- Email: [email protected]
Ultrasound for Diagnosis of Appendicitis in a Community Hospital Emergency Department has a High Rate of Nondiagnostic Studies
The Journal of Emergency Medicine, Vol. -, No. -, pp. 1–6, 2017 Ó 2017 Elsevier Inc. All rights reserved. 0736-4679/$ - see front matter
http://dx.doi.org/10.1016/j.jemermed.2017.01.003
Selected Topics: Emergency Radiology
ULTRASOUND FOR DIAGNOSIS OF APPENDICITIS IN A COMMUNITY HOSPITAL EMERGENCY DEPARTMENT HAS A HIGH RATE OF NONDIAGNOSTIC STUDIES Scott M. Alter, MD,* Brian Walsh, MD,† Patrick J. Lenehan, MD,‡ and Richard D. Shih, MD* *Division of Emergency Medicine, Charles E. Schmidt College of Medicine, Florida Atlantic University, Boca Raton, Florida, †Department of Emergency Medicine, Morristown Medical Center, Morristown, New Jersey, and ‡Department of Emergency Medicine, California Hospital Medical Center, Los Angeles, California Reprint Address: Scott M. Alter, MD, Division of Emergency Medicine, Florida Atlantic University at Bethesda Health, 2815 South Seacrest Blvd, Boynton Beach, FL 33435
, Abstract—Background: Radiation concerns are changing the way emergency physicians evaluate patients. This is especially prevalent in pediatrics, and exemplified by abdominal pain management. Large academic centerbased studies suggest appendix ultrasound (U/S) is sensitive and specific for appendicitis, with low nondiagnostic rates. Objectives: We sought to determine the diagnostic rate of appendix U/S and incidence of follow-up computed tomography (CT) imaging for pediatric patients at a community hospital. Methods: Design: Retrospective cohort. Setting: Emergency department with 85,000 annual visits. Population: Patients younger than 21 years old that had an appendix U/S over a 12-month period. U/S were performed by technicians and interpreted by radiologists. Investigators classified readings as ‘‘diagnostic’’ (‘‘positive’’ and ‘‘negative’’) or ‘‘non-diagnostic’’ (‘‘borderline’’ and ‘‘appendix not visualized’’) and identified follow-up CT studies and interpretations. Results: There were 441 pediatric appendix U/S performed; 26% were diagnostic (14% positive for appendicitis, 12% negative) and 74% nondiagnostic (5% borderline, 69% appendix not visualized). Follow-up CT scans were obtained in 19% of all patients, including 8% with positive U/S, 4% negative, 32% borderline, and 22% not visualized. Follow-up CT was nearly four times more likely in the nondiagnostic group than the diagnostic group (23% vs. 6%, p < 0.0001). Conclusion: The utility of U/S to diagnose appendicitis at a community hospital is limited by a high rate of nondiagnostic studies. Some patients with diagnostic U/S even had follow-up CT imaging. To minimize
radiation exposure in children, improvements should be made in the performance and acceptance of U/S as the primary modality of abdominal pain imaging at community hospitals. Ó 2017 Elsevier Inc. All rights reserved. , Keywords—ultrasound; appendix; pediatrics; emergency; community
INTRODUCTION The most common atraumatic surgical emergency in the pediatric age group is appendicitis (1). Physical examination alone is often not reliable enough for a surgeon to take a patient to the operating room. Therefore, additional diagnostic testing and imaging are utilized to determine a definitive diagnosis. Without the use of imaging, clinical examination and laboratory findings are the predominant tools utilized in the diagnosis of pediatric appendicitis. The most useful sign associated with appendicitis with the greatest likelihood ratio is fever; rebound tenderness, migration of pain, and leukocytosis are also helpful (2). Presence of pain upon hopping or walking increases the odds of appendicitis (3). White blood cell count >12 103/mL and leukocyte left shift in pediatric patients with nontraumatic abdominal pain have been shown to have high sensitivity
RECEIVED: 6 July 2016; FINAL SUBMISSION RECEIVED: 1 January 2017; ACCEPTED: 4 January 2017 1
2
and specificity for appendicitis (3,4). C-reactive protein level >3 mg/dL also has been shown to be predictive of appendicitis, and in combination with elevated white blood cell count has an odds ratio of 7.75 (3). Utilizing white blood cell count with signs and symptoms, the Alvarado score was devised as a tool in appendicitis diagnosis (5). Although this score may identify predictive factors, some studies have found it not to be reliable enough to exclude appendicitis in patients with low probability scores (6,7). Historically, computed tomography (CT) imaging of the abdomen/pelvis has been the best diagnostic test for imaging of the appendix. This CT for diagnosis of acute appendicitis has excellent sensitivity, specificity, and positive predictive value, all with percentages in the mid to high 90s, with fair negative predictive value, from 71–96% (6–9). However, concerns over radiation and contrast exposure have detracted from this being the first-line imaging for pediatrics. Contrast-induced nephropathy, generally defined as a >25% increase in serum creatinine levels over baseline, is not uncommon. Rates can vary from 0% to 90%, depending on risk factors, but the incidence in a healthy general population may be as low as 1–2% (10–12). As most children are generally without major health problems, the greater concern is the potential for malignancy from exposure to ionizing radiation. Pediatric CT imaging results significantly increased lifetime radiation risk when compared with adult CT imaging (13). It is estimated that CT imaging may lead to one malignant transformation for every 500 pediatric CT scans performed, which is around a 10-times greater risk than in adults (14,15). Due to this radiation risk, some research has suggested reducing radiation exposure during CT of the abdomen/ pelvis (16). One study with a 39% reduction in median absorbed radiation dose found no change in sensitivity or specificity for appendicitis diagnosis (17). However, even with reduced radiation exposure there continues to be risk of malignancy, and the potential side effects have continued to limit CT imaging in pediatrics. Over the past three decades, ultrasonography has emerged as a viable alternative for imaging the appendix and is frequently used in the diagnostic evaluation of appendicitis. There is evidence to suggest, however, that after an initial surge, its use had been waning in favor of CT in the late 2000s (18). As a response to overuse of CT imaging, collaborators in the State of Washington created the Safe and Sound campaign as an attempt to continue to reduce CT imaging and promote highquality ultrasound (U/S) imaging (18). More and more studies suggest increasing accuracy of U/S of the appendix, with sensitivities percentage generally in the high 80s and higher specificities percentage in the mid–low 90s (8,9,19,20). These studies tend to be at large, academic
