Constipation is a common condition with frequent treatment failures and persistent abdominal pain (1). Children with constipation require emergency department (ED) care significantly more often than those without constipation (2). Most children seeking ED care because of symptoms secondary to constipation have moderate degrees of fecal loading and are believed to require disimpaction and maintenance therapy (3,4). Disimpaction, which is usually performed with the aid of either oral or rectal medications, or a combination of the 2, has been shown to be effective in small, uncontrolled studies (4). Although the oral approach is less invasive and more empowering for the child, adherence can be problematic. Although the rectal approach may be faster, it is more invasive (5).
Although polyethylene glycol (PEG)–based laxatives can be effectively and safely used to treat chronic constipation in children presenting for outpatient care (eg, clinics, primary care) (6), data regarding their use in children presenting for ED care are limited (5). The largest prior pediatric ED study, which included 121 children, found that only one-third of this population received an enema (7). Although the authors found no association between clinical response, as reported 4 to 6 weeks later, and ED treatment (7), a subsequent trial involving 79 children concluded that enema disimpaction provides superior, immediate symptom relief compared with PEG (5). Enema administration, however, is not an entirely benign treatment. In addition to causing children to become “upset” (5), sodium phosphate enemas have the potential to result in electrolyte abnormalities (eg, hyperphosphatemia, hypocalcemia, hypernatremia, hypokalemia, metabolic acidosis) (8). Although infrequent, such events have resulted in severe neurologic deficits and death (9). Thus, given the limited knowledge surrounding the pharmacologic management of pediatric constipation in the ED setting and the success of selected interventions, we conducted a large, retrospective cohort study to determine whether enema administration is associated with 7-day ED revisits for persistent symptoms of constipation. We also sought to identify other diagnostic investigations and therapeutic interventions associated with ED revisits in children with constipation to guide clinical care and identify future research targets.
Study Design and Setting
This secondary analysis used a database created to conduct a retrospective cohort study related to diagnostic error in children with constipation (10). The dataset includes a consecutive series of children <18 years old who presented to the ED of The Hospital for Sick Children, a tertiary care hospital in Toronto, Canada, between November 2008 and October 2010, and were assigned a diagnosis consistent with constipation by the treating physician.
The primary objective was to determine whether ED enema administration (vs no enema) is associated with success as measured by 7-day ED revisits for persistent symptoms related to constipation. The latter was deemed to have occurred if the discharge diagnosis assigned at the revisit was in keeping with symptoms attributable to constipation (ie, abdominal pain, constipation). Secondary outcomes included assessing associations between discharge laxative medications (any medication vs no medication) and revisits and identifying other predictors of ED revisits.
Selection of Participants
Research ethics board approval was obtained for the conduct of a record review of children assigned, in the ED, an International Classification of Diseases–10th Revision discharge code consistent with constipation (F98.1, nonorganic encopresis; K59.0, constipation). Although not required, 84% of participants had previously been diagnosed as having constipation (Table 1) (11). Patients who left before medical evaluation, those hospitalized at the index visit or with a significant (ie, surgical) alternative diagnosis within 7 days (ie, at revisit), were excluded. Data were abstracted from electronic patient charts. All of the visits during the subsequent 7 days were reviewed.
Data abstractors were trained by the principal investigator. To minimize bias associated with data abstraction, we used specific, restrictive key words for subjective data fields. Unavailable data were coded as missing except for particular presenting symptoms (eg, fever, vomiting, abdominal pain), medical history, and the performance of diagnostic imaging, laboratory testing, and medical/surgical consultations, for which the absence of a specific description, was interpreted as “not present” or “not done.” When multiple documentation sources were present, that of the most senior physician was used. If physician documentation was unavailable, nursing documentation was reviewed and abstracted. Historical variables were documented before reviewing the outcome, investigations, and laboratory results.
We also abstracted data related to laxative/enema administration and the performance of abdominal x-rays (AXRs). The amount of stool on AXR was categorized, as we have previously reported, according to key words in the final radiology report, as “normal,” “small,” “moderate,” and “large” (10). The reports from other imaging modalities (ie, ultrasound and computed tomography) were not analyzed owing to the small numbers performed, the challenges with conflicting reports, and an inability to use standardized terminology. Rectal examination documentation, if performed, was classified as consistent (“firm,” “hard,” “full,” “loading,” “impacted,” and “large”) or inconsistent (“soft,” “none,” “empty,” “small,” and “liquid”) with constipation. A random number generator (Microsoft Office Excel 2007, Redmond, WA) identified 10% of charts for an independent blinded review to enable interobserver reliability assessment. Agreement was evaluated for the following variables: history of abdominal pain, significant medical history (categorized as none or chronic illness but unlikely to cause abdominal disease vs chronic illness potentially associated with abdominal disease) (10), general appearance, abdominal tenderness, and AXR performed.
