My home hospital is small. In a town of just over 20,000 people, this hospital has 25 beds and is designated “critical access” by Medicare because it is felt to be necessary to the health care of the community. Critical access is a designation which was introduced in 1997 when modernization of Medicare payment systems threatened to close a large proportion of hospitals in small communities which were unable to benefit from economies of scale. A small hospital needs to remain capable of providing services to small but significant floods of patients when everyone gets sick at once, and maintain skills and staff to care for a variety of different diseases. This can make such a hospital less financially efficient (though various other factors can partly offset this) and the critical access program spares my hospital some of the Medicare belt tightening that would make admitting Medicare insured patients prohibitively unprofitable.

Most larger hospitals are now paid for their Medicare insured patients according to the Prospective Payment System, in which a fixed amount of money is paid to the hospital according to what disease the patient has. This means that if a patient presents with pneumonia the hospital will get maybe $8,000 for all of the expenses involved in treating the patient, excluding doctor fees which are billed separately. If the patient does very well and only stays in the hospital 2 days, the hospital does very well and if the patient languishes and requires more resources, the hospital loses money.

It is designed, ideally, to work out so the hospital survives, covers its costs and makes a profit. This encourages hospitals to make patients well as soon as possible by making sure that they don’t profit from taking poor care of patients. It’s generally a good system. Small hospitals, though, are very sensitive to individual patient disasters, and if a patient or two uses more resources than the prospective payment covers, the hospital can fail. For this reason critical access hospitals are paid “fee for service.” Medicare pays 101% of its allowable fee for each service a patient uses in the hospital. Overall, this can end up costing Medicare more per patient, but it does keep small hospitals afloat.

In order to be critical access, a hospital must have 25 or fewer beds, though it can also have separate units for rehabilitation or psychiatric services, at 10 beds each. It must be 35 miles by primary road or 15 miles by secondary road from another hospital, or 15 miles in mountainous terrain. Up until 2006, it was also possible for the states to designate a hospital as a “necessary provider.” Hospitals so designated remain critical access, though this loophole is no longer available. The 2014 budget will address this definition again, proposing to remove critical access status from hospitals within 10 miles of another hospital.

Because critical access hospitals don’t have the usual built in incentives of the prospective payment system to limit the length of hospital stays, theoretically they can just make more money by keeping patients longer. Thus the 96-hour rule. Originally it was stipulated that critical access hospitals couldn’t keep patients longer than 4 days. Since this was patently ridiculous, the law was changed to allow hospitals to average no more than 96 hours, which was a wee bit better.

So why do we even want to have little hospitals? They don’t have very many medical specialists, they don’t have dialysis units or cancer centers or cardiac cath labs. Sometimes they can’t do MRI scans or nuclear stress tests, and the littlest ones can’t even do surgeries. When you put all critical access hospitals (there are about 1300 in the US) in a big pot and average their results they have a higher mortality for many diagnoses than the bigger hospitals.

Some of these hospitals are probably not good. But so are many larger hospitals. Communities often depend on the existence of hospitals. Businesses don’t want to locate themselves far from a hospital. When a local hospital closes, the chronically ill and the poor who have trouble with transportation are prohibitively far from health care. Small hospitals support the work of primary care doctors with labs and radiology suites and emergency services. Communities without a hospital often have only very few and basic outpatient providers. Small hospitals often provide more appropriate care for patients because everyone knows everyone else, so we remember what happened last time, what worked, what didn’t. The nurses know who drinks on the sly or is abused by their husband. The specialists and the hospital physicians know each other well and also know and share patients which makes care more personal and makes patients feel more secure. Communication is excellent and things get done right away.

For those patients who have problems that are beyond the scope of a small facility, the emergency department can stabilize and transport. In our hospital we have a heliport on the roof and our patients can be up and out and at the major referral center that is 90 miles away in next to no time. The very old and very sick who want comfort rather than life prolonging procedures can get that right near home, with physicians who know them and often know their family and friends, and share their concerns.

So I guess I am a fan of good little hospitals. But I am not necessarily a fan of the critical access system. Our hospital is occasionally so full we need to turn away patients, but we have more actual physical room and could definitely squeeze in a few more patients if we weren’t limited to 25 beds. Also the fee for service system really does not encourage us to develop innovations that make patients get better faster. I think we do a great job at my local hospital of reducing over-testing and over-treatment, but there is no financial incentive to do so since we make more money if we do more. We would be better at creative thriftiness if it actually saved us money. Our creativity is stifled by the need to be little and by being paid fee for service.

Lately we are being encouraged to take the 96-hour rule more seriously. We need to certify when we admit a patient that they will only require 96 hours in the hospital. If we think they are likely to be in our hospital longer than that, we theoretically need to transfer them to the larger hospital 90 miles away.

Seriously? And if we think they will only need 96 hours but we are wrong, theoretically we should then transfer them. So our dear sweet 96-year-old who is admitted with diarrhea and turns out to have a small bowel obstruction and then develops pneumonia would need to take an ambulance ride or a $40,000 helicopter ride on day 4 just as we have her really nicely stabilized. The alcoholic with pancreatitis who never gets better in 4 days but is perfectly appropriate for our level of care would go straight from the emergency room to the tertiary care facility. The overworked doctors at those hospitals are so definitely not going to love that, as they deal with their own sick patients.

I’ve been on both sides of this situation, as the doctor transferring a patient and as the one receiving the patient. If a patient is being transferred solely because it is difficult to design an appropriate discharge, it is more difficult to do that from a hospital 90 miles away from their home. It is also extremely likely that subtleties of the history will be lost and that procedures and tests and mistakes will be repeated. Transfer to a different hospital is necessary if the patient’s recovery is stalled because of lack of expertise or technology, but if it is only to save the payer (Medicare) money, I expect it will do the opposite.

What to do? I’m thinking that the whole critical access system needs to be reconsidered. Managing all small community hospitals the same way doesn’t make sense. Clearly little hospitals need support in order to survive, but giving them arbitrary rules and paying them fee for service is hardly the way to make them function better and certainly doesn’t encourage creative innovation.

Janice Boughton is a physician who blogs at Why is American health care so expensive?