VANCOUVER, B.C. - A simple scoring of five bedside assessments when Clostridium difficile infection is first diagnosed correlates significantly with cure rate.
"The higher the score, the lower the cure rate," said Dr. Mark Miller, head of the division of infectious diseases and chief of the department of microbiology at SMBD–Jewish General Hospital, McGill University, Montreal, who presented the findings.
The five parameters are age, temperature, leukocytosis, albumin, and systemic concomitant antibiotic use, ATLAS for short. The first four are rated on a 0-2 scale; 2 is added to the score if the patient is on systemic antibiotics, 0 if not. ATLAS scores range from 0 to 10 (see box).
The score also correlates with recurrence, but the correlation is not statistically significant.
Dr. Miller said there is a need to be able to categorize patients by C. difficile infection (CDI) severity to determine who should be treated aggressively, assign and assess outcomes in clinical studies, and communicate with other medical workers.
ATLAS Bedside Scoring System
"If someone calls up and says ‘I have a case of moderate CDI,’ it’s pretty much left up to the imagination about what they are talking about," at present, he said.
Although much work has been done previously to create a prognostic system for CDI, proposed systems have not been adequately validated, Dr. Miller said.
However, "if you look at all these publications, it’s all the same risk factors," he added.
So Dr. Miller and his colleagues combined them. "What we came up with was a simple combination of the bedside risk factors that are easy to collect and, we feel, should be most associated with cure and recurrence."
C. difficile strain type was omitted because it’s not usually known at the time of diagnosis; baseline serum creatinine isn’t either, so its elevation above baseline also was excluded.
ATLAS was tested using patient data from a large North American trial comparing fidaxomicin to vancomycin for CDI.
The ATLAS scores of 516 patients were calculated at their time of diagnosis and matched against their cure rates following 10 days of study treatment.
There was "an excellent correlation with cure rate," Dr. Miller said. (R2 0.88, P value less than .001).
Patients with an ATLAS score of 0 had a 98% cure rate; the rate dropped incrementally with higher scores. ATLAS scores of 7 corresponded to a 55% cure rate.
Dr. Miller and his colleagues then checked the 450 subjects cured after treatment to see who had gotten another C. difficile infection.
"With recurrence, the ATLAS score didn’t fair quite so well," he said.
Recurrence rates climbed with higher scores; 11% of patients with a 0 score had a recurrence, 43% with a score of 6.
But the correlation was weak (R2, 0.32) and insignificant (P, .14).
A subgroup analysis found that 229 patients assigned to the vancomycin arm threw the recurrence results off (R2, 0.02, P, .762). ATLAS scores predicted recurrence better in 221 fidaxomicin subjects (R2, 0.70, P, .009).
Recurrence rates in the vancomycin arm were much higher, not neatly distributed along a curve, which might have thrown off the results, Dr. Miller said.
Perhaps, there may also "be some additional refinement of the systemic antibiotics score that would improve" ATLAS’s correlation with recurrence, he said.
A second study presented in Vancouver showed significant correlation between ATLAS scores and 30-day CDI mortality in 308 adults aged 60 years or older.
"ATLAS score appears to ... predict severity in CDI in our patient population," according to the abstract, of which Dr. Miller was a coauthor.
Dr. Miller disclosed that he is a scientific adviser and grant investigator to several pharmaceutical companies, including Merck & Co., Novartis Pharmaceutical, and Optimer Pharmaceuticals, makers of fidaxomicin.