A sample size of 2795 patients was estimated to achieve 90% power to detect a 5% difference between groups (ie, those administered an enema vs those not administered an enema) in the primary outcome (ie, 7-day ED revisit), assuming the baseline proportions to be 15% in the no-enema group and that 25% of children would receive an enema. Calculations were performed using a 2-sided test, with significance set at 0.05. Sample size calculations were conducted with the use of PASS 2008 (NCSS, Kaysville, UT).
Frequency counts and percentages are given for discrete variables; means, medians, standard deviations (SDs), and interquartile ranges are provided for continuous variables. Between-group differences in continuous variables were analyzed using the 2-sample t-test and the Mann-Whitney U test for normally and non-normally distributed data, respectively. The association between ED treatment provided (ie, enema vs no enema) and home medications and the dichotomous outcome of 7-day ED revisits for constipation were assessed using the χ2 test. When the number of observations in any given cell of the contingency table was <10, the Fisher exact test was used. Further analysis of the association of enema administration in the ED with 7-day ED revisit was conducted using multiple logistic regression, with the presence or absence of a 7-day ED revisit as the dependent variable and enema administration as the independent variable. Variables considered for inclusion in the model were identified based on clinical plausibility and prior work and included age, sex, vomiting and abdominal pain in the past 24 hours, significant medical history, days since last stool, prior episodes of constipation, pain score, blood test and diagnostic imaging performance, enema administration, and discharge medications. General appearance was included and children were categorized as well or unwell as has been previously performed (12). The chosen variables were assessed for interaction. The model was the refined using backward stepwise elimination. Variables in the initial model with P > 0.05 were removed sequentially. The remaining variables were inspected to ensure that confounding bias was not reintroduced. A variable was considered confounded if the β coefficient changed by >20%.
Interobserver agreement was evaluated with the Cohen kappa (κ) statistic. Agreement was considered moderate if κ values were between 0.41 and 0.6, strong if between 0.61 and 0.8, and those >0.8 were deemed to represent near-complete agreement (13). All of the tests were conducted using SPSS 19.0 (IBM SPSS Statistics, Armonk, NY). Owing to the conduct of multiple analyses within each table, a Bonferroni correction was applied (see individual tables) (14).
A total of 112,381 ED visits occurred during the study period. Coding identified 4109 potentially eligible visits. After exclusion of children admitted, those with alternative diagnoses on chart review, those who left before medical evaluation, and those assigned a significant alternative surgical diagnosis, 3894 eligible visits remained (Fig. 1; the supplemental table shows diagnoses that resulted in subject exclusion, http://links.lww.com/MPG/A324). Three hundred two represented revisits within 7 days; therefore, 3592 visits formed the final cohort (Table 1).
Children meeting our definition of a revisit for constipation (n = 225, 6%) more frequently presented with vomiting, had a significant medical history, and reported having more pain when assessed at triage (Table 1). These returning children also underwent more investigations, including blood tests (19% [42/225] vs 11% [367/3367], 95% confidence interval [CI] of the difference 3%–14%, P < 0.001) and diagnostic imaging (62% [140/225] vs 47% [1580/3367], 95% CI of the difference 9%–22%, P < 0.001) (Table 2). Surgical and/or medical consultations were also more common among those who had a revisit (13/225 [6%] vs 93/3367 [3%], P = 0.01). The most frequently consulted specialties were pediatric surgery (67), neurosurgery (7), urology (6), hematology/oncology (5), cardiology (4), gynecology (4), and others (13).
ED Enema Administration
The most common treatment administered in the ED was a rectal enema. Twenty-five percent (877/3592) of the children received at least 1 enema while in the ED. Children administered an enema (any type and number) were 1.54 times more likely to revisit the ED within 7 days than those who did not receive an enema (8.6% vs 5.5%, 95% CI of the difference 1.1%–5.2%, P = 0.001).
Those who had an AXR performed were more likely to be administered an enema (513/1628 [32%] vs 364/1964 [19%], P < 0.001). Among those who underwent AXR, children with a moderate-to-large amount of stool were even more likely to receive an enema (318/923 [35%] vs 195/705 [28%], P = 0.005); however, among these children, enema administration did not reduce the likelihood of a revisit (27/318 [9%] vs 47/558 [8%], P = 0.70). Forty-six children received a second enema (1.3%).
