Table 1.
Key unit costs and probabilistic sensitivity analysis parameters
| Parameter | Base case input | Reference | Detail | Low | High | Assumed distribution* |
|---|---|---|---|---|---|---|
| Reimbursement/cost ratio, % | 87.0 | 27 | — | 43.5 | 1.31 | β |
| Clinic/HOPD cost ratio, % | 38.0 | 28 | — | 19 | 57 | β |
| HOPD/IP cost ratio, % | 35.0 | 29 | — | 17.5 | 52.5 | β |
| HOPD office visit, $ | 115.85 | 21 | APC 5012 | 57.93 | 173.78 | γ |
| Standard inpatient bed day, $ | 2420.77 | 19 | Adjusted to 2019 US$ | 1210.38 | 3631.15 | γ |
| ICU bed day, $ | 7181.84 | 25 | Adjusted to 2019 US$ | 3590.92 | 10 772.75 | γ |
| CT head scan with contrast, $ | ||||||
| Nonfacility TC | 107.76 | 20 | HCPCS 70460 | 53.88 | 161.64 | γ |
| 26; phys. comp. | 58.38 | 20 | HCPCS 70460 | 29.19 | 87.57 | γ |
| TC; APC | 480.77 | 21 | APC 8006 | 240.39 | 721.16 | γ |
| MRI brain scan with contrast, $ | ||||||
| Nonfacility TC | 227.77 | 20 | HCPCS 70552 | 113.89 | 341.66 | γ |
| 26; phys. comp. | 91.54 | 20 | HCPCS 70552 | 45.77 | 137.31 | γ |
| TC; APC | 855.60 | 21 | APC 8008 | 427.80 | 1283.40 | γ |
| PET scan with contrast, $ | ||||||
| Nonfacility TC | 1375.61 | 20 | HCPCS 78608 | 687.81 | 2063.42 | γ |
| 26; phys. comp. | 73.52 | 20 | HCPCS 78608 | 36.76 | 110.28 | γ |
| EEG, $ | ||||||
| Nonfacility TC | 375.89 | 20 | HCPCS 95819 | 187.95 | 563.84 | γ |
| 26; phys. comp. | 59.46 | 20 | HCPCS 95819 | 29.73 | 89.19 | γ |
| TC; APC | 252.31 | 21 | APC 85722 | 126.16 | 378.47 | γ |
| Lumbar puncture, $ | ||||||
| Nonfacility | 152.09 | 20 | HCPCS 62270 | 76.05 | 228.14 | γ |
| Facility | 80.37 | 20 | HCPCS 62270 | 40.19 | 120.56 | γ |
| Mechanical ventilation, $ per d | 2644.91 | 25 | Adjusted to 2019 US$ | 1322.46 | 3967.37 | γ |
| Dialysis, $ per d | 611.73 | 20 | APC 5401 | 305.87 | 917.60 | γ |
| Vasopressin, $ | 301.25 | 24 | NDC: 63323-0302-09 | 150.63 | 451.88 | γ |
| Norepinephrine, $ | 77.67 | 24 | NDC: 55390-0002-10; 25 μg/min | 38.84 | 116.51 | γ |
| Siltuximab, $ | 8571.74 | 24 | NDC: 50242-0135-01 | 4285.87 | 12 857.61 | γ |
| Anakinra, $ | 155.01 | 24 | NDC: 66658-0234-07; 100 mg | 77.51 | 232.52 | γ |
| Tocilizumab, $ | 4477.30 | 24 | NDC: 50242-0135-01; 800 mg | 2238.65 | 6715.95 | γ |
26; phys. comp., physician component; APC, Ambulatory Payment Classification; CT, computed tomography; EEG, electroencephalogram; HCPCS, Healthcare Common Procedure Coding System; HOPD, hospital outpatient department; IP, inpatient hospital; MRI, magnetic resonance imaging; NDC, National Drug Code; PET, positron emission tomography; TC, technical component.
The β distribution is applied to model the behavior of random variables limited to finite length. Given that estimates cannot be <0, the β distribution was selected. The β distribution is a suitable model for the random behavior of percentages and proportions. γ distribution is assumed for cost data, specifically for using small sample sizes that often have nonnormal distribution (mean and median are not similar); thus, cost studies often use a γ distribution to address the skewness of cost data that are subject to wide variance and likely a wide range, with a few outliers that affect the mean cost.