Background
Material and methods
Selection criteria
Search strategy
Data extraction
Results
General characteristics
Model or study characteristic | Number of articles (%) | References | |
---|---|---|---|
Number of patients | < 100 | 5 (6.3) | |
100–999 | 13 (16.5) | ||
1000–9999 | 39 (49.4) | ||
10,000+ | 21 (26.6) | ||
Not reported | 1 (1.3) | [80] | |
Number of time points | Median 2 | 1 (1.2) | [90] |
3 | 27 (33.8) | ||
≥3 (median, mean or maximum) | 47 (58.8) | ||
Not reported | 5 (6.2) | ||
Follow-up for longitudinal and survival length (years) | < 5 | 16 (20.0) | |
5 to 10 | 17 (21.2) | ||
10 to 20 | 29 (36.2) | ||
> 20 | 18 (22.5) | ||
Follow-up for longitudinal length (years) | < 5 | 24 (30.0) | |
5 to 10 | 25 (31.2) | ||
10 to 20 | 23 (28.8) | ||
> 20 | 8 (10.0) | ||
Follow-up for survival length (years) | < 5 | 19 (23.8) | |
5 to 10 | 26 (32.5) | ||
10 to 20 | 23 (28.8) | ||
> 20 | 4 (5.0) | ||
No survival analysis | 8 (10.0) | ||
Time-period for start of data collection | 1950s | 2 (2.5) | |
1960s | 6 (7.5) | ||
1970s | 5 (6.2) | ||
1980s | 20 (25.0) | ||
1990s | 13 (16.2) | ||
2000s | 26 (32.5) | ||
2010s | 4 (5.0) | ||
Not reported | 4 (5.0) | ||
Decade of publication | Prior to 2000 | 8 (10.0) | |
2000s | 7 (8.8) | ||
2010s | 63 (78.8) | ||
2020 | 2 (2.5) | ||
Baseline Age - mean/median | < 40 | 5 (6.2) | |
40–49 | 12 (15.0) | ||
50–59 | 18 (22.5) | ||
60–69 | 17 (21.2) | ||
70–79 | 7 (8.8) | ||
> 80 | 2 (2.5) | ||
Not reported | 19 (23.8) | ||
Region of dataset | Asia | 16 (20.0) | |
Europe | 22 (27.5) | ||
International | 3 (3.8) | ||
Middle East | 3 (3.8) | ||
North America | 33 (41.2) | ||
Australia & New Zealand | 3 (3.8) | ||
Males (%) | < 40 | 7 (8.8) | |
40s | 28 (35.0) | ||
50s | 7 (8.8) | ||
60–99 | 20 (25.0) | ||
All male | 11 (13.8) | ||
Not reported | 7 (8.8) | ||
Survival outcome type | Binary | 5 (6.2) | |
Continuous | 8 (10.0) | ||
Rate | 4 (5.0) | ||
Time to event | 63 (78.8) | ||
Longitudinal outcome type | Binary | 3 (3.8) | |
Categorical | 5 (6.2) | ||
Continuous | 69 (86.2) | ||
Ordinal | 3 (3.8) | ||
Survival analysis adjusted for age | Unadjusted | 8 (10.0) | |
Yes, Stratified | 3 (3.8) | ||
Yes, Baseline hazard | 1 (1.2) | [39] | |
Yes, Covariate | 61 (76.2) | ||
No survival analysis | 7 (8.8) | ||
Survival analysis adjusted for sex | Unadjusted | 4 (5.0) | |
Single sex | 12 (15.0) | ||
Yes, separate models | 9 (11.2) | ||
Yes, stratified | 3 (3.8) | ||
Yes, covariate | 45 (56.2) | ||
No survival analysis | 7 (8.8) | ||
Longitudinal Analysis adjusted for age | Unadjusted | 30 (37.5) | |
Yes, covariate | 17 (21.2) | ||
No longitudinal analysis | 33 (41.2) | ||
Longitudinal Analysis adjusted for sex | Unadjusted | 28 (35.0) | |
Single sex | 6 (7.5) | ||
Yes, separate models | 4 (5.0) | ||
Yes, covariate | 9 (11.2) | ||
No longitudinal analysis | 33 (41.2) | ||
Disease area | Chronic kidney disease | 1 (1.2) | [87] |
Cushing’s disease | 1 (1.2) | [16] | |
Cardiovascular disease | 61 (76.2) | ||
Diabetes | 1 (1.2) | [25] | |
