Table 1. Input Parameters.
| Parameter | Base-case value (range in sensitivity analyses) | Distribution for probabilistic analyses | Source |
|---|---|---|---|
| Transitional probabilities for the GDMT (control) arm | |||
| Rate of HF hospitalizations, per person year | Petrie et al,11 2020 | ||
| No diabetes | 0.080 (0.071 to 0.089) | β | |
| Diabetes | 0.122 (0.111 to 0.133) | β | |
| Proportion of HF hospitalizations that are fatal | 0.0954 (0.035 to 0.105) | β | Wadhera et al,12 2018 |
| Probability of 30-d readmission after a HF hospitalization | 0.20 | β | Wadhera et al,12 2018 |
| Proportion of HF-specific readmissions | 0.37 | β | Krumholz,13 2013 |
| Rate of urgent HF visits, per person year | Petrie et al,11 2020 | ||
| No diabetes | 0.006 (0.003 to 0.009) | β | |
| Diabetes | 0.007 (0.004 to 0.010) | β | |
| Rate of incident diabetes, per person year | 0.037 (0.030 to 0.044) | β | Inzucchi et al,14 2020 |
| All-cause mortality in first 24 mo, per person yeara | Petrie et al,11 2020 | ||
| No diabetes | 0.078 (0.069 to 0.087) | β | |
| Diabetes | 0.117 (0.106 to 0.128) | β | |
| Death from any cause (beyond 24 mo) | Ratio for all-cause mortality comparing the control arm to the US general population (see eMethods in the Supplement) | ||
| Effectiveness of dapagliflozin (intervention arm) | |||
| HR for HF hospitalizations, compared with GDMT | Petrie et al,11 2020 | ||
| No diabetes | 0.63 (0.48 to 0.81) | Log normal | |
| Diabetes | 0.76 (0.61 to 0.95) | Log normal | |
| HR for urgent HF visits, compared with GDMT | Petrie et al,11 2020 | ||
| No diabetes | 0.25 (0.07 to 0.89) | Log normal | |
| Diabetes | 0.62 (0.24 to 1.59) | Log normal | |
| HR for death from any cause compared with GDMT (first 24 mo) | Petrie et al,11 2020 | ||
| No diabetes | 0.88 (0.70 to 1.12) | Log normal | |
| Diabetes | 0.78 (0.63 to 0.97) | Log normal | |
| HR for incident diabetes compared with GDMT | 0.68 (0.50 to 0.94) | Log normal | Inzucchi et al,14 2020 |
| Costs | |||
| Dapagliflozin therapy, $/y | 4192 (953 to 6188) | Log normal | Base case: FSS-Big 4; lower bound: heavily discounted price; upper bound: wholesale acquisition (all prices estimated August 2020)15,16 |
| Background health care costs, $/y | HCUP17; Peterson et al,18 2015; Kazi et al,19 2020 | ||
| No diabetes | |||
| Age, y | |||
| <75 | 20 629 (16 503 to 24 755) | Log normal | |
| 75-85 | 22 512 (18 010 to 27 015) | Log normal | |
| >85 | 30 811 (24 648 to 36 973) | Log normal | |
| Diabetes | |||
| Age, y | |||
| <75 | 28 923 (23 139 to 34 708) | Log normal | |
| 75-85 | 26 430 (21 144 to 31 716) | Log normal | |
| >85 | 34 249 (27 400 to 41 099) | Log normal | |
| HF hospitalization costs, $ | 11 827 (8899 to 15 591) | Log normal | Medicare Provider Utilization and Payment Data 201720 |
| Urgent HF visit cost, $ | 807 (646 to 968) | Log normal | Charges for services provided during an urgent care visit and a cost-center–specific charge-to-payment ratio |
| Quality of life | |||
| Baseline KCCQ-OSS in the GDMT (control) arm | 68.6 (68.1 to 69.1) | Normal | Kosiborod et al,21 2020 |
| Baseline KCCQ-OSS in the dapagliflozin (intervention) arm | 68.4 (68.1 to 69.1) | Normal | Kosiborod et al,21 2020 |
| KCCQ-OSS in the GDMT (control) arm at 8 mo | 72.7 (72.0 to 73.2) | Normal | Kosiborod et al,21 2020 |
| KCCQ-OSS in the dapagliflozin (intervention) arm at 8 mo | 75.0 (74.4 to 75.4) | Normal | Kosiborod et al,21 2020 |
| Quality-of-life penalty applied for diagnosis of diabetes | −0.0351 (−0.0350 to −0.0352) | Normal | Sullivan et al,22 2006 |
| Quality-of-life penalty applied for HF hospitalization | −0.0066 (−0.0135 to 0) | Normal | Jaagosild et al,23 1998 |
| Quality-of-life penalty applied for urgent HF visit | −0.0045 (−0.009 to 0) | Normal | Jaagosild et al,23 1998 |
Abbreviations: FSS, Federal Supply Schedule; GDMT, guideline-directed medical therapy; HCUP, Healthcare Costs and Utilization Project; HF, heart failure; HR, hazard ratio; KCCQ-OSS, Kansas City Cardiomyopathy Questionnaire–Overall Summary Score.
a
Ratio for all-cause mortality comparing the control arm with the US general population beyond 24 mo available in the eMethods in the Supplement.