- Email: [email protected]
Benchmarking the value of ultrasound for acute appendicitis in children
Journal of Pediatric Surgery xxx (2016) xxx–xxx
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Benchmarking the value of ultrasound for acute appendicitis in children☆ Thomas P Cundy a,b,⁎, Roger Gent c, Claire Frauenfelder a, Laura Lukic c, Rebecca J Linke c, Day Way Goh a,d a
Department of Paediatric Surgery, Women's and Children's Hospital, South Australia Discipline of Surgery, University of Adelaide, South Australia c Department of Radiology, Women's and Children's Hospital, South Australia d Discipline of Paediatrics, School of Medicine, University of Adelaide, South Australia b
a r t i c l e
i n f o
Article history: Received 7 August 2016 Accepted 12 September 2016 Available online xxxx Key words: Pediatric Ultrasound Appendicitis
a b s t r a c t Background: This study appraises the diagnostic quality of ultrasound for acute appendicitis in children and consequently challenges the perception of inferior accuracy and suitability compared to computed tomography (CT). Methods: Radiologist reports for consecutive “query appendicitis” ultrasound studies were retrieved from a hospital database for the study period 2009–2014. Children who subsequently underwent appendicectomy were identified. Corresponding operative and histopathology findings were evaluated. Diagnostic accuracy of ultrasound was determined by analyzing overall accuracy, sensitivity, specificity, predictivity, and likelihood ratios. Results: A total of 3799 ultrasound examinations were evaluated. Mean age was 11.5 ± 3.8 years. The proportion of patients investigated with preoperative ultrasound was 59.9% (1103/1840). Appendix visualization rate was 91.7%. Overall diagnostic accuracy was 95.5%. Sensitivity and specificity values were 97.1% (95.9–98.1; 95% CI) and 94.8% (93.9–95.6; 95% CI), respectively. Separate analysis of only ultrasound positive and negative examinations (i.e., excluding nondiagnostic examinations) confirmed sensitivity and specificity values of 98.8% and 98.3%. Conclusion: In this largest reported single institution series of ultrasound examinations for appendicitis, we report benchmark standard quality of diagnostic accuracy and visualization rates. Given the radiation and cost implications of CT, there is a strong argument to recommend ultrasound as the primary imaging modality. Diagnostic Study–Level II. © 2016 Elsevier Inc. All rights reserved.
Appendicitis is the most common surgical emergency in children [1]. Diagnosis of this pathology can be challenging owing to variability in clinical presentation and numerous other surgical and nonsurgical diagnoses that may share similar clinical features, especially in atypical cases that represent approximately one-third of presentations [2]. Imaging is playing an increasing diagnostic role in acute appendicitis [2–5]. As an adjunct to clinical assessment, it has contributed to significant improvements in rates of negative appendicectomy [5–7]. Ultrasound and computed tomography (CT) are the two most utilized advanced imaging modalities for investigation of acute appendicitis [5]. Distinguishing superiority between these two diagnostic tools is a topic of contention due to respective advantages and disadvantages that are often mutually exclusive [8–10]. In general, CT is favored in North American centers, while ultrasound is more frequently used in most other countries [5,6,9–12]. The literature reports marginally higher diagnostic accuracy for CT; however this is at the expense of higher financial cost and exposure to ionizing radiation [2,4,10]. Estimated lifetime risk of radiation-induced malignancy
☆ Conflicts of interest: none. ⁎ Corresponding author at: Department of Paediatric Surgery, Women's and Children's Hospital, 72 King William Road, Adelaide, South Australia, 5006. Tel.: +61 8161 7328. E-mail address: [email protected] (T.P. Cundy).
associated with an abdominal CT scan in childhood ranges from 13 to 26 per 100,000 [4,6,13]. The aim of this study is to evaluate the diagnostic accuracy for a large series of consecutive unselected abdominal ultrasound examinations performed for clinical suspicion of acute appendicitis in children. In doing so, we seek to establish a benchmark standard for ultrasound that could challenge the perception of inferior accuracy compared to CT and therefore may empirically justify ultrasound as the preferred primary imaging investigation for suspected appendicitis in children. 1. Materials and methods 1.1. Study setting The study was undertaken at a single university teaching hospital following ethics review board approval. Our tertiary pediatric surgery service provides care for a local population of approximately 1.6 million persons. At our hospital, ultrasound is used almost exclusively as the primary imaging modality for acute appendicitis. A dedicated pediatric ultrasonography service is available with provision for scans to be conducted at all hours of the day as clinical need arises. Our sonography service is staffed by an experienced group of specialist pediatric sonographers. Investigations for “query appendicitis” are identified with a separate coding number.
http://dx.doi.org/10.1016/j.jpedsurg.2016.09.009 0022-3468/© 2016 Elsevier Inc. All rights reserved.
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
2
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
The sonographic technique at our institution is a standard graded compression examination. The ultrasound used for all examinations was an iu22 Ultrasound System with a C8-5 tightly curved array probe transducer (Philips Healthcare, Amsterdam, Netherlands). In special circumstances of obese or older children with large body habitus, a C9-4 curved array transducer was used. There are no fixed sonographic diagnostic criteria for acute appendicitis. Instead, the radiological diagnosis was made using a combination of primary and secondary factors in a case-dependent manner that included; 1) appendix diameter N 6 mm, 2) degree of localized hypervascularity, 3) presence of surrounding free fluid, 4) compressibility of the appendix, 5) presence of a luminal fecolith, 6) focal ultrasound probe tenderness and 7) appearances of surrounding soft tissue.
only by histopathology confirmation such that ultrasound positive cases treated nonoperatively (with or without antibiotic therapy) were coded as negative appendicectomies for the purpose of diagnostic accuracy calculations. Additionally, the denominator value for all analyses was assigned as the total number of ultrasound examinations for each subgroup (positive, negative or equivocal) and not adjusted to exclude repeat ultrasound examinations per patient. Kruskal–Wallis one-way analysis of variance was used to compare negative appendicectomy rates between ultrasound positive, negative and nondiagnostic groups. If significant differences were determined, then the Chi-square test was used to further compare outcomes between 2 groups. The Chi-square test was also used to compare the rates of appendicitis between equivocal and nonvisualization ultrasound examinations. Statistical significance was regarded as P b 0.05. Statistical analysis was performed using SPSS version 21.0 (IBM Corp, NY).
1.3. Data collection
2. Results
The study period comprised six complete calendar years from 1st January 2009 to 31st December 2014. Radiologist reports for consecutive ultrasound studies requested for “query appendicitis” were retrieved from a central hospital database. Outcomes of ultrasound investigations were categorized as positive, negative, or nondiagnostic. Nondiagnostic ultrasounds were further subcategorized as either equivocal or nonvisualization examinations. If the vermiform appendix was unable to be visualized in its entirety, then the ultrasound examination was categorized as not visualized. If the vermiform appendix was visualized but there was diagnostic uncertainty, then the ultrasound examination was categorized as equivocal. If the same patient underwent repeat ultrasound examinations, then each examination was recorded as an independent event in the database. Electronic medical records were evaluated to identify children who proceeded to appendicectomy following ultrasound investigation. To corroborate ultrasound findings, both the operative findings and histopathology findings were evaluated. Histopathology findings were categorized as positive or negative relative to a final pathological diagnosis of acute appendicitis. Other diagnoses were also documented when confirmed in the presence or absence of acute appendicitis at the time of planned appendicectomy.
