A Model Predicting Perforation and Complications in Paediatric Appendicectomy

Obinna Obinwa1*,#a, Colin Peirce1,#b, Michael Cassidy1, Tom Fahey2, John Flynn1,2,

1Department of Surgery, Portiuncula Hospital, Ballinasloe, County Galway, Ireland 2Department of Pathology, Portiuncula Hospital, Ballinasloe, County Galway, Ireland

3Division of Population Health Sciences, Royal College of Surgeons in Ireland, Dublin 2, Ireland

*Corresponding author

E-mail: [email protected]

#a Current Address: Department of Vascular Surgery, Mater Misericordiae University Hospital, Dublin, Ireland.

Telephone: +353 1 803 2000

#b Current Address: Department of Surgery, St Vincent University Hospital, Dublin, Ireland

1 ______

Abstract

Purpose: To analyse the diagnostic value of simple clinical measurements in ensuring an early and accurate detection of advanced appendicitis (perforation, mass and peri-appendicular abscess) and possible complications. Methods: A retrospective, single centre study of all paediatric (age 0-14 years) appendicectomies was conducted over a 14-year period. Preoperative symptoms, signs and laboratory results, intra-operative findings and postoperative complications were analyzed. Receiver Operating Characteristic (ROC) curves were used to estimate sensitivity and specificity of significant (p ≤ 0.05) predictor variables based on multivariate logistic regression models. Results: One thousand and thirty-seven patients were included. Perforations were seen in 88 (8.5%) cases, and abscesses were seen in 35 (3.4%) cases. Of all the clinical variables evaluated, preoperative temperature ≥ 37.5oC was most discriminatory for advanced appendicitis. Significant other discriminatory clinical variables were WBC count ≥ 15,100/ʮL, preoperative anorexia and rebound tenderness. Postoperative complications occurred in 74 (7.1%) patients, and were associated with preoperative temperature ≥ 37.5 oC and advanced appendicitis Conclusion: Independent clinical predictors of advanced appendicitis exist but lack individual accuracy. In this study, preoperative pyrexia is shown to be highly associated with both advanced appendicitis and development of postoperative complications. This independent factor may point to early need for antibiotic treatment, urgent imaging and subsequent intervention in patients with appendicitis.

Key Words: antibiotic therapy, paediatrics, outcomes assessment, appendicitis, risk factors

______

2 Introduction

Advanced appendicitis (perforation, appendicular mass or abscess) is a significant health burden. The worldwide incidence of acute appendicitis is estimated to be 86 cases annually per 100 000, with approximately

70 000 paediatric cases in the United States alone. The incidence of perforation ranges from 20% to 76% of cases in the paediatric population. Post-appendicectomy complication rates range from 6.4% to 29.3% with perforation accounting for the majority of cases. Mortality from appendicectomies is estimated at 0.5 per 100

000 population.

The protocols for the diagnosis and management of a pediatric advanced appendicitis vary across the globe. The classical clinical presentation of the disease is well known, however, the significance of individual symptoms, signs and laboratory measurements in predicting outcomes requires further investigation. The presentation of a child with appendicitis may also be delayed resulting in an advanced appendicitis and increased risk of morbidity. An early and accurate diagnosis, therefore, is paramount in order to decrease the risks of peritonitis and sepsis.

The present study focuses on the diagnostic usefulness of elements of disease history, clinical findings and white blood cell count in predicting advanced appendicitis and postoperative complications. The primary aim of the study was to evaluate to what extent the preoperative clinical features predict advanced appendicitis. A secondary aim was to evaluate to what extent these preoperative clinical features are predictive of postoperative complications in the same cohort.

Materials and Methods

A retrospective chart review of all paediatric (age 0-14 years) appendicectomies performed between January

1995 and December 2008 in a single institution was conducted. Data were collected in an anonymized format, and the collected data were maintained in a Microsoft ® Access 2007 database. The preoperative data recorded were: demographic information (age and sex), clinical symptoms at presentation (anorexia, nausea/vomiting, migratory pain) and clinical signs (guarding and rebound tenderness). The white blood cell (WBC) count and the maximum recorded preoperative vital signs (temperature, pulse, and respiratory rate) prior to surgery were

3 also recorded. Operative findings were graded G0 for a macroscopically normal appendix, G1 for acute appendicitis, G2 for gangrenous acute appendicitis, G3 for perforation or phlegmon and G4 for a peri- appendicular abscess. The duration of empirical antibiotic treatment was recorded in all patients.

