In this analysis, the tolterodine ER group showed statistically significant and clinically meaningful improvements in UUI episodes, subjects' perceptions of the extent of their bladder-related problems, and HRQL compared with placebo. These results are consistent with previous reports demonstrating that tolterodine ER significantly improved UUI episodes [26
], PPBC scores [18
], and KHQ scores [16
] in subjects with OAB. However, correlational analyses showed that improvement in UUI episodes was significantly correlated with improvements in PPBC scores and most domains of the KHQ in subjects receiving either tolterodine ER or placebo. The results were similar regardless of whether median percentage or mean changes in UUI episodes were used in the analyses. Changes in PPBC scores were also correlated with changes in scores on each domain of the KHQ, and categorical improvements in PPBC scores were positively associated with the magnitude of improvement on each KHQ domain. These data suggest that the PPBC and KHQ are relevant measures of OAB treatment efficacy.
Although the correlations among improvements in UUI episodes, PPBC scores, and KHQ scores after treatment with tolterodine ER were small to moderate, they were statistically significant. Similarly, a study that used data from primary care patients collected during an open-label trial of tolterodine ER found that improvements in bladder diary variables were significantly correlated with improvements on the PPBC, the Overactive Bladder Questionnaire (OAB-q) Symptom Bother scale, and all HRQL domains of the OAB-q [28
]. As in the present study, the correlations were small to moderate but statistically significant. Collectively, these findings suggest that PROs provide information about OAB treatment efficacy that is related to, but not redundant with, bladder diaries. This interpretation is consistent with a study by Abrams et al [20
] in which symptom-based outcomes (ie, bladder diaries, urodynamics) were compared with questionnaire outcomes from 27 clinical trials of antimuscarinics. The results showed that, although there were often parallel improvements in these 2 outcomes, there were cases in which improvements in symptom-based measures were not accompanied by corresponding changes in PRO measures.
Abrams et al [20
] concluded that statistically significant changes in the frequency of symptoms may not always reflect subjects' assessments of treatment success, and that both symptom-based measures and patient-centric measures of treatment impact should be used in studies of OAB pharmacotherapy [20
]. The findings reported here support the conclusion that symptom-based measures and patient-centric measures are complementary and should be used together in clinical trials. The KHQ, PPBC, and other instruments, such as the OAB-q, Overactive Bladder Symptom Score (OABSS) and Urgency Perception Scale (UPS), have been validated among patients with OAB and are reliable and responsive to treatment [9
In the current study, correlations between improvements in PPBC and KHQ scores were also of small to moderate strength, suggesting that these instruments may reflect different aspects of OAB treatment efficacy. The PPBC, a single-item measure that is easy to use and interpret, provides a global integrated measure of the subjects' perception of the extent of their bladder-related problems. Other validated instruments, such as the KHQ or OAB-q, provide detailed and multidimensional information concerning patient outcome or condition that is not captured by a single-item measure. The choice of questionnaires administered in clinical trials should reflect the research objectives of the trial [9
Notably, these instruments are patient-friendly and may help clinicians diagnose OAB and assess treatment efficacy in the clinic without the inconvenience of a bladder diary. The importance of measuring patients' assessment of treatment impact has been demonstrated by evidence of disparity between the opinions of clinicians and patients regarding the level of bother and the impact of urinary symptoms on HRQL [33
Although changes from baseline for each endpoint were significantly smaller among placebo- versus tolterodine ER-treated subjects, relationships among endpoints were generally similar in the active and placebo groups. This is not unexpected because the correlations should be similar for changes from baseline that are in the same direction, whether they occur in the treatment or placebo group.
The current analysis suggests that PROs are relevant and complementary to changes in UUI episodes in the assessment of OAB treatment; however, the study has some limitations. Inclusion criteria for this study required subjects to have UUI; therefore, the analysis is limited to only patients with UUI. However, a large proportion of patients with OAB are not incontinent [3
]. Because UUI was the only variable used to assess these relationships, we cannot extrapolate these results to other OAB symptoms.