A total of 3799 ultrasound examinations were evaluated. The mean (± SD) patient age was 11.5 (± 3.8) years. Gender distribution was 1:1.2 for males:females. Approximately one-quarter (29.0%, 1103/3799) of those investigated with ultrasound for query appendicitis proceeded to an appendicectomy. A total of 1840 appendicectomies for suspected acute appendicitis were performed during the study period. The proportion of patients who were investigated with ultrasound prior to appendicectomy was 59.9% (1103/1840). The prevalence of appendicitis in the study cohort was 27.6% (1049/3799). There were 14 patients with positive ultrasound examinations that were managed nonoperatively, and 5 patients who later underwent interval appendicectomy. Sonographic, clinical and pathological findings are summarized in Fig. 1. The vermiform appendix was confidently seen in its entirety in 91.7% (3484/3799) of ultrasound examinations. Ultrasound diagnosis was positive, negative and nondiagnostic in 27.3%, 61.1%, and 11.6% of examinations respectively. Alternative sonographic findings of symptomatic or incidental pathologies were demonstrated in 13.0% of ultrasound examinations and are summarized in Table 1. The proportion of patients with intraoperative findings of perforated appendicitis was 15.1%, 8.3% and 14.3% for positive, negative and nondiagnostic ultrasound groups respectively. Overall diagnostic accuracy was 95.5% (3627/3799). Sensitivity and specificity values for ultrasound were 97.1% (95.9–98.1; 95% CI) and 94.8% (93.9–95.6; 95% CI) respectively. Positive and negative predictive values for ultrasound were 87.8% (85.8–89.6%; 95 CI) and 98.9% (98.4–99.2; 95% CI) respectively. Positive and negative likelihood ratios for ultrasound were 18.8 (16.0–22.1; 95% CI) and 0.03 (0.02–0.04; 95% CI) respectively. Separate analysis of only ultrasound positive and negative examinations (i.e. excluding nondiagnostic examinations) yielded sensitivity and specificity values of 98.8% (97.9–99.4%; 95% CI) and 98.3% (97.7–98.8%; 95% CI). Among patients who had a nondiagnostic ultrasound, 13.9% underwent an appendicectomy, and acute appendicitis was confirmed in 9.1% (Fig. 1). The incidence of appendicitis in children with equivocal ultrasound examinations was significantly higher than that in children with nonvisualization ultrasound examinations (16.9% versus 5.1%; P b 0.001). The overall negative appendicectomy rate was 4.9% (54/1103). There was a significant difference in negative appendicectomy rates between the ultrasound positive, negative and nondiagnostic groups (1.8%, 50.0%, and 38.1% respectively; P b 0.001, Fig. 1). There was no difference in negative appendicectomy rates between the ultrasound negative and nondiagnostic groups (P = 0.325). Repeat ultrasound examinations of the same patient were performed in 3.9% (149/3799) of cases. In the majority of circumstances (94.0%, 140/149), repeat imaging was requested based on ongoing or recurring symptoms suspicious of acute appendicitis in the context of a previous ultrasound examination that was either negative (54.3%,
1.2. Standard ultrasound technique
1.4. Statistical analysis Diagnostic accuracy of ultrasound was determined by analyzing overall accuracy, sensitivity, specificity, predictivity and likelihood ratios. These test characteristics were calculated using conventional formulae with 95% confidence intervals (95% CI). Methodological handling of equivocal or nonvisualization ultrasound examination data is inconsistent in the literature and can be misrepresentative [14]. Processing of these data was therefore modeled on consensus methodology among the most transparent and high quality related literature available [14–16]. For the main analysis of test characteristics, ultrasound findings were dichotomized as either positive or negative. Post-hoc processing was performed to recategorize equivocal examinations as positive, and nonvisualization examinations as negative. This recategorization was based on assumptions of 1) equivocal examinations demonstrating abnormal findings that were genuinely suspicious but not convincingly diagnostic for acute appendicitis, and 2) nondiseased vermiform appendixes being typically less apparent on ultrasound and more likely to be nonvisualized. To highlight the impact that is introduced by ignoring equivocal or nonvisualization examinations in calculations of diagnostic accuracy, separate analysis was performed that only included ultrasound examinations that were truly test positive or test negative (i.e. excluding equivocal and nonvisualization examinations). In addition to including nondiagnostic ultrasound examinations in the main analysis, data were further interpreted based on intentionto-treat principles by categorizing acute appendicitis (true positive)
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
3
Fig. 1. Decision tree outlining ultrasound, clinical and pathological outcomes for the study sample (n = 3799 ultrasound examinations).
Table 1 Alternative or additional sonographic findings in children who underwent ultrasound examinations for query appendicitis (n = 3799). Colitis 108 Mesenteric adenitis 98 Ovarian cyst 59 Terminal ileitis 35 Lymphoid hyperplasia of terminal ileum 21 Pyelonephritis 17 Inflammatory phlegmon or abdominal collection 14 Omental infarction 14 Fecal loading 13 Ovarian torsion 10 Small bowel obstruction 8 Ovarian mass or nodule 7 Pneumonia 7 Cystitis 6 Hip synovitis or joint effusion 5 Typhlitis 5 Intussusception 5 Renal cyst 4 Cholelithiasis 4 Pelvicalyceal dilatation 3 Free intraabdominal fluid with appendix unable to be visualized 3 Intestinal malrotation 3 Ureteric stone 3 2a Cholecystitis, abdominal solid or cystic tumor, ascites and splenomegaly, abdominal wall cellulitis, pleural effusion, urachal remnant, osteitis pubis, pancreatitis, renal stone, uterine didelphys and hematometrocolpos 1a Pelvic inflammatory disease, enthesitis or myositis of the lower limb, intraabdominal testis, nephritis, tubo-ovarian abscess, liver cyst or other lesion, intestinal parasites, salpingitis, accessory spleen torsion, hepatomegaly, hepatomegaly, rectus sheath hematoma, hematosalpinx, hepatic hemangioma, intrahepatic biliary tree dilatation, Meckel's diverticulum, vesicoureteric junction obstruction, epididymo-orchitis, renal abscess, duodenitis, pyosalpinx, pelviureteric junction obstruction, Gartner’s duct remnant or epoophoron a
Number of occurrences for each listed sonographic finding.
