Background
Methods
Study design and population
Pre-selected variables
Outcome measurement
Statistical methods
Sensitivity analysis
Results
Study population
Study population n = 2211 | Not includedan = 502 | ||||
---|---|---|---|---|---|
Baseline characteristics | n | % | n | % | p-valueb |
Age at interview: years (median, IQR) | 82.1 (78.8; 86.5) | 82.6 (79.2; 87.3) | 0.044 | ||
Baseline GARS (median, IQR) | 31 (24; 41) | 34 (25; 43) | 0.006 | ||
Sex | |||||
Male | 707 | 32.0 | 150 | 29.9 | 0.362 |
Female | 1504 | 68.0 | 352 | 70.1 | |
Polypharmacy | n = 501 | ||||
< 4 per day | 710 | 32.1 | 185 | 36.9 | 0.039 |
4 or more per day | 1501 | 67.9 | 316 | 63.1 | |
Multimorbidity | |||||
Yes (> 1 chronic disease) | 2011 | 91.0 | 458 | 91.2 | 0.843 |
No | 200 | 9.0 | 44 | 8.8 | |
Living situation | n = 501 | ||||
Independent, alone/with others | 1985 | 89.8 | 429 | 85.6 | 0.007 |
Home for older persons | 226 | 10.2 | 72 | 14.4 | |
GP opinion on vulnerability | n = 323 | ||||
Not vulnerable | 894 | 40.4 | 93 | 28.8 | < 0.001 |
Possibly vulnerable | 617 | 27.9 | 96 | 29.7 | |
Vulnerable | 700 | 31.7 | 134 | 41.5 | |
ISCOPE-score on ISCOPE screening questionnaire | |||||
0 (no domain with problems) | 198 | 9.0 | 49 | 9.8 | 0.099 |
1 (1 domain with problems) | 183 | 8.3 | 34 | 6.8 | |
2 (2 domains with problems) | 614 | 27.8 | 119 | 23.7 | |
3 (3 domains with problems) | 843 | 38.1 | 195 | 38.8 | |
4 (4 domains with problems) | 373 | 16.9 | 105 | 20.9 |
Lost to follow-up and missing data
Univariate association between pre-selected variables and functional decline
Relevant functional decline | ||||||
---|---|---|---|---|---|---|
Without decline (n = 1817; 82.2%) | With decline (n = 394; 17.8%) | |||||
Characteristics | n | % | n | % | Odds ratio | 95% CI |
Age in years (median, IQR) | 81.8 (78.5–86.2) | 83.5 (79.8–87.7) | 1.06 | 1.04; 1.08 | ||
Baseline GARS (median, IQR) | 30 (24–39) | 36 (28–49) | 1.04 | 1.03; 1.05 | ||
Sex | ||||||
Male | 559 | 30.8 | 148 | 37.6 | ref | |
Female | 1258 | 69.2 | 246 | 62.4 | 0.74 | 0.59; 0.93 |
Polypharmacy | ||||||
< 4 per day | 603 | 33.2 | 107 | 27.2 | ref | |
4 or more per day | 1214 | 66.8 | 287 | 72.8 | 1.33 | 1.05; 1.70 |
Multimorbidity | ||||||
No | 169 | 9.3 | 31 | 7.9 | ref | |
Yes (> 1 chronic disease) | 1648 | 90.7 | 363 | 92.1 | 1.20 | 0.81; 1.79 |
Living situation | ||||||
Independent, alone/with others | 1656 | 91.1 | 329 | 83.5 | ref | |
Home for older persons | 161 | 8.9 | 65 | 16.5 | 2.03 | 1.49; 2.77 |
GP opinion on vulnerability | ||||||
Not vulnerable | 802 | 44.1 | 92 | 23.4 | ref | |
Possibly vulnerable | 518 | 28.5 | 99 | 25.1 | 1.67 | 1.23; 2.26 |
Vulnerable | 497 | 27.4 | 203 | 51.5 | 3.56 | 2.72; 4.67 |
ISCOPE-score on ISCOPE screening questionnaire | ||||||
0 (no domain with problems) | 179 | 9.9 | 19 | 4.8 | ref | |
1 (1 domain with problems) | 172 | 9.5 | 11 | 2.8 | 0.60 | 0.28; 1.30 |
2 (2 domains with problems) | 508 | 28.0 | 106 | 26.9 | 1.97 | 1.17; 3.30 |
3 (3 domains with problems) | 687 | 37.8 | 156 | 39.6 | 2.14 | 1.29; 3.54 |
4 (4 domains with problems) | 271 | 14.9 | 102 | 25.9 | 3.55 | 2.10; 5.99 |
Predictive value of pre-selected variables on a functional decline
Variables included in the model | AUC | compared to | delta AUC | p-value | R2a | |
---|---|---|---|---|---|---|
Model 1 | Age and sex | 0.602 | 0.033 | |||
Model 2 | model 1, polypharmacy, multimorbidity and living situation | 0.620 | model 1 | 0.018 | 0.029 | 0.042 |
Model 3 | model 1 and ISCOPE-score | 0.644 | model 1 | 0.024 | < 0.001 | 0.066 |
Model 4 | model 1 and GP opinion on vulnerability | 0.669 | model 1 | 0.049 | < 0.001 | 0.086 |
Model 5 | model 1, ISCOPE-score and GP opinion on vulnerability | 0.684 | model 4 | 0.015 | 0.009 | 0.102 |
Model 6 | model 2 and ISCOPE-score | 0.649 | model 2 | 0.029 | 0.007 | 0.069 |
Model 7 | model 2 and GP opinion on vulnerability | 0.672 | model 2 | 0.052 | < 0.001 | 0.090 |
Model 8 | model 2, ISCOPE-score and GP opinion on vulnerability | 0.686 | model 7 | 0.014 | 0.016 | 0.105 |