Unicondylar knee arthroplasty: a cementless perspective.

OBJECTIVE: To compare the results of cementless unicondylar knee arthroplasty (UKA) with those already reported in a similar study on cemented UKA. DESIGN: A case-series cross-sectional study. SETTING: The Queen Elizabeth II Health Sciences Centre, Dalhousie University, Halifax. PATIENTS: Fifty-one patients who underwent a total of 57 UKAs between May 1989 and May 1997. Inclusion criteria were osteoarthritis involving the predominantly the medial compartment of the knee, relative sparing of the other compartments, less than 15 degrees of varus, minimal knee instability, and attendance at the postoperative clinical visit. INTERVENTION: Cementless UKA. MAIN OUTCOME MEASURES: Clinical parameters that included pain, range of motion and the Knee Society Clinical Knee Score. Roentgenographic parameters that included alpha, beta, gamma and sigma angles and the presence of periprosthetic radiolucency or loose beads. RESULTS: Age, weight, gender and follow-up interval did not significantly affect the clinical results in terms of pain, range of motion or knee score. Knees with more than 1 mm of radiolucency had significantly lower knee scores than those with no radiolucency. Knees that radiologically had loose beads also had significantly lower knee scores. The clinical outcomes of cementless UKA were comparable to those already reported on cemented UKA. Cementless femurs had less radiolucency than the cemented femurs, whereas cementless tibias had more radiolucency than their cemented counterparts. CONCLUSIONS: Cementless UKA seems to be as efficacious as cemented UKA. However, there is some concern about the amount of radiolucency in the cementless tibial components. A randomized clinical trial comparing both cementless and cemented tibial components with a cementless femur (hybrid knee) is needed to further assess this controversial issue in UKA.

In vitro analysis of periprosthetic strains following total knee arthroplasty.

Clinical case studies have disclosed certain risk factors associated with periprosthetic fracture in elderly patients. How the mechanical strength of the distal femur is changed by total knee arthroplasty (TKA) has not been elucidated. Using elderly cadaveric femora, this study evaluated both periprosthetic strains and associated fracture patterns arising from an in vitro simulation of a fall onto the distal femur. The data showed a significant increase in anterior and posterior mechanical strain following TKA. Neither stemless nor stemmed versions of two cemented Howmedica prostheses (Rutherford, NJ) reduced distal femur strains to baseline values. However, neither produced a disproportionate frequency of periprosthetic fractures. Although not formally evaluated herein, bone geometry/density may contribute more profoundly to the occurrence of periprosthetic fracture than the implants tested.

Graft dilatation following abdominal aortic aneurysm resection and grafting.

BACKGROUND: It has been suggested that graft dilatation following repair of abdominal aortic aneurysm (AAA) is associated with complications such as anastomotic aneurysm and graft rupture. The purpose of the present study was to document the degree of dilatation observed in grafts after aneurysm repair and to correlate this with any graft-related complications. METHODS: Between January 1987 and December 1992, 219 patients had elective repair of their AAA at St George Hospital. A follow-up ultrasound scan was available for 154 of these patients. The following factors were examined: age, sex, size of aneurysm, type and size of graft, time of follow-up scan, size of graft at follow-up and any graft-related complications. RESULTS: The mean graft dilatation observed in knitted grafts (42.6%; 95% CI: 39.1-46.1%) was significantly greater than that observed for woven grafts (25.5%; 95% CI: 19.0-32.1%; P < 0.0001). There were no graft-related complications. CONCLUSIONS: Graft dilatation is a predictable phenomenon following AAA repair. It is more pronounced in knitted than in woven grafts, but does not necessarily lead to graft-related complications or failure.

The cost of elective and emergency repair of AAA in patients under and over the age of 80.

