MRSA Dx Tool in Osteomyelitis Called Faulty
Published: February 09, 2012
Reviewed by Zalman S. Agus, MD; Emeritus Professor
University of Pennsylvania School of Medicine.
SAN FRANCISCO -- A common clinical method for gauging risk that theStaphylococcus aureus causing a child's osteomyelitis will be methicillin-resistant is unreliable, a researcher said here.
The so-called Boston algorithm may have been accurate when and where it originated, but it was no help at all in assessing MRSA risk more recently in a major pediatric hospital in Arizona, according to a poster presented at the American Academy of Orthopaedic Surgeons annual meeting.
M. Wade Shrader, MD, of Phoenix Children's Hospital, said that children with osteomyelitis showing all four of the algorithm's criteria had the same risk of confirmed MRSA infection -- 50% -- as those with just one of the factors.
He and his colleagues concluded that genetic markers in clinical samples "may be the best early test to identify MRSA infections."
The algorithm in question was published in 2001 by Mininder Kocher, MD, MPH, of Children's Hospital in Boston, and colleagues, based on a series of pediatric patients evaluated there.
It identified four factors as highly predictive of MRSA versus methicillin-sensitive staph infections in pediatric osteomyelitis:
- Temperature above 38° C
- Hematocrit below 34%
- White blood count greater than 12,000 per microliter
- C-reactive protein above 13 mg/L
Kocher and colleagues found that 91% of pediatric osteomyelitis patients with all four factors were infected with MRSA as opposed to methicillin-sensitive S. aureus, whereas only 1% of those with a single factor had MRSA infections.
But the MRSA label covers multiple S. aureus strains, the distribution of which varies widely from one region to another. Moreover, individual manifestations of infection may vary from one person to another even when the same strain is involved.
Consequently, Shrader and colleagues examined records of 58 children seen recently in their hospital to determine whether the algorithm would apply in the population it serves. Only patients with culture-based confirmation of methicillin resistance status were included in the analysis.
They found that the Boston algorithm was essentially useless in the patients seen in Phoenix.
Half the children meeting all four criteria had confirmed MRSA, as did half of those with only one factor.
Among those meeting three of the Boston criteria, MRSA was confirmed in 42%; it was confirmed in 21% of those meeting two criteria.
Chad T. Price, MD, a pediatric orthopedic surgeon in Orlando, Fla., who was not involved in the study, commented that the desire for a predictive algorithm is understandable.
"It is nice to begin to predict MRSA because you might add vancomycin or some other medication at the beginning," he said.
Unfortunately, he said, the Phoenix study confirms the geographic and temporal variation in MRSA manifestations.
"Basically, what [the study] says is you can't predict," Price said.
In his clinic, he said, skin lesions are often a clue to MRSA in children with septic joints -- at least, to community-acquired MRSA strains that frequently carry the Panton-Valentin leukocidin toxin. He added that he had heard of vancomycin being administered as a matter of course in all children with osteomyelitis.
The study had no external funding.
Shrader and Price indicated they had no relevant financial interests.
Primary source: American Academy of Orthopaedi
