Article, Pediatrics

Diagnostic approach to constipation impacts pediatric emergency department disposition

a b s t r a c t

Objectives: Constipation is a common cause of abdominal pain in children presenting to the emergency depart- ment (ED). The objectives of this study were to determine the diagnostic evaluation undertaken for constipation and to assess the association of the evaluation with final ED disposition.

Methods: A retrospective chart review of children presenting to the pediatric ED of a quaternary care children’s

hospital with abdominal pain that received a soap suds enema therapy.

Results: A total of 512 children were included, 270 (52.7%) were female, and the median age was 8.0 (IQR: 4.0- 11.0). One hundred and thirty eight patients (27%) had a digital rectal exam (DRE), 120 (22.8%) had bloodwork performed, 218 (43%) had urinalysis obtained, 397 (77.5%) had abdominal radiographs, 120 (23.4%) had abdom- inal ultrasounds, and 18 (3.5%) had computed tomography scans. Children who had a DRE had a younger median age (6.0, IQR: 3.0-9.25 vs. 8.0, IQR: 4.0-12.0; p b 0.001) and were significantly less likely to have radiologic im- aging (OR = 0.50, 95% CI 0.32-0.78; p = 0.002), but did not have an increased odds of being discharged home. After adjusting for gender, ethnicity, and significant past medical history those with an abdominal radio- graph were less likely to be discharged to home (aOR = 0.56, 95% CI 0.31-1.01; p = 0.05).

Conclusions: The diagnostic evaluation of children diagnosed with fecal impaction in the ED varied. Abdominal imaging may be avoided if children receive a DRE. When children presenting to the ED with abdominal pain had an abdominal radiograph, they were more likely to be admitted.

Introduction

Constipation has a high prevalence in children with estimates as high as 29.6% [1]. Children with constipation seek care in the emergency department (ED) nearly three times more frequently than children who do not have constipation [2,3]. One study showed that constipation was the most common cause of abdominal pain among children presenting to the ED for care [4].

Despite the frequency of pediatric ED visits due to constipation, the diagnosis of constipation poses a challenge to ED healthcare providers,

? Sources of support: This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

?? Prior presentation: This study was presented at the Pediatric Academic Society

Annual Meeting in Washington D.C. May 2013 and at Digestive Disease Week May 2013 in Orlando, FL.

* Corresponding author at: Pediatric Emergency Medicine, Baylor College of Medicine, Texas Children’s Hospital, 6621 Fannin Street, Suite A2210, Houston, TX 77030-2399, USA. E-mail addresses: [email protected] (C.E. Chumpitazi), chris.rees@ bcm.edu (C.A. Rees), [email protected] (E.A. Camp), [email protected]

(E.B. Henkel), [email protected] (B.P. Chumpitazi).

particularly when constipation is accompanied by abdominal pain. In these cases, the ED provider must address the possibility of emergent etiologies of abdominal pain (e.g. appendicitis) while accounting for the wide variation in constipation symptoms [5]. Common symptoms of constipation in children may include: passing stool less than three times per week, fecal incontinence, hard stools, excessive straining, or feeling of incomplete evacuation of stool [6-8].

Pediatric gastroenterology society guidelines recommend that providers perform a thorough medical, surgical, drug, and dietary history along with a detailed physical exam and often screening lab- oratory tests as part of the diagnostic evaluation for constipation [9]. Similarly, the Rome IV diagnostic criteria and the Paris Consensus on Childhood Constipation Terminology support a Clinical approach to the diagnosis of constipation [10-13]. As the clinical diagnosis of constipation in children may be challenging because of children’s difficulty describing and reporting symptoms, digital rectal examina- tion (DRE) can provide useful clinical information such as the presence of fecal impaction [14,15]. The reliability and utility of abdominal radio- graphs for the diagnosis of constipation has been a point of debate in the past [16,17].

http://dx.doi.org/10.1016/j.ajem.2017.04.060

C.E. Chumpitazi et al. / American Journal of Emergency Medicine 35 (2017) 1490–1493 1491

Few studies describe the Diagnostic approach undertaken to identify constipation as a likely contributor in children presenting to the ED with abdominal pain [18]. To our knowledge, none have previously evaluated the potential influence of the diagnostic evaluation on ED disposition. Therefore, the objective of this study in children with abdominal pain who received enema therapy for constipation in the pediatric ED was to 1) determine the diagnostic evaluation undertaken for constipation and 2) to assess the association of this diagnostic evaluation with final ED disposition.

