Table 2.

Subgroup analysis of diagnostic performance of NLR

Covariate/Subgroup	Studies, n	Sensitivity(95%CI)	P-value	Specificity(95%CI)	P-value
Number of samples			0.05		0.75
> 130	4	0.58[0.31–0.85]		0.87[0.74-1.00]
≤ 130	8	0.83[0.72–0.95]		0.78[0.63–0.94]
Country			0.59		0.60
China	3	0.72[0.43-1.00]		0.79[0.54-1.00]
Other countries	9	0.78[0.63–0.93]		0.83[0.71–0.96]
Study design			0.24		0.75
Prospective	5	0.85[0.71–0.99]		0.83[0.68–0.99]
Retrospective and Clinical	5	0.61[0.35–0.87]		0.83[0.67-0.0.98]
Gold standard			0.95		0.03
Blood culture	4	0.80[0.58-1.00]		0.64[0.37–0.91]
Clinic, blood culture	6	0.76[0.56–0.96]		0.91[0.81-1.00]
Type of Sepsis			0.01		0.11
EOS&LOS	5	0.57[0.33–0.80]		0.72[0.50–0.94]
EOS	5	0.87[0.76–0.89]		0.89[0.78-1.00]
Specimen			0.94		0.68
Blood	11	0.76[0.62–0.90]		0.82[0.70–0.94]
Venous cord blood	1	0.83[0.48-1.00]		0.82[0.44-1.00]
cut-off			0.54		0.04
> 3	3	0.69[0.35-1.00]		0.96[0.90-1.00]
≤ 3	8	0.78[0.62–0.94]		0.72[0.57–0.87]

For all studies on NLR, the median cut off value is 2.12, and then the integer 3 is taken