Table 1
Demographic and clinical features of the participants
|
Development Cohort
|
Validation Cohort
|
Diagnosis
|
Not or non-reactive NS
|
Reactive NS
|
P value
|
Not or non-reactive NS
|
Reactive NS
|
P value
|
N
|
44
|
104
|
|
40
|
27
|
|
Sex, n (%)
|
|
|
0·219
|
|
|
0·432
|
Male
|
33 (75·0)
|
87 (83·7)
|
|
26 (65·0)
|
20 (74·1)
|
|
Female
|
11 (25·0)
|
17 (16·3)
|
|
14 (35·0)
|
7 (25·9)
|
|
Age, mean (sd), year
|
46·4 (13·7)
|
47·3 (11·8)
|
0·694
|
48·75 (13·40)
|
52·96 (16·32)
|
0·252
|
Height, mean (sd), cm
|
164·3 (7·8)
|
166·5 (7·7)
|
0·208
|
165·50 (8·46)
|
166·63 (7·98)
|
0·585
|
Weight, mean (sd), kg
|
61·4 (9·6)
|
62·9 (11·4)
|
0·542
|
71·13 (20·89)
|
68·92 (12·59)
|
0·642
|
Education level, n (%)
|
|
|
0·110
|
|
|
0·197
|
High school or beyond
|
19 (44·2)
|
30 (30·3)
|
|
13 (32·5)
|
13 (48·1)
|
|
Beyond high school
|
24 (55·8)
|
69 (69·7)
|
|
27 (67·5)
|
14 (51·9)
|
|
Address, n (%)
|
|
|
0·088
|
|
|
0·202
|
City
|
24 (55·8)
|
33 (33·0)
|
|
9 (22·5)
|
5 (18·5)
|
|
Urban-rural fringe area
|
5 (11·6)
|
14 (14·0)
|
|
6 (15·0)
|
5 (18·5)
|
|
Village and town
|
14 (32·6)
|
47 (47·0)
|
|
18 (45·0)
|
17 (63·0)
|
|
Other province of China
|
0 (0·0)
|
4 (4·0)
|
|
6 (15·0)
|
0 (0·0)
|
|
Aboard
|
0 (0·0)
|
2 (2·0)
|
|
1 (2·5)
|
0 (0·0)
|
|
Treatment before, n (%)
|
15 (43·9)
|
28 (27·7)
|
0·207
|
22 (55·0)
|
12 (44·4)
|
0·347
|
Clinical symptoms, n (%)
|
|
|
|
|
|
|
No symptom
|
6 (13·6)
|
8 (7·7)
|
0·259
|
5 (12·5)
|
3 (11·1)
|
0·863
|
Psychiatric symptom
|
9 (20·5)
|
52 (50·0)
|
<0·001
|
8 (20·0)
|
7 (25·9)
|
0·568
|
Memory change
|
9 (20·5)
|
40 (38·5)
|
0·033
|
7 (17·5)
|
5 (18·5)
|
0·915
|
Sleep difficulty
|
5 (11·4)
|
23 (22·1)
|
0·127
|
1 (2·5)
|
1 (3·7)
|
0·776
|
Photophobia
|
6 (13·6)
|
8 (7·7)
|
0·259
|
2 (5·0)
|
1 (3·7)
|
0·801
|
Blurred Version
|
4 (9·1)
|
6 (5·8)
|
0·462
|
3 (7·5)
|
1 (3·7)
|
0·520
|
TP CLIA, mean (sd)
|
-
|
-
|
-
|
57·3 (42·0)
|
93·3 (49·9)
|
0·002
|
1 < X <100, n (%)
|
7 (16·3)
|
1 (1·0)
|
|
24 (60·0)
|
7 (25·9)
|
|
X > 100, n (%)
|
36 (83·7)
|
99 (99·0)
|
|
16 (40·0)
|
20 (74·1)
|
|
Creatine kinase, (sd), U/L
|
106·5 (137·3)
|
196·5 (373·0)
|
0·132
|
-
|
-
|
-
|
Serum TRUST, n (%)
|
|
|
<0·001
|
|
|
0·008
|
Negative
|
9 (22·5)
|
3 (3·0)
|
|
15 (37·5)
|
1 (4·0)
|
|
1
|
14 (35·0)
|
9 (9·0)
|
|
19 (47·5)
|
11 (44·0)
|
|
8
|
5 (12·5)
|
7 (7·0)
|
|
4 (10·0)
|
5 (20·0)
|
|
16
|
6 (15·0)
|
18 (18·0)
|
|
1 (2·5)
|
3 (12·0)
|
|
32
|
4 (10·0)
|
30 (30·0)
|
|
0 (0·0)
|
1 (4·0)
|
|
≥64
|
2 (5·0)
|
33(33·0)
|
|
1 (2·5)
|
4 (16·0)
|
|
Serum TPPA > 1:320, n (%)
|
36 (81·8)
|
96 (92·3)
|
0·060
|
35 (87·5)
|
25 (96·2)
|
0·130
|
Note: sd, standard deviation; Q, Quartile; Neurosyphilis, NS; TP CLIA, Treponema pallidum chemiluminescence assay; TRUST, Toluidine red unheated serum test; TPPA, Treponema pallidum particle agglutination assay; Immunoglobulin, IgG. *Data with normal distribution were described using mean (sd). |
