Multi-criteria decision making to validate performance of RBC-based formulae to screen $$\beta$$ -thalassemia trait in heterogeneous haemoglobinopathies

Jain, Atul Kumar; Sharma, Prashant; Saleh, Sarkaft; Dolai, Tuphan Kanti; Saha, Subhas Chandra; Bagga, Rashmi; Khadwal, Alka Rani; Trehan, Amita; Nielsen, Izabela; Kaviraj, Anilava; Das, Reena; Saha, Subrata

doi:10.1186/s12911-023-02388-w

BMC Medical Informatics and Decision Making

Table 2 Performance analysis of discriminant formulae

From: Multi-criteria decision making to validate performance of RBC-based formulae to screen $\beta$-thalassemia trait in heterogeneous haemoglobinopathies

Study	ACC.	SE.	SP.	YI	AUC-ROC	PPV	NPV	FOR
S & B	0.884	0.348	0.977	0.325	0.662	0.719	0.897	0.101
E & F	0.870	0.212	0.983	0.196	0.598	0.685	0.879	0.119
Mentzer	0.898	0.464	0.973	0.437	0.718	0.745	0.913	0.084
RBC	0.871	0.433	0.947	0.380	0.690	0.583	0.907	0.089
S & L	0.783	0.950	0.754	0.705	0.852	0.399	0.989	0.008
RDW	0.660	0.011	0.772	-0.217	0.391	0.008	0.819	0.145
Ricerca	0.374	0.760	0.307	0.067	0.534	0.159	0.881	0.040
G & K	0.869	0.270	0.972	0.241	0.621	0.621	0.886	0.112
Das Gupta	0.695	0.527	0.724	0.251	0.626	0.247	0.899	0.075
Telmissani-MCHD	0.317	0.939	0.210	0.149	0.575	0.170	0.952	0.010
Telmissani-MDHL	0.859	0.164	0.979	0.143	0.571	0.571	0.872	0.126
Jayabose-RDWI	0.876	0.392	0.959	0.350	0.675	0.621	0.902	0.095
Huber-Herklotz	0.837	0.013	0.979	-0.008	0.496	0.097	0.852	0.145
Sirdah	0.889	0.338	0.984	0.322	0.661	0.787	0.896	0.102
Kerman-I	0.892	0.630	0.937	0.568	0.784	0.634	0.936	0.060
Kerman-II	0.899	0.488	0.969	0.457	0.729	0.731	0.917	0.081
Ehsani	0.898	0.475	0.971	0.446	0.723	0.740	0.915	0.083
Keikhaei	0.877	0.311	0.975	0.286	0.643	0.681	0.892	0.106
Wongprachum	0.846	0.286	0.942	0.229	0.614	0.460	0.885	0.109
Nishad	0.892	0.570	0.947	0.517	0.759	0.650	0.928	0.069
Sehgal	0.887	0.497	0.942	0.439	0.720	0.549	0.930	0.067
Sargolzie	0.832	0.269	0.929	0.198	0.599	0.396	0.881	0.112
Pornprasert	0.758	0.387	0.822	0.209	0.605	0.272	0.886	0.095
Sirachainan	0.631	0.111	0.721	-0.169	0.416	0.064	0.825	0.133
Bordbar	0.719	0.832	0.699	0.531	0.766	0.322	0.960	0.028
Hameed	0.150	0.992	0.006	-0.002	0.499	0.147	0.804	0.001
Hisham	0.882	0.351	0.973	0.324	0.662	0.694	0.897	0.100
Matos	0.825	0.302	0.915	0.217	0.609	0.380	0.884	0.107
Ravanbakhsh-F1	0.870	0.449	0.942	0.391	0.696	0.572	0.909	0.087
Ravanbakhsh-F2	0.690	0.330	0.752	0.082	0.541	0.186	0.867	0.103
Ravanbakhsh-F3	0.866	0.390	0.948	0.338	0.669	0.565	0.900	0.095
Ravanbakhsh-F4	0.818	0.891	0.805	0.697	0.848	0.441	0.977	0.018
Zaghloul-1	0.149	0.992	0.004	-0.003	0.498	0.146	0.763	0.001
Zaghloul-2	0.151	0.990	0.006	-0.004	0.498	0.146	0.774	0.002
Kandhro-1	0.372	0.191	0.403	-0.406	0.297	0.052	0.743	0.122
Kandhro-2	0.689	0.360	0.745	0.106	0.553	0.196	0.871	0.099
Merdin-1	0.864	0.585	0.912	0.497	0.748	0.533	0.927	0.067
Merdin-2	0.869	0.366	0.956	0.321	0.661	0.587	0.897	0.098
Cruise	0.363	0.755	0.296	0.051	0.526	0.156	0.876	0.040
Janel (11T)	0.887	0.283	0.991	0.274	0.637	0.840	0.889	0.110
Index26	0.893	0.332	0.989	0.321	0.660	0.839	0.896	0.103
SCS$_{BTT}$	0.726	0.974	0.684	0.658	0.829	0.346	0.993	0.004
Best	Kerman-II	Zaghloul-1, Hameed	Janel (11T)	S &L	S &L	Janel (11T)	SCS$_{BTT}$	Zaghloul-1, Hameed
Worst	Zaghloul-1	RDW	Zaghloul-1	Kandhro-1	Kandhro-1	RDW	Kandhro-1	RDW, Huber-Herklotz

RDW, Huber-Herklotz, Sirachainan, Hameed, Zaghloul-1, Zaghloul-2, and Kandhro-1 leads to negative YI, i.e., extensively high misclassification costs for those formulae

Back to article page

ISSN: 1472-6947

Contact us

General enquiries: journalsubmissions@springernature.com