BACKGROUND: As Australia's population ages, the number of elderly patients presenting for surgery of abdominal aortic aneurysms (AAA), both elective and ruptured, will increase. The aim of the present study was to compare the costs of treatment of patients with AAA, under and over the age of 80, in the elective and emergency settings in a hospital with a divisional structure in which the true costs can be accurately obtained. METHODS: A total of 40 patients were selected at random from a series of 267 patients treated with open surgery for AAA between January 1987 and December 1994, 10 in each of four groups: group A, elective repair in patients aged < 80 (171/267); group B, elective AAA repair in patients aged > 80 (25/267); group C, emergency AAA repair in patients aged < 80 (50/267); and group D, emergency AAA repair in patients aged > 80 (11/267). A retrospective analysis of the hospital costs of treatment of these patients at St George Hospital was conducted. These true costs were then compared to Australian National Diagnostic Related Group (AN-DRG) costs. RESULTS: Group A and B had no mortality. In Group C and D the mortality was 20 and 60%, respectively. The emergency treatment groups also had longer lengths of stay. A statistically significant difference in cost of AAA repair between elective and emergency groups in both age groups was seen; that is, group A cost less than group C and group B cost less than group D. Costs per survivor, however, showed a dramatic difference between the cost of group C patients ($30000) and group D patients ($60000). In comparison with AN-DRG calculated costs, the true costs of groups A and B were equivalent to AN-DRG costs. In the emergency groups, however, there were marked discrepancies between the true cost ($61000) and that calculated by the DRG ($25000) in group D, with similar differences seen in group C to a lesser extent. CONCLUSION: Emergency repair of AAA is significantly more expensive and has a high mortality in the over-80 age group. Also, there is a substantial shortfall between the true costs of treating these patients and the funds allocated for treatment in this group.

Expansion rates of small abdominal aortic aneurysms.

BACKGROUND: The present study was carried out in order to examine those factors that influence the rate of expansion of small abdominal aortic aneurysms. METHODS: A retrospective study was undertaken of 112 patients who attended the St George Vascular Laboratory between 1987 and 1997. These patients had abdominal aortic aneurysms that were considered to be too small to warrant surgical repair at the time of presentation. Sequential ultrasound examinations were used to measure maximal anteroposterior aneurysm diameter. From these data, annual growth rates were calculated. Growth rate per annum was then compared with gender, age, initial aortic aneurysm diameter, presence of hypertensive disease, cardiac disease, family history of aneurysmal disease, diabetes mellitus, smoking, beta-adrenergic blockade and lipid lowering drugs. RESULTS: Univariate analysis showed that three factors were significantly related to growth rate: the initial size of the aortic aneurysm, the presence of cardiac disease and the presence of beta-adrenergic blockade. CONCLUSIONS: The presence of beta-adrenergic blockade appeared to have an independent effect on aneurysm growth rate, and suggests a possible role for beta-adrenergic blockade as a therapeutic strategy in controlling expansion rates of small abdominal aortic aneurysms. A controlled double-blind clinical trial is required to demonstrate this conclusively.

Treatment of abdominal aortic aneurysm disease in the 9th and 10th decades of life.

BACKGROUND: The appropriate management of patients who are older than 80 years of age and who present with an abdominal aortic aneurysm (AAA) remains controversial. While it appears that elective repair can be performed safely, appropriate management of these patients in the emergency situation is unclear. The purpose of the present study was to examine the results obtained in treating this elderly group in the elective and emergency setting, by operation and conservative techniques at St George Hospital, Kogarah. METHODS: Between January 1987 and December 1994 85 patients older than 80 years of age were treated for AAA. These patients were divided into four groups: I, elective presentation/no surgery; II, elective presentation/elective surgical repair; III, emergency presentation/surgical repair; and IV, emergency presentation/conservative treatment. We examined age, sex, size of AAA, mode of presentation, type of treatment, length of survival and cause of death. RESULTS: The mean age of the total group (n = 85) of patients was 84 years (range: 80-94). The mean AAA diameter for this group was 5.6 cm (95% CI: 5.2-6 cm). The diameters for group I (n = 40), II (n = 22), III (n = 16) and IV (n = 7) were 4.9 cm (4.4-5.5, 95% CI), 5.7 (4.9-6.5 CI), 7.0 (6.1-7.7 CI) and 6.2 (5.2-7.2 CI), respectively. The median survival for groups I, II, III and IV was 18, 38.5, 0.25 and 0 months, respectively. Group II had a longer survival than any other group (P = 0.015), and group IV had a shorter survival than the total group (P = 0.001). However, the length of survival was no different for III versus IV (P = 0.146). Deaths in each group were due to the following reasons. I: cardiopulmonary events (14), rupture (3), malignancy/sepsis (3); II: cardiopulmonary events (3), rupture (thoracic aneurysm) (2), malignancy (I); III: rupture (10), malignancy (I); and (IV): rupture (6), malignancy (1). CONCLUSIONS: Elective surgical repair offers the best management option for AAA in patients older than 80 years of age. Death may still occur from progression of aneurysmal disease at other sites. An aggressive surgical approach to the management of haemodynamically unstable patients in this age group is of questionable benefit.

