ABC of Urology
BLADDER OUTFLOW OBSTRUCTION
Chris Dawson, Hugh Whitfield
Symptoms
Bladder outflow obstruction due to benign prostatic hyperplasia is a
Obstructive symptoms common problem in elderly men, and by the age of 80 years most men
* Hesitancy will have developed symptoms attributable to this disorder. The
* Poor flow symptoms may be either obstructive or irritative in type. Less common
* Intermittent flow causes of bladder outflow obstruction include bladder neck obstruction
* Post-micturition dribble
* Incomplete emptying and urethral stricture.
The aetiology of the irritative symptoms is poorly understood.
Obstruction of the urethra by an enlarged prostate leads to a poor
urinary flow. The bladder may take longer to generate a high enough
Irritative symptoms pressure to start the flow of urine, and this pressure may not be
sustained over the (prolonged) voiding cycle. This leads to hesitancy
* Frequency of micturition and intermittency. Other causes of irritative symptoms which merit
* Urgency attention are bladder cancer, urinary tract infection, urethral stricture,
* Nocturia
* Urge incontinence bladder diverticula, bladder calculus, and neuropathic bladder
dysfunction.
Evaluation
The assessment of a patient with outflow
obstruction begins with a thorough history
and examination. Symptom scores such as the
American Urological Association's score are
seldom used except in clinical trials.
The clinical examination of a patient
should focus on the urinary tract, especially
the external genitalia, and should also include
a digital rectal examination. A digital rectal
examination is minimally invasive, allows a
rough estimation of the size of the prostate,
will detect locally advanced prostate cancer,
and permits an assessment of rectal sphincter
tone (which may be relevant to the presenting
history).
A urine culture should be performed to
exclude urinary infection or haematuria, and
the serum electrolytes should be checked for
Assessment procedure for patients with bladder outflow obstruction. satisfactory renal function.
BMJ VOLUME 312 23 MARCH 1996 767
Uroflometry is an important procedure that
is often neglected but should be performed
before surgery is considered. In some centres
this is performed at the same time as
ultrasound assessment of the bladder. Patients
with large residual volumes are not candidates
for a policy of watchful waiting-they should
be referred to a urologist. Many men with
symptoms of outflow obstruction are now
being managed entirely in the community by
their general practitioner, and some practices
,,s,~-7-_. run shared care systems with their local
Diagram showing typical flow patterns: normal flow-short time to maximum urology department.
flow, high peak flow, and rapid emptying; obstructed bladder outflow-delay in
starting urination, slow rise to maximum flow (noticeably reduced), and slow
decline with prolonged emptying time; and urethral stricture-reduced peak flow
with typical plateau phase.
Prostate specific antigen is produced by all types of prostate tissue,
Factors affecting prostatic and therefore a change in its value is not specific for prostate cancer.
specific antigen Although the normal range of prostate specific antigen with
Cause of rise in value monoclonal assays is 0-4 ng/ml, many men without prostate cancer will
* Increasing age have a value above the upper limit of normal. Several developments
* Acute retention
* Catheterisation have recently been described to try to increase the specificity of this
* Transurethral resection of the prostate test, particularly in the range of 4-10 ng/ml, where diagnostic
* Prostatitis confusion between prostate cancer and benign prostatic hyperplasia is
* Prostate cancer greatest. Prostate specific antigen density is calculated by dividing the
* Prostatic biopsy serum antigen value by the volume of the gland as determined by
* Benign prostatic hyperplasia
transrectal ultrasonography. Although the serum values of prostate
Cause of fall in value specific antigen in prostate cancer and benign prostatic hyperplasia
* Use of enzyme 5ac-reductase inhibitors
(finasteride) often overlap, the density values have been shown to be significantly
Digital rectal examination does not cause a rise in different for these two groups.
prostatic specific antigen
Prostate specific antigen velocity can be determined by plotting the
antigen values against time and determining the slope of the curve.
Velocity changes have been shown to be higher in men with prostate
Age specific reference ranges for prostatic disease (both benign prostatic hyperplasia and cancer) than in men
specific antigens without, but the technique has little practical application at present.
Age (years) Normal range (ng/mI)
Prostate specific antigen values increase with age, in the absence of any
prostatic disease. Age specific reference ranges have been developed to
40-49 2.5 try to increase the specificity of prostate specific antigen. Patients with
50-59 -3.5 an antigen value above the age adjusted maximum or an abnormal
60-69 z4.5 result on digital rectal examination should be referred for transrectal
70-79 6.5
ultrasonography of the prostate and prostatic biopsy.
