Hematochezia In Hospitalized Children Older Than 28 Days:Clinical Analysis Of 1122 Cases

ObjectiveTo explore the etiology and related influencing factors through a retrospective analysis of hematochezia in children older than 28 days admitted to our hospital in the past year,so as to improve the cognition,diagnosis,therapy and prognosis of children's hematochezia.Objects and methodsAccording to the inclusion and exclusion criteria,1122 children older than 28 days with chief complaint of hematochezia admitted to Children's Hospital of Chongqing Medical University during October 1,2017 to October 1,2018 were included and retrospectively analyzed by the statistical software SPSS 21.0.Excel 2016 was used to draw charts.Results1.Among 1122 cases of hematochezia in children,LGIB of 793 cases?70.7%?was the main hemorrhage site,followed by 223 cases?19.9%?of systemic diseases,90 cases?8.0%?of UGIB,and 16 cases?1.4%?of OGIB.The top 10 etiologies of hematochezia were successively intussusception of447cases?39.8%?HSP of 160 cases?14.3%?,enteritis of 115 cases?10.3%?,intestinal polyp of 94 cases?8.4%?,crissum diseases of 77 cases?6.9%?,digestive tract malformation of 58 cases?5.2%?,PU of 46 cases?4.1%?,hematological diseases of 43 cases?3.8%?,EGV of 27 cases?2.4%?,and OGIB of 16 cases?1.4%?.2.There were statistically significant differences in the hemorrhage sites and etiology of hematochezia in children among different age groups??2=345.280,P<0.001;?2=791.217,P<0.001?.LGIB was dominant in infant group,toddler group and pre-school group,while systemic diseases were dominant in school-age group and adolescent group.The top 3etiologies in each age group were:intussusception,enteritis and digestive tract malformation in infant group;intussusception,intestinal polyp,enteritis in toddler group;HSP,intestinal polyp,crissum diseases in preschool group;HSP,crissum diseases,intestinal polyp in school-age group;HSP,PU,hematological diseases in adolescent group.3.There was no statistically significant difference in the hemorrhage sites and etiology between different genders??2=5.022,P=0.170;?2=15.536,P=0.077?,the top 3 etiologies of hematochezia were intussusception,HSP,and enteritis in both male and female group.4.There were statistically significant differences in the course of hematochezia and hospital days among different age groups?H=273.647,P<0.001;H=200.595,P<0.001?.The course of hematochezia in the infant group was the shortest among the five age groups,and the length of hospital days was shorter than that in the pre-school age,school-age and adolescent groups.Meanwhile,there were also statistically significant differences in the course of hematochezia and hospital days among different etiologies?H=682.423,P<0.001?;H=543.089,P<0.001).Intestinal polyp had the longest course of 165?1⁷30?days,followed by crissum diseases?60?1¹825?days?,both of which were longer than other etiologies except digestive tract malformation and OGIB.Intussusception had shorter course of 1?1³65?days than other etiologies,and its length of hospital days??2?1²1?days??was not statistically significant compared with crissum diseases?3?2¹0?days?,but both of which were shorter than other etiologies.What's more,digestive tract malformation,OGIB HSP,and EGV caused the longest hospital stay.5.There was no statistically significant difference in the onset seasons among different age and gender groups However,there were statistically significant differences in the hemorrhage sites and etiology among different seasons??2=38.928,P<0.001;?2=63.869,P<0.001?:The incidence of LGIB in summer?81.5%?was higher than that in spring?68.3%?,autumn?65.5%?and winter?61.4%?,but the incidence of systemic disease in summer?11.7%?was lower than other seasons.Besides,the incidence of HSP in summer?6.8%?was lower than that in spring?17.0%?,autumn?16.7%?and winter?20.8%?.6.There were statistically significant differences in stool property among different hemorrhage sites and etiologies??2=443.731,P<0.001;?2=861.337,P<0.001?:the rate of melena in UGIB?64.4%?was higher than that in LGIB?3.5%?and systemic diseases?46.2%?,the rate of bloody stool in LGIB?83.7%?and bloody stool plus drop of blood after defecation?9.8%?were both higher than that in UGIB and systemic diseases.The top 3etiologies of different stool property were:PU?73.9%?,HSP?50.6%?,EGV?40.7%?in melena group;intussusception?96.9%?,enteritis?89.6%?,OGIB?81.3%?in bloody stool group;crissum diseases?9.1%?,EGV?7.4%?and hematological diseases?4.7%?in drop of blood after defecation group;digestive tract malformation?15.5%?,EGV?11.1%?,PU?10.9%?in melena plus bloody stool group;intestinal polyp?40.4%?,crissum disease?40.3%?,enteritis?5.2%?in bloody stool plus drop of blood after defecation group.7.The first 10 common accompany symptoms of hematochezia in children were vomiting?50.3%?,abdominal pain?29.1%?,fever?19.8%?,hematemesis?19.4%?,rash?19.0%?,diarrhea?11.5%?,pallor?6.1%?,constipation?4.5%?,dizziness?4.3%?,and perianal prolapse?2.2%?.The differences in common symptoms among different hemorrhage sites and etiologies were statistically significant:the rate of vomiting in LGIB?56.7%?was