Clinical trial designs based on sequential conditional probability ratio tests and reverse stochastic curtailing.

We propose a group sequential method based on the sequential conditional probability ratio test and show that it has the conservatism desired in practice. We also propose calculating the discordant probability, that is, the probability that the sequential test concludes differently from a fixed-sample test at the planned end of the trial, recognizing that this probability could be substantial, even if the sequential test has the same size and power as the fixed-sample size test at the planned end of the study. In addition, we show that the proposed method can be used as a stochastic curtailing tool. Thus, the method accommodates unplanned interim analyses as well as those deemed necessary based on data trends, virtually without inflating the type I error, but it is less conservative than the usual stochastic curtailing. The method is implemented through an interactive computer program.

Random effects models in latent class analysis for evaluating accuracy of diagnostic tests.

When the results of a reference (or gold standard) test are missing or not error-free, the accuracy of diagnostic tests is often assessed through latent class models with two latent classes, representing diseased or nondiseased status. Such models, however, require that conditional on the true disease status, the tests are statistically independent, an assumption often violated in practice. Consequently, the model generally fits the data poorly. In this paper, we develop a general latent class model with random effects to model the conditional dependence among multiple diagnostic tests (or readers). We also develop a graphical method for checking whether or not the conditional dependence is of concern and for identifying the pattern of the correlation. Using the random-effects model and the graphical method, a simple adequate model that is easy to interpret can be obtained. The methods are illustrated with three examples from the biometric literature. The proposed methodology is also applicable when the true disease status is indeed known and conditional dependence could well be present.

Emory Angioplasty Versus Surgery Trial (EAST): design, recruitment, and baseline description of patients.

In patients with multivessel coronary artery disease who require revascularization there is uncertainty as to the selection of the appropriate patients for percutaneous transluminal coronary angioplasty (PTCA) as opposed to coronary artery bypass grafting (CABG). To define the relative roles of PTCA and CABG for multivessel disease, 392 patients were randomly assigned to revascularization by PTCA or CABG and followed for 3 years. This is a single-center randomized study in which individual group assignment was known but grouped data remained blinded until a full 3 years of follow-up was complete on all patients. The patients were randomized in 4 strata: (1) 2-vessel disease with 1 lesion per vessel; (2) 2-vessel disease with multiple lesions in at least 1 vessel; (3) 3-vessel disease with 1 lesion per vessel; and (4) 3-vessel disease with multiple lesions in at least 1 vessel. All data were collected on study-specific forms and sent to an independent biostatistical coordinating center for entry into a computerized database. All data will be analyzed by intention-to-treat. The primary endpoint of the trial is the composite of death of any cause, Q-wave myocardial infarction within 3 years, or a large reversible thallium defect at 3 years. Multiple secondary endpoints will include each of the components of the primary endpoint, the need for additional procedures, angiographic status at 1 and 3 years measured by an independent quantitative coronary arteriography laboratory, and measures of quality of life. A total of 5,118 patients were screened, of whom 3,371 were excluded for angiographic reasons, 191 because the angioplasty operators or surgeons believed that the patients were anatomically unsuitable, and 714 for clinical exclusions, leaving 842 eligible patients, of whom 392 were randomized. Of these, 40% had triple-vessel disease and 60% had double-vessel disease. There was no baseline difference between the 2 treatment arms for any clinical or major angiographic variable. This randomized trial will permit an in-depth comparison of coronary angioplasty and coronary surgery in comparable patients suitable for either procedure. Clinical, angiographic, and quality-of-life comparisons will be made and should be helpful in medical decision making between the 2 procedures.

Effect of distal splenorenal shunt on survival of patients with primary biliary cirrhosis.

Distal splenorenal shunt is known to effectively control bleeding from esophageal and gastric varices; however, the effect of this selective shunt on liver function is less well understood. We examined retrospectively the effect of distal splenorenal shunt on the survival of 19 patients with primary biliary cirrhosis subjected to surgery for bleeding varices over a 20-yr period and had been followed for at least 1 yr. Actual Kaplan-Meier survival curve was compared with predicted survival curve based on the Mayo Clinic model using clinical data collected at the time of surgery. The patients median length of follow-up was 65.9 mo. Ten of the 19 patients died or underwent orthotopic liver transplantation during the period of observation. The actual Kaplan-Meier and predicted Mayo Clinic model survival curves were similar and did not differ significantly. Survival was best in patients with good liver function (i.e., low Mayo risk scores). Distal splenorenal shunt, therefore, did not appear to have an adverse effect on the survival of patients with primary biliary cirrhosis. We conclude that variceal bleeding in primary biliary cirrhosis patients with good liver function should not be considered an indication for liver transplantation. Instead, if treatment with sclerotherapy or beta-blockers fails then distal splenorenal shunt will prevent recurrent bleeding in 90% of patients and leave them with an excellent prognosis.

