This EHR "glitch" killed three-year-old Samuel Starr - but patient safety was not compromised

Actually, safety was compromised.

I have written repeatedly that the term "glitch", an often-seen euphemism (usually accompanied with the phrase "but patient care/patient safety was not compromised"), is really a term for life-threatening health IT bugs and malfunctions.  See query link http://hcrenewal.blogspot.com/search/label/glitch for examples of health IT "glitches" (retrieves multiple posts).

Here's a "glitch" that tragically killed this boy:

Three-year-old boy dies after new NHS computer system delays heart treatment

Mar 05, 2014
By James Tute

http://www.mirror.co.uk/news/uk-news/samuel-starr-death-nhs-computer-3209365

Coroner Maria Voisin rules that new booking system meant he did not receive life-saving treatment

A three-year-old boy died in his parents' arms after a new NHS computer system failed to schedule him for a vital heart scan, a coroner has ruled.

Samuel Starr was born with a congenital cardiac defect and needed surgery not long after he was born in 2010.

Although he made a good recovery doctors said he would need regular tests to check on his progress.

But he did not have a scan until 20 months after his first major operation because of a delay following the introduction of a new computer system, Cerner Millennium.

When he finally had the appointment doctors found Samuel needed open heart surgery. During the procedure, Samuel had a stroke.

A month later, after further complications, he died in the arms of his devastated parents, Catherine Holley and Paul Starr, at Bristol Royal Hospital for Children.

There are some in my field who would (and have) called incidents like this "anecdotes" or "bumps in the road."

Yet, as I've recently posted, an internationally-renowned healthcare safety tester, ECRI Institute (http://hcrenewal.blogspot.com/2013/02/peering-underneath-icebergs-water-level.html), and the medical malpractice insurer of the Harvard medical community, CRICO (http://hcrenewal.blogspot.com/2014/02/patient-safety-quality-healthcare.html), have confirmed cases like this are no mere anecdotes.  I estimate from their numbers, and from cases known to me personally and from descriptions of others, that on the order of ten thousand (that's likely low) patients annually are being injured and/or die as a result of bad health IT in the U.S. alone.  Lack of transparency about such incidents impairs knowing the true numbers.

Ms Holley wept as she told the hearing of the moment she kissed her son goodbye, saying: "He was a kind and gentle boy with a great understanding of the world.

"As we read Samuel his favourite stories, he died in our arms."

Avon Coroner Maria Voisin, recording a narrative conclusion, ruled the booking system meant Samuel was not seen and did not receive treatment.

This case somewhat reminds me of the case of a newborn whose x-ray test was done, but never looked at due to a bug in the software that scheduled films to be read, see http://hcrenewal.blogspot.com/2011/06/babys-death-spotlights-safety-risks.html, case #2

Concluding a three-day inquest at Flax Bourton Coroner's Court in North Somerset, Ms Voisin said: "Samuel Starr was born with a complex cardiac disease. "He required surgery and regular check-ups at outpatients to manage his conditions.  "Due to the failure of the hospital outpatients booking system there was a five month delay in Samuel being seen and receiving treatment. "Samuel's heart was disadvantaged and he died following urgent surgery."

Just a "glitch."

Doctors diagnosed that Samuel had a restricted pulmonary artery - causing blood to easily flow away from his heart to his lungs - during a 20-week scan in pregnancy ... On March 3, 2010, Samuel underwent open heart surgery at Bristol Royal Hospital for Children. He was discharged six days later and referred to the Paediatric Cardiac Clinic at the Royal United Hospital (RUH) in Bath for check-ups.

In October 2010, Samuel had his first check-up, in which an echocardiogram, also known as an "echo", was carried out. Ms Holley said she was told the procedure would be next carried out early in the new year.

But it did not happen until April 2011 and an echo was not performed.  Paediatric cardiologist Dr Andrew Tometzki told the inquest he did not "deem it necessary" on that day to carry out the echo scan. Dr Tometzki then ordered a further review - where an echo scan would be carried out - in nine months' time.

Samuel's parents did not receive information about the check-up and raised their concerns with community children's nurse Clare Mees.

Ms Mees, who works for Sirona Lifetime Services, said she would chase up the matter with Dr Tometzki's secretary.

Medical secretary Annabelle Attridge insisted she had taken details and forwarded them on to a dedicated appointments team.

However, the inquest heard "glitch problems" with a new booking system meant the appointment was not logged. 

Apparently not due to human fault:

Paediatric clinic co-ordinator Donna McMahon said while Samuel's medical records had been created on the new Millennium computer programme, no appointments had been transferred across.

Just a "glitch."  Just a "bump in the road."

Ms McMahon said she was first aware of the appointment delay when she received an email from nurse Ms Mees in May 2012.

Samuel did not have the vital check-up until June 21.

