Role of Gray Scattergram in Detecting Blasts Compared with Microscopic Peripheral Smear in Acute Leukemia Patients Authors: Novi Rostikasari, Leni Lismayanti, Adhi Kristianto Sugianli, Tiene Rostini Published: 2026 ABSTRACT AbstractPurposeEarly and accurate diagnosis of acute leukemia is essential for timely treatment and improved patient outcomes. Gray scattergram patterns generated by hematology analyzers may indicate the presence of blasts and serve as an early screening tool. This study evaluated the correlation and agreement between gray scattergram findings and microscopic detection of blasts on peripheral blood smears in newly diagnosed acute leukemia patients.Patients and MethodsThis retrospective cross-sectional study used secondary data from inpatients with acute leukemia treated at Dr. Hasan Sadikin General Hospital between April 2024 and March 2025. Diagnoses were identified using ICD-10 codes for acute lymphoblastic leukemia and acute myeloid leukemia. Eligible patients had complete hematological data at diagnosis, including complete blood count and peripheral blood smear results. Collected variables included demographic characteristics, hematological parameters, peripheral blood blast percentage, analyzer-generated flags, and scattergram findings. Leukocyte differential counts and blast percentages were validated by manual microscopic examination. Scattergram findings were classified as gray or non-gray, with gray defined by the presence of ≥35.5% blasts. Agreement between gray scattergram findings and microscopic blast detection was analyzed using Cohen’s kappa statistic. However, the moderate level of agreement indicates that gray scattergram findings should be interpreted cautiously and require confirmation by microscopic examination.ResultsA total of 106 patients were included. Gray scattergram findings were observed in 66 patients, of whom 52 (83.9%) had ≥35.5% blasts on microscopic examination. Among non-gray scattergram cases, 10 patients (16.1%) also showed blast counts ≥35.5%. Agreement between gray scattergram findings and microscopic blast identification was moderate (κ=0.527, p<0.001).ConclusionGray scattergrams may provide early indications of blast presence and serve as a supportive screening tool in suspected acute leukemia. However, confirmation by peripheral blood smear or bone marrow examination remains necessary in clinical settings.Keywords: acute leukemia, gray scattergram, blasts, hematology analyzers, Cohen’s kappa Introduction Leukemia is a hematological malignancy characterized by the rapid proliferation of immature cells, most commonly blasts, which often results in hyperleukocytosis. In early stages, however, leukocyte counts may remain within normal limits or present as leukopenia. Leukemia is classified into two main categories: acute and chronic, with the acute form being associated with a more aggressive clinical course and higher mortality compared to the chronic type.1 The presence of circulating blasts is a key indicator for the diagnosis of acute leukemia, which can be confirmed through peripheral blood smear and bone marrow aspiration.2,3 However, identifying and quantifying blasts through microscopic examination is a manual process that demands considerable effort and relies heavily on skilled laboratory personnel, which may contribute to delays in diagnosis and timely initiation of treatment. Recent developments within the field have been made in 5-differential (5-diff) and 6-differential (6-diff) complete blood count (CBC) assays which allow detection of a blast’s cells for easier identification in scattergram plots. Hematology analyzers give out graphs called scattergrams that show the distribution of the different classes of cells in blood samples. The hematology analyzers displays a non-gray scattergram with specific colors: purple for lymphocytes, green for monocytes, light blue for neutrophils, yellow for basophils, orange for eosinophils, and dark blue for immature granulocytes. In contrast, the gray scattergram appears in cases where leukocyte types cannot be identified, falling outside the classification of the non-gray scattergram. Blasts that have similar nuclear size and shape to lymphocytes and monocytes and are stained by fluorescent dyes will appear as part of the non-gray scattergram. Meanwhile, blasts that are unstained and located outside the areas designated for lymphocytes and monocytes will appear as a gray scattergram. The presence of blasts sometimes appears as a gray area due to the high leukocyte count and is influenced by the lipid content in the blasts. Blasts have more lipid rafts on the cell membrane, which results in