I was deeply involved in hospice software, and specifically developing hospice revenue cycle solutions, about 15 years ago, and I can certainly confirm that the Medicare/Medicaid reimbursement formulas -- while in most areas arguably about as well-designed as you're going to get with such a program -- incentivize fraud because of the odd combination of up-front incentives and retrospective clawbacks. For well-run hospices, these aren't major issues (clawbacks aren't uncommon, but aren't financially controlling), but for hospice systems built to drain public cash out of dying patients, it's every bit as bad as the article says. (I did find that, using public Medicare hospice data, it was very easy to find providers that -- comparing my model to DOJ prosecutions -- were engaged in patient-stuffing and fraud by looking at their patient population data. Unsurprisingly, there was a significant correlation between questionable patient censuses and financial performance.)

Generally, hospice is a hard case -- I'd never put a loved one (and I didn't) through one of the major providers, while smaller hospices have their own issues even if they're well-run, because Medicare/Medicaid reimbursement isn't really sufficient at the small scale (many are non-profit for this reason). I'd like to see more money directed towards hospice while restricting hospice size (such as putting caps on total patient census) to try and reduce the incentives for fraud and gaming the system, but to be honest I don't think there's a good way to fix the problems inherent in the model.

(Also, the government should be a lot more aggressive, and licensing bodies more strict, on penalties against these companies. Yank the medical licenses from medical directors, slam some FTC decrees on named individuals, and put some execs in prison -- rather than, say, letting them go on to get elected to the Senate -- and the risks to medical fraud will start to outweigh the value for a lot of people.)