‹ ›
Table of Contents

Prevalence of pneumonia in children under 12 years of age who undergo abdominal radiography in the emergency department

Pediatrics

Valérie Homier, MD;* Colette Bellavance, MD;*† Marianne Xhignesse, MD†

From the *Faculty of Medicine, University of Sherbrooke, Sherbrooke, Que. and the †Centre hospitalier universitaire de Sherbrooke, Sherbrooke, Que.

CJEM 2007;9(5):347-351

Abstract

Objective: Pneumonia is a well-known cause of acute abdominal pain in children. However, the utility of chest radiography in this setting is controversial. We sought to determine the prevalence of pneumonia in children under 12 years of age who had abdominal pain and underwent abdominal radiography when visiting an emergency department (ED). We also aimed to describe the signs and symptoms of children diagnosed with pneumonia in this context.

Methods: We conducted a retrospective analysis of electronic data from ED visits to a tertiary care centre by children 12 years of age and under who were seen between June 1, 2001, and June 30, 2003, and who underwent both an abdominal and a chest radiograph during the same visit, or an abdominal x-ray at a first visit as well as a chest x-ray in the 10 days following the initial visit.

Results: Of 1584 visits studied, 30 cases of pneumonia were identified, for a prevalence of 1.89% (95% confidence interval 1.22%-1.56%). If chest radiography had been limited to children who presented with fever, cough and symptoms of an upper respiratory tract infection (URTI), the diagnosis of pneumonia would have been missed in only 2/1584 visits (0.13%).

Conclusion: Children aged 12 years and under presenting to the ED with acute abdominal pain and in whom an abdominal radiograph is requested need only undergo a chest radiograph in the presence of cough, fever or other symptoms of a URTI.

Résumé

Objectif : La pneumonie est une cause bien connue de douleur abdominale aiguë chez les enfants. L'utilité de la radiographie pulmonaire dans ce contexte est toutefois controversée. Nous avons cherché à déterminer la prévalence de la pneumonie chez les enfants de moins de 12 ans qui souffraient de douleur abdominale et ont eu une radiographie abdominale à l'urgence. Nous voulions aussi décrire les signes et les symptômes des enfants chez lesquels on a diagnostiqué une pneumonie dans ce contexte.

Méthodes : Nous avons procédé à une analyse rétrospective des données électroniques tirées des visites effectuées à l'urgence d'un centre de soins tertiaires par des enfants de 12 ans et moins examinés entre le 1er juin 2001 et le 30 juin 2003 et qui on eu une radiographie abdominale et pulmonaire au cours de la même visite, ou une radiographie abdominale au cours d'une première visite et une radiographie pulmonaire dans les 10 jours suivant la visite initiale.

Résultats : Sur les 1584 visites étudiées, on a repéré 30 cas de pneumonie, ce qui représente une prévalence de 1,89 % (intervalle de confiance à 95 %, 1,22 %-1,56 %). Si on avait limité la radiographie pulmonaire aux enfants qui avaient de la fière, toussaient et présentaient des symptômes d'infection des voies respiratoires supérieures (IVRS), on aurait raté le diagnostic de pneumonie dans seulement 2 cas sur 1584 (0,13 %).

Conclusion : Les enfants de 12 ans et moins qui se présentent à l'urgence avec une douleur abdominale aiguë et pour lesquels on demande une radiographie abdominale n'ont besoin que d'une radiographie pulmonaire s'ils toussent, ont de la fière ou présentent d'autres symptômes d'IVRS.

Introduction

Pneumonia is a well-known cause of abdominal pain in children.1-8 A review of the literature suggests that there is still disagreement about the necessity of performing a chest radiograph in children who present to the emergency department (ED) with abdominal pain. While some authors favour this approach,9-12 others do not recommend that a chest radiograph be performed routinely.13-17

In November 2000, the Centre hospitalier universitaire de Sherbrooke (CHUS), Hôpital Fleurimont, instituted a policy concerning children 12 years of age or under presenting to the ED with abdominal pain. This policy mandated a chest radiograph with every abdominal series in order to rule out pneumonia in these children.

The objectives of this study were to determine the prevalence of pneumonia in children who presented to the ED with abdominal pain and who underwent abdominal radiography, and to describe the signs and symptoms found in those with positive chest x-rays.

Methods

A retrospective analysis was carried out using the hospital's computerized database (Nucleus). The study targeted children 0 to 12 years of age who presented to the ED with a chief complaint of abdominal pain between June 1, 2001, and June 30, 2003, and who underwent a chest and abdominal radiograph at the time of the ED visit (group 1), or who underwent an isolated abdominal radiograph performed at the time of the ED visit and had a chest radiograph performed in the 10 days following the initial visit (group 2).