Pediatric Fleet phosphate enemas (65 mL) represented the most common initial enema administered (602/877, 69%) followed by 130 mL adult Fleet phosphate enemas (235/877, 27%). Normal saline enemas were the least common enema administered (40/877, 5%) and they were the most likely to be followed by a second enema; 35% (14/40) of those initially administered saline enemas received a second compared with 4% of those administered a Fleet enema initially (34/837, P < 0.001).
Laxative medications were recommended at discharge to 2736 children overall (76%; Table 3) and to 75% (659/877) of those who were administered an enema in the ED. Seven hundred ninety-eight children (22%) were instructed to take ≥2 medications (any combination) and such children were less likely to have a revisit compared with children receiving ≤1 medication (4.6% vs 6.7%, 95% CI of the difference 0.2%–3.7%, P = 0.03). Children who received a combination of the most common laxative discharge recommendation, PEG and Pico-Salax (sodium picosulfate, magnesium oxide, and citric acid; Ferring Pharmaceuticals, Toronto, Canada), however, had a 7.7% revisit rate (19/246).
Four variables were determined to be independent predictors of revisits: the administration of an enema in the ED (odds ratio [OR] 1.40, 95% CI 1.05–1.88), vomiting in the last 24 hours (OR 1.45, 95% CI 1.07–1.98), the administration of an enema in the ED (OR 1.40, 95% CI 1.05–1.88), and a significant medical history (OR 1.26, 95% CI 1.04–1.53) (Table 4).
Interobserver agreement was good (N = 360), ranging from 0.70 (95% CI 0.61–0.74) for unwell appearance to 0.93 for AXR performance (95% CI 0.91–0.95).
Independent predictors of ED revisits for ongoing symptoms of constipation include the performance of a diagnostic imaging test, the presence of vomiting at the index visit, enema administration, and the presence of a significant medical history. A large proportion of children received 1 or more enemas in the ED, and a variety of medications were recommended at discharge. Although certain medication combinations were associated with ED revisits, our findings highlight the fact that the optimal ED treatment regimen remains unknown.
The benefit obtained from enema administration, as compared with oral laxative administration, remains a topic of active research. Recently, the 2 approaches were found to be equally effective in treating rectal fecal impaction in children (6). In our cohort, enema administration was a significant predictor of ED revisits. A recent clinical trial reported that disimpaction by enema may be superior to PEG administration in terms of providing immediate symptomatic relief, but the benefit was not sustained beyond 24 hours (5). The lack of sustained benefit may be because enema administration only relieves pressure locally, in the rectum, but has little impact on stool impaction throughout the rest of the gastrointestinal tract. A further downside of enema administration is discomfort, which was reported by 54% compared with 0% administered PEG (5).
The practice variation seen in our study is not unique (7,15); however, the choice of enema is not a trivial matter because significant complications have been reported following enema administration to children with constipation (8,16,17). A systematic review identified 28 pediatric cases with severe hyperphosphatemia and hypocalcemia following sodium phosphate–containing enema administration (9). Such events result from pooling of liquid in the bowel, which attracts large amounts of water, causing dehydration. The retained phosphate is absorbed, leading to severe hypernatremia and hyperphosphatemia (18). These occurrences have led to recommendations against the use of phosphate-containing enemas in children of age <2 years (19,20).
Our study design did not permit us to determine follow-up recommendations at discharge owing to inconsistent charting. Although we have a single-payer system and in our population >90% of patients have a primary care provider, we cannot rule out that access to primary care may have influenced future care seeking behavior (11). We do not know how many children were seen during follow-up by their primary care provider or at an alternate institution. The latter, however, is unlikely because there are no other pediatric EDs in the city, and we have previously reported that caregivers rarely seek care elsewhere (21).
Although we have documentation indicating the 84% of participants had a history of constipation, we are unaware of prior ED visits and gastroenterology evaluations. These limitations may have introduced bias into our study. It should also be noted that follow-up with or the ED consultation of a pediatric gastroenterologist is not routinely performed at the study institution. Most children with refractory symptoms are studied by a pediatrician; in Canada, the latter serve as consultants and manage a large proportion of children with constipation, referring them for further evaluation (eg, pediatric gastroenterology) when they have been refractory to conventional management or when concerns arise regarding an organic etiology.
Although a convenient, symptom-based, diagnostic tool with clinical applicability (the Rome III criteria) has been designed to assist with assessing children with functional gastrointestinal disorders (22,23), its use by emergency medicine physicians at our institution is unknown. Consequently, we used a pragmatic discharge diagnosis–based approach; thus, not all of the participants who were assigned a diagnosis of constipation would meet diagnostic criteria (eg, Rome III). The correlation between International Classification of Diseases–10th Revision diagnostic coding and Rome III diagnostic criteria (24) is unknown.