A randomized trial comparing coronary angioplasty with coronary bypass surgery. Emory Angioplasty versus Surgery Trial (EAST)

BACKGROUND: The clinical benefit of percutaneous transluminal coronary angioplasty (PTCA) as compared with coronary-artery bypass grafting (CABG) for patients with multivessel coronary artery disease has not been established. To determine the outcomes of these treatments in patients referred for the first time for coronary revascularization, we conducted a three-year prospective, randomized trial comparing the two procedures. METHODS: Revascularization was performed by accepted methods. Follow-up clinical information was collected every six months, and coronary arteriography and thallium stress scanning were performed at one and three years. The primary end point was a composite of death, Q-wave myocardial infarction, and a large ischemic defect identified on thallium scanning at three years. Secondary end points included clinical and angiographic status and the need for additional revascularization procedures. Data were analyzed according to the intention-to-treat principle. RESULTS: Of the 5118 patients screened for the trial, 842 (16.5 percent) were eligible for enrollment, and 392 (7.7 percent) agreed to participate. A total of 194 patients were randomly assigned to the CABG group, and 198 to the PTCA group. The primary end point occurred in 27.3 percent of the CABG group and 28.8 percent of the PTCA group (P = 0.81). Death occurred in 6.2 percent of the CABG group and 7.1 percent of the PTCA group (P = 0.73 by log-rank test). At three years, the proportions of patients in the CABG group who required repeated bypass surgery (1 percent) or angioplasty (13 percent) were significantly lower than the proportions in the PTCA group (22 and 41 percent, respectively; P < 0.001). Angiographic studies at three years showed a greater degree of revascularization in the CABG group. Angina was more frequent in the PTCA group (20 percent) than in the CABG group (12 percent). CONCLUSIONS: We found that CABG and PTCA did not differ significantly with respect to the occurrence of the composite primary end point. Consequently, the selection of one procedure over the other should be guided by patients' preferences regarding the quality of life and the possible need for subsequent procedures.

Screening for cognitive impairment in older individuals. Validation study of a computer-based test.

OBJECTIVE: This study examined the validity of a computer-based cognitive test that was recently designed to screen the elderly for cognitive impairment. DESIGN: Criterion-related validity was examined by comparing test scores of impaired patients and normal control subjects. Construct-related validity was computed through correlations between computer-based subtests and related conventional neuropsychological subtests. SETTING: University center for memory disorders. PARTICIPANTS: Fifty-two patients with mild cognitive impairment by strict clinical criteria and 50 unimpaired, age- and education-matched control subjects. Control subjects were rigorously screened by neurological, neuropsychological, imaging, and electrophysiological criteria to identify and exclude individuals with occult abnormalities. RESULTS: Using a cut-off total score of 126, this computer-based instrument had a sensitivity of 0.83 and a specificity of 0.96. Using a prevalence estimate of 10%, predictive values, positive and negative, were 0.70 and 0.96, respectively. Computer-based subtests correlated significantly with conventional neuropsychological tests measuring similar cognitive domains. Thirteen (17.8%) of 73 volunteers with normal medical histories were excluded from the control group, with unsuspected abnormalities on standard neuropsychological tests, electroencephalograms, or magnetic resonance imaging scans. CONCLUSIONS: Computer-based testing is a valid screening methodology for the detection of mild cognitive impairment in the elderly, although this particular test has important limitations. Broader applications of computer-based testing will require extensive population-based validation. Future studies should recognize that normal control subjects without a history of disease who are typically used in validation studies may have a high incidence of unsuspected abnormalities on neurodiagnostic studies.

Volumetric and functional liver blood flow are both increased in the human transplanted liver.

