Randomized, prospective, double-blind, placebo-controlled trial of effect of nitroglycerin ointment on pain after hemorrhoidectomy.

PURPOSE: Fissure-in-ano is characterized by pain, bleeding, and internal anal sphincter hypertonicity. Spasm of the internal sphincter also plays a role in hemorrhoidal disease and may be a source of anal pain after hemorrhoid surgery. Inducing sphincter relaxation with a nitroglycerin ointment has shown promise in healing anal fissures and relieving symptoms of pain. Our study attempts to test the hypothesis that topical nitroglycerin applied to the perianal region is beneficial in reducing pain after hemorrhoidectomy. METHODS: After hemorrhoidectomy 39 patients were randomly assigned to receive 0.2 percent nitroglycerin ointment (n = 19) or placebo (n = 20). Ointments were applied to the perianal region three times daily for seven days. Patients were prescribed hydrocodone bitartrate to take as needed. Visual analog scales were used to measure postoperative pain intensity and ointment benefits. Patients completed questionnaires to record medication morbidity and number of prescribed or nonprescribed medications taken. RESULTS: Patients using nitroglycerin had less pain and greater benefit from ointment than those did in the placebo group, but differences were not significant. Narcotic use was higher in the placebo group when considered on a daily basis, but was statistically significant on the second postoperative day only (P < 0.05). Morbidity from ointment application was significantly higher in the nitroglycerin group (P < 0.002) and included a headache in 8 of 19 patients. Nonsteroidal anti-inflammatory drugs and acetaminophen were not prescribed, but were taken more frequently in nitroglycerin patients (P < 0.0003). CONCLUSION: Perianal application of 0.2 percent nitroglycerin ointment after hemorrhoidectomy significantly reduced narcotic requirements on the second postoperative day. Headaches and a subsequent need for nonnarcotic medications may limit benefits of nitroglycerin.

Open versus percutaneous dilatational tracheostomy: efficacy and cost analysis.

The economic advantages of percutaneous dilatational tracheostomies versus open tracheostomies in the operating room have been thoroughly evaluated. We are now reporting our comparison of the costs and charges of percutaneous dilatational tracheostomies with those of open bedside tracheostomies at our institution. The current literature comparing the two open techniques and the percutaneous method of placing tracheostomies was reviewed and the charges and costs for these procedures at our institution were compared. Patients were placed into one of three groups for analysis: open tracheostomies in the operating room (Group I), open tracheostomies in the intensive care unit (Group II), and percutaneous dilatational tracheostomies in the intensive care unit (Group III). Based on our own experience and a literature review it is evident that all three approaches to tracheostomies are safe. Economic analysis showed a savings of $180 in cost per procedure and a $658 savings in charges per procedure for the open method at the bedside when compared with the percutaneous method at the bedside. The professional fee for bronchoscopy was not included in this calculation; including this would lead to greater savings with the open method over the percutaneous method. Open tracheostomy in the operating room increased costs over the bedside procedure by $2194 and increased charges by $2871. For the 150 to 180 tracheostomies done each year at our institution utilization of the open technique at the bedside results in a cost savings of approximately $31,500 and a charge savings of $109,000 compared with the percutaneous dilatational tracheostomy. Both the open bedside and percutaneous dilatational methods are reasonable and safe options. However, the open bedside tracheostomy is a better utilization of resources and is more cost effective, and it is the procedure of choice at our institution.

Transrenal fixation of aortic stent-grafts for the treatment of infrarenal aortic aneurysmal disease.