Outcome for patients with abdominal aortic aneurysms that are treated non-surgically.

BACKGROUND: There are few reports in the literature describing the outcome for patients with abdominal aortic aneurysm who are not treated by surgical repair. This is in spite of the fact that this group of patients often defines the success of surgical treatment. The purpose of this report is to review those patients from St George Hospital Kogarah who have been rejected for surgical therapy and to examine the long-term outcome and mode of demise of these patients. METHODS: At the end of December 1992 we completed a computerized list of all patients seen at St George Hospital Kogarah with abdominal aortic aneurysm. Since that time we have continued to accrue patients to this list and obtain follow-ups prospectively. End points examined in this study were aortic aneurysm transverse diameter, sex, age, intercurrent illnesses, reasons for not undertaking surgical treatment, length of survival and cause of death. RESULTS: The mean age of patients in this series was 77 +/- 8.29 years (SD). Survival at yearly intervals for 5 years in our 101 patients were 69, 55, 44, 35 and 33%, respectively. For patients with an abdominal aortic aneurysm of < or = 5 cm, the 5-year survival rate was 42% while for patients with an abdominal aortic aneurysm of > 5 cm, the 5-year survival rate was 25%. There were 66 deaths in this series; 40% were due to cardiopulmonary events and 30% were due to ruptured abdominal aortic aneurysm. The median time between presentation and death was 12 months. CONCLUSION: The patients from this report were significantly older than those reported from previous series. We believe that it will become increasingly important to develop methods of selection of patients for current and newer modalities of treatment for abdominal aortic aneurysm. It will be increasingly important for groups offering these treatments to be able to explain to patients what their prognosis is likely to be should they not be selected for those treatments.

Duplex scanning for recurrent varicose veins.

BACKGROUND: Most studies of recurrent varicose veins have been based on the findings clinically, at operation, or on phlebography. Occasionally these findings have been compared with Doppler findings produced by hand-held continuous-wave Doppler. Duplex scanning is a more refined approach to the assessment of recurrent varicose veins. METHODS: 202 patients (267 legs) have been examined consecutively for recurrent varicose veins between January 1990 and December 1995 at St George Vascular Laboratory. This was a retrospective study of some aspects and patterns of recurrence of varicose veins in this group. RESULTS: The ratio of female to male was 3:1. The mean age of the group was 52 years and mean time to recurrence was 13 years (1 year 95% CI). There were six patterns of recurrence accounting for 95.2% of legs. In descending order of frequency, these were: (i) the saphenofemoral junction and long saphenous vein were intact and incompetent (44.6%); (ii) an incompetent thigh perforator and long saphenous vein remained intact; there was no saphenofemoral junction (16.5%); (iii) the long saphenous vein remained intact and incompetent; there was no saphenofemoral junction (10.5%); (iv) there was an incompetent saphenofemoral junction only (9.74%); (v) there was an intact and incompetent saphenopopliteal junction alone (9.74%); and (vi) an isolated thigh perforator was incompetent, with no more proximal site of incompetence detected (4.12%). Incompetent calf perforator (69.2%) and gastrocnemius veins (9.3%) were frequently detected, but rarely existed in isolation (seven legs in total). CONCLUSIONS: Duplex scanning is an important recent adjunct to the management of recurrent varicose veins in order to define the pathway of incompetence'. The saphenofemoral junction and long saphenous vein remain the key to recurrence of varicose veins. Calf perforator and gastrocnemius vein incompetence are of secondary importance in recurrent varicose veins.

An assessment of tonometry and regional splanchnic blood flow during aortic cross-clamping in the pig.

