Xu M, Lin G, Dong Z, Wang Q, Ma L, Su J. Logistic-Nomogram model based on red blood cell parameters to differentiate thalassemia trait and iron deficiency anemia in southern region of Fujian Province, China.
J Clin Lab Anal 2023;
37:e24940. [PMID:
37386931 PMCID:
PMC10431415 DOI:
10.1002/jcla.24940]
[Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/07/2023] [Revised: 06/01/2023] [Accepted: 06/19/2023] [Indexed: 07/01/2023] Open
Abstract
BACKGROUND
Differentiation between thalassemia trait (TT) and iron deficiency anemia (IDA) is challenging and costly. This study aimed to construct and evaluate a model based on red blood cell (RBC) parameters to differentiate TT and IDA in the southern region of Fujian Province, China.
METHODS
RBC parameters of 364 TT patients and 316 IDA patients were reviewed. RBC parameter-based Logistic-Nomogram model to differentiate between TT and IDA was constructed by multivariate logistic regression analysis plus nomogram, and then compared with 22 previously reported differential indices.
RESULTS
The patients were randomly selected to a training cohort (nTT = 248, nIDA = 223) and a validation cohort (nTT = 116, nIDA = 93). In the training cohort, multivariate logistic regression analysis identified RBC count, mean corpuscular hemoglobin (MCH), and MCH concentration (MCHC) as independent parameters associated with TT susceptibility. A nomogram was plotted based on these parameters, and then the RBC parameter-based Logistic-Nomogram model g (μy ) = 1.92 × RBC count-0.51 × MCH + 0.14 × MCHC-39.2 was devised. The area under the curve (AUC) (95% CI) was 0.95 (0.93-0.97); sensitivity and specificity at the best cutoff score (120.24) were 0.93 and 0.89, respectively; the accuracy was 0.91. In the validation cohort, the RBC parameter-based Logistic-Nomogram model had AUC (95% CI) of 0.95 (0.91-0.98); sensitivity and specificity were 0.92 and 0.87, respectively; accuracy was 0.90. Moreover, compared with 22 reported differential indices, the RBC parameter-based Logistic-Nomogram model showed numerically higher AUC, net reclassification index, and integrated discrimination index (all p < 0.001).
CONCLUSION
The RBC parameter-based Logistic-Nomogram model shows high performance in differentiating patients with TT and IDA from the southern region of Fujian Province.
Collapse