higher than that in UGIB?31.1%?and systemic diseases?36.8%?.The rate of hematemesis?36.7%?in UGIB was higher than LGIB?20.6%?and systemic diseases?9.4%?,and the rate of dizziness?22.2%?was higher than LGIB?2.1%?and systemic diseases?3.1%?.The incidence of abdominal pain in systemic diseases?76.7%?and was higher than that of UGIB?45.6%?and LGIB?13.7%?.The incidence of rash?86.5%?was also higher than the other three groups.In the first 5 common accompany symptoms,the causes of the top 5 etiologies were:intussusception?88.6%?,HSP?46.9%?,PU?34.8%?of vomiting;HSP?100.0%?,OGIB?62.5%?,EGV?51.9%?of abdominal pain;enteritis?40.9%?,intussusception?27.5%?,EGV?22.2%?of fever;intussusception?36.2%?,PU?34.8%?,EGV?33.3%?of hematemesis;HSP?98.8%?,hematological diseases?23.3%?,OGIB?6.3%?of rash.8.Major laboratory index of different etiologies were statistically significantly different?all P<0.001?.The d-dimer of systemic disease was higher than the other three groups.Both BUN and Cr in UGIB were higher than those in LGIB,but the differences of BUN/Cr between the two groups were not statistically significant.To explore whether BUN/Cr could identify UGIB and LGIB,the ROC curve was drawn with the area of 0.565,and the95%confidence interval of 0.499⁰.630,P=0.045.Taking BUN/Cr=31.5 as the cut-off point,the corresponding Jordan index was the largest?the corresponding Jordan index was 0.143,the corresponding sensitivity was58.9%,and the specificity was 55.4%?.The WBC counts of HSP and intussusception were higher than other etiologies,as well as the neutrophil ratio of HSP.And PLT counts of hematological diseases and EGV were significantly lower than other etiologies.The mean value of Hb of digestive tract malformation,PU and EGV was lower than 90g/L,significantly lower than other etiologies;while the mean value of Hb of HSP and crissum diseases was higher than 120g/L.The EOS count in enteritis was higher than that in other etiologies,and there was no statistically significant difference in EOS counts among other etiologies.9.The age of children with intussusception was 9 months?from 2months to 14 years old?,74.1%occurred within 1 year old,and 95.1%occurred within 3 years old.Among them,8 cases were secondary to rotavirus enteritis,1 case was secondary to HSP,and 4 cases was intraoperatively detected as colonic villous tubular adenoma,ascending colon tumor and descending colon giant polyp.The rate of intestinal necrosis in ileo-ileal-colic intussusception?33.3%?was higher than that in ileocolic intussusception?14.8%?.Intestinal necrosis group had longer hospitalization time,WBC counts and PT than the other group.And the failure rate of air enema in premature infants,children with vomiting or hematemesis was higher than full-term infants,children without vomiting or hematemesis.The failure rate of air enema was different in different age groups.Compared with the successful group,the failure group had higher N percentage and d-dimer,but lower Hb counts and albumin.10.The majority of HSP patients were school-age children,with 89cases?55.6%?.In 110 cases?68.7%?,lower extremity rash was the first symptom,and in the remaining 50 cases?31.3%?,abdominal pain was the first symptom.Of the 50 patients,48 developed a rash within 5?1²0?days after abdominal pain,and 2 did not develop a rash during the course of the disease.A total of 32 patients had gastroscopy,all of them had submucosal hyperemia,and most had varying degrees of edema,erosion,3 had ulcer?2cases of gastric ulcer,1 case of duodenal ulcer?.Inflammation was mainly found in gastric antrum,gastric body,duodenal bulb and descending part of duodenal,without esophagus.92 cases?57.5%?were combined with purpura nephritis.The age,length of hospital days,WBC counts,N ratio,and 24hours total urine protein in the purpura nephritis group were all higher than the other group,and the difference was statistically significant.11.Infectious enteritis?47.0%?and food protein allergic enteritis?40.0%?were the main etiologies of enteritis.Of the 54 cases of infectious enteritis,28 cases were positive in fecal culture with Salmonella of 25 cases?Salmonella typhimurium in 12 cases?,Escherichia coli of 1case,and Candida albicans of 2 cases.And amoeba was found in 2 cases by feces smear.Infants were both high risk population for infectious enteritis and food protein allergic enteritis,but the age of children with food protein allergic enteritis?1?1¹38?months?was less than that of children with infectious enteritis?12.5?1¹63?months??P<0.001?.The hematochezia course,the rate of first-degree relative who had allergic history,PLT counts,EOS ratio and EOS counts of food protein allergic enteritis were all higher than that of the infectious enteritis group,but the fever rate and N ratio were lower than that of the infectious enteritis group.And there was no significant difference in allergy history,rash rate,WBC counts and Hb between the two groups?P>0.05?.12.Intestinal polyps were most common in children of pre-school age?37.2%?.Among 94 cases of intestinal polyps,63 cases?67.0%?were rectal polyps,15 cases?16.0%?were sigmoid colon polyps,9 cases?9.6%?were descending