Original article / research
|
|||||||||||||||||||||||||||||||||||||||||
Role of Impedance-Based Platelet Parameters and Related Ratios in Screening for Coronavirus Infection- An Institutional Experience |
|||||||||||||||||||||||||||||||||||||||||
Preethi A Rai, Jayaparkash Chandrashekar 1. Assistant Professor, Department of Pathology, Father Muller Medical College, Mangalore, Karnataka, India. 2. Professor and Head, Department of Pathology, Father Muller Medical College, Mangalore, Karnataka, India. |
|||||||||||||||||||||||||||||||||||||||||
Correspondence
Address : Dr. Preethi A Rai, Assisrtant Professor, Department of Pathology, Father Muller Medical College, Kankanady, Manglore-575002, Karnataka, India. E-mail: drpreethirai@gmail.com |
|||||||||||||||||||||||||||||||||||||||||
ABSTRACT | |||||||||||||||||||||||||||||||||||||||||
: Impedance based platelet parameters can be easily obtained from most automated haematology analysers. Few of these parameters, especially Platelet Count (PLT), Mean Platelet Volume (MPV) and Plateletcrit (PCT), have been shown to predict disease progression in Coronavirus Disease 2019 (COVID-19). However, data on usefulness of these parameters as predictors of COVID-19 infection, in suspected cases, is limited. Aim: To evaluate the role of impedance-based platelet parameters, like PLT, MPV, PCT, Platelet-Distribution Width (PDW), and related ratios, as predictors of COVID-19 infection in suspected cases, using real-time Reverse Transcription Polymerase Chain Reaction (rRT-PCR) as standard. Materials and Methods: Retrospective analytical cross-sectional study was conducted in September 2020, at a teaching hospital located in Mangalore, Karnataka, India. The study parameters were retrieved and/or calculated, from the medical records of 328 patients (from July 2020 to August 2020) who had undergone COVID-19 rRT-PCR testing of their respiratory samples, comprising equal number (n=164) of patients with positive and negative test results, matched for age and sex. The data were compared and analysed. Statistical Package for Social Sciences (SPSS) version 23.0 was used in statistical analysis of this study. The significance level (p-value) was set to 0.05. Results: Data of total 328 subjects (age range: 18-85 years, 164 COVID positive and 164 COVID negative subjects) was collected and analysed for comparison. Differences in PLT, PCT, and most of the calculated ratios were statistically significant between the two groups; however, Area Under Curve (AUC) for all were <0.7. PCT values were significantly lower in COVID-19 positive group, despite normal PLT (p-value=0.005). Conclusion: Impedance based platelet parameters have limited role as predictors of COVID-19 infection in suspected patients, despite of their established prognostic value. Decrease in PCT values possibly precedes occurrence of thrombocytopenia following COVID-19 infection. | |||||||||||||||||||||||||||||||||||||||||
Keywords : platelet counts, Haematological tests, Mean platelet volume, Predictive value of tests, Sensitivity and specificity, Thrombocytopenia | |||||||||||||||||||||||||||||||||||||||||
INTRODUCTION | |||||||||||||||||||||||||||||||||||||||||
The PLT is a component of routine haematological investigation, which can be estimated using various cost-effective manual and automated methods. Among automated methods, the one using impedance technology is popularly employed. Along with PLT, most of the automated haematology analysers working on the principle of impedance, also simultaneously derive other related parameters like the MPV, PCT and PDW, collectively referred to as Platelet indices. With regard to COVID-19, various studies have been conducted to determine the characteristics and prognostic implications of PLT and related indices like MPV, in COVID-19 patients (1),(2),(3),(4). Most of these studies have suggested an association of decreased PLT with increased disease severity and mortality in COVID-19 disease, thus serving as a prognostic biomarker (1),(2). This study was primarily conducted to compare, and determine if there are statistically significant differences in, PLT and related impedance based indices and ratios, between patients who were positive and negative for ‘Severe Acute Respiratory Syndrome Corona virus 2’ (SARS-CoV-2) Ribonucleic Acid (RNA) in respiratory samples, tested by COVID-19 rRT-PCR test. The study variables included MPV, PCT, PDW, Platelet-to-Lymphocyte ratio (PLR), Red Blood Cells-to-PLT ratio (RBC-PLT ratio), ratio of MPV to PCT (MPV/PCT ratio), ratio of PDW to PLT (PDW/PLT ratio), ratio of MPV to PLT (MPV/PLT ratio), ratio of PDW to PCT (PDW/PCT ratio) and ratio of product of MPV and PDW to product of PLT and PCT {(MPVxPDW)/(PLTxPCT) ratio}. The secondary objective of the study was to determine the sensitivity and specificity of the above parameters that had statistically significant differences among the two groups of patients, in predicting the results of the COVID-19 rRT-PCR test. | |||||||||||||||||||||||||||||||||||||||||