Postoperative data included any postoperative complications recorded within the 30-day postoperative period.

Prolonged ileus was defined as persistence of ileus (i.e. a decrease or absence of intestinal peristalsis with abdominal pain, nausea, vomiting or the absence of flatus necessitating a delay in beginning diet or stopping diet) after the third postoperative day. Intra-abdominal abscess was confirmed by imaging and microbiological samples. Wound infection was confirmed clinically and by microbiological samples. The study was approved by the Galway University Hospitals Clinical Research Ethics Committee in April 2010.

Data Analysis

Data analysis was performed with Microsoft Excel and STATA statistical software version 11 (StataCorp LP,

College City, Texas). Continuous data were expressed as means or medians where appropriate. WBC count, preoperative temperature, pulse and respiratory rate were further converted to categorical variables

(leucocytosis, pyrexia, tachycardia, and tachypnea respectively) by using a cut point that maximized the

Youden’s index. The Youden’s index is the value that maximizes the sensitivity and specificity of any continuous variable, helping to choose an appropriate cut point for dichotomization. All categorical data were expressed as the number of patients and percentages. The trend between the pathological grade and age was determined by linear regression analysis, and that between the pathological grade and sex-ratio was assessed using Mantel-Haenszel chi-square test. The Mantel-Haenszel chi-square test was also used to assess the trend between preoperative clinical features, postoperative complications and the pathological grade.

The Regression Models

The predictors of post-appendicectomy complications were identified using a binary logistic regression modelling technique, and that of advanced appendicitis were described using ordinal logistic regression modelling. Odds ratios (OR) and 95% confidence intervals (CI) were determined in the regression models, and the result was statistically significant when p ≤ 0.05. The performance of the clinical diagnositic model was assessed with respect to its discrimination The concordance-statistic (c-statistic) was used to quantify the discriminative ability and value is identical to the area under the receiver operating characteristic curves (AUC).

4 A high value of AUC implies a high discriminatory ability. The likelihood ratio, sensitivity, specificity and

AUC were calculated for each or other combination of clinically significant predictor variables.

Results

Data were retrieved for 1,037 patients who underwent appendicectomy between January 1995 and December

2008. The distribution of the pathological findings is shown in Table 1. The negative appendicectomy rate was

21% (220/1037). Different pathologies were encountered in 112/220 of the G0 cases, and the aetiology was unknown in the remaining 108/220 cases. G0 pathologies included 98 mesenteric adenitis, six ovarian cysts and eight others. The latter pathologies were: Meckel’s diverticulum, endometriosis, mesenteric cyst, omental torsion, terminal ileitis, torsion of the appendices epiploicae, volvulus, and intraabdominal adhesion.

Advanced appendicitis (G3/G4) was more frequent in males and very young children. The incidence of advanced appendicitis was 71% among 1-year olds. Incidence rates of 19-67% were seen in the 2-5year old group, 10-16% among 6-9 year olds and 7-9 % in the 10-14 year old group. A significantly higher proportion of males to females had appendicitis (G1/G2/G3) compared to those with a G0 finding (P < 0.01, Table 1).

Patients with elevated temperature, pulse and respiratory rate were more likely to have advanced appendicitis than those without (P <0.001). Similarly those with increased white blood cell count were also more likely to have advanced appendicitis than those without (p < 0.001).

There were 86 postoperative complications in 74 (7.1%) patients (Table 2). Compared to a negative appendicectomy, the risk of complications were 4.7 times higher if advanced appendicitis (P <0.001) and 1.5 times less if an appendicitis were accurately detected (P = 0.168). Complications requiring surgical intervention included three cases of intestinal obstruction that did not settle with conservative management, and five wound explorations. None of the patients required a stay in intensive care for organ dysfunction and no mortalities were recorded.