colon polyps,2 cases?2.1%?were transverse colon polyps,4cases?4.3%?were colorectal multiple polyps,and 1 case?1.1%?was multiple polyps in colon.The diameter of all intestinal polyps was 2?0.8⁵?cm.There were no statistically significant differences in the location of polyps among different age groups and in the Hb counts among different location??2=16.343,P=0.663;F=1.648,P=0.169?.However,the diameter of different location of polyps was different?P<0.001?:the incidence of large polyps?diameter?2cm?in descending colon?100.0%?was higher than that in rectum?39.0%?,and no difference among other location of polyps.The N ratio and PT in the large polyp group was higher than that in the small polyp group,but the difference in Hb count was not statistically significant.13.Among the 58 cases of digestive tract malformation,46 cases?79.3%?were MD with the median age of 4.8 years?from 3months to 12.8years?,the ration of male to female was 3.18:1,and the average Hb count was 82.0±2.8g/L.The sensitivity,specificity,positive predictive value and negative predictive value of Technetium-99 nuclide scan in the diagnosis of MD were 76.3%,76.9%,90.6%,47.4%and 76.5%,respectively.The diagnostic rate to MD of abdominal ultrasound examination cannot be considered consistent with the Technetium-99 nuclide scan due to their McNemar Test P=0.043<0.05 and Kappa=0.062<0.2,P=0.628>0.05.26cases?56.5%?of MD had ectopic gastric mucosa?male of 21 cases?80.8%??,1 case?2.2%?had ectopic pancreatic mucosa and 1 case?2.2%?had ectopic gastric plus pancreatic mucosa.So the sensitivity,specificity,positive predictive value negative predictive value,and diagnostic accuracy of Technetium-99 nuclide scan in the diagnosis of ectopic gastric mucosa in MD were 72.7%,18.8%,55.2%,33.3%,50.0%respectively.14.PU,which included 13 cases of gastric ulcer?28.3%?and 33 cases of duodenal ulcer?71.7%?.,was common in adolescent group and the ratio of male to female is 37:9=4.11:1,accounts for 51.1%of UGIB.There was statistically significant difference of the proportion of gastric ulcer and duodenal ulcer among different age groups??2=18.339,P<0.001?:the incidence of duodenal ulcer?100.0%?in adolescent group was higher than that in toddler group?41.7%?and pre-school age group?0.0%?.The rate of hematemesis of gastric ulcer?69.2%?was higher than that of duodenal ulcer?21.2%?.The BUN and Cr values of duodenal ulcer were higher than those of gastric ulcer.However,there were no statistically significant differences in Hb count and BUN/Cr ratio between the two groups.There was no consistency between ¹³C-UBT and RUT for the detection rate of Hp infection?Kappa=-0.070,P=0.361;McNemar Test=0.001<0.05?;But RUT and pathological examination showed strong consistency in the detection rate of Hp infection?Kappa=0.667,P=0.006?.McNemar Test=0.5>0.05).ConclusionsLGIB was the main hemorrhage site of hematochezia in children,followed by systemic disease and UGIB.The top 10 etiologies of hematochezia were successively intussusception,HSP,enteritis,intestinal polyp,crissum diseases,digestive tract malformation,PU,hematological diseases,EGV and OGIB.The etiologies of hematochezia in different age groups were different:the most common etiology in infants was intussusception,and in older children was HSP.Hematemesis and melena suggested a high possibility of UGIB,while drop of blood after defecation accompanied by constipation or perianal prolapse should be considered for rectal polyps or crissum diseases,and intussusception rather than UGIB should be considered when accompanied by vomiting.HSP and intussusception with intestinal necrosis should be considered when WBC count significantly increased.In addition to hematological diseases,EGV should also be considered when PLT significantly decreased.Dgestive tract malformation,PU,and EGV should be considered in the case of moderate to severe anemia,whereas crissum diseases,HSP and other diseases with relatively small blood loss should be considered in the case of normal Hb.However,whether BUN/Cr ratio can predict UGIB or LGIB still needs to be confirmed by further studies.In addition,intussusception in older children should take into account secondary factors such as large polyps and neoplastic diseases.The atypical HSP without skin rash could be early diagnosed by gastroscopy.In enteritis,infectious factors such as salmonella infection should be considered first,if the EOS ration or count is increased,it should be distinguished from food protein allergic enteritis and enosinophilic gastroenteritis.Intestinal polyps were more common in school-age children,and most of them were rectal polyps.Technetium-99 nuclide scan was the first choice for the suspected MD.PU was more common in male adolescents.Hematemesis was more likely to indicate gastric ulcer rather than duodenal ulcer.RUT and pathological examination showed strong consistency in the detection rate of Hp infection.In short,in order to decrease missed diagnosis and misdiagnosis,it is of great significance to strengthen the understanding that UGIB and systemic diseases can also cause hematochezia,and perform early endoscopy after evaluation.