Coroner Ms Voisin added: "This overall failure led to a five month delay in Samuel being seen by his cardiologist.

"This meant Samuel had not been seen for five months and had not had an echo for 20 months."

His parents were horrified when Dr Tometzki informed them Samuel required heart surgery that spring.

Just a "glitch."  Just a "bump in the road."

... "Just hours before his operation Samuel was dancing around the ward and telling the nurses all about Spiderman - we had to remind him to quieten down," Ms Holley said.

Samuel's chest was closed on August 8. As he was brought out of sedation on August 9, he suffered a stroke and cardiac arrests.

"It is clear that if the surgery had been performed earlier, the procedure would have been the same," Ms Voisin said.

But she said expert evidence found that as time went on "the more complicated the surgery is, as the more disadvantaged the heart is and the more difficult it is to get over the operation".

Delays in care of fragile patients - in this case a young child, but in many other cases, ED and ICU patients of all ages - is very unwise.

Samuel's condition continued to deteriorate and on September 6 his parents were recommended for treatment to be withdrawn after he had a second cardiac arrest.

The coroner said she would not write a report making recommendations to the hospital, as changes had already been implemented.

Likely, workarounds to bad health IT, inviting future catastrophe.

In my view, many cases such as this never make headlines, as the findings ignore the role of the IT and blame people, cases are settled with gag clauses, patients (who survive) or family are never told of the role of the IT, or lawsuits prove too expensive or too unpleasant to pursue.

But patient safety is never compromised by "glitches", and regulation will just stifle innovation.

My condolences to the parents.  I've suffered a death in my own family as a result of health IT, and I'm a specialist in the field.  Laypeople have no chance.

One should pray the next child affected in this way is not one's own.

-- SS

Mar. 25, 2014 addendum:  

Also see these related posts, the first mentioning "teething problems", another euphemism for potentially-deadly EHR malfunctions:

Cerner's Blitzkrieg on London: Where's the RAF?
 http://hcrenewal.blogspot.com/2010/08/cerners-blitzkrieg-on-london-wheres-raf.html 

North Bristol Hits Appointment Problems: Another "Our Lousy IT Systems Screwed Up, But Patient Safety Was Never Compromised" Story
http://hcrenewal.blogspot.com/2012/01/another-our-crappy-computers-screwed-up.html

-- SS

UK: Another Example of IT Malpractice With Bad Health IT (BHIT) Affecting Thousands of Patients, But, As Always, Patient Care Was "Not Compromised"

At my Dec. 2011 post "IT Malpractice? Yet Another "Glitch" Affecting Thousands of Patients. Of Course, As Always, Patient Care Was "Not Compromised" and others, I noted:

... claims [in stories regarding health IT failure] that "no patients were harmed" ... are both misleading and irrelevant:

Such claims of 'massive EHR outage benevolence' are misleading, in that medical errors due to electronic outages might not appear for days or weeks after the outage ... Claims of 'massive EHR outage benevolence' are also irrelevant in that, even if there was no catastrophe directly coincident with the outage, their was greatly elevated risk. Sooner or later, such outages will maim and kill.

Here is a prime example of why I've opined at my Sept. 2012 post "Good Health IT (GHIT) v. Bad Health IT (BHIT): Paper is Better Than The Latter" that a good or even average paper-based medical record keeping system can facilitate safer and better provision of care than a system based on bad health IT (BHIT).

Try this with paper:

NHS 'cover-up' over lost cancer patient records

Thousands awaiting treatment were kept in the dark for five months when data disappeared

Sanchez Manning
The Independent
Sunday 30 September 2012

Britain's largest NHS trust took five months to tell patients it had mislaid medical records for thousands of people waiting for cancer tests and other urgent treatments. Imperial College Healthcare NHS Trust discovered in January that a serious computer problem and staff mistakes had played havoc with patient waiting lists.

It's quite likely the "serious computer problem" far outweighed the impact of "staff mistakes", as disappearing computer data does so in a "silent" manner.  One does not realize it's missing as there's not generally a trail of evidence that it's gone.

About 2,500 patients were forced to wait longer on the waiting lists than the NHS's targets, and the trust had no idea whether another 3,000 suspected cancer patients on the waiting list had been given potentially life-saving tests. Despite the fact that the trust discovered discrepancies in January and was forced to launch an internal review into the mess, including 74 cases where patients died, it did not tell GPs about the lost records until May.

That is, quite frankly, outrageous if true and (at least in the U.S.) might be considered criminally negligent (failure to use reasonable care to avoid consequences that threaten or harm the safety of the public and that are the foreseeable outcome of acting in a particular manner).