Abdominal pain is the most common presenting complaint in children ultimately diagnosed as having constipation (25,26), and it is a frequently used outcome measure in studies evaluating the treatment of constipation. It has been shown to improve in >50% of children in the ED following enema administration (6,7). Thus, given the absence of a criterion standard diagnostic test for constipation, it was deemed appropriate to include “abdominal pain” as a discharge diagnosis in keeping with failure of treatment related to the index visit diagnosis of “constipation.” Other measures such as symptomatic relief following enema administration were not selected because of potential placebo effect bias and the fact that early improvements in children with constipation do not necessarily translate into long-term benefits (5).
Given the retrospective design of this study, the outcome measure had to be objective and available on all of the subjects; thus, ED revisits were selected because they are an accepted measure of quality of care across a range of conditions (27) and are an objective indicator of significant ongoing symptoms leading to ED care. Consequently, our cohort and analyses include a heterogeneous group of children who may be at increased risk of revisits because some of the included participants may have a combination of irritable bowel with constipation. Because we did find that children with revisits had more severe pain at presentation, it is possible that our grouping of children with predominantly abdominal pain with those with predominantly constipation may be affecting our findings. Children presenting primarily with abdominal pain may not improve significantly following enema administration, particularly if there is an alternate explanation for their abdominal pain. This issue has been highlighted by experts who have stated that although abdominal pain is common in children with functional constipation, the role that constipation plays in children who present with significant abdominal pain is unknown (28).
Although education and behavior modification aimed at ensuring that regular toilet habits are important initial steps in treating constipation (4), we could not capture those elements owing to the design of our study. Because the addition of medications to behavioral management is beneficial (29), we focused on the former because they are recorded in our electronic discharge database. We, however, do not have data related to compliance or the doses used. In keeping with the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition guidelines, which endorse many oral pharmacotherapeutic options (4), we identified the use of a multitude of medications in varying combinations. We found that laxative medications were recommended at discharge to only three-fourths of our cohort, yet their use was not more frequent among children administered an enema, the group most likely to require home laxative medications (7). In support of the North American Society for Pediatric Gastroenterology, Hepatology, and Nutrition recommendations, we found that children instructed to take multiple medications following discharge were less likely to have a revisit, whereas children who received a combination of magnesium citrate and PEG had an increased revisit rate. Although these children may simply have had more severe constipation, this could not be identified in our retrospective study and future research should focus on identifying the optimal maintenance regimen.
Although some aspects of data abstraction were limited by incomplete records, and it was challenging to determine the severity of illness at the time of the initial ED visit, we did assess interobserver reliability and found it to be good. Because the decision regarding enema administration was made by the treating physician (ie, there is no institutional protocol), we cannot determine what factors were responsible for the decisions at an individual patient level. In addition, we were unable to evaluate enema response because of a lack of consistent documentation. Because this was a single-center study, caution must be exercised when generalizing our findings to other sites where evaluation and treatment regimens may differ.
Enema administration, vomiting, a significant medical history, and the performance of diagnostic imaging are associated with revisits in children diagnosed with constipation. Although a large number of medications are recommended at discharge, the choice of medication does not significantly affect outcomes. Clinicians, however, should be aware that children instructed to take ≥2 medications had fewer ED revisits. Further research is needed to clarify the role of enemas, the optimal laxative regimen, and to facilitate primary care provider follow-up for children at risk for ED revisits, to improve outcomes in children with constipation.
18th Century Donum Dei for Children
Early American 18th century physiks used opiates (the Donum Dei of Sydenham) for just about any pediatric ailments, most commonly teething, worms and diarrhea. However, by the mid-century evidence of opiate toxicity in children became common wisdom among those who cared for them.1 Nevertheless, wet-nurses, nannies and mothers could freely purchase proprietary goods such as Godfrey's Cordial (an infusion of sassafras, treacle and ½ grain of opium) and Dalby's Carminative (aromatic peppermint oil with ⅛ grain of opium per oz).
Both of these products were advertized as good for colic, fluxes, fevers, colds and other ailments. Sadly, the most common use was by nurses to soporify infants and a mid-century legal report described a 14 month old who was “haggard and aged who appeared no larger than an infant of one or two months …and [who had been] kept on paregoric …”
J.B. Beck, Essay on Infant Therapeutics, 1849
Wrote John Cooke in A Plain Account of the Diseases Incident to Children (1769): “Opiates with infants ought to be used with the utmost caution. I lost a son above a year old, who was killed instantly, only with eight drops of liquid laudanum, when two drops are sufficient for a babe.”