greater resistance to lysing reagents used in hematology analyzers, preventing complete fluorescence staining of intracellular components, particularly nucleic acids within the cell nucleus.4,5 This method merits consideration as a preliminary detection and follow up tool for leukemia alongside peripheral smear examination.4,6 However, it is important to note that scattergram patterns and their interpretation may vary across different hematology analyzer models, depending on the underlying detection principles, fluorescence technology, and classification algorithms used by each system. The precision pertaining to scattergram accuracy depends on the skill and experience of the observer since most elements analysed from scattergram patterns rely on the observer. Previous studies have described scattergram patterns; however, most of the available evidence remains descriptive and does not specifically evaluate the agreement between gray scattergram findings and microscopic blast identification using a defined diagnostic threshold. Therefore, the diagnostic value of gray scattergram as a reliable screening tool in acute leukemia has not been fully established. The present study tested whether the gray scattergram produced by hematology analyzers could reliably spot blasts in patients newly diagnosed with acute leukemia. Delayed detection of blasts may lead to postponement of definitive diagnosis and initiation of appropriate therapy, which can adversely affect disease progression and patient outcomes, particularly in acute leukemia cases that require urgent clinical intervention. Results from the scattergram were measured against side by side microscopic examinations of peripheral blood smears. A good correlation with the gray scattergram is expected to make the scattergram from the hematology analyzers a rapid and cost-effective adjunct modality for screening patients with symptoms of acute leukemia. Material and Methods Study Design A cross-sectional observational analytic study was conducted. The subjects of the study were acute leukemia patients at RSUP Dr. Hasan Sadikin Bandung from April 2024 to March 2025, identified through ICD-10 codes C91.0 and C92.0, corresponding to acute lymphoblastic and acute myeloid leukemia, respectively. Inclusion criteria required complete blood count (CBC) and peripheral blood smear data from subjects who were newly diagnosed with acute leukemia, patients with incomplete documentation were excluded. Collected variables included age, sex, and blast percentages reported in both CBC results and gray scattergram files. CBC and differential counts were analyzed using automated hematology analyzers (Sysmex XN-1000 and XN-1500) equipped with fluorescence flow cytometry and multi angle light scatter technology, which generated gray scattergrams suggestive of blast cell presence. Microscopic Examination of Peripheral Blood Smear Blood samples intended for morphological evaluation were processed into peripheral blood smear slides, prepared according to institutional standard operating procedures consistent with national hematology guidelines.7 The slides were examined microscopically by multiple observers, and only results with concordant findings were included and subsequently recorded in the Laboratory Information System (LIS) for analysis. Although interobserver variability was not formally quantified, only concordant results were included in the analysis to minimize potential variability and ensure data reliability. The presence of blasts was documented based on these validated microscopic findings. As this study used a retrospective design, there was limited control over sample quality and data completeness. However, this limitation was mitigated by the use of validated peripheral blood smear results obtained from the LIS, where all examinations were performed according to standardized procedures and reviewed by multiple observers to ensure data reliability. The diagnostic cut-off for peripheral blast count was set at 35.5%, as determined by a previous study conducted by Lismayanti and Arif (2024), which evaluated the optimal threshold of WBC and peripheral blast percentages in predicting gray scattergram findings in leukemia patients. This cut-off demonstrated good diagnostic performance (AUC 0.872, sensitivity 76.2%) and was used in this study as a locally validated operational threshold for analysis, rather than as a standard diagnostic criterion.6 Scattergram Hematology Analyzers The leukocyte distribution on the hematology analyzers scattergram, showing lymphocytes (pink), monocytes (green), neutrophils and basophils (blue), and eosinophils (orange), was included in the