The accuracy of the frequency–volume chart: comparison of self-reported and measured volumes


Dr Palnæs Hansen Kronprinsessegade 50, DK-1306 Copenhagen, Denmark.



 To assess the accuracy of frequency–volume charts recorded by patients with voiding symptoms by comparing self-reported voiding data with the volumes of urine collected in same period.

Patients and methods

 The study comprised 18 patients (median age 63 years, range 20–80) and lasted 3 days, during which the patients recorded their fluid intake and voided volumes, and collected 24 h urine samples.


 The recorded volumes exceeded or underestimated the collected volume in 49% and 51%, respectively, of the samples, but the agreement between the accumulated daily voiding volumes on the charts and the volumes collected was acceptable. The median difference between volumes recorded on the charts and collected as samples was 100 (0–1450) mL per 24 h or 10 (0–117) mL per voiding (not significant). The median daily variation of creatinine in the urine was 1.1 (0.1–9.4) mmol (not significant), indicating that the 24 h collection periods were respected.


 Recordings on frequency–volume charts are valid and useful in the investigation of patients with voiding symptoms.


The frequency–volume chart (FV chart) is an important tool in the investigation of patients with voiding dysfunction [ 1]. The chart, which is completed daily by the patient over a certain number of days, facilitates history taking and allows the clinician to obtain information about the degree of frequency and nocturia, and the mean volume of urine at each episode of voiding [ 2]. The chart is generally well accepted by patients and when correctly executed, the method is a simple way to obtain an impression of the degree and character of any voiding problems. However, a prerequisite is the proper recording of data. To investigate the validity of self-reporting, we assessed the agreement between data recorded on FV charts and the volume of urine collected in the same period.

Patients and methods

The study comprised 16 men and two women (median age 63 years, range 20–80) who were admitted with voiding problems; all were mentally fit and self-reliant. As part of the diagnostic programme they were requested to record as accurately as possible the time and volume of fluid intake and urinary voids for 3 successive days on a FV chart. Before starting, patients received verbal and written instructions on the completion of the chart. Volumes were estimated using measuring cups provided by the laboratory. Instead of discarding the urine after the volume was measured, patients were instructed to collect it in numbered bottles each day. After the bottles were returned to the laboratory, the volume was determined and the results compared with the accumulated daily recordings on the FV chart. To assess whether the 24 h sampling periods were respected, each bottle was analysed for creatinine. Results are expressed as the median (range) and were compared using Wilcoxon’s matched-pairs signed-rank test, the Mann–Whitney test for unpaired samples, Spearman’s rho and Friedman’s two-way anova. Probability values of <0.05 were considered to indicate significance. The agreement between self-reported and measured volumes was assessed according to Bland and Altman [ 3] by plotting differences between the volumes from the charts and the samples against their mean, using 2 sd as the reference limit of agreement.


All FV charts were completed properly, although one patient kept his chart for only 2 days. The recorded daily fluid intake was 1900 (850–4350) mL and the fluid intake correlated with total voided volume (R=0.45, P<0.01). The daily voiding frequency was 10 (3–23) and the voided volume per micturition was 179 (72–463) ml ( Table 1). Nocturnal voiding accounted for 30% of the micturitions and 23% of the total urine volume. According to the FV charts, the accumulated daily urine output of the 53 daily recordings was 1700 (500–3100) mL and the urine volume of the respective samples was 1750 (510–2550) mL (not significant). The median of disagreement between the measurements was 100 (0–1450) mL per day, or 10 (0–117) mL per voiding. The limits of agreement are shown in Fig. 1a,b. The overestimation of urine output was just as common as underestimation, 26 (49%) vs 27 (51%) of the respective recordings. The difference between recorded and collected volume neither correlated with voiding frequency nor with patients’ age. The median excretion of creatinine in urine during the 3 days was 9.1 (3.3–13.9), 9.1 (3.1–14.6) and 9.8 (3.6–14.3) mmol, respectively, with a day-to-day variation of 1.1 (0.1–9.4) mmol (not significant).

Table 1.  Data from the frequency-volume charts and urine samples of 18 patients Thumbnail image of
Figure 1.

The difference plotted against the mean of a, self-reported and collected total 24 h urine volumes, and b, 24 h urine volume per voiding. 2 sd was taken as the limit of agreement.


It was possible to evaluate six variables from the FV chart used in the present study, i.e. the 24 h voiding frequency, the distribution of voiding, the voided volume, the mean voided volume, the variation of diuresis and the total daily volume. A 24 h sample may be difficult to obtain and is often associated with considerable inaccuracy. Despite the well-known problems with urine sampling, the present study showed a good agreement between the patients’ recordings on the FV chart and the collected urine volume. Although the study mainly comprised elderly patients, all FV charts were completed meticulously. Major inaccuracies in either recordings of frequency or voided volume or both would have resulted in a significant difference between self-reported and collected volume. Urine volume was over- and underestimated with the same frequency, and age, frequency or nocturia did not influence the difference between accumulated values on the charts and the sample urine volume. As the daily excretion of creatinine in urine is fairly constant during steady state conditions, it appeared that the 24 h sampling periods were generally respected.

 Unlike in other reports [ 4], there was a close correlation between fluid intake and total voided volume, but other factors such as food and metabolism contribute to total water consumption, and losses through perspiration are variable. Thus, accurate measurement is less important, but the recording of fluid intake is still valuable in the diagnosis of patients with polyuria and nocturia.

 The usefulness of the FV chart in the assessment of patients with voiding dysfunction has been shown in several studies [ 5[6]–7]. The correlation between simple questioning and the information extracted from the FV chart can differ. Nocturia, incontinence and pad usage usually correlate well, whereas frequency correlates less closely [ 8]. There was a poor agreement between cystoscopically and chart-determined bladder capacity, the former being up to twice that of the latter [ 9], and it has been concluded that the voided volume of the chart mainly represents the volume at ‘first desire to void’. Although computerized voiding diaries have been developed to simplify, augment and automate patients’ recordings of voiding symptoms [ 10], the FV chart remains an easy and reliable method in the diagnosis of patients with voiding disorders.

 In conclusion, the FV charts were completed properly with a good agreement between self-reported and collected urine volume of the same period. Deviations between reported and collected volumes were independent of void frequency and patients’ age. These results support the validity of the FV chart in the diagnosis of patients with voiding disorders.