Skip to main content

Table 2 Relevance of the automatically extracted sentences

From: Creating efficiencies in the extraction of data from randomized trials: a prospective evaluation of a machine learning and text mining tool

Report section

Data element

Reported in the trial, n (%)a

Found by Exact, n (%)b

Relevance, top sentence, n (%)c

Relevance, any sentence, n (%)c

Relevant sentences, n (%) of Total

Publication information

First author name

75 (100.0)

74 (98.7)

63 (85.1)

n/a

n/a

Date of publication

75 (100.0)

74 (98.7)

64 (86.5)

n/a

n/a

Digital object identifier

75 (100.0)

72 (96.0)

62 (82.7)

n/a

n/a

Meta information

Funding source

63 (84.0)

58 (92.1)

45 (77.6)

50 (86.2)

79/116 (68.1)

Funding number

29 (38.7)

22 (75.9)

20 (90.9)

22 (100.0)

35/110 (31.8)

Registration number

52 (69.3)

40 (76.9)

40 (100.0)

40 (100.0)

63/200 (31.5)

Enrollment

Eligibility criteria

75 (100.0)

75 (100.0)

38 (50.7)

47 (62.7)

110/375 (29.3)

Sample size

75 (100.0)

68 (90.7)

32 (47.1)

43 (63.2)

125/340 (36.8)

Enrollment start date

45 (60.0)

44 (97.8)

35 (79.5)

44 (100.0)

55/220 (25.0)

Enrollment end date

45 (60.0)

45 (100.0)

35 (77.8)

44 (97.8)

56/225 (24.9)

Early stopping

4 (5.3)

2 (50.0)

2 (100.0)

2 (100.0)

7/10 (70.0)

Intervention

Experimental arm(s)

75 (100.0)

74 (98.7)

43 (58.1)

65 (87.8)

123/370 (33.2)

Control arm(s)

75 (100.0)

75 (100.0)

49 (65.3)

65 (86.7)

121/375 (32.3)

Route of administration

29 (38.7)

14 (48.3)

12 (85.7)

14 (100.0)

32/70 (45.7)

Dose

37 (49.3)

32 (86.5)

19 (59.4)

28 (87.5)

50/160 (31.3)

Frequency of administration

43 (57.3)

28 (65.1)

23 (82.1)

27 (96.4)

45/140 (32.1)

Duration of treatment

55 (73.3)

41 (74.5)

25 (61.0)

30 (73.2)

57/205 (27.8)

Outcome

Primary outcome(s)

75 (100.0)

75 (100.0)

53 (70.7)

62 (82.7)

95/375 (25.3)

Primary outcome time point

74 (98.7)

50 (67.6)

27 (54.0)

39 (78.0)

76/250 (30.4)

Secondary outcome(s)

55 (73.3)

44 (80.0)

33 (75.0)

40 (90.9)

75/220 (34.1)

Secondary outcome time point

54 (72.0)

23 (42.6)

15 (65.2)

19 (82.6)

43/115 (37.4)

Summary measure

Median (IQR), n

55 (45 to 75)

44 (23 to 68)

35 (23 to 45)

40 (23 to 44)

57 (37 to 91)

Median (IQR), %

73.3 (60.0 to 100.0)

90.7 (74.5 to 98.7)

77.6 (61.0 to 85.1)

87.7 (82.6, 99.5)

32.0 (29.3 to 36.1)

  1. IQR  Interquartile range, n/a Not applicable (ExaCT presents only one solution for these elements). Values in italics typeface fall at or below the limit of the lowest quartile
  2. aAs identified during manual data extraction and verification
  3. bPertains to the studies where the data element was identified as reported in the study by the human reviewers (denominator, column 3)
  4. cPertains to the studies where the data element was correctly identified as reported in the study by ExaCT (denominator, column 4)
Back to article page

BMC Medical Research Methodology

ISSN: 1471-2288

Contact us