PURPOSE: We evaluated our early experience with the transrenal fixation of aortic stent-grafts to determine the efficacy of this procedure and its effects on renal artery patency and hemodynamics. METHODS: Twenty-eight patients (22 men) had endoluminally placed modular bifurcated stent-grafts with a bare spring structure at the proximal end crossing the origin of both renal arteries; no patient with infrarenal fixation was included for analysis. The mean age of the patients was 75 +/- 7 years (range, 58-86 years); the mean aneurysm size was 5.8 +/- 0.8 cm (range, 4.7-7.2 cm). Eight patients had preoperative or intraoperative angiographic evidence of renal artery atherosclerotic disease, but only four vessels had luminal narrowing of 50% or greater. No complications were noted during stent-graft placement, and all patients have returned for follow-up visits, ranging from 1 to 12 months (mean follow-up, 6 +/- 4 months). Follow-up evaluations included clinical assessment, duplex ultrasound scan of the renal arteries and kidneys, and computed tomographic angiography. RESULTS: No evidence of lobular or sublobular perfusion defects of the renal parenchyma was detected postoperatively. Two patients exhibited postoperative changes in renal artery hemodynamics-one progressing from a 30% diameter reduction to a greater than 60% diameter stenosis at the 12-month follow-up visit and one with a normal renal artery preoperatively having elevated flow velocities indicative of a greater than 60% stenosis at the 1-month visit. Of 19 patients with normal preoperative renal function, only one has had persistently elevated serum creatinine levels. CONCLUSION: We conclude from this experience that the transrenal placement of open stents is safe and effectively excludes the aneurysm, potentially expanding the availability of this technique to more patients with a short infrarenal aortic neck. Long-term follow-up is essential to determine the overall efficacy of this technique and to identify potential effects on renal artery hemodynamics or kidney function.

Nonoperative management of blunt splenic injury in adults 55 years and older: a twenty-year experience.

The nonoperative management of splenic injury secondary to blunt trauma in older patients remains controversial. We have reviewed our experience from January 1978 to December 1997 with selective nonoperative management of blunt splenic injury in adults 55 years and older. Criteria for nonoperative management included hemodynamic stability with any transient hypotension corrected using less than 2,000 cm3 crystalloid infusion, a negative abdominal physical examination ruling out associated injuries, and a blood transfusion requirement of no more than 2 units attributable to the splenic injury. During the study period, 18 patients over age 55 with radiographic confirmation of a splenic injury met the above criteria for nonoperative management. Their mean age was 72 years (range 56-86), and 13 of the 18 were female (72%). The mean Injury Severity Score was 15 (range 4-29), with the mechanism of injury equally divided between automobile crashes (9) and falls (9). During a similar time period, 15 patients 55 years or older with splenic injury composed an operative group; these patients did not differ with respect to age (mean 68 years), sex (60% female), or mechanism of injury. CT scans of 8 patients managed nonoperatively were available and graded using the American Association for the Surgery of Trauma classification, with a mean score of 2.3 (range 2-3). Eight of the 18 nonsurgical patients received blood transfusions. None of the 18 patients who met the criteria for nonoperative management "failed" the protocol, and none were taken to the operating room for abdominal exploration. Two patients (11%) died of associated thoracic injuries after lengthy hospital stays, one at 10 days and one at 24 days. We conclude from our data that nonoperative management of blunt splenic injury in patients age 55 years and older is indicated provided they are hemodynamically stable, do not require significant blood transfusion, and have no other associated abdominal injuries.

Hypothermia-induced coagulopathy during hemorrhagic shock.

A porcine model of hemorrhagic shock was used to study the effect of hypothermia on hemodynamic, metabolic, and coagulation parameters. The model was designed to simulate the events of severe blunt injury with hemorrhage occurring initially, to a systolic blood pressure of 30 mm Hg, followed by simultaneous hemorrhage and crystalloid volume replacement, followed by cessation of hemorrhage and blood replacement. Half of the animals were rendered hypothermic by external application of ice, and half remained normothermic. There was seven pigs in each group. Two deaths occurred in each during the hemorrhage phase. The hypothermic pigs demonstrated larger reduction in cardiac output than normothermic pigs. Volume replacement in the normothermic group restored cardiac output to baseline values. In the hypothermic group, cardiac output remained depressed despite volume replacement. Prothrombin times and partial thromboplastin times showed significantly more prolongation in the hypothermic group. Furthermore, this was not corrected by replacement of shed blood in the hypothermic group, as was seen in the normothermic group. We conclude that when shock and hypothermia occur together, their deleterious effect on hemodynamic and coagulation parameters are additive. The effects of hypothermia persist despite the arrest of hemorrhage and volume replacement. Thus, it is necessary to aggressively address both shock and hypothermia when they occur simultaneously.

Human endothelial cell cultures from progenitor cells obtained by leukapheresis.