Gout | 1 (1.2) | [59] | |
Hypertension | 1 (1.2) | [63] | |
Impaired sleep | 1 (1.2) | [24] | |
Mortality | 5 (6.2) | ||
Systemic lupus erythematosus | 1 (1.2) | [46] | |
Stroke | 7 (8.8) | ||
Primary Outcome | Acute coronary syndrome | 4 (5.0) | |
Atrial fibrillation | 2 (2.5) | ||
Cardiovascular mortality | 7 (8.8) | ||
Cardiovascular Mortality/acute coronary syndrome/stroke | 1 (1.2) | [93] | |
Cardiovascular disease | 36 (45.0) | ||
Cardiovascular disease risk | 8 (10.0) | ||
Cardiovascular disease/cancer/mortality | 1 (1.2) | [40] | |
Cardiovascular disease/mortality | 2 (2.5) | ||
Hospitalization/heart failure/cardiovascular mortality | 1 (1.2) | [28] | |
Hypertension | 1 (1.2) | [24] | |
Mortality | 9 (11.2) | ||
Stroke | 8 (10.0) | ||
Population focus | Acute coronary syndrome | 4 (5.0) | |
Atrial fibrillation and chronic kidney disease | 1 (1.2) | [87] | |
Chronic kidney disease | 2 (2.5) | ||
Cushing’s disease | 1 (1.2) | [16] | |
Cardiovascular disease | 37 (46.2) | ||
Diabetes | 3 (3.8) | ||
General population | 27 (33.8) | ||
Gout | 1 (1.2) | [59] | |
Heart failure | 1 (1.2) | [28] | |
HIV | 1 (1.2) | [47] | |
Systemic lupus erythematosus | 1 (1.2) | [46] | |
Mental health | 1 (1.2) | [89] |
Outcome data
Adjusting for covariates
Statistical analysis
Characteristics of included studies
Model or study characteristic | No of papers n (%) | Simple statistical tests n (%) | Single modela n (%) | Two-stage modelb n (%) | Joint modelc n (%) | Complete case analysis n (%) | |
---|---|---|---|---|---|---|---|
Number of patients | < 100 | 5 (6.3) | 2 (66.7) | 2 (5.0) | 1 (3.4) | 0 (0.0) | 5 (100.0) |
100–999 | 13 (16.5) | 1 (33.3) | 3 (7.5) | 6 (20.7) | 3 (37.5) | 11 (84.6) | |
1000–9999 | 39 (49.4) | 0 (0.0) | 24 (60.0) | 11 (37.9) | 4 (50.0) | 29 (74.4) | |
10,000+ | 21 (26.6) | 0 (0.0) | 10 (25.0) | 10 (34.5) | 1 (12.5) | 18 (85.7) | |
Not reported | 1 (1.3) | 0 (0.0) | 0 (0.0) | 1 (3.4) | 0 (0.0) | 1 (100.0) | |
Number of time points | Median 2 | 1 (1.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (12.5) | 0 (0.0) |
3 | 27 (33.8) | 3 (100.0) | 13 (32.5) | 9 (31.0) | 2 (25.0) | 21 (77.8) | |
≥3 (median, mean or maximum) | 47 (58.8) | 0 (0.0) | 23 (57.5) | 19 (65.5) | 5 (62.5) | 39 (83.0) | |
Not reported | 5 (6.2) | 0 (0.0) | 4 (10.0) | 1 (3.4) | 0 (0.0) | 5 (100.0) | |
Follow-up for longitudinal and survival length (years) | < 5 | 16 (20.0) | 3 (100.0) | 8 (20.0) | 4 (13.8) | 1 (12.5) | 14 (87.5) |
5 to 10 | 17 (21.2) | 0 (0.0) | 8 (20.0) | 8 (27.6) | 1 (12.5) | 15 (88.2) | |
10 to 20 | 29 (36.2) | 0 (0.0) | 12 (30.0) | 11 (37.9) | 6 (75.0) | 23 (79.3) | |
> 20 | 18 (22.5) | 0 (0.0) | 12 (30.0) | 6 (20.7) | 0 (0.0) | 13 (72.2) | |
Follow-up for longitudinal length (years) | < 5 | 24 (30.0) | 3 (100.0) | 9 (22.5) | 11 (37.9) | 1 (12.5) | 21 (87.5) |
5 to 10 | 25 (31.2) | 0 (0.0) | 13 (32.5) | 10 (34.5) | 2 (25.0) | 23 (92.0) | |
10 to 20 | 23 (28.8) | 0 (0.0) | 12 (30.0) | 6 (20.7) | 5 (62.5) | 16 (69.6) | |
> 20 | 8 (10.0) | 0 (0.0) | 6 (15.0) | 2 (6.9) | 0 (0.0) | 5 (62.5) | |
