To assess the reliability of symptom reports in 3-day vs 7-day bladder diaries used in clinical trials of patients with overactive bladder (OAB) and to compare those results and related issues with previous reports.
To assess the reliability of symptom reports in 3-day vs 7-day bladder diaries used in clinical trials of patients with overactive bladder (OAB) and to compare those results and related issues with previous reports.
We analysed two large-scale, randomized, phase 3 clinical trials of the use of transdermal oxybutynin for treating patients with OAB. The first trial (Trial A, 520 patients) compared three doses of transdermal oxybutynin (1.3, 2.6 and 3.9 mg/day) with placebo. Patients documented their OAB symptoms in a 7-day diary. The second clinical study (Trial B, 361 patients) compared the efficacy of 3.9 mg/day transdermal oxybutynin with 4 mg/day extended-release tolterodine and with placebo; this trial required symptom recording for only 3 days. The internal consistency of the data from the 7-day trial was determined and then compared with the 3-day trial results.
Patients on transdermal oxybutynin or long-acting tolterodine for their OAB symptoms showed a clinically and statistically significant improvement, results that were documented in both 3-day and 7-day bladder diaries. However, compared with 7-day symptom records, 3-day diaries were associated with significantly better compliance with record-keeping (P < 0.001).
Seven-day diaries used in clinical trials supply accurate and reproducible data on clinical manifestations of OAB, but 3-day diaries are equally effective and have the potential for better accuracy through increased patient convenience. Three-day diaries may also reduce the tendency for patients to complete gaps in record-keeping from memory.
Overactive bladder (OAB) is a syndrome characterized by urgency, with or without incontinence, usually with frequency and nocturia . A national telephone-based survey using clinically validated, symptom-based criteria revealed that OAB is more prevalent than previously thought, affecting ≈ 16% of men and women in the USA . Patients with OAB exceed the number of adults with heart disease by 10 million, and the prevalence of OAB is significantly greater than that of diabetes mellitus or asthma in adults [3,4] as determined by the Centers for Disease Control. The safety and efficacy of behavioural, pharmacological and surgical interventions used in the treatment of this disorder have been evaluated in numerous randomized, controlled clinical trials. The antimuscarinics oxybutynin and tolterodine are most frequently prescribed for patients with OAB. These agents show consistently better improvement than in patients receiving placebo . In an effort to improve the tolerability of these agents, extended-release formulations of each compound have been developed. Recently, a transdermal formulation of oxybutynin was shown to have comparable efficacy and a better systemic safety profile [6–8]. Other recently approved pharmacological agents, trospium, solifenacin and darifenacin, are also effective [9–11].
Urodynamic studies can provide an objective evaluation of lower urinary tract function, but bladder volume, volume at first desire to void, and pressure rise during bladder filling may be poor indicators of the incidence or severity of OAB symptoms [12,13]. Also, serial urodynamic assessments are resource intensive in the context of randomized clinical trials. Therefore, in clinical trials and in clinical practice, patient diaries are generally the primary tool used to evaluate OAB symptoms and treatment interventions [14–16]. However, the success of patient diaries depends on many patient and trial design factors. Accurate patient diaries require that patients understand OAB symptoms to sufficiently differentiate between urge- and stress-related episodes. Also, patient cooperation is required to provide an accurate, real-time record of urinary habits. As might be anticipated, some patients asked to use diaries may not comply fully, which can lead to missing data and/or attempts by the patient to reconstruct data, particularly if the data-entering period extends beyond what they consider to be a convenient period.
In this study we examined the evidence from two large-scale, randomized clinical trials to determine the accuracy of data documented in 3-day diaries compared with that obtained from 7-day diaries. We also review data on the variability of information obtained from urinary diaries and detail the effect of the duration of diary-keeping on the quality of the evidence.
Two randomized, phase 3 clinical trials of the transdermal formulation of oxybutynin, a tertiary amine with anticholinergic and direct antispasmodic properties, were conducted in patients with OAB symptoms [7,8]. These trials differed in documentation of symptoms; one trial used a 7-day diary (Trial A)  while the other required that symptoms were recorded for only 3 days (Trial B) .
