Improvement of health care policy reguires an assessment of health car
e practices. In France, morbidity registries might be the best tool of
such an assessment. This study shows how the treatment of colorectal
cancer can be assessed by French cancer registries. Two studies were c
onducted. One cross-sectional on data from 7 cancer registries in 1990
and one longitudinal on data from 2 digestive cancer registries (depa
rtments of Calvados and Cote-d'Or) between 1978 and 1990. Cancer resec
tion rate significantly increased from 1978 so 1990. In 1990, mean res
ection rate was 85%, higher for colon (89%) than rectum (79%). Sphinct
er preservation rate also significantly increased to 40.8% in 1990. No
geographical variations were observed for surgical practices. The use
of adjuvant radiotherapy had significantly increased between 1978 and
1990, more rapidly in university centres. In more recent years, the u
se of radiotherapy concerned 50% of resected rectal cancers without di
fferences between the different types of health care centre. However,
in 1990, great geographical variations were observed in the use of adj
uvant radiotherapy. Similar geographical variations were observed for
the use of chemotherapy which did nor increased with time. Colorectal
cancers were diagnosed more and more earlier from 1978 to 1990 in both
departments of Calvados and Cote-d'Or. The French network of French c
ancer registries (FRANCIM) provides accurate and reliable knowledge on
medical practices, geographical variations and trends. Information is
also available on prognosis in cancer patients. Potentialities of can
cer registries are not actually well known although such information i
s required to plan health care policy.