non-gray scattergram group of the White Blood Cell Differential (WDF) channel. This group represents the typical, identifiable leukocyte populations that fall within normal distribution patterns. Leukocytes that could not be identified by the hematology analyzers WDF channel, often indicating atypical or immature cells, were classified into the gray scattergram group. The gray scattergram highlights the presence of these abnormalities, allowing for further investigation and differentiation of abnormal leukocyte populations in clinical diagnostics.8,9 Statistical Analysis Cohen’s kappa statistic was calculated to assess the agreement between interpreting gray scattergram results and identifying blasts in peripheral blood smears. A kappa value approaching 1 signified substantial agreement.10 Ethical Clearance This research was granted ethical approval by RSUP Dr. Hasan Sadikin Bandung Hospital under the approval number DP.04.03/D.XIV.6.5/124/2025. The study adheres to the ethical principles outlined in the Declaration of Helsinki and was reviewed to ensure that it met all required ethical standards for research involving human data. The requirement for informed consent was waived because the research involved retrospective analysis of existing de-identified medical records, which posed no more than minimal risk to participants and rendered individual consent impracticable. Since this study only utilised secondary data, no direct interaction with patients or clinical trials was involved, minimising any potential ethical concerns related to participant involvement. Results This study utilized secondary data from patients diagnosed with acute leukemia (ICD-10 codes C91.0 and C92.0) who received care at RSUP Dr. Hasan Sadikin Bandung between April 2024-March 2025. As part of the study, complete blood count (CBC) results and peripheral blood smear findings were retrieved and analyzed. Figure 1 provides a detailed illustration of the subject distribution. Flowchart of Study Subject Determination. Table 1 shows the characteristics of research subjects. While no notable differences of gender and platelet count disparity existed between gray scattergram groups and non-gray scattergram groups, stratification based on hemoglobin, hematocrit, erythrocytes and leukocytes did show significant differences. Within the gray scattergram group, there was more frequent flagging for “WBC abn scattergram,” “monocytosis”, “lymphocytosis”, “abn lympho?”, and “blasts.” In contrast, the non-gray scattergram group had more flagging for “blasts?”, “abn lympho?”, “lymphocytosis”, and “monocytosis”. Only significant differences for “WBC abn scattergram”, “monocyte”, and “lymphocyte” were observed. Although several WBC flagging parameters such as “monocytosis”, “lymphocytosis”, “abn lympho?”, and “blasts” appeared more frequently in one group than the other, statistically significant differences were limited. Except for the “WBC Abn Scattergram” flag, most flagging parameters did not differ significantly between the gray and non-gray scattergram groups. Table 1Characteristics of Research SubjectsCharacteristicGray Scattergram (n=66)Non-Gray Scattergram (n=40)p-ValueAge (year)– Median167 Min-max1–762–73Sex, n (%)0.994 Male38 (57.5)23 (57.5) Female28 (42.5)17 (42.5)Hematology Examination Hemoglobin (g/dL)8.22 ± 1.789.06 ± 1.680.018 Hematocrite (%)24.61 ± 5.4926.73 ± 4.940.047 Erythrocyte (million/µL)2.92 ± 0.733.23 ± 0.720.032 Leukocyte (/µL)54.595 (2.480–420.590)5.505 (890–26.920)<0.001 Thrombocyte (/mm3)42.000 (2.000–275.000)53.000 (3.600–580.000)0.126 Peripheral Blood Blast (%)65(2–92)12.5(0–82)<0.001Type of Acute Leukemia (Bone Marrow Smear) ALL34 (51.5)30 (75.0)- AML32 (48.5)10 (25.0)WBC Scattergram Flagging Characteristics, n (%) WBC Abn Scattergram65 (98.5)5 (12.5)<0.001 Lymphocytosis59 (89.4)10 (25.0)<0.001 Monocytosis63 (95.5)12 (30.0)<0.001 Blast/Abn Lympho?13 (19.7)1 (2.5)0.011 Abn Lympho?40 (60.6)17 (42.5)0.070 Blasts?46 (69.7)22 (55.0)0.126Notes: Values shown in bold indicate statistical significance (p < 0.05).Abbreviations: n, number of patients; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; WBC, white blood cell; Abn, abnormal. Characteristics of Research Subjects Notes: Values shown in bold indicate statistical significance (p < 0.05). Abbreviations: n, number of patients; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; WBC, white blood cell; Abn, abnormal. According to previous study, the presence of blasts was considered significant when blast counts on peripheral smear reached ≥35.5%, which served as the diagnostic threshold in evaluating gray scattergram performance, Table 2 illustrates that subjects