Follow-up for survival length (years) | < 5 | 19 (23.8) | 0 (0.0) | 9 (22.5) | 9 (31.0) | 1 (12.5) | 16 (84.2) |
5 to 10 | 26 (32.5) | 0 (0.0) | 12 (30.0) | 13 (44.8) | 1 (12.5) | 23 (88.5) | |
10 to 20 | 23 (28.8) | 0 (0.0) | 10 (25.0) | 7 (24.1) | 6 (75.0) | 16 (69.6) | |
> 20 | 4 (5.0) | 0 (0.0) | 4 (10.0) | 0 (0.0) | 0 (0.0) | 2 (50.0) | |
No survival analysis | 8 (10.0) | 3 (100.0) | 5 (12.5) | 0 (0.0) | 0 (0.0) | 8 (100.0) | |
Time-period for start of data collection | 1950s | 2 (2.5) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 0 (0.0) | 1 (50.0) |
1960s | 6 (7.5) | 0 (0.0) | 4 (10.0) | 2 (6.9) | 0 (0.0) | 6 (100.0) | |
1970s | 5 (6.2) | 0 (0.0) | 3 (7.5) | 2 (6.9) | 0 (0.0) | 3 (60.0) | |
1980s | 20 (25.0) | 0 (0.0) | 10 (25.0) | 7 (24.1) | 3 (37.5) | 16 (80.0) | |
1990s | 13 (16.2) | 0 (0.0) | 7 (17.5) | 3 (10.3) | 3 (37.5) | 9 (69.2) | |
2000s | 26 (32.5) | 1 (33.3) | 12 (30.0) | 11 (37.9) | 2 (25.0) | 22 (84.6) | |
2010s | 4 (5.0) | 1 (33.3) | 1 (2.5) | 2 (6.9) | 0 (0.0) | 4 (100.0) | |
Not reported | 4 (5.0) | 1 (33.3) | 2 (5.0) | 1 (3.4) | 0 (0.0) | 4 (100.0) | |
Decade of publication | Prior to 2000 | 8 (10.0) | 0 (0.0) | 6 (15.0) | 2 (6.9) | 0 (0.0) | 8 (100.0) |
2000s | 7 (8.8) | 1 (33.3) | 5 (12.5) | 1 (3.4) | 0 (0.0) | 6 (85.7) | |
2010s | 63 (78.8) | 2 (66.7) | 28 (70.0) | 25 (86.2) | 8 (100.0) | 49 (77.8) | |
2020 | 2 (2.5) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 0 (0.0) | 2 (100.0) | |
Baseline Age - mean/median | < 40 | 5 (6.2) | 0 (0.0) | 2 (5.0) | 3 (10.3) | 0 (0.0) | 5 (100.0) |
40–49 | 12 (15.0) | 1 (33.3) | 7 (17.5) | 2 (6.9) | 2 (25.0) | 8 (66.7) | |
50–59 | 18 (22.5) | 1 (33.3) | 10 (25.0) | 7 (24.1) | 0 (0.0) | 13 (72.2) | |
60–69 | 17 (21.2) | 0 (0.0) | 9 (22.5) | 6 (20.7) | 2 (25.0) | 14 (82.4) | |
70–79 | 7 (8.8) | 0 (0.0) | 2 (5.0) | 3 (10.3) | 2 (25.0) | 6 (85.7) | |
> 80 | 2 (2.5) | 0 (0.0) | 0 (0.0) | 1 (3.4) | 1 (12.5) | 2 (100.0) | |
Not reported | 19 (23.8) | 1 (33.3) | 10 (25.0) | 7 (24.1) | 1 (12.5) | 17 (89.5) | |
Data Region | Asia | 16 (20.0) | 0 (0.0) | 5 (12.5) | 10 (34.5) | 1 (12.5) | 15 (93.8) |
Europe | 22 (27.5) | 1 (33.3) | 15 (37.5) | 2 (6.9) | 4 (50.0) | 14 (63.6) | |
International | 3 (3.8) | 0 (0.0) | 2 (5.0) | 0 (0.0) | 1 (12.5) | 2 (66.7) | |
Middle East | 3 (3.8) | 1 (33.3) | 2 (5.0) | 0 (0.0) | 0 (0.0) | 2 (66.7) | |
North America | 33 (41.2) | 1 (33.3) | 15 (37.5) | 16 (55.2) | 1 (12.5) | 29 (87.9) | |
Australia & NZ | 3 (3.8) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 1 (12.5) | 3 (100.0) | |
Males (%) | < 40 | 7 (8.8) | 1 (33.3) | 2 (5.0) | 3 (10.3) | 1 (12.5) | 6 (85.7) |
40s | 28 (35.0) | 1 (33.3) | 14 (35.0) | 10 (34.5) | 3 (37.5) | 24 (85.7) | |
50s | 7 (8.8) | 0 (0.0) | 3 (7.5) | 3 (10.3) | 1 (12.5) | 6 (85.7) | |
60–99 | 20 (25.0) | 0 (0.0) | 10 (25.0) | 9 (31.0) | 1 (12.5) | 16 (80.0) | |
All male | 11 (13.8) | 0 (0.0) | 7 (17.5) | 2 (6.9) | 2 (25.0) | 8 (72.7) | |
Not reported | 7 (8.8) | 1 (33.3) | 4 (10.0) | 2 (6.9) | 0 (0.0) | 5 (71.4) | |