MATERIAL AND METHODS | |||||||||||||||||||||||||||||||||||||||||
This study was a retrospective analytical cross-sectional study conducted at a teaching hospital catering to all clinical specialties, located in Mangalore, Karnataka, India. Patient data that were available from the hospital medical records from July 2020 to August 2020 (two months period), were used for the study. Data analysis was done in September 2020. Ethical clearance was obtained for the study from the Institutional Ethics Committee (Reference number: FMIEC/CCM/446/2020). Sample size calculation: The minimum sample size needed for the study, as derived by the formula, n={2(Za+Zß) p (1-p)}/(p1-p2)2, with reference to study by Li Q et al., was 139 per group (5). The values used in the formula were Za=1.96 at 95% confidence interval; Zß=0.841 at 80% power; p1=4.3%; p2=14%. Inclusion criteria: First group (n=164) included all patients above 12 years of age, whose upper respiratory samples were tested and reported positive for SARS-CoV-2 Ribonucleic Acid( RNA) using COVID- 19 rRT-PCR test in the Microbiology Department of the hospital laboratory, and also had haematological analysis of their venous blood sample done as a part of routine investigation, using Coulter LH 750 automated haematology analyser (manufactured by Beckman Coulter, Canada L.P., Mississauga), in the Haematology Department of the hospital laboratory, within a span of maximum two days prior or after the day of RT-PCR testing, irrespective of the clinical indication for the testing. The second group (n=164) enrolled included whose upper respiratory samples were tested and reported negative for SARS-CoV-2 infection using the COVID-19 rRT-PCR test, during the same time-frame, and were age- and sex-matched with the previous group. Exclusion criteria: Positives cases where haematological data were unavailable and/or where age and sex-matched ‘COVID-19 Negative’ controls were not available were excluded from the study. Study Procedure The institutional standard operating procedures were followed for the sample collection and for conducting the tests. The venous blood samples for haemograms were collected in standard 4 mL Becton Dickinson (BD) Dipotassium Ethylene Diamine Tetraacetic Acid (K2 EDTA) vacutainer tubes and were run in Complete Blood Count (CBC) mode in the automated haematology analyser within two hours of collection. Values of PLT, Red Blood Cells Count (RBC), MPV, PCT, PDW, Total Leukocyte Count (TLC) and Absolute Lymphocyte Counts (ALC) were noted retrospectively from the computer database of the analyser using the ‘unique identification number’ of the test sample, from which parameters like PLR, RBC-PLT ratio, MPV/PCT ratio, PDW/PLT ratio, MPV/PLT ratio, PDW/PCT ratio and (MPVxPDW)/(PLTxPCT) ratio were calculated for each subject and the results were compiled and analysed. The haematology laboratory adheres to internal and external quality control checks, participates in the National external quality assurance program, and is accredited by the National Accreditation Board for Testing and Calibration Laboratories (NABL), as per ISO 15189 standards. Reference ranges used were 150×109/L to 450×109/L for PLT, 7.2-11.7 fL for MPV, 0.22-0.24% for PCT and 10.0%-17.9% for PDW (6),(7). STATISTICAL ANALYSIS The data analysis was done using International Business Machines (IBM) SPSS Statistics 23.0 software (Statistical Product and Service Solutions, developed by International Business Machines corporation). The numerical variables were expressed as mean, median, standard deviation and Inter-Quartile Range (IQR), and compared between the two study groups using Mann-Whitney U test. Categorical data were compared using Chi-squared test/Fisher’s-exact test. Receiver Operating Characteristic (ROC) curve analysis was done for the parameters that showed statistically significant differences between the two study groups, to estimate the AUC, to identify the best threshold for these variables with their sensitivity and specificity, and to evaluate their discriminative ability in predicting the results of the COVID-19 rRT-PCR test. A p<0.05 was considered statistically significant, with p<0.01 being statistically highly significant. | |||||||||||||||||||||||||||||||||||||||||
RESULTS | |||||||||||||||||||||||||||||||||||||||||