Demographic and clinical features
The demographic and clinical features of the 215 clinically suspected patients with NS with complete information of treponemal, non-treponemal serum, and CSF examination are summarised in Table 1. Among the 148 patients in the development cohort, 83·7% were male; 14 (21·3%) were non-symptomatic patients and required a CSF test. Further, 105 patients (70·9%) did not receive any prior antibiotic treatment. The most common symptoms of NS were psychiatric symptoms (50·0% in the development and 25·9% in the validation cohort). Other symptoms such as sleep disorders, photophobia, and blurred version were also observed (Supplement 1).
Table 2
Univariate logistic regression analysis of each indicator and diagnostic in the development cohort
|
Univariable Analysis
|
Characteristic
|
Odds Ratio
|
95% CI
|
P value
|
Male
|
1·71
|
(0·72, 4·02)
|
0·222
|
Age
|
1·01
|
(0·98, 1·04)
|
0·691
|
Treatment before
|
0·53
|
(0·25, 1·11)
|
0·094
|
Psychiatric symptoms
|
3·89
|
(1·70, 8·89)
|
0·001
|
No symptom
|
0·53
|
(0·17, 1·62)
|
0·263
|
Serum TPPA > 1:320
|
2·67
|
(0·93, 7·64)
|
0·068
|
Serum TRUST
|
|
|
|
Negative
|
Ref
|
|
|
1
|
1·93
|
(0·41, 9·10)
|
0·407
|
8
|
4·20
|
(0·74, 23·91)
|
0·106
|
16
|
9·00
|
(1·82, 44·59)
|
0·007
|
32
|
22·50
|
(4·23, 119·77)
|
0·000
|
64
|
49·50
|
(7·15, 342·77)
|
<0·0001
|
Serum creatine kinase
|
1·00
|
(1·00, 1·01)
|
0·109
|
Serum IgG
|
1·03
|
(0·90, 1·18)
|
0·687
|
Serum albumin
|
0·95
|
(0·88, 1·04)
|
0·284
|
CSF Protein
|
6·51
|
(1·77, 23·96)
|
0·005
|
<=0·45
|
Ref
|
|
|
>0·45
|
4·80
|
(2·22, 10·36)
|
<0·0001
|
CSF Glucose
|
0·66
|
(0·39, 1·09)
|
0·105
|
CSF Nucleated cells
|
1·00
|
(0·99, 1·01)
|
0·899
|
<=5
|
Ref
|
|
|
>5
|
2·89
|
(1·38, 6·08)
|
0·005
|
TPPA, Treponema pallidum particle agglutination assay; TRUST, Toluidine red unheated serum test; Cerebrospinal fluid, CSF; Immunoglobulin, IgG. *Data with normal distribution was described using mean (sd). |
Laboratory findings and diagnostic yield
The univariable logistic regression analysis revealed significant differences in psychiatric symptoms between the reactive NS group and control group (odds ratio [OR], 3·89; 95% confidence interval [CI], 1·70-8·89, P = 0·0001; Table 2). The serum TRUST titres above 1:8 showed associations with NS diagnosis. However, no difference in creatine kinase levels was observed between the confirmed reactive NS and control groups (OR, 1·0; 95% CI, 1·00-1·01). Moreover, there were no significant differences between these two groups in CSF protein >0·45 g/L (OR, 4·80; 95% CI, 2·22-10·36, P < 0.0001) and CSF nucleated cells > 5 cells/L (OR, 2·89; 95% CI, 1·38-6·08, P = 0·005).
