The value of hyperthermia as an adjuvant to radiotherapy in patients with malignant melanoma was studied in a European multicentre trial. 134 metastatic or recurrent lesions of malignant melanoma in 70 patients were randomly assigned to receive radiotherapy (three fractions of 8 Gy or 9 Gy in 8 days) alone or followed by hyperthermia (43 degrees C for 60 min). Overall, the 2-year actuarial local tumour control was 37 (SE 5)%. Univariate analysis showed a beneficial effect of hyperthermia (radiation alone 28% vs combined treatment 46%, p = 0.008) and radiation dose (24 Gy 25% vs 27 Gy 56%, p = 0.02), but no effect of tumour size (< or = 4 cm 42% vs > 4 cm 29%, p = 0.21). Cox multivariate regression analysis showed the most important prognostic variables to be hyperthermia (odds ratio for 2-year local control 1.73 [95% CI 1.07-2.78], p = 0.023), tumour size (0.91 [0.85-0.99], p = 0.05), and radiation dose (1.17 [1.01-1.36], p = 0.05). Addition of heat did not significantly increase acute or late radiation reactions. Heating was well tolerated, but because of difficulties with equipment only 14% of treatments achieved the protocol objective. The overall 5-year survival rate was 19%, but 38% of the patients for whom all known disease was controlled survived 5 years. Adjuvant hyperthermia significantly improved local tumour control when applied in association with radiation in treatment of malignant melanoma. Successful local treatment of patients with a single or a few metastatic malignant melanoma lesions has significant curative potential.

Feedback can be described as a way to provide information on doctors' performance to enable changes in future behaviour. Feedback is used with the aim of changing test-ordering behaviour. It can lead to reductions in test usage and cost savings. It is not sufficiently clear, however, whether feedback leads to more appropriate test use. Since 1985, the Diagnostic Coordinating Center Maastricht has been giving feedback on diagnostic tests as a routine health care activity to all family doctors in its region. Both quantity and quality of requests are discussed. In a randomised, controlled trial over 2.5 years, discussion of tests not included previously was added to the existing routine feedback. One group of family doctors (n = 39) received feedback on test-group A (electrocardiography, endoscopy, cervical smears, and allergy tests), the other (n = 40) on test-group B (radiographic and ultrasonographic tests). Thus, each group of doctors acted as a control group for the other. Changes in volume and rationality of requests were analysed. The number of requests decreased during the trial (p = 0.036). Request numbers decreased particularly for test-group A (p = 0.04). The proportion of requests that were non-rational decreased more in the intervention than in the control groups (p = 0.009). Rationality improved predominantly for test-group B (p = 0.043). Thus, routine feedback can change the quantity and quality of requests.

Interventions to avoid atherosclerosis might be more successful if launched early in life when eating and life-style patterns are formed, but dietary interventions have been limited by fears of diet-induced growth failure. We investigated the effects of a diet low in saturated fat and cholesterol on serum lipid concentrations and growth in 1062 healthy 7-month-old infants in a randomised study. Every 1-3 months, families in the intervention group received dietary advice aimed at adequate energy supply, with low fat intake (30-35% energy, polyunsaturated/monounsaturated/saturated fatty acid ratio 1/1/1, and cholesterol intake < 200 mg daily). Infants in control families consumed an unrestricted diet. 3-day food records were collected at ages 8 and 13 months. Growth was carefully monitored. Between 7 and 13 months serum cholesterol and non-high-density-lipoprotein cholesterol concentrations did not change significantly in the intervention group (mean change -0.03 [SD 0.72] mmol/L and 0.01 [0.67] mmol/L) but increased substantially in the control group (0.24 [0.64] mmol/L and 0.23 [0.60] mmol/L; p for difference in mean changes between groups < 0.001). Daily intakes of energy and saturated fat were lower in the intervention than in the control group at 13 months (4065 [796] vs 4370 [748] kJ, p = 0.033, and 9.3 [3.5] vs 14.5 [4.8] g, p < 0.001, respectively), and intake of polyunsaturated fat was higher (5.8 [2.2] vs 4.4 [1.4] g, p < 0.001). Growth did not differ between the groups and was as expected for children at this age. Serum cholesterol concentrations fell significantly in parents of intervention-group infants. The increases in serum cholesterol and non-high-density-lipoprotein cholesterol concentration that occur in infants between the ages of 7 and 13 months can be avoided by individualised diets, with no effect on the children's growth.

