- Email: [email protected]
Factors influencing morbidity in patients undergoing transurethral resection of the prostate
ADULT UROLOGY
FACTORS INFLUENCING MORBIDITY IN PATIENTS UNDERGOING TRANSURETHRAL RESECTION OF THE PROSTATE TOYOAKI UCHIDA, MAKOTO OHORI, SHIGEHIRO SOH, TAKEFUMI SATO, MASATSUGU IWAMURA, TERUAKI AO, AND KEN KOSHIBA
ABSTRACT Objectives. Transurethral resection of the prostate (TURP) has become the primary method to relieve bladder outlet obstruction for patients with benign prostatic hyperplasia (BPH). Data from 3861 consecutive patients with BPH who underwent TURP from 1971 to 1996 at our hospital were retrospectively analyzed. Methods. The patients were classified into two groups comprising 1930 patients who underwent TURP from 1971 to 1985 (early group) and 1931 patients who underwent TURP from 1985 to 1996 (late group). Risk factors associated with blood transfusions and perioperative complications were analyzed in these patients. Results. Mortality, morbidity, and blood transfusions were noted in 5 (0.1%), 516 (13.4%), and 507 (13.1%) patients, respectively. The blood transfusion and morbidity rates decreased over the 25-year period (P ,0.001, chi-square test for trends), which was reflected in a decrease in these rates in the late group (6.1% and 9.5%, respectively) compared with those of the early group (20.2% and 17.2%, respectively). Postoperative bleeding and morbidity were closely related to prostatic gland size and operating time. The most significant differences for the risk of a blood transfusion were related to resection time, the amount of tissue resected, age, and the decade (1970s, 1980s, or 1990s) in which the surgery was performed (P ,0.0005), whereas resection time was significantly correlated with morbidity (P ,0.0005). As risk factors for each complication, the time of surgical resection, the decade of surgery, and the amount of tissue resected directly correlated with the incidence of extravasation and hemostatic procedures (P #0.003), whereas the incidence of postoperative epididymitis positively correlated with a preoperative vasectomy and a closed drainage system (P ,0.0005). Conclusions. Since the 1970s, the rates of blood transfusions and morbidity have decreased for patients undergoing TURP. Advances in techniques, instrumentation, and surgical and perioperative management, including anesthesia, have made TURP a relatively safe procedure, and it remains an effective means for treating patients with BPH. UROLOGY 53: 98–105, 1999. © 1999, Elsevier Science Inc. All rights reserved.
F
or the past 50 years, transurethral resection of the prostate (TURP) has been the primary method for relieving bladder outlet obstruction in patients with benign prostatic hyperplasia (BPH).1– 4 Although TURP is an effective treatment for most men with symptomatic BPH, 20% to 25% of patients who undergo the operation do not have a satisfactory long-term outcome.5 Another concern is that the rate of reoperation is 15% to 20% in From the Department of Urology, Kitasato University School of Medicine, Kanagawa, Japan Reprint requests: Toyoaki Uchida, M.D., Department of Urology, Kitasato University School of Medicine, 1-15-1 Kitasato, Sagamihara, Kanagawa 228, Japan Submitted: March 31, 1998, accepted (with revisions): July 6, 1998
98
© 1999, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
men followed up for 10 years or longer.6 Also, in several preliminary retrospective studies, the life expectancy of men undergoing TURP was shorter than that of men undergoing an open prostatectomy as treatment for BPH.7,8 Reported complications include retrograde ejaculation in 70% to 75% of men, impotence in 5% to 10%, postoperative urinary tract infection in 5% to 10%, and some degree of urinary incontinence in 2% to 4% of patients. Furthermore, 5% to 10% of patients require a blood transfusion.1–3,9,10 Several minimally invasive therapies, such as transurethral incision, balloon dilation, hyperthermia, laser surgery, highintensity focused ultrasound, and drug therapy with 5-alpha-reductase inhibitors or alpha-blockers, have been used in the past 5 years as alterna0090-4295/99/$19.00 PII S0090-4295(98)00524-X
TABLE I.
Age (yr) Resection time (min) Amount of tissue resected (g) Mortality Morbidity Blood transfusion
Characteristics of 3861 patients undergoing TURP Early Group (1971–1985) (n 5 1930)
Late Group (1985–1996) (n 5 1931)
P Value
Total (n 5 3861)
70.2 6 7.7 65.0 6 32.8 25.9 6 21.7 4 (0.2%) 332 (17.2%) 389 (20.2%)
70.1 6 8.0 68.3 6 33.8 23.3 6 19.5 1 (0.5%) 184 (9.5%) 118 (6.1%)
0.6927 0.0021 ,0.0001 0.1792 ,0.0001 ,0.0001
70.1 6 7.9 66.6 6 33.3 24.6 6 20.7 5 (0.1%) 516 (13.4%) 507 (13.1%)
KEY: TURP 5 transurethral resection of the prostate. Data presented are mean value 6 SD or number (%) of patients.
tives to TURP.11–18 A recent survey indicated that the first choice of treatment for patients with BPH based on American Urological Association (AUA) symptom assessment was observation for an AUA score of 0 to 7 (77%), chemotherapy with alphablockers for an AUA score of 8 to 19 (65%), and TURP for an AUA score of 20 to 35 (41%).19 TURP is preferred most by American and Japanese urologists in patients with severe symptoms and a larger prostate gland.19,20 Since 1990, the frequency of TURP as a surgical treatment for patients with BPH has decreased compared with the period from 1960 to 1990. The availability of less invasive methods as well as economic concerns may account for this decrease.21 In the present study, 3861 patients treated with TURP from 1971 to 1996 were retrospectively examined and compared with regard to history, operative management, and complications. MATERIAL AND METHODS PATIENT CHARACTERISTICS Between August 1971 and December 1996, 3861 patients underwent TURP for BPH at our university hospital. Fortyfive variables were evaluated for each patient and included specific information on demographics, operative management, intraoperative complications, and immediate and longterm postoperative complications. Also, the patients were classified into two groups, with 1930 patients who underwent TURP from 1971 to 1985 in the early group and 1931 patients who underwent TURP from 1985 to 1996 in the late group (Table I).
SURGICAL TECHNIQUE Lumbar, epidural, or general anesthesia was used in all patients unless medically contraindicated, and all patients were placed in the dorsal lithotomy position. All resections were conducted using an Iglesias-type resectoscope (CIRCON ACMI, Santa Barbara, Calif), and nonhemolytic solutions were used as the irrigating solution. The prostatectomies were performed using the technique of Cockett and Koshiba,22 where the prostate adenoma was completely excised, thus exposing fibers of the surgical capsule throughout the prostatic fossa. The procedure was initiated and conducted for the first 15 to 60 minutes by senior surgical residents who were not familiar with the TURP procedure, under the supervision of faculty surgeons, who were then succeeded by an experienced staff surgeon who completed the operation. Routine serum electrolyte and blood cell count determinations were performed both UROLOGY 53 (1), 1999
during the procedure and postoperatively. Depending on the electrolyte imbalance, a 10% saline solution was administered intravenously to prevent hyponatremia. For patients undergoing surgery from 1971 to 1986, the procedure was terminated by insertion of a 22F two-way Foley balloon catheter into the bladder, which remained in the bladder for 3 days postoperatively. Since 1986, a 22F three-way Foley balloon catheter has been used to provide continuous irrigation with a saline solution for 3 days postoperatively, depending on the postoperative course. Gentle traction was applied at the bladder neck in most patients for 30 to 60 minutes postoperatively. From 1971 to 1984, nonclosed urinary drainage was used, which was changed to closed urinary drainage in 1985. Antibiotic agents were routinely administrated intravenously for 3 days postoperatively and then orally for 1 to 2 weeks to prevent postoperative infection.
