Table 4.

Subgroup analysis for the predicting value of ROX

Subgroups	Number of articles	Sen	Spe	PLR	NLR	DOR	AUC
Overall studies	9	0.67 (0.57–0.76)	0.72 (0.65–0.78)	2.4 (2.0–2.80)	0.46 (0.37–0.58)	5 (4–7)	0.75 (0.71–0.79)
Population
COVID-19	4	0.71 (0.56–0.82)	0.73 (0.63–0.81)	2.60 (2.10–3.30)	0.40 (0.28–0.56)	7 (5–9)	0.78 (0.74–0.82)
Other population	5	0.63 (0.48–0.75)	0.71 (0.60–0.80)	2.10 (1.70–2.70)	0.53 (0.40–0.70)	4 (3–6)	0.72 (0.68–0.76)
Acquisition time
6 h after receiving HFNC	6	0.67 (0.54–0.78)	0.70 (0.60–0.78)	2.20 (1.80–2.80)	0.47 (0.35–0.63)	5 (3–7)	0.74 (0.70–0.78)
Other time	3	0.73 (0.67–0.79)	0.74 (0.70–0.78)	2.84 (2.39–3.39)	0.36 (0.29–0.45)	7 (5–10)	0.79 (0.75–0.83)
Cut-off value
Cut-off ≤ 5	5	0.67 (0.65–0.76)	0.71 (0.61–0.80)	2.3 (1.8–2.9)	0.48 (0.37–0.63)	5 (3–7)	0.74 (0.70–0.78)
Cut-off > 5	4	0.59 (0.54–0.65)	0.83 (0.79–0.86)	3.5 (2.78–4.43)	0.49 (0.43–0.56)	7 (5–9)	0.78 (0.74–0.82)

Sen sensitivity; Spe specificity; PLR positive likelihood ratio; NLR negative likelihood ratio; DOR diagnostic odds ratio; AUC area under curve; HFNC high flow nasal cannula