Paying someone to mow your lawn is a pretty straightforward affair. Ryan the lawn guy will look at the lawn size and maybe the hilliness of your yard, and you’ll settle on a price for mowing and trimming it. When you decide to contract for Ryan’s services on a more regular basis, payment might get a little more complicated. If you pay Ryan every time he mows your lawn he might mow it more often than necessary. But that problem is easily addressed by paying him a fee to take care of your lawn for the entire growing season.

Paying a hospital to care for someone who had a stroke is not so straightforward. Imagine you are an insurance company, and you decide to pay the hospital for each day the patient is in residence. With that kind of payment scheme, the hospital visit might drag on indefinitely. Indeed several decades ago insurance companies in the United States primarily reimbursed hospitals on a “per diem” basis, cool kid lingo for per day. Incentivized by this reimbursement scheme, the length of stay in American hospitals was often surprisingly long for even relatively mild conditions. Think of the parallel to lawn care: Pay per mowing and you can expect lots of mowings!

Health care payers have developed several methods to overcome this per diem/per mowing problem. I will explain these methods shortly, but first the bottom line. Figuring out how to pay for hospital care is a hell of a lot more complicated than figuring out how to pay your lawn service.

To combat the unintended consequences of per diem payment coverage, Medicare switched to per diagnosis payments in the 80s, and the insurance companies followed shortly thereafter. Under this DRG program, a hospital taking care of a patient with a severe stroke would receive appropriate payment for that diagnosis, while a hospital taking care of someone with mild pneumonia would receive a smaller payment appropriate for the typical costs of paying for that condition. Following the implementation of this type of prospective payment, length of stay in American hospitals plummeted. In response, much of medical care was shifted to post-acute care, to rehabilitation hospitals, for instance, for stroke patients, or to outpatient clinics for people with pneumonia. These non-hospital services were not covered by the diagnosis-based DRG payments, so health care providers had little incentive to practice parsimoniously once their patients left the hospital.

Enter bundled payments. In 2013, the Center for Medicare and Medicaid Services, henceforth CMS, launched its Bundled Payments for Care Improvement Initiatives, henceforth BPCI (in case your life needs more acronyms). In the BPCI, CMS identified 48 clinical conditions that qualify for bundled payments, meaning participating health care providers would receive payments designed to cover not only hospital care for the condition in question, but money to pay for all health care related services they receive for the next 30 days. Unsurprisingly, not all hospitals are eager to join this program. For starters, the hospitals have to be financially integrated with post-hospital providers. If patients receiving stroke care at Our Lady of Acute Care Hospital receive post-acute care from a hodgepodge of rehabilitation facilities, many of which have no connection to the hospital, then coordinating payments will be a disaster.

According to a recent study, a little more than 10 percent of hospitals have signed up to participate in the bundled payment program. Participating hospitals tend to be larger and more often have academic affiliations than nonparticipating hospitals.

The study also provides a picture of where the bundled payments are going — the amount going for acute hospital care versus post-acute care versus outpatient care for instance. And interestingly, hospital payments account for less than half of expenditures, with post-acute care and provider fees making up much of the rest:

Moreover, when looking across hospitals, the biggest difference in Hospital A to Hospital B was typically in post-acute care:

The researchers also discovered that many hospitals that joined the program later dropped out. And this brings us back to just how much more complicated it is to pay for hospital care versus lawn service.  Think for a moment about what a daunting task it is to define a “bundle” of care. Suppose a patient is admitted with a stroke. The onset of her condition is clear. But which of the following health care problems, all occurring in the next 30 days, should be included in the bundled payment? Suppose the patient falls and breaks her hip. That one is easy. Services to treat this problem should be covered by the bundled payment, because good health care would have prevented the fall. Here’s another easy one. Suppose the patient, 25 days after the stroke, develops acute leukemia requiring emergent chemotherapy. That diagnosis is unconnected to the stroke, and should be billed separately. But now for some potentially tougher diagnoses. Should the hospital receive extra money to treat the patient for pneumonia she’s contracted 12 days after her stroke? What about coverage of insulin therapy for the spike in blood sugar she experienced on day 17? What about consultation at day 21 from a psychiatrist because she has become depressed? What about treatment for a new rash, unconnected (as far as anyone can tell) from the patient’s stroke care?

There is nothing simple about paying for health care. Health care is by necessity a mess. Administratively, some systems are more messed up than others. But all systems are flawed because every way we choose to pay for health care potentially lures health care providers away from offering the best possible care.

When people say they know a simple way to fix any health care system — whether in the USA or elsewhere — I have simple advice for how to evaluate their solution. Dismiss it.

Peter Ubel is a physician and behavioral scientist who blogs at his self-titled site, Peter Ubel and can be reached on Twitter @PeterUbel. He is the author of Critical Decisions: How You and Your Doctor Can Make the Right Medical Choices Together. This article originally appeared in Forbes.