Characteristics
|
TB-IRIS (45)
|
Non-TB-IRIS (154)
|
Total (199)
|
Crude ORb
|
P value
|
Adjusted ORc
|
P value
|
---|
N (%)
|
N (%)
|
N (%)
|
(95% CI)
|
(95% CI)
|
---|
Gender
|
Male
|
42 (93.3)
|
146 (94.8)
|
188 (94.5)
|
1
|
Ref.
| | |
Female
|
3 (6.7)
|
8 (5.2)
|
11 (5.5)
|
1.304 (0.331–5.133)
|
0.705
| | |
Age group (years)
|
< 25
|
9 (20.0)
|
16 (10.4)
|
25 (12.6)
|
1.744 (0.699–4.351)
|
0.233
|
2.024 (0.786–5.212)
|
0.144
|
25–44
|
30 (66.7)
|
93 (60.4)
|
123 (61.8)
|
1
|
Ref.
|
1
|
Ref.
|
45–64
|
5 (11.1)
|
42 (27.3)
|
47 (23.6)
|
0.369 (0.134–1.018)
|
0.054
|
0.355 (0.127–0.993)
|
0.049
|
> 64
|
1 (2.2)
|
3 (1.9)
|
4 (2.0)
|
1.033 (0.104–10.310)
|
0.978
|
1.603 (0.156–16.479)
|
0.692
|
Initial CD4 cell count (cells/mm3)
|
> 100
|
2 (4.4)
|
24 (15.6)
|
26 (13.1)
|
1
|
Ref.
| | |
50–100
|
3 (6.7)
|
25 (16.2)
|
28 (14.1)
|
1.440 (0.221–9.388)
|
0.703
| | |
< 50
|
40 (88.9)
|
105 (68.2)
|
145 (72.9)
|
4.571 (1.033–20.238)
|
0.045
| | |
CD4 cell count after ART (cells/mm3)
|
> 100
|
21 (46.7)
|
71 (46.1)
|
92 (46.2)
|
1
|
Ref.
| | |
50–100
|
16 (35.6)
|
42 (27.3)
|
58 (29.1)
|
1.288 (0.606–2.738)
|
0.511
| | |
< 50
|
8 (17.8)
|
41 (26.6)
|
49 (24.6)
|
0.660 (0.268–1.623)
|
0.365
| | |
Increase in CD4 cell count
|
< 4 folds
|
21 (46.7)
|
103 (66.9)
|
124 (62.3)
|
1
|
Ref.
|
1
|
Ref.
|
≥ 4 folds
|
24 (53.3)
|
51 (33.1)
|
75 (37.7)
|
2.308 (1.175–4.533)
|
0.015
|
2.614 (1.288–5.303)
|
0.008
|
HIV virial load (copies/mL)
|
≤ 1000 copies
|
24 (53.3)
|
94 (61.0)
|
118 (59.3)
|
1
|
Ref.
| | |
> 1000 copies
|
21 (46.7)
|
60 (39.0)
|
81 (40.7)
|
1.371 (0.702–2.677)
|
0.356
| | |
HAART regimena
|
First-line
|
38 (84.4)
|
136 (88.3)
|
174 (87.4)
|
1
|
Ref.
| | |
Second-line
|
7 (15.6)
|
18 (11.7)
|
25 (12.6)
|
1.392 (0.541–3.578)
|
0.493
| | |
- aFirst-line regimen includes tenofovir, lamivudine and efavirenz; Second-line regimen includes tenofovir, lamivudine and lopinavir/ritonavir
- bOR, odds ratio; CI, confidential interval
- cThe initial CD4 cell count is considered as a confounding variable of increase in CD4 cell count, which is removed in the multivariate analysis