Article Text
PDF
Abstract
OBJECTIVES To determine the proportion of paediatric surgical interventions that are evidence-based and to identify areas where randomised controlled trials (RCTs) or further research are required.
DESIGN Prospective review of paediatric general surgical inpatients.
SETTING A regional paediatric surgical unit.
SUBJECTS All consecutive paediatric general surgical patients admitted in November, 1995.
MAIN OUTCOME MEASURES Each patient on whom a diagnosis had been made was allocated a primary diagnosis and primary intervention (n=281). On the basis of expert knowledge, Plusnet Medline, and ISI Science Citation database searches, each intervention was categorised according to the level of supporting evidence: category 1, intervention based on RCT evidence; category 2, intervention with convincing non-experimental evidence such that an RCT would be unethical and unjustified; category 3, intervention without substantial supportive evidence.
RESULTS Of 281 patient interventions, 31 (11%) were based on controlled trials and 185 (66%) on convincing non-experimental evidence. Only 23% of interventions were category 3.
CONCLUSIONS In common with other medical specialties, the majority of paediatric surgical interventions are based on sound evidence. However, only 11% of interventions are based on RCT data, perhaps reflecting the nature of surgical practice. Further RCTs or research is indicated in a proportion of category 3 interventions.
- evidence-based medicine
- paediatric surgery.
Statistics from Altmetric.com
Evidence-based medicine aims to ensure that clinical practice is based on the best evidence available. In order to achieve this goal it is necessary to identify areas of current clinical practice where supportive evidence is lacking. For many years a popular if largely unsubstantiated dogma has been that only a small proportion of medical treatment has been proved to be beneficial.1 2 A recent study by Ellis et al has refuted this by demonstrating that 82% of acute inpatient general medicine is evidence-based, either on the basis of randomised clinical trial (RCT) data (53%) or convincing non-experimental evidence (29%).3 Similar results have been reported in the general practice setting.4 Whether other specialties can claim such a high proportion is open to question. This study aims to establish the proportion of paediatric surgery that is evidence-based and to identify areas where RCTs, or further research, is required.
Methods
All consecutive paediatric general surgical admissions to Alder Hey Children’s Hospital (n=326) during the month of November 1995 were prospectively analysed. The methodology of Ellis et al 3 was employed and each patient was allocated a primary diagnosis and primary intervention. Primary diagnosis was defined as the principal diagnosis leading to the patient’s admission. Primary intervention was defined as the main intervention (operative or non-operative) performed during admission. Patients admitted for investigation in whom no diagnosis was reached during the study period were excluded (n= 45, 14%). In the department of paediatric surgery there are five consultants and seven registrars. Only one consultant and one registrar were aware of the study during data collection.
Each primary diagnosis/intervention combination was discussed by a paediatric surgical panel consisting of one professor, two consultants, and two paediatric surgical trainees, who classified treatment according to three categories3 : category 1, treatment based on RCT evidence; category 2, there are no RCT data in support of the treatment modality but there is convincing non-experimental evidence or the validity of the intervention is such that an RCT would be unnecessary or unethical (an example of a self evident intervention is revision of a blocked ventriculoperitoneal shunt in a symptomatic patient); category 3, empirical treatment that was not evidence-based (an example of this is anal stretch in a patient with anal fissure).
Relevant published data were obtained from four sources: the panel’s collective knowledge; standard reference works; Plusnet Medline CD-Plus archives (1966–96); ISI Science Citation Index (1984–96) (Bath Information Data Services). Data were stored and analysed on an IBM-compatible personal computer on a Microsoft Access database (Microsoft Corporation, 1994).
