Table 3 Univariate analysis of ocular features recorded in the electronic health record system.

From: Evaluation of machine learning algorithms for trabeculectomy outcome prediction in patients with glaucoma

Features

Surgical failure (\(\hbox {n}=126\))

Surgical success (\(\hbox {n}=104\))

P value

(19) History of angle surgery (n, %)

11 (8.7)

9 (8.7)

1.000

(20) History of selective laser trabeculoplasty (SLT) (n, %)

44 (34.9)

33 (31.7)

0.712

(21) History of previous trabeculectomy (n, %)

5 (4.0)

5 (4.8)

1.000

(22) Pseudophakia (n, %)

43 (34.1)

36 (34.6)

1.000

(23) Diabetes mellitus (DM) (n, %)

36 (28.6)

35 (33.7)

0.492

(24) Study eye

  

1.000

   Left (n, %)

63 (50.0)

52 (50.0)

 

   Right (n, %)

63 (50.0)

52 (50.0)

 

Number of pre-operative topical glaucoma medications

(25) (0, 1) (n, %)

7 (5.6)

10 (9.6)

0.359

(26) (2, 3) (n, %)

82 (65.1)

71 (68.3)

0.712

(27) (4, 5) (n, %)

37 (29.4)

23 (22.1)

0.273

(28) Topical beta blocker (n, %)

92 (73.0)

74 (71.2)

0.868

(29) Topical prostaglandin analogue (PGA) (n, %)

103 (81.7)

73 (70.2)

0.057

(30) Topical alpha-agonist (n, %)

62 (49.2)

53 (51.0)

0.895

(31) Topical carbonic anhydrase inhibitor (CAI) (n, %)

67 (53.2)

51 (49.0)

0.623

(32) Oral CAI (n, %)

17 (13.5)

10 (9.6)

0.482

(33) Angle anatomy

  

0.647

   Open angle glaucoma (OAG) (n, %)

113 (89.7)

96 (92.3)

 

   Angle closure glaucoma (ACG) (n, %)

13 (10.3)

8 (7.7)

 

(34) Pre-operative visual acuity (VA) (Mean, SD)

0.49 (0.6)

0.46 (0.7)

0.720

Pre-operative IOP

(35) < 18 mmHg (n, %)

42 (33.3)

30 (28.8)

0.557

(36) \(>=\) 18 and \(<= 25\) (n,%)

40 (31.7)

34 (32.7)

0.991

(37) \(> 25\) (n,%)

44 (34.9)

40 (38.5)

0.676

(38) Surgery-trabeculectomy with cataract extraction (n, %)

21 (16.7)

13 (12.5)

0.484

(39) Central corneal thickness (CCT) (Mean, SD)

554.9 (48.3)

548.8 (45.2)

0.326

  1. \(\hbox {n}= \hbox {number}; \hbox {SD}= \hbox {standard deviation}\).
  2. The threshold for statistical significance was P \(< .05\).