood culture results, which may have provided more data to determine if these factors would influence the reliability of test results. Finally, our study failed to assess the patient’s historical data surrounding MRSA NAAT results and cultures. Given that a positive MRSA NAAT can be indicative of MRSA colonization anywhere in the body,^[14] this information would have been useful to rule out other confounding variables.

Close

Figure 2:

Isolates of Gram-positive pathogens at providence Oregon region 2024. PSVMC: Providence St Vincent Medical Center, PPMC: Providence Portland Medical Center, PMMC: Providence Medford Medical Center, PWFMC: Providence Willamette Falls Medical Center, PNH: Providence Newberg Hospital, PMH: Providence Milwaukie Hospital, PNRMC: Providence Hood River Medical Center, PSH: Providence Seaside Hospital, CLSI: Clinical and Laboratory Standards Institute, CSF: Cerebrospinal fluid.

Export to PPT

CONCLUSION

The MRSA NAAT NPV in patients with SSTI demonstrated in our study was lower than those previously reported. Given that MRSA is a common causative pathogen that can result in serious skin infections, this does raise concern for the reliable use of MRSA NAAT PCR as a sole tool to de-escalate vancomycin to an antibiotic with a narrower spectrum. The current research includes a majority of superficial wound cultures, which may have resulted in more discordance of culture growth, with concern for contamination. Additional data assessing subsets with high-risk past medical history is necessary to determine the clinical implications surrounding the use of MRSA NAAT PCR as a tool in the management of antibiotics used to treat SSTIs.