Outcome after combined carotid endarterectomy and coronary artery bypass is related to patient selection.

OBJECTIVE: The optimal management of patients with significant coronary and carotid artery disease remains controversial. Since reporting on a series of 100 patients undergoing combined carotid endarterectomy and coronary artery bypass (CEA/CAB) 4 years ago, we have liberalized our selection criteria for combined operation. We sought to compare outcomes of the recent cohort of 74 patients and the previous group. METHODS: All patients who underwent CEA/CAB since 1984 have been tracked in a database containing identifying information, demographic factors, anatomic information, details of surgery, and short- and long-term follow-up data. We compared the 74 patients (Group 2) undergoing CEA/CAB since 1994 with the previously reported group of 100 patients (Group 1) who underwent CEA/CAB between 1984 and 1994. We examined demographic and comorbidity factors, presence of cerebrovascular symptoms, degree of contralateral carotid stenosis, and perioperative stroke and death. Statistical comparisons were made with the chi(2) test. RESULTS: The groups had similar age and sex distributions and similar incidences of hypertension, diabetes, congestive heart failure, prior myocardial infarction, and hypercholesterolemia. More patients in Group 1 had preoperative transient cerebral ischemia or monocular blindness (55% vs 31%, P <.002) and preoperative stroke (18% vs 7%, P <.03). More patients in Group 2 had unilateral asymptomatic carotid artery stenosis (55% vs 18%, P <.001). The incidence of all perioperative strokes was higher in Group 1 (9% vs 1.4%, P <.035). There were fewer deaths (3% vs 8%) and ipsilateral strokes (0 vs 4%) in Group 2, though these were not statistically significant. CONCLUSION: We have liberalized our criteria for performing combined CEA/CAB, such that more than 50% of our recent patients have asymptomatic unilateral carotid stenosis. This practice is associated with a lower incidence of all perioperative strokes and a trend toward lower ipsilateral stroke and death. These observations suggest that perioperative stroke after CEA/CAB is related to patient selection and that low-risk patients can undergo CEA/CAB with the benefits of low morbidity, patient convenience, and cost savings from avoiding a second hospitalization and operation.

The carotid artery as alternative access for endoluminal aortic aneurysm repair.

Endograft treatment of aortic aneurysms has become a common procedure in many centers. However, not all patients are candidates for this new technology, because of their vascular anatomy and device limitations. One common problem is iliofemoral occlusive disease, which when present, even in a moderate degree, may preclude introduction of the large-diameter delivery devices currently in use. We present a case of a high-risk male patient with a thoracic aortic aneurysm and severe occlusive disease of the iliac arteries. An alternative approach for device delivery through the carotid artery was used and the procedure was successful with no neurologic complications. We recommend this technique for highly selected patients with an aneurysm who can undergo tube endograft repair without feasible access through the iliac or femoral arteries.

Factors predicting prolonged length of stay after carotid endarterectomy.

BACKGROUND: Over the last several years, implementation of critical pathways in patients undergoing carotid endarterectomy has decreased postoperative length of stay significantly. Discharge the day after surgery has become commonplace in many centers, including our own. Unfortunately, managed care may interpret this refinement as a standard of care and limit reimbursement or even disallow admissions extending beyond 1 day. We therefore examined our carotid registry to identify risk factors associated with postoperative length of stay exceeding 1 day. METHODS: We retrospectively reviewed all patients undergoing carotid endarterectomy at our academic center from May 1996 through April 1999. Combined procedures and patients undergoing subsequent noncarotid-related procedures on those admissions were excluded. The charts were inspected for atherosclerosis risk factors, including sex and age, specific attending surgeon, side of the surgery, use of intravenous vasoactive drugs, actual preoperative blood pressure, and presence of neurologic symptoms or postoperative complications. Multiple regression analysis was performed on all collected variables. Statistical significance was inferred for P less than.05. RESULTS: A total of 188 patients met the study criteria and had complete, retrievable medical records. A mean postoperative length of stay of 1.65 +/- 0.08 days and a mean total length of stay of 2.17 +/- 0.14 days were observed. Fifty-seven percent of patients went home the day after surgery. There was a 1.6% stroke-mortality rate. Significant predictors of a prolonged stay, listed in order of decreasing importance on the basis of their calculated contribution to prolonging the postoperative length of stay, are as follows (P value; beta coefficient): postoperative complications (<.0001; 1.03), age > 79 years (.008; 0.547), diabetes mellitus (.011; 0.407), female sex (.007; 0.398), and intravenous vasodilator requirement (. 035; 0.382). Other atherosclerosis risk factors, prior neurologic symptoms, the postoperative use of vasopressors, and reoperative surgery did not contribute to extended length of stay. CONCLUSIONS: Discharge on the first postoperative day is feasible in many, but not all, patients undergoing carotid endarterectomy. Our data help define subsets of patients at risk for prolonged postoperative stay. Targeting these subsets for preoperative medical and social interventions may allow safe early discharge more frequently.