STATISTICAL ANALYSIS All categoric variables were compared using the chi-square or Fisher’s exact test, with a P value less than 0.05 considered significant. The chi-square test for trends was used to determine whether there was a significant change in the proportion of patients with a blood transfusion and morbidity, and the effect of risk factors such as age, decade of surgery, resection time, and the amount of tissue resected was assessed by the multiple logistic regression test. All analyses were performed with a statistics software package by Stat View (Abacus Concepts, Berkeley, Calif) and STATA (College Station, Tex).
RESULTS The average patient age was 70.1 6 7.9 years at the time of surgery, and the average resection time and amount of tissue resected were 66.6 6 33.3 minutes and 24.6 6 20.7 g, respectively (Table I). The mean postoperative hospital stay and duration of catheter drainage were 14.1 6 6.8 and 4.0 6 1.9 days, respectively. Only 15 patients (0.4%) were less than 50 years old at the time of surgery, whereas 330 (8.5%) were 50 to 59 years old, 1385 (35.9%) were 60 to 69 years old, 1702 (44.1%) were 70 to 79 years old, 410 (10.6%) were 80 to 89 years old, and 19 (0.5%) were more than 90 years old. A 24F, 26F, or 28F resectoscope was used in 229 (5.9%), 2227 (57.7%), and 1405 (36.4%) patients, respectively. The 22F two-way Foley balloon catheter was used in 330 patients (8.5%), whereas a 22F three-way Foley balloon catheter combined with continuous irrigation for 3 days postoperatively, which was 99
FIGURE 1. Blood transfusion and morbidity data from 1971 to 1996. The incidence of blood transfusion and morbidity gradually decreased over time and was statistically significantly different (P ,0.001, chi-square test for trend).
introduced at our hospital in 1986, was used in the remaining 3531 patients (91.5%). Nonclosed urinary drainage was used in 1484 patients (38.4%) who underwent TURP from 1971 to 1984, whereas closed urinary drainage, which has been used since 1985, was initiated in 2377 (61.6%). Surgery was performed during spinal anesthesia in 1345 patients (34.8%), epidural anesthesia in 2286 (59.2%), and general anesthesia in 230 (6.0%). MORTALITY There were only five deaths in our study population, resulting in a mortality rate of 0.1%. A postoperative death was defined as that which occurred during the postoperative hospital period or within 30 days of surgery, regardless of the cause. Three patients died of myocardial infarction, and 1 died of cerebral infarction. In none of these patients were the deaths related to the surgical procedure. The other patient died of disseminated intravenous coagulation and sepsis after 37 days of uncontrollable postoperative bleeding. There was no statistically significant difference in mortality between patients in the early or late groups (Table I). MORBIDITY A total of 516 complications (13.4%) occurred in 465 patients (12.0%), and blood transfusions were necessary in 507 (13.1%). There was a statistically significant difference for both blood transfusion and morbidity and the year that the procedure was performed (P ,0.001, chi-square test for trends) (Fig. 1). This was evidenced by decreases in morbidity and blood transfusion rates in patients in the late group (9.5% and 6.1%, respectively) compared with these rates in the early group (17.2% and 20.2%, respectively) (Table I). Extravasation was the most common intraoperative complication and occurred in 136 (3.5%) patients in whom drainage was immediately instituted. A hemostatic procedure had to be repeated transurethrally in 94 patients (2.4%) and in 1 pa100
tient who required open suprapubic packing after unsuccessful transurethral hemostatic treatment. Acute epididymitis, urethral strictures (treated by a dilating metal bougie or an internal urethrotomy), and a secondary resection arising from postoperative contracture of the bladder neck were observed in 60 (1.6%), 59 (1.5%), and 28 (0.7%) patients, respectively. External meatal stricture, incontinence, and transurethral resection syndrome (less than 125 mEq/L) were noted in 20 (0.5%), 20 (0.5%), and 15 (0.4%) patients, respectively. The occurrence of extravasation (P ,0.0001) and epididymitis (P ,0.0001) or the need for a hemostatic procedure (P 5 0.0004) were significantly less in patients in the late group (Table II). RELATION BETWEEN AMOUNT OF TISSUE RESECTED AND TIME OF SURGICAL RESECTION The amount of tissue resected and resection time were divided into 30-g and 30-minute increments, respectively. Of all patients studied, 81.5% required less than 90 minutes for resection of the prostatic tissue, 73% had less than 30 g of tissue resected, and less than 7% had greater than 61 g resected (Table III). The incidence of blood transfusion and morbidity were in proportion to the resection time and gland size. Among patients with a resection time of greater than 90 minutes, the incidence of bleeding showed the most striking difference compared with patients with a resection time of less than 90 minutes (P ,0.0001). Also, patients with a gland larger than 60 g had a significantly higher incidence of bleeding (56.3%) and morbidity (21.1%) compared with patients with a gland smaller than 60 g (P ,0.0001). RISK FACTORS FOR A BLOOD TRANSFUSION AND MORBIDITY The relation of the necessity of a blood transfusion and morbidity according to the various surgical variables in all patients is shown in Table IV. UROLOGY 53 (1), 1999
TABLE II. Complication
Complications from TURP procedures
Early Group (1971–1985)
Extravasation Hemostatic procedure Epididymitis Urethral stricture Contracture of bladder neck External meatal stricture Incontinence TUR syndrome Cardiovascular disease Liver dysfunction Other
105 64 49 23 19 7 10 10 6 5 34
(5.4) (3.3) (2.5) (1.2) (1.0) (0.4) (0.5) (0.5) (0.3) (0.3) (1.8)
Total
332 (17.2)
Late Group (1985–1996)
P Value
Total
(1.6) (1.6) (0.6) (1.9) (0.5) (0.8) (0.5) (0.3) (0.3) (0.2) (1.6)
,0.0001 0.0004 ,0.0001 0.0888 0.0577 0.1790 0.9991 0.1955 0.7620 0.7380
136 94 60 59 28 20 20 15 11 9 64
184 (9.5)
,0.0001
516 (13.4)
31 30 11 36 9 13 10 5 5 4 30
(3.5) (2.4) (1.6) (1.5) (0.7) (0.5) (0.5) (0.4) (0.3) (0.2) (1.7)
KEY: TUR 5 transurethral resection; TURP 5 transurethral resection of the prostate. Data presented are number (%) of patients.
TABLE III. Resection time and amount of prostatic tissue resected versus blood transfusions, morbidity, and mortality in patients undergoing TURP Total Patients [No. (%)] Resection time (min) ,30 31–60 61–90 91–120 .120 Amount of tissue resected (g) ,30 31–60 61–90 .90
Blood Transfusions [No. (%)]
Morbidity [No. (%)]
Mortality [No. (%)]
577 1371 1200 505 208
(14.9) (35.5) (31.1) (13.1) (5.4)
13 86 164 152 98
(2.3) (5.8) (13.7) (30.1) (47.1)
47 141 160 74 43
(8.1) (10.3) (13.3) (14.7) (20.7)
0 2 (0.1) 2 (0.2) 0 1 (0.5)
2815 799 196 51
(72.9) (20.7) (5.1) (1.3)
157 211 100 39
(5.6) (26.4) (51.0) (76.5)
297 116 39 13
(10.6) (14.5) (19.9) (25.5)
2 (0.1) 2 (0.3) 1 (0.5) 0
KEY: Abbreviations as in Table I.