S. M. Alter et al.
centers and may not reflect the real-world practice of how the test is performed and interpreted. Our experience at a community hospital did not seem to be aligned with other published statistical measures of the utility of U/S of the appendix, as we were experiencing a greater rate of nondiagnostic U/S with high rates of follow-up CT imaging. Therefore, our goal of this study was to examine the diagnostic rate of U/S for appendicitis in a community setting, as well as the frequency of follow-up CT abdomen/pelvis imaging. MATERIALS AND METHODS Design and Setting This study is a retrospective cohort study performed at a community teaching hospital with 85,000 emergency department (ED) visits per year. Selection of Participants Consecutive pediatric ED patients (<21 years old) who underwent an appendix ultrasound between January 1, 2012 and December 31, 2012 were enrolled. Methods and Measurements We queried the picture archiving and communication system database for pediatric patients who had an ultrasound of the appendix performed. Patients were excluded if the ultrasound order was not placed by an emergency physician. For each patient, we also recorded any follow-up CT abdomen/pelvis imaging performed after the U/S appendix and during the same ED visit, also obtained from the picture archiving and communication system database. The decision to order an U/S, as well as any subsequent CT, was at the discretion of the treating emergency medicine attending physician. American Registry for Diagnostic Medical Sonography-certified sonographers performed all U/S based on standard department protocol: using the classic graded compression technique with 6–15-MHz linear-array probes. One of 14 boardcertified radiologists interpreted each U/S study, though some cases had initial preliminary readings provided by radiology residents prior to attending over-reading. Two study investigators reviewed all of the U/S final radiology reports and classified the reading for appendicitis as ‘‘positive,’’ ‘‘negative,’’ ‘‘borderline,’’ or ‘‘not visualized.’’ Positive was defined as a report that diagnosed appendicitis. Negative was defined as a report that either the radiologist diagnosed as not appendicitis or identified a normal appendix. Borderline was a report that mentioned any abnormal or equivocal findings possibly related to appendicitis, without the diagnosis
Ultrasound for Diagnosis of Appendicitis
3
being documented by the radiologist (eg, borderline tubular structure, minimal surrounding edema). Not visualized was defined as a report that did not identify the appendix. ‘‘Positive’’ and ‘‘negative’’ results were considered ‘‘diagnostic,’’ and ‘‘borderline’’ and ‘‘not visualized’’ results were considered ‘‘non-diagnostic.’’ Analysis We tallied the number of patients based on radiology report classification, and calculated percentages and 95% confidence intervals (CI). We then calculated the percentage of patients who had a follow-up CT abdomen/pelvis scan and 95% CIs for each report classification. Percent difference of follow-up CT between diagnostic and nondiagnostic was calculated, along with 95% CI and p-value, using a one-tailed test with alpha set at 0.05. We also examined and tabulated the follow-up CT results stratified by U/S interpretation. RESULTS During the 12-month period, 441 pediatric U/S of the appendix were performed in ED patients. The average age of patients was 11.7 years, with a range from 9 months to 20 years; 41% of patients were male. In total, 114 (26%, 95% CI 22–30) U/S studies were classified as diagnostic, whereas 327 (74%, 95% CI 70–78) were classified as nondiagnostic. Fourteen percent (95% CI 11–18) of all U/S were classified as positive, 12% (95% CI 9–15) were negative, 5% (95% CI 3–7) were borderline, and 69% (95% CI 65–73) were appendix not visualized (Table 1). Follow-up CT scans were obtained in 82 (19%, 95% CI 15–22) patients, including 8% (95% CI 1–15) of patients with diagnostic positive U/S, 4% (95% CI 1–9) of patients with diagnostic negative U/S, 32% (95% CI 12–52) of patients with nondiagnostic borderline U/S, and 22% (95% CI 18–27) of patients with nondiagnostic Table 1. Appendix Ultrasound (U/S) Results by Impression with Percentage of Follow-Up Abdominal CT Imaging Total U/S Impression Diagnostic Positive Negative Total Nondiagnostic Borderline Not visualized Total
Follow-up CT
n
%
95% CI
n
%
63 51 114
14% 12% 26%
11–18 9–15 22–30
5 2 7
8% 4% 6%
1–15 1–9 2–11
22 305 327
5% 69% 74%
3–7 65–73 70–78
7 68 75
32% 22% 23%
12–52 18–27 18–27
CT = computed tomography; CI = confidence interval.
95% CI
appendix not visualized U/S. Of all the diagnostic U/S studies, 6% (95% CI 2–11) had follow-up CT scans, and of all the nondiagnostic U/S studies, 23% (95% CI 18–27) had follow-up CT scans (Table 1). The percentage difference of follow-up CT scans performed between diagnostic and nondiagnostic U/S studies is 17% (95% CI 10–24, p < 0.0001). Of the 7 patients with diagnostic U/S imaging who had follow-up CT imaging, appendicitis was found on CT in all of the patients who had a positive U/S and none of the patients who had a negative U/S. Of the 7 patients with borderline U/S imaging who had follow-up CT imaging, 71% (95% CI 35–108) were found to have appendicitis on CT and 29% (95% CI 8–65) were found not to have appendicitis. Of the 68 patients with no visualized appendix on U/S who had a follow-up CT, 22% (95% CI 12–32) were found to have appendicitis on CT, 71% (95% CI 60–81) were found not to have appendicitis, and 7% (95% CI 1–14) had borderline findings (Table 2). Additionally, of the patients with no visualized appendix on U/S, 7 (10%, 95% CI 3–18) did not have their appendices visualized on CT either. One of these 7 was suspected of having early appendicitis based on secondary findings, whereas the other 6 had no secondary signs of appendicitis and were presumed negative. DISCUSSION Pediatric patients with abdominal pain are common occurrences in the ED, and often require imaging to determine the need for surgical intervention. Our study found a high rate of nondiagnostic appendix U/S imaging, with a four-times greater rate of follow-up CT imaging compared with patients with diagnostic U/S imaging. A previous study of children and adults conducted at a community hospital that routinely performs U/S for evaluation of appendicitis reported a sensitivity of 83% and a specificity of 95% (21). However, this study enrolled patients based upon radiology reports containing the word ‘‘appendicitis’’ (21). This methodology does not account for U/S performed to rule out appendicitis when ‘‘appendicitis’’ is not explicitly mentioned, which could lead to fewer negative (‘‘normal appendix’’) or nondiagnostic (‘‘appendix not visualized’’) reports included, and thus falsely elevated test performance measures. Despite having lower test characteristics at community hospitals vs. academic centers, most agree that U/S should still be the first-line study when evaluating a pediatric patient with signs and symptoms of appendicitis–mainly due to risks of CT imaging, including concerns about radiation and intravenous contrast exposure. Although diagnostic accuracy can be high for both CT and U/S, CT is better at diagnosing negative findings and U/S is better at diagnosing positive findings (9). Many researchers have
