By Carol A. Kemper, MD, FACP
Clinical Associate Professor of Medicine, Stanford University, Division of Infectious Diseases, Santa Clara Valley Medical Center
SOURCE: Peinado-Acevedo JS, Hurtado-Guerra JJ, Hincapie C, et al. Validation of VIRSTA and predicting risk of endocarditis using a clinical tool (PREDICT) scores to determine the priority of echocardiography in patients with Staphylococcus aureus bacteremia. Clin Infect Dis 2021;73:e1151-e1157.
These authors at two different facilities in Medellin, Columbia, compared two predictive scoring methods, VIRSTA and PREDICT, in the evaluation of 922 patients with Staphylococcus aureus bacteremia (SAB). The results were retrospectively compared in their ability to predict infective endocarditis (IE) according to modified Duke criteria and the need for echocardiography. Recall that modified Duke criteria employ a mixture of two major criteria (organism consistent with an IE-type bacteria and an abnormal echocardiogram), and five minor criteria (fever, predisposing heart condition, injection drug use, vascular or immunologic phenomena consistent with a diagnosis of IE, and serologic evidence of an IE-type organism).
PREDICT was developed at the Mayo Clinic based on a retrospective cohort of 678 patients. The VIRSTA scoring system was developed in France, based on a prospective evaluation of adults with SAB presenting to eight different university healthcare centers. VIRSTA scoring is based on 10 criteria assessed at 48 hours from the first positive blood culture, including cerebral or pulmonary embolism, meningitis, permanent indwelling cardiac devices, history of IE, injection drug use, pre-existing valvular disease, vertebral osteomyelitis, community acquisition, severe sepsis or septic shock, elevation of C-reactive protein, and (subsequently) persistence of bacteremia for > 48 hours of antibacterial therapy. A score of 3 or more points is considered high risk for IE, and a score of 2 or fewer points is considered low risk for IE. In contrast, PREDICT takes a different approach, requiring two days of evaluation: at day 2 of SAB, the site of acquisition of the bacteremia and the presence of cardiac devices; and at day 5, the presence of sustained bacteremia > 72 hours. A positive PREDICT score is 4 or more points at day 1 and 2 or more points at day 5; otherwise, it is considered negative/low risk for IE.
In the current Columbian patient group, SAB was defined as adult patients (> 18 years of age) with one or more blood cultures positive for SA (believed not to be a contaminant). The mean age was 57 years; 52.8% were considered nosocomial, 32.1% were healthcare-associated, and 16.7% were considered community-acquired; 26.3% were methicillin-resistant bacteremias. Echocardiography was performed in 600 cases, of which nearly three-fourths were transesophageal. Of the 922 adults with SAB, 62 were diagnosed with IE. Fifty-two cases met the modified Duke criteria for IE (47 cases met two major criteria, and 15 met one major criterion and at least three minor criteria).
When the VIRSTA scoring system was applied, the frequency of IE was 12.8% in those with a positive score and 0.44% in those with a negative score. (These two patients were believed to be falsely negative because they quickly died, and follow-up blood cultures could not be obtained.) In contrast, when the PREDICT scoring system was used in the same group, the frequency of IE was 10.6% in those with a positive score and 4.8% in those with a negative score. The frequency of IE was tenfold greater in those at low risk using PREDICT vs. VIRSTA, respectively. The median VIRSTA score for those with and without IE, according to Duke criteria, was 7 and 2, respectively. The median PREDICT score for those with and without IE at day 1 was 1 and 0, and at day 5, was 2 and 1, respectively. The sensitivity and negative predictive value of the VIRSTA scoring system was 96.7% and 99.5%, whereas the sensitivity and negative predictive value of PREDICT at day 5 was 51.6% and 95.1%.
When using the VIRSTA scoring system, echocardiography could be reasonably and safely deferred in those patients with a negative score, who have < 0.5% chance of IE. Thus, nearly half the echocardiograms performed could have been avoided. In contrast, the frequency of IE was 4.6% in those with a negative PREDICT score — too high to skip echocardiography. However, the utility of the scoring systems may depend on the populations studied. The Colombian patient group had fewer cases of IE (6.7%) than either that of VIRSTA (11%) or PREDICT (13%), and the patients were significantly younger, with fewer injection drug users and a lower frequency of cardiac devices and prior IE. Thirty-nine percent of the SABs were related to hemodialysis catheters. All these factors generally would confer a lower risk of IE in the population studied. It would be interesting to compare the two scoring systems in a higher-risk group.
Researchers in Columbia put predictive scoring methods through their paces.
You have reached your article limit for the month. Subscribe now to access this article plus other member-only content.
- Award-winning Medical Content
- Latest Advances & Development in Medicine
- Unbiased Content