Resection time, the amount of tissue resected, the year that TURP was performed, and patient age strongly correlated with the need for a blood transfusion (P ,0.0005), whereas a continuous irrigation system using a three-way Foley balloon catheter reduced the risk for a blood transfusion (P 5 0.045). The most striking difference in the risk for significant morbidity was related to the time of resection and the year that the surgery was performed (P ,0.0005). A preoperative vasectomy (P 5 0.002) and a closed urinary drainage system (P 5 0.014) also significantly reduced the risk of morbidity. We analyzed the risk factors between the various study variables and the three primary complications (extravasation, the need for a hemostatic procedure, and epididymitis) (Table V). The total resection time and the year that the surgery was performed significantly correlated to extravasation (P ,0.0005), whereas the year that the surgery UROLOGY 53 (1), 1999
was performed (P 5 0.001), the amount of tissue resected (P 5 0.003), and the time of resection (P 5 0.046) correlated with the need for a hemostatic procedure. Also, postoperative epididymitis strongly correlated with a preoperative vasectomy (P ,0.0005), the use of a closed drainage system (P ,0.0005), the size of the resectoscope (P 5 0.008), a three-way balloon catheter system (P 5 0.038), the amount of tissue resected (P 5 0.04), and resection time (P 5 0.042). COMMENT Three reports have detailed proper assessment of TURP.1–3 The mortality rate of the TURP procedure in the United States markedly decreased from 2.5% in the 1960s to 0.23% by the 1980s.10 –12 In the present study, only 5 deaths (0.1%) were noted in the total study population, and there was no significant difference between the early and late 101
TABLE IV. Blood transfusion and morbidity versus TURP operative variables Blood Transfusion Variable
OR (95% CI)
Resection time Amount of tissue resected Decade (1970, 1980, 1990) Age Balloon catheter* Drainage system† Size of resectoscope‡ Vasectomy
1.931 1.382 0.299 1.548 0.694 0.723 1.008
(1.701–2.913) (1.298–1.471) (0.228–0.391) (1.329–1.802) (0.485–0.991) (0.514–1.015) (0.900–1.129)
Morbidity
Z Value
P Value
10.156 10.103 28.796 5.623 22.008 21.870 0.142
,0.0005 ,0.0005 ,0.0005 ,0.0005 0.045 0.061 0.887
OR (95% CI) 1.268 1.044 0.671 0.999 0.795 0.689 0.962 0.651
Z Value
(1.317–1.421) (0.986–1.106) (0.536–0.839) (0.878–1.137) (0.567–1.113) (0.512–0.927) (0.866–1.068) (0.495–0.856)
P Value
4.092 ,0.0005 1.499 0.134 23.487 ,0.0005 20.003 0.998 21.333 0.182 22.456 0.014 20.719 0.472 23.064 0.002
KEY: OR 5 odds ratio; CI 5 confidence interval; other abbreviations as in Table I. * Two-way vs. three-way. † Open vs. closed. ‡ 24F, 26F, and 28F.
TABLE V.
Postsurgical complications versus TURP operative variables Extravasation
Variable Resection time Decade (1970, 1980, 1990) Age Amount of resected tissue Size of resectoscope* Vasectomy Balloon catheter† Drainage system‡ Extravasation Hemostatic procedure Urethral stricture Contracture of bladder neck
OR (95% CI)
Hemostatic Procedure
P Value
OR (95% CI)
P Value
Epididymitis OR (95% CI)
P Value
2.543 (1.989–3.250) ,0.0005 1.259 (1.004–1.578) 0.046 0.675 (0.462–0.986) 0.886 (0.591–1.326) ,0.0005 0.406 (0.240–0.685) 0.001 1.352 (0.693–2.640)
0.042 0.376
0.762 (0.599–0.969) 0.973 (0.866–1.092)
0.025 0.382
1.091 (0.827–1.438) 0.535 0.914 (0.644–1.295) 1.154 (1.050–1.269) 0.003 1.188 (1.007–1.401)
0.614 0.040
1.158 (0.953–1.406)
0.139
1.032 (0.835–1.277) 0.767 0.712 0.251 0.563 (0.311–1.018) 0.058 0.442 1.067 (0.544–2.094) 0.849 0.167 2.034 (0.962–4.297) 0.063 0.899 1.931 3.448 3.045
(0.553–0.916) 0.008 (0.139–0.452) ,0.0005 (0.205–0.954) 0.038 (0.066–0.424) ,0.0005 (0.188–4.303) 0.895 (0.613–6.077) 0.260 (0.911–13.040) 0.068 (0.538–17.230) 0.208
KEY: Abbreviations as in Tables I and IV. * 24F, 26F, and 28F. † Two-way vs. three-way. ‡ Open vs. closed.
groups. The mortality rate in this study was less than that reported for the United States (Fig. 2), where the predominant causes of death were myocardial infarction in the period from 1960 to 1970 and sepsis in the 1980s.1–3 Myocardial infarction may be predisposed by racial factors and differences in eating and smoking habits between American and Japanese men. The cause for the high rate of sepsis noted in the U.S. studies is unknown but may be based on the duration of prophylactic antibiotic use because we routinely prescribe prophylactic antibiotic agents postoperatively for 7 to 14 days. Furthermore, a different insurance system between Japan and other Western countries must be considered when mortality data are discussed. In Japan, many patients older than 65 years pay only 10% to 30% of their medical charges, and there is no charge for hospital admission at any time, so that the duration of the hospital period is 102
not a relevant consideration. Generally, all patients entered our hospital 2 days before surgery and received a preoperative check by the anesthesiologist. Surgery was postponed or canceled when the patient required further treatment. A combination of these factors may contribute to the difference in mortality rates between Japanese and American patients undergoing TURP. The frequency of blood transfusions is the most important factor in choosing a treatment regimen for patients with BPH. At present, a high rate of blood transfusions with TURP is one of the most disadvantageous factors in comparison to various minimally invasive treatments.11–18 In the present study, the incidence of blood transfusions in the late group (1985 to 1996) was only 6.1%, which was significantly less than that in the 1960s (36.6%) or 1970s (25.1%). This low rate in the late group was comparable to that reported in the UROLOGY 53 (1), 1999
FIGURE 2. Percentage of patients with surgical risk factors, morbidity, and mortality: published reports versus the current study. Solid bars 5 Holtgrewe and Valk1 (1955 to 1960); dotted bars 5 Melchior et al.2 (1965 to 1971); open bars 5 Mebust et al.3 (1978 to 1987); hatched bars (left) 5 present study (1971 to 1985); hatched bars (right) 5 present study (1985 to 1996).
