The psychometric properties of the Greek version of the State-Trait Anxiety Inventory in cancer patients receiving palliative care.

OBJECTIVES

To validate the Greek version of the State-Trait Anxiety Inventory (STAI) in a sample of cancer patients.

DESIGN

The scale was administered twice, with a 3-day interval, to 99 eligible patients with cancer. Together with the Greek version of STAI scale, the patients also completed the anxiety subscale from the Hospital Anxiety and Depression scale (HAD).

OBSERVATIONS

Factor analysis yielded a three-factor solution, explaining 47.143% of the variance. Cronbach alpha for three scales was between 0.729 and 0.852. Inter-scale correlations were moderate-to-high and ranged from 0.282 to 0.563 (p < 0.0005, p < 0.005). The assessment of the relationships among the Greek STAI scales and HAD-Anxiety showed statistically significant correlations between them (r ranged between 0.428 and 0.596, p < 0.0005). The test/retest reliability of scale (Pearson's 'r'), showed that the coefficient agreement ranged between 0.85 and 0.90 (p < 0.0005). Univariate analysis revealed significant correlations between female gender, metastasis, performance status, chemotherapy, mild opioids and low education level with increased anxiety.

CONCLUSIONS

These results support that the Greek version of STAI is an instrument with satisfactory psychometric properties, and is a valid research tool for Greek cancer patients.