Comparison of survival following coronary artery bypass grafting vs. percutaneous coronary intervention in diabetic and non-diabetic patients: retrospective cohort study of 6320 procedures.

AIM: To determine whether mortality following percutaneous coronary intervention vs. coronary bypass grafting varies according to whether or not patients have diabetes. METHODS: We used the Scottish Coronary Revascularization Register to identify all patients undergoing revascularization in Scottish NHS hospitals since 1997. We excluded single-vessel disease, left main stem stenosis, and bypass grafting performed at the same time as other operations. We used death certificate data from the Registrar General to identify all subsequent deaths. RESULTS: Of the 6320 eligible procedures, 5042 (80%) were bypass grafts and 1278 (20%) angioplasties. Overall 831 (13%) patients had diabetes with no significant difference by procedure (13% vs. 12%). A total of 382 deaths occurred over a mean follow-up of 2.3 years. Diabetic patients had a poorer prognosis following both surgery (adjusted hazards ratio (HR) 1.43, 95% confidence interval (CI) 1.08, 1.89) and percutaneous intervention (adjusted HR 2.58, 95% CI 1.43, 4.63). Among non-diabetic patients, no significant differences in mortality were detected between the two procedures. Among diabetic patients, no significant difference was detected in those with two-vessel disease. In those with impaired left ventricular function and triple-vessel disease, angioplasty was associated with a significantly higher risk of death (adjusted HR 3.58, 95% CI 1.40, 9.19). CONCLUSIONS: This is the first study to demonstrate statistically significant results that support the BARI trial findings. Our study demonstrated a significant difference for triple-vessel disease but not two-vessel disease. The former may be due to incomplete revascularization using percutaneous intervention. Our results require corroboration from randomized trials.

Outcomes following coronary artery bypass grafting and percutaneous transluminal coronary angioplasty in the stent era: a prospective study of all 9890 consecutive patients operated on in Scotland over a two year period.

OBJECTIVE: To determine current outcomes of percutaneous transluminal coronary angioplasty (PTCA) and coronary artery bypass grafting (CABG). DESIGN: The Scottish coronary revascularisation register provided prospectively collected data on case mix and in-hospital complications for all revascularisation procedures between April 1997 and March 1999 (4775 PTCA; 5115 CABG). Linkage to routine hospital discharge and death data provided follow up information on survival and repeat revascularisation. RESULTS: Stents were used in 51% of PTCA procedures. CABG patients were older, had more severe coronary disease, and had greater comorbidity. PTCA was more likely to be undertaken as an urgent or emergency procedure. Perioperative death and urgent surgery followed 0.3% and 0.6% of PTCA procedures, respectively. Case fatality rates were higher following CABG, with 6.7% dead within two years compared with 3.4% following PTCA. PTCA was more often followed by readmission for ischaemic heart disease, repeat angiography, or revascularisation: 22.8% of patients had repeat revascularisation within two years, compared with 1.8% following CABG. CONCLUSIONS: The severity of coronary heart disease was greater than in previously published registry studies and randomised trials. Despite this, overall survival figures were comparable and repeat revascularisation rates lower, particularly following PTCA. Perioperative death and urgent surgery following PTCA were also lower. These favourable outcomes may be attributable, in part, to increased use of bail out and elective stenting.

Coronary surgery in Europe: comparison of the national subsets of the European system for cardiac operative risk evaluation database.

