Clinical value of Leukocyte counts in evaluation of patients with suspected appendicitis in emergency department

Abstract

The objective of this study was to assess the value of total leukocyte count in patients presenting to the emergency department with sign and symptoms suggestive of acute appendicitis.
Patient demographics, presenting sign, and symptoms, initials total leukocyte count and discharge diagnosis was documented retrospectively. Admitted patients were followed for surgical and clinical outcomes, and discharged patients followed up by telephonic interview.
Out of 116 eligible patients 48 patients was found to have acute appendicitis. Total leukocyte count >=10.000 cell/mm3 yielded a sensitivity of 89.58% and a specificity of 64.71%.The positive predictive value (PPV) was 64.18%, and the negative predictive value (NPV) was 89.80%.The positive likelihood ratio (+LR) was 2.54, and the negative likelihood ration (-LR ) was 0.16. The area under the curve for the receiver operating characteristic (ROC) curve was 0.822, which was moderately accurate.
The total leukocyte counts are helpful in the diagnosis and exclusion of acute appendicitis. The elevated leukocyte count >=10,000 cells/mm3 was statistically associated with the presence of acute appendicitis. The ROC curve suggests there is value of total leukocyte counts and has sufficient sensitivity and specificity to be clinical value in the diagnosis of acute appendicitis.

Full text

1
Turkish Journal of Trauma & Emergency Surgery
Original Article Klinik Çalışma
Ulus Travma Acil Cerrahi Derg 2012;18 (6):000-000
Clinical value of leukocyte counts in evaluation of patients with
suspected appendicitis in emergency department
Acil serviste apandisitten kuşkulanılan hastaların değerlendirilmesinde
lökosit sayımlarının klinik değeri
Mohammad Zikrullah TAMANNA,1 Uzma ERAM,2 Turki Mohsen AL HARBI,1
Saleh Awad ALRASHDI,1 Shafkat Ullah KHATEEB,1
Saleem Ahmed ALADHRAI,1 Abdul Muthalib HUSSAIN3
1Department of Emergency Medicine, King Fahad Medical City, Riyadh,
Saudia Arabia; 2Department of Community Medicine, J.N Medical Collage,
AMU Aligarh, India; 3Department of Pulmonary and Critical Care
Medicine, King Fahad Medical City, Riyadh, Saudia Arabia.
1Kral Fahd Tıp Kenti, Acil Tıp Bölümü, Riyad, Suudi Arabistan;
2JN Tıp Koleji, Toplum Hekimliği Bölümü, AMU Aligarh, Hindistan;
3Kral Fahd Tıp Kenti, Göğüs Hastalıkları ve Yoğun Bakım Kliniği,
Riyad, Suudi Arabistan.
Correspondence (İletişim): Mohammad Zikrullah Tamanna, M.D. Post Box 59046, Riyadh, Saudia Arabia.
Tel: +90 - 0096 656 767 30 81 e-mail (e-posta): zikitam@gmail.com
BACKGROUND
The objective of this study was to assess the value of to-
tal leukocyte count in patients presenting to the emergency
department with sign and symptoms suggestive of acute
appendicitis.
METHODS
Patient demographics, presenting sign, and symptoms, ini-
tials total leukocyte count and discharge diagnosis was do-
cumented retrospectively. Admitted patients were followed
for surgical and clinical outcomes, and discharged patients
followed up by telephonic interview.
RESULTS
Out of 116 eligible patients 48 patients was found to have
acute appendicitis. Total leukocyte count ≥10.000 cell/mm3
yielded a sensitivity of 89.58% and a specicity of 64.71%.
The positive predictive value (PPV) was 64.18%, and the
negative predictive value (NPV) was 89.80%.The positive
likelihood ratio (+LR) was 2.54, and the negative likeliho-
od ration (-LR ) was 0.16. The area under the curve for the
receiver operating characteristic (ROC) curve was 0.822,
which was moderately accurate.
CONCLUSION
The total leukocyte counts are helpful in the diagnosis and
exclusion of acute appendicitis. The elevated leukocyte co-
unt ≥10,000 cells/mm3 was statistically associated with the
presence of acute appendicitis. The ROC curve suggests
there is value of total leukocyte counts and has sufcient
sensitivity and specicity to be clinical value in the diagno-
sis of acute appendicitis.
Key Words: Abdominal pain; appendicitis; leukocyte counts.
AMAÇ
Bu çalışmanın amacı acil servise akut apandisiti düşündü-
ren belirti ve bulgularla gelen hastalarda toplam lökosit sa-
yımının değerini belirlemekti.
