Platelets play a pivotal role in haemostasis, vascular integrity, and wound healing, and their quantitative and qualitative assessment holds significant clinical importance across a spectrum of haematological and medical conditions. Platelet count, along with platelet indices such as mean platelet volume (MPV), plateletcrit (PCT), and platelet distribution width (PDW), is associated with disorders ranging from thrombocytopenia and bleeding diatheses to thrombocytosis and thrombotic risk. Automated haematology analysers have become the standard modality; however, they arenot without limitations. Platelet clumping, small platelet size, the presence of micro-fragments or red cell fragments and giant platelets can affect the accuracy of automated counts. Conversely, manual platelet counting methods provide an important means of cross-verification, especially in samples flagged for abnormalities. (1,2)There remains potential for clinically relevant discrepancies between automated and manual platelet counts and indices. Discrepancies may influence key clinical decisions like determining the need for platelet transfusion, diagnosing pseudo thrombocytopenia, or monitoring platelet recovery in treatment settings. (3,4) In the context of a tertiary care centre where a high volume of samples, heterogeneous patient populations and varying sample pre-analytical problems exist, systematically analysing the extent, direction, and clinical implications of discrepancies between automated versus manual platelet counts and indices is of paramount importance. (5) Therefore, the present study aims to evaluate and quantify the agreement, bias, and potential systematic discrepancies between automated platelet counts/indices and manual counting methods in our setting. Ultimately, our goal is to contribute for the improved accuracy in platelet reporting and interpretation in a tertiary care environment.

This cross-sectional study was conducted at Mel maruvathur Adhi para sakthi Institute Of Medical Sciences And Research, Tamilnadu, India from December 2023 to May 2024 (a period of six months), assessing 250 samples. Ethical clearance was obtained with reference number 52 dated 08/12/2023. Inclusion criteria: Patients of all age group, both gender presenting during the study period consecutively were included in the study. Exclusion criteria: Blood samples with signs of clotting or improper handling were excluded from the study. Study Procedure The demographical details of the participants were obtained from the hospital's laboratory information system. Blood samples were analysed for platelet count and other platelet parameters using the automated haematology cell counter, Mindray BC-6200. The automated haematology analyzer was regularly monitored for quality using daily quality controls and yearly calibrated as per manufacturer guidelines. Peripheral blood smears were prepared from the blood samples collected in EDTA vacutainers and stained using Leishman stain according to standard procedures. Microscopy of the stained peripheral smears were assessed for platelet count, size, shape and platelet aggregation under oil immersion lens (100X) in 10 fields. where RBCs are just touching each other in monolayer sheet, and then indirect method of platet counting was done by taking average number of platlets in ten fields that were multiplied by 20,000. The pheripheral smears were screened by two pathologists. All statistical analyses were performed using SPSS 30. Comparisons between manual platelet counts and automated analyzer platelet counts were performed using the paired Student’s t-test for continuous variables. Categorical variables were compared using the Chi-square test or Fisher’s exact test, as appropriate. Correlation between the two platelet estimation methods were evaluated using Pearson’s correlation coefficient (r), and linear regression analysis were performed to assess the relationship and agreement between the methods. The diagnostic performance of manual platelet estimation were assessed by calculating sensitivity, specificity, positive predictive value(PPV),negative predictive value (NPV), and diagnostic accuracy, using the manual method as comparator reference method. A p-value < 0.05 were considered statistically significant, with a confidence level of 95%.

