Empiric therapy for patients with pneumonia is not as clear cut as one would think. Guidelines tend to differ in terms of when to start empiric MRSA coverage, with some guidelines recommending initial therapy that covers MRSA in those who are admitted to the ICU (1), while other guidelines suggest to start MRSA therapy if there are certain risk factors present (2). Recent data suggest the incidence of MRSA pneumonia in those being admitted to the hospital is lower than previously thought, and empiric MRSA therapy may be associated with higher mortality.
For instance, a multicenter, prospective study (3) found the prevalence of Staph aureus CAP was 1.6% in a cohort of 2259 patients, of which only 15 (or 0.7%) were due to MRSA:
Not surprisingly, those with Staph aureus CAP had higher morbidity and mortality when compared to non-staph aureus CAP patients, but should be noted these patients were much sicker up front, with 59.5% of patients with Staph Aureus CAP having a PSI of at least 4 or 5 compared to 34.3% in the non-staph aureus group.
Another finding was the number of patients who were given MRSA coverage in those who had non-staph CAP was about 30% in this cohort:
Another study (4) found the rate of MRSA coverage was increasing for those admitted with CAP, with vancomycin use increasing from 16% to 31% over a one year period:
A recent retrospective cohort study (5) evaluated the 30d disk of death in patients who received empirical anti-MRSA therapy who were admitted for community acquired pneumonia. Again, those who received MRSA therapy had greater number of co-morbidities and a greater number of risk factors for MRSA colonization, including prior MRSA infections, prior hospitalization, and therapy with prior antibiotics:
Mortality was significantly higher in the unadjusted analysis in the MRSA therapy group, however after propensity matching, anti-MRSA therapy was correlated with higher mortality across all patients with the exception of those who were MRSA culture positive:
Admittedly, there is a lot of bias that propensity scores are unlikely to eliminate, but this represents likely the highest quality evidence we will get on the subject. Because of the overall low prevalence of MRSA as an etiology of CAP, and the potential risk of mortality of empiric MRSA therapy, many tools have been implemented to guide empiric therapy. I won’t cover one of them in this post (screening nares PCR, that is a beast in and of itself), but will cover a less well known scoring system to risk stratify patients according to their MRSA risk: the Shorr score.
A retrospective, multi-center analysis across 62 hospitals (6) was done to develop a score to risk stratify patients for their risk of MRSA pneumonia. Nosocomial pneumonia was excluded. The cohort was divided into 2; two thirds was used in the development cohort and the remaining one third was put in the validation cohort. 5975 patients were included, and the prevalence of MRSA pneumonia was 14% (837 patients). Not surprisingly, patients with MRSA had more co-morbidities and were generally sicker, with more of those patients being admitted to the ICU:
A logistic regression procedure was done to arrive at the 8 parameters that made up the 8 variables that made up the score:
The prevalence of MRSA rose with higher scores, with prevalence being <10% if score was </= 1, and >30% if >/= 6.
A score <1 vs greater than 1 had a sensitivity and specificity of 59.1% and 60% respectively (positive likelihood ratio 1.48, negative likelihood ratio 0.68). PPV was 19.2% but NPV was 90.1%. With the prevalence of MRSA in the overall cohort, the post-test probability with a score of <1 was roughly 10%.