Lamotrigine has been licensed widely as adjunctive therapy for partial and secondary generalised seizures. Use of the drug as monotherapy was investigated in a double-blind, randomised, parallel-group comparison with carbamazepine in newly diagnosed epilepsy. After 4 weeks of planned, fixed dose escalation, doses were adjusted according to efficacy, adverse events, and plasma concentrations. 151 of 260 patients (131 lamotrigine, 129 carbamazepine) in eight UK centres completed the 48-week trial. No differences in efficacy between the drugs were found for partial seizures with or without secondary generalisation or for primary generalised tonic-clonic seizures. The proportion of patients maintained seizure-free during the last 24 weeks of treatment was almost the same in both groups (39% lamotrigine, 38% carbamazepine). More patients with primary generalised tonic-clonic seizures (47% both groups) than those presenting with a focal onset (35%, 37%) were fully controlled. Overall, fewer patients on lamotrigine than on carbamazepine withdrew because of adverse events (15 vs 27%). The commonest side-effect leading to withdrawal with either drug was rash (9%, 13%). Sleepiness was less common in lamotrigine than in carbamazepine recipients (12 vs 22%, p < 0.05). More lamotrigine than carbamazepine recipients (65 vs 51%, p = 0.018) completed the study (hazard ratio 1.57 [95% CI 1.07-2.31]). Lamotrigine and carbamazepine showed similar efficacy against partial onset seizures and primary generalised tonic-clonic seizures in newly diagnosed epilepsy. Lamotrigine, however, was better tolerated.

We investigated the effect of heparin on urinary albumin excretion in patients with insulin-dependent diabetes mellitus. 39 patients with persistent urinary albumin excretion of 30-300 mg/24 h were randomly treated for 3 months with subcutaneous injections twice daily of isotonic saline, 5000 IU unfractionated heparin, or 2000 anti-Xa IU low-molecular-weight heparin. Unfractionated and low-molecular-weight heparin induced a significant reduction in urinary albumin excretion (p = 0.04 and p = 0.004). The mechanism and clinical relevance is unknown but deserve further attention.

Case management arose in the USA as a solution to the difficulties of providing community care to people with severe mental disorders. The basic principle of the approach is that a case manager takes responsibility for a client; arranges an assessment of need, a comprehensive service plan, delivery of suitable services, and monitoring and assessment of services delivered. The case-management approach has been widely accepted, to the extent that recent legislation has made case-management the cornerstone of community care in the UK. We did a randomised controlled trial to evaluate a social services case-management team for people with long-term mental disorders. Subjects were referred from hostels for the homeless, night shelters, a general-practitioner clinic for the homeless, the Oxford City Council homelessness unit, and local voluntary-sector group homes. Of 103 subjects referred, 80 consented to be randomised to treatment or control groups. At 14-month follow-up, as assessed by standardised interviews, there were no significant differences between groups in number of needs, quality of life, employment status, quality of accommodation, social behaviour, or severity of psychiatric symptoms. In the case-management group there was a significant reduction in deviant behaviour on a standardised behaviour rating scale (REHAB) (mean = 0.79; 95% CI 0.26-1.32). It is unfortunate, in view of the limited effectiveness we have shown, that social services case-management was not evaluated in randomised controlled trials before its implementation in the UK.

The efficacy of adjuvant chemotherapy after surgery for colorectal cancer remains unproven. We have investigated the efficacy of a perioperative intraportal cytotoxic regimen in a randomised trial of 533 patients with operable colorectal carcinoma. Patients were randomly assigned either a single course of portal infusion with mitomycin (10 mg/m2, one dose) plus fluorouracil (500 mg/m2 per 24 h for 7 days) starting immediately after surgery, or no adjuvant treatment. 505 (94%) were evaluable. At median follow-up of 8 years, adjuvant therapy reduced the risk of recurrence by 21% (hazard ratio 0.79 [95% CI 0.62-1.00], p = 0.051) and the risk of death by 26% (0.74 [0.57-0.97], p = 0.026). The lower risk of relapse was observed in all subgroups based on node status or localisation of the tumour; the risk reduction was greatest in patients with tumour-involved lymph nodes (Dukes' C; 0.67 [0.45-0.99], p = 0.045) and for those with colon cancer (0.78 [0.56-1.09], p = 0.151). Most of the difference in overall and disease-free survival could be attributed to a consistent reduction of all kinds of tumour recurrences (local relapses, liver metastases, and other distant metastases) in the treated group, rather than to a reduction of liver relapses only. We conclude that part of the benefit obtained with a single course of adjuvant chemotherapy via the portal vein for patients with operable colorectal carcinoma might be due to the systemic effects of the portal chemotherapy.