OBJECTIVE: To compare the national samples of patients who underwent isolated coronary artery bypass grafting (CABG) during the European System for Cardiac Operative Risk Evaluation (EuroSCORE) trial in order to evaluate national differences in epidemiology, patient risk profile and surgical methods. METHODS: From September to November 1995, 11731 patients had CABG in the six largest contributing nations to the EuroSCORE project: Germany, UK, Spain, Finland, France and Italy. The Chi-square and Kruskal-Wallis tests were applied to obtain an international comparison of patient general status, including pre-operative risk factors, cardiac status, critical pre-operative states, rare conditions, urgency of surgery, angina status, coronary lesions, procedures and EuroSCORE risk assessment. RESULTS: Large national samples (from 984 patients in Finland to 3138 in Germany) identified significant differences in epidemiology, risk profile and surgical practice. Regarding epidemiology, CABG accounted for 62.8% of adult cardiac surgery, with a range of 46.2 in Spain to 77.7% in Finland (P<0.001). The mean age was 62.9 years (61.4 in Britain to 64.4 in France, P<0.001). The mean body mass index was 26.8 (26 in France to 27.5 in Finland, P<0.001). With regard to risk profile, diabetes was present in 20.3% of patients (11.8% in Britain to 27.7% in Spain, P<0.001). Chronic renal failure was present in 8.3% (6.8% in Germany to 10.6% in Spain, P<0.001). Chronic airway disease affected 3.8% (1.9% in Italy to 5. 1% in Germany, P<0.001). The mean ejection fraction was 0.56 (0.48 in Britain to 0.58 in Finland, P<0.001). The mean predicted mortality (according to EuroSCORE) was 3.3% (2.8% in Finland to 3.6% in France, P<0.001). The prevalence of chronic congestive heart failure, unstable angina and recent myocardial infarction also showed statistically significant differences. No differences were found for some critical preoperative states (such as immediate preoperative cardiac massage and pre-operative intubation), or for surgery for catheter laboratory complication. Regarding surgical practice, major differences were noted in preoperative intra-aortic balloon use (mean 1%, Finland 0%, Spain 2.3%, P<0.001), the number of mammary artery conduits used (mean 0.9, Spain 0.7, France 1.1, P=0.0001) and the number of distal anastomoses (mean 3, France 2.7, Finland 3.8, P=0.001). CONCLUSION: There are important epidemiological differences in the national cohorts of CABG patients in the EuroSCORE database. Any international comparison of European surgical results must therefore take into account the risk profile of patients by using a compatible risk stratification system.

Risk factors and outcome in European cardiac surgery: analysis of the EuroSCORE multinational database of 19030 patients.

OBJECTIVE: To assess risk factors for mortality in cardiac surgical adult patients as part of a study to develop a European System for Cardiac Operative Risk Evaluation (EuroSCORE). METHODS: From September to November 1995, information on risk factors and mortality was collected for 19030 consecutive adult patients undergoing cardiac surgery under cardiopulmonary bypass in 128 surgical centres in eight European states. Data were collected for 68 preoperative and 29 operative risk factors proven or believed to influence hospital mortality. The relationship between risk factors and outcome was assessed by univariate and logistic regression analysis. RESULTS: Mean age (+/- standard deviation) was 62.5+/-10.7 (range 17-94 years) and 28% were female. Mean body mass index was 26.3+/-3.9. The incidence of common risk factors was as follows: hypertension 43.6%, diabetes 16.7%, extracardiac arteriopathy 2.9%, chronic renal failure 3.5%, chronic pulmonary disease 3.9%, previous cardiac surgery 7.3% and impaired left ventricular function 31.4%. Isolated coronary surgery accounted for 63.6% of all procedures, and 29.8% of patients had valve operations. Overall hospital mortality was 4.8%. Coronary surgery mortality was 3.4% In the absence of any identifiable risk factors, mortality was 0.4% for coronary surgery, 1% for mitral valve surgery, 1.1% for aortic valve surgery and 0% for atrial septal defect repair. The following risk factors were associated with increased mortality: age (P = 0.001), female gender (P = 0.001), serum creatinine (P = 0.001), extracardiac arteriopathy (P = 0.001), chronic airway disease (P = 0.006), severe neurological dysfunction (P = 0.001), previous cardiac surgery (P = 0.001), recent myocardial infarction (P = 0.001), left ventricular ejection fraction (P = 0.001), chronic congestive cardiac failure (P = 0.001), pulmonary hypertension (P = 0.001), active endocarditis (P = 0.001), unstable angina (P = 0.001), procedure urgency (P = 0.001), critical preoperative condition (P = 0.001) ventricular septal rupture (P = 0.002), noncoronary surgery (P = 0.001), thoracic aortic surgery (P = 0.001). CONCLUSION: A number of risk factors contribute to cardiac surgical mortality in Europe. This information can be used to develop a risk stratification system for the prediction of hospital mortality and the assessment of quality of care.

Radiographic detection of single strut leg separations as a putative basis for prophylactic explantation of Björk-Shiley convexo-concave heart valves.

