The use of gastroscopy has been growing worldwide, as well as its diagnostic and therapeutic possibilities [
Performing a gastroscopy on a child is associated with risks. The gastroscopy itself is associated with a risk of bleeding or perforation of the esophagus [
The aim of this study was to map the routine use, outcomes, and ethical aspects of pediatric gastroscopy within the Department of Pediatric Surgery. The purpose was to be able to determine which gastroscopies have a favourable risk/benefit ratio. Our study sought to answer the following questions. Are any unnecessary gastroscopies performed? Should gastroscopy be used more or less frequently? What are the benefits for the children? Can outcomes be predicted?
The goal for our findings was to outline the volume of pediatric gastroscopies, their indications, changes in the frequencies of these indications over time, and their complications. This information is important for future decision-making in the clinical setting and for administrative planning.
To identify the indications for the use of pediatric gastroscopy and patient outcomes, the prospectively collected database containing all children admitted to the Department of Surgery was used. This department is a tertiary centre that provides free public health care for an area of 1.8 million inhabitants and 22,000 newborns each year. Surgeons receive a monthly salary with overtime payment related to their time on call and time on duty doing surgical work.
Gastroscopies were identified in the database using surgery codes “gastroscopy” and “gastroscopy with biopsy” over two time periods: January 2001 to December 2004 and July 2011 to May 2014. Gastroscopies were sorted into five different types of indications: obstruction in the upper gastrointestinal (GI) tract: dilatation and calibration of esophagus after surgery for esophageal atresia, after malignant disease, after esophagitis, and after intake of corrosives; dilatation and calibration of duodenum after duodenitis; investigation of obstruction in the pyloric region, in suspected high intestinal obstruction; investigation of suspected esophageal achalasia; and investigation of dysphagia; gastroesophageal reflux disease (GERD): investigation of suspected or diagnosed GERD; preoperative investigation; endoscopy for postoperative complications; and examination to evaluate the outcome of fundoplication surgery; investigation of gastroenterological conditions: investigation of suspected coeliac disease, inflammatory bowel disease, gastritis, laparoscopic gastrostomy: laparoscopic gastrostomy concurrent with an evaluating gastroscopy performed according to local routines [ other: treatment of oesophageal varices; removal of foreign bodies in the upper GI tract; and placement of capsule for endoscopic investigation.
Variables analysed included gender, age at examination, indication for the gastroscopy, whether or not a biopsy was taken, outcome, and complications related to the gastroscopy within 30 days.
Intention to treat was the main analysis strategy and encompassed all patients. The study was performed according to the Declaration of Helsinki and was approved by the regional research ethics committee (registration number 2010/49). Data were anonymized prior to analysis and are presented in such a way that no single patient can be identified. Therefore, it was not necessary to obtain approval from patients’ guardians. All evaluations, treatments, and procedures described in this paper were standard of care for patients and were conducted at the Department of Pediatric Surgery. No protocols that would have required appropriate informed consent or approval by an institutional review board were used. The ethical questions that may arise with gastroscopy in a child were considered according to the guidelines published in 2014 by the Swedish Council on Health Technology Assessment. The council published twelve questions in four different aspects of ethics to be used for consideration of whether a procedure is ethical [
Because this was a descriptive study, no power calculations were performed. SPSS Statistics was used to compare data from the two time periods, 2001–2004 and 2011–2014, using a chi-squared test with an alpha level of 0.05.
A total of 834 gastroscopies were performed from January 2001 through December 2004 and from July 2011 through May 2014. There was a statistically significant difference in the distribution of indications between the two time periods. The latter period was mapped in detail because it best represented current routines and should provide more accurate information on future routines. In total, 379 gastroscopies were performed in this period. Medical records were not accessible for three examinations, which were then excluded. Hence, the sample consisted of 376 examinations on 314 patients, of whom 289 had only one examination and 25 had multiple examinations.
The median age of patients was four years. Figure
Age distribution across 376 gastroscopies performed from July 2011 to May 2014.
Age distribution across 132 gastroscopies performed on children under 2 years of age from July 2011 to May 2014.
Figure
Distribution of indications for all gastroscopies performed on children from 0 to 15 years of age from July 2011 to May 2014.
Age distribution among indication groups for all 376 patients from 0 to 15 years of age from July 2011 to May 2014.