In animals, intermittent sympathomimetic stimulation with dobutamine produces benefits analogous to those of physical conditioning. Longer intermittent or continuous beta-stimulant therapies have not, however, been successful in managing patients with chronic heart failure. We have investigated the role of beta-receptor stimulants in patients with severe chronic heart failure by changing the method of administration to intermittent, very short-duration pulsed intrope therapy (PIT). We studied 10 patients (mean age 64 [SE 2] years) with stable moderate to severe chronic heart failure (ejection fraction 23 [3]%) who received PIT, and 10 control patients matched for age and severity. We infused sufficient dobutamine to raise heart rate to 70-80% maximum for 30 min per day, 4 days per week for 3 weeks. PIT increased exercise tolerance (from 10.4 [1.2] min at baseline to 13.0 [1.5] min at 3 weeks; p < 0.001, 95% CI for difference 1.6 to 3.9) and lowered peripheral vascular resistance (19.8 [3.1] to 17.7 [2.4] mm Hg.min.L-1; p < 0.05, -4.1 to -0.1). PIT produced significant increases in lymphocyte beta-receptor density (502 [110] to 1200 [219] per cell, p < 0.02, 258 to 1138) and chronotropic responsiveness to exercise (change in heart rest to peak exercise 51.0 [3.2] to 57.5 [3.9] beats per min; p < 0.01, 2.9-10.1). Plasma noradrenaline concentrations (2.39 [0.28] to 1.65 [0.19] nmol/L, p < 0.05) were reduced. The patients' symptoms were also improved. By contrast, no change in autonomic function or exercise capacity was seen in the control group. Short-duration PIT induces pharmacological conditioning with improved symptoms, autonomic balance, exercise tolerance, beta-receptor up-regulation, and enhanced chronotropic responsiveness in chronic heart failure.

In developed countries, use of oral rehydration salts (ORS) solution with osmolarity lower than that of plasma has been recommended because of the risk of hypernatraemia. We compared the clinical efficacy of reduced-osmolarity ORS and standard ORS solutions in children with acute diarrhoea in four developing countries. 447 boys aged 1-24 months, admitted to hospitals in four countries with acute diarrhoea and signs of dehydration, were randomly assigned either standard ORS (311 mmol/L) or reduced-osmolarity ORS (224 mmol/L) solution. Total stool output was 39% greater (95% CI 11-75), total ORS intake 18% greater (3-33), and duration of diarrhoea 22% longer (2-45) in the standard ORS group than in the reduced-osmolarity ORS group. The risk of requiring intravenous infusion after completion of the initial oral rehydration was greater in children given standard ORS solution than in those given reduced-osmolarity ORS solution in three of the four countries (all-country relative risk 1.4 [0.9-2.4]). This relative risk was significantly increased only in non-breastfed children (2.0 [1.0-3.8], p < 0.05). In breastfed children, the relative risk of requiring intravenous infusion was not affected by the ORS solution (0.9 [0.4-2.0]). The mean sodium concentration 24 h after admission was significantly lower in the reduced-osmolarity ORS group than in the standard ORS group (135 [134-136] vs 138 [136-139] mmol/L, p < 0.01). Reduced-osmolarity ORS solution has beneficial effects on the clinical course of acute diarrhoea. Our findings support the use of reduced-osmolarity ORS solution in children with acute non-cholera diarrhoea in developing countries. Further studies are needed to find the best formulation and whether such a solution would be satisfactory for the treatment of cholera.

