Oesophageal atresia: Are "long gap" patients at greater anesthetic risk?

BACKGROUND: Long gap oesophageal atresia occurs in approximately 10% of all oesophageal atresia infants and surgical repair is often difficult with significant postoperative complications. Our aim was to describe the perioperative course, morbidity, and early results following repair of long gap oesophageal atresia and to identify factors which may be associated with complications. METHODS: This is a single center retrospective cohort study of consecutive patients with oesophageal atresia undergoing surgical repair at The Royal Children's Hospital Melbourne from January 2006 to June 2017. RESULTS: Two hundred and thirty-nine consecutive oesophageal atresia infants included 44 long gap oesophageal atresia infants and 195 non-long gap infants. A high rate of prematurity (24.7%), major cardiac (17%), and other surgically relevant malformations (12.6%) was found in both groups. The median age at oesophageal anastomosis surgery was 65.5 days for the long gap group vs 1 day for the oesophageal atresia group (mean difference 56.8 days, 95% CI 48.1-65.5 days, P < .01). Surgery for long gap oesophageal atresia included immediate primary anastomosis (n = 10), delayed primary anastomosis (n = 11), oesophageal lengthening techniques (n = 12) and primary oesophageal replacement (n = 6). Long gap oesophageal atresia was not associated with an increased incidence of difficult intubation (OR 2.8, 95% CI 0.6-22.1, P = .17), intraoperative hypoxemia (OR 1.6, 95% CI 0.6-4.5, P = .32), or hypotension (OR 0.9, 95% CI 0.5-1.8, P = .81). The surgical duration (177.7 vs 202.1 minute, mean difference [95% CI], 28 [5.5-50.4 minutes], P = .04) and mean duration of postoperative mechanical ventilation (107 vs 199.8 hours, mean difference [95% CI], 91.8 [34.5-149.1 hours], P < .01) were shorter for the non-long gap group. Overall in-hospital mortality was 7.5% (15.9% long gap vs 5.6% non-long gap oesophageal atresia OR 1.1, 95% CI 0.4-3.4, P = .85). CONCLUSION: Long gap oesophageal atresia infants have a similar incidence of perioperative complications to other infants with oesophageal atresia. Current surgical approaches to long gap repair, however, are associated with longer anesthetic exposures and require multiple procedures in infancy to achieve oesophageal continuity.

Achalasia: Outcome in children.

BACKGROUND: Oesophageal achalasia is well-recognized but relatively rare in children, occasionally appearing as the "triple A" syndrome (with adrenal insufficiency and alacrima). Treatment modalities, as in adult practice, are not curative, often needing further interventions and spurring the search for better management. The outcome for syndromic variants is unknown. We sought to define the efficacy of treatments for children with achalasia with and without triple A syndrome. METHODS: We conducted a retrospective analysis of presentation and outcomes for 42 children with achalasia presenting over three decades to a major pediatric referral center. Long term impact of the diagnosis was assessed by questionnaire. RESULTS: We identified 42 children including six with triple A syndrome. The median overall age at diagnosis was 10.8 years and median follow-up 1593 days. Initial Heller myotomy in 17 required further interventions in 11 (65%), while initial treatment with botulinum toxin (n = 20) was ultimately followed by myotomy in 17 (85%). Ten out of 35 patients who underwent myotomy required a repeat myotomy (29%). Patients with triple A syndrome developed symptoms earlier, but had delayed diagnosis, were more underweight at diagnosis and at last follow up. Questionnaire results suggested a significant long term deleterious impact on the quality of life of children and their families. CONCLUSION: Many children with achalasia relapse after initial treatment, undergoing multiple, different procedures, despite which symptoms persist and impact on quality of life. Symptoms develop earlier in patients with triple A syndrome, but the diagnosis is delayed and this has substantial nutritional impact.

A review of anesthetic techniques and outcomes following minimally invasive repair of pectus excavatum (Nuss procedure).

