VARIATION IN SURVIVAL OF CHILDREN WITH CANCER WITHIN A REGION OF THE UNITED-KINGDOM

BACKGROUND. A population-based study of survival of 678 children, age less than 15 years, diagnosed with cancer in the Southwest region of t he United Kingdom showed an improvement in 5-year survival from 53% be tween 1976 and 1980 to 64% between 1981 and 1985 (P = 0.008). Survival varied significantly among the five counties that make up the region (P = 0.0008); the differences were greatest for central nervous system (CNS) tumors. METHODS. The expertise of local hospitals and entrance into national trials was examined. The region has nine hospitals actin g as primary treatment hospitals for children with cancer, two neurosu rgical units, and one regional oncology unit. Four of the primary hosp itals saw more than six children per year and were designated ''large' ' centers. Five saw fewer than six patients per year and were designat ed ''small'' hospitals. RESULTS. There was a significant difference in survival of patients according to the number of patients treated per year per hospital. A hospital that treated fewer than six patients per year was designated ''small'' and six or more designated ''large.'' A t 5 years, patients with CNS tumors experienced a 58% survival rate in large hospitals and a 41% survival rate in small hospitals. (P = 0.03 ). The rate of entrance into trials for all malignancies did not diffe r between large (30%) and small (27%) centers. Only 5% of children wit h CNS tumors were placed on national trials. There was no difference i n survival rates for CNS tumors in the two neurosurgical centers. CONC LUSIONS. There were significant differences in survival by county for children with cancer, especially those with CNS tumors. In the first h alf of the study decade, in hospitals treating fewer than six cases of childhood cancer per year, children with CNS tumors were rarely place d on national protocols and did poorly. (C) 1996 American Cancer Socie ty.