Screening and carotid endarterectomy have been advocated for asymptomatic carotid stenosis. However, the risk of stroke without treatment has not been adequately defined. We investigated the risk of stroke in the distribution of the asymptomatic carotid artery in 2295 patients randomised in the European Carotid Surgery Trial. During a mean follow-up of 4.5 years, there were 69 carotid territory strokes, nine of which were fatal, giving three year Kaplan-Meier risks of stroke and fatal stroke of 2.1% (95% Cl, 1.5-2.8) and 0.3% (95% Cl, 0.06-0.56) respectively. The stroke risk in the 127 patients with severe (70-99%) carotid stenosis was 5.7% (95% Cl, 1.5-9.8). Given these low stroke risks, the potential benefit of endarterectomy for asymptomatic carotid stenosis is small. Population screening is not justified and endarterectomy for asymptomatic carotid stenosis should only be performed in the context of well organised randomised controlled trials.

Observational studies have suggested that the prevalence of trachoma is lower in children with clean faces than in those with ocular or nasal discharge or flies on the face. We carried out a community-based randomised trial in three pairs of villages to assess the impact on trachoma of a face-washing intervention programme following a mass topical antibiotic treatment campaign. Six villages in Kongwa, Tanzania, were randomly assigned mass treatment plus the face-washing programme or treatment only. 1417 children aged 1-7 years in these villages were randomly selected and followed up for trachoma status and observations of facial cleanliness at baseline and 2, 6, and 12 months. At 12 months, children in the intervention villages were 60% more likely to have had clean faces at two or more follow-up visits than children in the control villages. The odds of having severe trachoma in the intervention villages were 0.62 (95% Cl 0.40-0.97) compared with control villages. A clean face at two or more follow-up visits was protective for any trachoma (odds ratio 0.58 [0.47-0.72]) and severe trachoma (0.35 [0.21-0.59]). This community-based participatory approach to face-washing intervention had variable penetration rates in the villages and was labour intensive. However, we found that, combined with topical treatment, community-based strategies for improving hygiene in children in trachoma-endemic villages can reduce the prevalence of trachoma.

The UK Medical Research Council trial MRC UKALL X was designed to investigate the benefit of one or two courses of additional intensification therapy in children with acute lymphoblastic leukaemia receiving standard treatment. From 1985 to 1990 1612 children, comprising more than 90% of eligible cases in the UK, were treated with intensive induction therapy, central nervous system directed therapy with cranial irradiation and intrathecal methotrexate, and continuing treatment for 2 years. 1171 children were randomised to receive additional intensification therapy at 5 weeks, 20 weeks, both, or neither. At follow-up of at least 3 years disease-free survival for all children at 5 years was 62% (95% confidence interval [Cl] 60.0-64.4), a significant improvement over the 56% (53.0-59.6) found in the preceding MRC UKALL trial. The 5-year disease-free survival was 71% (65.5-76.1) for children randomised to two blocks of intensification therapy, this being significantly better than the 62% (56.6-68.0), 61% (55.7-67.1), and 57% (50.9-62.7) rates for the groups randomised to one intensification block at 5 weeks, one at 20 weeks, and no intensification, respectively. The benefits of intensification therapy were seen irrespective of clinical factors known to influence outcome such as age, sex, and initial leucocyte count. We conclude that the addition of two courses of intensification therapy has produced a 14% improvement in disease-free survival and an 11% improvement in overall survival for the randomised patients. This additional treatment is of benefit to all children with acute lymphoblastic leukaemia, even those traditionally deemed at lower risk of relapse.

Approximately 15% of registered pregnancies end in spontaneous abortion, and for 50 years or so dilatation and curettage (D&C) has been the usual management. In a prospective randomised trial we compared the clinical results after either expectant management or D&C, for miscarriages of less than 13 weeks' gestation in which transvaginal ultrasound examination showed intrauterine tissue and/or blood clots with a diameter 15-50 mm. 103 patients were randomised to expectant management and spontaneous resolution of pregnancy occurred within 3 days in 81 cases (79%). 52 patients were randomised to D&C. 3 infections were diagnosed among patients who underwent expectant management (3%); 5 infections and 1 case of postoperative anaemia were observed among patients randomised to D&C (11%). The duration of vaginal bleeding was a mean of 1.3 days longer in the expectant management group (p < 0.02). Convalescence time, time during which patients experienced pain, and packed-cell volume 3 and 14 days after inclusion did not differ significantly between the groups. This study shows that expectant management of selected cases of spontaneous abortion has a similar outcome to D&C.