Rating scales pervade health care.   Once again, we learn that they can kill.

The latest scandal again comes from our single national, regionalized health care system, the VA. The newest story involves a VA hospital in a logging town in Eastern Oregon.

There will be many others.

The Trouble With Roseburg

The  VA hospital system has had many problems.  To help fix them, they devised a five point rating scale for hospitals. As described in a report from the investigative team at the NY Times,  Roseburg was ranked 1 out of 5, the lowest.

The new administrator of Roesburg was determined to improve these “metrics.” Enlisting the aid of his chief of mental health services, he reviewed the numbers and recognized hospital admissions  made up a big part of his hospital’s low ranking.

A new plan was devised: keep sick patients out.  They would be transferred to other VAs or  private hospitals, a very, very expensive process.  But those pots of money lay elsewhere, not in his administrative budget.

So Walter Savage, a former Air Force Mechanic, was refused admission though his body was covered with ulcers, he had just broken ribs in a fall, was dehydrated and malnourished.

The ER doctors all wanted him admitted.  They were overruled. Mr. Savage was sent home.

He came back.  This time a doctor refused the administration’s request and admitted him.

Within 24 hours, he was sent to a nursing home.  While over half the hospital’s beds were empty.

During his tenure, Roseburg’s administrator got their ratings up from one to two stars.  His hospital was recognized as a “rising star” within the VA system.

Another way to game the ratings was to make sure patients did not die in the hospital or within 30 days of discharge.  Roseburg came up with an ingenious solution – declare the patient a hospice case.

Often patients were declared hospice cases against the desires of the physicians.  Sometimes the patients were switched to hospice care without their knowledge.

Hospice patients don’t receive active treatment.  If that treatment might have saved you, you were out of luck.

Another  administrative gambit was to have patients with congestive heart failure classified as  “hypervolemic.” That diagnosis didn’t appear in the rating scales.

In 2015, Roseburg lost 17 out of 23 primary care physicians, while its ratings went up.

It’s Not Just Roseburg

The Times team also interviewed Dr. Michael Mann, who used to run Thoracic Surgery at the San Francisco VA, a major teaching hospital.  He helped in putting into practice a new VA ranking system for surgical complications.  From 1997 to 2007, surgical complications “dropped” in the VA system by 47%.

A remarkable result.
“Of course quality had not really improved by that much. People had just learned to make it appear that it had,” Dr. Mann noted.

How did the numbers get massaged?  You just stopped doing high risk procedures because the results might not look good.  “I’m very ashamed,” Dr. Mann explained.  “I colluded.  I was told not to operate and pulled back, and at least one of my patients died of it.”

Ratings in a Time of Guidelines

Is gaming health care rating systems new?  Not at all.  Over thirty years ago I was teaching at the University of Texas, Houston.  New headlines declared that a hospital none of us had heard of had the best cardiovascular program in the city.

Better than Baylor.  Better than St. Luke’s.  Two programs that were ranked not as national but global leaders under doctors like Michael DeBakey.

How had Rose Hospital managed this feat?  The answer was simple – they took no complicated cases.

The surgeons I asked declared it was generally acknowledged Rose as one of the worst cardiovascular programs in the County.  They would not send anyone there – especially if they were sick.

But Rose had the best numbers.

Recently ratings games hit closer to home when I and my colleagues were audited by Medicare.

After passage of the Mental Health Parity Act in 2004, psychiatric treatment was supposed to be on par with medical treatment.  In part due to poor reimbursement, people had not flocked to careers in psychiatry.  That would hopefully change.

So I was surprised to find that the auditors had “split” my chart into a medical and psychiatric parts.  They did not look at the “medical” part.

As many of my patients have complex med-psych problems like sleep apnea, bipolar disorder, metabolic syndrome and hypersomnia, I was surprised.   One point of the Mental Health Parity Act was to have medical and psychiatric issues dealt with coherently, as a unity.

No, I was told, that’s not correct.  I would be audited as if I were two separate physicians, doing two separate charts on two separate patients.

Except there was only one doctor and one patient.

Does anyone imagine splitting physicians and patients in two is good for care?

It may work for the bottom line.  The auditing company gets paid more for noting “deficiencies.” That’s easier when you retrospectively make up new rules on what constitutes “medical errors.”  Making documentation requirements more onerous means fewer psychiatrists and other physicians taking Medicare.  Presently only 26.5% of medical subspecialists nationwide do take Medicare.  If you have few or physicians in a region taking Medicare, your regional payouts go down.

You’ve gotten under budget!

The only problem is that when you do this in psychiatry, medical costs of patients skyrocket. The already high medical costs of psychiatric patients is one famous reason why people with mental health diagnoses have such a tough time getting insurance. But the auditors have their money.  And the increased costs come out of a non-mental health pot.   Not our problem.

Except it’s yours.

Rating the Raters

In a time when physicians are being converted to “data entry technicians” or in my terms, checklist monkeys, it’s necessary for independent clinicians and clinical epidemiologists to rate the raters. Are these checklists and rating scales improving public health outcomes?

If they’re not, we’ll all be sliding to Roseburg.