GEREÇ VE YÖNTEM
Hastanın demograk özellikleri, görülen belirtiler ve bul-
guları, toplam lökosit sayıları ve hastaneden çıkıştaki tanı-
ları geriye dönük olarak belgelendi. Kabul edilen hastalar
cerrahi ve klinik sonuçlar açısından izlendi, çıkışı yapılan
hastalar telefon görüşmeleriyle takip edildi.
BULGULAR
Yüz on altı hastanın 48’inde akut apandisit saptantı. Top-
lam ≥10,000 /mm3 şeklindeki lökosit sayısı %89,58 duyar-
lılık ve %64,71 özgüllüğe sahipti. Pozitif (PPV) ve negatif
(NPV) öngörü değerleri sırasıyla %64,18 ve %89,80 idi.
Pozitif (+LR) ve negatif (-LR) olasılık oranları ise sıra-
sıyla 2,54 ve 0,16 şeklindeydi. Algılayıcı işletim eğrisi-
nin (ROC-receiver operating characteristics) altında kalan
alanı 0,822 olup orta derecede doğruluk derecesini sergi-
lemekteydi.
SONUÇ
Akut apandisit tanısı koyma veya bu tanının dışlanmasın-
da toplam lökosit sayıları yardımcı olur. Yüksek lökosit
sayısı (≥10.000/mm3 ) istatistiksel açıdan akut apandisit
varlığıyla ilişkiliydi. ROC eğrisi akut apandisit tanısında
toplam lökosit sayılarının klinik değer taşıyacak şekilde
yeterli duyarlılık ve özgüllüğe sahip olduğunu düşündür-
mektedir.
Anahtar Sözcükler: Karın ağrısı; apandisit; lökosit sayıları.
doi: 10.5505/tjtes.2012.83652

Ulus Travma Acil Cerrahi Derg
Acute appendicitis is a common consideration in
the differential diagnosis of patients with right lower
abdominal pain presenting to the emergency depart-
ment (ED). Both routine history and physical exami-
nation remain the most effective and practical diag-
nostic modalities. To increase diagnostic accuracy,
inammatory variables such as leukocyte count, C-
reactive protein (CRP), and body temperature have
been studied.[1,2,] None of these is specic for acute
appendicitis, and their ability to improve on the pre-
operative diagnosis is doubtful. The prevailing per-
ception is that the patients with appendicitis will have
a high leukocyte count. Most of the ED physicians,
while evaluating acute appendicitis, obtain a complete
blood cell count (CBC), with the expectation that a
high leukocyte count will support the diagnosis. Con-
sidering the overlap with other inammatory process-
es mimicking acute appendicitis, the diagnostic value
of the leukocyte count remains controversial. Several
literatures recommend total leukocyte count as part
of the evaluation with suspected appendicitis.[3] How-
ever, some studies do not recommend total leukocyte
count because of the limitation of this test.[4] Most of
the studies have calculated the sensitivity and speci-
city at arbitrarily chosen threshold values for acute
leukocyte count. Only a few have used receiver oper-
ating characteristic (ROC) curve analysis in patients
with suspected acute appendicitis, but either without
calculating the area under the ROC curve (AUC) or
without dening the cut-off values with highest sen-
sitivity and specicity.[5,6] However, in this continuous
variable, ROC curve analysis is a statistically more ap-
propriate approach.
In this study, we tried to evaluate retrospectively
the diagnostic value of total leukocyte count by per-
forming ROC curve analysis in consecutive patients
presenting to an ED with signs and symptoms sugges-
tive of acute appendicitis.
MATERIALS AND METHODS
This retrospective study was conducted in a 1500-
bed tertiary care teaching hospital in Saudi Arabia.
The institutional review board approval was obtained.
The present investigation included patients who pre-
sented to the ED between June 2011 and January 2012
with right lower quadrant pain who were suspected
as having acute appendicitis. A total of 159 patient
charts that met the inclusion criteria were reviewed.
The study was designed to assess the impact of Al-
varado score and computed tomography (CT) scan in
the diagnosis of acute appendicitis. This investigation
represents a planned subanalysis of the total leukocyte
count obtained as part of routine data collected from
the enrolled patients.
Patient demographics and presenting signs and
symptoms were documented from the pro forma. All
laboratory tests were obtained at the discretion of the
managing physicians. If appendectomy or surgical in-
tervention was performed, all appendices and other
specimens were submitted to the pathology depart-
ment. Ofcial radiology reports, the surgical pathology
report, and medical records were reviewed retrospec-
tively. Patients were excluded if lost to follow-up or
transferred to another hospital. Other alternative diag-
noses were recorded. If acute appendicitis was consid-
ered unlikely and no other acute etiology was found,
patients were discharged with discharge instruction to
return if symptoms worsened, with appropriate fol-
low-up. All patients discharged from the ED were fol-
lowed up by telephone interview after the index visit.
