Influence of adequacy of the sample on detection of the precursor lesions of the cervical cancer

PURPOSE: to evaluate whether the sample adequacy influences the detection of precursor cervical cancer lesions. METHODS: a transversal study from January 2004 to December 2005. A number of 10,951 results of cervical cytotopathological exams from users of the National Health System (Sistema Unico de Saude, SUS) in Goiania, Goias, Brazil, was studied. These women had spontaneously looked for the services from the Family Health Program or from the Basic Units of Health. Samples were collected by medical doctors and nurses, through the conventional technique to detect cervical cancer. The analyzed smears were classified by the Bethesda System, the sample adequacy being defined along the routine screening and categorized as: satisfactory, satisfactory but presenting factors that might partially jeopardize the analysis, and unsatisfactory. Results were stored in the Epi-Info 3.3.2 program. The chi(2) test was used to compare altered results with the adequacy of the samples from cytopathological smears. Differences with probability of rejection of the null hypothesis lower than 5% (p< 0.05) were considered as significant. RESULTS: From 10,951 smears, 51.1% were classified as having satisfactory adequacy for analysis, 46.6% as satisfactory, but presenting some limiting factors, and 2.3%, as unsatisfactory. The main factors which have partially jeopardized the analysis were: lack of endocervical cells (52.2%), dried smears (22.8%), purulence (14.9%), or smears with some thick areas (9.5%). There was a higher rate of altered smears when the sample had been classified as satisfactory for analysis and with representation of endocervical cells ASC-US (2.3%), ASC-H (0.6%), LSIL (3.2%), HSIL (1.7%) and 0.3% of AGC. Differences were significant when p=0.001. The rate of low and high grade lesions was higher when the smears were satisfactory for analysis. CONCLUSIONS: the rate of precursor uterine cervix cancer lesions varies according to the sample adequacy, and the main adequacy limitations of the sample are mainly related to the collection condition.