|
One of the approximately 6,000,000 men in your age group in the
United States who suffers from an enlarged prostate. Or perhaps you know
someone who has the symptoms of what is medically known as Benign Prostatic
Hyperplasia (BPH). This non-cancerous but often difficult condition can now, in
most instances, be treated without prescribed medications or surgery.
The American Urological Association (AUA) Symptom Index is
recommended as the symptom scoring instrument to be used in the initial
assessment of each patient presenting with symptoms of prostatism. When the AUA
system is used, symptoms should be classified as mild (0 to 7), moderate (8 to
19), or severe (20 to 35). The symptom score should be the primary determinant
of treatment response or disease progression in the followup period.
Most patients seeking treatment for BPH do so because of
bothersome symptoms that affect the quality of their lives. Tools to quantify
those symptoms are important to determine the severity of the disease and to
document the response to therapy, to assess the patient's symptoms, and to
follow them over time to determine the progression of the disease and points of
necessary intervention. Such assessment tools also allow comparison of the
effectiveness of various interventions. To the patient, of course, relief of
symptoms is the single most important outcome, not flowrate, detrusor pressure,
or urethral resistance factors.
At least two significant attempts to develop tools for assessing
a patient's symptom status were made prior to the development of the AUA
Symptom Index. In one system, by Madsen and Iversen (1983), patients are
interviewed specifically about symptoms that include the quality of their
urinary stream, straining to void, hesitancy, intermittency, bladder emptying,
stress incontinence or postvoid dribbling, urgency, frequency, and nocturia.
All symptoms are graded on a scale of 0 to 4, and the scores are added for a
total score that can reach 27 points. Patients scoring less than 10 points are
considered mildly symptomatic. Patients scoring from 10 to 20 points are
considered moderately symptomatic. Patients scoring above 20 points are
considered severely symptomatic. The symptom-score assessment is integrated in
a comprehensive clinical evaluation sheet including residual urine, cystoscopic
findings, presence or absence of urinary tract infections, urinary retention,
and renal failure.
The Madsen-Iversen scoring system cannot be regarded as an ideal
system to evaluate BPH patients. One especially important detraction is that
differential weights for symptoms were assigned on the basis of expert opinion,
not empirical data from patients.
Another tool to quantify patient symptoms is the Boyarsky system
(Boyarsky, Jones, Paulson, et al., 1976). An ad hoc group convened by the Food
and Drug Administration (FDA) in 1975 developed guidelines for the
investigation of BPH. The main part of these guidelines includes the Boyarsky
tool. This symptom-scoring system evaluates the severity of nocturia,
frequency, hesitancy, intermittency, terminal dribbling, urgency, impairment of
size and force of stream, dysuria, and sensation of incomplete voiding. The
system allows 0 to 3 points for each of 9 questions, for a maximum of 27
points.
The Boyarsky scoring system, like the Madsen-Iversen system, is
not ideal. Although the symptoms in the Boyarsky system are each scored
similarly, from 0 to 3, symptoms such as terminal dribbling are weighted
equally with symptoms that appear to be better predictors of the presence or
absence of BPH. As in the Madsen-Iversen tool, the equal-weighting scheme is
based on opinion rather than data. Moreover, the Boyarsky tool asks about
several dimensions of some symptoms, such as frequency and severity, resulting
in response frames that are not collectively exhaustive.
Recently, during the time the BPH Guideline Panel was meeting,
the AUA formed a Measurement Committee to develop a symptom and quality-of-life
questionnaire to provide outcome measures for a prospective study of BPH
treatment. The committee initially drafted a composite questionnaire consisting
of 73 questions, in part by reviewing prior instruments including the
Madsen-Iversen and Boyarsky scoring systems. Although 18 questions dealt with
the frequency and severity of urinary symptoms, another 17 dealt with the issue
of how much the patient was bothered by these symptoms. Other items covered
health-related quality of life, sexual function, and continence.
In a pilot study, the full questionnaire was administered to
patients with a clinical diagnosis of BPH who were drawn from urologic
practices and to younger control subjects without urinary complaints who were
drawn from a general medical practice (Barry, Fowler, O'Leary, et al., 1992a).
