Symptom-based scoring technique by machine learning to predict COVID-19: a validation study

BMC Infectious Diseases

Table 2 Diagnostic performances

Setting	Test	Contingency (n)				Evaluation metric ± 95% CI (%⁾^a
Setting	Test	TPs	FNs	FPs	TNs	Sens	Spec ^b	PPV	NPV ^b
Emergency department	Index		10		9				48 ± 4
	Index or comparator 1	2			3
	Comparator 1	2			3
	Comparator 2
Referral center	Index	30	150	5	105	17 ± 2	95 ± 1	85 ± 2	42 ± 2
	Index or comparator 1	55	7	5	2	88 ± 1	30 ± 6	92 ± 2	21 ± 3
	Comparator 1	29	12		6	71 ± 2			32 ± 3
	Comparator 2	24	33	6	50	39 ± 3	90 ± 2	79 ± 3	17 ± 2
	Comparator 3 ^c	327		203				61 ± 1

^a Interval estimates by 30 times bootstrapping; ^b patients with a negative outcome were likely underrepresented in the referral center; ^c under the assumption of positive COVID-19 by clinical judgment from primary care. CI Confidence interval, FNs False negatives, FPs False positives, NPV Negative predictive value, PPV Positive predictive value, Sens. Sensitivity, Spec. specificity, TNs True negatives, TPs, True positives

ISSN: 1471-2334