Sixteen patients with pyogenic osteomyelitis of the ribs are reported; ages ranged from 3 months to 42 years and 10 were female. Right sided ribs were affected in 10 cases and single ribs in 12. Antecedent causes included empyema thoracis (56%) and blunt chest trauma (19%). Fourteen out of 16 patients presented with discharging chest wall sinuses and five patients had associated swellings on the chest wall. The duration of symptoms ranged from two to 36 months. The most common microorganism isolated was Staphylococcus aureus (50%). Rib excision (with drainage of the empyema in two cases) was curative in all 16 cases in this study.Osteomyelitis of the ribs appears to be a rare disease in both the tropical and the temperate areas of the world.13 Adeyokunnu and Hendrickse,4 in a review of 63 cases of salmonella osteomyelitis seen in Nigerian children, found no case of rib osteomyelitis. Similarly, Dich et al1 reviewed 163 cases of osteomyelitis in infants and children over 15 years and found only one case (0.6%) affecting the ribs.Osteomyelitis of the ribs has been found in association with childhood empyema, -7 Staphylococcus aureus infection,1 sickle cell anaemia, and infection due to Salmonella and Klebsiella species and Proteus rettgeri.3 Osteomyelitis arising in ribs adjacent to sites of intubation in the chest in children with empyema has been documented by Osinowo et al.7 The present report reviews 16 cases of rib osteomyelitis treated by our cardiothoracic unit over seven and a half years. From this experience and a review of the published reports we consider the aetiology, pathogenesis, clinical presentation, and results of treatment of the disease. MethodsThe records of all patients with rib osteomyelitis admitted to the unit from January 1975 to May 1983 formed the basis of this prospective study.The data extracted from each patient's record included the age, sex, clinical presentation, radiological findings, mode and results of treatment, and the results of bacteriological and routine haeAddress for reprint requests: Dr 0 Osinowo,

The case of a 46-year-old man with congenital broncho-oesophageal fistula is presented. The patient had had recurrent chest infections since childhood and a recent history of heartburn and flatulence. A barium-swallow examination showed a small sliding hiatal hernia and an oesophageal diverticulum communicating via a fistula with the apical and posterior segmental bronchi of the left lower lobe. Bronchography showed bronchiectasis in this lobe. At thoracotomy resection of the broncho-oesophageal fistula and left lower lobectomy were performed. The patient's recovery was uneventful. Twenty published cases are reviewed. The classification of congenital broncho-oesophageal fistulae should take account not only of the type of fistula but also of the type and size of the bronchus with which the fistula communicates, the age of the patient at the first appearance of symptoms, and the duration of symptoms.The common origin of the respiratory tract and the oesophagus from the same embryonic structures sometimes gives rise to anomalous communications such as tracheo-oesophageal fistula and bronchooesophageal fistula. Most patients with tracheooesophageal fistula associated with oesophageal atresia present in infancy.' In patients who have congenital fistula without oesophageal atresia symptoms are usually insidious in onset and the patients may present in adult life."' The rarity of congenital broncho-oesophageal fistula in adults has been emphasised by various authors and by 1979 there were probably no more than 75 published cases. A barium-swallow examination showed a small sliding hiatal hernia and an oesophageal diverticulum in the middle third of the oesophagus. The contrast medium spilled into the left lower lobe through a fistulous communication between the oesophageal diverticulum and the apical and posterior basal segmental bronchi (fig 1). Bronchography showed bronchiectasis in the left lower lobe, particularly noticeable in the apical and posterior segments. There was no evidence of sequestration or leakage of the contrast medium into the oesophagus. Bronchoscopy showed an inflamed orifice of the apical segmental bronchus of the left lower lobe. At oesophagoscopy the oesophagus appeared normal and the broncho-oesophageal fistula could not be located. Methylene blue (3 ml) injected into the left-lower-lobe bronchus did not appear in the oesophagus.At operation (a left thoracotomy) there was a fistulous communication between the mid-portion of the oesophagus and the apical segment of the lower lobe (fig 2). The fistula was about 2 cm long and 0-4 cm in diameter. The fistula was excised and the oesophageal end repaired. Left lower lobectomy was performed. The small hiatal hernia was not repaired. The postoperative period was uneventful

No abstract

Seventeen patients were treated for major peripheral injuries at the Univer sity College Hospital, Ibadan from May 1977 to April 1984. Stab wounds and blunt trauma were the commonest modes of injury. Diagnosis was based on clinical features with arteriography in 10 patients. Arterial transection was found in 8 patients, thrombosis in 4, arteriovenous fistulae in 3 and aneurysm in 2. The ischaemic time in the acute cases averaged 53.5 hours. Arterial repair was performed in 11 cases, embolectomy in 2, ligation in 2, primary amputation and non-operative treatment in 1 case each. Repair failed in 2 cases of popliteal injuries grafted after 48 hours of ischaemia; both were amputated. Functional results were excellent in 13 patients. One patient died from a delayed rupture from the distal end of an ilio-femoral dacron graft. The need for early presenta tion of cases and avoidance of prosthetic graft is reconfirmed.

Aortic root enlargement with a patch is sometimes indicated either to prevent aortic homograft valve distortion during implantation or to facilitate easy, tension-free closure of the aortotomy. Patches made of prosthetic material have been widely used for this purpose. The use of autogenous pericardium has recently been reported. Although dura mater has been shown to have great strength, low antigenicity, athrombogenicity, easy availability in large sizes and rapid bonding to most tissues, its use for patch enlargement of the aortic root has not been previously documented. From 1979 to 1983, 38 patients had dura mater aortic root gussets placed during aortic valve replacement at the Southampton General Hospital. In all cases, the patches were placed to facilitate aortic closure, or to prevent homograft valve distortion by enlarging the non-coronary sinus. Aortic homografts were implanted in 11 patients, Carpentier Edwards' Xenograft valves in 16, Björk-Shiley valves in 8 and Wessex Xenografts in 3 patients. All the patients survived and in a mean follow-up of 30 +/- 12.8 months (range 3 to 48 months) there has been no clinical evidence of patch failure due to leakage, rupture or aneurysm formation. These results suggest that glycerol-preserved dura mater is a satisfactory patch material for aortic root enlargement during aortic valve replacement.