Assessing data availability for answering bedside questions in the critical care settings

With ubiquitous implementation of Electronic Health Records (EHRs), healthcare organizations hold large volumes of clinical data. Historic, observational data from routine care may provide useful information for bed-side decisions in patient care. We take the initial step toward this goal through analysis of 53 clinical questions relevant to clinical decision making in the Intensive Care Unit. Each question was decomposed into population, intervention and outcome statements then into data elements. Overall, 92.5 % of the questions were supported by data elements. However, algorithms were needed for population determination for 98% of the questions with available data elements. Thirty-one (63%) of the interventions required algorithms. Seven of the standard outcomes required algorithms. The work reported here is the initial step in evaluating the feasibility of observational data for use in clinical decision support. The results are encouraging enough to support further analysis.