Table 1 shows how often study subjects used various examinations and tests for initial evaluations and evaluations before the surgical treatment of men who had symptoms suggestive of benign prostatic hyperplasia (See Appendixes 4 and 5). For initial diagnostic assessments, the most commonly used tests were urinalysis (90.3%) and digital rectal examination (86.6%). Symptom assessment was used in 46.9% of cases. Of respondents, 163 (58.8%) reported that treatment decisions were not based on the symptom questionnaire (see Appendix 6). Although sonography of the prostate was performed by 45.5% of the urologists, less than 10% utilized other imaging studies, such as abdominal x-rays (7.9%), intravenous urography (0.7%) or renal sonography (3.6%). Prostate specific antigen examinations were performed in 62.5% of diagnoses, but cytologic examination in only 3.6%. Uroflowmetry and post-void residual measurement were performed in 34.7% and 33.6%, respectively, but pressure-flow studies in only 1.1%. Almost all the urologists reported that they used urinalysis, digital rectal examination and prostate specific antigen examinations for evaluations before surgical treatment. Before surgical treatment, nearly 80% of the respondents "almost always" measured transrectal ultrasonography, uroflowmetry and post-void residual, creatinine and symptom score.

Table 2 lists the use of various treatments for men with moderate symptoms and for men who request or want treatment (see Appendix 7). According to most guidelines, medical therapy was the main treatment for these patients. Of the pharmacological options, almost all urologists reported that they prescribed alpha-blockers. Of the respondents, 55.6% had prescribed alpha-blockers as the primary treatment and 41.9% prescribed alpha-blockers with 5-alpha reductase inhibitors as the primary treatment. Physicians with a practice duration of less than 10 years who had also worked at private clinics prescribed alpha-blockers and alpha-blockers with 5-alpha reductase inhibitors more frequently than their counterparts. 81.2% of urologists perceived that selective alpha-blockers are different in terms of efficacy and 82.7% felt they also differed in safety (see Appendixes. 8 and 9). When respondents were asked, "What is the prostate volume threshold at which you would prescribe 5-alpha reductase inhibitors?" One hundred three (37.2%) urologists replied "30-39 g" and 90 (32.5%) respondents reported "40-49 g" as their threshold. The other survey choices and responses were "10-19 g" chosen by 0 doctors (0.0%), "20-29 g" chosen by 15 doctors (6.4%), "> 50 g" chosen by 9 doctors (3.2%), "regardless of prostate volume" chosen by 40 doctors (14.4%) and "I do not prescribe 5-alpha reductase inhibitor" chosen by 20 doctors (7.2%) (see Appendix 10).

Digital rectal examination, urinalysis, and use of a symptom questionnaire to assess the patient's symptoms have been recommended by most of the guidelines.⁷ The digital rectal examination is emphasized as an important test for identifying prostatic abnormalities. Several guidelines also suggested that the patient's quality of life be measured. In this survey, although digital rectal examination and urinalysis were performed routinely, symptom assessments have not been routinely used by a significant number of respondents. In addition, a significant number of respondents reported that treatment decisions were not based on the symptom questionnaire although, in most guidelines, treatment decisions were recommended on the basis of the severity of the patient's symptoms alone or, in addition, on the basis of how bothersome these were. Furthermore, the use of a symptom questionnaire can present an opportunity to more objectively monitor patient response to therapy.

The use of uroflowmetry and residual volume measurements, which are optional, according to most guidelines, varied among study subjects. However, a small minority seemed to perform upper tract imaging or cystoscopy routinely, according to the recommendation of selective use by most guidelines. Pressure-flow measurements also appeared to be determined infrequently.

