Deadline for ICU credentialing approaches
Marsh survey - CEO conducts Gallup poll, dialogue with employees
Remembrance and aid offered to Tsunami victims
'State of the School' addressed
PAWS isn't only for patients ...
Getting wired: EMR
Bruce T. ADORNATO
We have heard a lot about medical "errors" and preventable deaths in the past few years. There is a revolution occurring in our midst.
A famous and often quoted report from the Institute of Medicine (IOM) in 1999 said between 44,000 and 98,000 preventable deaths in the U.S. occur each year due to a variety of preventable actions and inactions. (If you want to think about how people in other professions might view this, let's look at another profession where errors count: A pilot attempting to take off without lowering the flaps, has made a preventable error.)
Although we can argue about the methodologies of the 1999 study and criticize the numbers, we all agree there is room for improvement.
The follow-up 2003 IOM report does show improvement in numbers and parallels the suggestions found in a survey of physicians sponsored by the Kaiser Family Foundation in Menlo Park. This follow-up was published in the New England Journal of Medicine in 2002. The physicians suggested improvement in overall hospital systems, increases in nurse-to-patient ratios, limiting high risk procedures to centers of excellence (where a high volume of these procedures are performed), limiting ICU care to intensivists, and using physician order entry and electronic medical records (EMR).
We are in the process of seeing those recommendations put into practice at Stanford and nationally. The Jan. 17 issue of the Wall Street Journal published a piece on "office technology" which was actually a detailed report on the process and effects of the University of Pittsburgh Medical Center's (UPMC) efforts at technological transformation. If you think we have a complex medical system, consider UPMC with 19 hospitals, 400 outpatient clinics and doctors offices, and more than a dozen retirement homes and skilled nursing facilities. UPMC has spent more than $500 million over five years but has converted only 500 of its 2,000 employed physicians to EMR. (Some sites are more converted than others). The figures don't even count another affiliated 2,000 outside physicians who need to be converted to EMR.
One of the most difficult implementation problems is getting the patchwork of computer systems from radiology, from pharmacy, from EMR, from POE, from accounting, etc. to recognize and communicate with each other. This is an IT and programmer nightmare.
Yet despite the enormity and difficulty of the project the rewards so far have been measurable.
The article looked at pre- and post-POE data at Children's Hospital of Pittsburgh: Medication errors dropped 75 percent from 4 to only 1 per 100,000 doses; errors in doctor note transcription fell from 5 to 0 per month; physician satisfaction with radiology turnaround improved from 47 percent to 68 percent; physician satisfaction with lab turnaround rose from 64 to 81 percent; and verbal orders signing within 24 hours increased from 79 to 97 percent.
At Stanford, we are well into the implementation of a similar process. In the past few years, we have instituted physician order entry, and the process is getting better with each iteration. We now have a system of procedural safety for patients that factors in physician consent. We have solid procedures for operating room "time-outs," blood product safety procedures, and a pilot program for computerized referrals. We also have off-campus laboratory and radiology reporting. The digital radiology library is a wonder in itself, allowing archiving of each patient's x-ray and scan file. Stanford has required that physicians have credentials or equivalent experience in intensive care to treat in the ICU [see related story this issue]. Nurse ratios are going up and we are hiring more RN's.
One area of computerization that is troublesome to me as a consultant is the EMR and the use of "structured notes." As a consultant often called in at some time after the "event" (read stroke, seizure, coma, altered level of consciousness, weak leg, numb hand, dysarthric speech, visual loss, aphasia, etc.) I depend on the recorded history - especially the physical examination. Too often, it seems, we already abbreviate our description of the physical findings with arcane abbreviations (PERRLA, A&O X4, etc), vague and ambiguous generalities (neuro exam physiologic, in no acute distress, etc.) and hieroglyphic pictorials (stickmen for deep tendon reflexes). Clearly it would be more useful to write a simple sentence describing whether or not the patient can walk, get dressed, and use a knife and fork. I know the days of the quill pen and cursive are over but there are limits to our data reduction.
The other trap is the overuse or inappropriate use of boilerplate and templates. Driven by the need for speed and documentation (one of the unintended consequences of Medicare legislation), templates are becoming standard fare. Need a normal neuro exam? Just hit F1. Need a normal rheumatology exam for 74 joint range of motions? Hit F2. Need a routine high school physical? Hit F3. It would be OK if we really knew that the printed line always reflected reality. But we don't. And something is being lost in patient care and in the creativity of the physician observer and examiner.
Fortunately, there is hope for the process of EMR. We look forward to voice recognition software eventually allowing us to quickly record what we see and think with an instant copy available to our consultants, our fellow caregivers and to our payors. There are already versions used by radiology at some institutions and used by some ER physicians elsewhere, but these prototypes lack the versatility to wax eloquently. Hopefully, we won't have to wait much longer.
On another topic, please be sure to fill out the questionnaire that is included with this issue of the Medical Staff Update. As physicians, it's our opportunity to make ourselves heard on issues that affect us.
