Wells and Geneva scores evaluated for risk stratification for PE in primary care
Discriminative ability was comparable for the original Wells, modified Wells, simplified Wells, revised Geneva, and simplified revised Geneva models, but the Wells rules had lower failure rates.
A recent study tested 5 clinical decision models for stratifying risk in patients with possible pulmonary embolism (PE) in the primary care setting.
Discriminative ability was comparable for the original Wells, modified Wells, simplified Wells, revised Geneva, and simplified revised Geneva models. However, the Wells rules had lower failure rates, defined as the presence of a PE in a patient with a low predictive model score, according to the study. Results were published online on Sept. 8 by The BMJ.
The sensitivity of the different models ranged from 88% (simplified revised Geneva) to 96% (simplified Wells), and specificity ranged from 48% (revised Geneva) to 53% (simplified revised Geneva). The overall efficiency of all models in identifying patients at low risk for pulmonary embolism was between 43% and 48%.
However, failure rates differed, especially between the simplified Wells and the simplified revised Geneva models (failure rates, 1.2% [95% CI, 0.2% to 3.3%] and 3.1% [95% CI, 1.4% to 5.9%], respectively). Each model missed the same 3 patients with PE.
An accompanying editorial warned that clinicians must ask if these results are generalizable to their own practice setting. “Doing so quickly identifies an important limitation of this study: the specific D-dimer test that was used (Simplify D-dimer; Clearview, Inverness Medical, Bedford, UK) is not universally available,” the editorialist wrote.
The editorial also criticized the sensitivity of the models, which was about 95% altogether, asking, “Can the practicing office clinician accept a strategy that misses one in 20 patients harboring a pulmonary embolus? I suspect that many clinicians, particularly those who are risk averse, would judge this sensitivity (the ability of this strategy to ‘rule out’ a pulmonary embolus) to be inadequate and would likely avoid the proposed strategy.”
The study also did not include 3 other validated clinical decision rules for PE: the Charlotte rule, PE rule-out criteria (PERC), and Pisa rules, the editorial stated. The PERC rule, in fact, had a sensitivity of 97.4% (without any D-dimer test) when studied in a large, diverse emergency department population, according to the editorial.
Also, if the PERC rule had been used in the study, the 3 false negatives would have been correctly identified, according to the editorial. “I suspect that if (and when) the PERC rule is tested in primary care, it will prove as useful for primary care doctors as it has for doctors in the emergency department,” the editorialist wrote, noting that primary care physicians will still need to rely on their clinical judgment when managing low-risk patients with suspected PE.