Date Published: May 1, 2019
Publisher: Public Library of Science
Author(s): Choong Shi Hui Clarice, Visula Abeysuriya, Sanjay de Mel, Basuru Uvindu Thilakawardana, Primesh de Mel, Chandima de Mel, Lal Chandrasena, Suranjith L. Seneviratne, Christina Yip, Eng Soo Yap, Dermot Cox.
The early identification of patients at risk of severe dengue infection (DI) is critical to guide clinical management. There is currently no validated laboratory test which can predict severe complications of DI. The Atypical lymphocyte count (ALC) is a research parameter generated at no extra cost when an automated Full Blood Count (FBC) is performed. The purpose of this study was to assess the association of ALC with the severity of DI.
We prospectively collected data on patients admitted to Nawaloka Hospital Sri Lanka (NH) with DI between December 2016 and November 2017. DI was diagnosed based on a positive Non-structural antigen 1 (NS1) or dengue IgM antibody. ALC (absolute ALC and percentage) data were extracted from the Sysmex XS500i automated full blood count (FBC) analyzer (Sysmex Corporation Kobe, Japan). Clinical data was recorded from medical records and the computerized data base maintained by NH.
530 patients were enrolled. Patients with clinical manifestations of severe dengue have a significantly higher AL % compared to dengue without warning signs. Patients who presented with respiratory compromise had statistically significantly higher AL% compared to those without. (AL%; 8.65±12.09 vs 2.17±4.25 [p = 0.01]). Similarly, patients who developed hypotension had higher AL% compared to those who did not suffered from shock (AL%; 8.40±1.26 vs 2.18±4.25 [p = 0.001]). The AL% of dengue patients presenting with bleeding, at 4.07%, is also higher than those without bleeding complications, at 2.15%. There was a significant negative association between platelet count and AL% (p = 0.04).
Clinical manifestations of severe dengue have a significantly higher AL % compared to dengue without warning signs. AL % at presentation may be predictive of severe DI and future larger prospective longitudinal studies should be done to determine if AL % on admission is predictive of the complications of DI.
Dengue infection (DI) is the most common arthropod-borne virus disease, with 2.5 billion people worldwide, across all age groups, at risk of infection . The clinical manifestations of DI are heterogeneous, with the most severe and life-threatening forms being dengue hemorrhagic fever (DHF) and dengue shock syndrome (DSS). Between 2001 to 2010, an estimated 5,906 DI related deaths were reported.
Table 1 shows the socio-demographic characteristics of study population. 52.5% of patients were female. The median age was 34 years. 77.9% did not have any co-morbidity. 43.7% of the patients were admitted to hospital following two days of fever, while 84.3% were admitted after four days of fever. Confirmation of DI was based on positive Dengue NS1 (81.1%), Dengue IgM (16.9%) and both (2.0%). The percentage of patients who suffered from severe dengue is 1.32%.
We showed that patients with clinical manifestations of severe dengue have a significantly higher AL % compared to dengue without warning signs. Similarly, the AL % among those with respiratory compromise, hypotension and bleeding symptoms is higher as compared to those without. These results suggest that a higher AL % level is associated with increased complications of dengue infection, leading to severe dengue. From Tables 3 and 4, it is evident that patients with severe dengue infection, have significantly more AL % in their peripheral blood, compared to DI without warning signs. In addition, there is a statistically significant 5-fold difference in AL % between severe and non-severe DI. From subgroup analysis between DI with and without warning signs in Table 3, there is a positive association of AL % with severity of DI. The predictive cut-off AL % value for predicting severe DI is > 0.55, as shown in Fig 1, with the sensitivity of 95.5% and specificity of 71.5%. For our study, this is a point estimate which limits its predictive value, therefore, it is likely that consecutive measurements may improve on this and future studies would be required. This is clinically relevant as this may be used as a simple and dynamic predictive measurement for DI severity.
Clinical manifestations of severe dengue have a significantly higher AL % compared to dengue without warning signs. AL % at presentation may be predictive of severe DI and future larger prospective longitudinal studies should be done to determine if AL % on admission is predictive of the complications of DI later in the course of dengue infection.