BACKGROUND: Pectus excavatum (PE) is the most common congenital chest wall deformity, occurring in 1 : 1000 children with a male to female ratio of 4 : 1. Several procedures have been described to manage this deformity, including cartilage resection with sternal osteotomy (the Ravitch procedure) and a minimally invasive repair technique (the Nuss procedure). While initially described as a nonthoracoscopic technique, the current surgical approach of the Royal Childrens Hospital involves thoracoscopic assistance. Postoperative pain is significant in patients undergoing the pectus repair and multiple analgesic regimens have been advocated with continuous thoracic epidural infusions and opioid infusions the most common. Some authors have advocated patient-controlled analgesia (PCA), paravertebral nerve blocks (PVNB), and wound infusion catheters as alternatives. AIMS: The primary aim of this study was to assess our experience with postoperative pain and analgesia requirements associated with the minimally invasive repair of pectus excavatum in children. METHODS: This is a retrospective cohort study with a contemporaneous comparison group examining patients treated between January 2005 and December 2015 for minimally invasive repair of pectus excavatum by the Nuss procedure. RESULTS: Two hundred and seventeen patients [mean age 14.9 (sd 1.9) years] with pectus excavatum treated at the Royal Childrens Hospital between 2005 and 2015 were identified. All patients were managed with thoracic epidural analgesia and intravenous morphine infusions. The epidural was effective in the postanesthesia care unit in 97.3% (failure to place an epidural rate was 4 (1.9%); no block on awakening 0.9%). A further 4 (1.8%) were removed within 24 h. The mean morphine equivalent dose in the first 24 h was 0.8 mg·kg-1 ·day-1 . PCA was continued for a mean of 3.8 days and the total mean morphine equivalent dose was 2.2 mg·kg-1 ·day-1 . Minor complications occurred in 67 (30.9%) with postoperative nausea and vomiting in 36 patients (16.6%) and urinary retention requiring an indwelling catheter in 40 patients (18.4%). CONCLUSION: An epidural-based analgesic regime is associated with low pain scores and few acute complications. The continuation of morphine analgesia after the first postoperative day is common but associated with an increased incidence of urinary retention and nausea and vomiting.

Central venous access device-related infections in patients with haemophilia.

AIMS: Fully implantable central venous access devices (CVADs) can offer long-term reliable venous access to facilitate regular factor replacement therapy in haemophilia. However, CVAD-related infection remains a major deterrent to the optimal use of CVAD in this population. This report represents the first review of CVAD use in haemophilia in Australia and aims to examine the rate of complications including CVAD-related infections. METHODS: A retrospective review of medical records was conducted of all haemophilic patients with fully implantable CVADs at the Royal Children's Hospital (RCH), Melbourne, between 1 June 1992 and 30 June 2009. CVAD-related bloodstream infection was defined based on the guidelines from the Centre of Disease Control and Victoria National Nosocomial Infection Surveillance. To further enhance identification of CVAD-related infection in this study, a third criterion of 'suspected infection' was added by the authors. RESULTS: Eighty-one CVADs in 56 patients were managed at the RCH during this time period resulting in a combined study period of 94 756 CVAD days. Median age at first CVAD insertion = 2.16 years (range 0.66 to 13.98 years). CVADs were inserted predominantly due to difficult venous access and prophylaxis initiation (70.4%). Median life-span of a CVAD was 1227 days, equivalent to 3.36 years (n = 50; range 0.22 to 9.44 years). Fifty-seven CVAD-related infections occurred in 37 CVADs (46.3%) in 29 patients (51.8%). Overall incidence of confirmed CVAD-related bloodstream infection = 0.42 per 1000 CVAD days (95% confidence interval (CI): 0.31 to 0.58 per 1000 CVAD days) and indicate better performance compared with the published benchmark of 0.66 per 1000 CVAD days (0.44 to 0.97 per 1000 CVAD days). The incidence of both confirmed (criteria 1, 2) and suspected (criterion 3) CVAD-related infection is 0.60 per 1000 CVAD days (95% CI: 0.46 to 0.78), which is comparable to the international benchmark. The majority of CVAD-related infections (73.7%) were successfully treated with intravenous antimicrobials without necessitating CVAD removal. Klebsiella pneumoniae was the most common organism found in positive blood cultures. CONCLUSION: CVAD-related infection in this Australian population was comparable to rates described in the medical literature. Ongoing surveillance for infection rates is important to provide an up-to-date assessment of risks associated with CVAD use in this population.

Pancreatic trauma in children.