This study measured volumetric liver blood flow and galactose clearance concurrently during orthotopic liver transplant in human subjects. Ultrasound transit time flowmeters measured hepatic artery and portal vein flow 1-3 h after reperfusion. Galactose (100 mg/min) was infused over 45-60 min to steady state for calculation of clearance. Mean +/- S.D. total volumetric flow was 1966 +/- 831 ml/min with portal flow contributing 86%. Mean galactose clearance was 1988 +/- 641 ml/min. There was a significant correlation (p < 0.05, r = 0.61) between volumetric total liver blood flow and galactose clearance. The data show that: (i) the newly transplanted liver is capable of metabolizing galactose within 1-3 h of reperfusion; and (ii) liver blood flow is high in the newly implanted liver. The clinical importance of this observation is that there is increased clearance of high first pass substances by the transplanted liver which may be of importance in patient management.

Hemodynamics during liver transplantation: the interactions between cardiac output and portal venous and hepatic arterial flows.

Liver blood flow and systemic hemodynamics were measured intraoperatively in 34 patients after liver transplantation. Ultrasound transit-time flow probes measured hepatic arterial and portal venous flow over 10 to 75 min 1 to 3 hr after reperfusion. Cardiac output was measured by thermodilution. Mean cardiac output was 9.5 +/- 2.8 L/min; the mean total liver blood flow of 2,091 +/- 932 ml/min was 23% +/- 11% of cardiac output. Mean portal flow of 1,808 +/- 929 ml/min was disproportionately high at 85% +/- 10% of total liver blood flow. Correlation analysis showed a significant (p less than 0.01; r = 0.42) correlation between cardiac output and portal venous flow and a trend toward negative correlation (p = 0.087) between cardiac output and hepatic arterial flow. These data show that increased flow in the newly transplanted liver is predominantly portal venous flow and is associated with high cardiac output and reduced hepatic arterial flow. In the last 13 patients studied, portal flow was reduced by 50% and the hepatic artery response was measured. We saw a significant (p less than 0.05) increase in hepatic artery flow from 322 +/- 228 to 419 +/- 271 ml/min, indicating an intact hepatic arterial buffer response. The hepatic artery response also showed that it is a reversible rather than a fixed resistance that contributes to the low hepatic artery flow in these patients.

Selective shunt in the management of variceal bleeding in the era of liver transplantation.

This study reports the Emory experience with 147 distal splenorenal shunts (DSRS) and 110 orthotopic liver transplants (OLT) between January 1987 and December 1991. The purpose was to clarify which patients with variceal bleeding should be treated by DSRS versus OLT. Distal splenorenal shunts were selected for patients with adequate or good liver function. Orthotopic liver transplant was offered to patients with end-stage liver disease who fulfilled other selection criteria. The DSRS group comprised 71 Child's A, 70 Child's B, and 6 Child's C patients. The mean galactose elimination capacity for all DSRS patients was 330 +/- 98 mg/minute, which was significantly (p less than 0.01) above the galactose elimination capacity of 237 +/- 82 mg/minute in the OLT group. Survival analysis for the DSRS group showed 91% 1-year and 77% 3-year survival, which was better than the 74% 1-year and 60% 3-year survivals in the OLT group. Variceal bleeding as a major component of end-stage disease leading to OLT had significantly (p less than 0.05) poorer survival (50%) at 1 year compared with patients without variceal bleeding (80%). Hepatic function was maintained after DSRS, as measured by serum albumin and prothrombin time, but galactose elimination capacity decreased significantly (p less than 0.05) to 298 +/- 97 mg/minute. Quality of life, measured by a self-assessment questionnaire, was not significantly different in the DSRS and OLT groups. Hospital charges were significantly higher for OLT (median, $113,733) compared with DSRS ($32,674). These data support a role for selective shunt in the management of patients with variceal bleeding who require surgery and have good hepatic function. Transplantation should be reserved for patients with end-stage liver disease. A thorough evaluation, including tests of liver function, help in selection of the most appropriate therapeutic approach.

Role of sodium in mediator release from human basophils.