1980s (6.4%) in the United States.1–3 Our analysis also confirmed the hypothesis that a large amount of tissue resected and increased resection times were associated with an increasing number of blood transfusions and associated morbidity. Additionally, the year that TURP was performed and age strongly correlated with the incidence of blood transfusions. Preoperative factors, including underlying conditions and a generally lower level of blood cell counts in older patients, may predispose older patients to receive blood transfusions. It is probable that improvements in techniques and instrumentation, including changes in the type of balloon catheter and the institution of continuous irrigation in 1987, and anesthesiologic management in recent patients may result in the observed decreases in perioperative morbidity and the need for blood transfusions over time. Although mortality was markedly reduced from the 1960s to the 1980s, morbidity remained at a similar level, ranging from 17.3% to 25.0% over this period.1–3 Previous investigators postulated that morbidity was strongly correlated to a large gland and increased surgical time. In our series, the morbidity rate in the late group (1985 to 1996) was reduced compared with that of the early group (1971 to 1985), from 17.2% to 9.5%, and strongly correlated with the time of surgical resection and the year that TURP was performed. When we analyzed multiple factors for each surgical complication, extravasation and associated hemostatic procedures occurred most frequently, but the incidence was significantly lower in the late group. Similarly, the amount of resected tissue, the total resection time, the year of surgery, and age were decreased for patients in the late group. New techniques involving teaching microscopes and a video system and advanced instrumentation may be effective in decreasing the incidence of extravasation and hemostatic procedures in recent years. The incidence of postoperative infections is a significant morbidity factor. Rates of epididymitis (0.2% to 6%), pneumonia/pulmonary infection UROLOGY 53 (1), 1999
(0.49% to 6.7%), and pyelonephritis (0.4% to 1.1%) have been reported for patients undergoing TURP.1–3 Additionally, from these studies, patients who had a preoperative vasectomy were at an increased risk (10%) for postsurgical epididymitis.1–3 The incidence of epididymitis in the present study was also reduced in the late group. All patients were given prophylactic antibiotic therapy for at least 14 days postoperatively to prevent postoperative infection, including epididymitis, and the introduction of a closed urinary drainage system in 1985 also significantly reduced the risk of morbidity. It is not clear whether these factors had a beneficial effect in TURP for preventing postoperative infection, but they may be important for the observed reduction in infection in the late group in the present study. Alternatives to TURP may be equally effective, with less morbidity, but they are not sufficient to totally replace standard TURP treatment.11–18 Although new types of surgical loops have been examined for the potential to reduce bleeding during TURP, conventional TURP remains fundamental to the transurethral treatment for BPH, and surgeons should become adept in this procedure.23 CONCLUSIONS The incidence of blood transfusions and morbidity in patients undergoing TURP has decreased compared with that during the 1970s. Advances in techniques, instrumentation, and surgical and perioperative management, including anesthesia, have made TURP a relatively safe procedure. TURP remains an effective treatment option for patients with BPH. REFERENCES 1. Holtgrewe HL, and Valk WL: Factors influencing the mortality and morbidity of transurethral prostatectomy: a study of 2,015 cases. J Urol 87: 450 – 459, 1962. 103
2. Melchior J, Valk WL, Foret JD, et al: Transurethral prostatectomy: computerized analysis of 2,223 consecutive cases. J Urol 112: 634 – 646, 1974. 3. Mebust WK, Holtgrewe HL, Cockett ATK, et al: Transurethral prostatectomy: immediate and postoperative complications: a cooperative study of 13 participating institutions evaluating 3,885 patients. J Urol 141: 243–247, 1989. 4. Uchida T, Adachi K, Ao T, et al: Clinical analysis in 2266 cases of transurethral resection of the prostate. Jpn J Urol 84: 890 – 896, 1993. 5. Lepor H, and Rigaud G: The efficacy of transurethral resection of the prostate in men with moderate symptoms of prostatism. J Urol 143: 533–537, 1990. 6. Wennberg JE, Roos N, Sola L, et al: Use of claims data systems to evaluate health care outcomes: mortality and reoperation following prostatectomy. JAMA 257: 933–936, 1987. 7. Roos NP, Wennberg JE, Malenka DJ, et al: Mortality and reoperation after open and transurethral resection of the prostate for benign prostatic hyperplasia. N Engl J Med 320: 1120 –1124, 1989. 8. Malenka DJ, Roos N, Fisher ES, et al: Further study of the increased mortality following transurethral prostatectomy: a chart-based analysis. J Urol 144: 224 –227, 1990. 9. Holtgrewe HL, Mebust WK, Dowd JB, et al: Transurethral prostatectomy: practice aspects of the dominant operation in American urology. J Urol 141: 248 –253, 1989. 10. Fowler FJ Jr, Wennberg JE, Timothy RP, et al: Symptoms status and quality of life following prostatectomy. JAMA 259: 3018 –3022, 1988. 11. Holtgrewe HL: Benign prostatic hyperplasia (editorial). J Urol 152: 2030, 1994. 12. Kaplan SA, Wasson JH, Kirby RS, et al: BPH: where are we in 1996? AUA News 1: 25–27, 1996. 13. Holtgrewe HL: Guidance for clinical investigations of devices used for the treatment of benign prostatic hyperplasia. J Urol 150: 1588 –1590, 1993. 14. Kaplan SA, and Olsson CA: State of the art: microwave therapy in the management of men with benign prostatic hyperplasia: current status. J Urol 150: 1597–1602, 1993. 15. Schatzl G, Madersbacher S, Lang T, et al: The early postoperative morbidity of transurethral resection of the prostate and of 4 minimally invasive treatment alternatives. J Urol 158: 105–111, 1997. 16. Uchida T, Yokoyama E, Iwamura M, et al: High intensity focused ultrasound for benign prostatic hyperplasia. Int J Urol 2: 181–185, 1995. 17. Uchida T, Egawa S, Iwamura M, et al: A non-randomized comparative study of visual laser ablation and transurethral resection of the prostate in benign prostatic hyperplasia. Int J Urol 3: 108 –112, 1996. 18. Oesterling JE: Benign prostatic hyperplasia: medical and minimally invasive treatment options. N Engl J Med 332: 99 –109, 1995. 19. Gee WF, Holtgrewe HL, Albertsen PC, et al: Practice trends in the diagnosis and management of benign prostatic hyperplasia in the United States. J Urol 154: 205–206, 1995. 20. Egawa S, Uchida T, and Koshiba K: Current and future trends in the interventional therapy for benign prostatic hyperplasia. J Urol 159: 1958 –1960, 1998. 21. Holtgrewe HL: Current trends in management of men with lower urinary tract symptoms and benign prostatic hyperplasia. Urology 51(suppl): 1–7, 1998. 22. Cockett ATK, and Koshiba K: Manual of Urologic Surgery. New York, Springer-Verlag, 1979, p 237. 23. Uchida T, Ohkawa A, Shibata Y, et al: New transurethral resection of the prostate using WEDGE® loop for benign prostatic hyperplasia. Jpn J Clin Urol 51: 1013–1017, 1997. 104
EDITORIAL COMMENT The present article provides insight into the adverse outcomes of 3861 transurethral resection of the prostate (TURP) procedures conducted at a single institution over a 25-year period, in itself an interesting analysis. The authors divided the cohort into the first 1930 and the second 1931 patients, a numerical split that classified the patient population into those operated on from 1971 to 1985 and those operated on from 1985 to 1996. This in itself is rather curious, because in most institutions the number of TURPs performed has dramatically decreased in the 1990s. In the present institution, however, the first 1930 cases were done over 14 years, whereas the second half was performed over only 11 years, resulting in an apparent increase in the number of procedures done per year. Although there may be many reasons for this increase, which should not be interpreted as an increase in the incidence of TURP per unit of population at risk, it is an interesting fact to note. Decreases in the morbidity rate (17.2% versus 9.5%) of the procedure and the need for blood transfusions (20.2% versus 6.1%) over time were noted when the two groups were compared. The decrease in morbidity was mostly due to a decrease in the rate of extravasation, need for hemostatic procedures, and postoperative epididymitis; other complications did not decrease at the same rate. Of course, one must realize that there was no dramatic decrease in any of these complications from before to after 1986; rather, a steady decline is the underlying cause, as becomes clear from Figure 1. Of note, during the 25 years, no apparent change in patient age or amount of resected tissue (25.9 versus 23.3 g) occurred. The analysis of individually reported series of TURP conducted for the Agency for Health Care Policy and Research (AHCPR) guidelines,1 spanning the better part of the 20th century, resulted in an overall transfusion rate of 12.5%. However, when sorted by date of publication, there was a similarly clear trend toward a reduction in transfusion rates noted over time. Table VI compares the 90% confidence interval (CI) from the AHCPR guidelines and the two groups of patients analyzed in the present report. It is clear that for the most part, the observed complication rates fall well into the 90% CI reported in the guidelines, which cover nearly the entire time period of observation reported in the present study. Several additional points are noteworthy. For one, the mortality rate is exceedingly low, even in the early group, a fact that the authors explain by the comorbidity status of Japanese versus American men and the routine use of prolonged antibiotic therapy. This, however, does not appear immediately intuitive because many studies have shown that even a single oral dose of antibiotic agents surrounding the time of surgery is effective in preventing infections or possibly lethal sepsis episodes. The administration of a 10% saline solution intravenously in all patients having electrolyte imbalance is in most parts of the world a very rarely used measure, if not regarded as outright dangerous. One can only speculate on the impact that this intervention in an unknown number of patients had in either a positive or negative sense. The one analysis that the AHCPR meta-analysis did not permit by focusing on studies rather than individual patients is perhaps the most significant message worth repeating. The larger the gland, the longer the resection time, and the longer the resection time, the higher the morbidity and mortality rates and the greater the need for transfusion. Most graduating residents finishing their training with fewer than 50 TURPs should take note and undertake a TURP only if the amount of resectable tissue is less than 60 g. This statement is supported by the AHCPR guidelines1 as well as the World Health Organization benign prostatic hyperplasia consensus conference.2 UROLOGY 53 (1), 1999
FACTORS INFLUENCING MORBIDITY IN PATIENTS UNDERGOING TRANSURETHRAL RESECTION OF THE PROSTATE TOYOAKI UCHIDA, MAKOTO OHORI, SHIGEHIRO SOH, TAKEFUMI SATO, MASATSUGU IWAMURA, TERUAKI AO, AND KEN KOSHIBA
ABSTRACT Objectives. Transurethral resection of the prostate (TURP) has become the primary method to relieve bladder outlet obstruction for patients with benign prostatic hyperplasia (BPH). Data from 3861 consecutive patients with BPH who underwent TURP from 1971 to 1996 at our hospital were retrospectively analyzed. Methods. The patients were classified into two groups comprising 1930 patients who underwent TURP from 1971 to 1985 (early group) and 1931 patients who underwent TURP from 1985 to 1996 (late group). Risk factors associated with blood transfusions and perioperative complications were analyzed in these patients. Results. Mortality, morbidity, and blood transfusions were noted in 5 (0.1%), 516 (13.4%), and 507 (13.1%) patients, respectively. The blood transfusion and morbidity rates decreased over the 25-year period (P ,0.001, chi-square test for trends), which was reflected in a decrease in these rates in the late group (6.1% and 9.5%, respectively) compared with those of the early group (20.2% and 17.2%, respectively). Postoperative bleeding and morbidity were closely related to prostatic gland size and operating time. The most significant differences for the risk of a blood transfusion were related to resection time, the amount of tissue resected, age, and the decade (1970s, 1980s, or 1990s) in which the surgery was performed (P ,0.0005), whereas resection time was significantly correlated with morbidity (P ,0.0005). As risk factors for each complication, the time of surgical resection, the decade of surgery, and the amount of tissue resected directly correlated with the incidence of extravasation and hemostatic procedures (P #0.003), whereas the incidence of postoperative epididymitis positively correlated with a preoperative vasectomy and a closed drainage system (P ,0.0005). Conclusions. Since the 1970s, the rates of blood transfusions and morbidity have decreased for patients undergoing TURP. Advances in techniques, instrumentation, and surgical and perioperative management, including anesthesia, have made TURP a relatively safe procedure, and it remains an effective means for treating patients with BPH. UROLOGY 53: 98–105, 1999. © 1999, Elsevier Science Inc. All rights reserved.