76/140) or equivocal (45.7%, 64/140). The remaining indications for repeat ultrasound examination were for interval assessment of acute appendicitis that was being managed nonoperatively with antibiotic therapy (4.7%, 7/149), or for further evaluation of a positive ultrasound that poorly corresponded with clinical findings (1.3%, 2/149). There was a low positive diagnostic yield for repeat ultrasounds with only 8 new positive ultrasound diagnoses arising from previously negative or equivocal ultrasound examinations (5.7%, 8/140). Repeat ultrasound was only able to clarify previously nondiagnostic ultrasounds as positive or negative in approximately half of cases (51.6%, 33/64). There were only 27 patients (0.7%, 27/3799) who proceeded to CT scan examinations following initial ultrasound examination. The most frequent reason for subsequent CT scan was for further investigation of abnormal ultrasound findings after demonstration of a sonographically normal appendix (40.7%, 11/27). Other reasons for CT examinations were for large body habitus (18.5%, 5/27) or inability to tolerate ultrasound probe pressure (7.4%, 2/27) that resulted in a nondiagnostic ultrasound examination. A CT diagnosis of appendicitis occurred on only 3 occasions. 3. Discussion Advanced imaging has transformed diagnostic processes for acute appendicitis. Past reliance on clinical assessment alone has evolved toward current practice where some argue that imaging should be undertaken in all children before appendicectomy [12]. In this series, preoperative ultrasound was used in 59.9% of appendicectomies. In a large administrative database study involving 30 pediatric hospitals in the United States, the median percentage of patients who had preoperative imaging was 69% [17]. Other smaller series report preoperative ultrasound usage that ranges widely from 18% to 99% [5,15,17–21]. Reliance on imaging has increased owing to greater access to radiological services, improvements in imaging accuracy, and clinicians striving to achieve lower negative appendicectomy rates [5,12].
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
4
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
Perceptions of ultrasound and CT imaging modalities, and weighting of their corresponding pros and cons are well described [2]. Diagnostic accuracy is the most important discriminator when evaluating any diagnostic test or investigation. A recent meta-analysis by Doria et al. evaluated the comparative diagnostic accuracy of ultrasound versus CT for appendicitis in children [4]. Pooled analysis of results from 9356 patients determined that CT was significantly more sensitive compared to ultrasound (P = 0.001) [4]. Sensitivity and specificity values for ultrasound were 88% and 94% respectively, and for CT were 94% and 95% respectively [4]. While CT remains the preferred imaging choice for pediatric acute appendicitis in North America compared to ultrasound (median 34% for CT versus 6% for ultrasound), increased use of ultrasound since 2007 is encouraging and should be further supported based on the findings of this study [5,9]. Our entirely unselected study population is representative of the wide spectrum of patients who undergo ultrasound examination for clinically suspected appendicitis. As such, we feel that our study cohort fairly reflects actual daily clinical practice. According to Bayes theorem, observed diagnostic performance of a test such as ultrasound is dependent and proportional to the pretest probability of the disease in the patient population [12,14]. Lower disease prevalence in an unselected patient population such as this study will infer lower expected test accuracy compared to a more selected population such as those who are stratified to imaging according to various scoring algorithms or those who are initially clinically assessed by experienced surgical staff [12,14]. Various clinical assessment scoring algorithms such as the Alvarado and Pediatric Appendicitis Scores have been developed to risk stratify patients with suspected appendicitis and assign more selective use of imaging investigations [6,22–24]. Limited evidence for these score-based diagnostic aids indicates modest accuracy and reproducibility [6,25]. For these reasons, scoring algorithms are not yet recommended as isolated diagnostic tools in routine clinical practice and were not used at our institution during the study period [6]. 3.1. Benchmarking Numerous series in the literature describe diagnostic accuracies of ultrasound for acute appendicitis [4,6,9]. Existing gold-standard accuracies are represented by those reported by Goldin et al. and Baldisserotto et al. who published their series of 304 and 425 patients respectively [26,27]. Sensitivity and specificity values reported by Goldin et al. were 98.7% and 95.4%, however these were derived from post-hoc determined diagnostic criteria of maximal outside appendix diameter ≥ 7 mm or appendix wall thickness N 1.7 mm, and with an appendix visualization rate of 68% [27]. Baldisserotto et al. reported sensitivity and specificity values of 98.5% and 98.2% respectively, with positive and negative predictive values of 98.0% and 98.7% [6,26]. Nondiagnostic ultrasound examinations are an accepted reality in clinical practice and should therefore be appropriately incorporated into diagnostic accuracy calculations. Goldin et al. selectively excluded nondiagnostic ultrasounds for accuracy calculations and Baldisserotto et al. did not disclose their method of handling these data [26,27]. Considering the relative conservativeness of our highly competitive accuracy values with inclusion of nondiagnostic examination data, our unrivaled accuracy values with exclusion of nondiagnostic examination data, and excellent visualization rates in a large unselected consecutive series, we propose that our results should be considered among benchmark standards.
proportional relationship [14,15]. With approximately 60% of patients undergoing preoperative imaging, approximately 2 scans are performed each day. This case volume facilitates consistent exposure enabling skill acquisition, refinement and maintenance among our staff. Visualization of the appendix is obligatory for a diagnostic ultrasound examination. Our visualization rate of 91.7% is among the highest reported and is considerably higher than many other contemporary series that report visualization rates b 50% [14,15,21]. We attribute much of this high visualization rate to the virtues of patience and dedication among our staff. A detailed examination requires patience and available time to employ different visualization strategies such as patient repositioning, distraction techniques, emptying and filling the urinary bladder, and breath-holding maneuvers. Dedication and quarantined time are required for a more intensive personalized approach to each individual examination. In contrast to rigid criterion based diagnostic approaches, an individualized approach to each examination puts into context the constellation of sonographic features with available clinical information. For the diverse presentations of acute appendicitis, we feel that the flexibility of this approach is advantageous and improves diagnostic performance. An accurate investigation may require longer time, and therefore ultrasonography staff resources; however it facilitates targeted patient care and may be more cost effective overall for the patient's admission. A linear array transducer is routinely used in most centers internationally (Fig. 2). In our experience, a tightly curved array transducer is more suitable as it allows more targeted graded compression and relative size appropriateness for younger children (Fig. 2). For these reasons, we would advocate that the tightly curved array transducer should be considered as standard. A commonly cited drawback of ultrasound compared to CT is lower capability to diagnose alternate pathologies [2,8,9]. As highlighted in Table 1, we have found ultrasound to be highly versatile in diagnosing a diverse range of pathologies aside from acute appendicitis. This is facilitated by the dynamic real-time nature of ultrasound in that visualization of a normal appendix encourages more detailed assessment of other visceral structures that might yield pathological features.
3.2. Ultrasound technique Diagnostic quality of ultrasound is operator dependent. Experience is both necessary and invaluable in developing operator expertise. As experience is gained through case exposure, quality and volume of ultrasound examinations share a strong directly
Fig. 2. C8-5 tightly curved (left) and L17-5 linear (right) array transducers (Philips Healthcare, Amsterdam, Netherlands).