The final radiologist's report was reviewed by the main investigator in order to determine the number of cases of pneumonia. There was no review of x-rays by a second blinded radiologist at the time of the data collection. For each case of pneumonia, the following information was gathered: the age of the patient, the type of radiograph that established the diagnosis of pneumonia (abdominal or chest), the location of the pneumonia, the patient's white blood cell and granulocyte count, the date of the visit and the encrypted medical record number.

Data were entered into distinct Excel files (Microsoft). The first file (group 1) consisted of patients who underwent chest and abdominal radiography at the initial visit (n = 1603). The second (group 2) consisted of cases in which a chest radiograph was performed in the 10 days following the ED visit (n = 31).

In cases where the radiologist's interpretation was not diagnostic (e.g., there was a presence of non-specific retrocardiac opacity that could represent atelectasis or the beginning of a pneumonia), the chart was reviewed to confirm or refute the possibility of pneumonia. This approach permitted the retrieval of clinical information, such as the presence and localization of abdominal pain, vital signs at triage, history of fever, presence of cough and symptoms of an upper respiratory tract infection (URTI) (i.e., rhinorrhea, nasal congestion, sore throat and otalgia). In total, 60 charts were reviewed. The findings were discussed and consensus was reached among researchers.

Fever at triage was defined as a temperature greater than 38.5ºC rectally or 38ºC orally. Normal white blood cell and granulocyte counts as a function of age were provided by the CHUS (Hôpital Fleurimont) in July 2003 for the purpose of defining abnormal values. Age-specific tachypnea was defined according to the reference range provided in the Guide de l'examen clinique.18

There were no statistical analyses performed in this study.

Ethical considerations

This study was approved by the director of professional services at the CHUS in the context of a quality improvement project.

Results

In group 1, we identified 1401 children aged 0 to 12 years, who made 1603 visits to the ED between June 1, 2001, and June 30, 2003, and underwent both an abdominal and a chest x-ray during the same visit. Nineteen visits were excluded because of missing radiographic reports. Ten additional cases were excluded without chart review because they represented nosocomial pneumonias or because they were already being treated for pneumonia at the time of presentation. Three additional cases were excluded after record review; 2 because they were already being treated for pneumonia, and the other because there was no documentation of abdominal pain in the physician's notes.

In group 2, 31 children underwent abdominal radiography at the time of the ED visit, followed by chest x-rays in the 10 days following the initial visit. Two charts were excluded from this group because the radiographic reports were missing. A single case of pneumonia was identified among the 29 remaining charts. This case was excluded because it involved the transfer from another institution of a patient who had undergone x-rays at the referring hospital. An overview of the study methods is represented in Table 1. All of these exclusions were based on criteria established a priori as part of the study protocol.

Table 1. Incidence of pneumonia in children aged 12 years and under who underwent an abdominal radiograph in the emergency department*
Variable No. of patients†
Group 1 Group 2
Radiography, no. of cases‡ 1603 31
    Exclusions§ 19 2
    Total 1584 29
Radiographic interpretations    
    Pneumonia 43 1
    Exclusions without chart review¶ 10 0
    Total 33 1
    Additional exclusions after chart review      3**      1††
    Total 30 0
ED = emergency department. *Groups 1 and 2 include all children aged 0 to 12 years with abdominal pain who were seen in the ED during the study period and who underwent radiographic examination. †Unless otherwise indicated. ‡1603 patients in group 1 underwent abdominal and chest radiography at the index visit; 31 patients in group 2 underwent abdominal radiography at the initial visit and pulmonary radiography in the 10 subsequent days. §Missing radiology reports. ¶Patients had nosocomial pneumonia and were hospitalized or already being treated for pneumonia on presentation to the ED. **2 patients were excluded because they had nosocomial pneumonia and were hospitalized or already being treated for pneumonia on presentation to the ED; the other patient was excluded because of the absence of abdominal pain, as noted in the chart. ††Radiographs were obtained from another hospital.

Of the 1584 visits analyzed in group 1, 30 cases of pneumonia were identified, providing a prevalence of 1.89% (95% confidence interval [CI] 1.22%-1.56%). Four of these radiographically-diagnosed cases of pneumonia were initially missed by the emergency physician. Table 2 provides descriptions of the patient characteristics, signs and symptoms of the 30 children with pneumonia.