The studyincluded250participantswithanage range of1to80years. Themeanage was 41.14 years, with 124 (49.6%) females and 126(50.4%) males. [Table 1]. Table1:Distribution of participants by gender Gender Frequency Percent Male 126 49.6% Female 124 50.4% Total 250 100.0% Theplateletcounts rangedfrom4×10⁹/Lto 1153×10⁹/Lbytheanalyzerandfrom20 × 10⁹/L to 1100× 10⁹/Lmanually. The mean platelet count by the automated analyzer (210 ± 160) was significantly lower than the manual count (270 ± 160) (p < 0.001). A very strong positive correlation was observed between the two methods with a Pearson correlation coefficient (r = 0.989, p < 0.001). This confirms excellent agreement between the automated andmanualplateletcounts,thoughminor underestimation by the analyzer was noted. Among the 250 samples, 49 (19.6%) were classified as normal by both methods, 56 (22.4%) as thrombocytosis by both methods, and 40 (16.0%) as thrombocytopenia by both methods. However, 105 (42.0%) samples identified as thrombocytopenic by the automated analyzer were found to have normal plateletcount son manual examination,suggestingahigh proportion of pseudothrombocytopenia by the automated method [Table 2]. Table2:Comparison of Automated and Manual Platelet count Plateletcount Automated Manual Normal 49cases(19.6%) 49cases(19.6%)+105 (42.0%) Thrombocytosis 56cases(22.4%) 56cases(22.4%) Thrombocytopenia 40cases(16.0%) 40cases(16.0%) Psuedothrombocytopenia 105cases(42.0%) 0 (normal) Among the 42% of pseudothrombocytopenia cases, the causes of discrepancy found out on peripheral examination wereas follows: 42%ofthepatients had platelet clumps, 39% had giant platelets, and 19% had both platelet clumps and giant platelets. [Figure 1], [Figure 2]. Regarding the platelet indices the Mean Platelet Volume (MPV) and Platelet Distribution Width (PDW) were evaluated for all subjects and compared between thrombocytopenic and non-thrombocytopenic groups. [Table 3&4]. Table3:The values of MPV and PDW across all categories Catagory MPV(fl) PDW Normal 8.6 – 10.4 15.3 – 16.4 Thrombocytosis 7.0 – 8.8 15.2 – 16.2 Thrombocytopenia 9.4 – 14.4 15.0 – 17.2 Table4:MPVandPDWvaluesbetweenthrombocytopenicandnonthrombocytopenic groups Group MeanMPV (fL) MeanPDW pvalue (MPV) pvalue(PDW) Thrombocytopenic 13.57 ± 1.4 16.63 ± 0.9 0.397 0.846 Non-Thrombocytopenic 10.35 ± 1.9 16.78 ± 0.8 - - An inverse trend was evident—lower platelet counts were associated with higher MPVvalues though the relationship was not statistically significant in this dataset. [Figure 3]. PDW values showed minimal variation between groups, suggesting a relatively uniformed greeof platelet size heterogeneityregardles softotal platelet number.This lack of significant difference reinforces that PDW alone is not a reliable discriminator of thrombocytopenic etiology in mixed clinical populations [Figure 4] When manual platelet count was considered as comparator reference standard, the automated platelet count demonstrated a sensitivity of 100% and a specificity of 50.0% for detecting thrombocytopenia. The PPV was 27.6%, indicating that only about one-quarter of thrombocytopenic results obtained by the automated analyzer were confirmed by manual counting, whereas the NPV was 100%, indicating that samples classified as non-thrombocytopenicbytheautomatedanalyzerwereallconfirmedasnon-thrombocytopenicon manual examination.

The accurate estimation of platelet count remains acritical component of hematological evaluation, particularly in patients with thrombocytopenia, infectious diseases, hematological malignancies, and bleeding disorders While automated hematology analyzers offer rapid and reproducible results, discrepancies between automated and manual platelet counting methods continue to pose diagnostic challenges. (6)In the present study, a very strong positive correlation was observed between automated and manual platelet counts (r = 0.989, p < 0.001),confirming that automated analyzers provide reliable platelet estimates in the majority of cases. However, automated counts were significantly lower than manual counts, indicating a tendency toward under estimation in selected samples. Our findings are consistent with the study by Tariq et al and Pathak et al, who reported that automated platelet counts significantly underestimated platelet levels in thrombocytopenic patients compared with manual peripheral smear verification. They demonstrated a mean automated count of 58×10⁹/Lcomparedwithamanualcountof117×10⁹/L, emphasizing the importance of smear review in thrombocytopenic samples. (7,8)However there was a concordance of all normal, all thrombocytosis and 40 cases of thrombocytopenia between automated and manual platelet count method. A noteworthy observation in the present study was the presence of a large “grey zone” comprising 42.4% of samples that were categorized as thrombocytopenic by the analyzer but demonstrated normal platelet counts on manual examination. Peripheral smear review identified platelet clumps (42%), giant platelets (39%), and a combination of both (19%) as the major causes of pseudo thrombocytopenia. These findings are in agreement with the review by Lardinois et al, Macro and Bahri et al., who highlighted EDTA-dependent pseudo thrombocytopenia, platelet aggregation, and giant platelets as the most common causes of spuriously low automated platelet counts. (9,10,11) The most common cause of pseudo thrombocytopenia induced by platelet clumps are EDTA based which by calcium chelation induces conformational change in GP IIb/ IIIa receptor on the platelets. Natural antibodies present in some individual binds with this modified receptor and triggers its clumping. The other causes are improper sample collection in the form of difficulty in venous access, improper mixing, delayed processing and rarely due to clinical situations. The present study also evaluated platelet indices, particularly MPV and PDW, in relation to platelet counts. Although thrombocytopenic patients demonstrated lower platelet counts with relatively higher MPV values, the difference was not statistically significant. Nevertheless, the inverse trend observed between platelet count and MPV supports the biological concept that increased peripheral platelet destruction stimulates compensatory bone marrow release of larger, younger platelets which are concordance with other studies. These findings suggested that MPV may serve as an indirect marker of platelet turn over and marrow response.(12,13) PDW showed minimal variation between thrombocytopenic and non-thrombocytopenic groups and exhibited only weak inverse correlations with platelet counts. This finding was comparable to previous studies that demonstrated substantial overlap of PDW values across different thrombocytopenic conditions, limiting its utility as an independent discriminator of thrombocytopenia etiology. (12) Our study revealed 100% sensitivity & NPV by automated method which signified the efficacy of this method in identifying non thrombocytopenic cases. However specificity of 50% & PPV of 27%indicated relative limitation of automated method in identifying and confirming thrombocytopenia when compared to manual method. Similar findings have been reported by Mishra etal., who observed discrepancies between automated and manual platelet counts, particularly among samples flagged as thrombocytopenic by automated analyzers. In their study, 8.6% of samples flagged as thrombocytopenic by the automated method were found to have normal platelet counts on manual examination, highlighting the possibility of false-positive thrombocytopenia with automated counting. (1) The strong correlation observed between manual and automated platelet counts in our study supports the continued use of automated analyzers as the primary method for routine platelet enumeration. However, the substantial proportion of pseudo thrombocytopenia cases highlights the indispensable role of manual peripheral smear review, particularly in samples with unexpectedly low platelet counts, abnormal histograms, or analyzer flags. (14)Recent recommendations for platelet count verification emphasize that abnormal thrombocytopenic results should be systematically reviewed to identify pre analytical sample & analytical interferences and prevent diagnostic errors. (15).