PURPOSE: To document the demographics, mechanisms and outcome of traumatic pancreatitis in children at a single large tertiary referral centre in Australia. METHODS: We undertook a 10-year retrospective audit of children admitted to the Royal Children's Hospital, Melbourne, Australia with a hospital coded diagnosis which included pancreatic injury between 1993 and 2002. Data included patient demographics, source of admission, mechanism of injury, pancreatic complications, associated injuries, intensive care unit admission, results of any operative findings, results of any acute computed tomography and/or ultrasound imaging of pancreas, selected laboratory findings and length of stay. RESULTS: We identified two distinct groups of patients in the 91 documented cases of pancreatic trauma (median age 8.0 years, range 0.6-15.8 years; M:F 2.5:1.0): 59 had a history of abdominal trauma and elevated serum lipase but no CT or ultrasound evidence of pancreatic injury (Group A); 32 had a history of abdominal trauma, elevated serum lipase but also had CT scan and/or ultrasound evidence of pancreatic injury (Group B). Patients with "less severe" injury based on normal imaging had a lower initial lipase level [Group A, median 651 U/L (interquartile range 520-1,324) vs. Group B, 1,608 U/L (interquartile range 680-3,526); p = 0.005] and shorter admission time [Group A, 9.0 days (interquartile range 5.5-15.5) vs. Group B, 13.4 days (interquartile range 6.8-23.8); p = 0.04]. There were no differences with respect to mortality (Group A, 13.5% vs. Group B, 12.5%), but patients with evidence of injury on imaging were more likely to have surgical intervention (p = 0.0001). The single most important overall cause of pancreatic trauma was involvement in a motor vehicle accident as a passenger or pedestrian. However, in children with high-grade ductal injury, bicycle handlebar injuries were most common. Associated injuries were common in both groups. CONCLUSION: Significant pancreatic injury can occur in the absence of abnormality on medical imaging. Pancreatic trauma commonly occurs in the context of multiple injuries after motor vehicle accidents in children and bicycle handlebar injuries, especially in boys. Most children can be treated conservatively, with surgical intervention being limited to high-grade ductal injury.

Reducing "cry wolf"--changing trauma team activation at a pediatric trauma centre.

BACKGROUND: To improve utilization of scarce surgical resources, we changed from a single tier trauma paging system (TPS) to a three tiered TPS at a tertiary pediatric trauma center. We investigated if patients were appropriately classified into the three levels of trauma team activation. METHODS: Trauma registry data were used to review data 12 months before and after implementation of a three tiered TPS (level I entire team present, level II surgical subspecialties within 10 minutes, level III emergency department team only at patient arrival). We correlated TPS activation with proxies of injury severity (admission status and major/nonmajor trauma). RESULTS: There were 192 activations during 12 months of the single tier TPS and 216 during the three tier TPS (33 level I, 49 level II, and 134 level III). The entire team was to attend in all 192 single tier and in 82 (40%) level I and II three tier TPS activations i.e., there were 60% fewer surgical team activations. During single tier TPS, 96% patients were admitted and 23% classified as major trauma. Three tiered TPS level I, II and III were admitted in 97%, 94%, and 81% and classified as major trauma in 58%, 35%, and 15%, respectively. Of the 20 level III patients classified as major trauma, TPS level was deemed appropriate in 18 and inappropriately low in 2, although patient care had not been compromised. CONCLUSION: Our results suggest that a three tiered TPS more efficiently utilizes limited surgical resources without leading to major misclassifications.

Congenital cystic adenomatoid malformation: monitoring the antenatal and short-term neonatal outcome.

OBJECTIVE: To determine the antenatal and short-term neonatal outcome of antenatally detected congenital cystic adenomatoid malformation (CCAM). METHODS: A retrospective review was conducted on all women with an antenatal diagnosis of CCAM who attended the Royal Women's Hospital, Melbourne, between January 1995 and December 2005. RESULTS: An antenatal diagnosis of CCAM was made in 38 singleton pregnancies. Serial ultrasounds were performed in 34 cases. Thirteen lesions (38%) appeared to resolve, ten lesions (29%) appeared to decrease, four lesions (12%) appeared to remain unchanged and seven lesions (21%) appeared to increase in size. Four pregnancies (10.5%) were complicated by hydrops fetalis which was associated with a poor outcome. Thirty-seven babies were liveborn. Seven babies (18.9%) developed respiratory distress. Two of these babies died within two days of birth resulting in three deaths in total. Two babies were lost to follow up (n=33). Sixty per cent of babies in whom the CCAM appeared to resolve on antenatal ultrasound had an abnormal chest X-ray (CXR). All computed tomography (CT) scans (eight of eight) in this group were abnormal. Of the surviving babies, 27.3% (nine of 33) have had surgery to date. CONCLUSIONS: The antenatal diagnosis of CCAM is associated with a good short-term prognosis. The pregnancy should initially be managed at a tertiary centre with serial ultrasound. Asymptomatic babies should have a postnatal CT even if the CCAM appears to have resolved or decreased on antenatal ultrasound.

