Figure 1
Flowchart of the ALL-IN trial. LTFU = Lost to follow-up.
Table 1
Baseline characteristics of included patients. Numbers are counts (%) unless stated otherwise. The frailty index consists of the presence or absence of 36 health deficit items (scale 0–1, higher value indicating more frailty). EQ5D-5L, EuroQol 5D questionnaire; IQR, interquartile range; NOAC, non-vitamin K antagonist oral anticoagulant; RAAS, renin–angiotensin–aldosterone system; TIA, transient ischaemic attack; VKA, vitamin K antagonist.
| INTEGRATED CARE(N = 522) | USUAL CARE (N = 425) | |
|---|---|---|
| Age (years), median (IQR) | 76.0 (71.0–80.0) | 77.0 (72.0–82.0) |
| Female sex | 236 (45.2) | 211 (49.6) |
| Hypertension | 308 (59.0) | 230 (54.1) |
| Diabetes mellitus | 130 (24.9) | 110 (25.9) |
| Prior stroke/TIA | 81 (15.5) | 49 (11.5) |
| Coronary artery disease | 93 (17.8) | 73 (17.2) |
| Prior myocardial infarction | 36 (6.9) | 28 (6.6) |
| Heart failure | 72 (13.8) | 66 (15.5) |
| Peripheral vascular disease | 35 (6.7) | 29 (6.8) |
| Prior venous thromboembolism | 25 (4.8) | 10 (2.4) |
| Chronic renal impairment | 59 (11.3) | 61 (14.4) |
| Chronic obstructive pulmonary disease | 71 (13.6) | 62 (14.6) |
| History of cancer | 94 (18.0) | 82 (19.3) |
| Pacemaker | 34 (6.5) | 36 (8.5) |
| Frailty index, median (IQR) | 0.14 (0.11–0.22) | 0.14 (0.11–0.19) |
| Polypharmacy (≥5 chronic drugs) | 134 (25.7) | 86 (20.2) |
| Anticoagulant use | ||
| VKA | 386 (73.9) | 340 (80.0) |
| NOAC | 83 (15.9) | 57 (13.4) |
| Antiplatelet therapy | 48 (9.2) | 22 (5.2) |
| Beta-blockers | 373 (71.5) | 312 (73.4) |
| Calcium channel antagonists | 149 (28.5) | 111 (26.1) |
| Digoxin | 96 (18.4) | 79 (18.6) |
| Class I and III antiarrhythmic drugs | 32 (6.1) | 31 (7.3) |
| Diuretics | 194 (37.2) | 186 (43.8) |
| RAAS-inhibitors | 278 (53.3) | 248 (58.4) |
Table 2
Imputed, unadjusted costs for the intervention versus usual care. Mean number of procedures and costs (in euros) per patient throughout the 2-year follow-up period. ‡ For hospital admissions, the mean length of stay in days (summed for all admissions per patient) is shown. Except for the number of INR measurements in the usual care group (for which an assumption was made, see text), all number of procedures were observed or, if indicated with *, derived through questionnaires. ¥ The bootstrap p-value was calculated as the proportion of bootstraps in which the mean value in the intervention arm was higher than the mean value in the control arm. ECG = Electrocardiography; ACT = Ambulant compression therapy for crural ulcers; ECV = electrocardioversion.
