| Reuchlin-Vroklage et al. Diagnostic value of abdominal radiography in constipated children: a systematic review. 2005. Archives of Pediatrics and Adolescent Medicine 159[7], 671-678 | Study type:
Systematic Review
Evidence level:
1+
Study aim:
to evaluate the additional diagnostic value of the plain abdominal radiography in the diagnosis of constipation in children | 6 studies (3 case series, 2 case-control studies, 1retrospective re-examination of abdominal radiographs
Inclusion criteria:
Controlled, observational studies investigating the relationship between faecal loading on plain abdominal radiography and symptoms and signs related to constipation in otherwise healthy children aged from 1 to 18 years old
Exclusion criteria:
Lack of control group, no data on diagnostic value presented, symptoms of constipation not related to the outcomes of a plain abdominal radiography
Setting:
all 6 studies hospital based | Otherwise healthy children aged from 1 to 18 years old with signs and symptoms related to constipation. Some studies included children with soiling or encopresis, while others exclude this group
Country:
The Netherlands | Test and Reference Standard
(studies could treat either test as the reference standard)
- -
Faecal loading on plain abdominal radiography according to a predefined scoring system (reference test in 3 studies) - -
Clinical diagnosis of constipation according to the presence or absence of predefined symptoms and signs (reference test in 3 studies)
In the 6 studies included, 3 different scoring systems for assessing impaction on abdominal radiography were used:
| Diagnostic value:
(LR: Likelihood ratio)
- -
Ability of the abdominal radiography to discriminate between clinically constipated and non constipated children (4 studies): Sensitivity: 76 (95% CI: 58 to 89) Specificity: 75 (95% CI: 63 to 85) LR: 3.0 (95% CI: 1.6 to 4.3) Sensitivity: 60 (95% CI: 46 to 72) Specificity: 43 (95% CI: 18 to 71) LR: 1.0 (95% CI: 0.5 to 1.6) Sensitivity: 80 (95% CI: 65 to 90) Specificity: 90 (95% CI: 74 to 98) LR: 8.0 (95% CI: 0.7 to 17.1) Accuracy 80% (95% CI: 50 to 100)
Ability of the clinical examination to discriminate between radiographically constipated and non constipated children (1 study): Sensitivity: 77 (95% CI: 70 to 84) Specificity: 35 (95% CI: 27 to 44) LR: 1.2 (95% CI: 1.0 to 1.4)
- -
Association between a history of hard stool and faecal impaction on radiography: - -
Association between a finding of absence of rebound tenderness and faecal impaction on radiography: - -
Association between stool present on rectal examination and faecal impaction on abdominal radiography: LR: 1.6 (95% CI, 1.2 to 2.0) LR: 1.5 (95% CI, 0.8 to 2.3)
Interobserver reliability:
Intraobserver reliability:
Evaluated in 3 studies, ranged from moderate (k=0.52) to excellent (k≥0.85) | MEDLINE searched from inception to April 2004, search terms reported and comprehensive. Results of this search combined with search strategy specific to identify diagnostic studies.
References lists of reviews articles and included studies checked for further relevant articles. Experts in the field contacted and asked to identify published and unpublished studies. No language restrictions applied
Two reviewers independently screened the titles and abstracts f studies identified by the searches for eligibility. All potentially relevant studies were retrieved as full papers and independently screened by two reviewers. Any disagreements were resolved through consensus or by arbitration of a third reviewer
Methodological quality of studies assessed using the QUADAS tool. An overall methodological quality value was assigned to studies by calculating the number of positive scores (maximum value 14). Studies with scores of 9 or higher >60%) were arbitrarily regarded as being of “high” methodological quality. Two reviewers independently assessed the methodological quality of the independent studies. Any disagreements were resolved by consensus or through consultation with third reviewer. Reviewers scored 84 items and agreed on 65 item (77.4%, k=0.54)
Structured data extraction performed independently by two reviewers and any disagreement resolved by consensus
Source of funding: Not reported |
| de Lorijn et al. The Leech method for diagnosing constipation: intra- and interobserver variability and accuracy. 2006. Pediatric Radiology 36[1], 43-49 | Study type:
Diagnostic. Case control
Evidence level:
III
Study aim:
to assess intra- and interobserver variability and determine diagnostic accuracy of the Leech method in identifying children with functional constipation | 89 non selected consecutive children
Inclusion criteria:
