J G Hakim

OBJECTIVE: To determine the effect of adjunctive prednisolone on morbidity, pericardial fluid resolution, and mortality in HIV seropositive patients with effusive tuberculous pericarditis. DESIGN: Double blind randomised placebo controlled trial. SETTING: Two medical school affiliated referral hospitals in Harare, Zimbabwe. PATIENTS: 58 HIV seropositive patients aged 18–55 years with tuberculous pericarditis. INTERVENTIONS: All patients received standard short course antituberculous chemotherapy and were randomly assigned to receive prednisolone or placebo for six weeks. MAIN OUTCOME MEASURES: Clinical improvement, echocardiographic and radiologic pericardial fluid resolution, and death. RESULTS: 29 patients were assigned to prednisolone and 29 to placebo. After 18 months of follow up there were five deaths in the prednisolone treated group and 10 deaths in the placebo group. Mortality was significantly lower in the prednisolone group (log rank χ² = 8.19, df = 1, p = 0.004). Resolution of raised jugular venous pressure (p = 0.017), hepatomegaly (p = 0.007), and ascites (p = 0.015), and improvement in physical activity (p = 0.02), were significantly more rapid in the prednisolone treated patients. However, there was no difference in the rate of radiologic and echocardiographic resolution of pericardial effusion. CONCLUSIONS: Adjunctive prednisolone for effusive tuberculous pericarditis produced a pronounced reduction in mortality. It is suggested prednisolone should be added to standard short course chemotherapy to treat HIV related effusive tuberculous pericarditis.

Echocardiographic study of sixty patients on maintenance haemodialysis (MHD) was undertaken to determine the prevalence and factors associated with left ventricular (LV) hypertrophy (LVH), LV diastolic dysfunction and pericardial disease. The mean age was 34.4 (standard deviation 13.0), range 14-66 years with 31 (51.7%) men. LVH was found in 41 (68%) patients. Of the factors analysed, serum calcium and calcium-phosphate product were significantly associated with LVH (t = 2.01, df = 58, p = 0.046; t = 2.18, df = 58, p = 0.032 respectively). Hypertension in this study was not significantly associated with LVH (p = 0.169). LV diastolic dysfunction was found in 23/41 (56%) patients with LVH, and in 9/19 (47%) patients without LVH (difference is not statistically significant, X2 = 0.12, df = 1, p = 0.725). Small pericardial effusions were detected in 4/60 (7%) patients and two patients had pericardial thickening. We conclude that in our MHD patients LVH, is very common and that diastolic dysfunction is observed equally in patients with and without LVH. However, haemodynamically significant pericardial effusions are rare in patients who have been on dialysis for at least six months.

To estimate non-condom user rate and to characterize non-condom users among female nurses. Cross sectional. Health institutions. Data from 640 (86.5pc) out of 740 consenting female nurses were available for analyses. Non-condom user rate. Non-condom user rate (per 100) among the female nurses was 73.1pc (95pc CI69.7 to 76.5). Non-condom users tended to be aged above 30 years (OR 1.57; 95pc CI 1.02 to 2.40), midwives (OR 1.56; 95pc CI 1.03 to 2.37), married (OR 2.70; 95pc CI 1.73 to 4.21), not to think that spouse/partner has had sex with other partners (OR 1.72; 95pc CI 1.11 to 2.68) and to have had a sexually transmitted disease (OR 2.61; 95pc CI 1.25 to 5.43). There was a surprising high level of non-condom users among female nurses, probably due to the inability for females to initiate or negotiate condom use.

To determine whether the factors associated with human immunodeficiency virus (HIV) infection are the same among the general nurses, nurse midwives and office workers/teachers. Cross sectional. Schools, offices and hospitals. A total of 370 consenting females from each of the study groups. HIV seropositivity, medical and behavioural risk factors for HIV infection. HIV seropositive general nurses were more likely to report a history of sexually transmitted disease since 1980 (OR 5.17; 95pc CI 1.60 to 15.05), particularly genital ulcer (OR 3.59; 95pc CI 1.30 to 10.85), and they were also more likely to have more sexual partners (OR 4.08; 95pc CI 1.54 to 13.11). HIV seropositive nurse midwives were more likely to have had a blood transfusion (OR 2.17; 95pc CI 1.13 to 4.14), and HIV seropositive office workers/teachers were more likely to have scarification (OR 1.66; 95pc CI 1.07 to 2.57). Our findings support the importance of genital ulcers, blood transfusion and scarification as risk factors for HIV intervention studies should be specifically targeted to the sector of the population under study.

From 1988 to 1993 (six years), 127 suspected cases of acute myocardial infarction (AMI) were admitted to the Parirenyatwa Hospital coronary care unit. AMI was confirmed in 76 cases, 37 were Black, 27 White, six Indian and six Coloured. For Blacks the male to female ratio was 5:1. The clinical and laboratory features and complications of AMI were similar in all ethnic groups. Compared to other groups, Blacks presented to hospital late, an observation which has important implications for thrombolytic therapy. With the increasing number of cases of AMI now being seen among Black Zimbabweans, the time has come for the evaluation of the changing risk factor profile and the initiation of education and intervention programmes which could contain this rise before it spirals into a major health problem.

Fine needle aspiration (FNA) of the liver without ultrasound guidance was performed on 110 patients with hepatocellular carcinoma (HCC). The median age was 52 years, with a range of 16 to 86 years. There were 90 males and 20 females (a male: female ratio of 4.5:1), with a median age of 51.5 years (range 16 to 86 years) and 55.5 years (range 17 to 72 years) respectively. FNA was reported as showing malignancy in 92 (84 pc, 95 pc CI 77 to 91 pc) patients; 80 (73 pc) were definite HCC, 12 (11 pc) were malignant unspecified, seven (6 pc) were suspicious of malignancy, seven (6 pc) had no malignant cells and four (4 pc) were non-diagnostic. The only complication observed was dizziness in one patient. We conclude that FNA of the liver for the diagnosis of HCC is a safe, simple and accurate procedure which can be undertaken in settings that would otherwise not be suitable for formal liver biopsy.