We aimed to evaluate the tonometer in the assessment of gastrointestinal ischaemia induced by an infrarenal aortic cross-clamp. Nine anaesthetized pigs were cannulated for haemodynamic monitoring and radionuclide labelled microsphere (RLM) injection. Gastric and sigmoid tonometers were positioned. After haemodynamic stabilization an infrarenal aortic cross-clamp was applied. Animals were sacrificed at the completion of the study and tissue sampled from the stomach and sigmoid colon for regional blood flow measurements. Measurements were made pre-clamp, post-clamp, pre-release and post-release. Haemodynamic parameters, gastric intramucosal pH (pHi) and blood flow did not change throughout the experiment. Arterial pH increased during cross-clamp and returned to baseline post-release. Arterial bicarbonate fell post release. Sigmoid blood flow fell during cross-clamp. The sigmoid pHi fall, delayed until pre-release, remained low post-release. Although there was a consistent fall in sigmoid pHi, 63% of post-clamp values remained within the baseline range. We conclude that maintaining haemodynamic parameters around baseline values resulted in maintenance of gastric mucosal perfusion as indicated by a steady gastric pHi. However, below the aortic cross-clamp, delay between change in sigmoid colon blood flow and change in pHi and wide variation in sigmoid pHi limits the value of an individual pHi measurement in detecting ischaemia.

Sigmoid and gastric tonometry during infrarenal aortic aneurysm repair.

BACKGROUND: The value of the sigmoid tonometer in predicting sigmoid ischaemia and postoperative enteric organism infection has been reported but the value of tonometric measurement has been challenged. The purpose of this study was to examine the use of tonometric measurements in a series of patients undergoing infrarenal aortic aneurysm repair. METHODS: We assessed the results obtained when sigmoid (n=11) and gastric (n=8) tonometry were performed in patients undergoing infrarenal aortic aneurysm repair (n=11). We measured blood flow ultrasonically (n=6) in the inferior mesenteric artery(IMA) and IMA stump pressures. Sigmoid and gastric tonometry were measured prior to clamping of the infrarenal aorta, during cross clamping and after clamp release at 1, 4, 16 and 20h. Ultrasonic flow was measured before clamping. Stump pressures in the IMA were measured before, during and after clamping. RESULTS: The IMA was chronically occluded in five patients. The IMA flow was 37.5 +/- 8.7 mL/min (mean +/- s.e.). The mean IMA stump pressures before, during and after clamping were 64 +/- 13, 48 +/- 8 and 69 +/- 10 mmHg, respectively, and did not differ significantly. Mean systematic arterial pressures at these times were 89 +/- 7, 95 +/- 5 and 86 +/- 8 mmHg. These did not differ significantly or when compared with IMA stump pressure. The gradient between systemic arterial pressure and IMA stump pressure did not vary significantly at any of these times. Sigmoid and gastric intramucosal pH (pHi) did not differ significantly at any of the above times. Both sigmoid gastric pHi dropped on clamp application but 4 h afterwards had returned to baseline levels. Systemic arterial pH reflected significant ischaemia during clamping and shortly after release of the clamp(P=0.008). CONCLUSIONS: Tonometry may reflect systemic events as much as regional ischaemia. Useful tonometry results may depend on the development of a trend rather that individual measurements. The routine use of tonometry to detect intestinal ischaemia may not be cost-effective in aortic surgery.

Cryotherapy treatment of patients with hepatic metastases from neuroendocrine tumors.

BACKGROUND: Liver metastases from neuroendocrine tumors often present with disabling symptoms due to syndromes of hormonal excess. A locally destructive technique such as hepatic cryotherapy not only alleviates symptoms but may improve survival in this group of patients. METHODS: Six patients with metastatic neuroendocrine tumors were treated with hepatic cryotherapy. Four patients were symptomatic and three of these had elevated tumor markers from ectopic hormone production. RESULTS: All patients are alive and asymptomatic, with a median follow-up of 24 months (range, 6 months to 6 years). All have had a complete radiologic response. All with elevated preoperative markers have had a greater than 89% decrease in tumor markers. Coagulopathy occurred in two patients necessitating additional surgery, but there was no other morbidity attributable to the cryotherapy. CONCLUSION: To the authors' knowledge, this study demonstrates for the first time that hepatic cryotherapy offers supportive treatment for patients with neuroendocrine tumors metastatic to the liver. Cryotherapy alleviates symptoms and may improve survival.

The effect of infrarenal aortic cross clamping and declamping on renal function in the pig.

