Tincello, Douglas G. MD, MRCOG1; Williams, Kate S. RN, PhD2; Joshi, Miland MSc, CStat2; Assassa, R Phillip MD, MRCOG3; Abrams, Keith R. PhD2
From the 1Reproductive Science Section, Cancer Studies and Molecular Medicine, and 2Department of Health Sciences, University of Leicester, Leicester, United Kingdom; and 3Mid Yorkshire National Health Service Trust, Pontefract, Yorkshire, United Kingdom.
Corresponding author: Mr. Douglas Tincello, Reproductive Science Section, Cancer Studies and Molecular Medicine, University of Leicester, Robert Kilpatrick Clinical Sciences Building, Leicester Royal Infirmary, PO Box 65, Leicester LE2 7LX; e-mail: firstname.lastname@example.org.
OBJECTIVE: To compare the quality of data recorded in urinary diaries completed for 3 days compared with those completed for 7 days.
METHODS: Men and women enrolled in a randomized study of a nurse-led intervention for urinary incontinence were eligible for the study. Participants completed either a 7-day diary or a 3-day diary at baseline and after 8 weeks of the trial intervention. Data quality was assessed by categorizing each day as complete or incomplete. Incomplete days were those with fewer than four entries and those with obvious discrepancies from the rest of the diary days. Data were compared by percentage differences with 95% confidence intervals (95% CI).
RESULTS: Two hundred forty-eight patients completed the 3-day diary, and 40 patients competed the 7-day diary. The majority of the respondents were women. There were more completed 3-day diaries (90.7%) compared with 7-day diaries (50%) (difference 40.7%, 95% CI 28.8–52.6%). When compared with the first 3 days of the 7-day diary, the 3-day diary contained more complete data (90.7% versus 65%; difference 25.7%, 95% CI 14.6–36.9%). These differences were seen in all data types except nocturnal incontinence. There was no evidence of association between diary completion and demographic characteristics.
CONCLUSION: Our data demonstrated that the 7-day diary included less complete data than the 3-day diary, even in the first 3 days of the longer diary. This finding suggests elements of both diary “fatigue” and also diary “despair” may be acting. We would suggest that a 3-day diary is superior in terms of data quality.
LEVEL OF EVIDENCE: III
Urinary diaries provide an essential element of the assessment process for the individual with urinary symptoms1 and are a recommended component in the assessment of incontinent patients. Diaries and charts are also useful in clinical practice to allow the determination of a baseline from which future symptom evaluation can be made. They also provide diagnostic information alongside the clinical history and an appropriate battery of assessment tests.
While the International Continence Society recommends the use of a diary or chart to record voiding parameters, there is no formal recommendation for the duration of record keeping.2 The most frequently evaluated diary has been the 7-day diary3,4 although durations from 24 hours to 2 weeks have been reported.4–6 The variability and reliability of data from diaries of different lengths have been extensively studied.2,4,5,7,8 The authors reached different conclusions, from a 24-hour diary being sufficient,5,7 through to 7 days being the minimum recording interval necessary.8 The majority of the studies excluded diaries with incomplete data and so their conclusions are based on a “perfect” data set. The quoted rates of non-completion varied from 4%8 to 32%.7 In a clinical setting, one needs to find a balance between accuracy and reliability of data collection, and the completeness of data entry. This study is a secondary analysis of diary data collected from a randomized clinical study comparing a new nurse-led intervention for continence care with standard care.9 The objective of the secondary analysis was to compare the performance of 7- and 3-day diaries in terms of completeness of recorded data.
MATERIALS AND METHODS
This study used diary data from a randomized trial conducted under the auspices the Leicestershire Medical Research Council Incontinence Study, which explored the prevalence of urinary symptoms and delivery of treatment interventions in the community population aged more than 40 years. Institutional review board approval (local ethics committee) was obtained for the main study and for this analysis. Data collection was undertaken during the pilot phase of the evaluation study comparing a new continence nurse practitioner–led service with existing care, which is reported elsewhere.9
Individuals were identified randomly from the Leicestershire Health Authority patient register and sent a validated postal questionnaire.10 Individuals who satisfied eligibility criteria for a randomized controlled trial of a new nurse-led service compared with existing care9 were approached to be interviewed at home. Five continence nurses undertook this assessment. All individuals seen by a nurse during the pilot study were asked to complete a 3-day urinary diary, and a sample of 10 patients from each of the five nurses’ caseloads was asked to complete a 7-day diary instead of the 3-day diary. These 10 patients were identified as the first 10 to be recruited by each nurse during a defined 8-week period during late 1998. No formal sample size estimations were performed for this comparison.