surpassing this threshold were predominantly found within the gray scattergram subgroup (p<0.001).6 Moreover, a moderate agreement (κ=0.527) was noted between the hematology analyzers result derived from blocks interpretation compared to microscopy of blasts counting. These findings support the role of gray scattergram, generated by currently available hematology auto analyzers, as a screening tool for identifying significant blast populations in acute leukemia. Table 2Agreement of Hematology Analyzer Gray Scattergram with Microscopic Blast Presence in Acute Leukemia PatientsBlast Presence≥35.5% Blast Presence<35.5% Blast Presencep-ValueKappa Value95% Confidence IntervalGray Scattergram (Hematology Analyzers), n (%)52 (83.9%)14 (31.8%)<0.0010.5270.364–0.690Non-Gray Scattergram (Hematology Analyzers), n (%)10 (16.1%)30 (68.2%) Agreement of Hematology Analyzer Gray Scattergram with Microscopic Blast Presence in Acute Leukemia Patients Discussion In this study, the median age of patients with gray scattergrams was 16 years, while in the non-gray scattergram group it was 7 years, indicating that both groups were predominantly pediatric. Most of the study subjects were children, with 65 out of 106 (61.3%) diagnosed with acute lymphoblastic leukemia (ALL), and 30 out of 106 (28.3%) with acute myeloid leukemia (AML). This predominance of pediatric cases may be due to the fact that bone marrow examination is more frequently performed in the laboratory for pediatric patients. In our institution, bone marrow smears both in pediatric and adult patients are examined and interpreted by clinical pathologists. As a result, ALL cases were more frequently identified than AML. This is consistent with a study by Warzocha (2015), which stated that ALL is more commonly found in children, whereas AML predominantly occurs in adults.11 There was no significant difference in gender distribution between the gray and non-gray scattergram groups (p=0.994). However, the overall incidence of acute leukemia in this study was higher in males than in females. This finding aligns with global epidemiological data showing that acute leukemia is more frequent in males. A study by Lin et al (2017) reported the incidence rate of acute leukemia to be 36.33 per 100,000 in males and 28.75 per 100,000 in females.12,13 Hemoglobin, hematocrit, and erythrocyte counts showed statistically significant differences between the gray and non-gray scattergram groups (p=0.018; p=0.047; p=0.032, respectively). Hemogram parameters were presented as mean ± standard deviation for normally distributed data and as median (minimum–maximum) for non-normally distributed data, based on normality testing. The lower values observed in the gray scattergram group (Hb 8.2 g/dL; Hct 24.61%; RBC 2.92 million/µL) compared to the non-gray scatttergram group (Hb 9.1 g/dL; Hct 26.73%; RBC 3.23 million/µL) reflect disrupted erythropoiesis due to bone marrow infiltration by blasts. Massive blast proliferation in the bone marrow interferes with the differentiation of normal hematopoietic cells, including the erythroid lineage, resulting in anemia due to suppressed erythropoiesis. In the gray scattergram group, a higher percentage of blasts contributes to more severe suppression of red cell formation, whereas in the non-gray scatttergram group, the lower blast burden allows relatively preserved bone marrow function. Thus, the differences in erythrocyte parameters reflect the severity of erythropoietic disruption caused by blast infiltration, with more prominent impairment observed in the gray scattergram group.14,15 In the gray scattergram group, the median blast count was 65% (range 2%–92%), markedly higher than in the non-gray scatttergram group at 12.5% (range 0%–82%). The leukocyte count was also significantly elevated in the gray scattergram group (54,595/µL) compared to the non-gray scatttergram group (5,505/µL, p<0.001), indicating more massive blast proliferation. Leukocytosis in acute leukemia results from clonal blast proliferation with impaired maturation, leading to the release of immature cells into the peripheral circulation and a markedly elevated white blood cell (WBC) count.14,15 There was no statistically significant difference in platelet counts between the two groups (p=0.126), and both groups showed thrombocytopenia. In leukemia, abnormal blast proliferation in the bone marrow disrupts normal hematopoiesis, including thrombopoiesis. The dominance of blasts in the marrow suppresses the differentiation and maturation of megakaryocytes, the precursor cells for platelets, leading to significantly reduced platelet production. This is consistent with findings by