Recorded data included age, gender, total leuko-
cyte counts from the CBC count obtained during the
initial ED presentation, presence of appendicitis (op-
erative and pathologic report), and nal diagnosis. To-
tal leukocyte counts or white blood cell count (WBC)
>10,000 cells/mm3 were a priori classied as abnor-
mal. Total leukocyte counts of up to 10,000 cells/mm3
were chosen empirically to maximize sensitivity.
Data analysis
Analysis of the data on total leukocyte counts in-
cluded determination of sensitivities, specicities,
positive predictive value (PPV), negative predictive
value (NPV), positive likelihood ratio (+LR), and
negative likelihood ratio (-LR). All results were re-
ported with 95% condence intervals (95% CIs). ROC
curves were plotted for these variables along with cal-
culation of the AUC with 95% CI. We calculated the
AUC for leukocyte counts using the Hanley-McNeil
method for paired data.[7] The AUC is a summary sta-
tistic of diagnostic accuracy (based on all possible cut-
off values). The AUC can have values between 0 and
1. The practical lower limit for the AUC of a diagnos-
tic test is 0.5. The test results are classied as unin-
formative (AUC=0.5), less accurate (0.5<AUC≤0.7),
moderately accurate (0.7<AUC≤0.9), highly accurate
(0.9<AUC<1), and perfect (AUC=1).[8] To calculate
sensitivity and specicity, we used cut-off values for
the leukocyte counts. The selected cut-off value pro-
vided an ideal balance between sensitivity and speci-
city for the diagnosis of acute appendicitis. Clinical
and analytical variables of the acute appendicitis and
normal appendix group were compared using chi-
square test (Table 1). Statistical analyses were per-
formed using MedCalc for Windows version 8.2.1.0.
RESULTS
The study began with retrieval of 159 patients who
presented to the ED with right lower quadrant pain.
Forty-three were excluded because they were trans-
ferred to another hospital after being diagnosed as ap-
2

Clinical value of leukocyte counts in evaluation of patients with suspected appendicitis in emergency department
pendicitis due to non-availability of beds in the surgi-
cal ward. Two were discharged against medical advice,
and these two patients were lost to follow-up. One hun-
dred and sixteen patients were included in the study.
Total leukocyte count was measured in 116 pa-
tients. Table 2 shows a standard 2×2 table for total leu-
kocyte count in patients with and without appendici-
tis. Total leukocyte count ≥10,000 cells/mm3 yielded a
sensitivity of 89.58% (95% CI: 77.34%-96.53%) and
a specicity of 64.71% (95% CI: 52.17%-75.92%).
The PPV was 64.18% (95%CI: 51.53%-75.3%) and
the NPV was 89.80% (95% CI=77.77%-96.60%). The
+LR was 2.54 (95% CI: 1.81-3.55) and the -LR was
0.16 (95% CI: 0.07-0.38) (Table 3).
Table 4 shows the diagnosis of acute appendicitis
for different ranges of total leukocyte counts, which
3
was statistically signicant (p<0.0001).
Table 5 shows an ROC curve for total leukocyte
counts. The AUC for the ROC curve was 0.822 (95%
CI: 0.740-0.887), which was moderately accurate. If
one chose to maximize both sensitivity and specic-
ity, an optimum cut-off for an abnormal total leukocyte
count would be 10,700 cells/mm3, which yielded a sen-
sitivity of 89.6% and a specicity of 73.5% (Fig. 1).
DISCUSSION
Inammatory markers have long been used to im-
prove the diagnostic performance in acute appendici-
tis. Measurement of total leukocyte count is usually
considered a routine part of the work-up for acute
appendicitis. Several studies suggest that an elevated
total leukocyte count is usually the earliest laboratory
nding of appendiceal inammation.[1,2,6,9] Neither
leukocyte count nor neutrophil percentage is specic
for acute appendicitis. Leukocytosis is a nonspecic
reaction caused by acute or chronic inammation,
acute physical or emotional stress, and several other
conditions. Most of the previous studies have tried
to compare the relationship between total leukocyte
count and acute appendicitis using sensitivity, speci-
city, PPV, and NPV using arbitrary cut-off points
with the upper reference limit.[1,2,6,7,10,11] The difculty
with this type of analysis is that PPV and NPV are
directly dependent on the prevalence of the disease.
