Role of hematological parameters in the diagnosis of influenza virus infection in patients with respiratory tract infection symptoms.

Abstract Background The differential diagnoses of patients hospitalized for respiratory infections due to influenza virus vs other pathogens are challenging. Our study investigated whether hematological parameters such as neutrophil (N), lymphocyte (L), platelet (PLT), and neutrophil‐to‐lymphocyte ratio (NLR) contributed in diagnosing influenza virus infections and in discriminating other respiratory infections. Methods We retrospectively analyzed the laboratory characteristics of 307 patients with respiratory infections caused by influenza/non‐influenza virus and bacteria. The diagnostic abilities of hematological indexes were evaluated in the patients compared with 100 healthy people. Results The hematological parameters in patients with influenza virus infection were dramatically altered compared with those in the controls. Additionally, among the systemic inflammatory markers, the sensitivity of NLR for influenza detection was higher than that of N and L. PLT was significantly lower in influenza virus‐positive infection than in influenza virus‐negative infection. Moreover, when patients with influenza virus infection were cured, PLT returned to a normal level. The red blood cell (RBC) and hemoglobin (Hb) levels of influenza virus infection were higher than those of bacterial infection. Compared with traditional N and L, NLR and platelet‐to‐neutrophil (PNR) showed greater significance between influenza virus and bacterial infection (P < .01). Conclusion Neutrophil‐to‐lymphocyte ratio with high sensitivity is a preferable diagnostic tool to screen influenza virus‐infected patients than N and L. PLT accounts in the differential diagnoses of respiratory infections due to influenza virus and other pathogens among patients. In addition, RBC, Hb, NLR, and PNR can significantly differentiate between influenza virus infections and bacterial infections.


Results:
The hematological parameters in patients with influenza virus infection were dramatically altered compared with those in the controls. Additionally, among the systemic inflammatory markers, the sensitivity of NLR for influenza detection was higher than that of N and L. PLT was significantly lower in influenza virus-positive infection than in influenza virus-negative infection. Moreover, when patients with influenza virus infection were cured, PLT returned to a normal level. The red blood cell (RBC) and hemoglobin (Hb) levels of influenza virus infection were higher than those of bacterial infection. Compared with traditional N and L, NLR and platelet-toneutrophil (PNR) showed greater significance between influenza virus and bacterial infection (P < .01).
Conclusion: Neutrophil-to-lymphocyte ratio with high sensitivity is a preferable diagnostic tool to screen influenza virus-infected patients than N and L. PLT accounts in the differential diagnoses of respiratory infections due to influenza virus and other pathogens among patients. In addition, RBC, Hb, NLR, and PNR can significantly differentiate between influenza virus infections and bacterial infections.

| INTRODUC TI ON
Annual seasonal influenza epidemics of variable severity result in significant morbidity and mortality worldwide. 1 Since avian influenza was first identified in Shanghai, China, in March 2013, there have been a total of five epidemics. 2 Patients with influenza virus mostly present with fever and cough are prone to progression to viral pneumonia. Moreover, in the event of acute respiratory distress and important organ dysfunction, the fatality rate is even higher.
In addition, late diagnosis of community-acquired influenza A virus infection is associated with a delay in ICU admission, greater possibilities of respiratory and renal failure, and higher mortality rate.
Delay in diagnosis of flu is an independent variable related to death. Therefore, early diagnosis and treatment with antivirals are critical for achieving effective therapeutic outcomes. 3 At present, molecular assays (eg, nucleic acid amplification tests and antigen tests) targeting respiratory tract specimens are recommended as critical diagnostic tests for clinical decision-making according to influenza clinical practice guidelines in different countries. 4,5 However, these tests are limited by some technical provisions and specifications for molecular assay utilization, especially in underdeveloped areas and community hospitals. Furthermore, antigen tests have poor sensitivity to some influenza viruses. 6,7 Recently, it was reported that hematological inflammatory indexes based on blood cell analysis had an important predictive value for the prognosis of infections, cancers, and many other diseases. [8][9][10] This study aimed to conduct a retrospective analysis of hematological inflammatory parameters (eg, neutrophils, lymphocytes, platelets) and blood cell count indexes, particularly the NLR, in hospitalized patients with suspected influenza. Moreover, this study aimed to validate hematological indexes as potential indicators to discriminate influenza virus infection from non-influenza infection, which would facilitate a clear diagnosis and enable the initiation of antiviral treatment as early as possible when molecular assays for respiratory specimen tests are not performed.

| Study population
The present study retrospectively enrolled 307 suspected influenza virus-infected patients with typical acute onset of respiratory A total of 91 patients were identified as negative controls with typical respiratory symptoms; these patients were negative for both the nucleic acid and bacterial culture results (58 male and 33 female; age range: 23-95 years; mean age: 56.04 ± 18.53 years).

| Blood analysis for the determination of infection
The complete blood count was routinely examined in hospitalized patients with suspected influenza virus infection in the study. Two milliliters of peripheral blood was placed in hemogram tubes with ethylenediaminetetraacetic acid (EDTA). The blood count was determined using an automated hematology analyzer (Sysmex-20 instrument, Sysmex). The NLR was calculated as neutrophils/lymphocytes, the PLR as platelets/lymphocytes, and the PNR as platelets/neutrophils.

| Statistical analysis
Continuous variables are summarized as the mean ± SD (standard deviation

| Comparison of diagnostic sensitivity of N, L, and NLR
The results showed that WBC, N, RDW, NLR, and PLR increased significantly in the influenza virus infection group compared with the healthy control (P < .01 for all), while L, RBC, Hb, PLT, and PNR decreased (P < .01 for all) (Table 1). Specifically, the sensitivity of NLR detection was higher than the common systemic inflammatory markers, including neutrophils and lymphocytes (

| Comparison of hematological parameters between influenza virus-positive infection and influenza virus-negative infection
In the present paper, we found that PLT was lower in the influenza virus infection group (169.869 ± 93.379) than in the negative control group (226.209 ± 119.245) (P < .001), while Hb was higher (P < .05).
Moreover, the other routine blood parameters did not differ significantly between the two groups (P > .05) ( Table 3). In addition, L and PLT increased to normal levels concomitantly when influenza patients were completely cured (P < .05). Furthermore, no other hematological parameters were found to have significance with a complete cure of influenza (P > .05) ( Table 4).

| The differential sensitivity of hematological parameters between influenza virus infection and bacterial infection
Compared with bacterial infection, RBC, Hb, and PNR markedly increased in influenza virus infection, while WBC, RDW, neutrophils, and NLR decreased (Table 5). Blood cell count analysis is a simple, effective, and rapid laboratory diagnostic basis for evaluating infectious inflammatory responses. 11 The influenza virus-infected patients analyzed in this paper were mainly elderly people. Additionally, elderly people were mainly involved as the controls in this study. Therefore, the age and sex distribution of the selected controls were as close as possible to those of the influenza virus-infected group. The results showed that routine blood indicators were independent of age and sex, consistent with the findings of previous literature reports. 12  and may be why bacterial infections cause more significant reduction in red blood cells.

| D ISCUSS I ON
In conclusion, NLR was found to have a high sensitivity in de-