Determinants of Preterm Birth among Women Referred from Community Health Facilities for Delivery: A Comparative Cross- Sectional Study

Abstract

Anemia is a significant health burden for pediatric patients, particularly those undergoing cancer treatment in resource-limited settings like Nepal. This study evaluated the diagnostic performance of automated hematology systems by Coulter against manual peripheral blood smear (PBS) analysis for classifying anemia types in a pediatric population. A total of 454 pediatric patients were included, comprising 274 oncology (60.4%) and 180 non-oncology (39.6%) cases. The overall mean hemoglobin was 10.76 ± 2.11 g/dL, with moderate anemia (7.0–9.9 g/dL) being the most prevalent category (118 patients). Anemia severity was concentrated in the non-oncology group (95 of 118 moderate; 19 of 20 severe), while oncology patients predominantly had normal hemoglobin (125 cases). Significant differences in red cell indices were observed: oncology patients had higher mean MCV (8.7 fL higher, p <0.001) and MCH (2.6 pg higher, p <0.001) compared to non-oncology patients. A highly significant association was found between anemia type and clinical status (χ² = 207.012, p <0.001). Normocytic profiles (normocytic normochromic anemia, NNA / normocytic normochromic picture, NNP) dominated the oncology group (51.7% NNP), consistent with anemia of chronic disease. In contrast, microcytic hypochromic anemia (MHA) was more common in the non-oncology group, consistent with iron deficiency. The automated Coulter diagnosis showed a strong agreement with manual PBS findings (r = 0.970, χ² = 922.47, p <0.001). In contrast, MHA was significantly more prevalent in the non-oncology group, suggesting iron deficiency. Given the strong agreement between the systems, the automated Coulter diagnosis serves as a reliable initial screening tool for anemia classification, but PBS confirmation remains essential to differentiate anemia types and guide effective clinical management in developing country laboratories.