We studied the effect of extracellular sodium concentration on histamine release (HR) from human basophils initiated by immunologic and nonimmunologic stimuli. We found that lowering extracellular sodium markedly enhances HR induced by an immunologic stimulus from these cells. In buffer in which sodium had been replaced with univalent ions of strong bases, enhancement of HR increased as extracellular sodium decreased. Enhancement was the result of increased duration of release. When sucrose was used for replacement of sodium, we also observed that enhancement of HR increased as extracellular sodium decreased, but there was some lessening of enhancement at [Na+]e between 5 and 10 mmol/L. Ouabain, which is an inhibitor of the Na+/K+ adenosine triphosphatase, and bumetanide and furosemide, which are inhibitors of Cl(-)-dependent Na(+)-K+ cotransport, caused small increases in enhancement of HR by sodium-deficient buffers; 4,4'-diisothiocyanostilbene-2-2'-disulfonic acid, an anion transport inhibitor, caused some inhibition of enhancement of HR. Analogues of amiloride, such as 5-(N-N-hexamethylene) amiloride (HMA) and 5-(N-4-chlorobenzyl)-2'-4'dimethylbenzamil (CBDMB), inhibit Na+/H+ exchange, Na+/Ca++ exchange, and Na+ channels. Interestingly, at higher doses, HMA and CBDMB caused marked enhancement of HR in both normal and sodium-deficient buffers. These results suggest that several cellular regulatory mechanisms potentially are important for normal basophil secretion. The most likely are pH regulatory mechanisms that include Na+/H+ exchange and anion exchangers that transport alkaline equivalents. Our findings enhancement of basophil HR by HMA and CBDMB is particularly noteworthy in light of the recent interest in use of amiloride by inhalation for therapy of lung disease in patients with cystic fibrosis.

Prevalence of invasive cytomegalovirus disease with administration of muromonab CD-3 in patients undergoing orthotopic liver transplantation.

OBJECTIVE: To assess the association of cytomegalovirus (CMV) disease with the administration of muromonab CD-3 (OKT-3) in patients undergoing liver transplant; specifically, to assess the risk of OKT-3 use as an agent for rejection prophylaxis and as an agent for therapy of rejection. DESIGN: Retrospective review of medical records. STUDY POPULATION: 83 liver transplant recipients (43 men, 40 women) with a mean age of 41.5 years (range 16-62). DATA EXTRACTION: The medical record for each liver transplant recipient was reviewed and analyzed for the following variables: (1) preoperative recipient CMV serology, (2) donor CMV serology, (3) incidence of invasive CMV disease, (4) administration of OKT-3, (5) postoperative administration time of OKT-3, and (6) the relationship between the administration of OKT-3 and the prevalence of invasive CMV disease. RESULTS: OKT-3 was administered to 34 of 83 (40.9 percent) liver remainder received OKT-3 as rejection rescue. All patients received OKT-3 5 mg iv for 14 days. Seventeen of the 34 patients receiving OKT-3 (50 percent) developed invasive CMV disease; 58.8 percent of the patients (20/34) receiving OKT-3 were given the agent within the first 14 postoperative days. Sixteen of these 20 patients (80 percent) developed invasive CMV disease. One of 14 patients (7.1 percent) who received OKT-3 after the first 14 postoperative days developed invasive CMV disease. Of those patients 94 percent (16/17) received OKT-3 in the first 14 postoperative days. This prevalence differed significantly from those receiving OKT-3 after the 14th postoperative day and those who did not receive OKT-3 at any time during their hospital course. CONCLUSIONS: The patients who received early administration of OKT-3 in our study had a greater risk of invasive CMV disease than did those who received OKT-3 later in the hospital course.

High cardiac output of advanced liver disease persists after orthotopic liver transplantation.

This study measured cardiac output before and 1 or 2 yr after orthotopic liver transplantation in 23 patients. Cardiac output was measured by thermodilution before transplantation and by first-pass radionuclide angiocardiography at follow-up. Study patients were selected as those doing well clinically and by standard laboratory tests at 1-yr and 2-yr reevaluations with no evidence of rejection: six had mild recurrent hepatitis shown on biopsy samples. Hepatocyte function was normal at the time of the study as shown by galactose elimination capacity of 442 +/- 90 mg/min. Medications were cyclosporine and prednisone in all patients, azathioprine in 10 patients and a combination of antihypertensive therapy to maintain diastolic blood pressure less than 90 mm Hg in 20 patients. Mean (+/- S.D.) pretransplantation cardiac output was 9.1 +/- 3.1 L/min and remained elevated at 8.3 +/- 2.1 L/min 1 yr, and 9.6 +/- 2.6 L/min (n = 13) 2 yr after transplantation. A significant (p less than 0.001) correlation was found between pretransplant and follow-up cardiac output. End diastolic, end systolic and stroke volumes are all increased in a pattern similar to that seen in end-stage cirrhosis. These data show that the high cardiac output of the hyperdynamic state of advanced liver disease persists after liver transplantation. The mechanisms and consequences of this require further study.

The effect of liver denervation on hepatic hemodynamics during hypovolemic shock in swine.

