Controlling bias in complex nursing intervention studies: a checklist.

Bias is defined as any systematic error resulting in an inaccurate estimate of the outcome of a study. In studies of complex nursing interventions, bias is particularly difficult to control because of issues related to blinding and choice of controls. The authors identify strategies to control bias in intervention studies. They conduct a literature review and consult expert opinion to identify 6 areas of study protocol development that have potential for reducing bias: study concept, definition of intervention, selection of comparisons, randomization/allocation, integrity of intervention, and ascertainment of outcomes. They provide a checklist to help researchers reduce the potential for bias in preparing protocols for complex nursing intervention trials. Use of the checklist can enhance scientific rigour and thus help to ensure that clinicians are ultimately provided with reliable information.