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
3.3. Limitations Cost-effectiveness of ultrasound in suspected appendicitis is a contentious topic and was not specifically evaluated in this study. Cost-effectiveness and clinical effectiveness are deemed to be most favorable in the child with equivocal diagnosis on clinical assessment. In these circumstances, the opportunity to more confidently exclude appendicitis with use of ultrasound can contribute to considerable cost saving through either avoidance of hospital admission for a period of observation, or negative appendicectomy [1,7]. Previous studies have sought to further define the cohort of patients with truly equivocal clinical findings that are best suited for radiological investigation [3]. A limitation of this study is that the certainty of clinical diagnosis was not assessed, although the subjectivity of this parameter limits its reliability. An acceptable negative appendicectomy rate has traditionally been estimated as 10%–25% [8,12,25,28]. In contemporary series, rates of negative appendicectomy are reported to be as high as 17%–56% [3,23,29]. Although our 4.9% negative appendicectomy rate in this series was low, we are unable to empirically attribute this to the benefit of ultrasound, as this study does not include a comparator cohort of children who underwent appendicectomy during the same study period without ultrasound examination. Higher negative appendicectomy rates were observed in the subset of patients with negative or nondiagnostic ultrasound examinations. Based on our findings, the clinical corollary for nondiagnostic ultrasounds is that they should be interpreted as low risk for acute appendicitis, particularly nonvisualization examinations. Lastly, the application of our results may be limited to the circumstances of clinical practice at our institution. We recognize that there are considerable variations in clinical, technical, financial, resource allocation and medicolegal practices between healthcare institutions. Individually and collectively, each of these factors can impact on the delivery of clinical service and may account for differences in advanced imaging use and performance for acute appendicitis in children. We contend that our benchmark standard quality for ultrasound is tangibly achievable in most clinical environments with the appropriate case volume, staff skill, resource allocation, and attitude of involved clinicians.
4. Conclusions In this largest reported single institution series of ultrasound examinations for acute appendicitis in children, we report benchmark standard quality of diagnostic accuracy and visualization rates. Our findings also highlight the versatility of ultrasound and comparative accuracy compared to CT. Given the radiation and cost implications of CT, the role of ultrasound should be promoted as the primary imaging modality for acute appendicitis in children.
5
References [1] Axelrod DA, Sonnad SS, Hirschl RB. An economic evaluation of sonographic examination of children with suspected appendicitis. J Pediatr Surg 2000;35:1236–41. [2] Doria AS. Optimizing the role of imaging in appendicitis. Pediatr Radiol 2009; 39(Suppl. 2):S144–8. [3] Bullapur HM, Deshpande AV, Phin SJ, et al. Adjunct ultrasonography in children with suspected acute appendicitis: identifying the optimal target group. ANZ J Surg 2014; 84:326–30. [4] Doria AS, Moineddin R, Kellenberger CJ, et al. US or CT for diagnosis of appendicitis in children and adults? A meta-analysis. Radiology 2006;241:83–94. [5] Bachur RG, Hennelly K, Callahan MJ, et al. Advanced radiologic imaging for pediatric appendicitis, 2005–2009: trends and outcomes. J Pediatr 2012;160:1034–8. [6] Dingemann J, Ure B. Imaging and the use of scores for the diagnosis of appendicitis in children. Eur J Pediatr Surg 2012;22:195–200. [7] Rice HE, Arbesman M, Martin DJ, et al. Does early ultrasonography affect management of pediatric appendicitis? A prospective analysis. J Pediatr Surg 1999;34:754–8. [8] Hernanz-Schulman M. CT and US in the diagnosis of appendicitis: an argument for CT. Radiology 2010;255:3–7. [9] Strouse PJ. Pediatric appendicitis: an argument for US. Radiology 2010;255:8–13. [10] Frush DP, Frush KS, Oldham KT. Imaging of acute appendicitis in children: EU versus U.S. ... or US versus CT? A North American perspective. Pediatr Radiol 2009;39:500–5. [11] Muehlstedt SG, Pham TQ, Schmeling DJ. The management of pediatric appendicitis: a survey of North American pediatric surgeons. J Pediatr Surg 2004;39:875–9. [12] Holscher HC, Heij HA. Imaging of acute appendicitis in children: EU versus U.S. ... or US versus CT? A European perspective. Pediatr Radiol 2009;39:497–9. [13] Wan MJ, Krahn M, Ungar WJ, et al. Acute appendicitis in young children: costeffectiveness of US versus CT in diagnosis—a Markov decision analytic model. Radiology 2009;250:378–86. [14] Trout AT, Sanchez R, Ladino-Torres MF, et al. Critical evaluation of US for the diagnosis of pediatric acute appendicitis in a real-life setting: how can we improve the diagnostic value of sonography? Pediatr Radiol 2012;42:813–23. [15] Mittal MK, Dayan PS, Macias CG, et al. Performance of ultrasound in the diagnosis of appendicitis in children in a multicenter cohort. Acad Emerg Med 2013;20: 697–702. [16] Saucier A, Huang EY, Emeremni CA, et al. Prospective evaluation of a clinical pathway for suspected appendicitis. Pediatrics 2014;133:e88–95. [17] Newman K, Ponsky T, Kittle K, et al. Appendicitis 2000: variability in practice, outcomes, and resource utilization at thirty pediatric hospitals. J Pediatr Surg 2003;38:372–9. [18] Rice-Townsend S, Barnes JN, Hall M, et al. Variation in practice and resource utilization associated with the diagnosis and management of appendicitis at freestanding children's hospitals: implications for value-based comparative analysis. Ann Surg 2014;259:1228–34. [19] Neufeld D, Vainrib M, Buklan G, et al. Management of acute appendicitis: an imaging strategy in children. Pediatr Surg Int 2010;26:167–71. [20] Kotagal M, Richards MK, Flum DR, et al. Use and accuracy of diagnostic imaging in the evaluation of pediatric appendicitis. J Pediatr Surg 2015;50:642–6. [21] Scrimgeour DS, Driver CP, Stoner RS, et al. When does ultrasonography influence management in suspected appendicitis? ANZ J Surg 2014;84:331–4. [22] Alvarado AA. Practical score for the early diagnosis of acute appendicitis. Ann Emerg Med 1986;15:557–64. [23] Lintula H, Kokki H, Kettunen R, et al. Appendicitis score for children with suspected appendicitis. A randomized clinical trial. Langenbecks Arch Surg 2009;394:999–1004. [24] Samuel M. Pediatric appendicitis score. J Pediatr Surg 2002;37:877–81. [25] Morrow SE, Newman KD. Current management of appendicitis. Semin Pediatr Surg 2007;16:34–40. [26] Baldisserotto M, Marchiori E. Accuracy of noncompressive sonography of children with appendicitis according to the potential positions of the appendix. AJR Am J Roentgenol 2000;175:1387–92. [27] Goldin AB, Khanna P, Thapa M, et al. Revised ultrasound criteria for appendicitis in children improve diagnostic accuracy. Pediatr Radiol 2011;41:993–9. [28] Kaneko K, Tsuda M. Ultrasound-based decision making in the treatment of acute appendicitis in children. J Pediatr Surg 2004;39:1316–20. [29] Sivit CJ. Imaging children with acute right lower quadrant pain. Pediatr Clin North Am 1997;44:575–89.