Wilson et al (2020), who reported that 74 patients with acute leukemia experienced thrombocytopenia (platelet count <50,000/mm3) and venous thromboembolic complications.16,17 In the study by Lismayanti and Arif, the median blast count on SADT in the gray scattergram group was 55% (range 2%–86%), while in the non-gray scatttergram group it was 16% (range 0%–47%). These findings are in line with our study, which showed a median blast count of 65% (2%-92%) in the gray scatttergram group and 12.5% (0%–82%) in the non gray scatttergram group, with a statistically significant difference (p<0.001). The higher blast counts in both groups compared to the Lismayanti and Arif study may be attributed to our inclusion of only newly diagnosed acute leukemia patients, where peripheral blast percentages are typically higher, especially at early presentation. In contrast, Lismayanti and Arif included all types of leukemia, including chronic subtypes that generally show lower blast percentages. This difference in study populations explains the higher median blast counts observed in our cohort.6 Our study focused exclusively on newly diagnosed acute leukemia patients (ALL and AML), whereas the study by Lismayanti and Arif included both acute and chronic leukemia cases. Therefore, the present study is not a subgroup analysis but represents an independent investigation with a distinct study population and objective. In addition, this study specifically evaluates the agreement between gray scattergram findings and microscopic blast counts using a defined diagnostic threshold, providing a more focused validation of its role as a screening tool in acute leukemia. Differences in sample population characteristics between the two studies are essential in interpreting scattergram results, particularly in patients with <35.5% blasts who still exhibited a gray scattergram. It is important to note that the 35.5% blast threshold used in this study does not represent the standard diagnostic criterion for acute leukemia, which is defined by ≥20% blasts according to WHO classification. Instead, this threshold was applied as a locally validated operational cut-off to evaluate the performance of gray scattergram findings. This occurs because blasts are large cells with high intracellular complexity and elevated RNA content. In hematology analyzers such as the Sysmex XN Series, these blasts generate abnormal scatter and fluorescence signals, distinguishing them from normal leukocyte clusters and causing them to appear as an unclassified population visible as a gray scattergram. Our study focused exclusively on patients with acute leukemia (ALL and AML), while the study by Lismayanti and Arif included all leukemia types, including chronic subtypes. As a result, peripheral blood blast burden and morphology differ between cohorts. Acute leukemia typically presents with a high percentage of immature blasts, whereas chronic leukemia generally has lower and more stable blast levels. This explains potential discrepancies in scattergram interpretation in patients with low blast counts.4,6,14 The distribution of blast percentages in the non-gray scattergram group also aligns with findings from previous studies. In the study by Lismayanti and Arif, 57.1% (12/21) of non-gray scattergram cases had <35.5% blasts, 23.8% (5/21) had ≥35.5% blasts, and 19.1% (4/21) had no detectable peripheral blasts. In our study, 68.2% (30/106) of non-gray scattergram cases had <35.5% blasts, 16.1% (10/106) had ≥35.5% blasts, and 5.67% (6/106) had no peripheral blasts. These findings are consistent with the fact that our cohort included patients at the initial diagnosis of acute leukemia, typically characterized by high peripheral blast counts. In contrast, Lismayanti and Arif included both acute and chronic leukemia cases, with bone marrow morphology used for diagnosis regardless of treatment status, resulting in more heterogeneous blast distributions. These population differences may explain the variations in gray scattergram and non-gray scattergram proportions.4,6 Nevertheless, both studies consistently demonstrated that the majority of non-gray scattergram cases had low or absent peripheral blasts, indicating limitations in scattergram sensitivity for detecting small blast populations. Scattergrams detect blasts and other leukocyte types on haematology analysers based on intracellular complexity (side scatter), cell size (forward scatter), and RNA content in the cytoplasm (side fluorescence).4 In addition to blasts, other atypical or unclassified cells such as abnormal lymphocytes, reactive lymphocytes, immature granulocytes, plasma cells, and nucleated red blood