The effect of thoracoscopy upon the repair of pectus excavatum.

BACKGROUND/AIMS: The Nuss procedure is the most commonly performed operation to correct pectus excavatum (PE). Thoracoscopic assistance has been anecdotally noted to improve the safety of this operative approach. This study aimed to compare complications and clinical outcomes before and after the introduction of thoracoscopy in a single-center. METHODS: A retrospective review was performed of all patients who underwent the Nuss procedure at The Royal Children's Hospital over an 11-year period (2005-2015), collecting data on all intra-operative and post-operative outcomes. RESULTS: A total of 217 Nuss procedures were performed (122 non-thoracoscopic pectus repairs, 95 thoracoscopic pectus repairs). Median patient age was 14.9years, with the majority male (185/217, 84.3%). Patient demographics (age, gender, defect severity) and postoperative recovery were comparable between the two groups. Major complications included cardiac arrest requiring internal cardiac massage, hemothorax, pneumothorax, empyema, bar displacement and infection. The overall major complication rate was low (19/217, 8.8%); however, there was a significant reduction in major complications in the thoracoscopic pectus repair group (13.1% versus 3.2%, p=0.02). CONCLUSIONS: Thoracoscopic vision during the Nuss procedure reduces the risk of major complications. LEVEL OF EVIDENCE: Treatment study - Level III (Retrospective comparative study).

Gastroesophageal reflux and related pathology in adults who were born with esophageal atresia: a long-term follow-up study.

BACKGROUND & AIMS: Esophageal atresia (EA) is the most common congenital anomaly of the esophagus. There are few long-term follow-up data on adults who had surgery for EA as infants. The primary aims were to evaluate the prevalence of esophageal symptoms and pathology and second to develop recommendations for follow-up. METHODS: This is a descriptive study of individuals attending a clinic in an adult tertiary referral hospital, established to provide care for adults who had surgery for EA as infants. Individuals aged 20 years or older were identified from an existing database and invited by telephone to attend the clinic. One hundred thirty-two patients attended the clinic from 2000-2003. Individuals were assessed by using a structured questionnaire. Endoscopy was performed in 62 patients because of symptoms. RESULTS: Reflux symptoms were reported by 83 (63%), including 25 (19%) with severe symptoms. Dysphagia was reported by 68 patients (52%). Of those who underwent endoscopy, reflux esophagitis was present in 36 patients (58%), Barrett's esophagus in 7 (11%), and strictures in 26 (42%) patients. One patient was diagnosed with esophageal squamous cell carcinoma. Men who were 35 years or older and individuals with severe reflux symptoms were at high risk of having severe esophagitis or Barrett's metaplasia. CONCLUSIONS: Reflux symptoms, esophagitis, and Barrett's esophagus are common in these individuals. We recommend clinical assessment as adults and upper endoscopy for reflux symptoms or dysphagia. Transition of young adults from pediatric care to an adult gastroenterology clinic with expertise in EA appears to be highly beneficial.

Chest x-ray as a screening tool for blunt thoracic trauma in children.

BACKGROUND: With the increasing use of thoracic computed tomography (CT) to screen for injuries in pediatric blunt thoracic trauma (BTT), we determined whether chest x-ray (CXR) and other clinical and epidemiologic variables could be used to predict significant thoracic injuries, to inform the selective use of CT in pediatric BTT. We further queried if these were discrepant from factors associated with the decision to obtain a thoracic CT. METHODS: This retrospective cohort study included cases of BTT from three Level I pediatric trauma centers between April 1999 and March 2008. Pre-CT epidemiologic, clinical, and radiologic variables associated with CT findings of any thoracic injury or a significant thoracic injury as well as the decision to obtain a thoracic CT were determined using logistic regression. RESULTS: Of 425 patients, 40% patients had a significant thoracic injury, 49% had nonsignificant thoracic injury, and 11% had no thoracic injury at all. Presence of hydrothorax and/or pneumothorax on CXR significantly increased the likelihood of significant chest injury visualized by CT (adjusted odds ratio 10.8; 95% confidence interval, 6.5-18), as did the presence of isolated subcutaneous emphysema (adjusted odds ratio, 19.8; 95% confidence interval, 2.3-168). Although a normal CXR finding was not statistically associated with a reduced risk of significant thoracic injury, 8 of the 9 cases with normal CXR findings and significant injuries involved occult pneumothoraces or hemothoraces not requiring intervention. Converse to features suggesting increased risk of significant injury, the decision to obtain a thoracic CT was only associated with later period in the study and obtaining a CT scan of another body region. CONCLUSION: CXR can be used to screen for significant thoracic injuries and direct the selective use of thoracic CT in pediatric BTT. Prospective studies are needed to validate these findings and develop guidelines that include CXR to define indications for thoracic CT in pediatric BTT. LEVEL OF EVIDENCE: Prognostic study, level III.