Over a period of two months during which this study was conducted, a total of 8149 patients were tested for SARS-CoV-2 RNA by rRT-PCR. Out of these, SARS-CoV-2 RNA was detected in 1388 cases, not detected in 6750 cases, and, the result was inconclusive in 10 cases. After application of inclusion and exclusion criteria, statistical analysis was carried out using haematological data collected from A total of 328 subjects (164 COVID-19 positive and COVID-19 negative subjects) were analysed and results were tabulated. The age of the subjects in the study ranged from 18-85 years, with a mean age±standard deviation of 40.7±15.6 years in each group. Each group had 87 females (53%) and 77 males (47%). (Table/Fig 1) compares the values of PLT, MPV, PCT, PDW, PLR, RBC-Platelet ratios, MPV/PCT ratios, PDW/ PLT ratio, MPV/PLT ratio, PDW/PCT ratio and (MPV×PDW)/(PLT×PCT) ratio between both the groups. Statistically significant differences were found in values of PLT, PCT, TLC, ALC, RBC-PLT ratio, MPV/PCT ratio, PDW/PLT ratio, MPV/PLT ratio, PDW/PCT ratio and (MPV×PDW)/(PLT×PCT) ratio between the two study groups, with all except MPV/PLT ratio being highly significant (p-value <0.01). No statistically significant difference was found in the PLR, MPV and PDW values between the two groups. In both the groups, PLT were within normal range in 89.6% of the study population. Distribution of the normal and abnormal result values of PLT, MPV, PCT, PDW, RBC count and TLC in both the study groups were compared (Table/Fig 2). The proportion of subjects with normal PLT, mild, moderate and severe thrombocytopenia, and thrombocytosis in both the groups were comparable (p-value=0.927). With regard to the values of MPV and PDW, the proportion of subjects with result values falling within the normal reference range, below the lower limit of normal and above the upper limit of normal respectively were also comparable between the two study groups, with p-values of 0.47 for MPV, and 0.582 for PDW. With regard to PCT, both the groups had equal percentage of subjects having result values within normal reference range (17.1%). However, a relatively higher percentage (67%; n=110) of COVID-19 positive subjects were having result values below the lower limit of normal reference range compared to the COVID-19 negative subjects (52.4%; n=86), and, a relatively higher percentage (30.5%; n=50) of COVID-19 negative subjects were having result values above the upper limit of normal reference range compared to the COVID-19 positive subjects (15.9%; n=26). These differences in distribution of PCT values among both the study groups were found to be statistically highly significant (p-value=0.005). Within each of the study groups, the proportion of subjects with lower than normal PCT levels was not dependent on the proportion of subjects with low PLT levels (thrombocytopenia) (p-value >0.05 for each group). With regard to the TLC, majority of the subjects in both the groups had TLC values within normal reference range; however, a significantly higher percentage of COVID-19 negative subjects (34.1%; n=56) had leucocytosis (TLC>11×109/L), and a significantly higher percentage (6.7%; n=11) of COVID-19 positive subjects had leucopenia (TLC<3.5×109/L), compared to the other group, respectively (Table/Fig 2). On analysing the platelet related variables further with ROC curve, PLT values of less than or equal to a cut-off value of 2,50,550/cumm had a sensitivity of 64.6% and a specificity of 57.2% for predicting cases positive for SARS-CoV-2 RNA by RT-PCR (AUC=0.61). PCT values of less than or equal to a cut-off value of 0.2% had a sensitivity of 61% and specificity of 62% (AUC=0.62), PDW/PLT ratio values of greater than or equal to a cut-off value of 0.068 had a sensitivity of 60.4% and specificity of 57.3% (AUC=0.597), (MPV×PDW)/(PLT×PCT) ratio values of greater than or equal to a cut-off value of 2.87 had a sensitivity and specificity of 60% (AUC=0.602), RBC-PLT ratio values of greater than or equal to a cut-off value of 16.67 had a sensitivity of 70.7% and specificity of 50.6% (AUC =0.628), MPV/PCT ratio values of greater than or equal to a cut-off value of 38.6 had a sensitivity of 69.5% and specificity of 51.8% (AUC=0.607), MPV/PLT ratio values of greater than or equal to a cut-off value of 0.035 units had a sensitivity of 53.7% and specificity of 61% (AUC=0.588), and PDW/PCT values of greater than or equal to a cut-off value of 78.3 had a sensitivity of 66.5% and specificity of 50.6% (AUC=0.605) for predicting cases positive for SARS-CoV-2 RNA by RT-PCR. The standard error, asymptotic significance (p-value), and asymptotic 95% confidence interval for these parameters are depicted in (Table/Fig 3). The ROC curves of PLT and PCT values are shown in (Table/Fig 4). The ROC curves of PDW/PLT ratio, RBC/PLT ratio, MPV/PCT ratio, MPV/PLT ratio, PDW/PCT ratio and MPV*PDW/(PLT*PCT) ratio are shown in (Table/Fig 5). An AUC values below 0.7 for all these parameters suggests that these tests have a poor discriminative ability in predicting the positivity for SARS-CoV-2 RNA by RT-PCR, individually (8). | |||||||||||||||||||||||||||||||||||||||||