Although improved prosthetic graft patency with endothelial cell (EC) seeding has been shown in animal models, the clinical application of this technique requires a convenient source of ECs. We have evaluated EC cultures derived from the mononuclear cell (MNC) fraction obtained during large-volume leukapheresis and compared this with cultures grown from peripheral blood cells obtained by phlebotomy. Leukapheresis was performed in healthy adult volunteers (n = 7) using software designed to increase the percentage of MNCs harvested. Blood (40-293 mL) was drawn from a peripheral vein in healthy adult volunteers (n = 13), and the MNCs were obtained by differential centrifugation using a Lymphoprep gradient. Significantly more MNCs were obtained by leukapheresis than by phlebotomy. Each leukapheresis procedure yielded 12.5 to 23 mL, which contained 2.29 +/- 0.35 x 10(9) MNCs, compared with 2.16 +/- 0.50 x 10(8) MNCs, for each phlebotomy (P < 0.001). EC colonies developed in significantly more cultures from leukapheresis-derived MNCs (6 of 7) than phlebotomy-derived MNCs (4 of 13; P = 0.008). Leukapheresis-derived cells developed EC morphology at 15.5 +/- 2 days compared with 21 +/- 3.4 days for cells obtained by phlebotomy (P = not significant). EC were identified by positive factor VIII and vascular endothelial growth factor receptor immunostaining. Leukapheresis significantly increases the number of progenitor cells available for differentiation into EC compared with phlebotomy and avoids the need for any surgical procedure to harvest a peripheral vein as a direct source of ECs.

Isolation of endothelial cells and their progenitor cells from human peripheral blood.

PURPOSE: We have developed techniques to isolate endothelial cell (EC) progenitors from human peripheral and umbilical cord blood. METHODS: Human adult peripheral and umbilical cord blood monocytes were isolated by centrifugation, and progenitor cells were separated with the use of magnetic polystyrene beads that were coated with a monoclonal antibody specific for the CD34 cell-membrane antigen. Cells were propagated in selective media, and developing cultures were immunostained for CD31, CD34, factor VIII, and vascular endothelial growth factor cell receptors. ECs that developed were transfected with a gene for prourokinase and used to line ePTFE grafts, which were evaluated in vitro in a pulsatile flow system. RESULTS: Umbilical cord monocyte cultures demonstrated colonies that resembled ECs at approximately 2 weeks, with growth being best supported by EC growth media plus 20% calf serum with iron. Immunostaining of colonies was positive for CD31 and factor VIII. After 18 days in culture, CD34(+) cells from adult peripheral blood were noted, which had the typical cobblestone appearance of ECs and immunostained positively for CD31 and factor VIII-related antigens. Cultures of umbilical cord-derived cells and adult peripheral blood-derived cells developed complex line formations within 1 week in culture that stained positively for vascular endothelial growth factor receptor-2. Urokinase-transfected ECs were shown to overexpress urokinase. Prosthetic grafts lined with transfected cells showed 87.33% +/- 4.97% cell adherence after 2 hours in a pulsatile flow system at clinically relevant shear stress. CONCLUSION: We conclude that endothelial progenitor cells can be isolated from human adult peripheral and umbilical cord blood and developed into EC cultures as a source of cells for vascular graft seeding and gene therapy.

Octreotide in the treatment of thoracic duct injuries.

Anecdotal reports support the use of octreotide in the treatment of traumatic thoracic duct injuries and chylothorax, but no prospective studies have proved its efficacy. We evaluated the effects of octreotide in treating thoracic duct transection in a canine model. Eight mongrel dogs (27.8+/-5.1 kg) were fed one pint of 10.5 per cent milkfat 2 hours before operation. Through a left supraclavicular neck incision, the thoracic duct was identified and transected, producing free flow of chyle. A quarter-inch drain was tunneled subcutaneously from the wound and attached to closed suction. After wound closure dogs were randomized to a control group (n = 4) receiving sham injections of saline subcutaneously three times per day, or a treatment group (n = 4) given 3 microg/kg octreotide three times per day. Postoperatively all dogs were fed a standard low-fat (5-7%) crude fat diet. Drain output was measured each day, and on odd-numbered postoperative days the drainage was analyzed for cholesterol, triglycerides, albumin, and total protein. Fistula closure was defined as drainage <10 ml/24-hour period. Treated dogs achieved fistula closure significantly faster than controls: 3.5+/-1.3 days versus 7.8+/-1.0 days (P = 0.0037). Whereas equivalent amounts of drainage occurred on the day of surgery and on postoperative day one in both groups, by postoperative day 2 the treatment group had significantly less drainage over 24 hours: 63+/-69 ml versus 195+/-79 ml (P = 0.046); this significant difference persisted through postoperative day 5 when drainage began to decrease in the control group. No significant differences between groups were seen in levels of cholesterol, triglycerides, albumin, or protein in the drainage at any time point. We conclude that octreotide is effective in treating thoracic duct injury, leading to an early decrease in drainage and early fistula closure. The mechanism for this effect remains to be clarified.