4
S. M. Alter et al.
Table 2. Abdominal CT Imaging Results for Appendicitis by Initial Ultrasound (U/S) Impression CT Impression Positive U/S Impression Diagnostic Positive Negative Nondiagnostic Borderline Not visualized
n
n
%
5 2
5 0
100% 0%
7 68
5 15
71% 22%
Negative 95% CI
(35–108) (12–32)
n
%
0 2
0% 100%
2 48
29% 71%
Borderline 95% CI
8–65 60–81
n
%
0 0
0% 0%
0 5
0% 7%
95% CI
1–14
CT = computed tomography; CI = confidence interval.
proposed using an algorithm in which CT is utilized only if the initial U/S examination is indeterminate (20). Unfortunately, as our study demonstrates, many appendix U/S readings are indeterminate. There may be many factors for this, including imaging adequacy. Although all of our sonographers are certified by the American Registry for Diagnostic Medical Sonography, U/S is an operator-dependent test. Compared with sonographers in academic centers, community sonographers may not have as much experience performing pediatric ultrasounds. Therefore, the appendix may not be identified or only partially visualized. Other patient-related predictors of nondiagnostic U/S imaging include increased body mass index, retrocecal location, normal appendix, perforation, and distal tip inflammation (22). Another possible reason for increased indeterminate readings is variability in radiologist interpretations. Radiologists at community hospitals may have less experience reading pediatric ultrasounds when compared with academic radiologists with subspecialized training. At our hospital, there is only one board-certified pediatric radiologist who reads all the pediatric U/S performed on weekdays during normal business hours. At other times, these radiologic studies were interpreted by general radiologists. One way to attempt to mitigate interpretation variability is for the radiologist to use standardized interpretative categories. For example, by including presence or absence of secondary signs when the appendix was not visualized, diagnostic accuracy is improved to 96.8% when compared with 94.1% for a binary interpretative scheme (23). Some studies have created protocols in an attempt to limit CT use. Protocols utilizing CT only after indeterminate appendix U/S have been found to have higher sensitivities and specificities than U/S alone, and a low negative appendectomy rate (19). By following a clinical practice guideline, one children’s hospital was able to reduce CT imaging by 41% without loss of diagnostic accuracy (24). In cases where U/S is nondiagnostic with persistent suspicion of appendicitis, follow-up CT imaging is a reasonable next step in the diagnostic algorithm.
Although nonvisualization of the appendix traditionally has not ruled appendicitis in or out, previous research has suggested that a nondiagnostic U/S is still useful when there are associated ultrasonographic inflammatory markers. The specificity of moderate-to-large free fluid collections was 98%; phlegmon: 100%; pericecal inflammatory fat changes: 98%; and any free fluid with prominent lymph nodes: 81%; with multiple markers greatly increasing the odds ratio of appendicitis (25). In another study, 15.6% of patients with incompletely viewed appendices had laboratory-confirmed appendicitis, with secondary U/S signs having a sensitivity of 40% and specificity of 91% (26). These findings suggest that when utilizing U/S alone, patients with incompletely visualized appendices and presence of sonographic secondary signs most likely have appendicitis and do not need additional imaging. Our study showed similar trends among nondiagnostic U/S, with borderline studies having a greater rate of appendicitis diagnosed on follow-up CT (71% positive vs. 29% negative), whereas U/S studies with nonvisualized appendices tended to have a higher rate of negative CT scans (22% positive vs. 71% negative). Other protocols and clinical decision rules have been created to identify low-risk patients for discharge. Leeuwenburgh et al. determined that patients with negative or inconclusive U/S of the appendix and fewer than two predictors of male gender, migration of pain to the right lower quadrant, vomiting, and white blood cell count higher than 12 103/mL can be discharged home with next-day follow-up, without the need for additional imaging (27). This clinical decision rule had a 94% negative predictive value for appendicitis. Another study concluded that patients with a nondiagnostic U/S without leukocytosis do not require any further diagnostic imaging (28). Limitations As this was a retrospective study, we did have some limitations. First, during the study period, there was no
Ultrasound for Diagnosis of Appendicitis
standard protocol used by the treating emergency physicians in their clinical decision-making when ordering imaging studies. This lack of standardization may have led some physicians to order more CT imaging due to individual clinical practice. Second, although not directly examined in this study, radiology readings were frequently accompanied by a ‘‘disclaimer.’’ These statements, such as ‘‘if there is suspicion for appendicitis, please correlate clinically,’’ seem to impact subsequent emergency physicians’ ordering of CT imaging. This may represent a real-world effect of how ultrasound reports are written and acts as an effect modifier that needs further investigation. Third, our study did not have clinical or pathologic outcome of each patient’s medical course. Although a limitation of this study, our focus was on the impact of initial U/S test results and the subsequent utilization of CT scan testing. Finally, this study was not blinded. The investigators classifying radiology reports for appendicitis (‘‘positive,’’ ‘‘negative,’’ ‘‘borderline,’’ or ‘‘not visualized’’) were aware of the research questions of the study, and may have introduced bias. However, the investigators were not aware of which patients had follow-up CT imaging performed at the time of U/S radiology report review. CONCLUSIONS The utility of U/S to diagnose appendicitis at a community hospital seems to be limited by a very high rate of nondiagnostic studies. Even some patients with a diagnostic U/S had follow-up CT imaging, which were all in agreement with the initial U/S findings. When obtaining an appendix U/S, if it is diagnostic, there should be no need to obtain follow-up imaging. However, until U/S demonstrates improved and consistent test characteristics, it would be prudent to have a standardized approach, including alternative diagnostic tests for work-up of pediatric abdominal pain. REFERENCES 1. Rothrock SG, Pagane J. Acute appendicitis in children: emergency department diagnosis and management. Ann Emerg Med 2000;36: 39–51. 2. Bundy DG, Byerley JS, Liles EA, et al. Does this child have appendicitis? JAMA 2007;298:438–51. 3. Kwan KY, Nager AL. Diagnosing pediatric appendicitis: usefulness of laboratory markers. Am J Emerg Med 2010;28:1009–15. 4. Wang LT, Prentiss KA, Simon JZ, et al. The use of white blood cell count and left shift in the diagnosis of appendicitis in children. Pediatr Emerg Care 2007;23:69–76. 5. Alvarado A. A practical score for the early diagnosis of acute appendicitis. Ann Emerg Med 1986;15:557–64. 6. Apisarnthanarak P, Suvannarerg V, Pattaranutaporn P, et al. Alvarado score: can it reduce unnecessary CT scans for evaluation of acute appendicitis? Am J Emerg Med 2015;33:266–70.
