Bredin, M., Corner, J., Krishnasamy, M., Plant, H., Bailey, C., & A’Hern, R. (1999). Multicentre randomised controlled trial of nursing intervention for breathlessness in patients with lung cancer. BMJ, 318(7188), 901–904.
Patients diagnosed with lung cancer were randomized to attend a nursing clinic offering interventions for breathlessness versus best supportive care. The intervention consisted of
Control: received standard of care for breathlessness; best supportive care involved standard management and treatment available for dyspnea. Nurse specialists used and were taught the intervention in same way, using a practice guideline to ensure correct delivery of the intervention.
The study reported on a sample of 119 patients with lung cancer; the goal was to recruit 150 patients for 90% power. Data from 16 patients were lost because a center was dropped from the study. Final sample included 51 in the intervention group and 52 in the control group.
The study was conducted in nursing clinics in six hospital settings in London, United Kingdom.
The study design included randomization to either intervention or control group.
Outcome measures were conducted at baseline and at four and eight weeks.
The intervention group improved significantly at eight weeks in 5 of 11 items assessed.
Final sample included 51 in the intervention group and 52 in the control group. Attrition: 16 patients died, and 28 withdrew. Reasons for withdrawal: 1 improved condition, 16 had deterioration, 4 were unhappy with the assigned treatment arm, and 7 were for unknown reasons. Survival of patients who withdrew was significantly worse than the intervention arm (hazard ratio = 2.5; p < 0.05). Survival of all who withdrew versus nonwithdrawals was also significantly worse (p < 0.01). Overall survival of the intervention and control groups was not different. At baseline, both groups had high level of distress caused by breathlessness and associated functional impairment.
Overall conclusion is that patients attending nursing clinics for breathlessness experienced improvements in breathlessness, performance status, and physical and emotional status.
The analysis was based on the assumption that missing data from patients who withdrew because of poor outcomes were assigned a lower change score .The findings would have been more credible if outcomes were actually assessed. The method assumed that patients were able to show a change in either direction on rating scales, but patients who were on the extreme could only go in one direction. As stated in the Corner et al. (1996) pilot study, the contribution of each component of the intervention is unknown. Are all components needed to achieve the same outcome, or was one aspect of the intervention most significant?