The Balanced Budget Act: potential implications for the practice of vascular surgery.

PURPOSE: Previous study results have shown a favorable impact on stroke rate with an increasing hospital volume of carotid endarterectomies (CEAs). This is not only the most frequently performed peripheral vascular procedure in the United States but also perhaps the most widely dispersed procedure relative to hospital type. Medical centers have adopted various strategies to lower the cost of hospitalization by reducing the length of stay (LOS), the major component of hospital cost. By 2002, the Balanced Budget Act is projected to reduce Medicare provider payments to academic medical centers (AMCs) by 15.5%, a reduction that is twice that for minor or nonteaching hospitals. We assessed the relationships between hospital costs, CEA volume, and stroke-mortality rates in AMCs and non-AMCs in Massachusetts. METHODS: With patient level data from the Massachusetts Division of Health Care Finance and Policy and with hospital cost and charge reports from the Health Care Financing Administration, HealthShare Technology provided data for all the patients discharged from a Massachusetts hospital who underwent CEA (n = 10,211) during the fiscal years 1995, 1996, and 1997, including cost, LOS, and disposition. The outcomes were further defined with in-hospital stroke and mortality rates. Five high volume AMCs (HVAMCs) were compared with all other nonacademic hospitals, which were further subdivided by annual volume into high volume non-AMCs (> or =50 cases), medium volume non-AMCs (24-49 cases), and low volume non-AMCs (12-23 cases). Statistical analysis was performed with analysis of variance to compare the means of all the cost and LOS data, and chi(2) test was used for comparison of incidence (significance assumed for P < or =. 05). RESULTS: Hospital costs were comparable among the four hospital types during individual years and averaged $6200, but HVAMCs were significantly more expensive overall, with a mean cost of $7882. The only centers to decrease their costs during the years evaluated were the HVAMCs, from $8706 to $6784. Length of stay did not differ among the groups in any year or overall, with a mean of 3.8 days, but did decrease between years at HVAMCs from 3.9 to 2.5 days. The combined stroke-mortality rates were significantly less at the HVAMCs (0.9%) than at either the high volume non-AMCs (1.9%) or the medium volume non-AMCs (2.5%). There was no significance in the analysis results of all the data within the low volume non-AMCs. CONCLUSION: Patients in HVAMCs have the best outcomes after CEA. Despite the achievement of significant efficiencies, AMCs have a small cushion to reduce further either LOS or resources to maintain a competitive cost position and to compensate for the fixed expenses of academic medicine. The Balanced Budget Act raises an equity concern for AMCs because it differentially affects the centers with the best outcomes. The financial implication of this may be a direct incentive for procedures to be done in centers with less optimal outcomes.

A proposed relationship between circumcision and neural reorganization.

Humans are intensely biocultural beings. The linkages and causal feedback loops among their symbolic world, their cultural world, and their physical bodies can be exquisitely complex and subtle. It is suggested in this article that one cultural event--circumcision--exemplifies that subtlety and complexity. It is hypothesized that circumcision reorganizes the male's sensory somato-cortex to raise the threshold of sexual excitability/distraction. This threshold shift thereby allows the young men of a social group (a) to be slightly more tractable in executing corporate activities beneficial to the community and (b) to be slightly more restrained sexually and more cooperative in the pair bond. The practice is accepted because the procedure is deeply enmeshed in the ritual and symbolic life of the social group and is applicable to all young males. Suggestions are made on how to test this hypothesis empirically.

