Use of questionnaire for assessment of nocturia

In this dissertation, the presence of nocturia is the main outcome. In the FINNO Study, we assessed occurrence and bother of LUTS by validated questionnaire instruments (primarily by DAN-PSS). In clinical trials, in addition to symptom questionnaires, FVCs (voiding diaries) have commonly been used as a primary tool for measuring LUTS. It is important to question whether a questionnaire is reliable for the assessment of nocturia episodes. Questionnaires have been reported to either overestimate (Ku et al. 2004, Yoshimura & Terai 2005) or underestimate (Blanker et al. 2000a), or neither (Jaffe et al. 2002, Stav et al. 2009) overestimate or underestimate prevalence estimates of nocturia.

Results from a recent retrospective US study suggested that women tend to overestimate their daytime urinary frequency, whereas they were accurate about their nocturnal frequency (Stav et al. 2009). The study was conducted among 601 women (aged 30 to 91) with completed FVC. The authors proposed that “awakening to the toilet is probably more bothersome and has more impact on patient memory” and

in patients who reported a daytime frequency greater than ten voids per day (Stav et al.

2009).

In the Krimpen study among a community-based sample of men aged 50 to 78, I-PSS nocturia item was compared to (3-day) FVC. Out of 3,924 contacted, 1,688 (50% of eligible men) responded to the questionnaire. Among those, 1,597 men completed the diary. However, nocturnal and diurnal voiding frequency could be estimated among 1,211 men (approximately 36% of eligible study sample; present author´s calculation) who also provided information on time of arising and bedtime. The results indicated a poor agreement between questionnaire and diary. Furthermore, questionnaires underestimated the prevalence of nocturia. A similar conclusion (although contradictorily stated) was also reported in a US study (Latini et al. 2004): poor agreement between questionnaire and FVC. In this particular study, the I-PSS nocturia question and FVC information were compared among 284 ambulatory men (aged 18 to 66, recruited by advertisements at hospitals). However, as stated by the authors, the main limitation of the study was comparing a 24-hour diary to the I-PSS assessing nocturia during the last month. Furthermore, there was no association between nocturia frequency and age thereby raising the question of the validity of the results (Latini et al.

2004).

There is no doubt as to whether questionnaires are vulnerable to recall bias:

questionnaires rely on the subject’s memory. He or she may totally forget an event or misrecall the timing. In earlier studies increasing accuracy of questionnaires (to estimate FVC result) has been found among those with few nocturia episodes (0, 1, or 2 voids/night) (Jaffe et al. 2002, Yoshimura & Terai 2005). These are ´the critical episodes´ in our nocturia case classification. Furthermore, all those who reported (by questionnaire) four, five, or six voids/night had at least two voids/night in the diary assessment in a recent US study (Jaffe et al. 2002). Hence, our classification of nocturia to two groups (subjects with <2 voids/night, and subjects with ≥2 voids/night) probably decreases the impact of recall uncertainty and bias. One can speculate that agreement between FVC and the questionnaire would have been better in earlier studies if fewer groups (of nocturia classification) had been used.

Questionnaires assess ´the typical frequency´ over a longer time period retrospectively (e.g. last month in the AUA-SI), whereas voiding diaries typically assess one, two, or three nights, or sometimes for one week prospectively. Hence, the studies compared two different time periods. The nights during FVC may not have been typical.

Substantial temporal variation in the frequency of nocturnal voidings has been shown (Yoshimura & Terai 2005). Various reasons for such variation have been proposed, such as unusual fluid intake, unusual duration of time in bed, feeling colder or warmer during time in bed and more difficulty getting to sleep (Yoshimura & Terai 2005).

Earlier studies comparing questionnaires to FVCs used the I-PSS nocturia question (identical with the AUA-SI nocturia item). In the FINNO Study, responses to nocturia questions from the DAN-PSS (Schou et al. 1993) and the AUA-SI (Barry et al. 1992) were combined to assess occurrence of nocturia. In the case of conflict between the responses to the two questions, the DAN-PSS was regarded as the gold standard so that the DAN-PSS response nearer to the AUA-SI response was chosen. We have shown excellent consistency between the two nocturia questions. However, using the PSS as the ´gold standard´ question might provide a more reliable answer as the DAN-PSS nocturia question is shorter and probably easier to understand (both in English and Finnish) than AUA-SI (i.e. I-PSS) nocturia item.

Owing to the subjective and symptomatic nature of urinary incontinence, there is no gold standard for outcomes assessment (Dmochowski 2001). However, FVC is a key tool in diagnosing nocturia (Blanker et al. 2000a, Weiss 2006). By FVCs, one can assess nocturia episodes and nocturnal polyuria, the latter cannot be discovered by questionnaires. Hence, FVCs are indispensable in clinical practice (Weiss 2006).

However, regarding the epidemiological research of nocturia, the case is unclear. Do diaries provide more accurate information in ´real-life research´? Besides the fluctuating character of nocturia, two other aspects are also important: 1) whether it is possible to obtain a representative sample when using diaries (of several days), and 2) whether FVCs impact on ´normal voiding habits´. When studying the prevalence and

et al. 2000a). Diaries were fully completed by 36% of the eligible study sample. In the FINNO Study, nocturia information was provided by more than 70% of men aged 50 to 79 years. The reason for this clear difference in response proportion is obvious.

Completing a FVC (especially documenting volumes voided) is much more troublesome and time-consuming than filling in a questionnaire. Studies using FVCs (with lower response proportions) use probably more selected samples (non-response and non-compliance bias). It is harder to recruit subjects with less symptoms: this is also supported by higher I-PSS score among participants (than among non-participants) in the Krimpen study (Blanker et al. 2000a). Another unclear aspect is the potential psychological impact of the diary on the subject. Participating in a prospective FVC evaluation may alter voiding habits. Remarkable placebo response, which is commonly seen in drug trials may partly be due to the effect of FVC assessment (´bladder training effect´) (Stav et al. 2009).

Is questionnaire or FVC more reliable for the assessment of nocturia episodes? In a population-based study of all adult ages (also including young subjects with much less LUTS), this is a very difficult question to answer. Ideally, FVCs provide more (accurate) information, including information on NP. In reality, one should take into account: 1) increased selection bias, 2) fluctuating character of nocturia, and 3) possible impact of prospective evaluation on nocturnal voiding habits.