Survival outcome type | Binary | 5 (6.2) | 0 (0.0) | 4 (10.0) | 1 (3.4) | 0 (0.0) | 3 (60.0) |
Continuous | 8 (10.0) | 3 (100.0) | 5 (12.5) | 0 (0.0) | 0 (0.0) | 8 (100.0) | |
Rate | 4 (5.0) | 0 (0.0) | 3 (7.5) | 1 (3.4) | 0 (0.0) | 3 (75.0) | |
Time to event | 63 (78.8) | 0 (0.0) | 28 (70.0) | 27 (93.1) | 8 (100.0) | 51 (81.0) | |
Longitudinal outcome type | Binary | 3 (3.8) | 0 (0.0) | 3 (7.5) | 0 (0.0) | 0 (0.0) | 2 (66.7) |
Categorical | 5 (6.2) | 0 (0.0) | 2 (5.0) | 3 (10.3) | 0 (0.0) | 4 (80.0) | |
Continuous | 69 (86.2) | 3 (100.0) | 32 (80.0) | 26 (89.7) | 8 (100.0) | 57 (82.6) | |
Ordinal | 3 (3.8) | 0 (0.0) | 3 (7.5) | 0 (0.0) | 0 (0.0) | 2 (66.7) | |
Survival analysis adjusted for age | Unadjusted | 8 (10.0) | 0 (0.0) | 4 (10.0) | 2 (6.9) | 2 (25.0) | 5 (62.5) |
Yes, Stratified | 3 (3.8) | 0 (0.0) | 2 (5.0) | 1 (3.4) | 0 (0.0) | 3 (100.0) | |
Yes, Baseline hazard | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 0 (0.0) | |
Yes, Covariate | 61 (76.2) | 0 (0.0) | 29 (72.5) | 26 (89.7) | 6 (75.0) | 50 (82.0) | |
No survival analysis | 7 (8.8) | 3 (100.0) | 4 (10.0) | 0 (0.0) | 0 (0.0) | 7 (100.0) | |
Survival analysis adjusted for sex | Unadjusted | 4 (5.0) | 0 (0.0) | 1 (2.5) | 2 (6.9) | 1 (12.5) | 4 (100.0) |
Single sex | 12 (15.0) | 0 (0.0) | 7 (17.5) | 2 (6.9) | 3 (37.5) | 8 (66.7) | |
Yes, separate models | 9 (11.2) | 0 (0.0) | 5 (12.5) | 3 (10.3) | 1 (12.5) | 7 (77.8) | |
Yes, stratified | 3 (3.8) | 0 (0.0) | 1 (2.5) | 2 (6.9) | 0 (0.0) | 3 (100.0) | |
Yes, covariate | 45 (56.2) | 0 (0.0) | 22 (55.0) | 20 (69.0) | 3 (37.5) | 36 (80.0) | |
No survival analysis | 7 (8.8) | 3 (100.0) | 4 (10.0) | 0 (0.0) | 0 (0.0) | 7 (100.0) | |
Longitudinal Analysis adjusted for age | Unadjusted | 30 (37.5) | 0 (0.0) | 3 (7.5) | 23 (79.3) | 4 (50.0) | 28 (93.3) |
Yes, covariate | 17 (21.2) | 0 (0.0) | 7 (17.5) | 6 (20.7) | 4 (50.0) | 13 (76.5) | |
No longitudinal analysis | 33 (41.2) | 3 (100.0) | 30 (75.0) | 0 (0.0) | 0 (0.0) | 24 (72.7) | |
Longitudinal Analysis adjusted for sex | Unadjusted | 28 (35.0) | 0 (0.0) | 3 (7.5) | 22 (75.9) | 3 (37.5) | 27 (96.4) |
Single sex | 6 (7.5) | 0 (0.0) | 1 (2.5) | 2 (6.9) | 3 (37.5) | 4 (66.7) | |
Yes, separate models | 4 (5.0) | 0 (0.0) | 1 (2.5) | 2 (6.9) | 1 (12.5) | 3 (75.0) | |
Yes, covariate | 9 (11.2) | 0 (0.0) | 5 (12.5) | 3 (10.3) | 1 (12.5) | 7 (77.8) | |
No longitudinal analysis | 33 (41.2) | 3 (100.0) | 30 (75.0) | 0 (0.0) | 0 (0.0) | 24 (72.7) | |
Disease area | Chronic kidney disease | 1 (1.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (12.5) | 1 (100.0) |
Cushing’s disease | 1 (1.2) | 1 (33.3) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Cardiovascular disease | 61 (76.2) | 2 (66.7) | 33 (82.5) | 21 (72.4) | 5 (62.5) | 47 (77.0) | |
Diabetes | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Gout | 1 (1.2) | 0 (0.0) | 0 (0.0) | 1 (3.4) | 0 (0.0) | 1 (100.0) | |
Hypertension | 1 (1.2) | 0 (0.0) | 0 (0.0) | 1 (3.4) | 0 (0.0) | 1 (100.0) | |
