A meta-analysis of intervention effectiveness for symptom management in oncology nursing research

CRD COMMENTARY: This is a well-written review that includes an excellent discussion of the problems found in the methodology of the primary studies. The authors correctly acknowledge that by limiting the primary studies to nurse-authored articles in nursing journals, other relevant studies may have been omitted, and that different results may have obtained by widening the literature search. Much consideration had been given to selecting and grading the studies for quality according to predefined criteria, although no details of the specific factors considered are given. Potential causes of heterogeneity between studies are considered to be the variation of factors across studies, including differences in interventions, differences within intervention category, age of patients, patients' characteristics, variety of scales, standardised instruments, observational guides, and investigator-constructed instruments used to measure the outcomes. As the authors state, the small sample size, i.e. 46 to 295 patients for the meta-analysis, limits the strength of any results. The conclusion that insufficient evidence exists at this time to recommend particular research-based nursing interventions to relieve symptoms in cancer patients, finds support in the discussion of the limitations of the included studies