Treatment of benign prostatic hyperplasia
Transurethral resection of the prostate remains the gold standard
Complications of transurethral treatment for benign prostatic hyperplasia, with most patients
resection of prostate experiencing a relief of symptoms. Although the complication rate is
low, patients should be counselled carefully before surgery. It is now
* Transurethral resection syndrome (50%)
occur after transurethral resection, although the mechanisms for this
are unclear and may be at least partly psychological. In some patients
erectile function may improve after resection.
768 BMJ VOLUME 312 23 MARCH 1996
Laser prostatectomy offers the prospect of day case prostate surgery.
A neodymium YAG laser is used-this produces light of a wavelength
of 1064 nm, which causes coagulative necrosis of tissue. Various
techniques have been developed for delivering the laser energy to the
prostate. The best coagulation is achieved by deflecting the laser beam
at an angle of 90° to the tissue. Laser prostatectomy may be carried out
either under ultrasound control or under direct vision down a
cystoscope.
The main advantage of laser prostatectomy is the absence of
complications-especially retrograde ejaculation-and of blood loss.
Hyperabsorption of hypotonic fluids during transurethral resection
may lead to profound hyponatraemia, hypotension, and coma (known
as the transurethral resection syndrome). Disadvantages of laser
prostatectomy include the lack of tissue available for histological
examination.
Laser ablation of the prostate has been shown to improve both
symptom score results and flow rates. The availability of many systems
K and techniques, however, cannot disguise the fact that laser
Cystoscopic view of benign prostatic hyperplasia. prostatectomy is in its infancy and that long term prospective studies
are needed to evaluate fully the role of this technique. Open
prostatectomy is now rarely performed but should be considered for
large prostate glands (>100 cm').
Drug treatment of benign prostatic
hypertrophy is increasing and is particularly
suitable for patients unwilling to have surgery,
those deemed unfit for surgery, and those
with only moderate symptoms.
Drug treatment of benign prostatic hVrpertrophy Urethral stents have recently been used in
* Finasteride blocks the enzyme 5 a-reductase in the prostate, preventing patients with benign prostatic hyperplasia.
conversion of testosterone to dihydrotestosterone (th ie more active form of The short term results with these prostheses
testosterone necessary for the development of benigrn prostatic are encouraging, but long term complications.
hypertrophy) particularly urinary tract infection and stone
* Several studies in progress suggest a benefit to sormne patients receiving formation, have been noted.
this treatment
* Drugs blocking the adrenergic receptors in the pros;tate's smooth muscle Balloon dilatation of the prostate results in
have also been shown to be effective a subjective improvement in more than two
* Patients with minor symptoms often need no treatrnent once the nature thirds of patients. Objective analysis has,
of their complaint has been explained to them-they can be followed up as
outpatients however, been less than impressive, and this
technique has fallen into disuse.
Recent studies have shown that microwave
treatment may relieve symptoms in some
patients. The durability of these results is not
known at present, but such treatment may be
suitable for patients unfit for surgery.
Less common causes of bladder outflow obstruction
Bladder neck dysfunction
Bladder outflow obstruction due to bladder neck dysfunction (also
known as bladder neck dyssynergia) is found almost exclusively in
young and middle aged men. It is characterised by an incomplete
opening of the bladder neck during voiding. The prostate is normal on
digital rectal examination. The condition is diagnosed urodynamically,
with characteristic trapping of contrast medium seen in the prostatic
urethra.
Definitive treatment is with cystoscopic bladder neck incision under
a general anaesthetic. Pharmacological treatment may be more
appropriate, however, in a man who wishes to remain fertile as the
incidence of retrograde or absent ejaculation after a bladder neck
incision varies between 15% and 50%. Treatment of this condition
with drugs that block a adrenergic receptors (such as indoramin) may
Cystoscopic view of bladder neck obstruction. relieve symptoms, but this treatment can be associated with unwanted
side effects, such as loss of libido and impotence.
BMJ vOLuME 312 23 MARCH 1996 769
Key references Urethral stricture
Carter HB, Pearson JD. PSA velocity for the diagnosis of early prostate Urethral stricture is also a cause of bladder
cancer. Urol Clin North Am 1993;20:665-70 outflow obstruction and is suggested by a
De Geeter P. The case against balloon dilatation of the prostate. European history of urethral trauma, previous
Urology Update Series 1993;2:146-51
catheterisation, or sexually transmitted disease
Denis LJ. Diagnosing benign prostatic hyperplasia versus prostatic cancer. (although it may occur de novo in the absence
Br J Urol 1995;76(suppl 1):17-23
Partin AW, Oesterling JE. The clinical usefulness of prostatic antigen:
of any of these factors). Uroflometry typically
update 1994. J Urol 1994;152:1358-68 shows a trace with a plateau and a prolonged
Seaman E, Whang M, Olsson CA, Katz A, Coner WH, Benson MC. PSA voiding cycle. Treatment initially is by
density (PSAD). Role in patient evaluation and management. Urol Clin urethrotomy, although complex or recurrent
North Am 1993;20:653-63 cases may need urethroplasty.
