Validity and reliability of global ratings of satisfaction with epilepsy surgery

Abstract

Objective We aimed to assess the reliability and validity of single‐item global ratings (GR) of satisfaction with epilepsy surgery. Methods We recruited 240 patients from four centers in Canada and Sweden who underwent epilepsy surgery ≥1 year earlier. Participants completed a validated questionnaire on satisfaction with epilepsy surgery (the ESSQ‐19), plus a single‐item GR of satisfaction with epilepsy surgery twice, 4–6 weeks apart. They also completed validated questionnaires on quality of life, depression, health state utilities, epilepsy severity and disability, medical treatment satisfaction and social desirability. Test‐retest reliability of the GR was assessed with the intra‐class correlation coefficient (ICC). Construct and criterion validity were examined with polyserial correlations between the GR measure of satisfaction and validated questionnaires and with the ESSQ‐19 summary score. Non‐parametric rank tests evaluated levels of satisfaction, and ROC analysis assessed the ability of GRs to distinguish among clinically different patient groups. Results Median age and time since surgery were 42 years (IQR 32–54) and 5 years (IQR 2–8), respectively. The GR demonstrated good to excellent test‐retest reliability (ICC = 0.76; 95% CI 0.67–0.84) and criterion validity (0.85; 95% CI 0.81–0.89), and moderate correlations in the expected direction with instruments assessing quality of life (0.59; 95% CI 0.51–0.63), health utilities (0.55; 95% CI 0.45–0.65), disability (−0.51; 95% CI −0.41, −0.61), depression (−0.48; 95% CI −0.38, −0.58), and epilepsy severity (−0.48; 95% CI −0.38, −0.58). As expected, correlations were lower for social desirability (0.40; 95% CI 0.28–0.52) and medical treatment satisfaction (0.33; 95% CI 0.21–0.45). The GR distinguished participants who were seizure‐free (AUC 0.75; 95% CI 0.67–0.82), depressed (AUC 0.75; 95% CI 0.67–0.83), and self‐rated as having more severe epilepsy (AUC 0.78; 95% CI 0.71–0.85) and being more disabled (AUC 0.82; 95% CI 0.74–0.90). Significance The GR of epilepsy surgery satisfaction showed good measurement properties, distinguished among clinically different patient groups, and appears well‐suited for use in clinical practice and research.