DISCUSSION | |||||||||||||||||||||||||||||||||||||||||
Prior studies have shown prognostic value of PLT, PCT and the ratios of MPV-to-PCT, PDW-to-PLT and MPV-to-PLT, in predicting disease progression and/or mortality in COVID-19 disease (1),(2),(9),(10),(11). However, studies on differences of these parameters between the patients with and without COVID-19 infection are limited. In this study, we found that platelet related parameters like PLT, PCT, RBC/PLT ratio, MPV/PCT ratio, PDW/PLT ratio, MPV/PLT ratio, PDW/PCT ratio and (MPVxPDW)/(PLTxPCT) ratio were significantly different for patients who were tested positive for SARS-CoV-2 RNA by rRT-PCR compared to those who were negative for the same. However, none of the parameters had a best cut-off value with a good combination of sensitivity and specificity in predicting positivity for coronavirus infection, precluding their use as good early screening tools for the same. In general, thrombocytopenia have been reported to be one of the common haematological changes seen in patients with COVID-19, apart from reduced lymphocyte count with normal white blood cell count, prolonged activated partial thromboplastin time and elevated D-dimer levels (12). However, it is important to realise that many patients with COVID-19 do not present with thrombocytopenia, and PLT level abnormalities are more significant in severe cases where serial follow-up may help predict worsening disease severity and/or mortality. The present study revealed that majority (89.6%; n=147) of the COVID-19 positive patients present with normal PLT levels, indicating that thrombocytopenia is not a sensitive biomarker of COVID-19 infection per se. In keeping with that, even a study done exclusively on 215 COVID positive patients who were admitted in hospital, observed thrombocytopenia in only 25.1% patients on the hospital admission day and in 24.1% patients on the third follow-up day (10). In other studies, done on 41, 99 and 1099 patients, thrombocytopenia was observed in 5%, 12% and 36.2% patients on admission, respectively (13),(14),(15). These variations in reported rates of thrombocytopenia in COVID-19 may be attributed to variations in study sample sizes and proportions of patients with severe disease manifestations in the study. Platelet activation and aggregation secondary to the damage to pulmonary endothelial cells have been implicated as cause for increased platelet consumption or thrombogenesis in severe COVID-19 disease. This along with bone marrow suppression and immune mediated destruction of the platelets has been hypothesised to cause thrombocytopenia in severe COVID-19 disease (3),(4). While some studies have reported correlation between decreasing platelet level and increasing disease severity or mortality in COVID-19 disease, few others did not find any such correlation (1),(2),(10). The established association of thrombocytopenia with severe or advanced COVID-19 disease suggests that most of the COVID-19 positive patients in the present study had milder disease severity and/or were likely in early course of COVID-19 infection with non severe disease. Occassionally, even thrombocytosis may be seen in COVID-19 positive patients. In the present study, three (1.8%) patients had PLT above normal reference range, and, Chen N et al., in their study observed the same in four (4%) COVID-19 positive patients on admission (14). However, as the frequency of thrombocytosis in COVID-19 positive group was not significantly different from the COVID-19 negative group in this study, it is possible that thrombocytosis was just a coincidental finding in these patients, with no direct causal relationship with COVID-19 infection. The finding in this study of abnormally low PCT values (below the lower-limit of normal reference range) in significantly higher proportion of COVID-19 positive patients compared to the negative group, despite of normal PLT levels in the majority and comparable rates of thrombocytopenia in the two groups, suggests possibility that alterations in PCT becomes evident earlier in the course of COVID-19 disease than thrombocytopenia (Table/Fig 2). However, as there was no follow-up of these parameters in the study subjects, there is lack of data on what