Paediatric and adolescent trauma care within an integrated trauma system.

BACKGROUND: The aim of this study was to establish the profile and outcomes of paediatric major trauma care (PTMC) within an integrated inclusive regionalised trauma system. METHODS: Prospectively collected data from July 2001 to June 2009 from the Victorian State Trauma Registry of patients aged <18 years were reviewed. RESULTS: There were 1634 major trauma cases with a median (IQR) age of 13 (6-16) years and 69% were male. The median ISS (IQR) was 18 (16-26). There were 1361 patients treated at a major trauma centre of which 69% (n=943) were treated at the PMTC. Head injury (AIS>2) was the most frequent injury (n=950, 58%). Surgery was required in 39% (n=637) of all cases; 437 patients in the 10-17 year old group and 200 patients in the 0-9 year old group; the mortality was 6.6%. There were 530 patients (32.4%) ventilated in ICU; these had a median ISS (IQR) of 25 (17-34) and mortality of 7.4%. Improvements in risk-adjusted mortality have occurred as the years have progressed [adjusted OR 95% CI: 0.87 (0.76, 0.99)] and being treated at a Level 1 trauma centre was associated with lower adjusted odds of mortality [adjusted OR 95% CI: 0.27 (0.11, 0.68)]. CONCLUSION: The establishment of this integrated inclusive regionalised trauma system has been associated with progressively improving risk-adjusted mortality. The relatively low volume of major trauma requiring surgery in the 0-9 year old age group is notable, creating a challenging environment for maintaining skills and institutional preparedness.

Lung function in children with repaired tracheo-oesophageal fistula using the forced oscillation technique.

BACKGROUND: Tracheo-oesophageal fistula (TOF) and oesophageal atresia (OA) are congenital anomalies commonly associated with pulmonary complications during early childhood. This study investigated the role of the forced oscillation technique (FOT) in assessing lung function in young children with repaired TOF/OA. METHODS: Forty children with repaired TOF/OA of median (range) age 8.0 (3.3-10.6) years, and 20 healthy children without TOF aged 6.1 (3.1-10.8) years were studied. FOT measurements were attempted in all subjects and spirometry only in those 6 years and above. Resistance and reactance (both hPasL(-1)) at 6 Hz (Rrs6 and Xrs6, respectively) and 8 Hz (Rrs8 and Xrs8) measured using FOT, and forced expired volume in 1 sec (FEV(1)), forced vital capacity, functional residual capacity, total lung capacity, and residual volume (all L) obtained from spirometry or plethysmography were compared with reference values and expressed as z-scores. RESULTS: Technically acceptable measurements of Rrs6, Rrs8, Xrs6, Xrs8, Fdep, and Fres were obtained in 37 children with TOF and 20 healthy children without TOF, respectively. Those with TOF had significantly higher mean (SD) z-scores for Rrs6 [0.99 (0.75)] versus healthy children without TOF [0.31 (0.69)] and lower mean (SD) z-scores for Xrs6 [-1.04 (1.07)] versus healthy children without TOF [-0.34 (0.83)]. Spirometry was successful in 24 of the 29 with TOF in whom it was attempted and all healthy children without TOF. Mean (SD) z-score for FEV(1) was significantly lower in those with TOF [-0.86 (1.13)] versus healthy children without TOF [0.67 (0.54)]. z-Scores for Rrs6 and FEV(1) were significantly correlated (r = -0.49; P = 0.003). CONCLUSIONS: Children with repaired TOF have diminished lung function compared with healthy children. FOT is sensitive and correlates well with standard spirometry. It can be used to measure lung function in younger children when spirometry is difficult to perform and should be considered as an objective method for monitoring clinical progress in young children with TOF.

The Victorian major trauma transfer study.

