Developing content for a process-of-care checklist for use in intensive care units: a dual-method approach to establishing construct validity

Background: In the intensive care unit (ICU), checklists can be used to support the delivery of quality and consistent clinical care. While studies have reported important benefits for clinical checklists in this context, lack of formal validity testing in the literature prompted the study aim; to develop relevant `process-of-care? checklist statements, using rigorously applied and reported methods that were clear, concise and reflective of the current evidence base. These statements will be sufficiently instructive for use by physicians during ICU clinical rounds. Methods: A dual-method approach was utilized; semi-structured interviews with local clinicians; and rounds of surveys to an expert Delphi panel. The interviews helped determine checklist item inclusion/exclusion prior to the first round Delphi survey. The panel for the modified-Delphi technique consisted of local intensivists and a state- wide ICU quality committee. Minimum standards for consensus agreement were set prior to the distribution of questionnaires, and rounds of surveys continued until consensus was achieved. Results: A number of important issues such as overlap with other initiatives were identified in interviews with clinicians and integrated into the Delphi questionnaire, but no additional checklist items were suggested, demonstrating adequate checklist coverage sourced from the literature. These items were verified by local clinicians as being relevant to ICU and important elements of care that required checking during ward rounds. Two rounds of Delphi surveys were required to reach consensus on nine checklist statements: nutrition, pain management, sedation, deep vein thrombosis and stress ulcer prevention, head-of-bed elevation, blood glucose levels, readiness to extubate, and medications. Conclusions: Statements were developed as the most clear, concise, evidence-informed and instructive statements for use during clinical rounds in an ICU. Initial evidence in support of the checklist?s construct validity was established prior to further prospective evaluation in the same ICU