The ABC ofUrology is edited by Chris Dawson, a senior registrar in urology at the
The two cystoscopic views were provided by Mr H N Blackford Edith Cavell Hospital, Peterborough, and Hugh Whitfield, a consultant urologist at the
of the Edith Cavell Hospital, Peterborough. Central Middlesex Hospital and the Institute of Urology and Nephrology, London.
Statistics Notes
Transforming data
J Martin Bland, Douglas G Altman
This is the 1 7th in a series of We often transform data by taking the logarithm, logarithms.2 So a variable which is the product of
occasional notes on medical square root, reciprocal, or some other function of the several factors has a logarithm which is the sum of
statistics data. We then analyse the transformed data rather than several factors and so will follow a normal distribution.
the untransformed or raw data. We do this because Thirdly, any relation between variance and mean
many statistical techniques, such as t tests, regression, over several groups is usually fairly simple. The
and analysis of variance, require that data follow a variance may be proportional to the group mean, the
distribution of a particular kind. The observations mean squared, the mean to the fourth power, etc. For
themselves must come from a population which such relations simple transformations can be found
follows a normal distribution,' and different groups of which will make the variance independent of the mean.
observations must come from populations which have If the variance is proportional to the mean we can use
the same variance or standard deviation. We need this the square root transformation. This is often the case
uniform variance because we estimate the variance for data which are counts of things or events-for
within the groups, and we can do this well only if we example, the number of cells of a particular type in a
can assume it to be the same in each group. Many given volume of blood or number of deaths from AIDS
biological variables do follow a normal distribution in a geographical area over one year. Such data tend to
with uniform variance. Many of those which do not can follow a Poisson distribution, which has its variance
be made to do so by a suitable transformation. equal to its mean. If the variance is proportional to the
Fortunately, a transformation which makes data follow mean squared-that is, the standard deviation is
a normal distribution often makes the variance proportional to the mean-we use the logarithmic
uniform as well, and vice versa. In this note we shall try transformation. This is the most frequent case
to explain why this is the case. in practice, suitable for variables such as serum
Firstly, the normal distribution and uniform cholesterol. If the variance is proportional to the mean
variance go together. It can be shown mathematically to the fourth power-that is, the standard deviation is
that if we take random samples from a population the proportional to the mean squared-we use a reciprocal
means and standard deviations of these samples will be transformation, used for highly variable quantities
independent (and thus uncorrelated) if the population such as serum creatinine. Thus we can transform the
has a normal distribution. In other words, the standard data to make the variance unrelated to the mean, in
deviation of the samples will not be related to the mean. which case the data are likely to follow a normal
Furthermore, if the mean and standard deviation are distribution.
independent the distribution must be normal. This is Some people ask whether the use of a transformation
Department of Public harder to credit, but it is true. is cheating. There is no reason why the "natural" scale
Health Sciences, St Secondly, if we add together many variables we should be the only, or indeed the best, way to present
George's Hospital Medical usually get a normal distribution. For example, the measurements. pH, for example, is always presented
School, London SW17 ORE
J Martin Bland, professor of central limit theorem shows that the means of large as a logarithmic measure, pH= -log10(H+), where H+
medical statistics samples will follow a normal distribution, whatever the is the concentration of hydrogen ions in moles per
distribution of the observations themselves.' Similarly, cubic decimetre. Thus the "natural" scale is 10_pH.
ICRF Medical Statistics if a biological variable is the result of the sum of many This natural scale is very awkward to use, and the
Group, Centre for influences, it will follow a normal distribution. Human logarithm is always used instead.
Statistics in Medicine, height is an example. Many biological measurements If we can transform data to follow a normal distri-
Institute ofHealth are not like this, however, but are the product of bution with variance independent of the mean, valid
Sciences, PO Box 777, several factors. Substances in blood, for example, may analyses can be carried out on this transformed scale.
Oxford OX3 7LF be removed at a rate depending on the level of some There is one drawback, however, as confidence
Douglas G Altman, head
other substance, which in turn is produced at a rate intervals on the transformed scale may be difficult to
Correspondence to: which depends on something else, and so on. We have interpret. We shall deal with this in a subsequent note.
Professor Bland. the product of several influences multiplied together,
rather than the sum. If we take the logarithm of the 1 Altman DG, BlandJM. The normal distribution. BMJ 1995;310:298.
BMJ 1996;312:770 product of several variables, we get the sum of their 2 Bland JM, Altman DG. Logarithms.BM_ 1996;312:700.
770 BMJ VOLUME 312 23 MARCH1996