Impaired sleep | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Mortality | 5 (6.2) | 0 (0.0) | 1 (2.5) | 3 (10.3) | 1 (12.5) | 5 (100.0) | |
Systemic lupus erythematosus | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Stroke | 7 (8.8) | 0 (0.0) | 3 (7.5) | 3 (10.3) | 1 (12.5) | 6 (85.7) | |
Primary Outcome | Acute coronary syndrome | 4 (5.0) | 0 (0.0) | 2 (5.0) | 2 (6.9) | 0 (0.0) | 2 (50.0) |
Atrial fibrillation | 2 (2.5) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 0 (0.0) | 2 (100.0) | |
Cardiovascular mortality | 7 (8.8) | 0 (0.0) | 6 (15.0) | 1 (3.4) | 0 (0.0) | 4 (57.1) | |
Cardiovascular mortality/acute coronary syndrome/stroke | 1 (1.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (12.5) | 1 (100.0) | |
Cardiovascular disease | 36 (45.0) | 0 (0.0) | 16 (40.0) | 15 (51.7) | 5 (62.5) | 28 (77.8) | |
Cardiovascular disease risk | 8 (10.0) | 3 (100.0) | 5 (12.5) | 0 (0.0) | 0 (0.0) | 8 (100.0) | |
Cardiovascular disease/cancer/mortality | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Cardiovascular disease/mortality | 2 (2.5) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 0 (0.0) | 2 (100.0) | |
Hospitalization/heart failure/cardiovascular mortality | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Hypertension | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Mortality | 9 (11.2) | 0 (0.0) | 4 (10.0) | 5 (17.2) | 0 (0.0) | 8 (88.9) | |
Stroke | 8 (10.0) | 0 (0.0) | 2 (5.0) | 4 (13.8) | 2 (25.0) | 7 (87.5) | |
Population focus | Acute coronary syndrome | 4 (5.0) | 0 (0.0) | 3 (7.5) | 1 (3.4) | 0 (0.0) | 2 (50.0) |
Atrial fibrillation and chronic kidney disease | 1 (1.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (12.5) | 1 (100.0) | |
Chronic kidney disease | 2 (2.5) | 0 (0.0) | 2 (5.0) | 0 (0.0) | 0 (0.0) | 1 (50.0) | |
Cushing’s disease | 1 (1.2) | 1 (33.3) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Cardiovascular disease | 37 (46.2) | 1 (33.3) | 17 (42.5) | 18 (62.1) | 1 (12.5) | 32 (86.5) | |
Diabetes | 3 (3.8) | 0 (0.0) | 1 (2.5) | 1 (3.4) | 1 (12.5) | 3 (100.0) | |
General population | 27 (33.8) | 1 (33.3) | 14 (35.0) | 8 (27.6) | 4 (50.0) | 20 (74.1) | |
Gout | 1 (1.2) | 0 (0.0) | 0 (0.0) | 1 (3.4) | 0 (0.0) | 1 (100.0) | |
Heart failure | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
HIV | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Systemic lupus erythematosus | 1 (1.2) | 0 (0.0) | 1 (2.5) | 0 (0.0) | 0 (0.0) | 1 (100.0) | |
Mental health | 1 (1.2) | 0 (0.0) | 0 (0.0) | 0 (0.0) | 1 (12.5) | 1 (100.0) |
Single-stage approaches
Method N(%)[refs] | Longitudinal outcome type | Disease outcome type | How the longitudinal data were used in the analysis N (%) [refs] | Reason for the use of method | Assumptions | Pros | Cons |
---|---|---|---|---|---|---|---|
Single-stage approaches (n = 40) | |||||||
Continuous, Categorical | Time to event | To clinically relevant time point to be used for prediction | PH | Simple method | Dependence between measurement times is ignored | ||