5 7. Ceydeli A, Lavotshkin S, Yu J, et al. When should we order a CT scan and when should we rely on the results to diagnose an acute appendicitis? Curr Surg 2006;63:464–8. 8. 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. 9. Johansson EP, Rydh A, Riklund KA. Ultrasound, computed tomography, and laboratory findings in the diagnosis of appendicitis. Acta Radiol 2007;48:267–73. 10. Berg KJ. Nephrotoxicity related to contrast media. Scand J Urol Nephrol 2000;34:317–22. 11. Gleeson TG, Bulugahapitiya S. Contrast-induced nephropathy. AJR Am J Roentgenol 2004;183:1673–89. 12. Journy N, Ancelet S, Rehel JL, et al. Predicted cancer risks induced by computed tomography examinations during childhood, by a quantitative risk assessment approach. Radiat Environ Biophys 2014;53:39–54. 13. Ogbole GI. Radiation dose in paediatric computed tomography: risks and benefits. Ann Ib Postgrad Med 2010;8:118–26. 14. Brenner DJ. Estimating cancer risks from pediatric CT: going from the qualitative to the quantitative. Pediatr Radiol 2002;32:228–31. discussion 242-4. 15. Brenner D, Elliston C, Hall E, et al. Estimated risks of radiationinduced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176:289–96. 16. Miglioretti DL, Johnson E, Williams A, et al. The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk. JAMA Pediatr 2013;167: 700–7. 17. Callahan MJ, Anandalwar SP, MacDougall RD, et al. Pediatric CT dose reduction for suspected appendicitis: a practice quality improvement project using artificial gaussian noise–part 2, clinical outcomes. AJR Am J Roentgenol 2015;204:636–44. 18. Kotagal M, Richards MK, Chapman T, et al. Improving ultrasound quality to reduce computed tomography use in pediatric appendicitis: the Safe and Sound campaign. Am J Surg 2015;209:896– 900. discussion 900. 19. Kaiser S, Frenckner B, Jorulf HK. Suspected appendicitis in children: US and CT—a prospective randomized study. Radiology 2002;223:633–8. 20. Teo EL, Tan KP, Lam SL, et al. Ultrasonography and computed tomography in a clinical algorithm for the evaluation of suspected acute appendicitis in children. Singapore Med J 2000; 41:387–92. 21. Chan I, Bicknell SG, Graham M. Utility and diagnostic accuracy of sonography in detecting appendicitis in a community hospital. AJR Am J Roentgenol 2005;184:1809–12. 22. Schuh S, Man C, Cheng A, et al. Predictors of non-diagnostic ultrasound scanning in children with suspected appendicitis. J Pediatr 2011;158:112–8. 23. Larson DB, Trout AT, Fierke SR, et al. Improvement in diagnostic accuracy of ultrasound of the pediatric appendix through the use of equivocal interpretive categories. AJR Am J Roentgenol 2015;204: 849–56. 24. Russell WS, Schuh AM, Hill JG, et al. Clinical practice guidelines for pediatric appendicitis evaluation can decrease computed tomography utilization while maintaining diagnostic accuracy. Pediatr Emerg Care 2013;29:568–73. 25. Estey A, Poonai N, Lim R. Appendix not seen: the predictive value of secondary inflammatory sonographic signs. Pediatr Emerg Care 2013;29:435–9. 26. Ross MJ, Liu H, Netherton SJ, et al. Outcomes of children with suspected appendicitis and incompletely visualized appendix on ultrasound. Acad Emerg Med 2014;21:538–42. 27. Leeuwenburgh MM, Stockmann HB, Bouma WH, et al. A simple clinical decision rule to rule out appendicitis in patients with nondiagnostic ultrasound results. Acad Emerg Med 2014;21: 488–96. 28. Cohen B, Bowling J, Midulla P, et al. The non-diagnostic ultrasound in appendicitis: is a non-visualized appendix the same as a negative study? J Pediatr Surg 2015;50:923–7.