The effect of graded compression elastic stockings on the lower leg venous system during daily activity.

PURPOSE: We evaluated the effects of wearing 20 to 30 mm Hg gradient thigh-length stockings during daily activity on the lower leg venous system. METHODS: Twenty-one healthy women volunteers, aged 39 +/- 12 years, were examined. One subject was clinical class 4, one was class 2, two were class 1, and the remainder were class 0. The diameter of the posterior tibial, peroneal, and greater saphenous veins were measured at midcalf by means of bilateral duplex ultrasound scanning; calf circumference was measured at the same level. In addition, the number and caliber of all detectable medial calf perforating veins were recorded. Each subject was examined on two separate days, one while wearing the gradient stockings and one not wearing them. Baseline measurements were made on each day before the subjects began their workday, with follow-up measurements made after approximately 4.5 hours of normal activity. RESULTS: Calf circumference increased an average of 23.8 +/- 10.1 mm without stockings (P <.001) and decreased by 5.2 +/- 7.0 mm when wearing stockings (P =.003). The number of detected perforating veins increased without stockings by 1.8 +/- 2.0 at the follow-up examination (P =.002); with stockings, the change was 0.4 +/- 1.2 (P = NS). CONCLUSION: Graded compression elastic stockings help preserve lower leg venous caliber and tone throughout the deep, superficial, and perforating venous systems during normal ambulatory activity, and this may, in part, explain their beneficial effects.

Prosecution and conviction of the injured intoxicated driver.

BACKGROUND: Previous studies have reported low conviction rates for drunk drivers injured in motor vehicle crashes and transported to the hospital. The purpose of this study was to evaluate this rate during a recent period and to investigate the variables that predict alcohol-related convictions for injured drunk drivers admitted to our hospital. METHODS: A retrospective review of medical records from January 1991 through May 1997 identified 71 patients who were legally intoxicated drivers injured in motor vehicle crashes. Court records, police reports, and driving records were also obtained. RESULTS: Overall, 51% of the drunk drivers were convicted of alcohol-related offenses and 32% escaped without any conviction. Blood alcohol level and a police officer's estimation of whether the driver had been drinking were significant predictors of an alcohol-related conviction. Age, Injury Severity Score, a police officer's estimation of injury, and the number of people or cars involved in the crash were not significantly associated with legal outcome. CONCLUSION: Although this study shows an important increase in alcohol-related conviction rates, responsibility for further progress will depend on the medical community, law enforcement agencies, and the judicial system working together.

Three-dimensional vascular imaging using Doppler ultrasound.

BACKGROUND: We have evaluated the efficacy of using three-dimensional reconstruction of amplitude Doppler imaging data to quantitatively assess carotid artery bifurcation stenoses. METHODS: Sixty-four consecutive frames of amplitude (power) Doppler images are stored to be reassembled into a three-dimensional image representing the patent lumen. These images can then be rotated by any angle necessary to clearly view the vascular anatomy and to make quantitative ultrasound caliper measurements of the stenotic lumen and normal vessel caliber. RESULTS: Three-dimensional Doppler images accurately classified 53 of 61 vessels (87%) into categories of stenosis compared with angiography. All stenoses with >60% diameter reduction were detected and classified as such, for a sensitivity of 100%. CONCLUSIONS: Three-dimensional vascular imaging based on amplitude (power) Doppler data provides an accurate noninvasive technique for quantitative diagnosis of carotid bifurcation atherosclerotic disease, with selectable viewing projections that eliminate vessel overlap and other artifacts, and complements the hemodynamic data already available with two-dimensional duplex ultrasound.

Improved adherence of genetically modified endothelial cells to small-diameter expanded polytetrafluoroethylene grafts in a canine model.