F
or the past 50 years, transurethral resection of the prostate (TURP) has been the primary method for relieving bladder outlet obstruction in patients with benign prostatic hyperplasia (BPH).1– 4 Although TURP is an effective treatment for most men with symptomatic BPH, 20% to 25% of patients who undergo the operation do not have a satisfactory long-term outcome.5 Another concern is that the rate of reoperation is 15% to 20% in From the Department of Urology, Kitasato University School of Medicine, Kanagawa, Japan Reprint requests: Toyoaki Uchida, M.D., Department of Urology, Kitasato University School of Medicine, 1-15-1 Kitasato, Sagamihara, Kanagawa 228, Japan Submitted: March 31, 1998, accepted (with revisions): July 6, 1998
98
© 1999, ELSEVIER SCIENCE INC. ALL RIGHTS RESERVED
men followed up for 10 years or longer.6 Also, in several preliminary retrospective studies, the life expectancy of men undergoing TURP was shorter than that of men undergoing an open prostatectomy as treatment for BPH.7,8 Reported complications include retrograde ejaculation in 70% to 75% of men, impotence in 5% to 10%, postoperative urinary tract infection in 5% to 10%, and some degree of urinary incontinence in 2% to 4% of patients. Furthermore, 5% to 10% of patients require a blood transfusion.1–3,9,10 Several minimally invasive therapies, such as transurethral incision, balloon dilation, hyperthermia, laser surgery, highintensity focused ultrasound, and drug therapy with 5-alpha-reductase inhibitors or alpha-blockers, have been used in the past 5 years as alterna0090-4295/99/$19.00 PII S0090-4295(98)00524-X
TABLE I.
Age (yr) Resection time (min) Amount of tissue resected (g) Mortality Morbidity Blood transfusion
Characteristics of 3861 patients undergoing TURP Early Group (1971–1985) (n 5 1930)
Late Group (1985–1996) (n 5 1931)
P Value
Total (n 5 3861)
70.2 6 7.7 65.0 6 32.8 25.9 6 21.7 4 (0.2%) 332 (17.2%) 389 (20.2%)
70.1 6 8.0 68.3 6 33.8 23.3 6 19.5 1 (0.5%) 184 (9.5%) 118 (6.1%)
0.6927 0.0021 ,0.0001 0.1792 ,0.0001 ,0.0001
70.1 6 7.9 66.6 6 33.3 24.6 6 20.7 5 (0.1%) 516 (13.4%) 507 (13.1%)
KEY: TURP 5 transurethral resection of the prostate. Data presented are mean value 6 SD or number (%) of patients.
tives to TURP.11–18 A recent survey indicated that the first choice of treatment for patients with BPH based on American Urological Association (AUA) symptom assessment was observation for an AUA score of 0 to 7 (77%), chemotherapy with alphablockers for an AUA score of 8 to 19 (65%), and TURP for an AUA score of 20 to 35 (41%).19 TURP is preferred most by American and Japanese urologists in patients with severe symptoms and a larger prostate gland.19,20 Since 1990, the frequency of TURP as a surgical treatment for patients with BPH has decreased compared with the period from 1960 to 1990. The availability of less invasive methods as well as economic concerns may account for this decrease.21 In the present study, 3861 patients treated with TURP from 1971 to 1996 were retrospectively examined and compared with regard to history, operative management, and complications. MATERIAL AND METHODS PATIENT CHARACTERISTICS Between August 1971 and December 1996, 3861 patients underwent TURP for BPH at our university hospital. Fortyfive variables were evaluated for each patient and included specific information on demographics, operative management, intraoperative complications, and immediate and longterm postoperative complications. Also, the patients were classified into two groups, with 1930 patients who underwent TURP from 1971 to 1985 in the early group and 1931 patients who underwent TURP from 1985 to 1996 in the late group (Table I).
SURGICAL TECHNIQUE Lumbar, epidural, or general anesthesia was used in all patients unless medically contraindicated, and all patients were placed in the dorsal lithotomy position. All resections were conducted using an Iglesias-type resectoscope (CIRCON ACMI, Santa Barbara, Calif), and nonhemolytic solutions were used as the irrigating solution. The prostatectomies were performed using the technique of Cockett and Koshiba,22 where the prostate adenoma was completely excised, thus exposing fibers of the surgical capsule throughout the prostatic fossa. The procedure was initiated and conducted for the first 15 to 60 minutes by senior surgical residents who were not familiar with the TURP procedure, under the supervision of faculty surgeons, who were then succeeded by an experienced staff surgeon who completed the operation. Routine serum electrolyte and blood cell count determinations were performed both UROLOGY 53 (1), 1999
during the procedure and postoperatively. Depending on the electrolyte imbalance, a 10% saline solution was administered intravenously to prevent hyponatremia. For patients undergoing surgery from 1971 to 1986, the procedure was terminated by insertion of a 22F two-way Foley balloon catheter into the bladder, which remained in the bladder for 3 days postoperatively. Since 1986, a 22F three-way Foley balloon catheter has been used to provide continuous irrigation with a saline solution for 3 days postoperatively, depending on the postoperative course. Gentle traction was applied at the bladder neck in most patients for 30 to 60 minutes postoperatively. From 1971 to 1984, nonclosed urinary drainage was used, which was changed to closed urinary drainage in 1985. Antibiotic agents were routinely administrated intravenously for 3 days postoperatively and then orally for 1 to 2 weeks to prevent postoperative infection.