6
S. M. Alter et al.
ARTICLE SUMMARY 1. Why is this topic important? Appendicitis is the most common pediatric surgical emergency, yet the commonly preferred initial imaging modality, ultrasound, does not seem to have great utility. Although previous studies have shown high accuracy, our experience has been different in a community emergency department. 2. What does this study attempt to show? This study demonstrates the low utility of ultrasound in the work-up of suspected pediatric appendicitis, and the rate of subsequent computed tomography (CT) scans performed. 3. What are the key findings? Pediatric ultrasound of the appendix was nondiagnostic 74% of the time, and patients who had nondiagnostic readings were almost four times more likely to have a subsequent abdominal CT scan. 4. How is patient care impacted? To avoid extraneous CT use and its associated risks, improvements should be made in the performance and acceptance of ultrasound imaging. Additionally, protocols should be developed and followed to limit CT use to the appropriate clinical settings for pediatric abdominal pain.
http://dx.doi.org/10.1016/j.jemermed.2017.01.003
Selected Topics: Emergency Radiology
ULTRASOUND FOR DIAGNOSIS OF APPENDICITIS IN A COMMUNITY HOSPITAL EMERGENCY DEPARTMENT HAS A HIGH RATE OF NONDIAGNOSTIC STUDIES Scott M. Alter, MD,* Brian Walsh, MD,† Patrick J. Lenehan, MD,‡ and Richard D. Shih, MD* *Division of Emergency Medicine, Charles E. Schmidt College of Medicine, Florida Atlantic University, Boca Raton, Florida, †Department of Emergency Medicine, Morristown Medical Center, Morristown, New Jersey, and ‡Department of Emergency Medicine, California Hospital Medical Center, Los Angeles, California Reprint Address: Scott M. Alter, MD, Division of Emergency Medicine, Florida Atlantic University at Bethesda Health, 2815 South Seacrest Blvd, Boynton Beach, FL 33435
, Abstract—Background: Radiation concerns are changing the way emergency physicians evaluate patients. This is especially prevalent in pediatrics, and exemplified by abdominal pain management. Large academic centerbased studies suggest appendix ultrasound (U/S) is sensitive and specific for appendicitis, with low nondiagnostic rates. Objectives: We sought to determine the diagnostic rate of appendix U/S and incidence of follow-up computed tomography (CT) imaging for pediatric patients at a community hospital. Methods: Design: Retrospective cohort. Setting: Emergency department with 85,000 annual visits. Population: Patients younger than 21 years old that had an appendix U/S over a 12-month period. U/S were performed by technicians and interpreted by radiologists. Investigators classified readings as ‘‘diagnostic’’ (‘‘positive’’ and ‘‘negative’’) or ‘‘non-diagnostic’’ (‘‘borderline’’ and ‘‘appendix not visualized’’) and identified follow-up CT studies and interpretations. Results: There were 441 pediatric appendix U/S performed; 26% were diagnostic (14% positive for appendicitis, 12% negative) and 74% nondiagnostic (5% borderline, 69% appendix not visualized). Follow-up CT scans were obtained in 19% of all patients, including 8% with positive U/S, 4% negative, 32% borderline, and 22% not visualized. Follow-up CT was nearly four times more likely in the nondiagnostic group than the diagnostic group (23% vs. 6%, p < 0.0001). Conclusion: The utility of U/S to diagnose appendicitis at a community hospital is limited by a high rate of nondiagnostic studies. Some patients with diagnostic U/S even had follow-up CT imaging. To minimize
radiation exposure in children, improvements should be made in the performance and acceptance of U/S as the primary modality of abdominal pain imaging at community hospitals. Ó 2017 Elsevier Inc. All rights reserved. , Keywords—ultrasound; appendix; pediatrics; emergency; community
INTRODUCTION The most common atraumatic surgical emergency in the pediatric age group is appendicitis (1). Physical examination alone is often not reliable enough for a surgeon to take a patient to the operating room. Therefore, additional diagnostic testing and imaging are utilized to determine a definitive diagnosis. Without the use of imaging, clinical examination and laboratory findings are the predominant tools utilized in the diagnosis of pediatric appendicitis. The most useful sign associated with appendicitis with the greatest likelihood ratio is fever; rebound tenderness, migration of pain, and leukocytosis are also helpful (2). Presence of pain upon hopping or walking increases the odds of appendicitis (3). White blood cell count >12 103/mL and leukocyte left shift in pediatric patients with nontraumatic abdominal pain have been shown to have high sensitivity
RECEIVED: 6 July 2016; FINAL SUBMISSION RECEIVED: 1 January 2017; ACCEPTED: 4 January 2017 1
2
and specificity for appendicitis (3,4). C-reactive protein level >3 mg/dL also has been shown to be predictive of appendicitis, and in combination with elevated white blood cell count has an odds ratio of 7.75 (3). Utilizing white blood cell count with signs and symptoms, the Alvarado score was devised as a tool in appendicitis diagnosis (5). Although this score may identify predictive factors, some studies have found it not to be reliable enough to exclude appendicitis in patients with low probability scores (6,7). Historically, computed tomography (CT) imaging of the abdomen/pelvis has been the best diagnostic test for imaging of the appendix. This CT for diagnosis of acute appendicitis has excellent sensitivity, specificity, and positive predictive value, all with percentages in the mid to high 90s, with fair negative predictive value, from 71–96% (6–9). However, concerns over radiation and contrast exposure have detracted from this being the first-line imaging for pediatrics. Contrast-induced nephropathy, generally defined as a >25% increase in serum creatinine levels over baseline, is not uncommon. Rates can vary from 0% to 90%, depending on risk factors, but the incidence in a healthy general population may be as low as 1–2% (10–12). As most children are generally without major health problems, the greater concern is the potential for malignancy from exposure to ionizing radiation. Pediatric CT imaging results significantly increased lifetime radiation risk when compared with adult CT imaging (13). It is estimated that CT imaging may lead to one malignant transformation for every 500 pediatric CT scans performed, which is around a 10-times greater risk than in adults (14,15). Due to this radiation risk, some research has suggested reducing radiation exposure during CT of the abdomen/ pelvis (16). One study with a 39% reduction in median absorbed radiation dose found no change in sensitivity or specificity for appendicitis diagnosis (17). However, even with reduced radiation exposure there continues to be risk of malignancy, and the potential side effects have continued to limit CT imaging in pediatrics. Over the past three decades, ultrasonography has emerged as a viable alternative for imaging the appendix and is frequently used in the diagnostic evaluation of appendicitis. There is evidence to suggest, however, that after an initial surge, its use had been waning in favor of CT in the late 2000s (18). As a response to overuse of CT imaging, collaborators in the State of Washington created the Safe and Sound campaign as an attempt to continue to reduce CT imaging and promote highquality ultrasound (U/S) imaging (18). More and more studies suggest increasing accuracy of U/S of the appendix, with sensitivities percentage generally in the high 80s and higher specificities percentage in the mid–low 90s (8,9,19,20). These studies tend to be at large, academic
