Discussion

This review identifies that the majority of institutions employing nasal MRSA PCR screens had established protocols allowing pharmacists to order the test without an order from a provider. Nasal MRSA PCR screens are used universally by these institutions to guide pneumonia therapy. Compelling evidence supporting the use of nasal MRSA PCR in pneumonia drives its widespread adoption in Kentucky hospitals’ policies. The performance of nasal MRSA swabs to predict MRSA pneumonia was reviewed in a meta-analysis of 22 studies (11 with nasal MRSA PCR, 4 utilized nasal MRSA cultures, one had both PCR and cultures, and 6 did not specify MRSA screen method) reported sensitivity of 70.9%, specificity of 90.3%, positive predictive value (PPV) of 44.8%, and negative predictive value (NPV) of 96.5% for all types of pneumonia. ^{Reference Parente, Cunha, Mylonakis and Timbrook12} In this analysis, nasal MRSA culture and PCR swabs had higher sensitivities in CAP compared to VAP (85% vs 40.3%). ^{Reference Parente, Cunha, Mylonakis and Timbrook12} PPV of nasal MRSA swab remained low regardless the type of pneumonia but has high NPV for both CAP and VAP (98.1% and 94.8% respectively). ^{Reference Parente, Cunha, Mylonakis and Timbrook12} Due to low PPV and high NPV of nasal MRSA swabs (culture and PCR) across different types of pneumonia, negative nasal MRSA PCR results prove useful for de-escalating therapy in pneumonia but less helpful when the result is positive. Although KASIC did not focus on impact of MRSA PCR protocols at their institutions, as it was outside the scope of this manuscript, other stewardship programs have reported a reduction in vancomycin duration after implementing pharmacist-led MRSA PCR nasal screening protocols ^{Reference Meng, Pourali and Hitchcock10,Reference Marinucci, Louzon, Carr, Hayes, Lopez-Ruiz and Sniffen11} Additionally, in a retrospective cohort study at Stanford HealthCare, no difference in clinical outcomes including hospital length of stays (LOS), ICU LOS, and in-hospital mortality was found despite shortened vancomycin duration in post MRSA PCR protocol period. ^{Reference Meng, Pourali and Hitchcock10} Similar findings were observed in a prospective study at Barnes Jewish Hospital, which used nasal MRSA culture to de-escalate empiric anti-MRSA therapy for critically ill patients with suspected pneumonia. ^{Reference Raush, Betthauser, Shen, Krekel and Kollef13}

Despite extensive evidence supporting the use of MRSA PCR for pneumonia, the evidence for nasal MRSA PCR in patients with complicated lung infections (lung empyema or lung abscess) remains limited. A retrospective cohort study at St’s Mary Medical Center in Wyoming evaluated the performance of nasal MRSA PCR in 93 patients with lung empyema and 46 patients with lung abscess. ^{Reference Lewis, Bridwell, Hambuchen, Clay and Orwig14} This study found 80% sensitivity, 84% specificity, 42% PPV, and 97% NPV for empyema and 0% sensitivity, 90% specificity, 0% PPV, and 90% NPV for lung abscess. ^{Reference Lewis, Bridwell, Hambuchen, Clay and Orwig14} The use of nasal MRSA PCR to rule out clinical MRSA infection in adult patients with complicated lung infections warrants caution.

A limited institutions in Kentucky adopted MRSA PCR for non-pneumonia indications, partly due to limited data. ^{Reference Carr, Daley, Givens Merkel and Rose9} Among non-pneumonia indications, the data is most consistent for intra-abdominal infection with a few retrospective cohort studies reported NPV of 97%. ^{Reference Carr, Daley, Givens Merkel and Rose9} The performance of MRSA PCR for SSTI is more heterogeneous depending on its location (non-extremity vs extremity), presence of ulcers, and presence of prosthetic implant. ^{Reference Carr, Daley, Givens Merkel and Rose9} There are no independent studies assessing performance of nasal MRSA PCR in urinary tract infection (UTI); however, a subgroup analysis of patients with urinary cultures at Veterans Affairs inpatient facilities showed 99.2% NPV. ^{Reference Carr, Daley, Givens Merkel and Rose9,Reference Mergenhagen, Starr, Wattengel, Lesse, Sumon and Sellick15} It should be noted that MRSA is extremely uncommon cause for UTI, and thus the use of nasal MRSA PCR is likely of low benefit. ^{Reference Carr, Daley, Givens Merkel and Rose9} No studies have investigated the performance of nasal MRSA PCR in CNS infections. ^{Reference Carr, Daley, Givens Merkel and Rose9} Given high mortality risk of CNS infections and the use of anti-MRSA agent to target other non-MRSA pathogens (such as drug-resistant Streptococcus pneumoniae), the use of nasal MRSA PCR cannot be recommended at this time. ^{Reference Carr, Daley, Givens Merkel and Rose9} This also aligns with practice in Kentucky, as no institutions endorsed nasal MRSA PCR screen to guide therapy for CNS indications.

