PROBABILITY APPROACH FOR IDENTIFYING A HIGH-RISK GROUP OF WOMEN FOR CERVICAL CYTOLOGY SCREENING

Cancer of uterine cervix is one of the leading malignancies seen in In dian women, In view of limited resources, large-scale cervical cytolog y is anear impossibility, thus the present communication attempts to s tudy the relevance of some sociodemographic and clinical variables in short listing the cases for cervical cytology by applying multivariate logistic regression analysis, It was observed that age, parity and ed ucation revealed statistical significance with multivariate odds ratio s of 4.8, 3.5 and 2.6 respectively, The presence of either clinical sy mptoms, i.e. cervical erosion that bleeds on touch, suspicious-looking cervix or unhealthy cervix, revealed an odds ratio of 15, It has been estimated that it is possible to predict about 60% of malignancies co rrectly by utilising the above variables, Thus a strategy of selective screening could be undertaken in the developing countries until resou rces are generated for large-scale cytology screening for cervical can cer.