This study tested the hypothesis that the denervated liver is more susceptible to hypovolemic shock than the normal liver. Fourteen swine, seven nondenervated and seven after liver denervation, were studied during hypovolemic shock to 50% of baseline blood pressure. Hepatic artery and portal vein flows were measured using transonic flow probes, and cardiac output and central venous pressure were measured using Swan-Ganz catheters. Hepatic artery flow fell equivalently in the two groups, from 132 +/- 71 to 94 +/- 17 ml/min in the nondenervated group compared with 149 +/- 56 to 91 +/- 55 ml/min in the denervated group. In contrast, portal flow in the denervated group (276 +/- 71 to 119 +/- 53 ml/min) fell significantly (p less than 0.001) more than in the nondenervated group (289 +/- 135 to 194 +/- 70 ml/min). The 58% reduction from baseline in portal flow in the denervated group compared with the 30% reduction in the nondenervated group suggests that the normal compensatory mechanism to maintain portal flow during hypovolemic shock is neurally mediated. It can be hypothesized that sensory afferent fibers might initiate a feedback to splanchnic vasodilatation in response to reduced portal flow. This study supports the hypothesis that the denervated liver is more susceptible to hypovolemic shock.

Medroxyprogesterone acetate, aggression, and sexual behavior in male cynomolgus monkeys (Macaca fascicularis).

Medroxyprogesterone acetate (MPA) is used clinically to treat male sex offenders, but there are conflicting reports about its effects on aggression. To investigate these matters in a nonhuman primate, four intact male cynomolgus monkeys were studied in a testing paradigm that involved the presence of a caged, aggression-arousing stimulus male either immediately before or during a pair-test with an ovariectomized, untreated female partner. After two 4-week periods of pretreatment baseline, males received weekly injections of 40 mg MPA either alone (two 4-week treatment periods) or in combination with testosterone replacement with sc implants (one period) and additional daily injections of 2 mg testosterone propionate (two periods). MPA was then withdrawn while testosterone replacement continued (three periods). The testing paradigm was effective in maintaining aggression, especially male-male aggression, for many months. Male-male aggression increased with MPA treatment, and increased further with testosterone replacement, whereas male-female aggression tended to change in the opposite direction. As in earlier studies, MPA decreased both plasma testosterone and male sexual activity, but restoring plasma testosterone levels in treated males failed to restore their sexual activity. MPA therefore has behavioral effects that are not mediated primarily by its suppression of circulating androgens.

Results of the Prospective Evaluation of Radial Keratotomy (PERK) Study five years after surgery. The Perk Study Group.

In the Prospective Evaluation of Radial Keratotomy (PERK) Study, 793 eyes of 435 patients with 2 to 8 diopters (D) of myopia received a standardized surgery consisting of 8 incisions with a diamond-bladed knife set at 100% of the thinnest paracentral ultrasonic corneal thickness measurement and a diameter of the clear zone of 3.0 to 4.5 mm; 97 eyes (12%) received an additional 8 incisions. There were 757 eyes (95%) followed for 3 to 6.3 years. After surgery, uncorrected visual acuity was 20/40 or better in 88% of eyes. The refractive error was within 1 D of emmetropia for 64% of eyes; 19% were myopic and 17% were hyperopic by more than 1 D. Between 6 months and 5 years after surgery, 22% of the eyes had a refractive change of 1 D or more in the hyperopic direction. For 25 eyes (3%) there was a loss of 2 or more lines of best spectacle-corrected visual acuity.

Stability of refraction during four years after radial keratotomy in the prospective evaluation of radial keratotomy study.

The Prospective Evaluation of Radial Keratotomy Study is a nine-center clinical trial of a standardized technique of radial keratotomy in 435 patients who had simple myopia with a preoperative refraction between -2.00 and -8.00 diopters. We studied the stability of the refractive error during four years after surgery for each of 341 eyes first operated on that had a single surgical procedure. Between baseline and two weeks after surgery, all eyes became less myopic; between two weeks and three months, 161 eyes (59%) lost 1.00 D or more of the initial effect; and between three and six months, 266 eyes (95%) had a stable refraction with less than 1.00 D change. The change from six months to four years was less than 1.00 D for 246 eyes (72%). There was 1.00 D or more decrease in surgical effect (increased minus power) for 13 eyes (4%), and 1.00 D or more increase in surgical effect (decreased minus power) for 82 eyes (24%). Eyes with larger amounts of preoperative myopia and smaller diameter of the clear zone were more likely to have an increasing effect of the surgery. The duration of this continued increasing effect of the surgery is unknown.