Trial A was a 12-week, multicentre, randomized, double-blind, placebo-controlled study . Eligible patients were required to have urge or mixed urge and stress urinary incontinence (UI), as shown by ≥ 10 episodes of urge UI per week and eight voids or more per day. About 80% of the participants had never received pharmacological treatment for their OAB symptoms. Patients were randomized to daily treatment with 1.3, 2.6 or 3.9 mg transdermal oxybutynin or placebo. A 7-day bladder diary, in conjunction with medical history, was used to confirm the diagnosis of urge or mixed UI at baseline. Patients were instructed on the proper methods for completing the diary, collecting and measuring urine volume, and distinguishing between urge and stress UI episodes. The primary efficacy variable was the change from baseline in the number of UI episodes per week, as recorded in the diary. Supporting secondary efficacy measures from the diary included changes from baseline in mean daily urinary frequency and mean urinary void volume. Patients documented these pieces of information in a 7-day diary at baseline and immediately before clinic visits at 3, 6, 9 and 12 weeks. At the discretion of the investigator, patients who failed to accurately complete their diary were allowed to repeat the diary evaluation. Those who failed to qualify after the second diary attempt were withdrawn from further study participation.
Trial B  was designed using a 3-day diary assessment period to minimize potential pitfalls that might be associated with lengthier record keeping. This 12-week, multicentre, randomized, double-blind, double-dummy, placebo-controlled trial compared 3.9 mg/day transdermal oxybutynin applied twice weekly, long-acting 4-mg tolterodine capsules, and placebo. In contrast to Trial A, Trial B required participants (361) to have had a previous beneficial response from antimuscarinic therapy. The primary efficacy variable was the change from baseline in the mean number of UI episodes per day, as recorded in the 3-day bladder diary. Supporting secondary efficacy measures from the diary included changes from baseline in mean daily urinary frequency and mean urinary void volume. As in Trial A, patients were instructed on the proper methods for completing the diary, collecting and measuring the urine volume, and distinguishing between urge and stress UI episodes. Patients documented the number of daily UI episodes, and voiding frequency and void volume, on 3 consecutive days at baseline and 4–6 days immediately before clinic visits at 2, 6 and 12 weeks. At the discretion of the investigator, patients who failed to accurately complete their diaries were allowed to repeat the diary evaluation. Those who failed to qualify after the second diary attempt were withdrawn from further study participation.
In Trial A, the effect of the duration of data collection was evaluated by comparing the mean daily number of UI episodes of an entire 7-day diary record with the mean of daily recordings made during the first 3 days of the same period. Repeated measures analysis was used within each week, separately for each treatment group, to determine if the number of UI episodes per day changed over time, while taking into account the covariance pattern of the repeated measurements. The covariance structure chosen for all analyses was compound symmetrical and chosen based on the Akaike and Bayesian information criterion ‘smaller is better’ fit statistics in Proc Mixed of SAS® (v8.1) criteria for model selection.
The total number of UI episodes was obtained from the diaries and the number of episodes per day calculated by dividing this total by the number of days with data recorded in the diary, rounding to the nearest whole number. To examine the effects of missing data on the primary and supporting efficacy analyses, treatment effect conclusions compared the last observation carried forward with the supporting analyses for both the intent-to-treat and evaluable cohorts.
Compared with placebo, both antimuscarinics produced clinically and statistically significant improvements in OAB symptoms, as documented in the 3- and 7-day bladder diaries. In Trial A, the median number of daily UI episodes decreased from 4 to 1 in the group receiving transdermal oxybutynin 3.9 mg/day (P = 0.0265 vs placebo). Further, daily urinary frequency decreased from baseline from a median of 11 to nine episodes (P = 0.0313 vs placebo) while void volume increased from 168 to 194 mL (P < 0.001 vs placebo) in the oxybutynin 3.9 mg/day group. Only the results for oxybutynin 3.9 mg/day differed significantly from the placebo in reducing UI episodes.
In Trial B, patients in the transdermal oxybutynin group had significantly (P = 0.0137) fewer UI episodes per day than those on placebo (median change −3 vs −2, respectively). There were also significant reductions in patients treated with tolterodine (median change −3 vs −2, P = 0.0011). There were no significant differences in primary efficacy measures between tolterodine and oxybutynin. In both trials, there was a large placebo effect for several reasons. Patients received information about bladder function, bladder control, and fluid management, each of which may result in an improvement in symptoms. In addition, patients in both studies were instructed to maintain normal fluid intake and to follow their usual programme of nonpharmacological management for UI (e.g. pelvic floor exercises, timed voiding/bladder training) . Finally, the act of recording the symptoms in a diary also contributes to the placebo effect in clinical studies of OAB.