STATISTICAL ANALYSIS All categoric variables were compared using the chi-square or Fisher’s exact test, with a P value less than 0.05 considered significant. The chi-square test for trends was used to determine whether there was a significant change in the proportion of patients with a blood transfusion and morbidity, and the effect of risk factors such as age, decade of surgery, resection time, and the amount of tissue resected was assessed by the multiple logistic regression test. All analyses were performed with a statistics software package by Stat View (Abacus Concepts, Berkeley, Calif) and STATA (College Station, Tex).
RESULTS The average patient age was 70.1 6 7.9 years at the time of surgery, and the average resection time and amount of tissue resected were 66.6 6 33.3 minutes and 24.6 6 20.7 g, respectively (Table I). The mean postoperative hospital stay and duration of catheter drainage were 14.1 6 6.8 and 4.0 6 1.9 days, respectively. Only 15 patients (0.4%) were less than 50 years old at the time of surgery, whereas 330 (8.5%) were 50 to 59 years old, 1385 (35.9%) were 60 to 69 years old, 1702 (44.1%) were 70 to 79 years old, 410 (10.6%) were 80 to 89 years old, and 19 (0.5%) were more than 90 years old. A 24F, 26F, or 28F resectoscope was used in 229 (5.9%), 2227 (57.7%), and 1405 (36.4%) patients, respectively. The 22F two-way Foley balloon catheter was used in 330 patients (8.5%), whereas a 22F three-way Foley balloon catheter combined with continuous irrigation for 3 days postoperatively, which was 99
FIGURE 1. Blood transfusion and morbidity data from 1971 to 1996. The incidence of blood transfusion and morbidity gradually decreased over time and was statistically significantly different (P ,0.001, chi-square test for trend).
introduced at our hospital in 1986, was used in the remaining 3531 patients (91.5%). Nonclosed urinary drainage was used in 1484 patients (38.4%) who underwent TURP from 1971 to 1984, whereas closed urinary drainage, which has been used since 1985, was initiated in 2377 (61.6%). Surgery was performed during spinal anesthesia in 1345 patients (34.8%), epidural anesthesia in 2286 (59.2%), and general anesthesia in 230 (6.0%). MORTALITY There were only five deaths in our study population, resulting in a mortality rate of 0.1%. A postoperative death was defined as that which occurred during the postoperative hospital period or within 30 days of surgery, regardless of the cause. Three patients died of myocardial infarction, and 1 died of cerebral infarction. In none of these patients were the deaths related to the surgical procedure. The other patient died of disseminated intravenous coagulation and sepsis after 37 days of uncontrollable postoperative bleeding. There was no statistically significant difference in mortality between patients in the early or late groups (Table I). MORBIDITY A total of 516 complications (13.4%) occurred in 465 patients (12.0%), and blood transfusions were necessary in 507 (13.1%). There was a statistically significant difference for both blood transfusion and morbidity and the year that the procedure was performed (P ,0.001, chi-square test for trends) (Fig. 1). This was evidenced by decreases in morbidity and blood transfusion rates in patients in the late group (9.5% and 6.1%, respectively) compared with these rates in the early group (17.2% and 20.2%, respectively) (Table I). Extravasation was the most common intraoperative complication and occurred in 136 (3.5%) patients in whom drainage was immediately instituted. A hemostatic procedure had to be repeated transurethrally in 94 patients (2.4%) and in 1 pa100
tient who required open suprapubic packing after unsuccessful transurethral hemostatic treatment. Acute epididymitis, urethral strictures (treated by a dilating metal bougie or an internal urethrotomy), and a secondary resection arising from postoperative contracture of the bladder neck were observed in 60 (1.6%), 59 (1.5%), and 28 (0.7%) patients, respectively. External meatal stricture, incontinence, and transurethral resection syndrome (less than 125 mEq/L) were noted in 20 (0.5%), 20 (0.5%), and 15 (0.4%) patients, respectively. The occurrence of extravasation (P ,0.0001) and epididymitis (P ,0.0001) or the need for a hemostatic procedure (P 5 0.0004) were significantly less in patients in the late group (Table II). RELATION BETWEEN AMOUNT OF TISSUE RESECTED AND TIME OF SURGICAL RESECTION The amount of tissue resected and resection time were divided into 30-g and 30-minute increments, respectively. Of all patients studied, 81.5% required less than 90 minutes for resection of the prostatic tissue, 73% had less than 30 g of tissue resected, and less than 7% had greater than 61 g resected (Table III). The incidence of blood transfusion and morbidity were in proportion to the resection time and gland size. Among patients with a resection time of greater than 90 minutes, the incidence of bleeding showed the most striking difference compared with patients with a resection time of less than 90 minutes (P ,0.0001). Also, patients with a gland larger than 60 g had a significantly higher incidence of bleeding (56.3%) and morbidity (21.1%) compared with patients with a gland smaller than 60 g (P ,0.0001). RISK FACTORS FOR A BLOOD TRANSFUSION AND MORBIDITY The relation of the necessity of a blood transfusion and morbidity according to the various surgical variables in all patients is shown in Table IV. UROLOGY 53 (1), 1999
TABLE II. Complication
Complications from TURP procedures
Early Group (1971–1985)
Extravasation Hemostatic procedure Epididymitis Urethral stricture Contracture of bladder neck External meatal stricture Incontinence TUR syndrome Cardiovascular disease Liver dysfunction Other
105 64 49 23 19 7 10 10 6 5 34
(5.4) (3.3) (2.5) (1.2) (1.0) (0.4) (0.5) (0.5) (0.3) (0.3) (1.8)
Total
332 (17.2)
Late Group (1985–1996)
P Value
Total
(1.6) (1.6) (0.6) (1.9) (0.5) (0.8) (0.5) (0.3) (0.3) (0.2) (1.6)
,0.0001 0.0004 ,0.0001 0.0888 0.0577 0.1790 0.9991 0.1955 0.7620 0.7380
136 94 60 59 28 20 20 15 11 9 64
184 (9.5)
,0.0001
516 (13.4)
31 30 11 36 9 13 10 5 5 4 30
(3.5) (2.4) (1.6) (1.5) (0.7) (0.5) (0.5) (0.4) (0.3) (0.2) (1.7)
KEY: TUR 5 transurethral resection; TURP 5 transurethral resection of the prostate. Data presented are number (%) of patients.
TABLE III. Resection time and amount of prostatic tissue resected versus blood transfusions, morbidity, and mortality in patients undergoing TURP Total Patients [No. (%)] Resection time (min) ,30 31–60 61–90 91–120 .120 Amount of tissue resected (g) ,30 31–60 61–90 .90
Blood Transfusions [No. (%)]
Morbidity [No. (%)]
Mortality [No. (%)]
577 1371 1200 505 208
(14.9) (35.5) (31.1) (13.1) (5.4)
13 86 164 152 98
(2.3) (5.8) (13.7) (30.1) (47.1)
47 141 160 74 43
(8.1) (10.3) (13.3) (14.7) (20.7)
0 2 (0.1) 2 (0.2) 0 1 (0.5)
2815 799 196 51
(72.9) (20.7) (5.1) (1.3)
157 211 100 39
(5.6) (26.4) (51.0) (76.5)
297 116 39 13
(10.6) (14.5) (19.9) (25.5)
2 (0.1) 2 (0.3) 1 (0.5) 0
KEY: Abbreviations as in Table I.