Results
Of 281 patient interventions in the month of November 1995, 77% were evidence-based, that is category 1 (11%) or category 2 (66%) interventions (tables 1 and 2). Congenital abnormalities requiring surgical correction are listed in table 3. The remaining 23% of interventions were judged to have no evidence basis (table 4). Several cases require further explanation. The patient with acute gastric distension (table 2) had adhesive obstruction, and previous fundoplication made passage of an endoscope or nasogastric tube impossible. Surgery for recalcitrant ingrowing toenails appears as both a category 1 and 2 intervention—there is an RCT in support of phenolisation in addition to wedge excision,19 (category 1), but in some patients phenolisation was not used (category 2). ‘Acute non-specific abdominal pain’, ‘chronic constipation’, ‘blunt abdominal trauma’, and ‘acutely tender testicle’ have been included as diagnoses. It is recognised that these terms embrace a number of diseases of differing aetiologies; however, in practical terms these are useful categories within which interventional strategies are uniform. For example, when children present with an acutely tender testicle it should be assumed to be due to testicular torsion until proved otherwise.20 No diagnostic tool has been shown to be infallible21 22 and therefore surgical intervention in these cases is self evident.
Category 1 interventions based on RCT data
Category 2 interventions (self evident interventions or where a RCT would be unethical)
Category 3 interventions (interventions with no self evident basis or supportive RCT data)
Category 2 interventions (congenital malformations requiring surgical correction)
Discussion
The majority of paediatric surgical interventions in our regional paediatric surgical unit are evidence-based and this is in broad agreement with findings from other medical specialties.3 4 However, the proportion of category 1 (RCT data) interventions was substantially lower than in the case of inpatient general medicine or general practice (fig 1). The relative lack of RCT data supporting paediatric surgical interventions probably reflects the different nature of surgical as against medical practice. By their very nature, many surgical interventions can be classified as ‘self evident’, for example, correction of life threatening congenital malformations. In addition, much paediatric surgical pathology is rare, with individual centres seeing relatively small numbers of certain conditions. This creates difficulty in RCT design and execution.
Analysis of the 23% of interventions lacking supportive evidence allows them to be subclassified into two further groups: (1) heterogeneous ill defined conditions such as chronic constipation in which the primary pathology is not well understood and is almost certainly multifactorial; treatment is thus empirical and RCT data are lacking; (2) conditions such as bleeding granulation tissue or anal fissure, where subjective clinical experience suggests that patients benefit from the intervention employed.
We recognise that the current study design has limitations. Although several search protocols were employed, it is likely that all relevant literature has not been identified.23 This would lead to an underestimate of the number of interventions that are based on sound evidence. The two investigators who were aware of the study during data collection were responsible for making diagnoses/ordering interventions and it could be argued that this could have altered clinical practice. However, if this were true it would only affect a small proportion of cases.
The reductionist approach necessarily adopted for both diagnosis and intervention belies the complexity of this particular group of patients. Patients often suffer from a number of disorders in addition to the primary diagnosis. For example, duodenal atresia may be associated with Down’s syndrome and a cardiac defect. Clinical practice is also influenced by less tangible but equally important psychological and social factors, thus, for an antegrade continent enema procedure to be successful, patients need to be well motivated and have specialised nursing support.24 Categorising interventions as 1, 2, and 3 should not be equated with ranking in terms of clinical value, for example closure of tracheo-oesophageal fistula is a life saving but category 2 intervention.
The limitations of RCTs, particularly within a surgical context, are well recognised25 ; furthermore, RCT findings are open to various interpretations. The RCT quoted in support of appendicectomy for acute appendicitis concluded that outcome from conservative treatment was similar to that for operative management.5However, it was the consensus opinion of the panel that the high rates of recurrent appendicitis coupled with unknown long term outcomes in terms of episodes of adhesive obstruction and female infertility did not justify this conclusion. Finally, some trial data cited here involve adult patients and the relevance of such findings to paediatric practice cannot be assumed.
This study has added weight to the argument that most current medical practice is ‘evidence-based’. As in all areas of medical practice, difficulties may be encountered in the application of RCT and other evidence to individual patients. Thus the results of interventions performed in large centres by enthusiasts with wide support may not be achievable in other centres.26 Evidence-based medicine is only a tool in the surgical decision making process.27This review has highlighted areas warranting further examination such as interventions for chronic constipation and the management of anal fissure. The findings from this study are to be incorporated into development of evidence-based critical care pathways where appropriate.