Van den Broek et al evaluated children with acute appendicitis. They developed a model using 3 parameters to determine management. The authors are from Medical Centre Alkmaar, RodeKruis Ziekenhuis and the Academic Medical Center Amsterdam in The Netherlands.
Patient selection: mean age 9 (standard deviation 2 years)
Parameters:
(1) rebound tenderness
(2) leukocyte (white blood cell) count
(3) temperature in °C
|
Parameter |
Finding |
Points |
|
rebound tenderness |
absent |
0 |
|
|
present |
2 |
|
leukocyte count |
< 10,000 per µL |
0 |
|
|
>= 10,000 per µL |
2 |
|
temperature in °C |
< 38°C |
0 |
|
|
>= 38°C |
1 |
total score =
= SUM(points for all 3 parameters)
Interpretation:
• minimum score: 0
• maximum score: 5
• A score of 0 or 1 indicates a low risk of appendicitis and observation is appropriate.
• A score of 4 or 5 indicates a high risk for appendicitis.
• A score of 2 or 3 has an uncertain risk.
• The prevalence of acute appendicitis affects the accuracy of the prediction, which suggests a Bayesian model might work with this data.
• Using the model for all patients rather than just those with suspected appendicitis results in a higher rate of negative appendectomies. This would reduce the prevalence of acute appendicitis, once again suggesting a Bayesian model.
• Temperature has little discriminatory impact, and the model can be simplified to the table below.
|
Rebound Tenderness |
Leukocyte Count |
Management |
|
absent |
< 10,000 per µL |
observe |
|
absent |
>= 10,000 per µL |
diagnostic laparoscopy |
|
present |
< 10,000 per µL |
diagnostic laparoscopy |
|
present |
>= 10,000 per µL |
appendectomy |
Specialty: Gastroenterology, Pedatrics, Surgery, general
