Intended for healthcare professionals

Letters

Incidence and remission of lower urinary tract symptoms

BMJ 2000; 321 doi: https://doi.org/10.1136/bmj.321.7268.1082 (Published 28 October 2000) Cite this as: BMJ 2000;321:1082

Authors should have used standardised questionnaire

  1. Simon Jackson, consultant obstetrician and gynaecologist. (jackson_simon{at}hotmail.com),
  2. Jenny Donovan, reader.,
  3. Paul Abrams, professor of urology.
  1. John Radcliffe Hospital, Oxford OX3 9DU
  2. Department of Social Medicine, Bristol University BS8 2PR
  3. Southmead Hospital, Bristol BS10 5NB
  4. Department of Accident and Emergency, Bromley Hospital, Kent BR2 9AJ
  5. Department of Obstetrics and Gynaecology, Glostrup County Hospital, University of Copenhagen, DK-2600 Glostrup, Denmark

    EDITOR—Møller et al did not describe in detail the questionnaire that they used in their longitudinal study of lower urinary tract symptoms in women,1 but it seems to have been based on two different instruments. 2 3 The Bristol female lower urinary tract symptoms questionnaire uses a five point scale for reporting symptoms.3 Respondents can reply “never,” “occasionally,” “sometimes,” “most of the time,” or “all of the time” when asked whether they have a particular symptom; when asked about frequency they can reply “never,” “once or less a week,” “2-3 times a week,” “once a day,” or “several times a day.”

    This raises the issue of reproducibility when the questionnaire is completed on more than one occasion. Reproducibility of our instrument was good when a test-retest analysis was performed with a two week interval, there being no apparent change in the underlying condition during that time; 78% of symptom questions were answered identically on both occasions, with no responses changing by more than two categories. The instrument used by Møller et al is reported to have fair to excellent reproducibility, but details are not supplied.

    The definition used for incidence in the authors' paper was “the proportion of women in whom symptoms arise or increase from sometimes to weekly or more.” Remission was defined as “the proportion of women with symptoms occurring weekly or more in whom symptoms decreased to less than weekly.” Thus, seemingly, a change in questionnaire response by one category could be recorded as incidence or remission. If reproducibility is similar to that of the Bristol female lower urinary tract symptoms questionnaire over 20% of women in whom there has been no apparent change in their underlying condition will change their response by one category or more over two weeks.

    Standardised questionnaires that have been tested for validity and reliability should be used whenever possible so that these types of measurement errors can be calculated. Møller et al may simply be confirming test-retest error, and the conclusion that there is an incidence and remission rate of 10.0% and 27.8% for female lower urinary tract symptoms over one year should be interpreted with caution.

    References

    1. 1.
    2. 2.
    3. 3.

    Results have practical implications

    1. Francis Lam, senior house officer. (Jin-Jin.Tang{at}gstt.sthames.nhs.uk),
    2. Stefan Nash, consultant.
    1. John Radcliffe Hospital, Oxford OX3 9DU
    2. Department of Social Medicine, Bristol University BS8 2PR
    3. Southmead Hospital, Bristol BS10 5NB
    4. Department of Accident and Emergency, Bromley Hospital, Kent BR2 9AJ
    5. Department of Obstetrics and Gynaecology, Glostrup County Hospital, University of Copenhagen, DK-2600 Glostrup, Denmark

      EDITOR—Møller et al found that the remission rate of lower urinary tract symptoms was as high as 27.8% in their study; no single treatment modality, including antibiotic treatment, was found to be beneficial on these symptoms overall.1 This has important practical implications, especially for doctors who see patients with this disorder in primary care. In our hospital we have conducted a study examining the extent of non-compliance among patients prescribed antibiotics in the accident and emergency department; we found that 31% of patients admitted to taking none of the antibiotics, or substantially less than the full course.2

      A quarter of women with lower urinary tract symptoms have a remission in one year with or without treatment, and on average one out of three patients will not be compliant with the antibiotic treatment. Given this, it seems rational to limit the use of empirical antibiotics, especially in those with equivocal evidence of infection. Perhaps more emphasis should be placed on communication with the patient, together with advice and reassurance—for example, information leaflets.

      When treatment is indicated we would advocate a short course (three days) of antibiotics, which has been shown to be as effective as a seven day3 or 10 day4 course. This is in line with the recommendation in the Standing Medical Advisory Committee's report striving to reduce the selection pressure for antibiotic resistance.5

      References

      1. 1.
      2. 2.
      3. 3.
      4. 4.
      5. 5.

      Authors' reply

      1. Lars Alling Møller, clinical research fellow. (LarsAM{at}Dadlnet.dk),
      2. Gunnar Lose, professor.
      1. John Radcliffe Hospital, Oxford OX3 9DU
      2. Department of Social Medicine, Bristol University BS8 2PR
      3. Southmead Hospital, Bristol BS10 5NB
      4. Department of Accident and Emergency, Bromley Hospital, Kent BR2 9AJ
      5. Department of Obstetrics and Gynaecology, Glostrup County Hospital, University of Copenhagen, DK-2600 Glostrup, Denmark

        EDITOR—Jackson et al ask about the reliability of our study. Our questionnaire was based on the Bristol female lower urinary tract symptoms questionnaire, but the number of response categories was reduced to “never,” “sometimes,” “once or more a week (often),” or “once or more a day (very often).” To test reproducibility a subgroup of 100 women from the study was asked to fill in identical questionnaires two weeks apart. The subgroup comprised 50 women reporting one or more lower urinary tract symptoms and 50 reporting none. Otherwise selection was randomised. The response rate was 77%.

        To calculate reproducibility, data were split into those for women with symptoms once or more a week (often and very often) and those for women with symptoms less often or with no symptoms. Classification into these groups was thought to be clinically relevant as it separated women with bothersome symptoms from those without.1 Regarding symptoms of urinary incontinence, test-retest analysis showed an agreement of 86.4% in women with bothersome symptoms and of 94.5% in women without (overall 92.2%). For lower urinary tract symptoms, agreements were 93.1% and 97.9% respectively (overall 96.1%).

        In comparison, Jackson et al found that overall 78% of symptom questions were answered identically on both occasions.2 As reproducibility is associated with the prevalence of a specific disease we believe that an overall estimate of reproducibility is a less useful variable. Moreover, by using an overall estimate Jackson et al assume that increasing frequency, as reported in different categories, reflects a continuous scale. We believe that this is not the case: we observed a sharp increase in bothersomeness when shifting category from women with symptoms sometimes to women with symptoms weekly (often).1 We therefore still believe that our design was adequate for the purpose.

        We agree with Lam et al that our study supports a conservative approach to treating lower urinary tract symptoms. Increased knowledge about the natural course of lower urinary tract symptoms is surely a way to allocate the relevant medical resources needed.

        References

        1. 1.
        2. 2.

        Log in

        Log in through your institution

        Subscribe

        * For online subscription