All patients were instructed how to complete the diary, and this was supported with written information. A calibrated jug was provided for the purpose of measuring output, and individuals were instructed to use the same size cup or mug for drinks which they would measure once. Details of the type and volume of fluid were recorded. Episodes of leakage were recorded by circling one of four options: almost dry, damp, wet, and soaked. These categories were established after early development of the diary which entailed discussion with diary users. After the 8 weeks of the intervention trial,9 patients were asked to complete a second diary, to allow assessment of sensitivity to change.
After completion, compliance was empirically assessed by examining entries in the chart. Essentially, a “common sense” approach was used. Data quality was assessed by categorizing each day as complete or incomplete. Days in which less than four entries were recorded were labeled as incomplete (such charts had usually been completed at the beginning of a day with one to four entries included, followed by no entries later in the day). Also, days in which there were obvious discrepancies of the timing and amount of data compared with the rest of the diary were labeled as incomplete (eg, a limited number of charts reported all fluid intake overnight which did not correlate with the patient’s reported sleeping times; an extra “0” appeared to have been added in some charts implying a single void of 3,000 mL). The proportion of incomplete diary days was calculated and compared for each diary. To look more closely at patterns in compliance, the diary information was subdivided into recorded intake, recorded output, number of voids recorded, and leakage episodes recorded. Leakage was subdivided into daytime and night time. Each category of data were reassessed for completion as described above.
Categorical data are presented as number (percentage), and comparisons of proportions in each group were done by calculating the percentage difference and the 95% confidence intervals (CIs) for this difference by means of the z test.11
Two hundred forty-eight patients completed the 3-day diary, and 40 patients completed the 7-day diary before any intervention. One of the five study nurses was unable to complete her allocation of patients. Of the 7-day diary and 3-day diary users, 82.5% and 57.7%, respectively, were women (difference 24.8%, 95% CI 8.6–41.1).
Compliance of diary completion is shown in Tables 1 and 2. There was no clear evidence of association between completion and demographic characteristics (data not shown). The percentage of diaries with complete data in the 3-day diary group was significantly greater than in the 7-day diary group when assessed overall (Table 1). Analysis of the five separate categories of diary data showed the same pattern of greater completion in the 3-day group, except night time leakage.
Data from the 3-day diary was compared with the first 3 days of the 7-day diary to check for evidence of worsening completion over the 7-day period. This comparison showed the same pattern of compliance, with the data from days 1–3 of the 7-day diary having lower completion rates than the 3-day diary, although the differences were smaller than for the whole 7-day dataset (Table 2).
There was an association between completion of the first 3 days of the 7-day diary and employment status in the post-treatment group (P=.024), suggesting that retired people were more likely to be completers.
Data from the diaries (where complete) was comparable in terms of reported frequency, nocturia, fluid intake, leakage episodes, and severity of leakage, with no significant differences between 3-day and 7-day diary (data not shown).
We have compared the completion rate of data recorded prospectively in 3-day and 7-day urinary diaries to assess whether there were any differences in the completeness of the data. We found that the 7-day diaries contained significantly more incomplete daily data than the 3-day diaries. This difference was still present, although of smaller size, in the first 3 days of the longer diary. These findings suggest that there was a component of diary “fatigue” across the 7-day study period but also that there was an element of diary “despair” such that the prospect of completing 7 days of data collection produced a negative influence upon completion rates from the outset.
Our figures for completion rates compare favorably with those of others.7,12 It is generally thought that increased involvement of both the patient and the supervising clinician will improve the compliance and utility of the charts,1 although there are few data to support this belief. Our compliance rate of 80% and higher was achieved as part of an intensive nurse-led intervention package13 and would support this conclusion.