cells may also appear in the gray scattergram region. This reflects the limitation of hematology analyzers in classifying cells with atypical morphological or fluorescence characteristics. Blasts in leukaemia exhibit more active extracellular communication compared to normal cells, leading to an increase in lipid rafts on the cell membrane. This makes them more resistant to lysing reagents, preventing complete fluorescence staining of intracellular components, particularly nucleic acids in the nucleus. Blasts with a size and nuclear shape resembling lymphocytes and monocytes, and that stain with fluorescence, appear as non-grey scattergrams, while those that do not stain and fall outside the lymphocyte and monocyte areas appear as grey scattergrams.5,18 In this study, the most frequent WBC flagging codes in the gray scattergram group were “WBC Abn Scattergram”, “Monocytosis”, “Lymphocytosis”, “Abn Lympho?”, and “Blasts” (p<0.001). The “WBC Abn Scattergram” flag indicates abnormal leukocyte distribution on the scattergram. This flag is automatically generated as a warning that an atypical or unclassified cell population may be present. In pathological conditions such as leukemia, the presence of blasts alters the distribution due to their larger cell size and strong fluorescence. However, because blast fluorescence patterns and DNA complexity may resemble mature leukocytes, these cells can occupy transitional or intermediate zones, overlapping with monocyte or lymphocyte areas on the scattergram.14,18 In the non-gray scattergram group, the most frequent flags were “Blasts” and “Abn Lympho”, followed by “Monocytosis”, “Lymphocytosis”, “WBC Abn Scattergram”, and “Blast/Abn Lympho”. The presence of these flags in non-gray scattergrams suggests that blasts may still be present, even without a typical gray scattergram appearance. Previous studies have indicated that these flags may appear when blast counts are low, the patient is leukopenic, or the blast population overlaps with reactive lymphocytes in light scatter channels, preventing the formation of a typical gray scattergram pattern. The presence of false-negative results, in which non-gray scattergrams still demonstrate high blast counts, has important clinical implications. Reliance solely on scattergram patterns may lead to missed or delayed detection of significant blast populations, particularly in cases with low leukocyte counts or overlapping cell populations. Therefore, scattergram findings should always be interpreted in conjunction with WBC flagging parameters and confirmed by microscopic examination to ensure accurate and timely diagnosis.4,19 The kappa coefficient in this study was 0.527, indicating moderate agreement between the gray scattergram on the hematology analyzer and the presence of blasts by microscopic examination on peripheral blood smear. This finding supports the potential utility of the gray scattergram as a useful indicator for detecting blasts at higher concentrations. The kappa result is consistent with the cut-off values used by Lismayanti and Arif, who reported an area under the curve (AUC) of 0.872 and sensitivity of 76.2% (p=0.001).6 A sensitivity of 76.2% indicates that 76.2% of patients with ≥35.5% blasts on peripheral smear were correctly identified as having a gray scattergram. Although the role of scattergram has been described in various hematological and non-hematological conditions, most previous studies are descriptive. In contrast, our study provides quantitative validation by assessing the agreement between gray scattergram findings and microscopic blast counts, supporting its clinical utility as a screening tool in hematological malignancies. Most subjects with gray scattergrams had ≥35.5% blasts (83.9%), although 14 subjects (31.8%) had <35.5% blasts. Among these 14, five subjects (7.6%) had leukocyte counts <16,380/µL, while the remaining nine had leukocyte counts ranging from 23,150 to 235,230/µL. The presence of a gray scattergram is often associated with both high leukocyte and high blast counts. According to WHO criteria, a diagnosis of leukemia requires ≥20% blasts in peripheral blood or bone marrow. In this study, cases with blast counts ≥35.5% and leukocyte counts ≥16,380/µL typically exhibited gray scattergrams. However, variations in leukocyte counts among subjects with high blast levels influenced the scattergram appearance, resulting in a moderate kappa value of 0.527, reflecting the degree of agreement between analyzer findings and microscopic evaluation.4,6 The kappa coefficient in this study was 0.527, indicating moderate agreement between gray scattergram findings and microscopic blast