Reduction of hospital resources utilization in vascular surgery: a four-year experience.

PURPOSE: Managed care whether through risk or through capitated contracts results in reduction in resources, reduced length of hospital stay, and reduced utilization of hospital resources (collectively referred to as resource reductions). These resource reductions will become even more noticeable as a greater proportion of Medicare patients who need vascular operations select a managed-care senior product. We examined the results of a 4-year experience with resource management in an academic vascular surgery practice during which best practice plans were developed and implemented. METHODS: We analyzed hospital cost data, which included both total hospital and intensive care unit length of stay, average units per operation for laboratory, pharmacy, and radiology services and operating room and direct hospital costs for 257 carotid endarterectomies performed over fiscal years (FY) 1994, 1995, 1996, and 1997 (6 month data) and 175 infrainguinal bypass procedures performed during the same period. RESULTS: For carotid endarterectomy, length of stay decreased 66% over the 4-year period to an average of 2.07 days in FY97. Both radiology and pharmacy utilization were reduced after the first year of institution of best practice plans (56% and 32% respectively) with 4-year total reductions of 86% and 55% by FY97. The most notable changes included elimination of routine postoperative laboratory testing, use of aspirin rather than low-molecular-weight dextran, emphasis on oral rather than intravenous vasoactive drugs, and routine use of duplex scanning alone rather than angiography for diagnosis after FY94-95. The length of operating room time for carotid endarterectomy remained relatively constant from FY94 to FY97. As a result of these multiple factors, our study showed a 30% decrease in total average direct hospital costs for carotid endarterectomy from $9974 to $7002 in this 4-year period. Infrainguinal bypass graft procedures showed a progressive decrease in total cost of 28% for patients without complications to $15,186 but remained unchanged for those with complications. Laboratory use, pharmacy use, and radiology use were not significantly different. CONCLUSIONS: Case management for patients undergoing carotid endarterectomy and infrainguinal bypass grafting involving an integrated team of vascular surgeons, surgical house staff, a dedicated vascular nurse, and a social work case manager resulted in dramatic reductions both in length of stay and hospital resource utilization. As these costs decreased, operating room expenses assumed increasing importance. Operating room costs account for 60% of the direct costs of carotid endarterectomy and a comparable percentage for uncomplicated infrainguinal bypass grafting. Further substantial reductions in direct hospital costs will depend primarily on reductions in operating room costs, particularly those related to length of time in the operating room.

Carotid and coronary disease: staged or simultaneous management?

Up to 12% of patients presenting for coronary bypass have critical carotid disease, and more than 50% of patients presenting for carotid endarterectomy have significant coronary disease. Patients requiring surgery for both carotid and coronary disease may be managed with carotid endarterectomy followed by coronary bypass (staged approach), with coronary bypass followed by carotid endarterectomy (reversed staged approach), or with simultaneous coronary bypass-carotid endarterectomy. There are no compelling data proving superiority of any of these three approaches. The staged approach is usually associated with lower stroke rates but higher myocardial infarction and mortality rates; the reversed staged approach with higher stroke rates but lower myocardial infarction and mortality rates; and the simultaneous approach with intermediate stroke, myocardial infarction, and mortality rates. Unfortunately, reported series vary widely in stroke and mortality rates because of wide variability in patient selection criteria, especially for simultaneous procedures. Management decisions in these patients should be based on the relative severity of their carotid and coronary lesions. Management guidelines are discussed in detail.

An additional facet of the incest taboo: a protection of the mating-strategy template.

It is suggested that sexual intercourse between adult males and girls tends to distort an inherent mating-strategy template: female choice of mating partner. The distortion seems to have long-range consequences in the form of myriad psychopathologies, which, in turn, reduce the afflicted individuals' chances for normative marriage and parenting profiles. In addition, a similar dynamic would hypothetically result from adult-male to boy incest. It is suggested that to minimize the chances of adult-child sexual intercourse, incest taboos have historically been reinforced and extended to nonparental adults, especially men, beyond the immediate nuclear family.