Model of spontaneous obesity in aging male Wistar rats.

We analyzed retrospectively data from 148 chow-fed male Wistar rats killed between the age of 6 wk and 2 yr while varying in body weight from 136 to 917 g. The purpose of this study was to clarify the relationship of body weight and body lipid content with the composition and cellularity of the epididymal and retroperitoneal fat depots. A positive linear association was found between body weight and body water or fat-free dry residue, whereas total body lipid exhibited a curvilinear relationship with body weight. The weight of the epididymal pads was linearly related to body weight but not to body lipid. In contrast, retroperitoneal pad weight was exponentially related to body weight and paralleled total body lipid. A strong linear correlation was found between total body lipid and weight (r = 0.959) or depot lipid content (r = 0.967) of the retroperitoneal fat pads. In this rat model of aging and spontaneous obesity, significant regional differences exist in adipose depot composition and cellularity. A practical outcome of this study is a simple and accurate prediction of body lipid content from the gravimetric determination of the retroperitoneal fat depots.

Kinetics of system A amino acid uptake by muscle: effects of insulin and acute uremia.

We found that acute renal failure (ARF) depresses maximal responsiveness to insulin of system A amino acid transport in muscle [Maroni, Karapanos, and Mitch. Am. J. Physiol. 251 (Renal Fluid Electrolyte Physiol. 20): F74-F80, 1986]. To examine mechanisms for this change, we measured the kinetics of system A in response to insulin and ARF in epitrochlearis muscle. Insulin stimulation increased the diffusion constant (KD) approximately twofold in muscles from both ARF and control rats, suggesting that insulin can modulate nonsaturable alpha-(methylamino)isobutyrate uptake. Compared with basal values, insulin did not significantly change maximal transport velocity (Vmax) in ARF or control rats; however, insulin decreased Michaelis constant (Km) by 79 and 63%, respectively. Inhibition of protein synthesis with cycloheximide did not prevent stimulation of system A by insulin. Acute uremia had no effect on nonsaturable uptake or basal system A kinetics. However, ARF produced opposing effects on insulin-stimulated system A kinetics. There was a approximately 47% decrease in Vmax, which was partially offset by a 66% reduction in Km. In summary, both insulin and ARF modulate system A kinetics by mechanisms that are independent of protein synthesis.

Endoscopic variceal sclerosis compared with distal splenorenal shunt to prevent recurrent variceal bleeding in cirrhosis. A prospective, randomized trial.

STUDY OBJECTIVE: To define the roles of endoscopic variceal sclerosis and distal splenorenal shunt in the prevention of recurrent variceal bleeding in patients with cirrhosis. DESIGN: A prospective, randomized clinical trial with crossover for those failing therapy. The median follow-up was 61 months. SETTING: A private, tertiary-referral university hospital. PATIENTS: Seventy-two patients fulfilling inclusion criteria were drawn from a total of 420 patients treated during a 4.5-year interval. TREATMENTS: Endoscopic variceal sclerosis or distal splenorenal shunt. MEASUREMENTS AND MAIN RESULTS: Survival was significantly (P = 0.02) improved in patients randomly assigned to receive sclerotherapy: 13 of these 37 (35%) patients failed sclerotherapy and required surgical rescue. A survival advantage (P = 0.01) was seen in patients with alcoholic cirrhosis who had this combined therapy; however, in patients with nonalcoholic cirrhosis, survival for those receiving sclerotherapy and surgical rescue was not significantly (P = 0.36) different from that of patients receiving distal splenorenal shunt. Control of variceal bleeding was significantly (P less than 0.001) better in the distal splenorenal shunt group (34 of 35 [97%] compared with 15 of 37 [41%] in the sclerotherapy group). Using death, uncontrolled rebleeding, or shunt thrombosis as the endpoints resulted in no significant difference between treatment groups. Hepatocyte function and portal perfusion were significantly better maintained in patients with alcoholic cirrhosis who were managed by sclerotherapy rather than shunt (P = 0.01 and P = 0.001, respectively). CONCLUSIONS: Endoscopic sclerotherapy with surgical rescue for uncontrolled bleeding is the optimum therapy for patients with alcoholic cirrhosis and variceal bleeding. Survival is similar in nonalcoholic patients treated with either distal splenorenal shunt or endoscopic sclerotherapy, but shunting provides better control of variceal bleeding.