The Schorr score was validated in 2 studies. In a retrospective single center cohort study (6), 134 patients with MRSA pneumonia where evaluated along with 134 with non-MRSA pneumonia. Similarly to the initial cohort, the MRSA patients tended to be sicker with more co-morbidities than the non-MRSA group. A risk score of >1 had a 93% sensitivity and 55% specificity for MRSA pneumonia with PPV of 68% and NPV of 89%. This translated roughly to a negative likelihood ratio of 0.13 and a positive likelihood ratio of 2.07 for a threshold of 1. Importantly, despite the presence of MRSA, in those with low risk, the mortality for MRSA and non-MRSA pneumonia was similar when compared to higher risk groups, though MRSA pneumonia tended to take longer prior to stabilizing:
Another VA retrospective, population based cohort study (8) also validated the Shorr score, but actually used all-cause 30d mortality as the primary outcome. They used multivariable logistic regression models to see if MRSA therapy was associated with 30d mortality. 80,330 patients were included, with 36% receiving MRSA therapy. The rate of MRSA isolation in the MRSA therapy group was 5.1% vs 2.3% in the non-MRSA group. Notably, 30d patient mortality increased with rising risk scores:
And MRSA therapy was associated with higher mortality in the low risk score patients, compared to the high risk patients, where it was associated with lower mortality:
This suggests that empiric MRSA therapy for all comers may not be best practice, and in fact it may be dangerous. The Shorr score may help with the clinical decision making, in conjunction with MRSA PCR nares, for empiric therapy. In other words, those who clinically do not look like they have MRSA pneumonia and have low risk scores, then maybe foregoing therapy is a reasonable choice here. Of course, clinical context matters. Withholding vancomycin from a young IVDA with multiple pulmonary lesions on their CXR may not be best practice.
- Liu C, Bayer A, Cosgrove SE, et al. Clinical practice guidelines by the infectious diseases society of america for the treatment of methicillin-resistant Staphylococcus aureus infections in adults and children [published correction appears in Clin Infect Dis. 2011 Aug 1;53(3):319]. Clin Infect Dis. 2011;52(3):e18-e55. doi:10.1093/cid/ciq146
- Kalil AC, Metersky ML, Klompas M, et al. Management of Adults With Hospital-acquired and Ventilator-associated Pneumonia: 2016 Clinical Practice Guidelines by the Infectious Diseases Society of America and the American Thoracic Society
[published correction appears in Clin Infect Dis. 2017 May 1;64(9):1298]
[published correction appears in Clin Infect Dis. 2017 Oct 15;65(8):1435]
[published correction appears in Clin Infect Dis. 2017 Nov 29;65(12):2161]. Clin Infect Dis. 2016;63(5):e61-e111. doi:10.1093/cid/ciw353
- Self WH, Wunderink RG, Williams DJ, et al. Staphylococcus aureus Community-acquired Pneumonia: Prevalence, Clinical Characteristics, and Outcomes. Clin Infect Dis. 2016;63(3):300-309. doi:10.1093/cid/ciw300
- Jones BE, Jones MM, Huttner B, et al. Trends in Antibiotic Use and Nosocomial Pathogens in Hospitalized Veterans With Pneumonia at 128 Medical Centers, 2006-2010. Clin Infect Dis. 2015;61(9):1403-1410. doi:10.1093/cid/civ629
- Jones BE, Ying J, Stevens V, et al. Empirical Anti-MRSA vs Standard Antibiotic Therapy and Risk of 30-Day Mortality in Patients Hospitalized for Pneumonia [published online ahead of print, 2020 Feb 17]. JAMA Intern Med. 2020;180(4):552-560. doi:10.1001/jamainternmed.2019.7495
- Shorr AF, Myers DE, Huang DB, Nathanson BH, Emons MF, Kollef MH. A risk score for identifying methicillin-resistant Staphylococcus aureus in patients presenting to the hospital with pneumonia. BMC Infect Dis. 2013;13:268. Published 2013 Jun 6. doi:10.1186/1471-2334-13-268
- Minejima E, Lou M, Nieberg P, Wong-Beringer A. Patients presenting to the hospital with MRSA pneumonia: differentiating characteristics and outcomes with empiric treatment. BMC Infect Dis. 2014;14:252. Published 2014 May 10. doi:10.1186/1471-2334-14-252
- Teshome BF, Lee GC, Reveles KR, et al. Application of a methicillin-resistant Staphylococcus aureus risk score for community-onset pneumonia patients and outcomes with initial treatment. BMC Infect Dis. 2015;15:380. Published 2015 Sep 18. doi:10.1186/s12879-015-1119-1