A reexamination of Gilligan's analysis of the female moral system : Distaff altruism will not succeed.

Gilligan's (1982) refinement of Kohlberg's theory on moral development operates on two theses: (1) females, more so than males, reach moral decisions based on the personalities of the relevant individuals; and (2) female behaviors stemming from moral decisions are based upon "care" and "responsibility for others." This article accepts the first thesis but argues that the second is incorrect. That is, self-interest-i.e., aiding "blood" kin and/or carefully monitoring reciprocity-rather than "altruism" is argued to be the operant dynamic in forging distaff morality and resultant behavior. Six empirical examples are presented as contraindicative of Gilligan's second thesis. Finally, it is suggested that selection for the psychological traits of independence and the mastery of subtle social chess yielded ancestral females who had more descendants-us-than did females with alternative profiles.

Establishing a link between cultural evolution and sexually transmitted diseases.

It is argued that archaic sexually transmitted diseases influenced cultural traditions by reducing multiple sexual partners within communities. In this article, the adverse consequences of current sexually transmitted diseases are surveyed: Such infections decrease fertility of women and increase infant mortality; those adverse consequences are especially potent when antibiotics are not readily available. Cultural (cross-generational transmission of learned) responses to the threat of widespread infertility and elevated infant mortality rates are hypothesized to include the implementation of expectations for restricted numbers of sexual partners. These expectations, formal or informal, have been instituted within the context of biological predispositions, the "certainty of paternity" model, already-established traditions, and the need for a social father to be aligned with the mother-child dyad. A case study of the contemporary United States is offered as a heuristic example of how and why cultural choices may be developed and sustained.

A biocultural analysis of circumcision.

The phenomenon of circumcision may well serve a range of religious and symbolic functions. In addition to these conceptual categories, we argue that circumcision also serves a more mundane, practical function of lowering excitability and distractibility quotients--sexual arousal--of pubescent males, i.e., biasing young males more toward increased tractability which would enhance group efforts and less toward individual goals of amorous exchanges. Neurological data suggest that early lesions of the prepuce/foreskin tissues would generate a re-organization/atrophy of the brain circuitry. This re-organization/atrophy, in turn, is suggested to lower sexual excitability. Epithelial data indicate that keratinization of the more exposed glans penis would lower the sensibility, hence sexual excitability, of the circumcised male's genitalia. In addition, circumcision removes the foreskin-prepuce which, by secreting smegma, would also minimize any pheromonic qualities which the smegma may generate. Inferential data support the hypothesis that a practical consequence of circumcision, complementary to any religious-symbolic function, is to make a circumcised male less sexually excitable and distractible, and, hence, more amenable to his group's authority figures.

Clinical results of common strategies used to revise infrainguinal vein grafts.

PURPOSE: Patients who have failing infrainguinal bypass grafts or failed grafts reopened with lytic therapy represent a group at high risk of subsequent failure. Previous studies suggest that vein patch angioplasty and jump grafting may be less durable than interposition grafting as a method of correcting graft lesions. Our objective was to assess the value of various technical strategies for graft revision in a series of autogenous infrainguinal bypass grafts and to assess how these variables might affect cumulative graft patency (CGP) rates. METHODS: We retrospectively reviewed the clinical course, anatomic sites of revision, and type of revision performed on 67 grafts in 58 patients who underwent at least one revision from 1991 to 1995. Results were assessed with regression analysis and Kaplan-Meier estimates of CGP rates (p < 0.05 was considered significant). RESULTS: Sixty-seven vein grafts underwent revision of 112 anatomical sites in 95 operations. Forty-nine of 67 grafts were single-segment greater saphenous vein grafts and 18 were composite (> 1 segment) grafts, with an overall 5-year CGP rate of 72%. No difference was observed between the 4-year CGP rate in grafts with hemodynamically significant distal anastomotic stenoses repaired primarily with jump grafts (n = 20, 71% CGP rate) and those with stenoses found only in the graft body (n = 41, 89% CGP rate). Vein patch angioplasty was used primarily, but not exclusively, for focal graft body stenoses (n = 35), whereas interposition grafts (n = 11) were reserved for more diffuse strictures; no significant difference in 3-year CGP rates was observed (94% and 73%, respectively). CONCLUSION: Using an appropriate revision strategy that favors vein patch angioplasty for graft body lesions and jump grafts for distal anastomotic lesions, acceptable assisted patency rates can be achieved in grafts that are at risk for repeated failure.