Material and methods

Study design

This was a secondary analysis of a retrospective database created to assess the efficacy of soap suds enema (SSE) in the treatment of consti- pation among children in the pediatric ED [19]. SSE is the primary ther- apy used in our pediatric ED for children with significant constipation. This study was approved by our institutional review board.

Study setting and population

This study evaluated subjects presenting to a quaternary care children’s hospital with an average annual volume of approximately 85,000 visits. All children between 8 months and 23 years of age who re- ceived a SSE for constipation between June 2011 and June 2012 in the ED were included. Patients were included if they presented with ab- dominal pain, were diagnosed clinically as having constipation (having fewer than three bowel movements per week, straining, hard stools, ab- dominal bloating or pain with defecation), and subsequently underwent SSE therapy. Patients were included if they were previously healthy and if they had a significant past medical history to maintain the generalizability of the study.

Study protocol

We used specific key words to search the electronic health record to identify patient charts for those who received SSE for constipation. Un- available data were coded as missing or not present. The principal inves- tigator trained the data abstractors to minimize potential variation.

Measurements

The diagnostic evaluation captured included: patient demographics, medical history (symptoms and significant past medical history), phys- ical examination (DRE), laboratory testing, and radiographic imaging. final disposition outcome was also captured. A significant medical his- tory was defined as the presence of a prominent comorbid conditions including: cystic fibrosis, cerebral palsy or muscle disorders, hypothy- roidism, spina bifida or other spine anomalies, gastric anomalies, anorectal malformations, Hirschsprung’s disease, cardiac anomalies, previous abdominal surgery, or previous diagnosis of constipation. For radiologic Imaging results, serious results were defined as “multiple air/fluid levels and distended small-bowel loops compatible with small-Bowel obstruction or Free air under the diaphragm suggestive of a perforated viscous” as defined by Kellow et al. [20].

Data analysis

Descriptive analyses were provided to identify potential con- founders for discharge status. Continuous variables were graphed to as- sess their distribution. All variables were skewed right (positive direction) and all Shapiro-Wilk tests for normality were significant (p b 0.001), therefore non-parametric testing (Mann-Whitney) was uti- lized to determine significance. Median and interquartile ranges (IQR) were calculated. For comparisons among categorical variables, Pearson

chi-square testing was utilized with percentages provided. For the de- scriptive analysis, a potential confounder was defined as p-value b

0.20 and included in subsequent models for further adjustment. Binary logistic regression was utilized to adjust the association analysis be- tween discharge outcome and diagnostic evaluations. All analyses were conducted using the Statistical Package for the Social Sciences, ver- sion 24 (IBM Corp., Armonk, NY).

Results

There were 512 children included in this study, all of whom met the study’s inclusion criteria. There were 270 (52.7%) female patients and the median patient age was 8.0 (IQR: 4.0-11.0). Four patients had re- peat visits that were not within 72 h and thus were included in the anal- ysis. Less than a third of patients (n = 138, 26.9%) had a DRE performed as part of their physical exam and 397 (77.5%) had abdominal X-ray.

