Application of metagenomic next-generation sequencing in the diagnosis of infectious diseases of the central nervous system after empirical treatment

Table 1 The performance of metagenomic next-generation sequencing and the conventional methods in the diagnosis of central nervous system virus infections

	Sensitivity	Specificity	Accuracy	Positive predictive value	Negative predictive value
Conventional methods+	53.5% (42.7%, 64.2%)	85.7% (69.0%, 94.6%)	62.8% (53.5%, 71.3%)	90.2% (77.8%, 96.3%)	42.9% (31.3%, 55.2%)
mNGS+	66.3% (55.2%, 75.9%)	88.6% (72.3%, 96.3%)	72.7% (63.7%, 80.2%)	93.4% (83.3%, 97.9%)	51.7% (38.5%, 64.6%)
P value	0.087	1.000	0.099	0.779	0.316

Table 2 Inconsistency between metagenomic next-generation sequencing and conventional methods in diagnosing central nervous system virus infections

mNGS	Conventional methods (+, -)	Total
+	38, 19	57
_	7, 22	29
Total	45, 41	86

mNGS: Metagenomic next-generation sequencing; +: Positive; −: Negative.

Table 3 Performance of metagenomic next-generation sequencing and the conventional methods in the diagnosis of central nervous system bacterial infections

	Sensitivity	Specificity	Accuracy	Positive predictive value	Negative predictive value
Conventional methods+	14.3% (8.5%, 23.6%)	84.6% (64.3%, 95.0%)	33.3% (24.2%, 43.8%)	71.4% (42.0%, 90.4%)	26.8% (17.9%, 37.9%)
mNGS+	65.7% (53.3%, 76.4%)	88.5% (68.7%, 97.0%)	71.9% (61.6%, 80.3%)	93.9% (82.1%, 98.4%)	48.9% (34.3%, 63.7%)
P value	< 0.001	1.000	< 0.001	0.061	0.011

Table 4 Inconsistency of metagenomic next-generation sequencing and conventional methods in diagnosing central nervous system bacterial infections

mNGS	Conventional test (+, -)	Total
+	8, 37	45
_	2, 23	25
Total	10, 60	70

mNGS: Metagenomic next-generation sequencing; +: Positive; −: Negative.

Table 5 Performance of metagenomic next-generation sequencing and the conventional methods in the diagnosis of central nervous system fungal infections

	Sensitivity	Specificity	Accuracy	Positive predictive value	Negative predictive value
Conventional methods	44.4% (26.0%, 64.4%)	83.3% (36.5%, 99.1%)	51.5% (33.9%, 68.8%)	92.3% (62.1%, 99.6%)	25.0% (9.6%, 49.4%)
mNGS	63.0% (42.5%, 79.9%)	100.0% (51.7%, 100.0%)	69.7% (51.1%, 83.8%)	100.0% (77.1%, 100.0%)	37.5% (16.3%, 64.1%)
P value	0.172	1.000	0.131	0.433	0.656

Table 6 Inconsistency between metagenomic next-generation sequencing and Conventional methods in diagnosing central nervous system fungal infections

mNGS	Conventional test (+, -)	Total
+	7, 10	17
_	5, 5	10
Total	12, 15	27

mNGS: Metagenomic next-generation sequencing; +: Positive; −: Negative.

Table 7 The results of meningitis in all patients were compared between the two methods

	Sensitivity	Specificity	Accuracy	Positive predictive value	Negative predictive value
Conventional methods	37.2% (30.2%, 44.6%)	85.1% (73.8%, 92.2%)	50.0% (43.7%, 56.3%)	87.2% (77.2%, 83.3%)	33.1% (26.3%, 40.8%)
mNGS	65.6% (58.2%, 72.3%)	89.6% (79.1%, 95.3%)	72.0% (65.9%, 77.4%)	94.5% (88.6%, 97.6%)	48.8% (39.7%, 57.9%)
P value	< 0.001	0.436	< 0.001	0.065	0.007

mNGS: Metagenomic next-generation sequencing.