Evaluation On Primary Screening And Preliminary Results Of Mass Screening For Colorectal Cancer In China

Backgrounds and objectiveColorectal cancer is the most common gastrointestinal tumor.According to GLOBOCAN 2012 from International Agency for Research on Cancer(IARC)of World Health Organization,about 1,361,000 new cases and 694,000 deaths of colorectal cancer were estimated in 2012.At present,mass screening programs for colorectal cancer have been implemented worldwide,while the strategies were different.In China,the target population for screening was the people aged 40-74 years old.High risk factors questionnaire(HRFQ)was conducted in primary screening.Besides,colorectal cancer quantitative risk-assessment model(QRAM)was considered as a kind of screening methods in pilot area.However,comparison of these methods was controversial.Jiashan County was recognized as a high risk area of colorectal cancer.There have been three mass colorectal cancer screening programs so far.The third screening program was launched in 2007,whereas there was no report about the performance and effect of this screening program systematically.To address these unresolved issues,we evaluated the present primary screening methods so as to provide references for the future screening program.We also analyzed the time trend changes of morbidity and mortality between screening and non-screening areas using quasi-stepped wedge design,in order to make a preliminary assessment of early effect of screening.Material and methodsExcluding 23,675 people who are not available for screening,a total of 113,292 permanent residents aged 40-74 were enrolled in the screening from 2007 to 2013.Based on data extracted from screening program records and Cancer Registration System of Jiashan County,we acquire screening information of each screening object,and the incidence and death cases were ascertained until December 2014.Using the colorectal cancer report records as criteria,we compared the sensitivity and specificity among iFOBT only,iFOBT&HRFQ,and iFOBT&QRAM.Meanwhile,the accuracy of the diagnosis of colorectal cancer patients in 3,5,and 7 years were analyzed.With the adjustment of natural growth rates,the morbidity and mortality of screening and non-screening areas were calculated using the quasi-stepped design.We compared the time trend of incidence and death rates in screening and non-screening area and calculated the annual percentage change(APC).In addition,stratified analyses were performed to minimize the impact of sex and age.The statistical analysis was performed with SAS Version 9.4(SAS Institute,Gary,NC,USA)and Microsoft Excel 2010.The trend analysis utilized Joinpoint Regression Program(Version 4.1.1.3-December 2014)developed by the American National Cancer Institute.The significance level was P<0.05.ResultsIn our study,a total of 113,292 residents(56,524 men and 56,768 women)were enrolled in the screening,with the compliance rates of 82.71%(men:81.93%;women:83.51%).There was significant difference between the genders(P<0.05).Compared iFOBT only,iFOBT&QRAM had a higher sensitivity and a lower specificity,while iFOBT&HRFQ had no impact on the performance of screening.The method of QRAM along with iFOBT showed a higher sensitivity in predicting the incidence risk after screening,while iFOBT only or combination of the iFOBT and HRFQ had a lower sensitivity,with false negative rate of over 70%.The natural growth rates from 1997 to 2006 in the population aged over 40 years old were analyzed.The morbidity increased during the decade,with the APC of 3.21%,while the trend of mortality was not significant.The increasing trend of morbidity remained to be 5.02%in male.However,we observed no time trend of morbidity in female and mortality in both genders.The morbidity and mortality showed no time trend after screening,while the corresponding rates increased in non-screening areas,with APC of 10.07%and 9.35%,respectively.The increasing trend of incidence rates in screening areas remained to be 9.04%in male and 10.63%in women,while there were no significant time trends in non-screening areas between both genders.The mortality showed no temporal trend in men and women.There was a similar variation trend of morbidity at age 40-49 group,while 82.14%reduction of death rate can attribute to screening in this group.For age 50-59 group,there was an increasing trend of morbidity in screening area and opposite trend in non-screening area.Screening can contribute to 70%reduction of incidence rate at age 60-69 group.For the people aged over 70 years old,screening can lead to a 42.75%reduction of morbidity and a stabilized mortality,while both of rates were increasing in non-screening areas.ConclusionGenerally,this program was well-organized and widely promoted with high compliance rate.QRAM&iFOBT showed a higher sensitivity,flexibility and applicability for colorectal cancer or forecasting short-term incidence.Screening has the effect on morbidity and mortality of colorectal cancer,especially for morbidity in early period.Men were more sensitive than women in screening.For the young,screening had great influence on death rates,and the effect on incidence rate increased with age.