The explosive growth in the use of implantable medical devices has been accompanied by an alarming growth in the number of adverse events and recalls related to them, costing payors and patients billions of dollars a year for revision surgeries, some of them life-threatening.
Just one model of a replacement hip generated 23,193 recalls necessitating revision surgery. The cost for each primary surgery averaged $31,000. The revision surgery added $54,000 in cost for the payor for a total expenditure of $85,560 per patient. The total additional cost to payors who had approved that specific replacement hip device was $1,984,393,084. Had a more costly but more effective device with an average cost of $42,000 per procedure been approved and used, no revision surgeries would have been required and the total cost to payors for those 23,193 patients would have been $974,106,000, saving payors over $1 billion.
Another example can be seen in the use of gastric lap bands that became a popular solution for weight loss in 2001. However, it wasn’t until 2009 that it finally became evident that an increasing number of problems with the device were requiring remedial surgeries. The average cost for the initial surgery was $14,500, but revision surgeries were costing payors an average of $25,520 per patient, or $2,552,000 for 100 patients. Between 2011 and 2014 these revision surgeries cost payors a total of $1 billion in additional expenditures.
Our technology, Formulary PREDICT™, the only commercially-available formulary for medical devices, offers payors the opportunity to avoid these unnecessary expenditures. Gathering real world evidence from millions of patients, claims data, CMS, and FDA data bases we identify and predict potential adverse events and recalls of devices, as well as the effectiveness of a device in patients with a specific medical and demographic profile using our machine learning models. Formulary PREDICT™ enables a payor to develop a risk-based predictive formulary for medical devices and build cost and risk sharing models with providers.