Continuous | Time to event | To incorporate change over time | PH; Change is linear | Incorporates more than one time point | Only looks at pairs of time points | ||
Continuous | Time to event | To incorporate constant change in the survival model | PH; Change is linear | Incorporates more than one time point | Only looks at pairs of time points | ||
Continuous | Time to event | Average (categorized before use),a N = 1 (2.5) [36] | To incorporate the average change over time | PH; Constant between time points; Change is linear | Incorporates the average impact over time | Interpretation unclear | |
Continuous, Categorical | Time to event | To incorporate change in exposure variable over time | PH; Change is constant between two consecutive time points; Longitudinal data are measured without error | Incorporates time-varying measures over the follow-up period | Computationally slower as compared to time-fixed covariates; Computationally infeasible if the longitudinal outcome is measured at different time points for different individuals; Interpretation is difficult; Can lead to great overfitting of the data; must be used with caution | ||
Continuous | Time to event | Summary measures(Standard deviation, number of drops between observations), N = 1 (2.5) [19] | To incorporate variability summaries of the longitudinal data | PH | Incorporates variability of measures into the model | Summary measures fairly specific to dataset | |
Continuous, Categorical | Time to event | Change in continuous variable between time points categorized with manually defined cut-offs, N = 1 (2.5) [56] | To summarise trajectories in an interpretable way | PH | Results interpretable | Groups manually selected based on data which could lead to bias | |
Hierarchical Cox model to adjust for multiple studies, N = 1 (2.5) [20] | Continuous | Time to event | Continuous measurements categorized. Multiple time points also categorised as consistent/non-consistent, N = 1 (2.5) [20] | To summarise trajectories in an interpretable way adjusting for combining multiple studies | PH | Results interpretable; Adjusts for use of multiple studies | Groups manually selected based on data which could lead to bias |
Continuous | Binary | Baseline only, N = 1 (2.5) [31] | Allows clinically relevant time point to be used for prediction | Not applicable | Simple method | Dependence between measurement times is ignored | |
Categorical | Binary | Separate time points, N = 1 (2.5) [30] | To include all predictive values in model | Not applicable | Simple method | Caution needed for multicollinearity | |
Continuous | Binary | Summaries of repeated measures • Standard deviation • Mean • Mean change from baseline • Average daily risk rangeb • Range N = 1 (2.5) [48] | Includes different measures of variation | Not applicable | Simple method | Interpretation of different summary measures non-trivial | |