Predictive model for advanced appendicitis

The univariate analysis showed that a large proportion of the patients presented with atypical features making clinical assessment difficult. A history of migratory pain, nausea or vomiting was insufficient in differentiating the different pathologies (P > 0.05). The classical clinical variables of anorexia, rebound tenderness and guarding each had a statistically significant discriminatory power for identifying the G3/G4 cases (ROC curve 5 areas 0.56 to 0.62). Preoperative pyrexia (temperature ≥ 37.5oC), tachycardia (pulse rate ≥ 102/min), and leukocytosis (WBC count ≥ 1,500/ʮL) showed a higher discriminatory power for identifying the G3/G4 cases than the other clinical variables (ROC curve areas of 0.69 to 0.72). There were fewer false positives than with the clinical assessessment variables (LR+ 2.18 to 6.35 versus LR+ 1.31 to 1.72). Advanced appendicitis was unlikely in the absence of anorexia (LR- 0.28), leukocytosis (LR- 0.35) and if the patient did not record a fever before surgery (LR- 0.36).

Multivariate analysis showed that independent clinical predictors of advanced disease were leucocytosis, pyrexia, rebound tenderness and anorexia (Table 3). Younger age affected the clinical assessment of patients with advanced appendicitis. This effect was most marked in children younger than 11 years (Figure 1). The adjusted OR associated with anorexia and rebound tenderness in the clinical model described were higher when adjusting for age compared to the values obtained without this (Table 3). The model strongly discriminated patients with advanced appendicitis from those without an advanced appendicitis (c-statistics = 0.828). Thirty- eight percent patients with advanced appendicitis (G3/G4) were correctly identified, and the predicted negative appendicectomy (G0) rate was 0.9%. Individual sensitivities of the identified predictors are displayed in Figure

Predictive model for postoperative complications

The univariate model showed that advanced appendicitis was a risk factor for the development of complications

(OR 4.84 95% CI 2.87 – 8.17). According to the univariate analysis, those with preoperative leukocytosis (OR

1.82), pyrexia (OR 2.22), tachycardia (OR 2.02) or tachypnea (OR 1.67) prior to surgery were also at significant risk of developing postoperative complications. An independent complication predictor on an initial unadjusted multivariate model was pyrexia (unadjusted OR 1.82 95% CI 1.07 – 3.08). Figure 3 shows that the effect of preoperative pyrexia became insignificant after controlling for the duration of postoperative antibiotic treatment

(adjusted OR 1.34 95% CI 0.76 -2.35).

Antibiotic treatment duration

All patients received empirical antibiotics as soon as the decision to proceed with operative management was made. Continuation of antibiotic treatment depended on the intraoperative findings and the clinical course of the patient. Intravenous antibiotics were administered for a median 2 (0 -3) days for G1 cases, 3 (2-4) days for the G2 cases and 4 (3-6) for G3/G4 cases. Of the 74 complicated cases, 21 (28.4%) patients settled without

6 additional antibiotic therapy, 45 required additional antibiotic treatment (60.8%), and eight (10.8%) required surgical or radiological intervention and further antibiotics.

Reducing G3/G4 complications

Preoperative pyrexia was the most discriminatory factor among other preoperative systematic inflammaroty response variables in predicting those at risk of complications in this study. An increased duration of postoperative antibiotics minimised the risk of complications in those with advanced appendicitis. (Fig 3)

Independent predictors of a negative appendicectomy

The independent predictors of a negative appendicectomy in the present cohort were anorexia [adjusted odds ratio (aOR) 0.35 95% CI 0.19 – 0.39], rebound-tenderness (aOR 0.54 95% CI 0.39 -0.79) and WBC count

<7,900 per ʮL (aOR 0.27 95% 0.19 – 0.31). The discriminatory ability of this model was 0.721

Discussion

This study shows younger children present more commonly with advanced appendicitis. Our cohort of >1,000 patients had a complication rate of 7% and a negative appendicectomy rate of 21%. A clinical model of anorexia, rebound tenderness, a WBC count of ≥ 15,100/ ʮL and a temperature ≥37.5oC were significant predictors of advanced appendicitis, with a sensitivity of 38%. The absence of these factors makes a diagnosis of advanced appendicitis less likely.