identification. This suggests that gray scattergrams are not sufficient for definitive diagnosis but may serve as an early screening tool to detect abnormal or blast populations and support timely referral. However, confirmation through peripheral blood smear and bone marrow examination remains essential. Although both methods are observer-dependent, this was minimized by using concordant results from multiple observers. Additionally, scattergram interpretation is supported by automated WBC flagging systems, such as “Blasts” and “WBC Abn Scattergram”. Therefore, gray scattergrams should be considered a complementary tool that enhances early detection rather than replacing conventional morphological assessment. The agreement test between gray scattergram appearance and the presence of peripheral blasts in acute leukemia yielded a kappa value of 0.527, representing moderate concordance. Gray scattergrams in the WDF channel were more frequently observed in patients with higher leukocyte counts (>16,380/µL) and blast percentages (>35.5%). However, their presence in some patients with lower leukocyte counts or blast percentages indicates variability in analyzer interpretation. Therefore, gray scattergram findings should be interpreted in conjunction with WBC flagging parameters to enhance diagnostic accuracy. The “Blasts” flag demonstrated high sensitivity (97%) and specificity (96%), while the “Abn Lympho” flag showed sensitivity of 80% and specificity of 95%, supporting their role as complementary indicators in blast detection. Importantly, scattergram findings cannot replace microscopic peripheral blood smear examination, particularly in cases with leukopenia or low blast percentages, where abnormal cells may not produce a typical gray scattergram pattern. Thus, scattergrams should be considered as a supportive screening tool and must always be confirmed by morphological evaluation to avoid missed or delayed diagnosis. This study has several limitations. First, the retrospective design may limit control over data quality and completeness. Second, although the sample size was sufficient for agreement analysis, it may be underpowered for subgroup comparisons, and these findings should therefore be interpreted with caution. Third, the use of a 35.5% blast threshold represents a locally validated operational cut-off rather than a standard diagnostic criterion. Fourth, interobserver variability in smear interpretation was not formally assessed, although only concordant results were included to ensure data reliability. In addition, differences between manual peripheral blood smear preparation and automated slide preparation and staining systems may influence morphological interpretation. While automated systems provide standardized and reproducible slide quality, manual methods allow more detailed morphological assessment but may be subject to observer variability. These factors should be considered when interpreting blast detection results and generalizing the findings. Despite these limitations, this study provides additional evidence supporting the role of gray scattergram as a useful adjunctive screening tool for detecting blasts in acute leukemia. Conclusion In conclusion, the agreement analysis between the gray scattergram generated by the hematology analyzer and the microscopic identification of blasts on peripheral blood smears yielded a Kappa value of 0.527, indicating a moderate level of concordance. These findings suggest that while the gray scattergram provides valuable preliminary information, its interpretation should be integrated with additional diagnostic indicators, such as WBC flagging alerts, to enhance the accuracy of blast detection. However, gray scattergram findings cannot replace microscopic examination and should be considered as a complementary screening tool in clinical practice. Further studies with larger cohorts and across multiple hematology analyzer platforms are needed to validate these findings and improve their generalizability. Acknowledgments The authors like to acknowledge the Department of Clinical Pathology at Hasan Sadikin Hospital, Faculty of Medicine, Universitas Padjadjaran, Bandung, West Java, Indonesia, for its operational support. Abbreviations 5-diff, 5-differential; 6-diff, 6-differential; CBC, complete blood count; ICD-10, 10th revision of the International Classification of Diseases; EDTA,ethylenediaminetetraacetic acid; LIS, Laboratory Information System; WDF, White Blood Cell Differential; ALL, acute lymphoblastic leukemia; AML, acute myeloid leukemia; WBC, white blood cell. Disclosure The author(s) report no conflicts of interest in this study. References