Sixty one percent (n = 313) of patients had significant past medical histories with 37.6% (n = 193) having a previous history of constipa- tion. Children with a significant past medical history for both constipa- tion (OR = 2.40, 95% CI 1.40-4.13; p = 0.001) and for all other conditions (OR = 1.58, 95% CI 1.04-2.39; p = 0.03) were more likely to have a DRE performed than patients without previous medical histo- ries. Children who had DREs had a younger median age (6.0, IQR: 3.0- 9.25 vs. 8.0, IQR: 4.0-12.0; p b 0.001) and were significantly less likely

to have radiologic imaging (OR = 0.50, 95% CI 0.32-0.78; p = 0.002), but did not have an increased odds of being discharged home (aOR = 1.29, 95% CI 0.78-2.13; p = 0.32). There were no other significant de- mographic or other characteristics associated with completion of a dig- ital rectal exam (data not shown).

Bloodwork was completed in 22.8% (n = 120) patients. Urinalysis was obtained in 43% (n = 218) and a urinary tract infection was diag- nosed in 5.8% (n = 30) patients. The majority of patients had some form of abdominal imaging performed with the most common modality being an abdominal radiograph (n = 398, 77.7%). Children who re- ceived abdominal imaging did not differ from those who did not in terms of their gender, age, ethnicity, past medical history, and history of constipation (data not shown). Minor abdominal findings on abdom- inal plain film such as ileus or dilated colon were found in 7.4% (n = 30), with none having serious findings. Twenty three percent (n = 120) of patients had Abdominal ultrasounds performed and 3.5% (n = 18) had Abdominal computed tomography performed. Nearly one fifth of pa- tients (n = 95, 18.5%) had more than one type of abdominal imaging performed.

Independent demographic and clinical history factors associated with being discharged from the ED included Female gender and Hispan- ic ethnicity, while a significant past medical history was associated with subsequent hospital admission (Table 1). Patients with a significant past medical history were twice as likely to have a positive X-ray finding as patients without a significant past history (OR = 2.00, 95% CI 1.12- 3.57, p-value = 0.02) (results not shown).

Having an abdominal X-ray performed significantly reduced the odds by approximately half of being discharged home but was no longer significant at the 0.05 level when adjusted for gender, ethnicity, and sig- nificant past medical history (aOR = 0.56 (95% CI 0.31-1.01); p-value =

0.05) (Table 2). Conversely, patients with a DRE performed had an in- creased the odds of being discharged home (aOR = 1.29 (95% CI 0.78- 2.13); p-value = 0.32) but was this was not statistically significant in unadjusted or adjusted models.

Discussion

Identifying effective diagnostic modalities for children presenting to the pediatric ED with abdominal pain and constipation is important. Previous studies have found a lack of uniformity in the diagnosis of con- stipation in the outpatient setting [21]. Our study adds to this under- standing by showing variation in the diagnostic workup of children

1492 C.E. Chumpitazi et al. / American Journal of Emergency Medicine 35 (2017) 1490–1493

Table 1

Disposition of pediatric patients that received enema therapy in the emergency department for constipation.

	Admitted n = 108 (21.1%) n (%) or median (IQR)	Discharged Home n = 404 (78.9%) n (%) or median (IQR)	p-Value
Age (years)	8.0 (3.25, 12.0)	7.0 (4.0, 11.0)	0.30
Female	46 (42.6)	224 (55.4)	0.02
Ethnicity: Hispanic	42 (38.9)	201 (50.3)	0.04
Positive past medical history for constipation	52 (63.4)	141 (61.6)	0.77
Significant past medical history	84 (77.8)	229 (56.7)	b0.001

Table 2

Effect estimates for receiving a digital rectal exam or abdominal X-ray on discharge status in patients who present to the emergency department.

	Unadjusted odds ratio	95% CI	p-Value	Adjusted^a odds ratio	95% CI	p-Value
Abdominal X-ray	0.54	0.30-0.95	0.03	0.54	0.29-0.98	0.04
Digital rectal exam	1.14	0.70-1.85	0.61	1.22	0.72-2.05	0.46

^a Odds ratios were adjusted for age, gender, and significant past medical history.

with constipation in the pediatric ED even in those children ultimately receiving an effective fecal impaction therapy [20]. New is our finding of an increased risk of hospital admission among patients who had ab- dominal radiographs as part of their diagnostic workup for constipation. This novel finding may prompt further investigation into the most ap- propriate evaluation paradigm for children presenting to the ED with abdominal pain.