Resection time, the amount of tissue resected, the year that TURP was performed, and patient age strongly correlated with the need for a blood transfusion (P ,0.0005), whereas a continuous irrigation system using a three-way Foley balloon catheter reduced the risk for a blood transfusion (P 5 0.045). The most striking difference in the risk for significant morbidity was related to the time of resection and the year that the surgery was performed (P ,0.0005). A preoperative vasectomy (P 5 0.002) and a closed urinary drainage system (P 5 0.014) also significantly reduced the risk of morbidity. We analyzed the risk factors between the various study variables and the three primary complications (extravasation, the need for a hemostatic procedure, and epididymitis) (Table V). The total resection time and the year that the surgery was performed significantly correlated to extravasation (P ,0.0005), whereas the year that the surgery UROLOGY 53 (1), 1999
was performed (P 5 0.001), the amount of tissue resected (P 5 0.003), and the time of resection (P 5 0.046) correlated with the need for a hemostatic procedure. Also, postoperative epididymitis strongly correlated with a preoperative vasectomy (P ,0.0005), the use of a closed drainage system (P ,0.0005), the size of the resectoscope (P 5 0.008), a three-way balloon catheter system (P 5 0.038), the amount of tissue resected (P 5 0.04), and resection time (P 5 0.042). COMMENT Three reports have detailed proper assessment of TURP.1–3 The mortality rate of the TURP procedure in the United States markedly decreased from 2.5% in the 1960s to 0.23% by the 1980s.10 –12 In the present study, only 5 deaths (0.1%) were noted in the total study population, and there was no significant difference between the early and late 101
TABLE IV. Blood transfusion and morbidity versus TURP operative variables Blood Transfusion Variable
OR (95% CI)
Resection time Amount of tissue resected Decade (1970, 1980, 1990) Age Balloon catheter* Drainage system† Size of resectoscope‡ Vasectomy
1.931 1.382 0.299 1.548 0.694 0.723 1.008
(1.701–2.913) (1.298–1.471) (0.228–0.391) (1.329–1.802) (0.485–0.991) (0.514–1.015) (0.900–1.129)
Morbidity
Z Value
P Value
10.156 10.103 28.796 5.623 22.008 21.870 0.142
,0.0005 ,0.0005 ,0.0005 ,0.0005 0.045 0.061 0.887
OR (95% CI) 1.268 1.044 0.671 0.999 0.795 0.689 0.962 0.651
Z Value
(1.317–1.421) (0.986–1.106) (0.536–0.839) (0.878–1.137) (0.567–1.113) (0.512–0.927) (0.866–1.068) (0.495–0.856)
P Value
4.092 ,0.0005 1.499 0.134 23.487 ,0.0005 20.003 0.998 21.333 0.182 22.456 0.014 20.719 0.472 23.064 0.002
KEY: OR 5 odds ratio; CI 5 confidence interval; other abbreviations as in Table I. * Two-way vs. three-way. † Open vs. closed. ‡ 24F, 26F, and 28F.
TABLE V.
Postsurgical complications versus TURP operative variables Extravasation
Variable Resection time Decade (1970, 1980, 1990) Age Amount of resected tissue Size of resectoscope* Vasectomy Balloon catheter† Drainage system‡ Extravasation Hemostatic procedure Urethral stricture Contracture of bladder neck
OR (95% CI)
Hemostatic Procedure
P Value
OR (95% CI)
P Value
Epididymitis OR (95% CI)
P Value
2.543 (1.989–3.250) ,0.0005 1.259 (1.004–1.578) 0.046 0.675 (0.462–0.986) 0.886 (0.591–1.326) ,0.0005 0.406 (0.240–0.685) 0.001 1.352 (0.693–2.640)
0.042 0.376
0.762 (0.599–0.969) 0.973 (0.866–1.092)
0.025 0.382
1.091 (0.827–1.438) 0.535 0.914 (0.644–1.295) 1.154 (1.050–1.269) 0.003 1.188 (1.007–1.401)
0.614 0.040
1.158 (0.953–1.406)
0.139
1.032 (0.835–1.277) 0.767 0.712 0.251 0.563 (0.311–1.018) 0.058 0.442 1.067 (0.544–2.094) 0.849 0.167 2.034 (0.962–4.297) 0.063 0.899 1.931 3.448 3.045
(0.553–0.916) 0.008 (0.139–0.452) ,0.0005 (0.205–0.954) 0.038 (0.066–0.424) ,0.0005 (0.188–4.303) 0.895 (0.613–6.077) 0.260 (0.911–13.040) 0.068 (0.538–17.230) 0.208
KEY: Abbreviations as in Tables I and IV. * 24F, 26F, and 28F. † Two-way vs. three-way. ‡ Open vs. closed.
groups. The mortality rate in this study was less than that reported for the United States (Fig. 2), where the predominant causes of death were myocardial infarction in the period from 1960 to 1970 and sepsis in the 1980s.1–3 Myocardial infarction may be predisposed by racial factors and differences in eating and smoking habits between American and Japanese men. The cause for the high rate of sepsis noted in the U.S. studies is unknown but may be based on the duration of prophylactic antibiotic use because we routinely prescribe prophylactic antibiotic agents postoperatively for 7 to 14 days. Furthermore, a different insurance system between Japan and other Western countries must be considered when mortality data are discussed. In Japan, many patients older than 65 years pay only 10% to 30% of their medical charges, and there is no charge for hospital admission at any time, so that the duration of the hospital period is 102
not a relevant consideration. Generally, all patients entered our hospital 2 days before surgery and received a preoperative check by the anesthesiologist. Surgery was postponed or canceled when the patient required further treatment. A combination of these factors may contribute to the difference in mortality rates between Japanese and American patients undergoing TURP. The frequency of blood transfusions is the most important factor in choosing a treatment regimen for patients with BPH. At present, a high rate of blood transfusions with TURP is one of the most disadvantageous factors in comparison to various minimally invasive treatments.11–18 In the present study, the incidence of blood transfusions in the late group (1985 to 1996) was only 6.1%, which was significantly less than that in the 1960s (36.6%) or 1970s (25.1%). This low rate in the late group was comparable to that reported in the UROLOGY 53 (1), 1999
FIGURE 2. Percentage of patients with surgical risk factors, morbidity, and mortality: published reports versus the current study. Solid bars 5 Holtgrewe and Valk1 (1955 to 1960); dotted bars 5 Melchior et al.2 (1965 to 1971); open bars 5 Mebust et al.3 (1978 to 1987); hatched bars (left) 5 present study (1971 to 1985); hatched bars (right) 5 present study (1985 to 1996).