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
Contents lists available at ScienceDirect
Journal of Pediatric Surgery journal homepage: www.elsevier.com/locate/jpedsurg
Benchmarking the value of ultrasound for acute appendicitis in children☆ Thomas P Cundy a,b,⁎, Roger Gent c, Claire Frauenfelder a, Laura Lukic c, Rebecca J Linke c, Day Way Goh a,d a
Department of Paediatric Surgery, Women's and Children's Hospital, South Australia Discipline of Surgery, University of Adelaide, South Australia c Department of Radiology, Women's and Children's Hospital, South Australia d Discipline of Paediatrics, School of Medicine, University of Adelaide, South Australia b
a r t i c l e
i n f o
Article history: Received 7 August 2016 Accepted 12 September 2016 Available online xxxx Key words: Pediatric Ultrasound Appendicitis
a b s t r a c t Background: This study appraises the diagnostic quality of ultrasound for acute appendicitis in children and consequently challenges the perception of inferior accuracy and suitability compared to computed tomography (CT). Methods: Radiologist reports for consecutive “query appendicitis” ultrasound studies were retrieved from a hospital database for the study period 2009–2014. Children who subsequently underwent appendicectomy were identified. Corresponding operative and histopathology findings were evaluated. Diagnostic accuracy of ultrasound was determined by analyzing overall accuracy, sensitivity, specificity, predictivity, and likelihood ratios. Results: A total of 3799 ultrasound examinations were evaluated. Mean age was 11.5 ± 3.8 years. The proportion of patients investigated with preoperative ultrasound was 59.9% (1103/1840). Appendix visualization rate was 91.7%. Overall diagnostic accuracy was 95.5%. Sensitivity and specificity values were 97.1% (95.9–98.1; 95% CI) and 94.8% (93.9–95.6; 95% CI), respectively. Separate analysis of only ultrasound positive and negative examinations (i.e., excluding nondiagnostic examinations) confirmed sensitivity and specificity values of 98.8% and 98.3%. Conclusion: In this largest reported single institution series of ultrasound examinations for appendicitis, we report benchmark standard quality of diagnostic accuracy and visualization rates. Given the radiation and cost implications of CT, there is a strong argument to recommend ultrasound as the primary imaging modality. Diagnostic Study–Level II. © 2016 Elsevier Inc. All rights reserved.
Appendicitis is the most common surgical emergency in children [1]. Diagnosis of this pathology can be challenging owing to variability in clinical presentation and numerous other surgical and nonsurgical diagnoses that may share similar clinical features, especially in atypical cases that represent approximately one-third of presentations [2]. Imaging is playing an increasing diagnostic role in acute appendicitis [2–5]. As an adjunct to clinical assessment, it has contributed to significant improvements in rates of negative appendicectomy [5–7]. Ultrasound and computed tomography (CT) are the two most utilized advanced imaging modalities for investigation of acute appendicitis [5]. Distinguishing superiority between these two diagnostic tools is a topic of contention due to respective advantages and disadvantages that are often mutually exclusive [8–10]. In general, CT is favored in North American centers, while ultrasound is more frequently used in most other countries [5,6,9–12]. The literature reports marginally higher diagnostic accuracy for CT; however this is at the expense of higher financial cost and exposure to ionizing radiation [2,4,10]. Estimated lifetime risk of radiation-induced malignancy
☆ Conflicts of interest: none. ⁎ Corresponding author at: Department of Paediatric Surgery, Women's and Children's Hospital, 72 King William Road, Adelaide, South Australia, 5006. Tel.: +61 8161 7328. E-mail address: [email protected] (T.P. Cundy).
associated with an abdominal CT scan in childhood ranges from 13 to 26 per 100,000 [4,6,13]. The aim of this study is to evaluate the diagnostic accuracy for a large series of consecutive unselected abdominal ultrasound examinations performed for clinical suspicion of acute appendicitis in children. In doing so, we seek to establish a benchmark standard for ultrasound that could challenge the perception of inferior accuracy compared to CT and therefore may empirically justify ultrasound as the preferred primary imaging investigation for suspected appendicitis in children. 1. Materials and methods 1.1. Study setting The study was undertaken at a single university teaching hospital following ethics review board approval. Our tertiary pediatric surgery service provides care for a local population of approximately 1.6 million persons. At our hospital, ultrasound is used almost exclusively as the primary imaging modality for acute appendicitis. A dedicated pediatric ultrasonography service is available with provision for scans to be conducted at all hours of the day as clinical need arises. Our sonography service is staffed by an experienced group of specialist pediatric sonographers. Investigations for “query appendicitis” are identified with a separate coding number.
http://dx.doi.org/10.1016/j.jpedsurg.2016.09.009 0022-3468/© 2016 Elsevier Inc. All rights reserved.
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
2
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
The sonographic technique at our institution is a standard graded compression examination. The ultrasound used for all examinations was an iu22 Ultrasound System with a C8-5 tightly curved array probe transducer (Philips Healthcare, Amsterdam, Netherlands). In special circumstances of obese or older children with large body habitus, a C9-4 curved array transducer was used. There are no fixed sonographic diagnostic criteria for acute appendicitis. Instead, the radiological diagnosis was made using a combination of primary and secondary factors in a case-dependent manner that included; 1) appendix diameter N 6 mm, 2) degree of localized hypervascularity, 3) presence of surrounding free fluid, 4) compressibility of the appendix, 5) presence of a luminal fecolith, 6) focal ultrasound probe tenderness and 7) appearances of surrounding soft tissue.
only by histopathology confirmation such that ultrasound positive cases treated nonoperatively (with or without antibiotic therapy) were coded as negative appendicectomies for the purpose of diagnostic accuracy calculations. Additionally, the denominator value for all analyses was assigned as the total number of ultrasound examinations for each subgroup (positive, negative or equivocal) and not adjusted to exclude repeat ultrasound examinations per patient. Kruskal–Wallis one-way analysis of variance was used to compare negative appendicectomy rates between ultrasound positive, negative and nondiagnostic groups. If significant differences were determined, then the Chi-square test was used to further compare outcomes between 2 groups. The Chi-square test was also used to compare the rates of appendicitis between equivocal and nonvisualization ultrasound examinations. Statistical significance was regarded as P b 0.05. Statistical analysis was performed using SPSS version 21.0 (IBM Corp, NY).
1.3. Data collection
2. Results
The study period comprised six complete calendar years from 1st January 2009 to 31st December 2014. Radiologist reports for consecutive ultrasound studies requested for “query appendicitis” were retrieved from a central hospital database. Outcomes of ultrasound investigations were categorized as positive, negative, or nondiagnostic. Nondiagnostic ultrasounds were further subcategorized as either equivocal or nonvisualization examinations. If the vermiform appendix was unable to be visualized in its entirety, then the ultrasound examination was categorized as not visualized. If the vermiform appendix was visualized but there was diagnostic uncertainty, then the ultrasound examination was categorized as equivocal. If the same patient underwent repeat ultrasound examinations, then each examination was recorded as an independent event in the database. Electronic medical records were evaluated to identify children who proceeded to appendicectomy following ultrasound investigation. To corroborate ultrasound findings, both the operative findings and histopathology findings were evaluated. Histopathology findings were categorized as positive or negative relative to a final pathological diagnosis of acute appendicitis. Other diagnoses were also documented when confirmed in the presence or absence of acute appendicitis at the time of planned appendicectomy.