Duplex assessment of venous reflux and chronic venous insufficiency: the significance of deep venous reflux.

PURPOSE: This study was undertaken to examine the role of superficial and deep venous reflux, as defined by duplex-derived valve closure times (VCTs), in the pathogenesis of chronic venous insufficiency. METHODS: Between January 1992 and November 1995, 320 patients and 500 legs were evaluated with clinical examinations and duplex scans for potential venous reflux. VCTs were obtained with the cuff deflation technique with the patient in the upright position. Imaging was performed at the saphenofemoral junction, the middle segment of the greater saphenous vein, the lesser saphenous vein, the superficial femoral vein, the profunda femoris vein, and the popliteal vein. Not all patients had all segments examined because tests early in the series did not examine the profunda femoris or lesser saphenous vein and because some patients had previous ligation and stripping or venous thrombosis. VCTs were examined for individual segment reflux, grouped into superficial and deep systems, and then correlated with the clinical stage as defined by the SVS/ISCVS original reporting standards in venous disease. Segment reflux was considered present if the VCT was greater than 0.5 seconds, and system reflux was considered present if the sum of the segments was greater than 1.5 seconds. Between-group differences were analyzed with analysis of variance and post hoc tests where appropriate. RESULTS: Sixty-nine limbs studied were in class 0, 149 limbs were in class 1, 168 limbs were in class 2, and 114 limbs were in class 3. VCTs in the superficial veins were significantly lower in class 0 than in the other clinical classes. There was no difference in superficial reflux in the symptomatic limbs (classes 1 to 3). Reflux VCTs in the superficial femoral and popliteal veins increased as the clinical symptoms progressed, with a significant increase in class 3 ulcerated limbs when compared with nonuclerated limbs. The incidence of deep venous reflux was 60% in class 3 limbs, compared with 29% in class 2 limbs, whereas the incidence of superficial venous reflux did not differ among the symptomatic limbs. Isolated superficial femoral and popliteal vein reflux was uncommon, even in class 3 limbs, but combined superficial femoral and popliteal vein reflux was found in 53% of class 3 limbs, compared with 18.5% of class 2 limbs. CONCLUSIONS: Reflux in the deep venous system plays a significant role in the progression of chronic venous insufficiency. Deep system reflux increases as clinical changes become more severe, with significant axial reflux contributing to ulcer formation.

Who needs surveillance of the contralateral carotid artery?

BACKGROUND: Although the value of carotid endarterectomy has been proven, postoperative surveillance remains controversial. The purpose of this study was to determine the natural history of disease progression in the contralateral carotid artery by duplex surveillance, and to assess the cost of stroke prevention on this contralateral side. METHODS: Vascular laboratory records were reviewed to identify carotid endarterectomy patients who had two or more duplex studies between 1984 and 1995. Critical stenosis was defined as > or = 75% area reduction. RESULTS: In all, 324 patients were followed up with duplex scans for 1 month to 11 years (mean 30.3 months). The only factors that correlated with progression to critical stenosis were age and initial stenosis. Overall, 19.5% of patients progressed to critical stenosis within 5 years while the high-risk groups with age > 65 years or initial stenosis > or = 50% progressed to critical disease in 27% and 39%, respectively (P < or = 0.05). The cost per stroke prevented ranged from $143,500 to $418,200 when stratified by initial stenosis. CONCLUSION: Patients who have undergone a carotid endarterectomy demonstrate a propensity for progression of carotid stenosis in the unoperated (contralateral) artery. The cost/benefit ratio may be improved by varying the intensity of duplex surveillance of the contralateral carotid based on the patient's age and initial degree of stenosis.