| TYPE OF PROCEDURE | INTEGRATED CARE (N = 522) | USUAL CARE (N = 425) | DIFFERENCE IN MEAN COSTS | BOOTSTRAPPED P-VALUE¥ | |||||
|---|---|---|---|---|---|---|---|---|---|
| MEAN NUMBER OF PROCEDURES | MEAN COSTS | SD | MEAN NUMBER OF PROCEDURES | MEAN COSTS | SD | ||||
| PRIMARY CARE COSTS | GP consults | ||||||||
| Consults | 12.73 | € 120.05 | € 5.75 | 9.11 | € 86.04 | € 3.49 | € 34.01 | 1.00 | |
| Double consult | 4.11 | € 271.25 | € 12.41 | 2.41 | € 158.85 | € 11.21 | € 112.40 | 1.00 | |
| Home visits | 7.17 | € 358.52 | € 35.92 | 3.21 | € 160.37 | € 14.71 | € 198.15 | 1.00 | |
| Telephone consults | 6.09 | € 103.54 | € 5.93 | 6.45 | € 109.59 | € 7.66 | –€ 6.04 | 0.27 | |
| Practice nurse consults (chronic conditions) | 4.78 | € 90.53 | € 7.95 | 0.74 | € 64.22 | € 7.82 | € 26.31 | 0.99 | |
| Practice nurse consults (mental health) | 0.09 | € 4.78 | NA | 0.01 | € 0.74 | € 0.50 | € 4.04 | 0.93 | |
| ECG | 0.19 | € 8.56 | € 1.41 | 0.10 | € 4.04 | € 1.71 | € 4.53 | 0.97 | |
| Small surgery, injections, ACT | 0.52 | € 31.73 | € 4.48 | 0.70 | € 46.17 | € 10.39 | –€ 14.43 | 0.10 | |
| Other | 49.16 | € 50.21 | € 2.57 | 44.29 | € 41.88 | € 1.73 | € 8.33 | 1.00 | |
| Subtotal primary care costs | € 1,039.18 | € 41.60 | € 671.88 | € 30.46 | € 367.30 | 1.00 | |||
| CARDIOLOGY OUTPATIENT CLINIC VISITS | € 99.05 | € 5.34 | € 119.02 | € 5.96 | –€ 19.97 | 0.02 | |||
| ANTICOAGULANT TREATMENT COSTS | € 2,174.48 | € 34.38 | € 2,279.26 | € 33.34 | –€ 104.78 | 0.01 | |||
| ADMISSIONS AND ECV | Hospital admissions‡ | 4.16 | € 1,980.60 | € 127.30 | 3.79 | € 1,804.32 | € 176.85 | € 176.28 | 0.32 |
| Temporary nursing home admissions | 3.88 | € 652.05 | € 260.55 | 4.48 | € 753.04 | € 198.18 | –€ 100.99 | 0.70 | |
| ECV | 0.09 | € 17.61 | € 3.32 | 0.11 | € 21.62 | € 5.56 | –€ 4.02 | 0.27 | |
| Subtotal admissions and ECV | € 2,650.25 | € 334.14 | € 2,578.98 | € 313.32 | € 71.27 | 0.55 | |||
| Other outpatient visits* | 6.97 | € 445.13 | € 54.74 | 7.09 | € 481.17 | € 54.37 | –€ 36.03 | 0.43 | |
| Day treatment* | 2.06 | € 569.13 | € 92.09 | 2.10 | € 580.44 | € 109.35 | –€ 11.31 | 0.47 | |
| Paramedic consults* | 17.96 | € 570.21 | € 77.98 | 24.83 | € 798.55 | € 95.66 | –€ 228.34 | 0.02 | |
| Home care* | 140.59 | € 6,047.90 | € 1,108.93 | 144.81 | € 6,391.06 | € 1,263.16 | –€ 343.17 | 0.42 | |
| Day care institution* | 4.42 | € 1,016.94 | € 288.48 | 3.67 | € 700.52 | € 302.35 | € 316.42 | 0.80 | |
| Emergency department visit* | 1.02 | € 264.16 | € 52.03 | 1.02 | € 263.27 | € 47.63 | € 0.89 | 0.50 | |
| Ambulance ride* | 0.73 | € 375.30 | € 85.29 | 0.66 | € 341.48 | € 77.10 | € 33.82 | 0.61 | |
| Assisted living facility* | 15.98 | € 2,684.22 | € 913.75 | 24.05 | € 4,039.91 | € 1,263.32 | –€ 1,355.69 | 0.16 | |
| Subtotal other direct costs | € 11,972.99 | € 1,464.16 | € 13,596.39 | € 1,853.53 | –€ 1,623.40 | 0.23 | |||
| INDIRECT COSTS | Informal care* | 235.46 | € 3,083.70 | € 286.63 | 220.26 | € 3,296.45 | € 321.76 | –€ 212.75 | 0.30 |
| Total | € 18,845.16 | € 1,677.69 | € 20,262.72 | € 2,144.58 | –€ 1,417.56 | 0.21 | |||
Table 3
Imputed EQ5D-5L at different time points and the QALY contribution over 2 years for the intervention versus control group.
| INTEGRATED CARE (N = 522) | SD | USUAL CARE (N=425) | SD | DIFFERENCE | BOOTSTRAP P-VALUE | ||
|---|---|---|---|---|---|---|---|
| IMPUTED TIMEPOINTS | T0 | 0.766 | 0.009 | 0.756 | 0.019 | 0.011 | 0.22 |
| T1 | 0.718 | 0.013 | 0.706 | 0.018 | 0.012 | 0.25 | |
| T2 | 0.676 | 0.014 | 0.662 | 0.010 | 0.014 | 0.23 | |
| QALY CONTRIBUTION OVER TWO YEARS | QALY contribution unadjusted, censored patients included | 1.439 | 0.023 | 1.416 | 0.022 | 0.022 | 0.23 |
| QALY contribution adjusted, censored patients included | 1.428 | 0.020 | 1.429 | 0.019 | 0.000 | 0.37 | |
Table 4
Results of the cost-utility analyses of the integrated care intervention compared to usual care. Adjusted = for baseline differences in age, sex, Frailty Index and clustering. QALYs were also adjusted for differences in baseline EQ5D-5L utility score. Δ is the mean difference between intervention – usual care patients of 100 bootstrapped samples. The colours correspond to the different colours in Figure 2, indicating whether patients who were censored due to permanent nursing home admission, and their follow-up time while admitted to the nursing home, were included or not. ECV=electrocardioversion.