There is conflicting evidence about the ideal length of data collection, from 24 hours5,7,12 to longer periods of 4 days or more,2,8 based on analysis of complete data only. Groutz et al12 found that compliance with recording measured voided volume fell from 92% in a 24-hour diary to 76% in a 3-day diary, supporting our concept of diary fatigue. There seems little merit in collecting data for 7 days when in practice this will result in a higher level of incomplete data. Taken with evidence that 3- and 4-day diaries do produce reliable data,2,14 our findings would support the use of a 3-day diary to obtain data of reasonably good quality and reliability. At the very least, the concept of “diary despair” deserves further consideration.
We acknowledge the weaknesses in this study. The 7-day diary group was significantly smaller than the 3-day group which will exaggerate the observed differences. However, the magnitude of the differences we observed are such that it seems likely that these are real differences. It is always difficult to exclude biases in diary data collection because patients complete them away from the investigators. However, our data were collected prospectively from a clinical trial which required repeated contact with a small group of motivated practitioners, and so we believe that the diaries were completed accurately and honestly.
Our data provide evidence that data quality of a 7-day diary is worse than a 3-day diary and this difference is due to both diary “fatigue” across the study period and diary “despair” at the prospect of a long collection period. These concepts should be considered in the debate about what represents the ideal collection period.
1. Abrams P, Klevmark B. Frequency volume charts: an indispensable part of lower urinary tract assessment. Scand J Urol Nephrol 1996;179:47–53.
2. Schick E, Jolivet-Tremblay M, Dupont C, Bertrand PE, Tessier J. Frequency-volume chart: the minimum number of days required to obtain reliable results. Neurourol Urodyn 2003;22:92–6.
3. Palnaes Hansen C, Klarskov P. The accuracy of the frequency-volume chart: comparison of self reported and measured volumes. Br J Urol 1998;81:709–11.
4. Wyman JF, Choi SC, Harkins SW, Wilson MS, Fantl JA. The urinary diary in evaluation of incontinent women: a test-retest analysis. Obstet Gynecol 1988;71:812–7.
5. Mazurick CA, Landis JR. Evaluation of repeat daily voiding measures in the National Interstitial Cystitis Data Base Study. J Urol 2000;163:1208–11.
6. FitzGerald MP, Brubaker L. Variability of 24-hour voiding diary variables among asymptomatic women. J Urol 2003;169:207–9.
7. van Melick HH, Gisolf KW, Eckhardt MD, van Venrooij GE, Boon TA. One 24-hour frequency-volume chart in a woman with objective urinary motor urge incontinence is sufficient. Urol 2001;58:188–92.
8. Homma Y, Ando T, Yoshida M, Kageyama S, Takei M, Kimoto K, et al. Voiding and incontinence frequencies: variability of diary data and required diary length. Neurourol Urodyn 2002;21:204–9.
9. Williams KS, Assassa RP, Cooper NJ, Turner DA, Shaw C, McGrother CW, et al. Clinical and cost-effectiveness of a new nurse-led continence service: a randomised controlled trial. Br J Gen Pract 2005;55:696–703.
10. Shaw C, Matthews RJ, Perry SI, Assassa RP, Williams K, McGrother C, et al. Validity and reliability of an interviewer-administered questionnaire to measure the severity of lower urinary tract symptoms of storage abnormality: the Leicester Urinary Symptom Questionnaire. BJU Int 2002;90:205–15.
11. Altman DG. Principles of statistical analysis. In: Altman DG, editor. Practical Stastistics for Medical Research. London (UK): Chapman & Hall; 1991. p 152–78.
12. Groutz A, Blaivas JG, Chaikin DC, Resnick NM, Engleman K, Anzalone D, et al. Noninvasive outcome measures of urinary incontinence and lower urinary tract symptoms: a multicenter study of micturition diary and pad tests. J Urol 2000;164:698–701.
13. Williams KS, Assassa RP, Smith NK, Jagger C, Perry S, Shaw C, et al. Development, implementation and evaluation of a new nurse-led continence service: a pilot study. J Clin Nurs 2000;9:566–73.
14. Matthiesen TB, Rittig S, Mortensen JT, Djurhuus JC. Nocturia and polyuria in men referred with lower urinary tract symptoms, assessed using a 7-day frequency-volume chart. BJU Int 1999;83:1017–22.