Based on how well the individual symptom questions correlated to two ratings of
the overall "bother" of each subject's urinary difficulties, a question set was
selected for further testing. It covered the symptoms of incomplete emptying,
frequency, intermittency, urgency, a weak stream, hesitancy, and nocturia.
This seven-question set was internally consistent (Cronbach's
alpha = 0.85). Moreover, the reliability of the index was high, with a
test-retest correlation of 0.93.
The index correlated strongly with patients' global ratings of
their urinary difficulties (r = 0.78), providing evidence of the construct
validity of the instruments. The tentative AUA index was also correlated with
the Madsen-Iversen and Boyarsky scores obtained on the same subjects. The
correlations, 0.85 and 0.93, respectively, were high. This provided additional
evidence of construct validity (Barry, Fowler, O'Leary, et al., 1992b).
Finally, as a test of criterion validity, the ability of the
index to separate the BPH patients from the control subjects in the validation
study was examined. The area under the Receiver Operating Characteristic (ROC)
curve for these indices, a measure of discrimination that uses each patient's
score as a diagnostic test for BPH, was 0.87. This measure suggests that a
randomly selected BPH patient and a randomly selected control subject from the
study population would be correctly classified 87 percent of the time.
Based on the results of this initial validation study, several
questions were modified (in response to subject feedback), and the 7-item index
was revalidated using the same design (this time with 107 BPH patients and 49
control subjects). The AUA Symptom Index as administered in the second
validation study is shown in Figure 1.
Each question on the AUA Symptom Index can yield 0 to 5 points,
producing a total symptom score that can range from 0 to 35. On revalidation,
the scores again demonstrated high internal consistency (Cronbach's alpha =
0.86) and high test-retest reliability (r =0.92). Scores were again correlated
with subjects' two global ratings of their urinary problem (r = 0.65 and 0.72)
and again discriminated BPH patients from control subjects (ROC area = 0.85).
As a final validation step, the sensitivity of the AUA Symptom
Index, its ability to capture clinically important changes in patients'
conditions, was assessed. Twenty-seven men with symptomatic BPH answered the
questionnaire before and 1 month after having a prostatectomy. Their scores
dropped from a mean of 17.6 to 7.1 over this period (95-percent CI 8.1-12.9
percent). This is statistically a highly significant result.
Data from the two validation studies were pooled to correlate
subjects' symptom scores with the subjects' global ratings of how bothered they
were by their condition (Table 4). These data can be used to divide the range
of the AUA score into "mild," "moderate," and "severe" symptom categories. As
can be seen from Table 4, only 1 of 120 men with scores from 0 to 7 was
bothered more than a little by his symptoms; these men can be considered the
mild symptom group. The majority of the 108 men with symptom scores from 8 to
19 were still bothered "not at all" or "a little." Only 4 of the 108 men were
bothered "a lot." These men can be labeled as having moderate symptoms. Most
men with scores from 20 to 35 were bothered by their condition "some" or "a
lot" and can be considered to have severe symptoms.
Clearly, symptom scores alone do not capture the morbidity of a
prostate problem as perceived by the individual patient. Symptom impact on a
patient's lifestyle must be considered as well. An intervention may make more
sense for a moderately symptomatic patient who finds his symptoms very
bothersome than for a severely symptomatic patient who finds his symptoms quite
tolerable.
The division of the AUA symptom scores into mild, moderate, and
severe ranges is somewhat arbitrary. Depending on how the scores are to be
used, different cutoff points may be more suitable for different purposes.
Using the cutoff points chosen, Table 5 shows the distribution of scores for
BPH patients and control subjects in the validation studies previously
discussed.
In conclusion, the panel considers the objective documentation
of a patient's symptom level the most essential part of its recommendation for
the diagnosis, evaluation, treatment planning, and followup of patients with
prostatism. The AUA Symptom Index, currently the best available instrument, is
recommended by the panel. However, the panel emphasizes that optimal treatment
decisions for individual patients will also need to take into account how a
given level of symptoms affects each patient's quality of life (bothersomeness)
|