PURPOSE: A significant limitation to using genetically modified endothelial cells (ECs) to seed prosthetic grafts before implantation has been poor cell adherence to the graft lumen. Methodologic changes to improve cell adherence were evaluated in a canine carotid interposition graft model using 4 mm interior diameter expanded polytetrafluoroethylene. METHODS: ECs harvested from external jugular veins were grown in culture, with 80% of the cells from each culture transduced by incubation with an LXSN-type retroviral vector carrying a gene for human prourokinase and a neomycin resistance gene for selection in antibiotic G418. Control grafts had passive luminal coating with fibronectin and were seeded with transduced ECs immediately after G418 selection; these grafts were incubated for 2 days before implantation. Experimental grafts had fibronectin forcefully squeezed through the interstices and were seeded with ECs that had recovered in culture for 5 days after G418 selection; these grafts were incubated for 4 days before implantation. For each control (n = 9) and experimental (n = 12) graft, a graft prepared in the same fashion but seeded with the remaining autologous nontransduced cells was placed in the contralateral carotid artery. Grafts were explanted after 30 days and were evaluated for patency, thrombus-free surface area, and cell-free surface area. RESULTS: No significant differences in patency rates were seen between any groups. The thrombus-free surface area was improved for experimental grafts (90%) compared with control grafts (76%), but this improvement did not achieve statistical significance. The cell-free surface area for transduced cells on experimental grafts was 65% compared with 96% for control grafts (p = 0.021) and was comparable with that for nontransduced cells on both control grafts (62%) and experimental grafts (51%; p = 0.201). CONCLUSIONS: Adherence of genetically modified endothelial cells to small-diameter expanded polytetrafluoroethylene grafts in an in vivo physiologic flow model is significantly improved when cells have a more prolonged recovery from G418 selection, when the graft lumen is more uniformly coated with fibronectin before EC seeding, and when seeded grafts are left longer in culture before implantation to develop cell lining stability. The short-term patency rate of these seeded grafts is not affected by increased cell retention; long-term graft patency data and luminal healing require further evaluation.

Birthweight prediction by three-dimensional ultrasonographic volumes of the fetal thigh and abdomen.

Our validation study examined a three-dimensional ultrasonographic phantom that contained irregularly shaped volume targets ranging from 0.5 to 76.1 milliliters. Four different examiners made blinded measurements from volume datasets that were acquired by 4 and 7 MHz transducers. Birthweight predictions using abdominal and thigh volumes from 18 term fetuses also were compared with two-dimensional ultrasonographic methods. In vitro volume measurements were accurate, precise, and repeatable despite a small systematic overestimation with increasing object size. Mean systematic error and precision for birthweight predictions by three-dimensional ultrasonography (-0.03% +/- 6.1%) were not significantly different from those by two-dimensional ultrasonography (-0.60% +/- 8.8%). Conventional prediction methods yielded three birthweights with greater than 15% error. By comparison, except for one infant whose birthweight indicated an 11% error, all predictions based on fetal volume parameters were within 10% of true values. Accurate birthweight predictions by fetal volume parameters appear to be technically feasible at term gestation although their practical clinical application requires further investigation. Birthweight predictions in this manner may allow remote consultants to evaluate the fetus over wide-area computer networks despite the physical absence of the patient.

Nonoperative management of adult blunt splenic trauma: a 15-year experience.

From January 1989 to December 1993, 40 consecutive adult patients with ruptured spleen from blunt trauma were examined. Fourteen patients (35%) were taken to the operating room initially because of hemodynamic instability and generalized peritoneal signs. Twenty-six patients (65%) were hemodynamically stabilized at admission and treated by nonoperative management, which included strict bed rest, intensive care unit monitoring, frequent physical examinations, and serial hematocrits. Four patients failed nonsurgical management and required a splenectomy, three because of clinical deterioration within 1 to 3 days of admission; the fourth patient had recurrent bleeding 7 days after injury. The patients in the operative group had a greater severity of injury with a mean injury severity score of 26.6, four deaths, and mean transfusion requirements of 3.7 to 4.0 units of blood, compared to a mean injury severity score of 14.6, one late death from cardiac causes, and average blood requirement of 0.4 to 0.7 units. Splenic injury grading averaged 3.2 in the surgical group (grade 1, one patient; grade 2, four patients; grade 3, eight patients; grade 4, no patients; and grade 5, one patient) and differed significantly from that of the nonoperative group (mean = 2.4; grade 1, 12 patients; grade 2, seven patients; grade 3, six patients; grade 4, two patients; and grade 5, no patients). Recent ultrasound analysis of select grades I to IV has shown excellent resolution or repair of these injuries. This report extends our series from 1978 to 1993 and includes 144 adult patients sustaining blunt splenic ruptures. Seventy-nine (55%) of these patients were treated nonsurgically. Seven patients (of 80) failed nonoperative management and required interval laparotomy, representing a 91 per cent success rate. Follow-up on more than 90 per cent of the patients has shown no sequelae from their splenic injuries. We conclude that adult patients with splenic injuries from blunt trauma who are hemodynamically stable and are without abdominal findings requiring celiotomy can be safely managed by a nonoperative approach.

