Graphical representation of lymphocyte–white blood cell count (L/WCC) ratios in all patients. All patients with tonsillitis fall below the L/WCC ratio of 0.35.
Histogram of frequency of lymphocyte–white blood cell count (L/WCC) ratios in patients with glandular fever and tonsillitis. It shows a clear separation in distribution of the ratios in the 2 illnesses.
Suggested flowchart to differentiate patients with glandular fever from those with tonsillitis. LFTs indicate liver function tests; L/WCC, lymphocyte–white blood cell count ratio; MST, mononucleosis spot test.
Wolf DM, Friedrichs I, Toma AG. Lymphocyte–White Blood Cell Count RatioA Quickly Available Screening Tool to Differentiate Acute Purulent Tonsillitis From Glandular Fever. Arch Otolaryngol Head Neck Surg. 2007;133(1):61-64. doi:10.1001/archotol.133.1.61
To find a quickly available screening tool for the differentiation of patients with glandular fever from those with acute purulent tonsillitis. The null hypothesis was that there was no difference between the lymphocyte–white blood cell count (L/WCC) ratio between the 2 patient groups.
Retrospective pilot study based on laboratory tests for lymphocyte counts, white blood cell counts, and the mononucleosis spot test.
Ear, Nose, and Throat Department, St George's Hospital, London, England.
One hundred twenty patients with glandular fever and 100 patients with bacterial tonsillitis.
Main Outcome Measures
Results from the mononucleosis spot test in conjunction with the clinical picture and the L/WCC ratio were analyzed. Significant differences were evaluated using the Mann-Whitney test and Fisher exact test.
The L/WCC ratio was significantly different in the 2 groups (P<.001). The mean L/WCC ratio in the glandular fever group was 0.54 and the mean L/WCC ratio in the bacterial tonsillitis group was 0.10. A ratio higher than 0.35 had a specificity of 100% and a sensitivity of 90% for the detection of glandular fever.
We recommend that the L/WCC ratio should be used as an indicator to decide whether mononucleosis spot tests should be requested. A ratio higher than 0.35 had a high specificity in our study group.
Acute tonsillitis, pharyngitis, and glandular fever are among the most common causes for ear, nose, and throat (ENT) admissions in hospitals in the United Kingdom. In 2003 to 2004, 24 221 patients were admitted with acute tonsillitis in England alone. The mean (median) length of stay for acute tonsillitis in England in 2003 to 2004 was 1.3 (1) days.1 A total of 30 708 bed days were funded.
Clinical presentation and medical history often make it difficult to differentiate these similar conditions. Early distinction between infectious mononucleosis and bacterial tonsillitis may alter treatment regimes, and this again could have certain cost benefits.
The aim of this study was to find a quickly available method of distinguishing patients with bacterial tonsillitis from those with infectious mononucleosis. This would prevent unnecessary requests for random expensive mononucleosis spot tests and facilitate the use of appropriate treatment regimes resulting in possible shorter hospital stays.
Results of the mononucleosis spot test in conjunction with the clinical picture and the lymphocyte–white blood cell count (L/WCC) ratio were analyzed in 120 patients with glandular fever (hereafter, glandular fever group) and 100 patients with bacterial tonsillitis (hereafter, tonsillitis group) attending the ENT Department of St George's Hospital, London, England. Symptoms of patients with tonsillitis or glandular fever included sore throat, pyrexia, difficulties in swallowing food, pain with swallowing, redness of throat and tonsils, and white plaques on the tonsils. Patients with quinsy, parapharyngeal or retropharyngeal abscesses, and compromised immune systems as well as patients with peritonsillitis were excluded from this study. To detect glandular fever we used the Monolatex test (Biokit SA, Barcelona, Spain), which detects heterophile antibodies in serum or plasma. The reagent is a suspension of latex particles coated with Paul-Bunnell antigen from bovine red cell membranes for a better detection of the agglutination reaction. Because of the high purity of the antigen, the reagent reacts only with infectious mononucleosis (IM) heterophile antibodies showing a clear agglutination in the positive samples. Latex particles allow visual observation of the antigen-antibody reaction. If IM heterophile antibodies are present in the sample, the latex suspension changes its uniform appearance, and a clear agglutination becomes evident. The Monolatex test kit contains latex reagent, positive control, negative control, and disposable slides. Fifty centiliters of undiluted serum sample are placed on the test slide, and 1 drop of reagent is added. The sample is then mixed and rotated for 3 minutes and read qualitatively and semiquantitatively. All mononucleosis spot tests were performed in the hematology laboratory at St George's Hospital following the standard protocol. The blood samples of all patients were analyzed with a Coulter counter (Gen S Series II; Beckman Coulter Inc, Fullerton, Calif) using flow cytometry following a standard protocol. These tests were also performed in the hematology laboratory of St George's Hospital.
