The Quality Of The Research On Nurse Decision Making And Use Of Evidence.

Most of the existing research on nurse decision making is of poor quality. Many studies use survey methods, often with the self-report questionnaire as a tool for data gathering. (Funk et al., 1995); (Robichaud-Ekstrand & Sherrard, 1994); (Shaffer, 1996); (Walckzack et al, 1994). As we have seen, studies of decision making in medicine that have combined observation with self-report tools highlight the over reporting that occurs with this type of approach to research design (Covell et al., 1985). More recently, Estabrooks (Estabrooks, 1999) has highlighted the lack of theoretical clarity associated with the concept of research utilisation itself. This work implies that different studies of the use of research evidence by nurses may not even be reporting the same phenomenon.

Other studies examining research utilisation via the survey method are limited in their generalizability due to the small non-random nature of their samples (Thompson & Sutton, 1985). Others, despite reasonably large randomly selected samples have poor response rates. For example, Bostrum and Suter’s (1993) examination of the correlates of research utilisation only secured a response from 23% of the original 7000 nurse sample. Some studies manage to combine all three of these characteristics. For example, Champion & Leach (1989), in their investigation of variables associated with research utilisation, used a battery of self-report scales on a convenience sample of 150 nurses, of whom only 59 yielded data (a response rate of just 39%).

Studies using qualitative methodologies fair little better in terms of quality. For example, few qualitative studies describe an explicit framework for sampling informants and settings (Luker & Kenrick, 1992); (Meah et al., 1996); (Rodgers, 1994).

There is much repetition in the literature and a paucity of good quality empirical studies examining information use in clinical decision making by nurses. Nethertheless, four themes, representing a typology of variables which may impact on nursing’s relationship with research evidence, can be advanced:

• Professional Cultural

• Environmental

• Individual decision maker related

• Information related