Issue link: https://nnumagazine.uberflip.com/i/198033
III • They're aware that their compliance statistics are automatically reported to management—another reason to think twice before rejecting the recommended path. Practitioners are, in effect, absorbed into the clinical decision support machine. Little by little, knowledge of the technology supplants knowledge of patient care. By deskilling health professionals, CDSS threaten to displace them. Only skilled professionals empowered to use their independent judgment can advocate in the exclusive interest of patients. Because CDSS are designed to put profit above patients, independent professionals who understand patient needs—not merely how to use CDSS—are essential. CDSS can never be adequately tested. They constantly incorporate new research findings without being retested. The process by which CDSS arrive at their recommendations is mysterious and unchallengeable. • HIT can only be fully understood by computer scientists. • Many of its operations are trade secrets protected from disclosure by law. Practitioners can't know how the decisions recommended by CDSS were arrived at, so they can't evaluate that process. CDSS hide the limitations of clinical studies. Formal medical research isn't complete or disinterested. • Historically, clinical studies have systematically excluded women and minority populations. • Patients with multimorbidity are usually disqualified. • Most pharmaceutical research is funded by drug companies with an interest in promoting their products. As a result, studies are designed to exaggerate the effectiveness of treatment, minimize its risks, and ignore alternatives. In medical journals, research limitations are generally explained. But EBM guidelines that synthesize research from multiple sources strip findings of their political and economic contexts. Consequently, practitioners can't evaluate the quality of research or its relevance to their particular patients. © Copyright IHSP 2009. All rights reserved. t 's m a n a g e m en Will er machine ev challenge an agement 's m values? H o w do y o u know whether the machine is right? at counts Wh e evidencd"? as " eci es Who d at counts? wh