Except for a lower reporting frequency on day 6 in one treatment group (that on 1.3 mg/day oxybutynin) there were no significant differences between the treatment groups in the number of episodes reported over time. There was no difference in the 3.9 mg/day treatment group; median values were either 4 or 5 episodes per day, with the mean (sd) ranging from 4.7 (3.0) on day 4 to 5.2 (3.4) on day 6 (P = 0.3478). The consistency of the daily diary reports during the 7 days emphasizes the chronic and consistent nature of OAB symptoms. Most importantly, the lack of a significant time effect indicates no obvious systematic changes in reporting or in the occurrence of UI episodes during the 7-day monitoring periods.
The change from baseline in the number of UI episodes in the 7-day diary study was re-evaluated using only diary days 1–3 (Table 1). The null hypothesis (no significant differences) between the active treatment groups and placebo was tested using the same statistical methods used for the 7-day diary evaluations. At the last observation carried forward, compared with placebo, only the 3.9 mg/day group had a significant improvement in UI episodes. This is consistent with results achieved with the 7-day diary evaluations in the trial. Comparison of only the placebo and the 3.9 mg/day treatment group (i.e. no multiple-comparison adjustment) also resulted in a significant difference (P = 0.013) based on the 3-day evaluation time period. This analysis suggests that the results obtained using the entire 7-day diary and over the first 3 days of these diaries are comparable.
|UI episodes* (change from baseline)||7-day diary||Diary days 1–3 only|
|Mean (sd)||−3.0 (2.5)||−3.0 (2.4)|
According to records in the diaries, the mean change in the number of daily UI episodes in patients treated with oxybutynin 3.9 mg/day from diary days 1–7 in Trial A were similar to those reported in 3-day diaries in Trial B (Table 1).
A chi-square analysis was used to compare the proportions of patients who had incomplete diaries in Trials A and B. Overall, 56 of 520 patients (11%) in Trial A (7-day diary) failed to document their symptoms on one or more days, compared with two of 382 patients (0.5%) in Trial B (3-day diary) (P < 0.001).
OAB is a prevalent disorder, diagnosed largely from a patient's history. In disease states where accurate diagnostic tools do not exist, patient self-reports are widely used to obtain information on subjective symptoms, such as severity of pain or fatigue, patient satisfaction, and quality-of-life assessments. In addition, certain disease states require self-observation reports, such as sleep disturbances, pulmonary function readings, medication usage, and urinary symptoms. When patients serve as the primary source of information, they may be asked to either recall their experiences after the fact, or to document their symptoms as they occur for later analysis by clinical investigators. Patient self-reported diaries are intended to document experiences close to the time that they occur for the purpose of yielding more accurate data. Indeed, diaries have been found to be reproducible and more accurate than patient recall . Despite the use of bladder diaries as a valid tool for measuring symptoms in clinical trials, diary-keeping may be underrated and is underused in clinical practice, perhaps because of the burden of traditional 7-day record-keeping on patients and physicians.
While numerous variables can be tracked with the use of bladder diaries, most trials specifically assess median or mean changes in UI episodes (urge, stress and total), micturition frequency (diurnal/nocturnal), and urinary void volume, the other variables being nocturia episodes, number of pad changes (diurnal/nocturnal), enuresis and largest single void, based on recorded occurrences. The mean void volume is the most reproducible of the standard diary variables . The frequency-volume curve correlates with lower urinary tract function in patients with OAB and can be easily constructed from the patient diary . The frequency and void volume measurements provide immediate feedback to patients on their urinary function and can serve as an incentive for them to actively participate in the management of their disorder. A clinical study was conducted to establish the statistical correlation between subjective (recalled) and objective (documented in a bladder diary) measures of severity of UI in patients with LUTS . This study showed that, while recalled data may be useful in evaluating outcomes in some instances, it cannot be relied upon for information in all patients with UI . In another study in patients with UI that compared patient recall with documentation in bladder diaries, bladder diaries were also more accurate than patient recall in evaluating OAB symptoms . Therefore, while the importance of bladder diaries is fairly well established, we have added much-needed data to the evidence suggesting that documentation for 7 days might not be necessary.
Patient compliance with frequency-volume charts follows a logarithmic curve that is highest when the duration of monitoring by the patient is short and decreases progressively as the time of data collection increases (Fig. 1) . Research has shown that reliability is low when the duration of monitoring is 1 day, increasing rapidly over the next several days, then decreasing progressively as the monitoring period exceeds 1 week. Therefore, the period that patients are required to keep diaries can affect data quality. The duration of data collection in clinical trials of various pharmacological agents currently approved for the treatment of OAB is typically 3–7 days (Table 2) [22–27]. Although diaries of 1–2 weeks provide more data, they are associated with lower compliance rates and, consequently, lower reliability .
