The development of the Canberra symptom scorecard: a tool to monitor the physical symptoms of patients with advanced tumours

Table 6 Selected results from stepwise ordinary least squares regressions showing the percentage of variance explained for each symptom associated with the cluster symptom measure¹.

Cluster 1 GI/pain/cogn	% Var	Cluster 2 Neural/emot	% Var	Cluster 3 Tension	% Var	Cluster 4 Respiratory	% Var
Vomiting	5.5	Difficulty sleeping	19.7	Tense	70.3	Coughing	18.9
Lack of appetite	0.8	Dizziness	66.8	Shortness of breath	29.7	Shortness of breath	81.1
Nausea	69.9	Sore/weak muscles	5.4			breath
Dry mouth	1.0	Numbness	1.0
Constipation	14.6	Depression	1.4
Heartburn	1.3	Tiredness	0.9
Pain	3.3	Lack of energy	4.9
Problems remembering things	3.7

1. A composite symptom measure was created for each of the four clusters (GI/pain/cognitive, Neurological/emotional, Tension, and Respiratory). The symptoms associated with each composite measure were then examined in a stepwise method for their contribution to explaining the variation in each cluster. The order symptoms were entered into the regression model follows the amount of variance explained, with the largest being entered first.

ISSN: 1471-2407