Revelations about the delay prompted a furious response yesterday from GPs, local authorities and patients' groups. Dr Tony Grewal, one of the GPs who had made referrals to Imperial, said doctors should have been told sooner to allow them to trace patients whose records were missing. "The trust should have contacted us as soon as it was recognised that patients with potentially serious illnesses had been failed by a system," he said. "GPs hold the ultimate responsibility for their patient care."

That is axiomatic.

The chief executive of the Patients Association, Katherine Murphy, added: "This is unacceptable for any patient who has had any investigation, but especially patients awaiting cancer results, where every day counts. The trust has a duty to contact GPs who referred the patients. It's unfair on the patients to have this stress and worry, and the trust should not have tried to hide the fact that they had lost these records. They should have let the GPs know at the outset."

Unfair to the patients is an understatement,  However, if one's attitude is that computers have more rights than patients, as many on the health IT sector seem to with their ignoring of patient rights such as informed consent, lack of safety regulation, and lack of accountability, then it's quite acceptable.

The trust defended the delay in alerting GPs, arguing that it needed to check accurately how much data it had lost before making the matter public. It said a clinical review had now concluded that no one died as a result of patients waiting longer for tests or care.

That would be perhaps OK if the subjects whose "data had been lost" through IT malpractice were lab rats.

Despite this, three London councils – Westminster, Kensington and Chelsea, and Hammersmith and Fulham – are deeply critical of the way the trust handled the data loss. Sarah Richardson, a Westminster councillor who heads the council's health scrutiny committee, said that trust bosses had attempted to "cover up" the extent of the debacle. "Yes, they've done what they can but, in doing so, [they] put the reputation of the trust first," she said. "Rather than share it with the GPs, patients and us, they thought how can we manage this information internally. They chose to consider their reputation over patient care."

As at my Oct. 2011 post "Cybernetik Über Alles: Computers Have More Rights Than Patients?", to be more specific, they may have put the reputation of the Trust's computers first. 

Last week, it was revealed that Imperial has been fined £1m by NHS North West London for the failures that led to patient data going missing. On Wednesday, an external review into the lost records said a "serious management failure" was to blame for the blunder.

Management of what, one might ask?

Imperial's chief financial officer, Bill Shields, admitted at a meeting with the councils that the letter could have been produced more quickly. He said that, at the time, the trust had operated with "antiquated computer systems" and had a "light-touch regime" on elective waiting times.

Version 2.0A will, as again is a typical refrain, fix all the problems.

Terry Hanafin, the leading management consultant who wrote the report, said the data problems went back to 2008 and had built up over almost four years until mid-2011. Mr Hanafin said the priorities of senior managers at that time were the casualty department and finance.

Clinical computing is not business computing, I state for the thousandth time.  When medical data is discovered "lost", the only response should be ... find it, or inform patients and clinicians - immediately.

He further concluded that while the delays in care turned out to be non-life threatening, they had the potential to cause pain, distress and, in the case of cancer patients, "more serious consequences" ... The trust said it had found no evidence of clinical harm and stressed that new systems have now been implemented to record patient data. It denied trying to cover up its mistakes or put its reputation before concerns for patients. "Patient safety is always our top priority," said a spokesman.

"More serious consequences" is a euphemism for horrible metastatic cancer and death, I might add.  The leaders simply cannot claim they "found no evidence of clinical harm" regarding delays in cancer diagnosis and treatment until time has passed, and followup studies performed on this group of patients.

This refrain is evidence these folks are either lying, CYA-style, or have no understanding of clinical medicine whatsoever - in which case their responsibilities over the clinic need to be ended in my opinion.

I, for one, would like to know the exact nature of the "computer problem", who was responsible, and if it was a software bug, how such software was validated and how it got into production.

-- SS

Oct. 1, 2012 Addendum:

What was behind the problems, according to another source?   

Bad Health IT (BHIT):

Poor IT behind Imperial cancer problems
e-Health Insider
28 September 2012
Rebecca Todd

An independent review of data quality issues affecting cancer patient referrals to Imperial College Healthcare NHS Trust has identified “poor computer systems” as a key cause of the problem.

The review’s report highlights the trust’s use of up to 17 different IT systems as causing problems for patient tracking.

However, it says the trust should be aware of the risks of [replacing the BHIT and] moving to a single system, Cerner Millennium, because of reported problems in providing performance data after similar moves at other London trusts.

In January 2012, the report says the NHS Intensive Support Team was reviewing the way reports on cancer waiting times were created from Imperial’s cancer IT system, Excelicare.

The team discovered that almost 3,000 patients were still on open pathways who should have been seen within two weeks. In May, letters were sent to GPs to try and ascertain the clinical status of around 1,000 patients.

BHIT must be forbidden from real-world deployments, and fixed rapidly or dismantled (as Imperial College Healthcare NHS Trust appears to be doing), although the "solution" might be just as bad, or worse, than the disease.

-- SS