Cineradiography, using higher kVp and two or more specified profiles for each outlet strut leg, was used to evaluate Björk-Shiley convexo-concave (C/C) heart valves with epidemiologically defined > 0.1% per year estimated fracture rates. Among 828 mitral valves, eight radiographs were assessed as definite single leg separations (SLS) and 23 were read as probable SLS. Explantation confirmed SLS in 24 valves; 4 probables were false positives, and 3 patients with probable SLS ratings decided against explantation. Four patients with SLS died, an operative mortality of 14%. Only 23 mitral and 6 aortic valves receiving ratings ranging from apparently normal to suspicious have become available for verification; 1 rated apparently normal was found to have a SLS. Two patients experienced fatal fractures and 1 SLS valve was explanted 3 to 15 months after apparently normal x-ray studies; it cannot be known if a SLS was or was not present at the time of these examinations. As 97% of negatively rated valves remain in situ, the sensitivity of the test is similarly not known. Only 1 SLS has been detected among 136 aortic valves. Uncertainty about test accuracy and SLS progression condition the clinical utility of radiographic SLS detection, particularly with respect to assurance from apparently normal readings. However, radiographic identification of an SLS substantially enhances epidemiologically derived risk categorization as a basis for consideration of prophylactic replacement for the approximately 12,000 valves with a > 0.1% estimated annual fracture risk, currently thought to be implanted in living patients.

The benefits of active rewarming after cardiac operations: a randomized prospective trial.

We conducted a prospective, randomized trial of three methods of rewarming patients after hypothermic cardiopulmonary bypass. Patients underwent either coronary artery bypass grafting or first-time valve replacement and were cooled to 32 degrees C during bypass. No significant differences existed among the groups as regards operative or preoperative parameters including hemodynamics and blood use. The patients actively warmed with a convective (Bair Hugger system, 3 hours) or a conductive blanket (electric overblanket, 4 hours) reached normothermia more quickly than those warmed with the space blanket (7 hours). This was reflected in significantly earlier extubation in the former two groups: Bair Hugger system 10.8 +/- 0.6 hours, electric blanket 11.3 +/- 1.0 hours, and space blanket 14.8 +/- 0.8 hours. Patients warmed with the space blanket required a higher dosage of morphine over the first 12 hours than those warmed with the electric blanket (10.4 vs 6.5 mg; p = 0.004), which may account for some of the differences between these two groups. No differences could be demonstrated between the two active blankets. On economic grounds we therefore recommend the reusable electric blanket for routine use.

Detection of microemboli by transcranial Doppler ultrasound.

Doppler ultrasound detection of abnormally high-pitched signals within the arterial waveform offers a new method for diagnosis, and potentially for prediction, of embolic complications in at-risk patients. The nature of Doppler "microembolic" signals is of particular interest in patients with prosthetic heart valves, where a high prevalence of these signals is observed. Monitoring the middle cerebral artery with 2-MHz transcranial Doppler ultrasound (TC-2000, Nicolet Biomedical; Warwick, UK), we looked for microemboli signals in 150 patients (95 women and 55 men), and found 1 or more signals during a 30-min recording in 89% of 70 patients with Bjork-Shiley valves (principally monostrut), 54% of 50 patients with Medtronic-Hall valves, and 50% of 30 patients with Carpentier-Edwards valves (p < 0.001, chi 2). In the patients with Bjork-Shiley valves, the mean number of signals per hour was 59 (range, 42-86; 95% confidence interval), which was significantly higher than the mean in patients with Medtronic-Hall and Carpentier-Edwards valves (1.5[range, 0.5-2.5] and 1 [range, 0-5.3], respectively; both p < 0.04, multiple comparisons. Bonferroni correction). In the patients undergoing serial pre- and postoperative studies, the causative role of the valve implant was emphasized. There was no correlation between the number of emboli signals and a prior history of neurologic deficit, cardiac rhythm, previous cardiac surgery, or the intensity of oral anticoagulation, in patients with prosthetic heart valves. In Bjork-Shiley patients, dual (mitral and aortic) valves were associated with more signals than were single valves. In Medtronic-Hall patients, the signal count was greater for valves in the aortic position than it was for valves in the mitral position. Comparative studies of Doppler emboli signals in other clinical settings suggest a difference in composition or size of the underlying maternal between prosthetic valve patients and patients with carotid stenosis. These studies also suggest that the signals are of gaseous origin in valve patients. The clinical significance of continuing microembolism remains to be determined.

Transcranial Doppler detection of microemboli in prosthetic heart valve patients: dependency upon valve type.