Continuous | Binary | Non-linear relationships considered through piecewise models or splines, N = 1 (2.5) [17] | To attempt to include a variety of shapes of relationships in the model using data from all time points | Not applicable | Includes all measured values of longitudinal variable with various relationships with risk | Splines harder to interpret; Produces population averages not individual predictions | |
Continuous | Binary | Multiple time points, N = 1 (2.5) [27] | To include values and change at all time points | Not applicable | Includes all measured values of longitudinal variable | Produces population averages not individual predictions | |
Continuous | Rates | Multiple time points, N = 1 (2.5) [37] Multiple time points categorized as stable, increasing (in the second or third time point), decreasing, unstable, N = 1 (2.5) [22] | To include all time points in predicting rates | Not applicable | Includes all measured values of longitudinal variable | Produces population averages not individual predictions | |
Continuous | Rates | To enable modelling of baseline rate | Not applicable | Enables modelling of baseline rate in a parametric manner | Dependence between measurement times is ignored | ||
Continuous, categorical | Continuous | To predict changes over time | Random effects are independent of covariates | Includes all measured values of longitudinal variable | None | ||
Fixed effects linear regression, N = 1 (2.5) [52] | Continuous, categorical | Continuous | The variable is transformed by subtracting patient-level mean to remove between patient variation. N = 1 (2.5) [52] | To predict changes over time | Not applicable | Includes all measured values of longitudinal variable; Relaxes assumption of independence of random effects from covariates; Computationally very easy to fit compared with mixed effects models | Lower statistical efficiency than mixed effects models |
Two-stage models
Method N(%)[refs] | Longitudinal outcome type | Disease outcome type | How the longitudinal data were used in the analysis N (%) [refs] | Reason for the use of method | Assumptions | Pros | Cons |
---|---|---|---|---|---|---|---|
Continuous | Time to event | Slope, N = 1 (3.4) [57] | To incorporate a constant change or variation in the survival model | PH; Change is linear | Incorporates information from all time points | Does not allow for adjustment by other covariates as it cannot calculate overall coefficients | |
Continuous, Categorical | Time to event | To find groups for the trajectories based on the data | PH; Population of trajectories arises from a finite mixture | Very effective at summarizing trajectories | Cannot place patients into trajectory groups easily in clinical practice; Computationally very hard model to fit | ||
Logistic Regression, N = 1 (3.4) [64] | Continuous | Binary | Latent class model used to calculate trajectory of longitudinal variable, N = 1 (3.4) [64] | To find groups for the trajectories based on the data | Population of trajectories arises from a finite mixture | Very effective at summarizing trajectories | Cannot place patients into trajectory groups easily in clinical practice; Computationally very hard model to fit |
