Diagnostic error is a major contributor to malpractice claims in the U.S. and the
Harvard medical system, where more than a quarter of the claims closed in 2006 involved
diagnosis-related issues. These represented more than half of the indemnity payments.
Further, the portion of diagnosis-related claims that were categorized as high-severity
went up 14 percent in the five-year period ending in 2007.
At a conference focused on diagnostic errors, sponsored by CRICO/RMF in June 2007
in Cambridge, one panel looked at practical issues and solutions. Panelist Carol
Murray is Senior Risk Management Specialist at OHIC Insurance Company, a professional
liability company in Columbus, Ohio. Murray’s focus was on the outpatient setting.
“As I look at solutions, we have primarily single practitioners, small to moderately
sized groups. They have a lot of different manual processes in terms in terms of
list logs, pending files, rescheduling patients for follow up, assigning a particular
staff member to work with the physician to be the person that oversees and using
things like requisitions and also getting their patients involved. So I see those
in different combinations.”
Murray described a malpractice case that involved a patient who presented to a small
group practice with chest pain after a fall. This practice did its own x-rays and
this man’s film was read as normal in the office. He was sent home with instructions
to return for follow up.
The system in that practice involved sending x-rays from the office to be overread
by the hospital radiologist, and picked up when someone goes over to drop off new
films to be overread, so they are returned with the radiologist report before a
follow-up visit.
That didn’t happen in this case. When the patient returned, he was seen by another
physician in the group who related his chest problems to his fall. Later the patient
presented at a hospital in another community with an empyema. He required surgery,
spent six months in the ICU, and the result was a substantial liability claim.
Murray said electronic medical record systems can both help and cause problems.
For the majority of practices without electronic records some basic processes can
help to prevent similar cases from falling through the cracks.
“In terms of the ones that I think really work well are the ones that have a person
assigned to be watching to make sure that the tests all come back and are brought
to the doctor’s attention. Now that manpower sometimes is not available in a really
small practice. The other thing that I think is helpful—and this primarily seems
to work in specialty practices—is rescheduling the patient at the time they are
first there to come back to review the results of the tests, which seems to be a
very helpful thing. And certainly patient involvement…. Some of the systems that
you probably have heard about are voice mailbox systems where doctors can leave
information for patients to pick up findings. That’s a good check. It also has from
a liability point of view usually a record of the communiqué so that later if there
were a lawsuit that they would be able to prove that the information was accessed
by the patient. But probably the overriding thing that I think would really be most
helpful would be to educate practices to use tools like RCA and FMEAs that are very
prevalent in hospitals so that they actually can analyze what is the weakness in
their system when they have a problem or a near miss situation and really make it
better. I think that they are not necessarily equipped with those analytical tools
and the way to apply them, and I think that’s something that we will be talking
to our physicians.”
Dr. John Kelly, is Chief of Staff at Abington Memorial Hospital in Pennsylvania.
He talked about a number of key policies that reduce risk at his hospital. Under
a rapid response protocol, nurses are empowered to communicate worrisome status
changes early up the chain of command.
“As we have all recognized the value of the rapid response teams and medical emergency
teams, I think we’ve recognized that the respiratory rate is a very cheap early
warning system for serious illness. That kind of vigilance that a nurse would take
a respiratory rate and note it to be 30 and say ‘this isn’t right, let me call somebody;
let me call the intern; if I don’t get the intern, let me call the resident; if
I don’t get the resident, I think I’m going to call Dr. Kelly at home.’ And they
have the right to do that at my hospital.”
Another technique—diagnostic checklists prior to discharge—was employed after some
difficult cases.
“We had a couple of bad cases where patients were sent home from our emergency room
with sort of very nonspecific complaints. But as do most people in the emergency
room, they had CBCs done, and theses CBCs showed varying total white counts from
normal to 20,000, but they all had significant bandemia in the range of 15 to 40
percent. Because it didn’t fit, it was discarded or because the emergency room doctor
sort of forgot it in the hubbub of the day when they are seeing between 200-300
patients in the emergency room per day, that was discounted. We sent patients home
who had very early presentations of entities such as necrotizing fasciitis and returned
gravely if not fatally ill. So we have insisted on checklists in the emergency room
that any unexplained physiological or laboratory abnormality has to be explained
on this checklist. The emergency room doctors resented it. They felt we were prescribing
to them, if you will. I’m pleased to report, though it has only been a couple of
years, we haven’t had anybody else going home with an unexplained bandemia. They
usually are admitted to the hospital recognizing that there is usually a serious
infection in the background.”
In the end, Dr. Kelly said patient safety must be driven throughout the institution
and not concentrated in one program or unit or department. He added that the development
of expertise in patient safety must be encouraged from the top.
According to Dr. Kelly, his board of trustees backs a rigorous training in patient
safety, sending numerous physicians in multiple disciplines to national conferences
and symposia, to help spread ideas throughout his system that reduce diagnosis-related
and other errors at every level of his institution.
