Validation of the Patient Global Impression scales for use in detrusor overactivity: secondary analysis of the RELAX study

Objective To assess construct validity of the Patient Global Impression scales (Severity [PGI-S], Bother [PGI-B] and Improvement [PGI-I]) for symptoms of detrusor overactivity (DO). Design Secondary analysis of a randomised trial of onabotulinum toxin A. Setting Eight UK urogynaecology departments. Population A total of 240 women with DO refractory to medical treatment randomised to receive 200 iu onabotulinum toxin A or placebo in the RELAX trial and followed up for 6 months. Main outcome measures Urinary diaries and disease-specific quality of life (QoL) questionnaires were completed at baseline and during follow up. Discriminatory ability of the PGI-S, PGI-B and PGI-I scales to identify symptom severity and change in severity was assessed by comparing mean diary and QoL outcomes across the response categories, analysed by one-way analysis of variance. Results Data were available from 237 women (98.8%) for validation of PGI-S and PGI-B at baseline, and 192 women (80%) at 6 weeks follow up for validation of PGI-I. Leakage episodes (P = 0.01), urgency episodes (P = 0.019), urgency severity (P = 0.012), and QoL scores (all P < 0.001) were greater in women with more severe problems on PGI-S. Similar results were seen for PGI-B: leakage (P = 0.051), urgency episodes (P < 0.001), urgency severity (P < 0.001), and QoL scores (all P < 0.001). PGI-I responses demonstrated significant relationships with size of change of all variables (P < 0.001). The generic instrument EQ-5D had weaker relationships (PGI-S, P = 0.09; PGI-B, P = 0.004; PGI-I, P = 0.06), suggesting that it was less sensitive. Conclusions The PGI scales are robust and valid instruments to assess disease severity, bother and improvement after treatment in women with detrusor overactivity.