Development of a 32-item scale to assess postoperative dysfunction after upper gastrointestinal cancer resection

The purpose of this study was to develop a 32-item scale to assess postoperative dysfunction in patients who underwent surgery for gastric and oesophageal cancer and to evaluate its reliability and validity.

For the objective assessment of postoperative dysfunction in patients with upper gastointestinal cancer, we performed a preliminary survey by mail using a 34-item questionnaire as a initial version. The results of the survey were assessed by item analysis of the scale. The scale items were further refined by researchers and specialists, and a 32-item scale for the assessment of postoperative dysfunction (initial scale) was developed.

Using this 32-item scale (initial scale), a mail survey was performed of 379 subjects selected by random sampling.

The questionnaire was returned by 292 patients (77.1%) and 283 responses (74.7%) were valid. Of these, 221 respondents had gastric cancer and 62 oesophageal cancer. The mean age of respondents was 64.9 SD 9.8 (range 35-89) years. The mean total score of the 32-items on the initial version for the assessment of postoperative dysfunction was 60.8 SD 16.7. The mean total score for gastric cancer patients and oesophageal cancer patients was 58.1 SD 15.8 and 70.1 SD 16.7 respectively. After the elimination of scale items regarded as irrelevant based on statistical considerations and the judgement of experts, factor analysis was performed. Seven factors were valid: 'regurgitation reflux', 'limited activity because of decreased food consumption', 'passage dysfunction immediately after eating', 'dumping-like symptoms', 'transfer dysfunction', 'hypoglycaemic symptoms' and 'diarrhoea-like symptoms'. The cumulative proportion of variance by scale reliability was confirmed by a Cronbach's alpha-coefficient of 0.926. The Cronbach's alpha-coefficient for all 32 items on the initial version was 0.926, the Cronbach's alpha-coefficient for sub-items was 0.705-0.856, and Pearson's correlation coefficient of re-test for the total score of the 32 items was 0.865, which confirmed a high degree of internal consistency. The construct validity of the scale was confirmed using the known-group technique by operative procedures, and from the result of factorial validity. This scale was named 'Postoperative Dysfunction for Upper Gastrointestinal Cancer 32; PODUGC-32'.

This scale is sufficiently reliable and valid and will be useful clinically. Relevance to clinical practice. We can use the new scale to assess postoperative dysfunction in patients with upper gastointestinal cancer for nursing practice.