Weighted pooled logistic regression, N = 1 (3.4) [59] | Continuous | Binary | Inverse probability weights calculated for each time-point. Each time-point had its own logistic regression model. Model results were pooled to produce HRs, N = 1 (3.4) [59] | To adjust for time-varying confounders | None considered | Accounts for variation in longitudinal data; Efficient to fit | Complex model that is not easy to understand or interpret |
Poisson regression, N = 1 (3.4) [79] | Continuous | Rates | Latent class model used to calculate trajectory of longitudinal variable, N = 1 (3.4) [79] | To find groups for the trajectories based on the data | Population of trajectories arises from a finite mixture | Very effective at summarizing trajectories | Cannot place patients into trajectory groups easily in clinical practice; Computationally very hard model to fit |
Joint models
Method N(%)[refs] | Longitudinal outcome type | Disease outcome type | How the longitudinal data were used in the analysis N (%) [refs] | Reason for the use of method | Assumptions | Pros | Cons |
---|---|---|---|---|---|---|---|
Continuous | Time to event | Longitudinal data were modelled in LME. Survival data were modelled in Cox PH. N = 5 (62.5) [86, 87, 91‐93] Association structures: 1st derivative, N = 1 (12.5) [87] | To predict changes in risk score over time using repeated measures | None considered | Includes all measured values of longitudinal variable | Computationally very hard model to fit | |
Continuous | Time to event | Structured equation model incorporated in survival model as covariate, N = 1 (12.5) [89] | To incorporate a constant change or variation in the survival model | PH; Change is linear | Incorporates information from all time points | Does not allow for adjustment by other covariates as it cannot calculate overall coefficients | |
Latent class model, N = 1 (12.5) [88] | Continuous | Time to event | Latent class model used to calculate trajectory of longitudinal variable. Trajectory class incorporated in model as covariate, N = 1 (12.5) [88] | To find groups for the trajectories based on the data | PH; Population of trajectories arises from a finite mixture | Very effective at summarizing trajectories | Cannot place patients into trajectory groups easily in clinical practice; Computationally very hard model to fit |
Bayesian approach, N = 1 (12.5) [90] | Ordinal | Time to event | Item response theory models were used to model ordinal data from a multi-question survey using a latent parameter. This latent parameter was modelled using a linear growth model and was incorporated in a multi-state Gompertz survival model as a covariate, N = 1 (12.5) [90] | To model ordinal survey data with the correct distribution | Values constant between observations | Incorporates data from complex survey accounting for ordinal data modelling the data directly rather than modelling the sum of the responses | Complex and requires Bayesian code to be used to define the model |