Methodological Issues in Outcomes Research

Abstract

With the increasing societal demand for professional and financial accountability, nurses and other health-care professionals are challenged to demonstrate that the care they provide "makes a difference" -that their interventions achieve positive client outcomes effectively and efficiently. Effective interventions produce the desired responses. In nursing, desired responses refers to measurable changes in the client's health status, condition, or behaviour that indicate the resolution of a presenting problem or diagnosis or the prevention of a condition (Hegyvary, 1993; Lang & Marek, 1990). Demonstrating that nursing interventions are effective rests on the ability to detect the expected changes. The ability to detect the changes depends, in turn, on selecting outcomes that are attributable to the antecedent care, sensitive to nursing care, and congruent with the unit of analysis, and on assessing the outcomes at the appropriate time (Bond & Thomas, 1991; Griffiths, 1995; Hegyvary, 1991; Jones, 1993; Stewart & Archbold, 1992). Even if the right outcomes are selected and assessed at the right time, detecting the expected changes requires that two methodological issues, selection of outcome measures and implementation of the intervention, be appropriately addressed; otherwise there is the likelihood that "real" intervention effects, when present, will not be detected and, subsequently, that the validity of conclusions regarding effectiveness of the intervention in achieving the desired outcomes will be threatened (Lipsey, 1990; Scheirer & Rezmovic, 1983).