Abstract
The possibilities of clustering between those electoral wards which display higher than expected incidences of cases of the lymphomas occurring between 1978 and 1982 are examined. Clusters are defined as being those wards with cases in excess (at a probability of less than 10%) which are geographically adjacent to each other. A separate analysis extends the definition of cluster to include high incidence wards that are adjacent or separated by one other ward. The results indicate that many high incidence lymphoma wards do occur close together and when computer simulations are used to compute expected results, many of the observed results are shown to be highly improbable both in the overall number of clustering wards and in the largest number of wards comprising a 'cluster'.
Full text
PDF
Selected References
These references are in PubMed. This may not be the complete list of references from this article.
- Barnes N., Cartwright R. A., O'Brien C., Richards I. D., Roberts B., Bird C. C. Rising incidence of lymphoid malignancies--true or false? Br J Cancer. 1986 Mar;53(3):393–398. doi: 10.1038/bjc.1986.64. [DOI] [PMC free article] [PubMed] [Google Scholar]
- Chen R., Mantel N., Klingberg M. A. A study of three techniques for time-space clustering in Hodgkin's disease. Stat Med. 1984 Apr-Jun;3(2):173–184. doi: 10.1002/sim.4780030210. [DOI] [PubMed] [Google Scholar]
- Craft A. W., Openshaw S., Birch J. M. Childhood cancer in the Northern Region, 1968-82: incidence in small geographical areas. J Epidemiol Community Health. 1985 Mar;39(1):53–57. doi: 10.1136/jech.39.1.53. [DOI] [PMC free article] [PubMed] [Google Scholar]