Management of recurrent carotid stenosis: should asymptomatic lesions be treated surgically?

PURPOSE: The purpose of this study was to determine factors that may influence patient selection for surgery in recurrent carotid stenosis (RCS) and to contrast the results of primary and secondary carotid endarterectomy (CENDX) with regard to operative morbidity and stroke prevention. METHODS: Forty-eight patients who underwent CENDX for RCS (RCS-OP group) were compared with a contemporaneous group of 40 patients who on at least one post-CENDX duplex ultrasonography study had a greater than 50% stenosis but did not undergo operation (RCS-NO-OP group). This latter group was drawn from 1053 follow-up duplex studies in 348 patients who underwent primary CENDX between the years 1983 and 1993. Each of these two groups was compared with a metanalysis of six key series derived from the literature. RESULTS: No significant differences were seen in the demographics or the incidence of risk factors between the two groups except for a higher incidence of coronary artery disease (p < 0.03) and peripheral vascular disease (p < 0.001) in the RCS-OP group. The operation-specific stroke rate was 2.1%, and the 30-day mortality was also 2.1%. Symptomatic RCS was the indication in 56% of cases. Important anatomic differences were found between groups. The duplex/arteriographic degree of stenosis was greater than 90% in 75% of the patients in the RCS-OP group, whereas only 10% of the patients in the RCS-NO-OP group had greater than 80% stenosis, most being in the 50% to 80% range. An unexpected finding was the sudden progression to occlusion in 10 (25%) of 40 in the RCS-NO-OP group, with 2 (5%) of 10 of the occlusions presenting as unheralded strokes. Overall, a stroke without an antecedent transient ischemic attack occurred in 3 (7.5%) of 40 of patients in the RCS-NO-OP group, all in patients with greater than 75% stenosis on their last documented scan preceding the stroke. CONCLUSION: Given the relatively low stroke rate with surgery in the RCS-OP group (2.1%) and the higher incidence of unheralded strokes (7.5%) in the RCS-NO-OP group, a more aggressive approach may be warranted in patients with asymptomatic high-grade (> 75%) RCS, a strategy not unlike that adopted for primary CENDX.

Evaluation of carotid artery stenosis: is duplex ultrasonography sufficient?

PURPOSE: The purpose of this study was to compare the results of duplex ultrasonography and magnetic resonance angiography in the evaluation of carotid artery stenosis to determine whether ultrasonography alone is sufficient for preoperative evaluation. METHODS: This study consisted of a retrospective review of 33 patients who underwent 35 carotid endarterectomies. A total of 66 vessels were studied by both duplex ultrasonography and magnetic resonance angiography, and an overall correlation between the two studies was determined. RESULTS: A high correlation was found between duplex and magnetic resonance angiography with an r coefficient equal to 0.87 (Pearson's correlation coefficient) and kappa = 0.75. Discrepancies between the two studies or the presence of intracranial disease did not alter surgical decision making. CONCLUSION: Duplex ultrasonography alone can accurately determine the degree of internal carotid artery stenosis and when paired with careful clinical evaluation is a reliable and cost-effective method for evaluating surgical carotid disease.

Simultaneous carotid endarterectomy and coronary bypass: perioperative risk and long-term survival.