A.R. Colón, Nurturing Children: A History of Pediatrics, 1999
—Submitted by Angel R. Colón, MD
1. Loening-Baucke V. Factors determining outcome in children with chronic constipation
and faecal soiling. Gut
2. Liem O, Harman J, Benninga M, et al. Health utilization and cost impact of childhood constipation
in the United States. J Pediatr
3. Pirie J. Management of constipation
in the emergency department. Clin Ped Emerg Med
Guideline Committee of the North American Society for Pediatric Gastroenterology, Hepatology, and NutritionEvaluation and treatment of constipation
in infants and children: recommendations of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition. J Pediatr Gastroenterol Nutr
5. Miller MK, Dowd MD, Friesen CA, et al. A randomized trial of enema
versus polyethylene glycol 3350 for fecal disimpaction in children presenting to an emergency department. Pediatr Emerg Care
6. Bekkali NL, van den Berg MM, Dijkgraaf MG, et al. Rectal fecal impaction treatment in childhood constipation
: enemas versus high doses oral PEG. Pediatrics
7. Miller MK, Dowd MD, Fraker M. Emergency department management and short-term outcome of children with constipation
. Pediatr Emerg Care
8. Mendoza J, Legido J, Rubio S, et al. Systematic review: the adverse effects of sodium phosphate enema
. Aliment Pharmacol Ther
9. Ladenhauf HN, Stundner O, Spreitzhofer F, et al. Severe hyperphosphatemia after administration of sodium-phosphate containing laxatives
in children: case series and systematic review of literature. Pediatr Surg Int
10. Freedman SB, Thull-Freedman J, Manson D, et al. Pediatric abdominal radiograph use, constipation
, and significant misdiagnoses. J Pediatr
11. Freedman SB, Degroot JM, Parkin PC. Successful discharge of children with gastroenteritis requiring intravenous rehydration. J Emerg Med
12. Schnadower D, Kuppermann N, Macias CG, et al. Febrile infants with urinary tract infections at very low risk for adverse events and bacteremia. Pediatrics
13. Landis JR, Koch GG. The measurement of observer agreement for categorical data. Biometrics
14. Sankoh AJ, Huque MF, Dubey SD. Some comments on frequently used multiple endpoint adjustment methods in clinical trials. Stat Med
15. Patel H, Law A, Gouin S. Predictive factors for short-term symptom persistence in children after emergency department evaluation for constipation
. Arch Pediatr Adolesc Med
16. Walker M, Warner BW, Brilli RJ, et al. Cardiopulmonary compromise associated with milk and molasses enema
use in children. J Pediatr Gastroenterol Nutr
17. McCabe M, Sibert JR, Routledge PA. Phosphate enemas in childhood: cause for concern. BMJ
18. Bowers B. Evaluating the evidence for administering phosphate enemas. Br J Nurs
20. Rowan-Legg A. Canadian Paediatric Society-Community Paediatrics CommitteeManaging functional constipation
in children. Paediatr Child Health
21. Freedman SB, Thakkar VA. Easing the strain on a pediatric tertiary care center: use of a redistribution system. Arch Pediatr Adolesc Med
22. Rasquin A, Di Lorenzo C, Forbes D, et al. Childhood functional gastrointestinal disorders: child/adolescent. Gastroenterology
23. Burgers R, Levin AD, Di Lorenzo C, et al. Functional defecation disorders in children: comparing the Rome II with the Rome III criteria. J Pediatr
24. Hyman PE, Milla PJ, Benninga MA, et al. Childhood functional gastrointestinal disorders: neonate/toddler. Gastroenterology
25. Loening-Baucke V, Swidsinski A. Constipation
as cause of acute abdominal pain
in children. J Pediatr
26. Caperell K, Pitetti R, Cross KP. Race and acute abdominal pain
in a pediatric emergency department. Pediatrics
27. Bardach NS, Vittinghoff E, Asteria-Penaloza R, et al. Measuring hospital quality using pediatric readmission and revisit rates. Pediatrics
28. Tabbers MM, Di Lorenzo C, Berger MY, et al. Evaluation and treatment of functional constipation
in infants and children: evidence-based recommendations from ESPGHAN and NASPGHAN. J Pediatr Gastroenterol Nutr
29. Nolan T, Debelle G, Oberklaid F, et al. Randomised trial of laxatives
in treatment of childhood encopresis. Lancet