With so much focus on surprise medical bills, new CFPB rules and pressure to shift to value-based care, it’s easy to gloss over the fact that hospital margins are razor-thin. As a quick (and painful) refresher:
- Hospital operating margins dropped nearly 40% between 2015-2017, roughly $6.8 billion.
- In 2018, the median operating margin for hospitals was 1.7% (described as “anemic” by Moody’s).
- Bad debt levels are between 8-9% driven by high deductible plans with ever-higher deductibles.
Anyone can argue about all the ways the system is broken — that doesn’t change the fact that hospitals need tools and innovations that improve their margins, now and into the future. Hospitals are in a position where they need to make every available effort to minimize costs and increase collections.
At Sift, we leverage healthcare financial data and artificial intelligence to improve the revenue cycle — increasing efficiencies and improving payments from both payers and patients. We are finding that every area of cost reduction in the revenue cycle and every magnitude of increase in collections is impactful for our partners.
We’ve recently concluded a pilot with a partner, a small test working on a small subset of their patient collections. As our team reviewed the results of this pilot, I was struck by how much of an impact AI can have for even small subsets of patient groups. Small improvements and seemingly small-sounding results have a significant impact. From this pilot, here is a quick illustration. Three improvements that could easily fall under the radar, but are material:
1) Increased Liquidation Rate
For our pilot partner, in patient segments where payment plans are not currently offered but should be offered, we’ve identified the best payment plan structure, contact method and contact frequency to get those patients on plans and drive payment. By proactively offering payment plans to these eligible accounts, liquidation rate will increase by 50% for these segments.
2) Strategic Outbound Calls
In this pilot, we segmented the patient population based on model scores and determined how to best influence payment. Some patients, who have a moderate likelihood of paying off their balance, need several phone calls (3+) spaced at strategic intervals, to maximize payment. With these patients, this methodical effort is worth it — they are influenced by a structured call cadence. We identified which patients to call, how many times and when.
3) Decrease Total Calls
By identifying who will always pay their bill and who will never pay their bill — regardless of receiving a phone call — we know which outbound calls to eliminate. In our test segment, we will decrease annual outbound phone calls by 50,000 (our early, most conservative estimate), without a decrease in liquidation. This frees up time and effort to work the accounts that will pay.
These improvements (and others, not listed here) will be made within our partner’s current systems, fully integrated into their revenue cycle workflows. At scale and with a feedback loop that drives continuous improvement, there is no doubt that these efforts to bring AI to the revenue cycle will have a powerful impact, not just in terms of payments, but efficiency and intelligence. These are measures that are increasingly valuable in healthcare and will positively impact margins now and as we trend towards consumer-driven healthcare.