S. M. Alter et al.
centers and may not reflect the real-world practice of how the test is performed and interpreted. Our experience at a community hospital did not seem to be aligned with other published statistical measures of the utility of U/S of the appendix, as we were experiencing a greater rate of nondiagnostic U/S with high rates of follow-up CT imaging. Therefore, our goal of this study was to examine the diagnostic rate of U/S for appendicitis in a community setting, as well as the frequency of follow-up CT abdomen/pelvis imaging. MATERIALS AND METHODS Design and Setting This study is a retrospective cohort study performed at a community teaching hospital with 85,000 emergency department (ED) visits per year. Selection of Participants Consecutive pediatric ED patients (<21 years old) who underwent an appendix ultrasound between January 1, 2012 and December 31, 2012 were enrolled. Methods and Measurements We queried the picture archiving and communication system database for pediatric patients who had an ultrasound of the appendix performed. Patients were excluded if the ultrasound order was not placed by an emergency physician. For each patient, we also recorded any follow-up CT abdomen/pelvis imaging performed after the U/S appendix and during the same ED visit, also obtained from the picture archiving and communication system database. The decision to order an U/S, as well as any subsequent CT, was at the discretion of the treating emergency medicine attending physician. American Registry for Diagnostic Medical Sonography-certified sonographers performed all U/S based on standard department protocol: using the classic graded compression technique with 6–15-MHz linear-array probes. One of 14 boardcertified radiologists interpreted each U/S study, though some cases had initial preliminary readings provided by radiology residents prior to attending over-reading. Two study investigators reviewed all of the U/S final radiology reports and classified the reading for appendicitis as ‘‘positive,’’ ‘‘negative,’’ ‘‘borderline,’’ or ‘‘not visualized.’’ Positive was defined as a report that diagnosed appendicitis. Negative was defined as a report that either the radiologist diagnosed as not appendicitis or identified a normal appendix. Borderline was a report that mentioned any abnormal or equivocal findings possibly related to appendicitis, without the diagnosis
Ultrasound for Diagnosis of Appendicitis
3
being documented by the radiologist (eg, borderline tubular structure, minimal surrounding edema). Not visualized was defined as a report that did not identify the appendix. ‘‘Positive’’ and ‘‘negative’’ results were considered ‘‘diagnostic,’’ and ‘‘borderline’’ and ‘‘not visualized’’ results were considered ‘‘non-diagnostic.’’ Analysis We tallied the number of patients based on radiology report classification, and calculated percentages and 95% confidence intervals (CI). We then calculated the percentage of patients who had a follow-up CT abdomen/pelvis scan and 95% CIs for each report classification. Percent difference of follow-up CT between diagnostic and nondiagnostic was calculated, along with 95% CI and p-value, using a one-tailed test with alpha set at 0.05. We also examined and tabulated the follow-up CT results stratified by U/S interpretation. RESULTS During the 12-month period, 441 pediatric U/S of the appendix were performed in ED patients. The average age of patients was 11.7 years, with a range from 9 months to 20 years; 41% of patients were male. In total, 114 (26%, 95% CI 22–30) U/S studies were classified as diagnostic, whereas 327 (74%, 95% CI 70–78) were classified as nondiagnostic. Fourteen percent (95% CI 11–18) of all U/S were classified as positive, 12% (95% CI 9–15) were negative, 5% (95% CI 3–7) were borderline, and 69% (95% CI 65–73) were appendix not visualized (Table 1). Follow-up CT scans were obtained in 82 (19%, 95% CI 15–22) patients, including 8% (95% CI 1–15) of patients with diagnostic positive U/S, 4% (95% CI 1–9) of patients with diagnostic negative U/S, 32% (95% CI 12–52) of patients with nondiagnostic borderline U/S, and 22% (95% CI 18–27) of patients with nondiagnostic Table 1. Appendix Ultrasound (U/S) Results by Impression with Percentage of Follow-Up Abdominal CT Imaging Total U/S Impression Diagnostic Positive Negative Total Nondiagnostic Borderline Not visualized Total
Follow-up CT
n
%
95% CI
n
%
63 51 114
14% 12% 26%
11–18 9–15 22–30
5 2 7
8% 4% 6%
1–15 1–9 2–11
22 305 327
5% 69% 74%
3–7 65–73 70–78
7 68 75
32% 22% 23%
12–52 18–27 18–27
CT = computed tomography; CI = confidence interval.
95% CI
appendix not visualized U/S. Of all the diagnostic U/S studies, 6% (95% CI 2–11) had follow-up CT scans, and of all the nondiagnostic U/S studies, 23% (95% CI 18–27) had follow-up CT scans (Table 1). The percentage difference of follow-up CT scans performed between diagnostic and nondiagnostic U/S studies is 17% (95% CI 10–24, p < 0.0001). Of the 7 patients with diagnostic U/S imaging who had follow-up CT imaging, appendicitis was found on CT in all of the patients who had a positive U/S and none of the patients who had a negative U/S. Of the 7 patients with borderline U/S imaging who had follow-up CT imaging, 71% (95% CI 35–108) were found to have appendicitis on CT and 29% (95% CI 8–65) were found not to have appendicitis. Of the 68 patients with no visualized appendix on U/S who had a follow-up CT, 22% (95% CI 12–32) were found to have appendicitis on CT, 71% (95% CI 60–81) were found not to have appendicitis, and 7% (95% CI 1–14) had borderline findings (Table 2). Additionally, of the patients with no visualized appendix on U/S, 7 (10%, 95% CI 3–18) did not have their appendices visualized on CT either. One of these 7 was suspected of having early appendicitis based on secondary findings, whereas the other 6 had no secondary signs of appendicitis and were presumed negative. DISCUSSION Pediatric patients with abdominal pain are common occurrences in the ED, and often require imaging to determine the need for surgical intervention. Our study found a high rate of nondiagnostic appendix U/S imaging, with a four-times greater rate of follow-up CT imaging compared with patients with diagnostic U/S imaging. A previous study of children and adults conducted at a community hospital that routinely performs U/S for evaluation of appendicitis reported a sensitivity of 83% and a specificity of 95% (21). However, this study enrolled patients based upon radiology reports containing the word ‘‘appendicitis’’ (21). This methodology does not account for U/S performed to rule out appendicitis when ‘‘appendicitis’’ is not explicitly mentioned, which could lead to fewer negative (‘‘normal appendix’’) or nondiagnostic (‘‘appendix not visualized’’) reports included, and thus falsely elevated test performance measures. Despite having lower test characteristics at community hospitals vs. academic centers, most agree that U/S should still be the first-line study when evaluating a pediatric patient with signs and symptoms of appendicitis–mainly due to risks of CT imaging, including concerns about radiation and intravenous contrast exposure. Although diagnostic accuracy can be high for both CT and U/S, CT is better at diagnosing negative findings and U/S is better at diagnosing positive findings (9). Many researchers have