The diagnosis of advanced appendicitis can be difficult in very young patients. The reasons for this observation include insufficient clinical history, aged-dependent communication difficulties and a rather large proportion of atypical and non-specific clinical presentations. The utilization of individual clinical symptoms may result in an early accurate detection of advanced appendicitis but at the expense of an increased rate of negative appendicectomy. However, when comparing a negative appendicectomy with an advanced appendicitis in this cohort, complication risks were 4.7 times higher in the advanced appendicitis group. The focus in management of appendicitis, therefore, should be in early and accurate detection of an advanced appendicitis

The clinical model of anorexia, rebound tenderness, leukocytosis and pyrexia accurately discriminates those with advanced appendicitis from those who did not have appendicitis. However, due to age-related difficulties in the clinical assessment, only 38.2% of those with advanced appendicitis were correctly identified. The age-

7 related adjustments improved the odds ratio of the history and examination finding variables, i.e., anorexia and rebound tenderness in detecting advanced appendicitis. (Table 3) The goal of defining this model was not to make a 100% clinical preoperative diagnosis of appendicitis but to assess the diagnostic value of simple bedside measurements in detecting early advanced appendicitis.

The current model employed bedside measurements only and leukocytosis was defined using a lower cut-off point of WBC count ≥ 15,100 per ʮL than the value of WBC count ≥ 19,400 per ʮL in a previous study. The variability in the cut-off points of leukocytosis set by the current model may relate to the timing of WBC assay in relation to the onset of symptoms. The significant implication is that the model may select cases of advanced appendicitis earlier in the clinical course than that using a cut-off point of 19,400 cells/ʮL. The clinical assessment model also permits a selective approach to the use of CT scan, thereby minimising risk of unnecessary ionising radiation while still achieving accuracy based on a predicted negative appendicectomy rate of 0.9%.

The American College of Radiologists currently recommends ultrasonography (US) as the preferred first imaging in children with a suspected acute appendicitis. The recommendation presumes that US is nearly as accurate as CT for the diagnosis of appendicitis in the paediatric population and without ionizing radiation exposure. When the US report is inconclusive, a focused CT is recommended by some prior to laparoscopy.

The alternative option is an MRI for these cases. [20]

Computerised Tomography (CT) is an excellent diagnostic tool for acute appendicitis. It has a specificity of more than 95% for diagnosing appendicitis and variable sensitivities for associated perforation. The major drawback to the use of CT is the need to balance the benefit against the risks and logistics of contrast administration, patient sedation and radiation risks in children. For example, a single abdominal CT study in a 5- year-old child would result in a lifetime risk of radiation-induced cancer of >20 per 100,000. In this regard, it might be safer to limit the use of CT to patients not already selected by this clinical model.

MRI is an alternative diagnostic modality with both 96% sensitivity and specificity for the diagnosis of appendicitis. However, it is expensive, sedation is often required, and it is not as readily available as CT.

Radiology is complimentary to clinical assessment, and may be helpful in children without anorexia, rebound tenderness, a WBC count of > 15,100/ ʮL and a temperature >37.5oC, as this cohort is unlikely to have severe appendicitis.

8 Laparoscopy offers a minimally invasive approach. It has both diagnostic and therapeutic advantages and is now the treatment of choice in experienced centres. The 2010 Cochrane review recommends using laparoscopy and laparoscopic appendicectomy (LA) in patients with suspected appendicitis unless laparoscopy itself is contraindicated or not feasible. Even though the analysis only included seven studies with children, the findings of this study are consistent with the review and show the potential benefit of laparoscopy in terms of length of stay and reducing the incidence of wound infections. Other potential benefits of LA include cosmesis, decreased postoperative pain with less narcotic use, lower risk of ileus, earlier return of gastrointestinal function and ambulation and a faster return to routine activity. However, the laparoscopic approach is not without disadvantages if it is conducted without prior imaging, including the option of leaving a macroscopically looking ‘normal’ appendix behind; the pathology report does not always correlate with the surgeon’s diagnosis.

More than 98.6% of children in our institution were operated exclusively by the open approach. In 2005, laparoscopy had not been firmly established as the therapeutic option of choice. The institution is not exclusively a paediatric hospital and did not have paediatric laparoscopic equipment available. The increased duration of surgery, higher costs and technical challenges of laparoscopy in children were other reasons for almost exclusive use of the open approach during that time.