|Drug||Comparators||Diary duration, days||Reference|
|Oxybutynin||Oxybutynin-IR vs placebo||7||Diokno et al.|
|Oxybutynin||Oxybutynin-ER vs placebo||7||Gleason et al.|
|Oxybutynin||Oxybutynin-TDS vs placebo; oxybutynin-TDS vs tolterodine-ER and placebo||7||Dmochowski et al.[7,8]|
|Tolterodine||Tolterodine-IR vs placebo||7||Appell |
|Tolterodine||Tolterodine-ER vs tolterodine-IR and placebo||7||Van Kerrebroeck et al.|
|Darifenacin||Darifenacin vs placebo||7||Haab et al.|
|Solifenacin||Solifenacin vs tolterodine-IR and placebo||3||Chapple et al.|
Several investigators have attempted to determine the most reliable duration of urinary diaries. Results from a behaviour-management trial evaluating 50 women with UI, using a 1-week bladder diary that was assessed for reliability of collected data, showed that a 1-week diary is reliable in assessing voiding frequency and UI episodes . Subsequently, these findings were generally accepted by clinical investigators engaged in studies of OAB, and data obtained from 7-day diaries have been accepted as a primary outcome measure in trials of patients with LUTS. However, reports also support data collection for <1 week.
Schick et al. provided evidence that supports the use of 4-day diaries; the findings of another study  in which participants were asked to document the occurrence of irritative and obstructive voiding symptoms over 3 days suggest that a 3-day chart provides data that can be accurately used to document the prevalence of lower urinary tract dysfunction, such as OAB and stress UI. Other researchers using frequency-volume charts to study micturition patterns in healthy women found that just 48 h of monitoring provided consistent data for assessing micturition patterns . A recent report of the reliability and validity of a 7-day bladder diary in patients with OAB compared data from completed 7-day diaries with data from 3- and 4-day diaries . The results of this evaluation show that the frequency of events calculated over 3 or 4 days did not significantly differ from those recorded in 7-day diaries, and are therefore reliable tools for assessing OAB symptoms and treatments. The use of diary entries for <7 days is also supported by the Standardization Committee of the ICS  and by the experimental observations of Schick et al. and Brown et al.. The clinical trials reported here involving transdermal oxybutynin provide further support for the reproducibility and reliability of data obtained from 3-day bladder diaries. The diary data from both studies of transdermal oxybutynin reported here indicate that patients with many UI episodes have persistent symptoms that have little day-to-day variability; therefore, they do not appear to be associated with detectable trends in reporting that would introduce bias through the use of a 3-day diary. This supports the previous observation that, while reproducibility increases with the number of days a diary is maintained, at some point the logistics of maintaining the diary by the patient can be outweighed by the inconvenience of the task . From the present data, diary keeping for 3 days appears to suffice for assessing the impact of pharmacological interventions for OAB symptoms.
Evidence suggests that the simple act of maintaining a bladder diary, even without the administration of pharmacological treatment, improves OAB symptoms . In a randomized, placebo-controlled trial conducted to assess the efficacy of biofeedback-assisted behavioural treatment in older women with urge or mixed UI, patients were randomized into three groups and received either behavioural therapy, drug therapy, or no therapy (control group). Control patients completed detailed bladder diaries, attended all clinic visits, completed an adverse-effects checklist at each visit, received therapeutic attention, and reviewed their diaries with a nurse practitioner in the same fashion as patients in the other arms of the trial. However, participants in the control group achieved a mean 39.4% reduction in the frequency of UI episodes. With the studies reported here involving transdermal oxybutynin, in Trial A, patients receiving placebo had a median reduction of 50% in UI episodes. Clearly, the placebo group benefited from modifying their behaviour, as required for participation in the trial.
In addition, it was reported that patients with a history of antimuscarinic treatment are less likely to respond to placebo than are their treatment-naive counterparts. In Trial A, 80% of participants had received no previous treatment for OAB. In the entire placebo group, the median weekly reduction in UI episodes per week was 15. However, a separate analysis of the subgroup who had previously received treatment for OAB and who received placebo in this trial (≈ 20% of the overall patient population) revealed a median reduction in weekly UI episodes of 11. Trial B was therefore constructed to include only patients who had benefited from previous pharmacological therapy. However, despite selection of a treatment-experienced population, there was a substantial placebo response that was comparable to the overall placebo response in Trial A and to those in other clinical studies of OAB . Maintaining a diary increases awareness of bladder habits and leakage patterns, thus leading to improved urinary symptoms. This explains the substantial ‘placebo response’ often seen in clinical studies of OAB.