4
S. M. Alter et al.
Table 2. Abdominal CT Imaging Results for Appendicitis by Initial Ultrasound (U/S) Impression CT Impression Positive U/S Impression Diagnostic Positive Negative Nondiagnostic Borderline Not visualized
n
n
%
5 2
5 0
100% 0%
7 68
5 15
71% 22%
Negative 95% CI
(35–108) (12–32)
n
%
0 2
0% 100%
2 48
29% 71%
Borderline 95% CI
8–65 60–81
n
%
0 0
0% 0%
0 5
0% 7%
95% CI
1–14
CT = computed tomography; CI = confidence interval.
proposed using an algorithm in which CT is utilized only if the initial U/S examination is indeterminate (20). Unfortunately, as our study demonstrates, many appendix U/S readings are indeterminate. There may be many factors for this, including imaging adequacy. Although all of our sonographers are certified by the American Registry for Diagnostic Medical Sonography, U/S is an operator-dependent test. Compared with sonographers in academic centers, community sonographers may not have as much experience performing pediatric ultrasounds. Therefore, the appendix may not be identified or only partially visualized. Other patient-related predictors of nondiagnostic U/S imaging include increased body mass index, retrocecal location, normal appendix, perforation, and distal tip inflammation (22). Another possible reason for increased indeterminate readings is variability in radiologist interpretations. Radiologists at community hospitals may have less experience reading pediatric ultrasounds when compared with academic radiologists with subspecialized training. At our hospital, there is only one board-certified pediatric radiologist who reads all the pediatric U/S performed on weekdays during normal business hours. At other times, these radiologic studies were interpreted by general radiologists. One way to attempt to mitigate interpretation variability is for the radiologist to use standardized interpretative categories. For example, by including presence or absence of secondary signs when the appendix was not visualized, diagnostic accuracy is improved to 96.8% when compared with 94.1% for a binary interpretative scheme (23). Some studies have created protocols in an attempt to limit CT use. Protocols utilizing CT only after indeterminate appendix U/S have been found to have higher sensitivities and specificities than U/S alone, and a low negative appendectomy rate (19). By following a clinical practice guideline, one children’s hospital was able to reduce CT imaging by 41% without loss of diagnostic accuracy (24). In cases where U/S is nondiagnostic with persistent suspicion of appendicitis, follow-up CT imaging is a reasonable next step in the diagnostic algorithm.
Although nonvisualization of the appendix traditionally has not ruled appendicitis in or out, previous research has suggested that a nondiagnostic U/S is still useful when there are associated ultrasonographic inflammatory markers. The specificity of moderate-to-large free fluid collections was 98%; phlegmon: 100%; pericecal inflammatory fat changes: 98%; and any free fluid with prominent lymph nodes: 81%; with multiple markers greatly increasing the odds ratio of appendicitis (25). In another study, 15.6% of patients with incompletely viewed appendices had laboratory-confirmed appendicitis, with secondary U/S signs having a sensitivity of 40% and specificity of 91% (26). These findings suggest that when utilizing U/S alone, patients with incompletely visualized appendices and presence of sonographic secondary signs most likely have appendicitis and do not need additional imaging. Our study showed similar trends among nondiagnostic U/S, with borderline studies having a greater rate of appendicitis diagnosed on follow-up CT (71% positive vs. 29% negative), whereas U/S studies with nonvisualized appendices tended to have a higher rate of negative CT scans (22% positive vs. 71% negative). Other protocols and clinical decision rules have been created to identify low-risk patients for discharge. Leeuwenburgh et al. determined that patients with negative or inconclusive U/S of the appendix and fewer than two predictors of male gender, migration of pain to the right lower quadrant, vomiting, and white blood cell count higher than 12 103/mL can be discharged home with next-day follow-up, without the need for additional imaging (27). This clinical decision rule had a 94% negative predictive value for appendicitis. Another study concluded that patients with a nondiagnostic U/S without leukocytosis do not require any further diagnostic imaging (28). Limitations As this was a retrospective study, we did have some limitations. First, during the study period, there was no
Ultrasound for Diagnosis of Appendicitis
standard protocol used by the treating emergency physicians in their clinical decision-making when ordering imaging studies. This lack of standardization may have led some physicians to order more CT imaging due to individual clinical practice. Second, although not directly examined in this study, radiology readings were frequently accompanied by a ‘‘disclaimer.’’ These statements, such as ‘‘if there is suspicion for appendicitis, please correlate clinically,’’ seem to impact subsequent emergency physicians’ ordering of CT imaging. This may represent a real-world effect of how ultrasound reports are written and acts as an effect modifier that needs further investigation. Third, our study did not have clinical or pathologic outcome of each patient’s medical course. Although a limitation of this study, our focus was on the impact of initial U/S test results and the subsequent utilization of CT scan testing. Finally, this study was not blinded. The investigators classifying radiology reports for appendicitis (‘‘positive,’’ ‘‘negative,’’ ‘‘borderline,’’ or ‘‘not visualized’’) were aware of the research questions of the study, and may have introduced bias. However, the investigators were not aware of which patients had follow-up CT imaging performed at the time of U/S radiology report review. CONCLUSIONS The utility of U/S to diagnose appendicitis at a community hospital seems to be limited by a very high rate of nondiagnostic studies. Even some patients with a diagnostic U/S had follow-up CT imaging, which were all in agreement with the initial U/S findings. When obtaining an appendix U/S, if it is diagnostic, there should be no need to obtain follow-up imaging. However, until U/S demonstrates improved and consistent test characteristics, it would be prudent to have a standardized approach, including alternative diagnostic tests for work-up of pediatric abdominal pain. REFERENCES 1. Rothrock SG, Pagane J. Acute appendicitis in children: emergency department diagnosis and management. Ann Emerg Med 2000;36: 39–51. 2. Bundy DG, Byerley JS, Liles EA, et al. Does this child have appendicitis? JAMA 2007;298:438–51. 3. Kwan KY, Nager AL. Diagnosing pediatric appendicitis: usefulness of laboratory markers. Am J Emerg Med 2010;28:1009–15. 4. Wang LT, Prentiss KA, Simon JZ, et al. The use of white blood cell count and left shift in the diagnosis of appendicitis in children. Pediatr Emerg Care 2007;23:69–76. 5. Alvarado A. A practical score for the early diagnosis of acute appendicitis. Ann Emerg Med 1986;15:557–64. 6. Apisarnthanarak P, Suvannarerg V, Pattaranutaporn P, et al. Alvarado score: can it reduce unnecessary CT scans for evaluation of acute appendicitis? Am J Emerg Med 2015;33:266–70.