The best treatment from the patient's viewpoint may differ from that believed by the physician to be the most efficacious treatment. Patients may prefer less effective therapy if it also has less risk or cost. Medical therapies are not as efficacious as surgical therapies but may provide adequate symptom relief with fewer and less serious associated adverse events. In this survey, nearly all urologists reported that they prescribed alpha-blockers (alone or combined with 5-alpha reductase inhibitors) for men with moderate symptoms and for men who request or want treatment. Meta-analyzed data from the Panel's evidence-based review suggest that alfuzosin, doxazosin, tamsulosin, and terazosin are similarly effective in partially relieving symptoms, producing, on average, a 4-to-6 point improvement in the American Urological Association Symptom Index.⁹ In this survey, however, 81.2% of the respondents perceived that selective alpha-blockers are different in terms of efficacy. In addition, 82.7% of urologists perceived that selective alpha-blockers also differ in terms of safety. The adverse event profile appears slightly different between the four alpha-blocking agents; tamsulosin, for example, appears to have a lower probability of orthostatic hypotension but a higher probability of ejaculatory dysfunction associated with it than do the other alpha blockers.⁹ Large, well-designed, direct comparator trials are needed to substantiate claims of superior safety.

Data over the last five years suggest that larger prostates are associated with more progressive disease, and a greater likelihood of symptom progression,¹⁶ flow rate deterioration,¹⁷ increased prostate growth^18,19 a urinary retention, and prostatic surgery.²⁰ Data from the Medical Therapy of Prostate Symptoms study indicates that alpha-blockers delay acute urinary retention but do not prevent it.¹⁶ In our study, although most respondents did not prescribe 5-alpha reductase inhibitors as a monotherapy, 41.9% of respondents reported that they prescribed alpha-blockers with 5-alpha reductase inhibitors as the primary treatment for the men with moderate symptoms and for men who request or want treatment. Generally, the combination of an alpha-blockers and 5-alpha reductase inhibitors is an appropriate and effective treatment for men with the lower urinary tract symptoms associated with demonstrable prostatic enlargement. At present, however, no absolute threshold values are provided. Most Korean urologists prescribed 5-alpha reductase inhibitors based on prostate size. Patients most likely to benefit from combination therapy are those in whom the baseline risk of progression is significantly higher, in general, than in patients with larger glands and higher prostate-specific antigen values.⁹

These findings suggest that local health care resources or cultural differences may influence practice patterns. In addition, although guidelines provide a framework for evaluation and treatment, these leave a great deal of room for a physician's personal opinions. Irani et al.¹⁰ suggested that countries or organizations with no resources to create their own high-quality Clinical Practice Guidelines should adapt their practice policies from Clinical Practice Guidelines that score highly when formally appraised. Although the process of creating the international guidelines involved a review of the worldwide literature on the diagnosis and treatment of benign prostatic hyperplasia, local health care issues, such as socialized medical systems, manpower issues, availability of extensive technology, the unique perspective of local doctors, or cultural differences in urological practices should be considered when adapting practice policies from Clinical Practice Guidelines.²¹

Some aspects suggest a need for caution when analyzing the present data. First, the frequencies of the various pretreatment investigations are estimates deduced from a 6-step rank scale, inevitably leading to some inaccuracy as to the single number. Second, the survey response was only about 32.6%. Non-response must necessarily result in an increase in random sampling error larger than that which would be expected if most of the questionnaires had been returned. The statistical consequence of increased sampling error is that it makes it more difficult to detect small but real differences as significant. Third, a more serious consequence of non-response is non-response bias. Non-response bias occurs if the subjects who respond to a survey are consistently different from those who do not respond. We suspect that the non-responders were not interested in this survey and, therefore, might not be eager to follow clinical practice guidelines. Therefore, the true indifference of Korean urological practitioners to benign prostatic hyperplasia might have been underestimated by this survey. Unfortunately, though, because we have no demographic information on the non-respondents which might allow a comparison with respondents, the importance of this effect in our survey is unclear. Finally, our findings must be interpreted cautiously because our data on urologists' practice patterns are based on self-reported behavior, not actual behavior as measured by audit. Questionnaire studies can be criticized if there are no cross-checks to assess the validity of the data.