AIM: To determine the predictability of the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN) and American Society for Gastrointestinal Endoscopy (ASGE) guideline with regard to appropriate endoscopic practice in children, positive endoscopic findings and contributive yield in clinical practice.

METHODS: This was a descriptive, retrospective analysis, conducted at the Department of Paediatrics, University Malaya Medical Centre, Malaysia. All children who had esophagogastroduodenoscopy (EGD) and colonoscopy from January 2008 to June 2011 were included. An endoscopy was considered appropriate when its indication complied with the NASPGHAN and ASGE guideline. All endoscopic findings were classified as either positive (presence of any endoscopic or histologic abnormality) or negative (no or minor abnormality, normal histology); effecting a positive contributive (a change in therapeutic decisions or prognostic consequences) or non-contributive yield (no therapeutic or prognostic consequences).

RESULTS: Overall, 76% of the 345 procedures (231 EGD alone, 26 colonoscopy alone, 44 combined EGD and colonoscopy) performed in 301 children (median age 7.0 years, range 3 months to 18 years) had a positive endoscopic finding. Based on the NASPGHAN and ASGE guideline, 99.7% of the procedures performed were considered as appropriate. The only inappropriate procedure (0.3%) was in a child who had EGD for assessment of the healing of gastric ulcer following therapy in the absence of any symptoms. The overall positive contributive yield for a change in diagnosis and/or management was 44%. The presence of a positive endoscopic finding was more likely to effect a change in the therapeutic plan than an alteration of the initial diagnosis. A total of 20 (5.8%) adverse events were noted, most were minor and none was fatal.

CONCLUSION: The NASPGHAN and ASGE guideline is more likely to predict a positive endoscopic finding but is less sensitive to effect a change in the initial clinical diagnosis or the subsequent therapeutic plan.

Endoscopy is a useful diagnostic tool in both adult and pediatric populations[1,2]. Endoscopy in the pediatric population is usually performed by a pediatric gastroenterologist, and occasionally by a pediatric surgeon. In settings where the expertise of a fully trained pediatric gastroenterologist is not available, an adult gastroenterologist, supported by a pediatrician, can perform simple, diagnostic endoscopy in children safely[3].

Esophagogastroduodenoscopy (EGD) and colonoscopy in children can be either diagnostic or therapeutic[2]. Common indications for diagnostic EGD and colonoscopy in children include the presence of symptoms indicative of an underlying organic pathology of the gastrointestinal (GI) tract[1,2,4,5].

Generally, diagnostic pediatric EGD and colonoscopy are safe[6]. The risks of therapeutic endoscopy depend on the nature of interventions, but if performed by a pediatric endoscopist with appropriate training, the complication rate is less than 1%[6,7]. Potential complications may be encountered in sedation and anesthesia provided during the procedure[8].

In 2000, the American Society for Gastrointestinal Endoscopy (ASGE) published a guideline on the appropriate use of GI endoscopy in the adult population[9]. Since then, many studies have found the ASGE guideline to be applicable in the adult population[10-12]. ASGE and the North American Society for Pediatric Gastroenterology, Hepatology and Nutrition (NASPGHAN) published a modification of the guideline for the pediatric population, where clear indications for both EGD and colonoscopy in children were recommended[2].

There are several studies on the appropriateness of endoscopy in the adult population[10-12]. However, similar studies in the pediatric population are limited[5,13,14]. We conducted a retrospective review to assess the appropriateness of GI endoscopy performed in children in our unit, based on the NASPGHAN and ASGE guideline[9]. In addition, the rates of positive and negative endoscopic findings as well as contributive and non-contributive yields to the diagnosis and management of the patients were also studied.

This was a retrospective, descriptive study conducted at the Gastroenterology and Nutrition Unit, Department of Paediatrics, University of Malaya Medical Centre (UMMC), Malaysia; from 1^st January 2008 to 30^th June, 2011. The present study was approved by the institutional ethical committee of UMMC.

During the study period, all children who required GI endoscopy in the unit, including those who were referred from outside the unit, were screened initially by one of the three practicing pediatric gastroenterologists (Lee WS, Chai PF, Boey CCM). All endoscopic procedures were performed by one of these three pediatric gastroenterologists.

During the study period, a total of 362 procedures were performed in 318 children. Of these, 17 procedures involving 17 patients were excluded from analysis: 8 had incomplete data (three for EGD, two each for colonoscopy and percutaneous endoscopic gastrostomy feeding tube, and one for foreign body removal), and 9 had an incomplete procedure (seven had colonoscopy abandoned because of poor bowel preparation, and two patients had EGD abandoned because of esophageal stricture). Thus, a total of 345 procedures involving 301 patients were analyzed. Of these, 231 patients had EGD alone, 26 had colonoscopy alone, while 44 had combined EGD and colonoscopy.

Generally, for a procedure to be considered as appropriate, its expected benefit should be greater than its expected negative consequences by a sufficiently wide margin to make the procedure worthwhile[15]. Benefit and negative consequences of a procedure are both defined in the broadest terms[10,15].