Although the importance of diaries cannot be overemphasized, there are some factors that must be considered to ensure their appropriate use in clinical trials. The terminology must be carefully defined and providers should establish that patients understand not only the meaning of various diary items, but the importance of full, timely and complete data entry. For clinical trials, patient convenience might be improved with effective layout as well as calendar reminders and frequent follow-up calls by the study centre. The type of diary in use may also affect patient compliance with data entry.
While paper diaries are most common, electronic diaries have recently been used and are proving to yield better compliance. Compliance with patient diaries in a nonurological setting has been challenged in a recent study supported by the US National Cancer Institute. Stone et al. recruited 80 patients with chronic pain to evaluate paper vs electronic diaries during a 21-day trial. To monitor compliance, diary binders were fitted with inconspicuous photosensors that recorded when the binder was opened and closed. The electronic diary was a PalmPilot® with software that presented identical pain questions and recorded the time and date of entries. Participants were instructed to complete daily entries within 15 min of target times at 10.00 hours, 16.00 hours and 20.00 hours. Patients were randomized on a 1:1 basis to paper or electronic diaries. The primary compliance efficacy variable was the number of entries made within the 30-min period. Completion within 90 min of the target time was designated as a secondary variable. With the electronic diary, entries could not be initiated outside the designated 30-min period. While enrolees randomized to the paper diary reported compliance rates of 90%, compliance based upon electronic surveillance indicated that the actual compliance was only 11% within the 30-min period and only 20% with a more liberal 90-min period. However, compliance among participants randomized to the electronic diaries was 94%. ‘Hoarding’ (days that a paper diary is opened but results not recorded) was identified in 75% of patients randomized to a paper diary. That almost all participants hoarded data may be a reflection of the 21-day duration of the trial. The potential liabilities of paper diaries suggest that efforts should be made to ensure that compliance is maximized in clinical trials using this method.
The present comparison of 3- and 7-day diaries has some limitations. Each study population was similar in age, gender and race but differed in treatment history. Most patients in Trial A had never been treated for OAB in the past; only 20% had a history of previous pharmacological treatment . In contrast, all patients in Trial B had a history of previous pharmacological treatment . It is possible that patients who have been treated in the past have also entered information into bladder diaries, unlike their treatment-naive counterparts. However, that all patients were trained in how to record their symptoms, and were not permitted to participate in the study unless they had demonstrated that they could do so, should have made any previous experience irrelevant. However, differences in medical and treatment history could still have influenced treatment expectations and symptom interpretation across the two studies. Also, while the inclusion of two separate trials led to the inclusion of more patients and therefore more diaries, a comparison of diaries in the same patients and in the same trial would have allowed each patient to serve as his or her own control. Finally, no telephone calls were made to remind patients of when to start data entry before their clinic visits. Therefore it is not known whether one group found it easier to remember when to start their data entry relative to the other population. It could be argued that it may have been relatively easier to remember data entry 7 days in advance rather than 3 days in advance. However, that only 0.5% of patients had incomplete diaries at week 12 with the 3-day diary, vs 11% of patients using a 7-day diary, strongly suggests that compliance with the 3-day diary was superior, even if some patients found it easier to remember to start their diaries 7 days rather than 3 days before the clinic visits.
In conclusion, numerous factors affect the accuracy and reproducibility of data from a bladder diary. The duration of data collection is an important variable and should be minimized to improve patient compliance. This review of two large-scale, randomized clinical trials of transdermal oxybutynin, with the results of previously published studies, suggest that a 3-day collection period provides accurate and reproducible data on symptoms of OAB in a clinical trial setting. Despite the objective of providing a placebo control group in therapeutic trials, diary-keeping makes patients more aware of their voiding habits. In clinical studies, the 7-day bladder diary is a commonly used primary tool and requires, for maximum results, that patients be educated, before the study is initiated, in maintaining accurate diary entries. In clinical practice, a bladder diary may be an important tool for documenting the frequency of UI before treatment; it can also be an invaluable means of helping clinicians to confirm a diagnosis of OAB and to determine the impact of therapy on symptoms. However, diary-keeping over 7 days in a community setting may affect patient compliance by increasing the duration of patient burden, and may prove burdensome to clinicians, who must rely on diary accuracy for symptom assessment. These results suggest that the use of 3-day diaries may increase patient compliance in clinical trials and may better allow for their routine use in the clinical setting.
R. Dmochowski, R. Appell, V. Nitti and G.W. Davila are paid consultants to sponsor; R. Appell is also a study investigator funded by sponsor; S. Sanders is an employee of sponsor.