Transcranial Doppler ultrasound has revealed the existence of cerebrovascular microemboli in asymptomatic patients with prosthetic heart valves. We investigated the relation between the presence and number of emboli signals and valve type. Patients with six types of prosthetic valves (Björk-Shiley monostrut, Medtronic-Hall, Carbomedics, ATS, Carpentier-Edwards standard, Carpentier-Edwards supraannular) were examined using transcranial Doppler ultrasound in two centers. The monitoring time was 30 min over the right middle cerebral artery. All patients were stabilized on warfarin at the time of study. Microemboli signals were identified by their characteristic audiovisual signal and on subsequent spectral analysis, based on accepted criteria. A standard neurologic questionnaire was completed by all patients. The prevalence of microemboli signals varied between 49% (Medtronic Hall) and 97% (Björk-Shiley monostrut), while their number varied between 1 [0-3] (Carpentier-Edwards standard) and 187 [136-240] (Björk-Shiley monostrut) per hour (median and 95% CI). Both parameters were significantly higher in patients with Björk-Shiley monostrut valves compared to the other patient groups. There were no significant differences in the prevalence of neurologic complications among the groups examined (overall 16%), or in emboli numbers between symptomatic and asymptomatic patients. The prevalence and quantity of microemboli signals in patients with prosthetic heart valves, as detected by transcranial Doppler, is dependent upon valve type. The clinical significance of these microemboli signals remains to be further evaluated.

Ultrasound detection of cerebral emboli in patients with prosthetic heart valves.

This study was undertaken in 64 patients, 50 with mechanical and 14 with porcine prosthetic valves, to evaluate the incidence of intracranial emboli and their distribution in the basal cerebral arteries. The patients were studied using transcranial Doppler (EME TC2-64B, Uberlingen, Germany), with a monitoring time of two minutes over each of the internal carotid arteries, middle and anterior cerebral arteries, vertebral arteries and the basilar artery. Sixty-three of the 64 patients were stabilized on warfarin at the time of the study. The incidence of emboli signals was significantly higher in patients with mechanical compared to porcine cardiac valves (88% versus 14%, p < 0.01). The number of emboli signals was significantly higher in the anterior compared with the posterior circulation, with a median of eight signals in the internal carotid arteries (95% confidence interval 5-15), 2.5 in the vertebral arteries (95% confidence interval 1-5.5)(p < 0.03). It was also significantly higher in those patients who had undergone double (aortic and mitral) as opposed to those who had undergone single aortic valve replacement: 18 versus two signals per minute (confidence intervals 5-30.5 versus 0.5-3.5) (p < 0.01). It is concluded that subclinical emboli signals are readily detectable using transcranial Doppler and are common in patients with prosthetic heart valves. Their number depends on both the type and the number of the prosthesis, while their distribution in the basal cerebral arteries is consistent with their cardiac source.

Prevalence and characteristics of intracranial microemboli signals in patients with different types of prosthetic cardiac valves.

BACKGROUND AND PURPOSE: Transcranial Doppler detection of microemboli is widely described, but there is no clear evidence of the clinical significance or nature of the embolic material in vivo. Thromboembolism is a major cause of morbidity in patients with prosthetic cardiac valves. We undertook this study to evaluate the prevalence and the acoustic characteristics of microembolic signals in three groups of patients with different prosthetic valves. METHODS: One hundred seventy-nine patients with prosthetic cardiac valves (85 Björk-Shiley, 56 Medtronic-Hall, and 38 Carpentier-Edwards) and 25 normal subjects were examined using transcranial Doppler. Monitoring time was 30 minutes over the right middle cerebral artery. RESULTS: The prevalence and numbers of embolic signals were significantly higher in patients with Björk-Shiley compared with those with Medtronic-Hall and Carpentier-Edwards valves (89% versus 50% and 53%, respectively; P < .001, chi 2; 156 [112, 204] versus 2 [1, 4] and 2 [1, 4] signals/h, respectively; median [95% nonparametric confidence interval], both P < .001, multiple comparisons, Bonferroni correction). The signal intensity was significantly higher in patients with Björk-Shiley and Medtronic-Hall valves than patients with Carpentier-Edwards valves (2435 [2345, 2527] and 2120 [1745, 2483] versus 225 [184, 287] power units, median [95% confidence interval], both P < .001). No correlation was found between embolic signal numbers and clinical parameters including history of neurological deficit, cardiac rhythm, duration of artificial valve, previous cardiac operations, or intensity of anticoagulation. Embolic signals were not detected in any of the control subjects. CONCLUSIONS: Our data showed that the prevalence, quantity, and acoustic characteristics of Doppler embolic signals differ in patients having three different types of prosthetic heart valves. However, no correlation with clinical parameters was identified.

