Table 2.
Total | Differences per time point | Adjusted difference | ||||
---|---|---|---|---|---|---|
Intervention | Control | p-value | CareWell Effect | p-value | ||
Clinical data | ||||||
Body Mass Index (BMI) | 0.099 | |||||
N | 620 | 335 | 285 | |||
Basal | 29.9 (5.4) | 30.4 (5.8) | 29.4 (4.9) | 0.016 | – | |
Final | 29.5 (5.4) | 29.9 (5.6)* | 28.9 (5) | 0.038 | –0.3 (–0.8,0.2) | |
Oxygen saturation (%) | 0.639 | |||||
N | 546 | 261 | 285 | |||
Basal | 95.8 (2.4) | 95.4 (2.6) | 96.2 (2.2) | <0.001 | – | |
Final | 96.1 (2.3) | 95.9 (2.1)* | 96.2 (2.4) | 0.114 | 0.1 (–0.4,0.5) | |
Blood glucose (mg/dl) | 0.677 | |||||
N | 546 | 261 | 285 | |||
Basal | 129.9 (45.9) | 131.7 (50.8) | 128.3 (41) | 0.401 | – | |
Final | 118.8 (50) | 111 (48.7)* | 123.2 (50.3) | 0.012 | –0.6 (–12.7,11.5) | |
HbA1c (%) | 0.131 | |||||
N | 364 | 177 | 187 | |||
Basal | 6.8 (1) | 6.9 (1.1) | 6.8 (1) | 0.236 | – | |
Final | 6.8 (1.2) | 6.6 (1.1) | 6.8 (1.2) | 0.227 | –0.4 (–0.8,–0.04) | |
Creatinine (mg/dl) | 0.182 | |||||
N | 364 | 177 | 187 | |||
Basal | 1.1 (0.4) | 1.1 (0.5) | 1.1 (0.4) | 0.252 | – | |
Final | 1.2 (0.6) | 1.2 (0.6)* | 1.1 (0.5)* | 0.220 | –0.1 (–0.1,0.02) | |
Geriatric Depression Scale | 0.233 | |||||
N | 757 | 403 | 354 | |||
Basal | 3.9 (3.7) | 3.9 (3.7) | 4 (3.7) | 0.607 | – | |
Final | 4 (3.8) | 3.7 (3.7) | 4.2 (3.8) | 0.089 | –0.6 (–1.6,0.3) | |
Physical functional status | ||||||
Barthel index | 0.889 | |||||
N | 757 | 403 | 354 | |||
Basal | 87.8 (20.3) | 86.7 (22) | 89.2 (18.2) | 0.085 | – | |
Final | 86.7 (21.4) | 85.4 (22.9)* | 88.1 (19.6)* | 0.085 | –1.1 (–3.2,1) | |
Data presented as mean, standard deviation or their corresponding 95% confidence interval; N, sample size used in the analysis (only included pilot sites which has the follow-up for the corresponding parameter); * indicates if there are pre-post differences within each arm (intervention or control); CareWell effect, indicates the difference between the intervention over the control group, by the interaction coefficient between the care group and the follow-up period. The effect coefficient was estimated considering follow-up of 12 months, and the mixed-effect models were adjusted using age, gender, and age-adjusted Charlson Comorbidity Index.