A total of 3799 ultrasound examinations were evaluated. The mean (± SD) patient age was 11.5 (± 3.8) years. Gender distribution was 1:1.2 for males:females. Approximately one-quarter (29.0%, 1103/3799) of those investigated with ultrasound for query appendicitis proceeded to an appendicectomy. A total of 1840 appendicectomies for suspected acute appendicitis were performed during the study period. The proportion of patients who were investigated with ultrasound prior to appendicectomy was 59.9% (1103/1840). The prevalence of appendicitis in the study cohort was 27.6% (1049/3799). There were 14 patients with positive ultrasound examinations that were managed nonoperatively, and 5 patients who later underwent interval appendicectomy. Sonographic, clinical and pathological findings are summarized in Fig. 1. The vermiform appendix was confidently seen in its entirety in 91.7% (3484/3799) of ultrasound examinations. Ultrasound diagnosis was positive, negative and nondiagnostic in 27.3%, 61.1%, and 11.6% of examinations respectively. Alternative sonographic findings of symptomatic or incidental pathologies were demonstrated in 13.0% of ultrasound examinations and are summarized in Table 1. The proportion of patients with intraoperative findings of perforated appendicitis was 15.1%, 8.3% and 14.3% for positive, negative and nondiagnostic ultrasound groups respectively. Overall diagnostic accuracy was 95.5% (3627/3799). Sensitivity and specificity values for ultrasound were 97.1% (95.9–98.1; 95% CI) and 94.8% (93.9–95.6; 95% CI) respectively. Positive and negative predictive values for ultrasound were 87.8% (85.8–89.6%; 95 CI) and 98.9% (98.4–99.2; 95% CI) respectively. Positive and negative likelihood ratios for ultrasound were 18.8 (16.0–22.1; 95% CI) and 0.03 (0.02–0.04; 95% CI) respectively. Separate analysis of only ultrasound positive and negative examinations (i.e. excluding nondiagnostic examinations) yielded sensitivity and specificity values of 98.8% (97.9–99.4%; 95% CI) and 98.3% (97.7–98.8%; 95% CI). Among patients who had a nondiagnostic ultrasound, 13.9% underwent an appendicectomy, and acute appendicitis was confirmed in 9.1% (Fig. 1). The incidence of appendicitis in children with equivocal ultrasound examinations was significantly higher than that in children with nonvisualization ultrasound examinations (16.9% versus 5.1%; P b 0.001). The overall negative appendicectomy rate was 4.9% (54/1103). There was a significant difference in negative appendicectomy rates between the ultrasound positive, negative and nondiagnostic groups (1.8%, 50.0%, and 38.1% respectively; P b 0.001, Fig. 1). There was no difference in negative appendicectomy rates between the ultrasound negative and nondiagnostic groups (P = 0.325). Repeat ultrasound examinations of the same patient were performed in 3.9% (149/3799) of cases. In the majority of circumstances (94.0%, 140/149), repeat imaging was requested based on ongoing or recurring symptoms suspicious of acute appendicitis in the context of a previous ultrasound examination that was either negative (54.3%,
1.2. Standard ultrasound technique
1.4. Statistical analysis Diagnostic accuracy of ultrasound was determined by analyzing overall accuracy, sensitivity, specificity, predictivity and likelihood ratios. These test characteristics were calculated using conventional formulae with 95% confidence intervals (95% CI). Methodological handling of equivocal or nonvisualization ultrasound examination data is inconsistent in the literature and can be misrepresentative [14]. Processing of these data was therefore modeled on consensus methodology among the most transparent and high quality related literature available [14–16]. For the main analysis of test characteristics, ultrasound findings were dichotomized as either positive or negative. Post-hoc processing was performed to recategorize equivocal examinations as positive, and nonvisualization examinations as negative. This recategorization was based on assumptions of 1) equivocal examinations demonstrating abnormal findings that were genuinely suspicious but not convincingly diagnostic for acute appendicitis, and 2) nondiseased vermiform appendixes being typically less apparent on ultrasound and more likely to be nonvisualized. To highlight the impact that is introduced by ignoring equivocal or nonvisualization examinations in calculations of diagnostic accuracy, separate analysis was performed that only included ultrasound examinations that were truly test positive or test negative (i.e. excluding equivocal and nonvisualization examinations). In addition to including nondiagnostic ultrasound examinations in the main analysis, data were further interpreted based on intentionto-treat principles by categorizing acute appendicitis (true positive)
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
3
Fig. 1. Decision tree outlining ultrasound, clinical and pathological outcomes for the study sample (n = 3799 ultrasound examinations).
Table 1 Alternative or additional sonographic findings in children who underwent ultrasound examinations for query appendicitis (n = 3799). Colitis 108 Mesenteric adenitis 98 Ovarian cyst 59 Terminal ileitis 35 Lymphoid hyperplasia of terminal ileum 21 Pyelonephritis 17 Inflammatory phlegmon or abdominal collection 14 Omental infarction 14 Fecal loading 13 Ovarian torsion 10 Small bowel obstruction 8 Ovarian mass or nodule 7 Pneumonia 7 Cystitis 6 Hip synovitis or joint effusion 5 Typhlitis 5 Intussusception 5 Renal cyst 4 Cholelithiasis 4 Pelvicalyceal dilatation 3 Free intraabdominal fluid with appendix unable to be visualized 3 Intestinal malrotation 3 Ureteric stone 3 2a Cholecystitis, abdominal solid or cystic tumor, ascites and splenomegaly, abdominal wall cellulitis, pleural effusion, urachal remnant, osteitis pubis, pancreatitis, renal stone, uterine didelphys and hematometrocolpos 1a Pelvic inflammatory disease, enthesitis or myositis of the lower limb, intraabdominal testis, nephritis, tubo-ovarian abscess, liver cyst or other lesion, intestinal parasites, salpingitis, accessory spleen torsion, hepatomegaly, hepatomegaly, rectus sheath hematoma, hematosalpinx, hepatic hemangioma, intrahepatic biliary tree dilatation, Meckel's diverticulum, vesicoureteric junction obstruction, epididymo-orchitis, renal abscess, duodenitis, pyosalpinx, pelviureteric junction obstruction, Gartner’s duct remnant or epoophoron a
Number of occurrences for each listed sonographic finding.
