Clinical decision support could become more personalized, if the results of a study recently published in the Journal of the American Medical Informatics Association are any indication. Researchers determined that a new data-driven and adaptive approach to CDS was found to be more reliable than some existing methods.
The new approach--dubbed ADAPT--does not use training data, and instead relies on confidence intervals (CIs) drawn from individuals. Each model uses different features, parameters and samples, according to the study's authors.
"While many predictive models have been developed and validated on data from cohort studies, little attention has been paid to ensure the reliability of a prediction for an individual, which is critical for point-of-care decisions," the authors wrote. "Because the goal of predictive models is to estimate outcomes in new patients, a critical challenge in prognostic research is to determine what evidence beyond validation is needed before practitioners can confidently apply a model to their patients."
InformationWeek Healthcare Editor Paul Cerrato argued last week that accurate data is imperative to clinical decision support systems. Without it, he said, patient outcomes and cost savings could go by the wayside.
"By way of contrast, no one questions the need to track patients' smoking status-a Meaningful Use reg-because the evidence behind such a recommendation is solid and universally accepted," he wrote. "You need to have a baseline headcount before you can measure how successful you are in getting patients to quit this life-threatening habit."
Additionally, Cerrato said, data needs to be more properly analyzed to reflect "the real world"; for instance, patients with co-morbidities, rather than just one ailment. "Sometimes, the results of [an] analysis should be: 'We don't have enough data to make any recommendation,'" he said.