Skip to main content

Table 2 Adherence by TRIPOD item among obstetrical prediction model studies, n (%)

From: The reporting of prognostic prediction models for obstetric care was poor: a cross-sectional survey of 10-year publications

Sections

TRIPOD items

Overall

Type of prediction model study

(N = 121)

Model development

External validation

Incremental value

Development and external validation

 

(N = 93)

(N = 10)

(N = 6)

(N = 12)

Title

1 Study presentation in title

16 (13.2)

5 (5.4)

1 (10.0)

2 (33.3)

8 (66.7)

Abstract

2 Summary of the study

7 (5.8)

2 (2.2)

3 (30.0)

0 (0.0)

2 (16.7)

Background and objectives

3a Key contents of background

114 (94.2)

87 (93.5)

9 (90.0)

6 (100.0)

12 (100.0)

3b objectives of model

119 (98.3)

91 (97.8)

10 (100.0)

6 (100.0)

12 (100.0)

Methods

 Source of data

4a Design/data source

119 (98.3)

91 (97.8)

10 (100.0)

6 (100.0)

12 (100.0)

4b Study dates

33 (27.3)

21 (22.6)

5 (50.0)

0 (0.0)

7 (58.3)

 Participants

5a Key elements of setting

97 (80.2)

71 (76.3)

9 (90.0)

6 (100.0)

11 (91.7)

5b Participant eligibility

115 (95.0)

88 (94.6)

9 (90.0)

6 (100.0)

4 (100.0)

5c Details of treatments

24/37 (64.9) a

16/26 (61.5) a

0/2 (0.00)

2/2 (100.0) a

6/7 (85.7) a

 Outcome

6a Outcome definition

96 (79.3)

75 (80.6)

8 (80.0)

3 (50.0)

10 (83.3)

6b Blind assessment of outcome

49 (40.5)

37 (39.8)

3 (30.0)

1 (16.7)

8 (66.7)

 Predictors

7a Predictor definition

82 (67.8)

61 (65.6)

8 (80.0)

4 (66.7)

83 (68.0)

7b Blind assessment

3 (2.5)

3 (3.2)

0 (0.0)

0 (0.0)

0 (0.0))

 Sample size

8 How to arrive at the size

18 (14.9)

12 (12.9)

0 (0.0)

1 (16.7)

5 (41.7)

 Missing data

9 Handling missing data

36 (29.8)

24 (25.8)

2 (20.0)

1 (16.7)

9 (75.0)

 Statistical analysis method

10a Predictor handling

15/111 (13.5) a

9 (9.7)

NA

0 (0.0)

6 (50.0)

10b Model-building procedures

5/111 (4.5) a

4 (4.3)

NA

0 (0.0)

1 (8.3)

10c Calculation for validation

23/28 (82.1) a

NA

9 (90.0)

2 (33.3)

12 (100.0)

10d Model performance

22/111 (19.8) a

15/87 (17.2) a

3/7 (42.9) a

0/5 (0.0) a

4 (33.3)

10e Model updating

7/11 (63.6) a

NA

1/4 (25.0) a

6 (100.0)

0 (0.0) a

 Risk groups

11 Details to create risk groups

18/41 (43.9) a

10/30 (33.3) a

3/3 (100.0)

NA

5/8 (62.5)

 Development vs. validation

12 Differences of validation from development

17/28 (60.7) a

NA

2 (20.0)

6 (100.0)

9 (75.0)

Results

 Participants

13a Flow of participants through the study

10 (8.3)

7 (7.5)

1 (10.0)

0 (0.0)

2 (16.7)

13b Characteristics of participants

40 (33.1)

24 (25.8)

4 (40.0)

3 (50.0)

9 (75.0)

13c Comparison of validation with development

15/28 (53.6) a

NA

2 (20.0)

6(100.0)

7 (58.3)

 Model development

14a Numbers

104/111 (93.7) a

87 (93.5)

NA

6 (100.0)

11 (91.7)

14b Unadjusted association of each candidate predictor and outcome

60/94 (63.8) a

48/80 (60.0) a

NA

4/4 (100.0) a

8/10 (80.0) a

 Model specification

15a Present the full model

49/111 (44.1) a

40 (43.0)

NA

2 (33.3)

7 (58.3)

15b How to use the model

15/111 (13.5) a

8 (8.6)

NA

0 (0.0)

7 (58.3)

 Model performance

16 Performance measures

16/111 (14.4) a

9/87 (10.3) a

3/7 (42.9) a

0/5 (0.0) a

4 (33.3)

 Model updating

17 Results from any updating

1/2 (50.0) a

NA

1/2 (50.0) a

NA

0 (0.0)

Discussion

 Limitations

18 Any limitations of the study

113 (93.4)

89 (95.7)

9 (90.0)

4 (66.7)

11 (91.7)

 Interpretations

19a Comparing results of validation with development

24/28 (85.7) a

NA

6 (60.0)

6 (100.0)

12 (100.0)

19b Overall interpretation of results

120 (99.2)

92 (98.9)

10 (100.0)

6 (100.0)

12 (100.0)

 Implications

20 Potential implication

75 (62.0)

59 (63.4)

4 (40.0)

2 (33.3)

10 (83.3)

Other information

 Supplementary information

21 Information about supplementary resources (Optional)

53 (43.8)

44 (47.3)

1 (10.0)

1 (16.7)

7 (58.3)

 Funding

22 Funding information

14 (11.6)

9 (9.7)

1 (10.0)

0 (0.0)

4 (33.3)

  1. Note: NA not applicable, there are some TRIPOD items are not applicable for all types of prediction model studies
  2. aPercentage is based on number of models for which that item was applicable (and should have been reported). Where this number deviates from the total number of models, the actual number of applicable models is presented as denominator, besides, denominator for the rest of percentage without specification is the total number for overall or each type of prediction models studies