Table 3
Multivariable logistic analysis for the construction of diagnostic models
|
Development Cohort
|
|
Validation Cohort
|
|
Multivariate
|
P
|
Multivariate
|
P
|
|
Male
|
1·0 (0·3, 3·1)
|
0·954
|
1·5 (0·5, 4·5)
|
0·43
|
|
Psychiatric symptoms
|
5·1 (1·7, 15·4)
|
0·004
|
1·4 (0·4, 4·5)
|
0·57
|
|
Serum TRUST
|
|
|
|
|
|
Negative
|
Ref
|
|
Ref
|
|
|
1
|
3·2 (0·6, 18·8)
|
0·194
|
8·7 (1·0, 75·0)
|
0·05
|
|
8
|
7·9 (1·1, 54·5)
|
0·036
|
18·8 (1·7, 209·6)
|
0·02
|
|
16
|
16·0 (2·6, 97·7)
|
0·003
|
45·0 (2·2, 937·4)
|
0·01
|
|
32
|
33·6 (5·4, 209·7)
|
<0·001
|
inf· (0·0, Inf)
|
0·99
|
|
64
|
71·9 (9·1, 570·0)
|
<0·001
|
60·0 (3·0, 1185.1)
|
0.01
|
|
TRUST, Toluidine red unheated serum test; Immunoglobulin, IgG; Cerebrospinal fluid, CSF Multivariable logistic analysis indicted in Table 3 that the following factors were more likely related to confirmed NS: serum TRUST 1:8 (OR, 7·9; 95% CI, 1·1-54·5), 1:16 (OR, 16·0; 95% CI,]·6-97·7), 1:32 (OR, 33·6; 95% CI, 5·4-209·7), and 1:64 (OR, 71·9; 95% CI, 9·1-570·0), psychiatric symptoms (OR, 5·1; 95% CI, 1·7-15·4) and male (OR, 1·00; 95% CI, 0·3-3 ·1).
Table 4
Accuracy of the diagnostic score of the nomogram for estimating the risk of reactive neurosyphilis
|
Value
|
Variable
|
Development Cohort
|
Validation Cohort
|
ROC area (AUC) 95% CI
|
0·85 (0·77, 0·91)
|
0·85 (0·74, 0·93)
|
Specificity, %
|
80·0
|
77·5
|
Sensitivity, %
|
79·0
|
81·5
|
Positive predictive value, %
|
90·8
|
71·0
|
Negative predictive value, %
|
60·3
|
86·1
|
Receiver operating characteristics (ROC) curve of the model was generated using gender, psychiatric symptoms, and serum TRUST for neurosyphilis diagnosis. Diagnose-odds ratio (OR), positive predictive value (pv), and negative pv were calculated. *Area Under Curve (AUC) confidence interval and significance test adopted 500-time bootstrap resampling
The sensitivity and specificity of the diagnostic model in the development cohort were 80·0 and 79·0, respectively, and those in the validation cohort were 77·5 and 81·5, respectively (Table 4). The receiver operating characteristics (ROC) curve of the model was similar in the development (0·85, 95% CI: 0·77-0·91) and validation cohorts (0·85, 95% CI: 0·74-0·93).
Figure 1 illustrates the nomogram with multi-variables for the diagnostic probability of reactive NS (Figure, A). The nomogram demonstrated good accuracy in estimating the diagnostic probability of reactive NS, with a bootstrap-corrected C index of 0·83 (Figure, B). The calibration plots were graphically good in the validation cohort; the nomogram displayed a C index of 0·71 for the estimation of reactive NS diagnosis (Figure, C).