Subclinical embolism in patients undergoing cardiac valve implantation and coronary artery bypass surgery.

Patients undergoing prosthetic valve insertion and coronary artery bypass surgery were examined with transcranial Doppler ultrasound, recently shown to be capable of detecting continuing subclinical emboli in patients with embolic sources. In 30 patients examined at least 1 year after valve surgery, and in whom warfarinisation was stable within defined limits, 20 of 24 patients (83%) with mechanical valves and 3 of 6 patients (50%) with porcine valves had embolic signals. In a serial preoperative and postoperative study in a further 30 patients, of whom 29 had native or bioprosthetic valves, only the one patient with a previous mechanical mitral valve prosthesis had embolic signals preoperatively. The incidence of embolic signals increased to 9 (30%) on the first postoperative day, and 20 (67%) on day 5. In a similar serial study in 25 patients undergoing coronary bypass surgery, 8 (32%) had preoperative embolic signals, which were explicable by cardiac and/or carotid disease in 6 cases. The embolus signal incidence and count did not increase postoperatively in this group. No embolic signals were found in 15 volunteer controls. The results indicate that prosthetic valves cause continuing microembolisation, detectable by transcranial Doppler; coronary artery bypass cases may have incidental embolic signals which are unaffected by cardiac surgery. This new application of Doppler ultrasound may improve the clinical assessment of embolic risk of new prosthetic valve types and deserves further examination.

Serum androgens and gonadotrophins in bronchial carcinoma.

The concentrations of testosterone, free testosterone index derived by two independent techniques (FTI1, FTI2), dehydroepiandrosterone sulphate (DHAS), androstenedione (AND), luteinizing hormone (LH), follicle stimulating hormone (FSH), and sex hormone binding globulin (SHBG) were measured in 38 healthy male controls aged between 40 and 60 years (Group 1) and in 52 male patients with bronchial carcinoma of whom 28 were aged between 40 and 60 years (Group 2) and 24 were aged over 60 years (Group 3). Compared with Group 1, testosterone, FTI1, FTI2, and DHAS were significantly lower (P < 0.001), AND and LH were significantly higher (P < 0.001) and there was no significant difference in FSH or SHBG, in Group 2. No significant difference between Groups 2 and 3 was found for any variable. Analysis of covariance for Groups 2 and 3 combined, showed testosterone, FTI1 and FTI2 were significantly lower in the presence of metastases. These results suggest a more widespread abnormality of androgen metabolism in male patients with bronchial carcinoma than a simple reduction in serum testosterone; DHAS is also significantly reduced while AND is significantly increased. The primary abnormality appears to be at testicular and/or adrenal level as pituitary function appears normal as witnessed by compensatory increases in LH.

Achalasia-like syndrome in systemic sclerosis.

Two methods of quantifying oesophageal emptying for liquids have been used to assess the dysphagia of patients with systemic sclerosis: the oesophageal infusion scintiscan and the timed Gastrografin swallow. Upper gastrointestinal endoscopy and oesophageal manometry were also performed. Thirteen patients with oesophageal symptoms were studied. Eight had dysphagia, and all of these had endoscopies with no evidence of oesophagitis or stricture. Four of these eight subjects had gross delay of oesophageal emptying for fluids, and manometry showed absence of oesophageal peristalsis and incomplete relaxation of the lower oesophageal sphincter. This abnormality is similar to achalasia. Two of these four patients have benefited from pneumatic dilatation with improvement in their severe dysphagia. We believe that pneumatic dilatation should be considered in patients with systemic sclerosis and severe dysphagia where reflux oesophagitis is not apparent.

Low serum testosterone as an indicator of metastatic bronchial carcinoma.

