This Figure provides a detailed illustration of the complexity of the input parameters for the hospital and DME stakeholders for patients on NIPPV vs no NIV.

Looking at the 2 lefthand columns in the Figure, there is no cost to the hospital related to acquisition of any device for patients who are not receiving any NIV because there is no device to be acquired, but in the same sense there is no cost to the hospital to acquire the NIPPV device either.  There is a cost related to readmissions that, based on a variety of sources, averages out at roughly $9200 per patient per readmission.  However we must account for the differential number of admissions that occur per patient per year, which was 3.2 for the no NIV population and 0.16 for the NIPPV population.  Of these readmissions, it is estimated that about 7% of these occur within the first 30 days, and the hospital would not receive any additional reimbursement for readmissions occurring within that 30-day window.  However admissions occurring outside the 30-day window would be seen as a new episode of care, for which reimbursement would apply.

There is a cost related to a split-night sleep study, which was estimated at $2500, and this cost would apply to both no NIV and NIPPV situations.  However no sleep studies were required, so this cost does not apply.

The mortality rate shown for years 0–3 are calculated at 29% for no NIV versus 18% for NIPPV.  The dropout rate is estimated to be roughly 17% for both populations; it remains to be seen if different technologies may actually provide greater adherence and therefore reduce that dropout rate, but for this model it is assumed that the dropout rates are comparable.

In the second column, looking at the DME provider, again there is no acquisition cost for no NIV, but for the NIPPV there is an acquisition cost for the device, the MSRP of which is $17,600.  There is a setup cost of roughly $500.00.  Repair and replacement after warranty for each device averages out to roughly $14.00 per device, and supplies are in the range of approximately $1000.

The readmission rates are the same here as shown for the hospital.  However there is a supportive care cost, which is typically provided for these home-ventilated patients and particularly patients who are treated with the NIPPV, and that cost is estimated at roughly $1900 per patient per year.

The mortality and dropout rates for DME providers are calculated as the same as for hospitals.  Regarding reimbursement (righthand column), then again new episodes occurring outside the 30 days count as new episodes and are reimbursed at roughly the same as it costs to treat these patients in-hospital, at roughly $9200.  When the costs of those unreimbursed 30-day readmissions are averaged, the average reimbursement is actually closer to $8500 per patient, and that reimbursement is the same whether the patients are receiving NIPPV or no NIV.

Finally, the DME provider perspective accounts for the reimbursement that the DME would receive for the device, which, based off the CMS fee schedule, is approximately $1600 per year per device that is in use.