An important aspect of MRSA PCR protocol/guidance is determining the utility of nasal MRSA PCR in patients receiving anti-MRSA therapies or MRSA de-colonization. Only two institutions in Kentucky limited the ordering of nasal MRSA PCR to within 48–72 hours of initiation of anti-MRSA antibiotics whereas another institution suggested avoiding using nasal MRSA PCR screen in patients undergoing de-colonization within last 30 days. This was in part due to scarce data examining durability of nasal MRSA PCR positivity for patients receiving anti-MRSA antibiotics or undergoing de-colonization. A study at Upstate University Hospital demonstrated that a positive nasal MRSA PCR typically persisted after 48–96 hours of continuous systemic anti-MRSA antibiotics. ^{Reference Wu, Kufel, Riddell, Steele and Asiago-Reddy16} Additionally, a retrospective observational cohort study showed reduced NPV of nasal MRSA PCR screen for pulmonary infections after mupirocin administration. ^{Reference Chaudhry, Allen, Paylor and Hayes17} Decisions to order nasal MRSA PCR screen and de-escalate based on negative result in patients receiving anti-MRSA therapies and/or MRSA de-colonization should be individualized based on exposure duration of anti-MRSA antibiotic and/or topical decolonization (ie, negative nasal MRSA PCR in patients received less than 48 h of anti-MRSA therapy is more reliable than those received a prolonged course of antibiotics), clinical severity, and patient risk factor for MRSA.

Another aspect in MRSA PCR protocol/guideline to deliberate is the necessity of repeating nasal MRSA PCR screening. Two institutions excluded patients who had previous nasal MRSA PCR result or MRSA in respiratory cultures within 7–14 days from their pharmacist nasal MRSA PCR protocols. A study of 736 patients from a community health-system examined NPV of nasal MRSA PCR based on the time between nasal PCR and respiratory samples. NPV of nasal MRSA PCR screen remained high regardless of the time elapsed between nasal MRSA PCR and respiratory culture collection. ^{Reference Turner, Seligson, Parag, Shea and Hobbs18} The NPVs were 93.8% in ≤ 24 hours group (n = 316), 98.6% in 25 to 48 hours group (n = 80), 95.7% in 49 hours to 7 days group (n = 190), 92.9% in 8–14 days group (n = 99), and 95.5% in >14 days group (n = 51). ^{Reference Turner, Seligson, Parag, Shea and Hobbs18} Another study found conversion rate from negative nasal MRSA PCR to positive within 14 days was rare (0.9%). ^{Reference Earl, Greenlee, Fong, Imlay and Spivak19} These data suggest that repeating nasal MRSA PCR within 14 days of a prior test is generally unnecessary, and may serve as an additional factor to consider when developing nasal MRSA PCR screen protocol/guideline. ^{Reference Turner, Seligson, Parag, Shea and Hobbs18,Reference Earl, Greenlee, Fong, Imlay and Spivak19}

The use of nasal MRSA PCR has become a widely adopted tool in antimicrobial stewardship, and guidelines and protocols are critical factors in its successful integration. Allowing all pharmacists—not just those in stewardship roles—to order nasal MRSA PCR can streamline the process and ensure quicker results. No institutions in Kentucky permit pharmacists to independently discontinue anti-MRSA therapy based on a negative nasal MRSA PCR screen. Future protocols can be designed to enable pharmacists to do so for specific indications, such as uncomplicated pneumonia for which data has consistently demonstrated high NPV. More expansive de-escalation criteria may allow pharmacists to facilitate timely clinical interventions in partnership with treating teams.