5 7. Ceydeli A, Lavotshkin S, Yu J, et al. When should we order a CT scan and when should we rely on the results to diagnose an acute appendicitis? Curr Surg 2006;63:464–8. 8. 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. 9. Johansson EP, Rydh A, Riklund KA. Ultrasound, computed tomography, and laboratory findings in the diagnosis of appendicitis. Acta Radiol 2007;48:267–73. 10. Berg KJ. Nephrotoxicity related to contrast media. Scand J Urol Nephrol 2000;34:317–22. 11. Gleeson TG, Bulugahapitiya S. Contrast-induced nephropathy. AJR Am J Roentgenol 2004;183:1673–89. 12. Journy N, Ancelet S, Rehel JL, et al. Predicted cancer risks induced by computed tomography examinations during childhood, by a quantitative risk assessment approach. Radiat Environ Biophys 2014;53:39–54. 13. Ogbole GI. Radiation dose in paediatric computed tomography: risks and benefits. Ann Ib Postgrad Med 2010;8:118–26. 14. Brenner DJ. Estimating cancer risks from pediatric CT: going from the qualitative to the quantitative. Pediatr Radiol 2002;32:228–31. discussion 242-4. 15. Brenner D, Elliston C, Hall E, et al. Estimated risks of radiationinduced fatal cancer from pediatric CT. AJR Am J Roentgenol 2001;176:289–96. 16. Miglioretti DL, Johnson E, Williams A, et al. The use of computed tomography in pediatrics and the associated radiation exposure and estimated cancer risk. JAMA Pediatr 2013;167: 700–7. 17. Callahan MJ, Anandalwar SP, MacDougall RD, et al. Pediatric CT dose reduction for suspected appendicitis: a practice quality improvement project using artificial gaussian noise–part 2, clinical outcomes. AJR Am J Roentgenol 2015;204:636–44. 18. Kotagal M, Richards MK, Chapman T, et al. Improving ultrasound quality to reduce computed tomography use in pediatric appendicitis: the Safe and Sound campaign. Am J Surg 2015;209:896– 900. discussion 900. 19. Kaiser S, Frenckner B, Jorulf HK. Suspected appendicitis in children: US and CT—a prospective randomized study. Radiology 2002;223:633–8. 20. Teo EL, Tan KP, Lam SL, et al. Ultrasonography and computed tomography in a clinical algorithm for the evaluation of suspected acute appendicitis in children. Singapore Med J 2000; 41:387–92. 21. Chan I, Bicknell SG, Graham M. Utility and diagnostic accuracy of sonography in detecting appendicitis in a community hospital. AJR Am J Roentgenol 2005;184:1809–12. 22. Schuh S, Man C, Cheng A, et al. Predictors of non-diagnostic ultrasound scanning in children with suspected appendicitis. J Pediatr 2011;158:112–8. 23. Larson DB, Trout AT, Fierke SR, et al. Improvement in diagnostic accuracy of ultrasound of the pediatric appendix through the use of equivocal interpretive categories. AJR Am J Roentgenol 2015;204: 849–56. 24. Russell WS, Schuh AM, Hill JG, et al. Clinical practice guidelines for pediatric appendicitis evaluation can decrease computed tomography utilization while maintaining diagnostic accuracy. Pediatr Emerg Care 2013;29:568–73. 25. Estey A, Poonai N, Lim R. Appendix not seen: the predictive value of secondary inflammatory sonographic signs. Pediatr Emerg Care 2013;29:435–9. 26. Ross MJ, Liu H, Netherton SJ, et al. Outcomes of children with suspected appendicitis and incompletely visualized appendix on ultrasound. Acad Emerg Med 2014;21:538–42. 27. Leeuwenburgh MM, Stockmann HB, Bouma WH, et al. A simple clinical decision rule to rule out appendicitis in patients with nondiagnostic ultrasound results. Acad Emerg Med 2014;21: 488–96. 28. Cohen B, Bowling J, Midulla P, et al. The non-diagnostic ultrasound in appendicitis: is a non-visualized appendix the same as a negative study? J Pediatr Surg 2015;50:923–7.
6
S. M. Alter et al.
ARTICLE SUMMARY 1. Why is this topic important? Appendicitis is the most common pediatric surgical emergency, yet the commonly preferred initial imaging modality, ultrasound, does not seem to have great utility. Although previous studies have shown high accuracy, our experience has been different in a community emergency department. 2. What does this study attempt to show? This study demonstrates the low utility of ultrasound in the work-up of suspected pediatric appendicitis, and the rate of subsequent computed tomography (CT) scans performed. 3. What are the key findings? Pediatric ultrasound of the appendix was nondiagnostic 74% of the time, and patients who had nondiagnostic readings were almost four times more likely to have a subsequent abdominal CT scan. 4. How is patient care impacted? To avoid extraneous CT use and its associated risks, improvements should be made in the performance and acceptance of ultrasound imaging. Additionally, protocols should be developed and followed to limit CT use to the appropriate clinical settings for pediatric abdominal pain.