76/140) or equivocal (45.7%, 64/140). The remaining indications for repeat ultrasound examination were for interval assessment of acute appendicitis that was being managed nonoperatively with antibiotic therapy (4.7%, 7/149), or for further evaluation of a positive ultrasound that poorly corresponded with clinical findings (1.3%, 2/149). There was a low positive diagnostic yield for repeat ultrasounds with only 8 new positive ultrasound diagnoses arising from previously negative or equivocal ultrasound examinations (5.7%, 8/140). Repeat ultrasound was only able to clarify previously nondiagnostic ultrasounds as positive or negative in approximately half of cases (51.6%, 33/64). There were only 27 patients (0.7%, 27/3799) who proceeded to CT scan examinations following initial ultrasound examination. The most frequent reason for subsequent CT scan was for further investigation of abnormal ultrasound findings after demonstration of a sonographically normal appendix (40.7%, 11/27). Other reasons for CT examinations were for large body habitus (18.5%, 5/27) or inability to tolerate ultrasound probe pressure (7.4%, 2/27) that resulted in a nondiagnostic ultrasound examination. A CT diagnosis of appendicitis occurred on only 3 occasions. 3. Discussion Advanced imaging has transformed diagnostic processes for acute appendicitis. Past reliance on clinical assessment alone has evolved toward current practice where some argue that imaging should be undertaken in all children before appendicectomy [12]. In this series, preoperative ultrasound was used in 59.9% of appendicectomies. In a large administrative database study involving 30 pediatric hospitals in the United States, the median percentage of patients who had preoperative imaging was 69% [17]. Other smaller series report preoperative ultrasound usage that ranges widely from 18% to 99% [5,15,17–21]. Reliance on imaging has increased owing to greater access to radiological services, improvements in imaging accuracy, and clinicians striving to achieve lower negative appendicectomy rates [5,12].
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
4
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
Perceptions of ultrasound and CT imaging modalities, and weighting of their corresponding pros and cons are well described [2]. Diagnostic accuracy is the most important discriminator when evaluating any diagnostic test or investigation. A recent meta-analysis by Doria et al. evaluated the comparative diagnostic accuracy of ultrasound versus CT for appendicitis in children [4]. Pooled analysis of results from 9356 patients determined that CT was significantly more sensitive compared to ultrasound (P = 0.001) [4]. Sensitivity and specificity values for ultrasound were 88% and 94% respectively, and for CT were 94% and 95% respectively [4]. While CT remains the preferred imaging choice for pediatric acute appendicitis in North America compared to ultrasound (median 34% for CT versus 6% for ultrasound), increased use of ultrasound since 2007 is encouraging and should be further supported based on the findings of this study [5,9]. Our entirely unselected study population is representative of the wide spectrum of patients who undergo ultrasound examination for clinically suspected appendicitis. As such, we feel that our study cohort fairly reflects actual daily clinical practice. According to Bayes theorem, observed diagnostic performance of a test such as ultrasound is dependent and proportional to the pretest probability of the disease in the patient population [12,14]. Lower disease prevalence in an unselected patient population such as this study will infer lower expected test accuracy compared to a more selected population such as those who are stratified to imaging according to various scoring algorithms or those who are initially clinically assessed by experienced surgical staff [12,14]. Various clinical assessment scoring algorithms such as the Alvarado and Pediatric Appendicitis Scores have been developed to risk stratify patients with suspected appendicitis and assign more selective use of imaging investigations [6,22–24]. Limited evidence for these score-based diagnostic aids indicates modest accuracy and reproducibility [6,25]. For these reasons, scoring algorithms are not yet recommended as isolated diagnostic tools in routine clinical practice and were not used at our institution during the study period [6]. 3.1. Benchmarking Numerous series in the literature describe diagnostic accuracies of ultrasound for acute appendicitis [4,6,9]. Existing gold-standard accuracies are represented by those reported by Goldin et al. and Baldisserotto et al. who published their series of 304 and 425 patients respectively [26,27]. Sensitivity and specificity values reported by Goldin et al. were 98.7% and 95.4%, however these were derived from post-hoc determined diagnostic criteria of maximal outside appendix diameter ≥ 7 mm or appendix wall thickness N 1.7 mm, and with an appendix visualization rate of 68% [27]. Baldisserotto et al. reported sensitivity and specificity values of 98.5% and 98.2% respectively, with positive and negative predictive values of 98.0% and 98.7% [6,26]. Nondiagnostic ultrasound examinations are an accepted reality in clinical practice and should therefore be appropriately incorporated into diagnostic accuracy calculations. Goldin et al. selectively excluded nondiagnostic ultrasounds for accuracy calculations and Baldisserotto et al. did not disclose their method of handling these data [26,27]. Considering the relative conservativeness of our highly competitive accuracy values with inclusion of nondiagnostic examination data, our unrivaled accuracy values with exclusion of nondiagnostic examination data, and excellent visualization rates in a large unselected consecutive series, we propose that our results should be considered among benchmark standards.
proportional relationship [14,15]. With approximately 60% of patients undergoing preoperative imaging, approximately 2 scans are performed each day. This case volume facilitates consistent exposure enabling skill acquisition, refinement and maintenance among our staff. Visualization of the appendix is obligatory for a diagnostic ultrasound examination. Our visualization rate of 91.7% is among the highest reported and is considerably higher than many other contemporary series that report visualization rates b 50% [14,15,21]. We attribute much of this high visualization rate to the virtues of patience and dedication among our staff. A detailed examination requires patience and available time to employ different visualization strategies such as patient repositioning, distraction techniques, emptying and filling the urinary bladder, and breath-holding maneuvers. Dedication and quarantined time are required for a more intensive personalized approach to each individual examination. In contrast to rigid criterion based diagnostic approaches, an individualized approach to each examination puts into context the constellation of sonographic features with available clinical information. For the diverse presentations of acute appendicitis, we feel that the flexibility of this approach is advantageous and improves diagnostic performance. An accurate investigation may require longer time, and therefore ultrasonography staff resources; however it facilitates targeted patient care and may be more cost effective overall for the patient's admission. A linear array transducer is routinely used in most centers internationally (Fig. 2). In our experience, a tightly curved array transducer is more suitable as it allows more targeted graded compression and relative size appropriateness for younger children (Fig. 2). For these reasons, we would advocate that the tightly curved array transducer should be considered as standard. A commonly cited drawback of ultrasound compared to CT is lower capability to diagnose alternate pathologies [2,8,9]. As highlighted in Table 1, we have found ultrasound to be highly versatile in diagnosing a diverse range of pathologies aside from acute appendicitis. This is facilitated by the dynamic real-time nature of ultrasound in that visualization of a normal appendix encourages more detailed assessment of other visceral structures that might yield pathological features.
3.2. Ultrasound technique Diagnostic quality of ultrasound is operator dependent. Experience is both necessary and invaluable in developing operator expertise. As experience is gained through case exposure, quality and volume of ultrasound examinations share a strong directly
Fig. 2. C8-5 tightly curved (left) and L17-5 linear (right) array transducers (Philips Healthcare, Amsterdam, Netherlands).