Our study concluded that there was an overall good agreement between automated and manual platelet counting method among non thrombocytopenic patients. However the values by machine were little bit lower than that of manual method. There was a high discrepancy among thrombocytopenic cases due to pre analytical factors interference which Highlighted the importance of manuals mear verification inc linically unexpected, discrepant, flagged cases. Though automated analysis is reliable for routine lab practice due to its rapidity, standardization, high throughput analysis, the comprehensive microscopic smear evaluation can improve the accuracy of platelet assessment by identifying pseudo thrombocytopenia & helpful for clinical decision making. Platelet indices were not statistically reliable parameter to predict platelet count and degree of thrombocytopenia in our study.

1. MishraS, Gupta A, KumarK. Automation versus manual platelet count: an audit of real-life scenario in a tertiary care centre of India. Blood. 2022;140(Suppl1):11259-11260. 2. Jain DK. Comparison of platelet count by manual and automated method. Int J Res Med Sci. 2020;8(10):3523-3527. doi:10.18203/2320-6012.IJRMS20204011. 3. Al-Hosni ZS, Al-KhaboriM ,Al-MamariS ,Al-QasabiJ, DavisH, Al-Lawati H, etal. Reproducibility of manual platelet estimation following automated low platelet counts. Oman Med J. 2016;31(6):409-413. doi:10.5001/omj.2016.83. 4. Lavanya M,Jayanthi C.Platelet estimation by manual and automated methods. AnnPathol Lab Med. 2019;6(11). doi:10.21276/APALM.2538. 5. Manickam N,Varghese RG. Evaluation of platelet count and its indices in various clinical conditions using an automated haematology analyzer in a tertiary care hospital. Ann Pathol Lab Med. 2018;5(7). doi:10.21276/APALM.2143. 6. Bhola A, Garg R, Sharma A, et al. Macro thrombocytopenia: role of automated platelet data indiagnosis. Indian J Hematol Blood Transfus .2023; 39(2):284-293. 7. TariqA, RashidA, Riaz MN. Comparison of platelet count by automated and manual methods in thrombocytopenia patients. J Haematol Stem Cell Res. 2023;3(1):17-20. 8. Pathak R, Pudasaini S, Kharel M, Kunwar S. A comparative study of platelet counts by automated and manual method in patients with thrombocytopenia and thrombocytosis. Nepal Med Coll J. 2025;27(2):113-118. 9. Lardinois B, Favresse J, Chatelain B, etal. Pseudo thrombocytopenia: are view on causes, occurrence and clinical implications. J Clin Med. 2021;10(4):594. doi:10.3390/jcm10040594. 10. Cattaneo M. Pseudo thrombocytopenia and other conditions associated with spuriously low platelet counts. Haematologica. 2025;110(8):1677-1692. 11. BahriR, Mohamed Amine A, Khayati S, etal. Pseudo thrombocytopenia: automate-blood smear confrontation. Int J Recent Innov Med Clin Res. 2021;3(1):21-29. 12. JurkK, Shirav and Y. Platelet pheno typing and function testing in thrombocytopenia. J Clin Med. 2021;10(5):1114. Doi :10. 3390 /jcm 10051114. 13. Aashna, Mahajan D, KoulKK, Jandial A. Platelet count correlation: automated versus manual on peripheral smear. Indian J Pathol Oncol. 2019;6(3):381-387. 14. Keswani AS, etal. Manual versus automated platelet count estimation: comparison. Natl J Lab Med. 2025;14(2):PO01-PO03. 15. MG,KM,P, Ullas C. A comparative study of platelet count by manual method and automated analyser: are trospective study. Int J Res Med Sci. 2024;12(9):3326-3330.