1980s (6.4%) in the United States.1–3 Our analysis also confirmed the hypothesis that a large amount of tissue resected and increased resection times were associated with an increasing number of blood transfusions and associated morbidity. Additionally, the year that TURP was performed and age strongly correlated with the incidence of blood transfusions. Preoperative factors, including underlying conditions and a generally lower level of blood cell counts in older patients, may predispose older patients to receive blood transfusions. It is probable that improvements in techniques and instrumentation, including changes in the type of balloon catheter and the institution of continuous irrigation in 1987, and anesthesiologic management in recent patients may result in the observed decreases in perioperative morbidity and the need for blood transfusions over time. Although mortality was markedly reduced from the 1960s to the 1980s, morbidity remained at a similar level, ranging from 17.3% to 25.0% over this period.1–3 Previous investigators postulated that morbidity was strongly correlated to a large gland and increased surgical time. In our series, the morbidity rate in the late group (1985 to 1996) was reduced compared with that of the early group (1971 to 1985), from 17.2% to 9.5%, and strongly correlated with the time of surgical resection and the year that TURP was performed. When we analyzed multiple factors for each surgical complication, extravasation and associated hemostatic procedures occurred most frequently, but the incidence was significantly lower in the late group. Similarly, the amount of resected tissue, the total resection time, the year of surgery, and age were decreased for patients in the late group. New techniques involving teaching microscopes and a video system and advanced instrumentation may be effective in decreasing the incidence of extravasation and hemostatic procedures in recent years. The incidence of postoperative infections is a significant morbidity factor. Rates of epididymitis (0.2% to 6%), pneumonia/pulmonary infection UROLOGY 53 (1), 1999
(0.49% to 6.7%), and pyelonephritis (0.4% to 1.1%) have been reported for patients undergoing TURP.1–3 Additionally, from these studies, patients who had a preoperative vasectomy were at an increased risk (10%) for postsurgical epididymitis.1–3 The incidence of epididymitis in the present study was also reduced in the late group. All patients were given prophylactic antibiotic therapy for at least 14 days postoperatively to prevent postoperative infection, including epididymitis, and the introduction of a closed urinary drainage system in 1985 also significantly reduced the risk of morbidity. It is not clear whether these factors had a beneficial effect in TURP for preventing postoperative infection, but they may be important for the observed reduction in infection in the late group in the present study. Alternatives to TURP may be equally effective, with less morbidity, but they are not sufficient to totally replace standard TURP treatment.11–18 Although new types of surgical loops have been examined for the potential to reduce bleeding during TURP, conventional TURP remains fundamental to the transurethral treatment for BPH, and surgeons should become adept in this procedure.23 CONCLUSIONS The incidence of blood transfusions and morbidity in patients undergoing TURP has decreased compared with that during the 1970s. Advances in techniques, instrumentation, and surgical and perioperative management, including anesthesia, have made TURP a relatively safe procedure. TURP remains an effective treatment option for patients with BPH. REFERENCES 1. Holtgrewe HL, and Valk WL: Factors influencing the mortality and morbidity of transurethral prostatectomy: a study of 2,015 cases. J Urol 87: 450 – 459, 1962. 103
2. Melchior J, Valk WL, Foret JD, et al: Transurethral prostatectomy: computerized analysis of 2,223 consecutive cases. J Urol 112: 634 – 646, 1974. 3. Mebust WK, Holtgrewe HL, Cockett ATK, et al: Transurethral prostatectomy: immediate and postoperative complications: a cooperative study of 13 participating institutions evaluating 3,885 patients. J Urol 141: 243–247, 1989. 4. Uchida T, Adachi K, Ao T, et al: Clinical analysis in 2266 cases of transurethral resection of the prostate. Jpn J Urol 84: 890 – 896, 1993. 5. Lepor H, and Rigaud G: The efficacy of transurethral resection of the prostate in men with moderate symptoms of prostatism. J Urol 143: 533–537, 1990. 6. Wennberg JE, Roos N, Sola L, et al: Use of claims data systems to evaluate health care outcomes: mortality and reoperation following prostatectomy. JAMA 257: 933–936, 1987. 7. Roos NP, Wennberg JE, Malenka DJ, et al: Mortality and reoperation after open and transurethral resection of the prostate for benign prostatic hyperplasia. N Engl J Med 320: 1120 –1124, 1989. 8. Malenka DJ, Roos N, Fisher ES, et al: Further study of the increased mortality following transurethral prostatectomy: a chart-based analysis. J Urol 144: 224 –227, 1990. 9. Holtgrewe HL, Mebust WK, Dowd JB, et al: Transurethral prostatectomy: practice aspects of the dominant operation in American urology. J Urol 141: 248 –253, 1989. 10. Fowler FJ Jr, Wennberg JE, Timothy RP, et al: Symptoms status and quality of life following prostatectomy. JAMA 259: 3018 –3022, 1988. 11. Holtgrewe HL: Benign prostatic hyperplasia (editorial). J Urol 152: 2030, 1994. 12. Kaplan SA, Wasson JH, Kirby RS, et al: BPH: where are we in 1996? AUA News 1: 25–27, 1996. 13. Holtgrewe HL: Guidance for clinical investigations of devices used for the treatment of benign prostatic hyperplasia. J Urol 150: 1588 –1590, 1993. 14. Kaplan SA, and Olsson CA: State of the art: microwave therapy in the management of men with benign prostatic hyperplasia: current status. J Urol 150: 1597–1602, 1993. 15. Schatzl G, Madersbacher S, Lang T, et al: The early postoperative morbidity of transurethral resection of the prostate and of 4 minimally invasive treatment alternatives. J Urol 158: 105–111, 1997. 16. Uchida T, Yokoyama E, Iwamura M, et al: High intensity focused ultrasound for benign prostatic hyperplasia. Int J Urol 2: 181–185, 1995. 17. Uchida T, Egawa S, Iwamura M, et al: A non-randomized comparative study of visual laser ablation and transurethral resection of the prostate in benign prostatic hyperplasia. Int J Urol 3: 108 –112, 1996. 18. Oesterling JE: Benign prostatic hyperplasia: medical and minimally invasive treatment options. N Engl J Med 332: 99 –109, 1995. 19. Gee WF, Holtgrewe HL, Albertsen PC, et al: Practice trends in the diagnosis and management of benign prostatic hyperplasia in the United States. J Urol 154: 205–206, 1995. 20. Egawa S, Uchida T, and Koshiba K: Current and future trends in the interventional therapy for benign prostatic hyperplasia. J Urol 159: 1958 –1960, 1998. 21. Holtgrewe HL: Current trends in management of men with lower urinary tract symptoms and benign prostatic hyperplasia. Urology 51(suppl): 1–7, 1998. 22. Cockett ATK, and Koshiba K: Manual of Urologic Surgery. New York, Springer-Verlag, 1979, p 237. 23. Uchida T, Ohkawa A, Shibata Y, et al: New transurethral resection of the prostate using WEDGE® loop for benign prostatic hyperplasia. Jpn J Clin Urol 51: 1013–1017, 1997. 104
EDITORIAL COMMENT The present article provides insight into the adverse outcomes of 3861 transurethral resection of the prostate (TURP) procedures conducted at a single institution over a 25-year period, in itself an interesting analysis. The authors divided the cohort into the first 1930 and the second 1931 patients, a numerical split that classified the patient population into those operated on from 1971 to 1985 and those operated on from 1985 to 1996. This in itself is rather curious, because in most institutions the number of TURPs performed has dramatically decreased in the 1990s. In the present institution, however, the first 1930 cases were done over 14 years, whereas the second half was performed over only 11 years, resulting in an apparent increase in the number of procedures done per year. Although there may be many reasons for this increase, which should not be interpreted as an increase in the incidence of TURP per unit of population at risk, it is an interesting fact to note. Decreases in the morbidity rate (17.2% versus 9.5%) of the procedure and the need for blood transfusions (20.2% versus 6.1%) over time were noted when the two groups were compared. The decrease in morbidity was mostly due to a decrease in the rate of extravasation, need for hemostatic procedures, and postoperative epididymitis; other complications did not decrease at the same rate. Of course, one must realize that there was no dramatic decrease in any of these complications from before to after 1986; rather, a steady decline is the underlying cause, as becomes clear from Figure 1. Of note, during the 25 years, no apparent change in patient age or amount of resected tissue (25.9 versus 23.3 g) occurred. The analysis of individually reported series of TURP conducted for the Agency for Health Care Policy and Research (AHCPR) guidelines,1 spanning the better part of the 20th century, resulted in an overall transfusion rate of 12.5%. However, when sorted by date of publication, there was a similarly clear trend toward a reduction in transfusion rates noted over time. Table VI compares the 90% confidence interval (CI) from the AHCPR guidelines and the two groups of patients analyzed in the present report. It is clear that for the most part, the observed complication rates fall well into the 90% CI reported in the guidelines, which cover nearly the entire time period of observation reported in the present study. Several additional points are noteworthy. For one, the mortality rate is exceedingly low, even in the early group, a fact that the authors explain by the comorbidity status of Japanese versus American men and the routine use of prolonged antibiotic therapy. This, however, does not appear immediately intuitive because many studies have shown that even a single oral dose of antibiotic agents surrounding the time of surgery is effective in preventing infections or possibly lethal sepsis episodes. The administration of a 10% saline solution intravenously in all patients having electrolyte imbalance is in most parts of the world a very rarely used measure, if not regarded as outright dangerous. One can only speculate on the impact that this intervention in an unknown number of patients had in either a positive or negative sense. The one analysis that the AHCPR meta-analysis did not permit by focusing on studies rather than individual patients is perhaps the most significant message worth repeating. The larger the gland, the longer the resection time, and the longer the resection time, the higher the morbidity and mortality rates and the greater the need for transfusion. Most graduating residents finishing their training with fewer than 50 TURPs should take note and undertake a TURP only if the amount of resectable tissue is less than 60 g. This statement is supported by the AHCPR guidelines1 as well as the World Health Organization benign prostatic hyperplasia consensus conference.2 UROLOGY 53 (1), 1999