Serum testosterone concentrations were measured preoperatively in 39 men undergoing thoracotomy for histologically proved bronchial carcinoma, in 10 patients with pulmonary opacities that transpired to be non-malignant (benign group) and in 23 men were undergoing minor elective surgical procedures (control group). Thirteen of the 39 patients with known bronchial carcinoma were considered to have had curative surgery and 26 a palliative procedure when operative and pathological findings were taken into consideration. Low serum testosterone concentrations (less than 12 nmol/l) were detected in four patients in the curative group, in 22 in the palliative group (chi 2 test: p less than 0.001), three in the benign group, and in two patients in the control group. A low serum testosterone concentration in patients with bronchial carcinoma may be an indicator of metastatic disease and sequential serum testosterone estimations may prove useful in the follow up of patients thought to have undergone curative surgery.

Wound colonisation following cardiac surgery. Reduction by adjuvant use of preincisional, presternal antibiotic infiltration: a double blind prospective randomised study.

In a double blind, prospective, randomised study of 100 patients undergoing elective cardiac surgery, a significant (p less than 0.01) reduction in wound colonisation, defined as positive culture of any wound discharge irrespective of wound appearance, occurred in those receiving preincisional presternal antibiotic infiltration (2%) as compared to a control group who received a similar volume of normal saline by the same route (24%). Both groups received, in addition, the same conventional intravenous regimen of broad spectrum antibiotic. A comparable concurrent group of patients, not entered into the study, demonstrated a wound colonisation rate similar to the trial control group (22%), thus excluding an adverse bias from the control saline infiltration. Analysis of control cases demonstrated a significant (p less than 0.001) discriminant effect in the degree of preoperative haemodilution with haematocrit falling on bypass by a mean of 25% in those who developed wound colonisation as compared to 13% in those who did not.

Drug sensitivity of non small cell carcinoma of lung by clonogenic assay in several media.

Lung tumours of non small cell pathology were cultured by clonogenic assay in several media. Culture was successful in spleen conditioned medium, but only 57% grew and low plating efficiencies (PE) meant that only 23% of the original number produced significant drug results. Comparison of rat erythrocyte lysate (REL) medium with serum free defined medium (HITES) and HITES + 10% FBS demonstrated clear enhancement of PE in REL although growth was 100% successful in all these media. Ninety-three percent of samples tested against drugs in REL produced significant results. A later comparison of REL with McCoy's 5A + rbc +/- hydrocortisone produced relatively poor culture success for these 3 media and equivocal growth patterns. Low PE was attributed to age of rats used for rbc. Vindesine and cis-platinum cytotoxicity in spleen conditioned medium were 61% and 15% sensitivity respectively. These do not concur with clinical experience but the figures for overt resistance, at 39% and 69%, correspond with expected non-responders to these regimes. Drug testing in REL produced figures correlating more closely with clinical performance at 45% sensitivity to platinum and 36% of patients sensitive to both drugs, but the vindesine sensitivity at 55% is again discrepant with performance of this drug as a single agent.

Mediastinal imaging in lung cancer.

Sixty-six patients with lung cancer underwent mediastinal staging with gallium scanning, CT scanning and mediastinal exploration at mediastinoscopy and/or thoracotomy. Histological findings at time of mediastinal exploration were correlated with the results of the non-invasive staging scans. Gallium scanning had an accuracy of 78.8 per cent and CT scanning had an accuracy of 77.3 per cent. There was no evidence of increased test accuracy by performing both scans in the same individual. Either scanning technique may be utilised as a simple non-invasive mediastinal staging procedure, and where negative it is appropriate to proceed directly to thoracotomy.

Wound prophylaxis in thoracic surgery: a new approach.

A prospective double blind, randomised study was performed in 100 patients undergoing major elective thoracic surgery to assess a new method of prophylaxis of wound infection using one preincisional intraparietal infiltration of cefuroxime sodium along the line of proposed incision as the sole protection against wound infection. A significant (p less than 0.01) reduction in the incidence of wound infection occurred in the antibiotic treated group (2%) compared with the control group (20%), who received by the same route the same volume of saline only. The groups were comparable with respect to age, sex, pathological condition, and operative variables. The use of additional antibiotics was significantly greater in the control group (p less than 0.01), largely owing to a much greater incidence of postoperative pulmonary infection in the control group (60%) than in the antibiotic treated group (40%). No morbidity was associated with this technique. The organisms found in oesophageal and bronchial operative luminal specimens did not correlate with postoperative wound or pulmonary infection or with organisms causing these infections. Reductions in wound and pulmonary infection rates equivalent to those produced by conventional multiple dose parenteral regimens were achieved by this technique.