Table 2. Characteristics, signs and symptoms among the 30 cases of pneumonia in group 1
Characteristics, signs or symptoms No. of patients* Characteristics, signs or symptoms No. of patients*
Sex Cough
    Male 20     Presence 20
    Female 10     Absence 1
Age     Information not documented 9
    Mean 4.7 yr URTI symptoms (rhinorrhea, congestion, sore throat or otalgia)
    Range 9 mo–12 yr     Presence of symptoms 23
Diagnosis of pneumonia     Absence of symptoms 2
    Visible on chest radiograph only 28     Information about symptoms not documented 5
    Also visible on abdominal radiograph 2 Fever (> 38.5°C rectal or > 38°C orally) as per history or documented in triage
Lobar involvement     Presence 25
    Left lower lobe 13     Absence 5
    Left middle lobe 4 Oxygen saturation
    Right lower lobe 4     < 95% 6
    Other 9     > 95% 17
Abdominal pain     Not documented 7
    Present as per chart 29 Respiratory rate‡
    No abdominal pain documented† 1     Increased for age 5
Location of abdominal pain     Normal for age 11
    Diffuse 7     Not documented 14
    Periumbilical 5 WBC count and granulocytes§
    Right lower quadrant 4     Abnormal WBC and granulocyte count 6
    Other 6     Normal WBC and abnormal granulocyte count 7
    Location not documented 8     Normal WBC and granulocyte count 8
      No values available 9
URTI = upper respiratory tract infection; WBC = white blood cell. *Unless otherwise indicated. †This patient was 1 year old. ‡Normal values taken from the Guide de l’examen clinique.15 §Normal values for WBC and granulocyte count from the hematology lab at the Centrehospitalier universitaire de Sherbrooke (Hôpital Fleurimont) in July of 2003.

The prevalence of URTI symptoms (i.e., cough, fever, tachypnea, low oxygen saturation, and WBC and granulocyte counts) were abstracted for the 30 children with pneumonia in group 1 (Table 3). Fever was the sign most closely associated with pneumonia and was present in 83% of the cases.

Table 3. Prevalence of signs and symptoms among the 30 cases of pneumonia in group 1
Sign or symptom Available data from group 1 (n = 30) No. of children presenting with sign or symptom Prevalence of sign or symptom
Cough 21 20 67%
URTI 25 23 77%
Fever 30 25 83%
Tachypnea 16 5 17%
Decreased saturation 23 6 20%
Increased WBC and/or granulocyte count 21 12 40%
URTI or cough 26 26 87%
URTI or fever 30 27 90%
Cough or fever 30 28 93%
URTI or cough or fever 30 28 93%
URTI = upper respiratory tract infection; WBC = white blood cell.

Missing data for each sign and symptom were considered negative, and prevalence was calculated by dividing the number of cases demonstrating those signs and symptoms by 30 and then converting this value into a percentage. This strategy was employed to minimize the impact of these indicators.

Clinical characteristics noted in this study and listed in decreasing order of prevalence are fever, URTI symptoms and cough. If a chest x-ray, in addition to an abdominal x-ray, had been restricted to those children with either fever or cough, or URTI symptoms, only 2 of 1584 visits (0.13%) with a final diagnosis of pneumonia would have been missed.

Discussion

The prevalence of pneumonia reported in this study is 1.89% (95% CI 1.22-1.56%) and corresponds to what is reported in the literature. Sims and Alexander19 reported a prevalence of 2%; Jones20 noted 2.2%; and Ravichandran and Burge15 found 1.6%. Given the low prevalence of pneumonia in children with abdominal pain, we suggest, as do Hayes,13 Spencer,14 and Ravichandran and Burge,15 that routine chest x-rays in those children presenting with isolated abdominal pain are unnecessary.

Three less recent studies recommend that all children with abdominal pain undergo chest radiography.9-11 Two of these studies reported a higher prevalence of pneumonia in their setting (4.8% and 4.2%).10,11 The third study does not report prevalence nor does it report the age of the children in the study.9

Neither Spencer14 nor Vendargon and colleagues16 support the notion of obtaining chest x-rays in children with acute abdominal pain, although their view is supported only by case reports. Ravichandran and Burge15 are of the same view, but failed to identify a gold standard for diagnosis in their work. Our study refutes the need for routine chest x-ray, thereby strengthening the position in the existing literature and providing a more recent perspective on this subject.

Certain study limitations warrant discussion. As in any retrospective study, chart data will be incomplete. It is possible that in some cases, the policy was not well applied and that a small number of children with abdominal pain had abdominal radiography without a chest x-ray. Furthermore, this study was completed in a single centre, thus limiting external validity. A prospective trial in multiple hospitals would permit confirmation of these results.