Guidelines on the appropriateness of endoscopic procedures have been devised to aid clinicians in selecting more appropriate patients for referral, especially to units with limited expertise and financial resources[2,9]. Recently, a guideline pertaining to the appropriate use of endoscopy in children was published by NASPGHAN and ASGE[2]. We believe that, although there are unavoidable socio-cultural and geographical differences as well as pattern of diseases, the NASPGHAN and ASGE guideline can be applied universally. Thus, for the present study the NASPGHAN and ASGE guideline was chosen.

In addition, little is known about pediatric endoscopic practice and its appropriateness in Asian countries, where human and financial resources, funding model, pattern of GI and liver diseases are different from the more advanced Western countries.

There have been several studies on the appropriateness of EGD in various clinical situations in children[13,14]. However, none are based on the NASPGHAN and ASGE guideline. For example, Jantchou et al[13], based on the recommendations by the French-language Pediatric Hepatology, Gastroenterology and Nutrition Group (GFHGNP), noted that 18% of the 251 EGD procedures performed were considered as inappropriate, a figure which was higher among outpatient referrals. Guariso et al[5], using a model of expert consensus from theoretical scenarios, noted that except in cases with a positive family history of peptic ulcer and/or H. pylori infection, children aged 10 years of older, or with persistent symptoms, not all EGD in children with dyspeptic symptoms could be considered as appropriate. Miele et al[14] found that the publication of Rome II criteria for functional GI disorders has a positive impact on the appropriateness of GI endoscopy, with inappropriate procedures reduced significantly after its publication. Nevertheless, 26% of all procedures were still considered as inappropriate[14].

In contrast, although using different standards, the overall inappropriateness for pediatric endoscopy in the present study was 0.3%, with an overwhelming 99.7% of the cases being considered as appropriate. The only case in the present study which was deemed to be inappropriate was an EGD reassessment of a healing gastric ulcer, in the absence of any symptoms and signs. This figure compares favorably with 18% of inappropriateness noted by Jantchou et al[13] and 26% found by Miele et al[14].

There is, at present, limited availability of human resources in pediatric gastroenterology practice in Malaysia. The pediatric gastroenterology and nutrition unit of UMMC is only one of two pediatric gastroenterology units in Malaysia providing regular pediatric endoscopic services. The model of practice is not an open-access system. Thus, in the present study, all referrals for GI endoscopy from office-based pediatricians were screened initially by one of the practicing gastroenterologists before being subjected to endoscopy, hence reducing potentially inappropriate cases.

Nevertheless, some authors argued that the probability of detecting a clinically relevant lesion is considered as important as the appropriateness of the procedure[16,17]. Gonvers et al[18] found that when applying the ASGE criteria to 450 outpatients who underwent EGD, there were no significant differences in clinically relevant findings in those patients who had an appropriate vs an inappropriate EGD.

Thus, we also studied the probability of finding a positive endoscopic finding in addition to studying the appropriateness of endoscopy. In the present study, the overall probability of detecting a positive endoscopic finding was 76%, higher in EGD (79%) than in colonoscopy (64%).

In the present study, the positive contributive yield for a change in the initial diagnosis was only 23% (Table 5). This is mainly because in over half of the cases (57%), a positive endoscopic finding confirmed the initial diagnosis, thus the contributive yield was considered as negative. However, what is equally important was a negative finding which has a positive contributive yield. Examples included the reassuring negative EGD finding in a child with upper GI bleeding. In the present study, a positive endoscopic finding was more likely to effect a change in the management plan than to effect a change in the initial diagnosis.

Although endoscopic procedures in the pediatric population are generally safe, adverse events and complications related to anesthesia and the endoscopic procedure itself are well documented[7,8,19]. Most of the adverse events encountered in the present study were minor and transient in nature. The perforation rate of colonoscopy in the present study was 2.9%, higher than similar figures in the literature[7,19]. Both cases were children with Crohn’s disease who had severe delay in referral, advanced malnutrition and total colonic involvement. Nevertheless, efforts should be initiated to reduce the complications rate further by improving the training of endoscopy in the unit[20].

The main shortcoming in the present study was its retrospective nature. Thus, it may not be entirely accurate in ascertaining whether an endoscopic finding effected any alteration in the initial diagnosis and subsequent therapeutic plan. In addition, the age range of the patients in the present study was wide, and the indications for endoscopy in young children may not be similar to adolescents. Thirdly, the present study was conducted in a university hospital setting and the procedures were performed by experienced pediatric gastroenterologists. Thus, the findings of the present study may not be entirely applicable in other settings.

In conclusion, the present study showed that the modified NASPGHAN and ASGE guideline is applicable universally, be it in a Western country or an Asian country. Although the NASPGHAN and ASGE guideline on the appropriateness of pediatric endoscopy is useful in helping clinicians selecting the most appropriate patient for GI endoscopic procedures, nevertheless its predictability of a positive endoscopic finding is moderate, and it is not very sensitive in predicting whether a procedure has any positive contributive yield in the diagnosis and management of the patients.