Gastroscopies revealed no pathological findings in 51% of GERD cases, 46% of investigations of gastroenterological conditions, 36% of investigations of obstruction in the upper gastrointestinal tract, and 100% of laparoscopic gastrostomies. In the “other” group, all examinations were interventional and thus had pathological findings that were treated.
The proportion of examinations that did not lead to an action or change in treatment was 100% in the laparoscopic gastrostomy group, 72% in the GERD group, and 45% in the investigation of gastroenterological conditions group. Table
Summary of the numbers and types of outcomes in the five indication groups.
Indication group | Outcome, percentage of patients ( |
---|---|
Obstruction in upper gastrointestinal tract | (i) Dilatation of stenosis, 55% (41) |
|
|
Gastroesophageal reflux disease | (i) No positive findings at either of the following aspects: inspection of mucosa/inspection of cardiac function/biopsy/24-hour pH measurement, 51% (29), of which 10% (3) led to dose change in proton pump inhibitor (PPI) treatment and 90% (26) led to no change of treatment |
|
|
Investigation of gastroenterological conditions | (i) No positive findings at inspection/biopsy, 46% (39), of which 23% (9) led to further investigation/treatment with gluten-free diet and 77% (30) led to no action taking/no change of treatment |
|
|
Laparoscopic gastrostomy | (i) Inspection of the placement of gastrostomy and signs of esophagitis, 100% (150) (of which none showed signs of esophagitis) |
|
|
Other | (i) Removal of foreign body, 60% (6) |
Of 376 examinations, complications related to the gastroscopy occurred four times (1%). Esophageal perforation occurred in two children, four and ten months of age. One child was examined with gastroscopy because of stricture after esophageal atresia and the other because of symptoms of GERD. Both were initially treated conservatively which was successful in one child, while the other child required a thoracotomy to drain the mediastinum. Aspiration and bronchial spasm occurred in two children, two and three years of age. Both were discharged from the hospital after 24 hours of uneventful observation in a pediatric surgical ward. Duration of hospital stay ranged from one day for the children with aspiration to 14 and 24 days for those with esophageal perforation. No lethal events occurred.
The use of gastroscopy increased from an average of 114 per year from 2001 to 2004 to an average of 129 per year from 2011 to 2014 (
Percentage distribution of indications from January 2001 to December 2004 (455 patients) compared to July 2011 to May 2014 (376 patients).
The four aspects of ethics outlined by the Swedish Council on Health Technology Assessment in 2014 [
This study demonstrates that there has been an increase in the number of gastroscopies performed as well as a change in the distribution of indications for gastroscopies over the past 10–15 years. Furthermore, laparoscopic gastrostomy was the predominant indication for gastroscopy. The outcome of gastroscopy was negative in many cases; that is, no pathological findings were reported. Additionally, in the GERD group, the vast majority of examinations did not lead to any further action or change in treatment. This trend was also observed to a lesser extent in the investigation of gastroenterological conditions group. The overall complication rate was 1%, and no lethal events occurred.
The current study is a unique mapping of gastroscopies performed in pediatric surgery care for several reasons. First, the country’s health care system is entirely free of cost for all children, which means that the sample represents all socioeconomic groups. Furthermore, the surgeons at the hospital are on salary and are not reimbursed based on the number of procedures they perform, which differs from some other countries, which often use the “fee for service” method [
The median age in the study group was four years. Recent studies on pediatric gastroscopy have reported a mean age of 6.9–9 years [
The group with an indication of obstruction in the upper GI tract had its first peak at age 0-1 year. This peak within the first year of life is not very surprising, since stricture as a complication of surgery for esophageal atresia is not unusual within the first months after surgery [
The current study also showed peaks in girls at age 5–10 years and 10–15 years. There is no explanation to these peaks to be found in the literature. However, our study included two female patients in these age groups who each underwent numerous gastroscopies, one of whom suffered from a corrosive esophageal stricture and one had an esophageal stricture due to malignant disease. This implies that the gender differences in both cases above could be explained by skewing of results by individual patients due to a relatively small number of examinations [
To our knowledge, no other investigations have been published recently on the indications for gastroscopy in pediatric surgery departments. Thus, there is no basis to compare the current study with previous results. At our centre, laparoscopic gastrostomy was the predominant indication, followed by investigation of gastroenterological conditions, obstruction in the upper gastrointestinal tract, GERD, and other indications.