Conclusion

In this study, the prevalence of pneumonia among children aged 0 to 12 years presenting to the ED of a tertiary care centre and having undergone abdominal radiography related to a chief complaint of abdominal pain between June 1, 2001, and June 30, 2003, was 1.89% (95% CI 1.22-1.56%). Fever was the sign most closely associated with pneumonia and was present in 83% of the cases. Children of this age who present to the ED with isolated abdominal pain need not be subjected to a chest x-ray in addition to abdominal radiography in a systematic manner. Instead, the chest x-ray may only be required if the child also shows evidence of cough, fever or a URTI.

References

  1. Purcell TB. Nonsurgical and extraperitoneal causes of abdominal pain. Emerg Med Clin North Am 1989;7:721-40.
  2. Kirschner BS. The gastrointestinal tract. In: Nelson essentials of pediatrics. 4th ed. Behrman RE, Kliegman RM, editors. Philadelphia (PA): W.B. Saunders Company; 2002. p 469-71.
  3. Black DD. The respiratory system. In: Nelson essentials of pediatrics. 4th ed. Behrman RE, Kliegman RM, editors. Philadelphia (PA): W.B. Saunders Company, 2002. p 541-4.
  4. Prober CG. Pneumonia. In: Nelson textbook of pediatrics. 16th ed. Behrman RE, Kliegman RM, Jenson HB, editors. Philadelphia (PA): W.B. Saunders Company, 2000. p 761-5.
  5. Ulshen M. Major signs and symptoms of digestive tract disorders. In: Nelson textbook of pediatrics. 16th ed. Behrman RE, Kliegman RM, Jenson HB, editors. Philadelphia (PA): W.B. Saunders Company; 2000. p 1102-6.
  6. Lillis KA. Gastrointestinal disorders. In: Barkin RM, editor. Pediatric emergency medicine: concepts and clinical practice. 2nd ed. St. Louis (MO): Mosby-Year Book Inc.; 1997. p 797-804.
  7. Letourneau MA, Schuh S, Gausche M. Respiratory disorders. In: Barkin RM, editor. Pediatric emergency medicine: concepts and clinical practice. 2nd ed. St. Louis (MO): Mosby-Year Book, Inc; 1997. p 1102-8.
  8. Ruddy RM. Pain-abdomen. In: Textbook of pediatric emergency medicine. 4th ed. Feisher GR, Ludwig S, editors. Philadelphia (PA): Lippincott Williams & Wilkins; 2000. p 421-8.
  9. Freimanis AK, Nelson SW. The chest roentgenogram in the diagnosis of acute abdominal disease. Radiol Clin North Am 1964; 2:3-20.
  10. Jona JZ, Belin RP. Basilar pneumonia simulating acute appendicitis in children. Arch Surg 1976;111:552-3.
  11. Jona JZ, Selke AC, Belin RP. Radiologic aids in the diagnosis of appendicitis in children. South Med J 1975;68:1373-6.
  12. Swischuk LE. Emergency imaging of the acutely ill or injured child. 3rd ed. Baltimore (MD): Williams & Wilkins; 1994. p 222-3.
  13. Hayes R. Abdominal pain: general imaging strategies. Eur Radiol 2004;14:L123-37.
  14. Spencer PAS. Pneumonia, diagnosed on the abdominal radiograph, as a cause for acute abdomen in children. Br J Radiol 1990; 63:306-8.
  15. Ravichandran D, Burge DM. Pneumonia presenting with acute abdominal pain in children. Br J Surg 1996;83:1707-8.
  16. Vendargon S, Wong PS, Tan KK. Pneumonia presenting as acute abdomen in children: a report of three cases. Med J Malaysia 2000;55:520-3.
  17. Cope's early diagnosis of the acute abdomen. 20th ed., revised by William Silen. New York (NY): Oxford University Press; 2000. p 58, 273-5.
  18. Bates B. Guide de l'examen clinique, 4e édition. Reuil- Malmaison: Arnette, 2001. p. 739.
  19. Sims DG, Alexander FW. Acute abdominal pain in childhood. BMJ 1976;2:880-1.
  20. Jones PF. Active observation in management of acute abdominal pain in childhood. BMJ 1976;2:551-3.
Competing Interests: 
None declared.
Correspondence to: 

Dr. Valérie Homier, Royal Victoria Hospital, 687 Pine Ave. West, Room A4.62, Montréal QC H3A 1A1; valerie.homier@mail.mcgill.ca

Received: Feb. 5, 2006; revisions received: Aug. 23, 2006; accepted: Oct. 15, 2006
‹ ›
Table of Contents