As these variables are obtained from various study
Table 2. Total leukocyte count and appendicitis
Count Diagnosis
Total WBC Acute No
count appendicitis appendicitis Total
WBC ≥10,000 43 24 67
<10,000 5 44 49
Total 48 68 116
A 2×2 table for patients with and without appendicitis. For the purpose of this
analysis, a WBC (leukocyte count) ≥10,000 cells/mm3 was considered abnormal.
Table 1. Chi-square tests
Value Df Asymp. Sig. Exact Sig. Exact Sig.
(2-sided) (2-sided) (2-sided)
Pearson chi-square 33.991a 1 .000
Continuity correctionb 31.802 1 .000
Likelihood ratio 37.630 1 .000
Fisher’s exact test .000 .000
N of valid cases 116
a0 cells (.0%) have expected count less than 5. The minimum expected count is 20.28.
bComputed only for a 2x2 table.
Table 3. Sensitivities, specicities, positive predictive values (PPV), negative predictive
values (NPV), positive likelihood ratios (+LR), and negative likelihood ratios
(-LR) with 95% condence intervals (95% CIs) at the cut-off value of total
leukocyte count ≥10,000 cells/mm3
Sensitivity 89.58% 77.34% - 96.53%
Specicity 64.71% 52.17% - 75.92%
Positive likelihood ratio 2.54 1.81 - 3.55
Negative likelihood ratio 0.16 0.07 - 0.38
Disease prevalence 41.38% 32.31% - 50.90%
Positive predictive value 64.18% 51.53% - 75.53%
Negative predictive value 89.80% 77.77% - 96.60%

Ulus Travma Acil Cerrahi Derg
4
Our present study has several limitations due to in-
herent aws of a retrospective study. Our study com-
prised a review of patients’ medical records and patho-
logical reports. The retrospective nature of the study
may limit the diagnostic performance, and was limited
to one institution. Another limitation was that we did
not study the diagnostic value of neutrophils. We also
did not study how total leukocyte counts might per-
form as part of a scoring system or combination of test
results.
There is a statistically signicant relationship be-
tween total leukocyte count and acute appendicitis in
ED patients presenting with signs and symptoms sug-
gestive of acute appendicitis. According to our study,
the relationship is modest and clinically very useful.
There is need to perform more investigations to pro-
vide further information for the diagnosis of acute ap-
pendicitis, such as imaging, short-term observation for
clinical evaluation, and repeated physical and labora-
populations with different prevalences of the disease,
the results have to be interpreted carefully. Moreover,
sensitivity and specicity alone do not allow clinicians
to directly apply the result of diagnostic tests to in-
dividual patients. ROC analysis allows calculation of
cut-off values with the highest sensitivity and specic-
ity. ROC analysis, however, is independent of disease
prevalence.
Only a few studies have done ROC analysis of in-
ammatory variables in acute appendicitis.[6,12] How-
ever, in some, only the appearance of the ROC curve
was used or the AUC was calculated without report-
ing the cut-off value with the highest sensitivity and
specicity.
In our study, while doing ROC analysis, the AUC
for the ROC curve was 0.822 (95% CI: 0.740-0.887),
which was moderately accurate. If one chose to maxi-
mize both sensitivity and specicity, an optimum
cut-off for an abnormal total leukocyte count would
be 10,700 cells/mm3, which yielded a sensitivity of
89.6% and a specicity of 73.5% (Fig. 1).
Other investigators have constructed ROC analysis
with comparable results. Andersson et al.[13] found an
AUC of 0.80, Paajanen et al.[6] found an AUC of 0.76,
Rodriguez-Sanjuan et al.[14] found an AUC of 0.67,
and Körner et al.[15] found an AUC of 0.69 (95% CI:
0.65-0.73).
Table 5. Receiver operating characteristic curve
analysis of total leukocyte count and area
under curve with 95% condence interval
Variable Count
Classication variable Diagnosis
Sample size 116
Positive group: diagnosis = 1 48
Negative group: diagnosis = 0 68
Disease prevalence (%) Unknown
Area under the ROC curve (AUC) 0.822
Standard Errora 0.0398
95% Condence intervalb 0.740 to 0.887
z statistic 8.092
Signicance level P (Area=0.5) <0.0001
aHanley & McNeil, 1982; bBinomial exact;
Criterion values and coordinates of the ROC curve;
Criterion corresponding with highest Youden index.
100
80
60
40
20
0
0 20 40
100-Specicity
count
Sensitivity
60 80 100
Sensitivity: 89.6
Specicity: 73.5
Criterion: >10.7
Fig. 1. ROC curve for total leukocyte counts. The AUC for
the ROC curve was 0.822 (95% CI: 0.740-0.887),
which was moderately accurate. If one chose to maxi-
mize both sensitivity and specicity, an optimum
cut-off for an abnormal total leukocyte count would
be 10,700 cells/mm3, which yielded a sensitivity of
89.6% and a specicity of 73.5%.