Abdominal radiographs were obtained in the majority of children in our study. Similarly, abdominal radiographs have previously been shown to be used in 50-90% of children who are diagnosed with consti- pation in the pediatric ED setting [4,18]. This occurred despite the poor validity and reproducibility of abdominal radiographs for assessing pe- diatric constipation [22,23]. The frequent use of abdominal radiographs is a potential cause for concern as abdominal radiographs result in mis- diagnoses in children presenting with abdominal pain [24]. Our study revealed an association between radiologic imaging and subsequent in- patient admission despite the lack of significant differences in patients’ age, gender, ethnicity, and past medical history. Though our study did specifically not evaluate for this, we hypothesize that the poor validity and reproducibility of abdominal radiographs leading to incorrect diag- noses of fecal impaction may in part contribute to the higher associated risk for admission. Given the ubiquity of their use, future studies assessing the accuracy of abdominal X-ray for diagnosing fecal impac- tion or using other radiologic modalities (e.g. abdominal ultrasound) in the pediatric ED setting may be helpful [25,26].

We found that children who received a DRE were less likely to un- dergo an abdominal radiograph. This finding is consistent with other studies which have attributed low numbers of DRE to the high number of abdominal radiographs used as a diagnostic tool to identify constipa- tion and fecal impaction [27]. Furthermore a recent educational module was found to decrease the use of abdominal radiographs in children with Suspected constipation through the encouragement of DRE for diagnosis of fecal impaction [28]. However it should be noted that DRE itself is not uniformly recommended as a recent pediatric gastroenterol- ogy society guideline stated that there is insufficient evidence to recom- mend the use of DRE to diagnose functional constipation [10]. In addition, others have found that both DRE and Abdominal radiography are often inconsistent [29]. Therefore though the use of DRE may de- crease the usage of abdominal radiography, future efforts evaluating the role of DRE in the acute pediatric ED setting are needed.

Though other studies have shown that bloodwork and urinalysis are not always needed to make the diagnosis of constipation [30], urinalysis was completed in nearly half of the patients evaluated for abdominal pain and subsequently treated for constipation in this study and nearly 6% had a comorbid UTI. This finding is in line with other studies that have shown up to 37% of children with urinary tract infections also have constipation [31]. Other studies have shown that urinary

symptoms, such as nocturnal enuresis, are common among children with functional constipation [32]. Given the relatively high frequency of UTIs, we recommend a careful urinary history with low threshold to obtain a urinalysis as part of the diagnostic workup for children with fecal impaction and abdominal pain in the pediatric ED.

This study is subject to several limitations. First as it was based on a retrospective chart review, the diagnostic evaluation and subsequent clinical decision to provide SSE therapy could not be standardized. Our findings are dependent on documentation and 73% of children in this study did not have a documented DRE. However, in surveying providers, the DRE rate is consistent with our reported practice and therefore sug- gests the study reflects actual clinical practice. Second, given limited numbers of those undergoing laboratory testing as part of their workup in the ED, we were unable to assess associations of laboratory testing and the utility therein. Future prospective studies are needed in this clinically prevalent area. Lastly, our methodology was unable to account for patient acuity which may have contributed to our finding of the as- sociation of abdominal radiographs and hospital admission.

Conclusions

The diagnostic evaluation of children diagnosed with fecal impaction in the ED varied. Abdominal imaging may be avoided if children receive a DRE during the evaluation. Children presenting to the ED with abdom- inal pain and receiving therapy for fecal impaction were more likely to be admitted to the hospital if they received an abdominal radiograph.

Acknowledgments

None.

References

AJEM