AIMS: To comprehensively examine the inter-hospital transfer of major trauma patients-including the reason for transfer, duration, escorts, interventions and unexpected events. METHODS: This was an detailed study of the transfer of major trauma cases in the State of Victoria, Australia, between April 16, 2003 and December 31, 2004. Twenty-three hospitals and seven transfer/retrieval services participated. Defined major trauma cases that were transferred between participating hospitals for the purpose of definitive care were eligible for enrolment. The transfer phase extended from 30 min before until 30 min after the transfer. The transferring and receiving hospitals and the transfer escorts were asked to record data on a specifically designed data collection form. RESULTS: A total of 451 cases were enrolled (mean Injury Severity Score 22.2). Transfers originated mainly from Regional Trauma (42.8%) and Metropolitan Trauma (31.3%) Services and most (90.5%) terminated at a Major Trauma Service. Median time from injury to arrival at the receiving hospital was 8 h 30 min. Median time from arrival at referring hospital to request for transfer was 3 h 25 min. Escorts comprised ambulance and medical/nursing staff in 67.0% and 30.4% of cases, respectively. Metropolitan retrieval services were involved in only 10% of cases. Medical escorts were mainly (62.9%) from the referring hospital and the majority of these were registrars (49.4%) and hospital medical officers (HMOs, 16.9%). Overall mortality was 6.2%. Mortality rates for cases escorted by referring hospital doctors, Mobile Intensive Care Ambulance (MICA), non-MICA and any other escorts were 14.5%, 6.0%, 2.6% and 4.3%, respectively. HMO escorts had the highest mortality risk (OR 3.67, 95%CI 1.00-13.49, p<0.001). Mortality risk was greatest for cases that required administration of vasopressor drugs (OR 11.4, 95%CI 3.78-34.36, p<0.001), intubation prior to arrival at the referring hospital (OR 10.36, 95%CI 3.51-30.52, p<0.001), any interventions at the referring hospital (OR 8.3, 95%CI 3.1-22.2, p<0.001), administration of blood at the receiving hospital (OR 4.91, 95%CI 1.5-16.1, p=0.01), and cases using escorts from the referring hospital (OR 3.8, 95%CI 1.69-8.39, p=0.001). CONCLUSION: Considerable variability in request for transfer and transfer times, transfer escorts and mortality risk exist. The single greatest issue identified that most severely injured group were escorted by the most junior doctors (HMOs) and had the highest mortality. This crucial issue must be addressed by the State Trauma System and by any redesigned retrieval service in Victoria. A detailed review of activation and responsiveness criteria and the nature of the transfer escort is indicated. The establishment of Adult Retrieval Victoria may address many of the concerns raised by this study.

Changing incidence of acute pancreatitis: 10-year experience at the Royal Children's Hospital, Melbourne.

BACKGROUND AND AIM: The aim of this study was to assess the incidence and etiology of acute pancreatitis at a major pediatric referral center in Australia. METHODS: A 10-year retrospective audit was conducted at The Royal Children's Hospital, Melbourne, Australia. All patients from 1993 and 2002 with a serum lipase level greater than three times the upper reference range and a history consistent with acute pancreatitis were included. RESULTS: During the 10-year period, 279 confirmed cases of acute pancreatitis were identified. The median age at presentation was 10 years (range, 0.2-15.9). In 209 (74.9%) patients, a likely cause of acute pancreatitis was found, including trauma (36.3%), systemic disease (22.2%), metabolic (5.8%), biliary (5.4%), drugs (3.2%), or viral illness (2.2%). In the remaining 70 (25.1%) cases, the pancreatitis was deemed idiopathic. Comparing data from 1993 to 1997 with data from 1998-2002, there was a significant increase in the annual incidence of pancreatitis (24.6 +/- 2.3 vs 31.2 +/- 6 cases per year; P = 0.04). A linear regression analysis showed a strong association between the incidence and the year of diagnosis (r(2) = 0.5775, P = 0.01). This increase was mainly due to a significant rise in idiopathic disease (r(2) = 0.83, P = 0.0002) and systemic disease (r(2) = 0.41, P = 0.048), whereas the incidence of other causes of acute pancreatitis remained unchanged. CONCLUSION: The incidence of acute pancreatitis in children has increased significantly over the past decade. The increase was greatest in children with idiopathic pancreatitis. It remains unclear whether this reflects a true incidence increase or improved clinical awareness.