Back to 2014 Annual Meeting Abstracts
Use of White Blood Cell Count and Left Shift to Improve the Predictive Value of Ultrasound for Suspected Appendicitis in Children.
*Seema Anandalwar, *Michael Callahan, *Richard Bachur, *Christina Feng, *Feroze Sidhwa, *George Taylor, *Shawn Rangel
Boston Children's Hospital, Boston, MA
Objective: The objective of this study was to examine the utility of WBC count and polymorphonuclear leukocyte differential (PMN%) for improving the predictive value of ultrasound (US) in the diagnosis of appendicitis in children.
Design: Retrospective cohort study.
Setting: Single children’s hospital.
Patients: Children receiving an US in the Emergency Department for suspected appendicitis between 1/1/2010–12/31/2012 (n=873).
Main outcome measures: Negative and positive predictive values (NPV & PPV) for pathology-proven appendicitis were calculated for the five most commonly encountered constellations of US findings (table). Predictive values were compared with and without the use of WBC/PMN% thresholds (WBC>9x103/µL and PMN%>65 for PPV; WBC≤9x103/µL and PMN%≤65 for NPV).
Results: 58% (504/873) of US’s were considered “non-diagnostic” (categories 1&2), and NPV’s significantly increased using WBC/PMN% thresholds (category 1: 48.2% vs. 97.0%, p<0.001; category 2: 51.7% vs. 98.6%,p<0.001). 36% (311/873) of US’s were considered “moderately or strongly positive” (categories 4&5), and PPV’s were significantly increased using WBC/PMN% thresholds (category 4: 71.8% vs. 88.9%, p<0.001; category 5: 91.1% vs. 97.8%, p<0.001). The appendix was visualized as “normal” in 6.6% of cases, and the NPV improved from 57.7% to 100.0% using WBC/PMN% thresholds (p=0.003). Guidelines recommending against the use of CT for very high and low-risk categories (NPV>97% & PPV>97%) using WBC/PMN% thresholds could have reduced the number of CT’s by 43.2% (203/470) during the study period (table-shaded cells).
Conclusion: Use of WBC count and differential can significantly improve the negative and positive predictive value of ultrasound in the diagnosis of suspected appendicitis in children. | | | | | | | | | | Ultrasound diagnostic category | N
(% cohort) | US findings
ONLY | US findings
+ WBC/PMN% thresholds | CT utilization rate | | | PPV
(%) | NPV
(%) | PPV*
(%) | NPV**
(%) | WBC≤9x103/µL + PMN%≤65
N/total (%)*** | WBC>9x103/µL + PMN%>65
N/total (%)*** | All patients
N/total (%) | 1.
Appendix not visualized, no primary1 or secondary2 signs
(prevalence of appendicitis = 16.5%) | 275 (31.5) | 16.4 | 48.2 | 26.0 | 97.0 | 103/131 (78.6) | 48/66 (72.7) | 208/275 (75.6) | 2.
Appendix not visualized, fluid present in RLQ or pelvis, no primary or secondary signs
(prevalence of appendicitis = 19.2%) | 229 (26.2) | 19.2 | 51.7 | 35.8 | 98.6 | 85/106 (80.2) | 47/69 (68.1) | 174/229 (76.0) | 3.
Normal appendix visualized, no primary or secondary signs
(prevalence of appendicitis = 17.2%) | 58 (6.6) | 17.2 | 57.7 | 33.3 | 100.0 | 11/21 (52.4) | 12/22 (54.5) | 30/58 (51.7) | 4.
Appendix visualized with primary signs only
(prevalence of appendicitis = 71.8%) | 142 (16.3) | 71.8 | 65.4 | 88.9 | 76.2 | 19/90 (21.2) | 6/21 (28.6) | 33/142 (23.2) | 5.
Secondary signs present +/- visualized appendix +/- primary signs
(prevalence of appendicitis = 91.1%) | 169 (19.4) | 91.1 | 71.4 | 97.8 | 53.3 | 20/128 (15.6) | 4/15 (26.7) | 25/169 (14.8) | | 1Primary signs of appendicitis= hyperemia of appendiceal wall or wall thickness>7mm. 2Secondary signs of appendicitis= fecalith or echogenic fat. *PPV= Positive predictive value for WBC>9K and PMNs>65%; **NPV= Negative predictive value for WBC≤9K and PMNs≤65%. ***The denominators in these two columns reflect the number of patients where both WBC and PMN% thresholds were met (not just one) whereas the final “all patients” column includes the total number of patients in a particular US category. |
Back to 2014 Annual Meeting Abstracts
|
|
