Friday, August 08, 2014

Electronic Health Records Are By No Means Totally Safe And Risk Free - Worrying Real World Experience.

This appeared a little while ago:

Hazards tied to medical records rush

Subsidies given for computerizing, but no reporting required when errors cause harm

By Christopher Rowland

Globe Staff   July 20, 2014
Dr. Larry Gordon and medical assistant Amanda Naimo discussed a procedure at the nurses’ work station at Family Medicine Associates in Hamilton. Some doctors say their data software is confusing to use.
Theresa Robertson’s bedside oxygen alarm sounded at 2:52 a.m., sending an urgent beep through the South Shore Hospital emergency room.
The alarm came as something of a surprise. The 46-year-old visual artist from Weymouth, who suffered from diabetes, congestive heart failure, and other serious health problems, had arrived at the nearby community hospital complaining of shortness of breath the previous afternoon. But by midnight, her vital signs were stable and she was resting peacefully.
Now, in the wee hours of Oct. 11, 2010, a new, more serious problem emerged. Four minutes after nurses responded to the oxygen alarm, Robertson had no pulse. The staff began pushing on her chest, racing to keep her alive.
The cause of Robertson’s crisis, it would later be learned, was a precipitous drop in her blood glucose — severe hypoglycemia. The deadly turn was the result of a medication error, lawyers for her family say, and that error can be traced in part to a major innovation in modern medical practice: electronic health records.
President Obama and Congress poured $30 billion in taxpayer subsidies into the push for digital medical records beginning in 2009, with only a few strings attached and no safety oversight of the vendors who sell the systems.
The move was touted as a way to improve patient care and help rein in medical costs. Five years later, the explosion in the use of the electronic records has created the potential for efficiencies and safety benefits but also new risks for patients, the scope of which still is not fully understood.
Academics, policy gurus, health care leaders, and patient safety advocates generally agree that the American medical system has long needed to abandon its old paper charts and catch up with the rest of the digital age. For instance, wider use of bar-coding in hospitals, matching patient wristbands with the drugs they are given, is creating a safer environment, specialists say.
But the scramble by doctors and hospitals to cash in on the incentives has thrust complex, balky, unwieldy, and error-prone computer systems into highly sensitive clinical settings at a record pace. From 2008 to 2013, the percent of US doctor’s offices with electronic health records rose from 17 to 48 percent. Growth in hospitals was even more dramatic, from 13 to 70 percent.
The scope and pace of change has been far beyond the capacity of medical institutions and government regulators to track, many officials and health care safety advocates agree.
Yet, facing staunch resistance to any regulation by the politically influential health records industry, the Obama administration has opted against mandatory reporting that would enable officials to track unsafe conditions, injuries, and deaths relating to these systems — to draw lessons, that is, from the tragedies of patients such as Theresa Robertson.
In her case, nurses gave her an excessive amount of insulin with three injections, over about two hours, that sent her glucose levels plummeting, her family alleges in a lawsuit.

Multiple orders issued

The nurses were acting in response to multiple orders for insulin that had been entered into two separate prescribing systems used by the hospital — one digital, one involving paper and a fax — and were listed under the names of different doctors, the lawyers said.
“The root cause is having two different systems that weren’t interacting with each other. It creates human error,” said Nicholas Cappiello, a malpractice lawyer at Lubin & Meyer in Boston who represents the family. “Essentially, the right hand didn’t know what the left hand was doing.”
South Shore Hospital said it stands behind the care it provided to Robertson. A spokeswoman for the Weymouth hopsital, said it is not accurate to claim that the use of electronic and paper health records at the same time was the “central issue” in the case, but she refused to discuss any details of the incident, citing requirements for patient confidentiality.
The hospital, pointing to the ongoing litigation, also declined to say what actions it had taken in the wake of Robertson’s death, including whether any doctors or nurses were disciplined.
Numerous adverse and even tragic incidents such as Robertson’s insulin overdose have been traced to errors related to electronic records.
One study of malpractice claims published this year, by safety researchers for a Harvard-affiliated malpractice insurance group, CRICO, produced an unsettling snapshot of the problem. The researchers sifted through a large database of malpractice claims from Boston and other health care centers around the country, looking for cases in which the use or misuse of electronic heath records was suspected of causing harm.
They found 147 instances in which electronic health records contributed to “adverse events” that affected patients — half of them designated as serious. The cases were culled from a one-year period of newly filed malpractice claims overlapping 2012 and 2013, a total pool of around 5,700 cases.
Forty-six of those events resulted in patient death, CRICO officials told the Globe — a toll that before now has not been disclosed publicly.
Faulty use of “hybrid” paper and electronic records (the sort of confusion-inducing circumstance at South Shore Hospital that is alleged by lawyers for Robertson’s family) was among the most frequently cited examples of harm in the review.
The analysts also identified as hazards: medical staff entering incorrect data; improper use of a computer’s “cut-and-paste” function to duplicate, without updating, daily notes on patients; and computer crashes that caused loss of data and left medical staff temporarily without access to critical records.
The study bolstered findings of an earlier review by a medical quality and safety group, the nonprofit ECRI Institute, which ranked electronic data hazards as the No. 1 patient safety concern for 2014. The Pennsylvania-based nonprofit analyzed voluntary reports of 171 events in 36 hospitals across the country — most causing harm to patients but also many “near misses.”
That study period spanned just nine weeks.
At least three of the 171 cases were suspected or confirmed to have resulted in a patient’s death.
Such devastating errors can be triggered not only by computer glitches that generate flawed information but also through human error compounded by poor systems design and sloppy implementation, both supporters and critics of the federal program told the Globe.
The systems can be complex, time-draining, and frustrating to use, especially early on. Hospital staff members routinely override automatic warnings, cut corners, and develop “workarounds” as they struggle to balance caring for patients with the demands of cumbersome drop-down menus and other forms that appear on their computer screens.
There is vastly more here:
This is a mandatory slow read just to discover how many ways things can go wrong and what has led to unreasonable haste to get systems in place that are not as well thought out and safe as we might hope.
The benefits case for eHealth surely is based on essentially zero harm and if this is not then it might be we need to rethink our approach both to incentives and EHR quality and use.
However, as we go forward the simplistic view is seemingly not one that can stand up.
David.

1 comment:

Anonymous said...

The benefits case for eHealth needs to be reduction of harm. Any new system (electronic or paper) will introduce new risks as well as remove some existing risks. See Grahame Grieve's diagram in http://www.healthintersections.com.au/?p=1673 .

Yes, rushing an implementation can and does lead to certain risks that a more measured approach will not be subject to. There are also risks associated with a more measured approach. Minimisation of risks should be the goal.