Table 4. Acute appendicitis for different ranges of total leukocyte counts
Diagnosis <8 8 to <10 10 to <12 12 to <15 ≥15
Appendicitis (48) 1 4 8 14 21
No appendicitis (68) 23 21 9 9 6
Total (116) 24 25 17 23 27
Chi-square 38.914; DF 4; Signicance level p<0.0001; Contingency coefcient 0.501.

Eur J Clin Chem Clin Biochem 1991;29:749-52.
6. Paajanen H, Mansikka A, Laato M, Kettunen J, Kostiainen
S. Are serum inammatory markers age dependent in acute
appendicitis? J Am Coll Surg 1997;184:303-8.
7. Hanley JA, McNeil BJ. The meaning and use of the area un-
der a receiver operating characteristic (ROC) curve. Radiol-
ogy 1982;143:29-36.
8. Greiner M, Pfeiffer D, Smith RD. Principles and practical
application of the receiver-operating characteristic analysis
for diagnostic tests. Prev Vet Med 2000;45:23-41.
9. Sasso RD, Hanna EA, Moore DL. Leukocytic and neutro-
philic counts in acute appendicitis. Am J Surg 1970;120:563-
6.
10. Bolton JP, Craven ER, Croft RJ, Menzies-Gow N. An assess-
ment of the value of the white cell count in the management
of suspected acute appendicitis. Br J Surg 1975;62:906-8.
11. Lau WY, Ho YC, Chu KW, Yeung C. Leucocyte count and
neutrophil percentage in appendicectomy for suspected ap-
pendicitis. Aust N Z J Surg 1989;59:395-8.
12. van Dieijen-Visser MP, Go PM, Brombacher PJ. The value
of laboratory tests in patients suspected of acute appendicitis.
Eur J Clin Chem Clin Biochem 1991;29:749-52.
13. Andersson RE, Hugander AP, Ghazi SH, Ravn H, Offenbartl
SK, Nyström PO, et al. Diagnostic value of disease history,
clinical presentation, and inammatory parameters of appen-
dicitis. World J Surg 1999;23:133-40.
14. Rodríguez-Sanjuán JC, Martín-Parra JI, Seco I, García-Cas-
trillo L, Naranjo A. C-reactive protein and leukocyte count
in the diagnosis of acute appendicitis in children. Dis Colon
Rectum 1999;42:1325-9.
15. Körner H, Söreide JA, Söndenaa K. Diagnostic accuracy of
inammatory markers in patients operated on for suspected
acute appendicitis: a receiver operating characteristic curve
analysis. Eur J Surg 1999;165:679-85.
Clinical value of leukocyte counts in evaluation of patients with suspected appendicitis in emergency department
5
tory examinations. Our ndings indicate that the total
leukocyte count offers a rapid and helpful method to
predict and discriminate appendicitis in the ED.
In conclusion, although elevated leukocyte count
is a nonspecic inammatory marker, high leukocyte
count is helpful in the diagnosis and exclusion of ap-
pendicitis. It is inexpensive, objective, and readily
available without the risk of radiation, and useful for
the emergency physician.
Acknowledgements
The author thanks Dr. Abdul Muthalib Hussain for
inspiring and giving proper guidance in writing this
article.
REFERENCES
1. Peltola H, Ahlqvist J, Rapola J, Räsänen J, Louhimo I,
Saarinen M, et al. C-reactive protein compared with white
blood cell count and erythrocyte sedimentation rate in the
diagnosis of acute appendicitis in children. Acta Chir Scand
1986;152:55-8.
2. Elangovan S. Clinical and laboratory ndings in acute ap-
pendicitis in the elderly. J Am Board Fam Pract 1996;9:75-8.
3. Kozar RA, Roslyn JJ. The appendix. In: Schwartz SI, Shires
SG, Spencer FC, Daley JM, Fischer JE, GallowayAC, edi-
tors. Principles of surgery. 7th ed. New York, NY: McGraw
Hill; 1999. p. 1383-95.
4. Hartman GE. Acute appendicitis. In: Behrman RE, Klieg-
man RM, Jenson HB, editors. Nelson Textbook of Pediatrics.
Phildephia, PA: WB Saunders Company; 2000. p. 1179.
5. van Dieijen-Visser MP, Go PM, Brombacher PJ. The value
of laboratory tests in patients suspected of acute appendicitis.