The study addressed a specific paediatric population aged between one month and fourteen years. It included more than 1000 consecutive patients in the analysis. It was conducted in a single centre, and a Consultant

General Surgeon was present for each case. All clinical symptoms and signs were analysed based on documentation by the surgical team and were near complete for the parameters reported. In the case of missing documentation, supplementation was made using information from the triage notes or notes from the emergency department clinician. The overall postoperative complication rate was 7.2 %. This compares favourably to the 6

– 29% complication rate reported for paediatric open appendicectomies in a recent systematic review. A clinical model was proposed to help diagnose advanced appendicitis without resort to sophisticated diagnostics.

Although the model has a sensitivity of 38%, children without these findings are more likely not to have advanced appendicitis.

This study was retrospective in design. The captured data represents the findings of all those who proceeded to surgery based on a clinical suspicion of appendicitis after an evaluation of the clinical and imaging findings. A capture of the total index of utilization of imaging and preoperative imaging in accurately detecting and

9 effectively ruling out the diagnosis of acute appendicitis during this period was beyond the scope of this retrospective study. Most cases were performed by the open approach.

Conclusion

Advanced appendicitis is common in young children due to age-related difficulties in the clinical assessment.

We present a clinical model based on a combination of anorexia, rebound tenderness, leukocytosis and pyrexia that may be useful in the early identification of patients with advanced appendicitis but due to limitations in clinical assessment in young patients, imaging prior to intervention is advised for atypical cases. The data presented here also demonstrate a negative appendicectomy is very likely in the absence of anorexia, rebound tenderness and if the WBC count < 7,900 cells per ʮL.

Acknowledgement

The authors thank the Late Mr. Nicola Motterlini for statistical support at the early stages of the research work.

The authors also thank Ms. Caitriona Canning and Mr. Ian Robertson for editing the revised manuscript.

Conflict of Interest

All authors have no conflict of interest to disclose.

10 References

1. Andersen SB, Paerregaard A and Larsen K. (2009) Changes in the epidemiology of acute appendicitis and appendectomy in Danish children 1996-2004. Eur J Pediatr Surg 19:286-289. doi:10.1055/s-0029- 1224199

2. Birnbaum BA and Wilson SR. (2000) Appendicitis at the millennium. Radiology 215:337-348. doi:10.1148/radiology.215.2.r00ma24337

3. Sivit CJ, Siegel MJ, Applegate KE et al. (2001) When appendicitis is suspected in children. Radiographics 21:247-262; questionnaire 288-294. doi:10.1148/radiographics.21.1.g01ja17247

4. Pearl RH, Hale DA, Molloy M et al. (1995) Pediatric appendectomy. J Pediatr Surg 30:173-178; discussion 178-181.

5. Nance ML, Adamson WT and Hedrick HL. (2000) Appendicitis in the young child: a continuing diagnostic challenge. Pediatr Emerg Care 16:160-162.

6. Newman K, Ponsky T, Kittle K et al. (2003) Appendicitis 2000: variability in practice, outcomes, and resource utilization at thirty pediatric hospitals. J Pediatr Surg 38:372-379; discussion 372-379. doi:10.1053/jpsu.2003.50111

7. Markar SR, Blackburn S, Cobb R et al. (2012) Laparoscopic versus open appendectomy for complicated and uncomplicated appendicitis in children. J Gastrointest Surg 16:1993-2004. doi:10.1007/s11605-012-1962-y

8. Schwartz MZ, Tapper D and Solenberger RI. (1983) Management of perforated appendicitis in children. The controversy continues. Ann Surg 197:407-411.

9. Stilling NM, Fristrup C, Gabers T et al. (2013) Acceptable outcome after laparoscopic appendectomy in children. Dan Med J 60:A4564.

10. Harrison MW, Lindner DJ, Campbell JR et al. (1984) Acute appendicitis in children: factors affecting morbidity. Am J Surg 147:605-610.

11. Stewart B, Khanduri P, McCord C et al. (2014) Global disease burden of conditions requiring emergency surgery. Br J Surg 101:e9-22. doi:10.1002/bjs.9329

12. Morrow SE and Newman KD. (2007) Current management of appendicitis. Semin Pediatr Surg 16:34- 40. doi:10.1053/j.sempedsurg.2006.10.005

13. Cappendijk VC and Hazebroek FW. (2000) The impact of diagnostic delay on the course of acute appendicitis. Arch Dis Child 83:64-66.

