Table 4.

Multivariable analysis of factors associated with a CDT of BMA in the subpopulation with no hematological malignancy or cancer known on admission (N = 160)

Variable	Odds ratio	95% confidence interval	P value
First model
Non-malignant hematological abnormality known on admission	7.75	[1.77–34.02]	0.007
Indication of BMA excluding isolated TPa,b	3.32	[1.16–9.51]	0.025
SOFA scorec	1.19	[1.06–1.34]	0.003
Pre-BMA HScorec,d	1.03	[1.01–1.04]	< 0.001
Second model
Non-malignant hematological abnormality known on admission	6.76	[1.51–30.38]	0.013
Indication of BMA excluding isolated TPa,b	3.82	[1.19–12.24]	0.024
SOFA—platelet-count SOFA subscorec	1.19	[1.04–1.35]	0.009
Pre-BMA HScorec,d	1.03	[1.01–1.04]	< 0.001
Platelet count SOFA subscore 0 versus other groups	2.14	[0.60–7.62]	0.239

BMA bone marrow aspiration, CDT contribution to diagnosis and/or treatment, Hscore reactive hemophagocytic syndrome diagnostic score, SOFA sequential organ failure assessment, TP thrombocytopenia

^aThrombocytopenia may be present or absent in these patients

^bIsolated thrombocytopenia, i.e., thrombocytopenia was the only indication for BMA

^cPer point

^dCalculated with no points assigned for the cytological variable