2007 OPEN FORUM Abstracts
A MATHEMATICAL MODEL FOR PREDICTING HOME OXYGEN TECHNOLOGY ATTRITION RATES AND RENTAL LIFE OPERATING UNDER A 36 MONTH PAYMENT CAP
J. Lewarski1, R. Chatburn2
Background: In 2005, Medicare imposed a payment cap for home oxygen, which stops payment and transfers ownership of the oxygen equipment (OE) from the home medical equipment (HME) provider to the patient following 36 consecutive rental months. The new payment policy significantly changes and complicates the OE business model, creating a high level of uncertainty for HMEs. The purpose of this study was to predict the equipment attrition rate and rental life expectancy for OE.
Methods: The fraction (F) of initial patients remaining on oxygen as a function of months (M) of use was calculated from Medicare patient claims data (n = 3,754): F = -0.218 x LN(M)+ 0.9997, where LN = natural logarithm (r2 = 0.998). A spreadsheet model was developed using Medicare oxygen patient claims data for dates of service covering 1-36 months. The same Medicare data were used to predict the months of OE use (M) for a random patient: M = 97.765*EXP(-4.5802*X), where X is a random number between 0.001 and 1. This was used to create a hypothetical sample of 1,000 consecutive patients being assigned OE. The sample was analyzed to create another data set (n= 200) of total months of useful service for unit at time unit removed. Assumptions were that M could not be > 36 (i.e., OE ownership would transfer to patient at 36 months) and if M < 36, the unit could be reused until ownership transferred or useful life (60 months) expired, whichever came first. The data set was re-sampled 1,000 times to determine expected rental months per OE unit. Resampling was repeated for businesses with different fleet sizes.
Results: Medicare oxygen claims data indicate that 78% of patients started on oxygen will remain on service for less than 36 consecutive months. The re-sampling procedure showed that although highly random and unpredictable on a patient to patient basis, 20% of a given fleet of OE will be owned by the first patient using them. The remaining 80% of OE can be re-deployed into the rental fleet after the first rental. The total rental months before leaving the fleet depends on the fleet size as shown in the table. Prediction uncertainty is greater with smaller fleet sizes, but the coefficient of variation was only 2% to 6%.
Conclusions: The 36-month payment cap and OE transfer of ownership creates a very complex but relatively predictable business model for HMEs required to manage large fleets of capital equipment serving a very dynamic patient population with varying lengths of need.