proportion of these subjects, with low PCT and normal PLT levels at the time of presentation, indeed developed thrombocytopenia later on. Hence, the utility of PCT as early predictor of impending thrombocytopenia cannot be conclusively established with the available study findings. Moreover, PCT can vary with variation in MPV, leading to significant overlap of PCT between thrombocytopenic patients and patients with normal PLT values (16). However, there were no statistically significant differences in MPV values among the two study groups, in this study. Güçlü E et al., in their study, reported an association of increase in MPV and PDW with increased disease severity and/or mortality rate in COVID-19 patients (10). PLR has also been reported as a parameter that indicates the severity of the infection some studies (11),(17),(18). Yun H et al., in their study comparing 32 patients with COVID-19 and 2337 negative patients, reported significantly higher levels of MPV in COVID-19 positive patients (19). In contrast, no statistically significant differences were found in the PLR, MPV and PDW values, between the patients with and without COVID-19 infection in this study, with majority of subjects in both the groups having MPV and PDW values within normal reference range (Table/Fig 2). This suggests a possibility that parameters like PLR, MPV and PDW may be minimally altered in the early course of the COVID-19 infection and/or in non severe version of the disease. The study findings of significantly higher levels of all the studied platelet related ratios, i.e., PDW/PLT ratio, RBC-PLT ratio, MPV/ PCT ratio, MPV/PLT ratio, PDW/PCT ratio and (MPVxPDW)/ (PLTxPCT) ratio, in COVID-19 positive subjects compared to the negative group, can be attributed to their inverse relationship with PLT and PCT, as all of these ratios have either PLT, PCT or both as their denominators. However, in the study by Ozcelik N et al., which compared platelet indices in patients with COVID-19 pneumonia and influenza pneumonia, MPV/PLT ratio was reported to be statistically significantly lower in the COVID-19 group (20). Overall, this study revealed poor performance of these parameters as independent predictors of COVID 19 infection. In a case-control study done on paediatric patients by Golwala ZM et al., MPV/PCT ratio, PDW/PLT ratio and MPV/PLT ratio, in the first sample after hospital admission, were shown to be predictors of in-hospital paediatric mortality and could predict 65% to 67% of deaths accurately (9). Studies on these ratios in COVID-19 patients are limited in the literature. Whether these parameters are better or early predictors of COVID 19 disease severity/mortality, than the PLT, PCT, PDW or MPV alone, needs further evaluation involving follow-up of patients. Limitation(s) The study had few limitations, apart from the small sample size and lack of follow-up of the study parameters in all the subjects. Factors like patient symptomatology, duration of illness, degree of disease severity, type of clinical manifestations at the time of testing and on follow-up, and presence or absence of any other illness/ co-morbidities in the participants of both the study groups, were not considered in the study. Certain co-morbidities that evoke systemic inflammatory response and/or impair immunity, may behave as potential confounding factors, as they may be associated with both alterations in platelet related parameters as well as an increased risk of acquiring the COVID 19 infection. | |||||||||||||||||||||||||||||||||||||||||
CONCLUSION | |||||||||||||||||||||||||||||||||||||||||
To conclude, though some of the platelet related parameters have been shown to have prognostic values in COVID-19 disease, they have a poor role as predictors of COVID-19 infection in suspected patients. The finding of normal platelet counts in majority of patients who were found positive for SARS-CoV-2 infection, suggests possibility that effects of COVID-19 disease on platelets, in early coarse of infection, is often not significant enough to be reflected as alterations in platelet counts and related ratios. However, abnormally low plateletcrit values may be seen earlier, compared to thrombocytopenia. More complex study designs involving study of serial follow-up data of platelet related parameters in COVID- 19 positive subjects may provide insight on the average time period taken for the occurrence of alterations in platelet related lab parameters, following initial detection of SARS-CoV-2 infection, with and without treatment. | |||||||||||||||||||||||||||||||||||||||||
|
|||||||||||||||||||||||||||||||||||||||||
TABLES AND FIGURES | |||||||||||||||||||||||||||||||||||||||||