Serial duplex ultrasound examinations for deep vein thrombosis in patients with suspected pulmonary embolism.

PURPOSE: We have prospectively evaluated the need for serial venous duplex ultrasound examinations in an inpatient population with an initially normal study result. METHODS: Patients were selected for study on the basis of clinical suspicion of pulmonary embolism and possible lower extremity deep vein thrombosis, a comorbid condition contributing to a nondiagnostic ventilation/perfusion lung scan, and an initially normal bilateral venous duplex ultrasound examination that included complete evaluation of the femoropopliteal system and the deep calf veins. Repeat duplex examinations were done during the same hospital admission between 5 and 14 days after the initial study. RESULTS: Ninety-four patients with an initially normal duplex ultrasound examination result had repeat studies done at an average of 7.9 +/- 2.6 days. Ninety-two examination results remained normal bilaterally. Two patients had isolated intramuscular calf vein deep vein thrombosis: one in the gastrocnemius system of both calves with associated calf tenderness at 11-day follow-up and one in a mid-calf soleal vein without associated symptoms at 10 days. No patients had any evidence of deep vein thrombosis in the femoropopliteal or tibioperoneal venous systems. CONCLUSIONS: Serial follow-up duplex ultrasound evaluation is unnecessary after an initially complete, normal study in patients with symptoms who have suspected pulmonary embolism and nondiagnostic ventilation-perfusion lung scans.

Progression of superficial venous thrombosis to deep vein thrombosis.

PURPOSE: We have evaluated the progression of isolated superficial venous thrombosis to deep vein thrombosis in patients with no initial deep venous involvement. METHODS: Patients with thrombosis isolated to the superficial veins with no evidence of deep venous involvement by duplex ultrasound examination were evaluated by follow-up duplex ultrasonography to determine the incidence of disease progression into the deep veins of the lower extremities. Initial and follow-up duplex scans evaluated the femoropopliteal and deep calf veins in their entirety; follow-up studies were done at an average of 6.3 days, ranging from 2 to 10 days. RESULTS: From January 1992 to January 1996, 263 patients were identified with isolated superficial venous thrombosis. Thirty (11%) patients had documented progression to deep venous involvement. The most common site of deep vein involvement was progression of disease from the greater saphenous vein in the thigh into the common femoral vein (21 patients, 70%), with 18 of these extensions noted to be nonocclusive and 12 having a free-floating component. Three patients had extended above-knee saphenous vein thrombi through thigh perforators to occlude the femoral vein in the thigh, three patients had extended below-knee saphenous disease into the popliteal vein, and three patients had extended below-knee thrombi into the tibioperoneal veins with calf perforators. At the time of the follow-up examination all 30 patients were being treated without anticoagulation. CONCLUSIONS: Proximal saphenous vein thrombosis should be treated with anticoagulation or at least followed by serial duplex ultrasound evaluation so that definitive therapy may be initiated, if progression is noted. More distal superficial venous thrombosis should be carefully followed clinically and repeat duplex ultrasound scans performed, if progression is noted or patient symptoms worsen.

Appropriate indications for venous duplex ultrasonographic examinations.

