THE PROBLEM OF KNOWLEDGE TRANSLATION IN HEALTHCARE: TOOLS FOR ITS SOLUTION IN THE AREA OF PATIENT SAFETY

Abstract

Adherence to evidence-based clinical guidelines enhances patient safety, but the level of knowledge implementation into routine practice remains unsatisfactory. The concept of knowledge translation (KT) was proposed as a «process that includes the synthesis, dissemination, exchange, and ethically sound application of knowledge» to define the framework of knowledge transfer from research to practical activities aimed at improving the quality and safety of healthcare. Although Russian authors pay much attention to translational medicine, the problem of implementation research is poorly discussed. Modern information technologies including clinical decision support systems (CDSS) play a crucial role in KT, but the evidence of their effectiveness is poor. The authors describe a systematic approach to the practical implementation of knowledge imbedded in clinical guidelines based on the use of CDSS. CDSS implementation was accompanied by organizational measures that ensured the overall success: creation of standard operating protocols and quality register, clinical audit with feedback, and staff training. The effectiveness of this approach in reducing the risk of inhospital complications in Moscow general hospital is illustrated by the example of hospital-acquired venous thromboembolism (HA-VTE) prevention. Implementation of a comprehensive program for HA-VTE prevention in 2014−2016 helped to increase the coverage with VTE and bleeding risk assessment from 15% to 80% (p0.0001), to reduce the frequency of pharmacological prophylaxis defects from 50.6% to 23% (p=0.01), to increase the compliance with anticoagulant use from 50% to 76% (p=0.0005), without increasing the rate of hemorrhagic complications. The HA-VTE rates decreased from 10 to 4.25 cases per 1000 overall hospital admissions (p=0.001), from 8 to 1.27 cases per 1000 operations in surgical patients (p=0.001) and from 12 to 6.38 cases per 1000 hospitalizations in therapeutic patients (p=0.06; with a statistically significant downward linear trend for HA-VTE rate, p=0.038).