[Benefit assessment in nursing research: illustrative examples and questions left unanswered].

Questions in nursing research are often studied in heterogeneous and instable fields such as nursing homes and use highly complex interventions. With the usual designs the mechanisms of their effects are difficult to evaluate. These interventions usually cannot be completely standardised. Comparison with 'care as usual' is hardly possible, since 'care as usual' can vary a lot. Moreover, the motivation to participate in a study with the chance of not getting an intervention is very low. Control groups with an active intervention, however, involve the risk that effects cannot be distinguished. This is why process evaluation and implementation research using multi-method approaches and triangulation are central to evaluating the effects of complex interventions. A solid theoretical foundation is the key to the proper interpretation of empirical results. (As supplied by publisher).