Studies that have mapped gastroscopies at pediatric hospitals have shown diverse results. A retrospective study from 2014, which analysed all gastroscopies performed at a pediatric tertiary centre in Pakistan over two years, identified the most common indication as failure to thrive with suspected coeliac disease (31%) [
A study from 2010 examined the changing indications of gastroscopy over a 20-year period in the US. Children that had undergone their first gastroscopy with biopsy were included. Abdominal pain was the predominant indication at 43% [
To sum up, the results of the current study concerning the distribution of indications are quite different from previous studies. A likely explanation is that the studies presented above included mostly gastroenterological, and not surgical, patients. Overall, the number of gastroscopies performed within a gastroenterology department is much higher than the number performed within a surgery department. Thus, when mapping gastroscopies at a pediatric hospital, data from pediatric surgery patients are likely to be obscured by the much greater volume of data from gastroenterological patients. Consequently, the results of the pediatric surgery patients alone cannot be seen. It is noteworthy that no study in the 21st century has mapped gastroscopies within a pediatric surgery department.
Of the gastroscopies performed for an indication of GERD, the majority had no positive findings and almost three quarters did not lead to any further action or change in treatment. For investigation of gastroenterological conditions, almost half of the examinations had no positive findings and almost as many did not lead to any further action or change in treatment. For laparoscopic gastrostomy, there was not a single pathological finding. This amount of negative gastroscopies is remarkable, and even though several other studies present similar results [
Before deciding to perform a gastroscopy examination, the physician should consider the potential outcomes and how the patient might gain from the examination. This information will help the practitioner to analyse the risk/benefit ratio. In this consideration, it is crucial to consider risks associated with performing a gastroscopy and the ethical aspects of the procedure. There is a risk not only of medical complications, but also of negative mental impact on the child. Considering possible gain, if the aim is to confirm a suspected condition, the practitioner must ask whether the examination will change the patient management. In many cases, the patient has started treatment for the condition before the gastroscopy is conducted. And if the diagnosis is verified, there will be no change in treatment. This is the case for many patients with GERD. In this group, it might be possible to cut down on gastroscopies in order to not subject children to the risks of a gastroscopy unless it is necessary.
Naturally, in some of the cases mentioned above, the gastroscopy is performed to rule out severe conditions such as malignancies. These examinations are of great value even though there is no positive outcome and they do not lead to further action. However, these conditions are very rare in pediatric patients and the explanation would not account for all of the above examinations. In fact, some “unnecessary” gastroscopies are performed in response to pressure from the patients’ guardians or referring doctor, not necessarily because the pediatric surgeon considered the examination obligatory [
To our knowledge, this is the first study published to apply the ethical guidelines as outlined by the Swedish Council on Human Technology Assessment. Thus, it cannot be compared with other studies in this respect. Application of these guidelines did not identify any ethical issues necessitating an immediate change in the use of gastroscopy in pediatric surgical practice.
In this study, complications occurred in four of 376 (1%) gastroscopies. Two of the complications were esophageal perforations and could be considered as major. One of these had to undergo surgical treatment and one had a hospitalization prolonged by several weeks. We could find no prior studies that mapped gastroscopy complications at a pediatric surgery department. A study of over 10.000 pediatric gastroscopies from 2007, where pediatric gastroenterologists reported the results, identified a complication rate of 2.3%. The authors reported the most common “immediate complications” to be hypoxia (1.5%) and bleeding (0.3%). No deaths or perforations were reported [
Over time, there was a statistically significant increase of the annual number of gastroscopies, from an average of 114 per year in 2004 to 129 per year in 2014. The increase is supported by another study that reported a 12-fold increase in the number of gastroscopies performed over a ten-year period [
There has been an increase in the number of gastroscopies performed the past years and the indications have changed. There was a modest diagnostic yield of gastroscopy to investigate GERD or gastroenterological conditions. Furthermore, gastroscopies performed in association with laparoscopic gastrostomy surgery did not have any positive findings. The overall complication rate was 1%. The implications of this study are that practitioners and administrators in each patient’s case should carefully consider whether a gastroscopy examination is necessary. The ethical analysis did not necessitate a change in the use of this procedure but should be considered in deliberations on pediatric gastroscopy across different indications.
The authors declare that there is no conflict of interests regarding the publication of this paper.
The authors wish to acknowledge the assistance of Filip Morén, statistician, Lund University, Sweden, and Susann Ullén, medical statistician, Department of Medical Statistics and Epidemiology, Lund University, Sweden, for statistical advice and BioMed Proofreading LLC, Cleveland, Ohio, US, for linguistic revision of the paper.