Renal insufficiency following periods of infrarenal aortic cross clamping has been reported by some investigators but not by others, and conflicting views have been expressed concerning the ability of renal autoregulation to overcome the adverse circulatory effects of cross clamping. The object of this study was to examine the blood flow distribution to four layers within the renal cortex (subcapsular to juxtamedullary) and measure global renal function following application and release of an aortic cross clamp after 90 min. Nine juvenile female pigs weighing 25 to 50 kg were anaesthetized and subjected to intensive physiological monitoring. Throughout the study the blood pressure and cardiac output were maintained as close as possible to control levels by fluid administration and varying the depth of anaesthesia. Renal cortical blood flow was estimated by means of radionuclide labelled microspheres and global renal function was determined by the measurement of creatinine clearance. The aortic cross clamp was applied for 90 min immediately distal to the renal arteries and proximal to the inferior mesenteric artery. Cardiovascular and renal parameters were recorded on four occasions during each experiment, prior to, 10 and 60 min after cross clamping, and 30 min after clamp release. No significant changes in cardiac output, systemic blood pressure of global renal function were recorded during the study. There was, however, a significant fall in renal blood flow following release of the aortic cross clamp but this was not associated with any significant redistribution of blood flow within the renal cortex. In the pigs studied, the application of an infrarenal aortic cross clamp did not have any adverse effects on the cardiovascular system or on global renal function.(ABSTRACT TRUNCATED AT 250 WORDS)

Incidence of rupture of aortic aneurysms after coincidental operation.

The clinical course of 76 patients with aortic aneurysmal disease undergoing 107 coincidental surgical procedures was analysed in order to examine the relationship between aortic aneurysmal rupture and coincidental treatment. Additionally the incidence of aneurysmal rupture was assessed following 82 endoscopic procedures in 42 patients with aortic aneurysms. Two patients ruptured an aortic aneurysm after operation, one after colonoscopy (maximal transverse diameter 7 cm) and one after coronary artery bypass grafting (maximal transverse diameter 5.6 cm). The mean maximal transverse diameter of aneurysms in 76 patients was 5.08 cm (95% confidence interval 4.7-5.4 cm). Both patients with ruptured aortic aneurysm were outside these confidence limits and were known hypertensives whose perioperative control of hypertension was questionable. The present series of patients is discussed with reference to induction of collagenase activity as a precipitating cause for postoperative rupture of aortic aneurysms, perioperative control of hypertension, transverse aneurysm diameter as a predictor of postoperative rupture and conduct of coincidental procedures in the presence of aneurysmal disease.

Aspects of central venous access catheter usage in patients with malignancy.

AIMS: A review of 5 years' experience with central venous access catheters (CVAC) to examine the results of catheter placement technique (open or percutaneous) on subsequent CVAC performance. In addition the catheter function and utilisation of both Hickman type and Access Port type of CVAC were reviewed to assess the indications and the types of access devices chosen in selected patients. METHODS: A retrospective review of CVAC between 1987 and 1991 was undertaken. During this time 113 CVAC were placed by the open (n = 76) or percutaneous (n = 37) method. The type of catheter used was Hickman (n = 79) or Access Port (n = 34). Indications for CVAC placement were haematological malignancy 74.2%, other malignancy 14.1%, and the rest were accounted for by TPN, HIV infection, osteomyelitis, anaemia and haemochromatosis. RESULTS: There was no difference between duration of function for CVAC placed by the open (5.1 months 95% CI, 3.8-6.4) and percutaneous methods (3 months, 95% CI 2.2-3.8). Nor was there a difference in duration of function between Hickmans (4.47 months, 95% CI 3.3-5.6) and Access Ports (4.3 months, 95% CI 2.8-5.8). CVAC morbidity included sepsis, accidental displacement, major venous thrombosis, bleeding and port erosion. There was no significant statistical difference in the incidence of complications between the open (25%) versus percutaneous (22%) techniques. The clinically significant difference in morbidity related to the type of catheter used: Hickman (18%) versus Access Port (35%). Infection and sepsis were a problem in the Port group accounting for 20.6% of the complications. CONCLUSIONS: We believe CVAC account for significant morbidity which is not often taken into account by those recommending their use, particularly in the management of malignant disease. In addition CVAC placement should be part of a service in consultation. There are cost-benefit advantages in using orthodox central lines in the sick patient with a view to placing a CVAC at a later date.

Renal angioplasty for renovascular disease: a reappraisal.

