摘要
Objective:Gastriccancer(GC)isoneoftheleadingcausesofdeathinChinaandotherAsiancountries.Recently,gastricendoscopyhasbecomethemainapproachforGCscreening,buttheidentificationofhigh-riskindividualsremainsachallengeinGCscreeningprograms.Methods:Therewere7,302patientswithchronicgastritisinvolvedinthisstudy.Endoscopicexaminationswereperformed,andtheirdemographiccharacteristicsandlifestyledatawerecollected.EachpossibleassociatedfactorofGC/premalignantandprecursorlesionswasevaluatedbyunivariateandmultivariatelogisticregressions.Nomogramswereusedforvisualizationofthosemodels,andreceiveroperatingcharacteristic(ROC)curveanalysiswasusedtopresentthepredictiveaccuracy.Results:Wedetected8(0.11%)gastricadenocarcinomas,17(0.23%)dysplasiacases,14(0.19%)hyperplasiacases,52(0.71%)intestinalmetaplasiacases,217(2.97%)inflammatorylesions,141(1.93%)gastriculcers,10(0.14%)atrophicgastritiscases,1,365(18.69%)erosivegastritiscases,and5,957(81.58%)superficialgastritiscasesin7,302patients.Theage(P<0.001),gender(P=0.086),laborintensity(P=0.018)andleekfoodintake(P=0.143)wereidentifiedasindependentpredictivefactorsofGC/premalignantlesionspossibility.Thecorrespondingnomogramexhibitedanareaunderthecurve(AUC)[95%confidenceinterval(95%CI)]of0.82(0.74–0.89)forthemodelinggroupand0.80(0.75–0.85)forthevalidationgroup.Theage(P=0.002),gender(P=0.024),smoking(P=0.002)andleekfoodintake(P=0.039)wereindependentpredictivefactorsofprecursorlesionspossibility.ThecorrespondingnomogramexhibitedanAUC(95%CI)of0.62(0.60–0.65)forthemodelinggroupand0.61(0.59–0.63)forthevalidationgroup.Conclusions:WeidentifiedseveralpotentialassociatedfactorsandprovidedapreclinicalnomogramwiththepotentialtopredictthepossibilityofGC/premalignantandprecursorlesions.
出版日期
2017年04月14日（中国期刊网平台首次上网日期，不代表论文的发表时间）