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009
T.P. Cundy et al. / Journal of Pediatric Surgery xxx (2016) xxx–xxx
3.3. Limitations Cost-effectiveness of ultrasound in suspected appendicitis is a contentious topic and was not specifically evaluated in this study. Cost-effectiveness and clinical effectiveness are deemed to be most favorable in the child with equivocal diagnosis on clinical assessment. In these circumstances, the opportunity to more confidently exclude appendicitis with use of ultrasound can contribute to considerable cost saving through either avoidance of hospital admission for a period of observation, or negative appendicectomy [1,7]. Previous studies have sought to further define the cohort of patients with truly equivocal clinical findings that are best suited for radiological investigation [3]. A limitation of this study is that the certainty of clinical diagnosis was not assessed, although the subjectivity of this parameter limits its reliability. An acceptable negative appendicectomy rate has traditionally been estimated as 10%–25% [8,12,25,28]. In contemporary series, rates of negative appendicectomy are reported to be as high as 17%–56% [3,23,29]. Although our 4.9% negative appendicectomy rate in this series was low, we are unable to empirically attribute this to the benefit of ultrasound, as this study does not include a comparator cohort of children who underwent appendicectomy during the same study period without ultrasound examination. Higher negative appendicectomy rates were observed in the subset of patients with negative or nondiagnostic ultrasound examinations. Based on our findings, the clinical corollary for nondiagnostic ultrasounds is that they should be interpreted as low risk for acute appendicitis, particularly nonvisualization examinations. Lastly, the application of our results may be limited to the circumstances of clinical practice at our institution. We recognize that there are considerable variations in clinical, technical, financial, resource allocation and medicolegal practices between healthcare institutions. Individually and collectively, each of these factors can impact on the delivery of clinical service and may account for differences in advanced imaging use and performance for acute appendicitis in children. We contend that our benchmark standard quality for ultrasound is tangibly achievable in most clinical environments with the appropriate case volume, staff skill, resource allocation, and attitude of involved clinicians.
4. Conclusions In this largest reported single institution series of ultrasound examinations for acute appendicitis in children, we report benchmark standard quality of diagnostic accuracy and visualization rates. Our findings also highlight the versatility of ultrasound and comparative accuracy compared to CT. Given the radiation and cost implications of CT, the role of ultrasound should be promoted as the primary imaging modality for acute appendicitis in children.
5
References [1] Axelrod DA, Sonnad SS, Hirschl RB. An economic evaluation of sonographic examination of children with suspected appendicitis. J Pediatr Surg 2000;35:1236–41. [2] Doria AS. Optimizing the role of imaging in appendicitis. Pediatr Radiol 2009; 39(Suppl. 2):S144–8. [3] Bullapur HM, Deshpande AV, Phin SJ, et al. Adjunct ultrasonography in children with suspected acute appendicitis: identifying the optimal target group. ANZ J Surg 2014; 84:326–30. [4] Doria AS, Moineddin R, Kellenberger CJ, et al. US or CT for diagnosis of appendicitis in children and adults? A meta-analysis. Radiology 2006;241:83–94. [5] Bachur RG, Hennelly K, Callahan MJ, et al. Advanced radiologic imaging for pediatric appendicitis, 2005–2009: trends and outcomes. J Pediatr 2012;160:1034–8. [6] Dingemann J, Ure B. Imaging and the use of scores for the diagnosis of appendicitis in children. Eur J Pediatr Surg 2012;22:195–200. [7] Rice HE, Arbesman M, Martin DJ, et al. Does early ultrasonography affect management of pediatric appendicitis? A prospective analysis. J Pediatr Surg 1999;34:754–8. [8] Hernanz-Schulman M. CT and US in the diagnosis of appendicitis: an argument for CT. Radiology 2010;255:3–7. [9] Strouse PJ. Pediatric appendicitis: an argument for US. Radiology 2010;255:8–13. [10] Frush DP, Frush KS, Oldham KT. Imaging of acute appendicitis in children: EU versus U.S. ... or US versus CT? A North American perspective. Pediatr Radiol 2009;39:500–5. [11] Muehlstedt SG, Pham TQ, Schmeling DJ. The management of pediatric appendicitis: a survey of North American pediatric surgeons. J Pediatr Surg 2004;39:875–9. [12] Holscher HC, Heij HA. Imaging of acute appendicitis in children: EU versus U.S. ... or US versus CT? A European perspective. Pediatr Radiol 2009;39:497–9. [13] Wan MJ, Krahn M, Ungar WJ, et al. Acute appendicitis in young children: costeffectiveness of US versus CT in diagnosis—a Markov decision analytic model. Radiology 2009;250:378–86. [14] Trout AT, Sanchez R, Ladino-Torres MF, et al. Critical evaluation of US for the diagnosis of pediatric acute appendicitis in a real-life setting: how can we improve the diagnostic value of sonography? Pediatr Radiol 2012;42:813–23. [15] Mittal MK, Dayan PS, Macias CG, et al. Performance of ultrasound in the diagnosis of appendicitis in children in a multicenter cohort. Acad Emerg Med 2013;20: 697–702. [16] Saucier A, Huang EY, Emeremni CA, et al. Prospective evaluation of a clinical pathway for suspected appendicitis. Pediatrics 2014;133:e88–95. [17] Newman K, Ponsky T, Kittle K, et al. Appendicitis 2000: variability in practice, outcomes, and resource utilization at thirty pediatric hospitals. J Pediatr Surg 2003;38:372–9. [18] Rice-Townsend S, Barnes JN, Hall M, et al. Variation in practice and resource utilization associated with the diagnosis and management of appendicitis at freestanding children's hospitals: implications for value-based comparative analysis. Ann Surg 2014;259:1228–34. [19] Neufeld D, Vainrib M, Buklan G, et al. Management of acute appendicitis: an imaging strategy in children. Pediatr Surg Int 2010;26:167–71. [20] Kotagal M, Richards MK, Flum DR, et al. Use and accuracy of diagnostic imaging in the evaluation of pediatric appendicitis. J Pediatr Surg 2015;50:642–6. [21] Scrimgeour DS, Driver CP, Stoner RS, et al. When does ultrasonography influence management in suspected appendicitis? ANZ J Surg 2014;84:331–4. [22] Alvarado AA. Practical score for the early diagnosis of acute appendicitis. Ann Emerg Med 1986;15:557–64. [23] Lintula H, Kokki H, Kettunen R, et al. Appendicitis score for children with suspected appendicitis. A randomized clinical trial. Langenbecks Arch Surg 2009;394:999–1004. [24] Samuel M. Pediatric appendicitis score. J Pediatr Surg 2002;37:877–81. [25] Morrow SE, Newman KD. Current management of appendicitis. Semin Pediatr Surg 2007;16:34–40. [26] Baldisserotto M, Marchiori E. Accuracy of noncompressive sonography of children with appendicitis according to the potential positions of the appendix. AJR Am J Roentgenol 2000;175:1387–92. [27] Goldin AB, Khanna P, Thapa M, et al. Revised ultrasound criteria for appendicitis in children improve diagnostic accuracy. Pediatr Radiol 2011;41:993–9. [28] Kaneko K, Tsuda M. Ultrasound-based decision making in the treatment of acute appendicitis in children. J Pediatr Surg 2004;39:1316–20. [29] Sivit CJ. Imaging children with acute right lower quadrant pain. Pediatr Clin North Am 1997;44:575–89.
Please cite this article as: Cundy TP, et al, Benchmarking the value of ultrasound for acute appendicitis in children, J Pediatr Surg (2016), http:// dx.doi.org/10.1016/j.jpedsurg.2016.09.009