14. Narsule CK, Kahle EJ, Kim DS et al. (2010) Effect of delay in presentation on rate of perforation in children with appendicitis. Am J Emerg Med. doi:S0735-6757(10)00184-1 [pii] 10.1016/j.ajem.2010.04.005

15. Williams N and Bello M. (1998) Perforation rate relates to delayed presentation in childhood acute appendicitis. J R Coll Surg Edinb 43:101-102.

16. Brender JD, Marcuse EK, Koepsell TD et al. (1985) Childhood appendicitis: factors associated with perforation. Pediatrics 76:301-306.

17. England RJ and Crabbe DC. (2006) Delayed diagnosis of appendicitis in children treated with antibiotics. Pediatr Surg Int 22:541-545. doi:10.1007/s00383-005-1625-0

11 18. Golladay ES and Sarrett JR. (1988) Delayed diagnosis in pediatric appendicitis. South Med J 81:38-42.

19. Andersson RE. (2007) The natural history and traditional management of appendicitis revisited: spontaneous resolution and predominance of prehospital perforations imply that a correct diagnosis is more important than an early diagnosis. World J Surg 31:86-92. doi:10.1007/s00268-006-0056-y

20. Humes DJ and Simpson J. (2006) Acute appendicitis. Bmj 333:530-534. doi:10.1136/bmj.38940.664363.AE

21. Rothwell PM. (2008) Prognostic models. Pract Neurol 8:242-253. doi:8/4/242 [pii] 10.1136/jnnp.2008.153619

22. Williams RF, Blakely ML, Fischer PE et al. (2009) Diagnosing ruptured appendicitis preoperatively in pediatric patients. J Am Coll Surg 208:819-825; discussion 826-818. doi:S1072-7515(09)00109-4 [pii] 10.1016/j.jamcollsurg.2009.01.029

23. Beltran MA, Almonacid J, Vicencio A et al. (2007) Predictive value of white blood cell count and C- reactive protein in children with appendicitis. J Pediatr Surg 42:1208-1214. doi:10.1016/j.jpedsurg.2007.02.010

24. Smith MP, Katz DS, Lalani T et al. (2014) ACR Appropriateness Criteria(R) Right Lower Quadrant Pain-Suspected Appendicitis. Ultrasound Q. doi:10.1097/RUQ.0000000000000118

25. Hennelly KE and Bachur R. (2011) Appendicitis update. Curr Opin Pediatr 23:281-285. doi:10.1097/MOP.0b013e32834604da

26. National Surgical Research Collaborative. (2013) Multicentre observational study of performance variation in provision and outcome of emergency appendicectomy. Br J Surg 100:1240-1252. doi:10.1002/bjs.9201

27. Aranda-Narvaez JM, Montiel-Casado MC, Gonzalez-Sanchez AJ et al. (2013) [Radiological support for diagnosis of acute appendicitis: use, effectiveness and clinical repercussions]. Cir Esp 91:574-578. doi:10.1016/j.ciresp.2013.01.009

28. Bixby SD, Lucey BC, Soto JA et al. (2006) Perforated versus nonperforated acute appendicitis: accuracy of multidetector CT detection. Radiology 241:780-786. doi:10.1148/radiol.2413051896

29. Tsuboi M, Takase K, Kaneda I et al. (2008) Perforated and nonperforated appendicitis: defect in enhancing appendiceal wall--depiction with multi-detector row CT. Radiology 246:142-147. doi:10.1148/radiol.2461051760

30. Alaedeen DI, Cook M and Chwals WJ. (2008) Appendiceal fecalith is associated with early perforation in pediatric patients. J Pediatr Surg 43:889-892. doi:10.1016/j.jpedsurg.2007.12.034

31. Lander A. (2007) The role of imaging in children with suspected appendicitis: the UK perspective. Pediatr Radiol 37:5-9. doi:10.1007/s00247-006-0304-1