The exact role of percutaneous transluminal angioplasty (PTA) for the treatment of renal artery stenoses (RAS) remains unclear. Therefore we analysed the outcome of unselected renal artery PTA in 30 subjects with angiographically proven unilateral RAS (Group I) compared with that of 19 who had RAS but were treated medically (Group II) and 21 who had essential hypertension but who underwent angiography (Group III). Good blood pressure control was achieved in all three groups but those who underwent PTA had a small but significant fall in their requirement for antihypertensive medications (p less than 0.05). However, no patient was "cured" of hypertension and blood pressure was able to be controlled in both Group II and Group III subjects without an increase in their number of antihypertensives. Serum creatinine did not improve significantly during follow-up in those who underwent PTA and was not different from that of Group II subjects at follow-up. Although unselected renal artery PTA for RAS may make blood pressure control easier it does not delay or prevent further deterioration in renal function and should not be used for this purpose.

Survival following perforation of peptic ulcer.

Between 1979 and 1989, 92 patients were treated at St George Hospital, Kogarah, for perforated duodenal or prepyloric ulcer. Mortality rate at 28 days was 18%. Life table analysis showed 1-, 5- and 10-year survival rates for this group to be 78%, 60% and 46% respectively. During the second half of the study period, increasingly older females with a history of cardiovascular disease, arthropathy, chronic renal impairment and non-steroidal anti-inflammatory drugs (NSAIDS) intake were identified and found to be at greater risk of dying from their perforation. Age, cardiovascular disease and chronic renal impairment were demonstrated to be independent factors affecting survival. Patients treated by simple closure of the perforation had a long-term survival rate equivalent to that of patients treated in other ways, although the number of these latter patients is small. Implications for the administration of NSAIDS are considered in the light of these findings. These results suggest that orthodox simple closure of perforated peptic ulcer and administration of H2 blocking agents is the most appropriate treatment for patients presenting with perforated peptic ulcer.

Is the outlook for the vascular amputee improved by striving to preserve the knee?

A trend to amputate below knee (BK) began in the late 1960's when the disadvantages of above knee (AK) amputation were recognised. In this study, the outcome of 189 consecutive patients who had major lower extremity amputations between 1978-1982 was compared to earlier reports from the same institution. Their cumulative survival of 52% at three years, was similar to the cumulative survival of 116 amputees whose surgery was done in 1966-1971. The risk of losing the second limb, almost 10% per year, was also similar to the earlier experience of 1966-1971. In 1964 one BK amputation was performed for every six above the knee. By 1980 this ratio had reversed to three BK for each AK amputation. When a ratio of BK:AK amputation greater than 2:1 was achieved in our patient population, using clinical criteria as the sole guide to amputation levels, one in four failed. The eventual ratio of healed BK to AK amputation achieved was little better than unity. A trend to below knee amputation was not associated with improvement of long term survival after lower extremity amputation for advanced arterial disease. These results indicate a need for better care of the vascular amputee and for a test to compliment clinical selection of amputation levels.

Blunt trauma to the internal carotid artery.

Blunt injury of the internal carotid artery is rare and easily overlooked. The injury should be considered in a conscious patient with dense neurological deficit after blunt trauma to the head and neck. The diagnosis was established in 17 patients (9 men, 8 women) by arteriography. There was a median delay of 19 hours in the onset of neurologic deficit in 15 patients. Follow up arteriography was obtained in 16 of the 17 patients with a median interval of three months between arteriograms. On repeat arteriography, the internal carotid artery was patent in three of the nine patients with internal carotid arteries occluded on initial arteriography. Surgical repair was attempted in six patients. No significant difference in long term neurological deficit occurred between patients treated conservatively and those treated operatively. Of the 15 patients with hemiparesis on presentation, eight made a complete recovery and six improved. This study supports nonoperative management for blunt injuries of the internal carotid artery.

Femoral false aneurysms: a recommendation for the method of repair.

False aneurysm formation is a major complication of vascular surgery. The most frequent site of anastomotic false aneurysm formation is the femoral artery. Between January 1974 and June 1986, 26 patients with 42 femoral false aneurysms were treated at the Princess Alexandra Hospital. Aneurysms developed following Dacron arterial grafting (29 aneurysms), saphenous vein grafting (10 aneurysms), umbilical vein grafting (one aneurysm) and femoral embolectomy (two aneurysms). Arterial wall failure (with intact suture and graft) was the most frequent operative finding. Ten recurrent aneurysms developed. There was a significantly greater number of recurrences when resuture or patch repair was employed than when an interposition graft was used as a repair. The development of a femoral anastomotic false aneurysm should be viewed as a total failure of that anastomosis and repair should be by replacement with an interposition graft rather than repair of the failed anastomosis by suture or patch.