patients referred for the evaluation of abdominal pain, constipation or faecal incontinence. Diagnosis of constipation: at least two of the following was present:
- -
defecation frequency less than 3 times/week - -
2/more episodes of faecal incontinence per week - -
production of large amounts of stool once over a period of 7-30 days - -
presence of palpable abdominal or rectal mass
(control children fulfilled criteria for functional abdominal pain (FAP) and for functional non-retentive faecal incontinence (FNRFI))
Exclusion criteria:
not reported
Setting:
tertiary gastroenterology outpatients clinic | 89 children
Median age: 9.8 years
Group 1 (constipation):
n=52 (28 boys)
Group 2 (controls):
N=37 (24 boys)
31: FNRFI
6: FAP
Diagnosis of functional non-retentive faecal incontinence (FNRFI) based on: 1) two/more faecal incontinence episodes/week with no signs of constipation 2) defecation frequency 3/more times/week 3) no periodic passage of very large amounts of stool at least once during a period of 7-30 days 4) no palpable abdominal or rectal mass on physical examination for a period of at least 1 week during the preceding 12 weeks. Faecal incontinence defined as the voluntary/involuntary loss of loose stools in the underwear after the age of 4 years Functional abdominal pain (FAP) defined as abdominal pain of at least 12 weeks duration 1)that was continuous or nearly discontinuous in a school-aged child or adolescent 2) that had no or only an occasional relationship with physiological events 3) that was accompanied by some loss of daily functioning 4) that was not feigned and ) for which there were insufficient criteria to indicate the presence of another functional gastrointestinal disorder
Country:
The Netherlands | Test:
Leech method to diagnose constipation in plain abdominal radiography
Reference test:
Colonic Transit Time (CTT)
Leech scoring method:
Colon divided into three segments: right, left and recto sigmoid Each segment provided with a score from 0-5
Colonic transit time:
Determined by the method of Bouchoucha. Radiography on day 7 used to count the number of markers in the colon. Number of markers × 2 produced total CTT in hours. Localization of markers and CTT calculated according to previously described formula. Normal range for total transit time based on the upper limits (mean ± 2×SD) from a study in healthy children. Based on this study a CTT > 62 h was considered delayed. | Mean Leech score (using the first score):
- -
Group 1 (constipation): 10.1 - -
Group 2 (controls): 8.5
p=0.002
Mean CTT:
- -
Group 1 (constipation): 92 h - -
Group 2 (controls): 37 h
p<0.0001
Diagnostic accuracy of Leech method vs. CTT method:
- -
Leech method: (cut-off point as per study comparable to 9 as per literature) Sensitivity: 75% Specificity: 59%
(cut-off point 9 as per literature) - -
CCT: (cut-off point 54h as per study) Sensitivity: 79% Specificity: 92%
(cut-off point 62h as per literature)
ROC analysis
- -
AUC (Leech method): 0.68 (95% CI 0.58-0.80) - -
AUC (CTT method): 0.90 (95% CI 0.83-0.96)
p=0.00015
AUC=Area Under the ROC curve
ROC=Receiving Operator Characteristic
Intraobserver variability (Leech score)
Systematic difference (Mean, 95% CI): - -
Scorer 1 0.7 (0.2-1.2) P=0.89 - -
Scorer 2 0.03 (-0.4-0.5) P=0.0005 - -
Scorer 3 -1.6 (-2.0-1.3) P<0.0001
Variability (SD) - -
Scorer 1: 2.2 Limits of agreement: -6.0-5.0 - -
Scorer 2: 2.2 Limits of agreement: -7.0-7.0 - -
Scorer 3: 1.5 Limits of agreement: -5.0-3.0
Interobserver variability (using the first score):
- -
Scorer 3 vs. scorer 1: Mean of differences 2.7 p<0.0001 - -
Scorer 3 vs. scorer 2: Mean of differences 2.9 p<0.0001 - -
Scorer 2 vs. scorer 1: no systematic differences found
| Children with clinical characteristics of FAP and FNRFI were classified as the control group, because according the authors they have “little or no faecal loading on an abdominal radiograph”
Treatment with oral/rectal laxatives was discontinued in each patient for at least 4 days. Thereafter the patient ingested one capsule with 10 small radiograph opaque markers on 6 consecutive days, in order to determine the CTT.
Subsequently, a plain abdominal radiograph was taken on day 7. this radiograph was both used in the Leech method and for CTT measurement
Three scorers independently scored the same radiography twice (4 weeks apart) using the Leech method, which was discussed amongst the three scorers previous to both readings
Scorers were three experienced doctors (a 5th year radiology resident, a paediatric radiologist and a senior paediatric gastroenterologist). No clinical information was about the patients was made available to them.
A Leech score of 9 or more was considered as suggestive of constipation.
