Systematic small-area variation in mortality for malignant neoplasms in Sweden 1975-1990

Cancer mortality has been stated to be the best single measure of progress in combatting cancer. The variation in total and cause-specific cancer mortality among health administrative areas in Sweden was analysed in order to find out if the mortality outcome has been equally distributed.

Data on underlying causes of death for ages 0-74 years were analyzed for the 26 health administrative areas in Sweden, 1975-1990. Analyses of systematic (non-random) variance were performed and measures chosen in order to make comparisons possible between different malignant neoplasms and different periods of time.

The systematic variance for all malignant neoplasms was reduced by about 40% during the study period. The largest systematic variations in mortality were found for cancer of the oesophagus and lung, although the systematic variation for these neoplasms was lower in the 1980s than in the 1970s. Large systematic variation was also found for cancer of the cervix uteri and of the bladder. For these causes the variation remained constant throughout the period.

Generally speaking, the outcome of cancer has become more equally distributed across the country. Preventive measures should be possible for the malignant neoplasms with the largest regional variation.