| Δ COSTS IN PRIMARY CARE | Δ CONSULTS CARDIOLOGIST | Δ ANTICOAGULANT COSTS | Δ ADMISSIONS AND ECV | Δ OTHER DIRECT COSTS | Δ INDIRECT COSTS | Δ PERMANENT NURSING HOME ADMISSION | Δ TOTAL COSTS, INCLUDING ASSUMED COSTS PERMANENT NURSING HOME ADMISSION | Δ TOTAL COSTS, WITHOUT ASSUMED COSTS FOR PERMANENT NURSING HOME ADMISSION | Δ EFFECTS (QALYS) INCLUDING PERMANENT NURSING HOME ADMISSION | Δ EFFECTS (QALYS) EXCLUDING PERMANENT NURSING HOME ADMISSION | |
|---|---|---|---|---|---|---|---|---|---|---|---|
| Base case | |||||||||||
| Imputed and adjusted | |||||||||||
| 522 vs 425 patients | € 363 | –€ 20 | –€ 105 | –€ 34 | –€ 941 | –€ 128 | –€ 478 | –€ 1,343 | –€ 865 | 0.002 | 0.000 |
| Bootstrapped 95% PI* | 254:435 | –32:–1 | –209:–24 | –737:1037 | –5532:2726 | –915:654 | –1226:138 | –6534:3109 | –5730:3641 | –0.035:0.046 | –0.040:0.042 |
| Sensitivity analyses | |||||||||||
| 522 vs 704 patients, imputed & adjusted | € 375 | –€ 17 | –€ 54 | –€ 337 | –€ 1,648 | –€ 1,013 | –€ 1,175 | –€ 3,868 | –€ 2,693 | 0.049 | 0.060 |
| Bootstrapped 95% PI* | 251:431 | –27:1 | –168:8 | –932:718 | –7618:1523 | –1957:–243 | –1904:383 | –9973:–50 | –8799:934 | 0.028:0.117 | 0.018:0.104 |
| Imputed, unadjusted | € 337 | –€ 17 | –€ 115 | € 71 | –€ 1,548 | –€ 213 | –€ 521 | –€ 2,006 | –€ 1,485 | 0.022 | 0.010 |
| Bootstrapped 95% PI* | 243:427 | –33:–1 | –210:–19 | –816:994 | –5973:2670 | –1017:594 | –1303:104 | –7273:2878 | –6606:3346 | –0.039:0.084 | –0.044:0.074 |
| Healthcare perspective** | € 151 | –€ 20 | –€ 105 | –€ 34 | –€ 941 | na | –€ 478 | –€ 1,427 | –€ 949 | 0.000 | 0.000 |
| Bootstrapped 95% PI* | 112:193 | –32:–2 | –209:–24 | –737:1037 | –5532:2726 | –1226:138 | –6594:2930 | –5904:3452 | –0.035:0.046 | –0.040:0.042 |
[i] * Bootstrapped 95% percentile intervals (contains 95% of the repeats).
** In the healthcare/third party payer perspective, informal care is not included and different unit costs for primary care consultations are used, as is specified by the Dutch Health Authority (see additional file, Table A2) [18]. Unit costs in the healthcare perspective are lower per consultation compared to unit costs in the societal perspective, as the residual costs are reimbursed separately through a fixed price per registered patient.
Figure 2
Cost-effectiveness plane. This figure shows the incremental costs (on the Y-axis) and incremental QALYs (on the X-axis) of integrated care compared to usual care of all the bootstrapped samples and, as is shown with the different colours, for the analyses with and without patients who were censored due to permanent nursing home admission. Negative costs (on the Y-axis) indicate cost-savings of integrated care compared to usual care, while positive costs (on the Y-axis) indicate additional spending. Negative QALYs (on the X-axis) indicate loss of QALYs due to integrated care compared to usual care, while positive QALYs (on the X-axis) indicate QALYs gained. The southeast quadrant therefore indicates the intervention to be dominant, i.e. more effective and less costly.