BACKGROUND: Because of the rapid growth in requests for lower extremity venous duplex ultrasonographic examinations, we have evaluated test results to determine the appropriate indications for testing and whether there is overuse of this technique. METHODS: We reviewed the records of all patients who had duplex ultrasonogram during an 18-month period to rule out deep vein thrombosis (DVT). Test outcome was evaluated with data on the age, gender, outpatient or inpatient status, duration of symptoms, physical findings, and risk factors of the patients. RESULTS: We reviewed 3474 examinations; 1265 outpatients and 1231 inpatients were evaluated to rule out DVT, and 978 patients were evaluated to rule out a source of pulmonary embolism. More outpatients were female, and they tended to be younger; outpatients also had more acute femoropopliteal DVT. The presence of sudden onset of unilateral swelling was a strong predictor of acute DVT in all three groups: 52% of such patients were found to have DVT. Mild unilateral symptoms with a coexistent risk factor was associated with a 10% to 20% prevalence of acute DVT. Chronic unilateral swelling was associated with acute DVT only with a coexistent risk factor and only in 1% of such patients. Bilateral swelling was only associated with acute DVT in the inpatient population, with 17% of studies having positive results. If calf tenderness was the only symptom with no associated risk factors, seven (1.6%) of 415 patients had acute DVT and six of these were tibioperoneal. No patients with cellulitis or isolated joint pain had acute DVT. With clinically suspected pulmonary embolism in patients without associated symptoms or risk factors, 7 (1.5%) of 447 studies showed tibioperoneal disease with no acute femoropopliteal disease. CONCLUSIONS: The physical examination and clinical history of the patient remain important indicators of the presence (or absence) of lower extremity DVT and can guide the appropriate use of duplex ultrasonography for the diagnosis.

Urokinase expression in transduced endothelial cells.

Increased thromboresistance through the release of lytic agents by endothelial cells may improve the patency of endothelial lined prosthetic grafts. We have evaluated the expression of urokinase from cells transduced with a retrovirus containing the gene for a human preprourokinase. Endothelial cells were enzymatically harvested from canine external jugular vein in nine animals and grown to confluence in culture. One-third of these cells served as controls, and the remaining two-thirds were transduced via incubation with an LXSN-type retroviral vector carrying the urokinase gene and a neomycin resistance gene. Successfully transduced cells were selected by incubation with 400 micrograms/mL G418 and pure cultures grown to confluence. Supernatants from confluent control and experimental cell cultures after 48 hours in defined, serum-free medium were assayed for human urokinase concentration and overall enzyme activity. ELISA quantitation of concentration using mouse antihuman urokinase antibody showed 0.15 +/- 0.11 ng/mL/hr/10(6) cells in the transduced cell supernatant; no measurable concentration was found in the control cells. (P < 0.01) Overall (human plus canine) enzyme activity of urokinase was determined using an indirect spectrophotometric assay based on plasminogen activation (ploug U/mL). Transduced cells showed activities of 0.12 at 10 days and 0.45 at confluence; control cell activity was 0.0 and 0.15, respectively. (P < 0.05) These data show that endothelial cells can be transduced with a urokinase expressing gene that increases the release of this thrombolytic agent. Lining small diameter prosthetic grafts with these cells may improve their thromboresistance and long-term patency.

Colorectal complications following cardiac surgery. Six-year experience.

PURPOSE: This study was undertaken to assess colorectal complications following cardiopulmonary bypass surgery. METHOD: This is a retrospective review of 5,801 patients who underwent 5,801 cardiopulmonary bypass procedures from 1985 to 1991. Patients were evaluated for type of bypass procedure, postoperative colorectal complications, age, sex, bypass time, aortic cross-clamp time, elective vs. emergency cases, uses of intra-aortic balloon pump, perioperative hypotension, and use of vasopressors. Statistical analysis was performed using chi-squared analysis and Student's t-test. RESULTS: Nineteen of the 5,801 patients developed 19 colorectal complications, a prevalence of 0.3 percent for the initial hospital stay following bypass surgery. Mortality in those with complications was 37 percent (7/19). Of the 19 complications, 9 (47 percent) followed coronary artery bypass grafting, whereas 10 (53 percent) followed valve replacement or combined coronary artery bypass grafting with other cardiac procedures. Five (26 percent) of the complications followed emergency cases, whereas 14 (74 percent) followed elective cases. Average age of those with complications was 69.8 years, compared with 63.2 years for those without complications. Average aortic cross-clamp time for those with complications was 71 +/- 25 minutes; pump time was 106 +/- 34 minutes. That was significantly higher than in those without complications. Nine (47 percent) patients with complications required vasopressors during the perioperative period, whereas eight (42 percent) suffered prolonged hypotension (systolic blood pressure, < 90 mmHg). CONCLUSIONS: It appears that increased age, valve replacement, or combined cardiac procedures, emergency procedures, and prolonged aortic cross-clamp and bypass pump times are risk factors for development of colorectal complications. Hypoperfusion, as suggested by prolonged pump times, clamp times, and emergency procedures may be a possible cause for development of colorectal complications.