Spectrum of Bone Marrow Pathological Findings in Patients Undergoing Bone Marrow Biopsy: A Cross-Sectional Study.

Introduction Bone marrow biopsy is a diagnostic cornerstone in the evaluation of unexplained cytopenias and suspected haematological neoplasms; however, diagnostic frequencies vary widely across populations. This study describes the spectrum of bone marrow pathology recorded in a single-centre registry and explores, as a secondary exploratory aim, whether it varies by sex and age. Methods  This retrospective cross-sectional study analysed 181 bone marrow examination records, which constituted the unit of analysis; no patient identifiers were available to confirm unique-patient status. Diagnoses were classified into 13 registry-derived categories using a prespecified, fully documented classification dictionary and subsequently collapsed into seven broad groups. The association between sex and diagnostic category was assessed using the chi-square test and the Fisher-Freeman-Halton exact test because of sparse cell counts. Age was compared across diagnostic groups using the Kruskal-Wallis test as the prespecified primary analysis because of non-normality, with ANOVA performed as a sensitivity analysis, followed by Dunn's post hoc test. Results  The median recorded age was 30 years (IQR, 20-50 years), and 66.9% of the records were from males. Acute leukaemia was the most frequent diagnosis (49.2%; 95% CI, 42.0%-56.4%), followed by aplastic anaemia/bone marrow failure syndrome (14.4%), inconclusive/deferred opinions (12.2%), non-neoplastic/reactive marrow (8.3%), myelodysplastic/myeloproliferative overlap (6.1%), lymphoproliferative/plasma cell neoplasms (6.1%), and other diagnoses (3.9%). Sex was not statistically significantly associated with diagnostic category (χ² = 5.995, df = 6, p = 0.424; exact p = 0.475; Cramér's V = 0.182). Males accounted for the majority of records in most, but not all, groups; females constituted the majority in the non-neoplastic/reactive marrow group (53.3%). Age differed significantly across the groups (Kruskal-Wallis H = 40.12, df = 6, p < 0.001, epsilon-squared = 0.196). The myelodysplastic/myeloproliferative overlap group had the highest median age (62 years), whereas the aplastic anaemia/bone marrow failure syndrome group had the lowest median age (20 years). Conclusions  Acute leukaemia predominated in this referral-heavy registry sample. Recorded age varied significantly by diagnostic category, whereas sex was not statistically significantly associated with diagnostic category. A substantial proportion of acute leukaemia records lacked a documented lineage subtype. Limitations related to the unit of analysis and sampling period are detailed in the text; therefore, the findings should be interpreted as descriptive and hypothesis-generating.