PURPOSE: The purpose of this article is to examine the outcome of simultaneous coronary bypass-carotid endarterectomy (CABG-CEA) and to compare it with the outcome of endarterectomy alone (CEA alone) in patients at high cardiac risk. METHODS: A retrospective review of the records and follow-up data for 100 consecutive patients who had undergone CABG-CEA and were at high risk and 114 patients who had undergone CEA, had overt coronary artery disease (angina, previous infarct, or ischemic electrocardiographic abnormalities), but had not undergone CABG was carried out. RESULTS: Our CABG-CEA group had a high incidence of symptomatic carotid disease (57%) and contralateral occlusion (28%) when compared with patients in other reports. Patients in the CABG-CEA group were older (67.9 +/- 8.3 years vs 63.6 +/- 15.7 years, p = 0.01) and more often smokers (81% vs 52.6%, p = 0.01) than patients in the CEA alone group. Perioperative mortality was 8% for the CEA-CABG group and for 1.8% for the CEA alone group (p = 0.035). Perioperative stroke morbidity was 9% for the CEA-CABG group and 2.6% for the CEA alone group (p = 0.05). Life table survival at 1,3, and 5 years was 90%, 82%, and 73% versus 96%, 84%, and 76% for the CABG-CEA and CEA alone groups, respectively (p = 0.30). CONCLUSIONS: Selection criteria for CABG-CEA greatly influence perioperative risk. Despite the greater age and more advanced coronary artery disease in the CABG-CEA group, long-term outcome differences are accounted for entirely by differences in perioperative morbidity and mortality. Prospective trials of strategies such as staged CEA and CABG to reduce perioperative risk are needed.

Brain edema after carotid surgery.

The postoperative hyperperfusion syndrome describes an abrupt increase in blood flow with loss of autoregulation in surgically reperfused brain. Reports described a spectrum of findings, including severe headache, transient ischemia, seizures, and intracerebral hemorrhage. Hypertension is common after carotid artery surgery and often plays a role in the pathophysiology. We now report five patients with severe white matter edema after carotid surgery, a finding not previously included in the hyperperfusion syndrome. Five to 8 days after carotid surgery and after hospital discharge, each patient developed hypertension, headache, hemiparesis, seizures, and aphasia or neglect due to severe white matter edema ipsilateral to the carotid surgery. One patient had a small hemorrhage within the edematous area. Hypertension was severe in four patients and moderate in one. The carotid artery was patent by ultrasound or angiography in each patient after surgery. Transcranial Doppler showed increased velocities ipsilateral to surgery in two patients and bilaterally in one. Computed tomographic abnormalities and neurologic signs resolved within 3 weeks in four of the five patients treated with antihypertensives and anticonvulsants. The fifth patient died from herniation secondary to massive edema. Brain edema with focal neurologic signs should be included as a serious but potentially reversible component of the postoperative hyperperfusion syndrome.

Foot infections in diabetic patients. Decision and cost-effectiveness analyses.

OBJECTIVE: To examine the cost-effectiveness of approaches to the diagnosis and treatment of patients with type II (non-insulin-dependent) diabetes mellitus (NIDDM) who have foot infections and suspected osteomyelitis. DESIGN: Decision and cost-effectiveness analyses were performed using a Markov model. We examined the prevalence of osteomyelitis, the major complications and efficacies of long-term antibiotic therapy and surgery, and the performance characteristics of four diagnostic tests (roentgenography, technetium Tc 99m bone scanning, indium in 111-labeled white blood cell scanning, and magnetic resonance imaging). Data were drawn from the English-language literature using MEDLINE searches and bibliographies from selected articles. SETTING: Primary care. PATIENTS: Patients with NIDDM who had foot infections and suspected osteomyelitis but no signs of systemic toxicity. INTERVENTIONS: Following hospitalization for surgical débridement and intravenous antibiotic therapy: (1) treatment for presumed soft-tissue infection, (2) culture-guided empiric treatment for presumed osteomyelitis, (3) 71 combinations of diagnostic tests preceding antibiotic therapy for osteomyelitis, (4) 71 combinations of tests preceding amputation, and (5) immediate amputation. MAIN OUTCOME MEASURES: Quality-adjusted life expectancy, average costs. RESULTS: Culture-guided empiric treatment for osteomyelitis with 10 weeks of oral antibiotic therapy has similar effectiveness to testing followed by a long course of antibiotic therapy if any test result is positive. However, empiric treatment is the least expensive strategy. CONCLUSIONS: Noninvasive testing adds significant expense to the treatment of patients with NIDDM in whom pedal osteomyelitis is suspected, and such testing may result in little improvement in health outcomes. In patients without systemic toxicity, a 10-week course of culture-guided oral antibiotic therapy following surgical débridement may be as effective as and less costly than other approaches.