CTT were assessed once by a single scorer. It was assumed that the counting of radiopaque markers would not lead to intra- or interobserver variability
In 5% of cases the Leech scores of the same patient produced by different scorers could differ by 4 points or more
Source of funding: not stated |
| van den Bosch et al. Systematic assessment of constipation on plain abdominal radiographs in children. 2006. Pediatric Radiology 36[3], 224-226 | Study type:
Diagnostic retrospective case series
Evidence level:
III
Study aim:
To assess the reproducibility of there scoring systems (Barr, Leech and Blethyn) for plain abdominal radiography, in order to determine which one is most useful in clinical practice | 40 patients
Inclusion criteria:
consecutive patients referred to hospital for assessment of constipation. Patients complained of infrequent defection, soiling, encopresis, or abdominal pain
Exclusion criteria:
None reported
Setting:
hospital | 40 patients
Mean age 7 years (range 3-12)
55% boys
Country:
The Netherlands | Test and Reference Standard
(all tests compared to each other)
- -
Barr scoring system - -
Leech scoring system - -
Blethyn scoring system
Barr scoring system:
Quantifies the amount of faeces in four different bowel segments (ascending colon, transverse colon, descending colon and rectum) and also the consistency of the faces i.e. granular or rocky stools Constipation defined as Barr score>10
Blethyn system:
Rough scoring system used to assess amount of faeces in large bowel
- -
Normal, grade 0: faeces in rectum and cecum only - -
Grade 1, mild constipation: faeces in rectum, cecum and discontinuous elsewhere - -
Grade 2, moderate constipation: faeces in rectum, cecum with continuous faeces affecting all segments - -
Grade 3, severe constipation: faeces in rectum and caecum, continuous elsewhere with dilated colon and rectal impaction
Leech method:
The colon is divided into three segments:
- 1.
ascending and proximal transverse colon - 2.
distal transverse and descending colon - 3.
rectosigmoid Amount of faces in each segment scored from 0 to 5. - 5.
O indicates no faeces and 5 severe faecal loading and bowel dilatation. With a possible score of 0-15, > 8 considered to indicate constipation
| Intraobserver variability (k values)
- -
Observer 1: Barr: 0.75 Blethyn: 0.61 Leech: 0.88
- -
Observer 2: Barr: 0.66 Blethyn: 0.65 Leech: 1.00
Interobserver variability (k values)
- -
Period 1 Barr: 0.45 Blethyn: 0.43 Leech: 0.91
- -
Period 2 Barr:0.71 Blethyn: 0.31 Leech: 0.84
All k values are statistically significant (p<0.05)
Kappa (k) coefficients (level of agreement):
<0.20: poor 021-0.40: fair 0.41-0.60: moderate 0.61-0.80: good 0.81-1.00: very good
| Masked abdominal radiographs of the children were independently evaluated by two observers, both experienced paediatric radiologists. Observers assessed each radiograph on two separate occasions, 6 weeks apart.
Each abdominal radiograph was scored according to the three different scoring systems
Intraobserver variability was determined for each scoring system by comparing data from the same observer at two different reading sessions.
Interobserver reproducibility was determined by comparing data from the two observers on one occasion. Thus two intraobserver and two interobserver variabilities could be derived for each parameter. Kappa coefficients were calculated as indicators of intra- and interobserver variability. |
| Giramonti et al. The association of constipation with childhood urinary tract infections. 2005. Journal of Pediatric Urology 1[4], 273-278United Kingdom. | Study type:
Diagnostic case control
Evidence level: III
Study aim:
To evaluate the relationship between a history of constipation, faecal loading on X-rays and a history of UTIs in an office practice | 133 children
Inclusion criteria:
Cases: Children with a history of UTIs who were already undergoing a VCUG(voiding cystourethrogram), who were on medications for the treatment of constipation
Controls: Children undergoing a plain film of the abdomen for reasons that did not include constipation/ UTIs (e.g. renal calculi, gastroesophage al reflux)
Exclusion criteria:
Neurological bowel and/or bladder dysfunction or lower gastrointestinal problems. Children with no history of UTI who were undergoing a plain film of the abdomen for constipation or encopresis
Setting:
office practice | 133 children
35 males
Mean age: 5.6 years
(range: from newborn to 14 years)
Group 1 (history of UTI
n=100
Group 2 (no history of UTI)
n= 33
Country:
USA | Test and Reference Standard
(not clear which one was what)
- -
Abdominal radiograph (KUB) - -
Clinical variables:
| Correlation between symptoms of constipation and faecal load on abdominal X-ray:
Correlation coefficient=0.08 | Authors defined constipation in the past as “at least 2 weeks of hard, rock-like stools passed less than 3 times/week without evidence of structural, endocrine or metabolic disease, other useful association include: abnormally large stools, and difficult or painful defecation, associated with stools accidents or faecal smearing in undergarments
Abdominal X-rays reviewed blindly by three physicians: two paediatric radiologists an one paediatric urologist and score for faecal loading based on a previously validated scoring system (Leech)
Data collected prospectively on several historical questions about constipation shortly after the X-ray was performed, but before they were reviewed with the family. An interviewer filled out the history questionnaire using consensus of the child's and parents' responses. Data were also obtained regarding a history of UTI. No data on the interviewer are reported
Constipation history responses were scored from 1 to 3 and a total history score was obtained scored were grouped as:
1-none or mild, 2-moderate, 3-severe
Data derived from scores on faecal loading were averaged for each patient and the scores then grouped in the same way as previous. Questionnaire not piloted previous to the study
As it was thought that children beyond toilet-training age would be more likely to have developed constipation related to overall elimination dysfunction and therefore UTIs as well, the data for children > 3 years were analysed separately |
