In the Trenches versus the Bird's Eye

Timothy Noah likes to state as a fact that the Veterans' Affairs Hospitals are unequivocably superior to conventional hospitals, in particular citing this NEJM article. Yet it is a doctors' truism that the VA Hospitals provide the worst patient care. They joke about the poorly trained, heavily organized nurses ("What's the difference between a VA nurse and a bullet? A bullet can be fired, a bullet draws blood, and a bullet only kills once."). This is a uniform opinion held by almost every doctor I know in every part of the country.

Anecdotes abound: I had a patient, in severe pain, who was being so ignored by his nurse he had to call his daughter on his cell phone to call the hospital to call the nurse manager to call his nurse.  I have, on multiple occasions, walked into a patient's room, to find him on the brink of death or other danger, unattended for hours.

But it isn't just the nursing. One of my colleagues, in trying to care for an inpatient having an acute stroke, unable to rouse any urgency from the staff,  called 911 to transport the patient to a better hospital.

It's not just the floor nurses or the transport staff. In the OR, cases can take 150% to 3 times longer, as in 6 hours instead of 2. Often needed equipment is missing, or broken; the staff are not properly trained to use the equipment. On several occasions, the OR table was missing. (Tricky to operate without one.)

It's not just staff incompetence. OR time is severely restricted (it's expensive to run one - our VA has 8 ORs and by policy, they are only allowed to run 6), so urgent cases get delayed for months. Specialists cost more money, so there is a shortage of them generally. There is, for example, almost no dedicated neuroradiology, so patients have to rely on general radiologists -- who are not competent. In 6 months, I reviewed ~1000 films, and I can think of dozens of errors, most resulting in delayed treatments and at least one unnecessary surgery.

So is Noah just wrong that the VA is superior? Or is there some phenomenon that studies can measure, but that doctors cannot perceive resulting in better outcomes? The NEJM article has at least 3 flaws:

1) the endpoints are all in the realm of general medicine, which may hide the VA's lack of specialty care. Focusing on resources the VA doesn't pay for might highlight some major problems.

2) the endpoints are all in the realm of general medicine, ignoring surgical care. Looking at surgical subspecialty care might also reveal inequities: it is rumored that there is a dataset showing that VA patients with brain tumors only live 60% as long as their private insurance counterparts.

3) the endpoints were all specific tasks the VA had been working to improve for many years. Hardly a fair comparison! If you decided to test 2 classes of students, and to 1 class you gave a copy of the test questions months in advance, you had better hope they did better! Moreover, before the VA began specifically improving these specific areas, they were performing horribly (e.g., ~27% flu vaccination), far below their private insurance counterparts. Which suggests that, in the absence of specific campaigns to improve specific measures, the VA may be performing far worse than fee-for-service.

So, is socialized medicine better than fee-for-service? Maybe at some things (e.g., organizing campaigns to improve certain targets) and maybe with time it can be on many counts ... but for me and mine, should they need health care near a VA, just drive on to the next hospital.