32. Tompane T, Leong CW, Bush R et al. (2013) Appropriateness of Radiology Procedures Performed in Children With Gastrointestinal Symptoms and Conditions. Clin Gastroenterol Hepatol. doi:10.1016/j.cgh.2013.08.044

33. Doria AS, Moineddin R, Kellenberger CJ et al. (2006) US or CT for Diagnosis of Appendicitis in Children and Adults? A Meta-Analysis. Radiology 241:83-94. doi:10.1148/radiol.2411050913

34. Adibe OO, Amin SR, Hansen EN et al. (2011) An evidence-based clinical protocol for diagnosis of acute appendicitis decreased the use of computed tomography in children. J Pediatr Surg 46:192-196. doi:10.1016/j.jpedsurg.2010.09.087

35. Brenner D, Elliston C, Hall E et al. (2001) Estimated risks of radiation-induced fatal cancer from pediatric CT. AJR Am J Roentgenol 176:289-296. doi:10.2214/ajr.176.2.1760289 12 36. Wan MJ, Krahn M, Ungar WJ et al. (2009) Acute appendicitis in young children: cost-effectiveness of US versus CT in diagnosis--a Markov decision analytic model. Radiology 250:378-386. doi:10.1148/radiol.2502080100

37. Koning JL, Naheedy JH and Kruk PG. (2014) Diagnostic performance of contrast-enhanced MR for acute appendicitis and alternative causes of abdominal pain in children. Pediatr Radiol. doi:10.1007/s00247-014-2952-x

38. Sauerland S, Jaschinski T and Neugebauer EA. (2010) Laparoscopic versus open surgery for suspected appendicitis. Cochrane Database Syst Rev:CD001546. doi:10.1002/14651858.CD001546.pub3

39. Esposito C, Calvo AI, Castagnetti M et al. (2012) Open versus laparoscopic appendectomy in the pediatric population: a literature review and analysis of complications. J Laparoendosc Adv Surg Tech A 22:834-839. doi:10.1089/lap.2011.0492

40. Katkhouda N, Mason RJ, Towfigh S et al. (2005) Laparoscopic versus open appendectomy: a prospective randomized double-blind study. Ann Surg 242:439-448; discussion 448-450.

13 Graphics

Tables

Table 1: Patient demographics

Table 2: Postoperative complications (N = 1,037)

Table 3: Age-adjusted multivariate model of paediatric advanced appendicitis

Figure Captions

Fig 1 Age versus advanced appendicitis

Fig 2 Predictors of advanced appendicitis

Fig 3 Effect of postoperative antibiotics on model

14 Table 1: Patient demographics

Pathology Grade

G0 G1 G2 G3/G4

N (%) 220 (21.2) 650 (62.7) 44 (4.2) 123 (11.9)

M:F ratio 87:133 361:289 29:15 67:56*

Age (yr, mean ± SD) 10.2 ± 2.9 10.2 ± 2.8 9.4 ± 3.1 8.4 ± 3.7†

*p = 0.02; †p < 0.001.

15 Table 2: Postoperative complications (N = 1,037)

Complications Number % Superficial wound infection 28 2.7 Wound dehiscence 6 0.6 Intra-abdominal abscess 10 1.0 Wound granuloma 1 0.1 Prolonged ileus 11 1.1 Small bowel obstruction 3 0.3 Lower gastrointestinal bleed 4 0.4 Enterocutaneous fistula 2 0.2 Pneumonia 10 1.0 Atalectasis 3 0.3 Urinary Tract Infection 6 0.6 Acute Kidney Injury 1 0.1 Allergic reaction 1 0.1

16 Table 3: Age-adjusted multivariate model of paediatric advanced appendicitis

Clinical variable N Unadjusted Analysis Adjusted Analysis OR (95% CI) p OR (95% CI) p Anorexia 731 4.24 (2.13 – 8.14) < 0.001 4.55 (2.38 – 9.11) <0.001 Rebound Tenderness 595 1.94 (1.23 -3.04) 0.004 2.22 (1.39 – 3.53) 0.001 WBC count ≥ 15,100 /ʮL 418 3.90 (2.44 – 6.26) <0.001 3.60 (2.24 –5.79) <0.001 Temperature ≥ 37.5o C 361 4.22 (2.67 – 6.65) <0.001 3.62 (2.27 – 5.77) <0.001