The data were collected and collated using Excel software (Microsoft, Redmond, Wash). Significant differences were evaluated using the Mann-Whitney test and Fisher exact test (SPSS Inc, Chicago, Ill).
Both study groups were similarly matched, and there was no statistical significance in the distribution of male and female patients in the cohort (Table 1).
However, there was a statistically significant variation in age between the 2 groups, which was expected considering the known epidemiological data.2 Analyzing our data, we found that the mean (SD) total white blood cell count was significantly raised (P<.001) in the tonsillitis group (16 560/μL [54 100/μL]) compared with the glandular fever group (11 400/μL [4670/μL]). The lymphocyte count was also significantly higher (P<.001) in the glandular fever group (6490/μL [3590/μL]) compared with the tonsillitis group (1590/μL [680/μL]) (Table 2). In addition to these findings, the neutrophil count was increased (P<.001) in the tonsillitis group (13 770/μL [5230/μL]) compared with the glandular fever group (3830/μL [1920/μL]) (Table 2). Calculating the L/WCC ratios in both groups, we found that the mean ratio was significantly higher (P<.001) in the glandular fever group (0.54 [0.14]) compared with the tonsillitis group (0.10 [0.08]) (Figure 1). A ratio higher than 0.35 would have a sensitivity of 90% and a specificity of 100% for the detection of glandular fever (Figure 2).
Differential blood tests are routine investigations for hospital admissions, and the test results are available within hours of hospital admission. In contrast, the mononucleosis spot test is done only once a day with the Monolatex test kit in batches at St George's Hospital. Therefore, it can take up to 48 hours after admission for the results to become available, which produces additional costs for the hospital.
The evaluation of the reliability of 6 commercially available kits for the rapid diagnosis of IM in comparison with Epstein-Barr virus (EBV)-specific serologic tests showed sensitivities and specificities of 70% to 92% and 96% to 100%.3 In another study by Linderholm et al,4 the sensitivities and specificities of 9 rapid kits varied from 63% to 84% and 84% to 100%.
Our data suggest that our test based on an L/WCC ratio higher than 0.35 had a sensitivity of 90% and a specificity of 100% for the detection of glandular fever. The specificity and sensitivity of this test seem to be better than the mononucleosis spot test itself. The mononucleosis spot test can produce false-negative results, and the true sensitivity of this test should be determined by EBV serologic tests. The importance in differentiating patients with tonsillitis from those with glandular fever is the prevention of spontaneous rupture of the spleen and acute intra-abdominal hemorrhage. Splenomegaly is common, but splenic rupture is very rare, occurring in 0.1% to 0.5% of all patients; however, it remains the most common lethal complication of IM.5,6 The convalescence period for patients with glandular fever tends to be longer than that of patients with tonsillitis. Patients with glandular fever should also have further follow-up with liver function tests and possible ultrasonography of the liver or spleen and should avoid minor injuries to the abdomen, contact sports, and alcohol to prevent life-threatening complications.7
Further prospective studies, including EBV serologic tests, may determine the exact sensitivity and specificity of our easily available test for differentiation of glandular fever from bacterial tonsillitis. However, we think that this can be confidently used to identify patients who should undergo further testing for the diagnosis of glandular fever (in particular, the mononucleosis spot test) to avoid unnecessary stress on limited laboratory resources (Figure 3).
In conclusion, we recommend that the L/WCC ratio should be used as an indicator to decide whether mononucleosis spot tests are required. Results from our retrospective pilot study suggest that the L/WCC ratio could be a quickly available alternative test for the detection of glandular fever. A ratio higher than 0.35 had a high specificity in our study group. However, further studies are needed to establish the true sensitivity and specificity of this test.
Correspondence: Dennis M. Wolf, BSc, DO-HNS, MRCS, Department of Otolaryngology, St George's Hospital, London SW17 0QT, England (email@example.com).
Submitted for Publication: May 11, 2006; accepted August 20, 2006.
Author Contributions: Drs Wolf, Friedrichs, and Toma had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Wolf and Friedrichs. Analysis and interpretation of data: Friedrichs and Toma. Drafting of the manuscript: Wolf, Friedrichs, and Toma. Critical revision of the manuscript for important intellectual content: Friedrichs and Toma. Study supervision: Friedrichs and Toma.
Financial Disclosure: None reported.