Thursday, September 24, 2015

Cambridge University Hospitals Trust IT Failures: An Open Letter to Queen Elizabeth II on Repeated EHR Failures, Even After £12.7bn Wasted in Failed NHS National IT Programme

Dear Queen Elizabeth,

I am an American citizen who has written for years about healthcare information technology mismanagement (IT malpractice), dangers to patients of this technology when faulty in healthcare, and the huge mania or bubble that has surrounded this technology in a layer of fairy tales that has cost your Kingdom's treasury, as well as that of the U.S., dearly.

Your subjects seem unable to learn from their mistakes, or learn even from free material at sites such as this, or at my academic site at Drexel University at http://cci.drexel.edu/faculty/ssilverstein/cases/.

Instead of being appropriately skeptical, they spend your citizen's money extravagantly and with abandon on grossly faulty computing.  This results in serious health care meltdowns such as I observed at my September 22, 2011 post on your now-defunct National Programme for IT in the National Health Service (NPfIT).  That post was entitled "NPfIT Programme goes 'PfffT'" and is at http://hcrenewal.blogspot.com/2011/09/npfit-programme-going-pffft.html.

In that post I observed:

... [NPfIT] also failed because of collective ignorance of these domains [e.g., healthcare informatics, social informatics, etc. - ed.] among its leaders, and among those who chose the leaders. For instance, as I wrote here:


The Department of Health has announced the two long-awaited senior management appointments for the National Programme for IT ... The Department announced in February that it was recruiting the two positions as part of a revised governance structure for handling informatics in the Department of Health.

Christine Connelly will be the first Chief Information Officer for Health and will focus on developing and delivering the Department's overall information strategy and integrating leadership across the NHS and associated bodies including NHS Connecting for Health and the NHS Information Centre for Health and Social Care.
Christine Connelly was previously Chief Information Officer at Cadbury Schweppes with direct control of all IT operations and projects. She also spent over 20 years at BP where her roles included Chief of Staff for Gas, Power and Renewables, and Head of IT for both the upstream and downstream business.

Martin Bellamy will be the Director of Programme and System Delivery. He will lead NHS Connecting for Health and focus on enhancing partnerships with and within the NHS. Martin Bellamy has worked for the Department for Work and Pensions since 2003. His main role has been as CIO of the Pension Service.

Excuse me. Cadbury Schweppes (candy and drink?) The Pension Service? As national leaders for healthcare IT?

Also see my August 2010 post "Cerner's Blitzkrieg on London: Where's the RAF?" at http://hcrenewal.blogspot.com/2010/08/cerners-blitzkrieg-on-london-wheres-raf.html.

It's clear medical leaders in the UK learned little from the £12.7bn NPfIT debacle.  Now we have this:

Addenbrooke's Hospital consultants concerned over online records
BBC News
31 July 2015
http://www.bbc.com/news/uk-england-cambridgeshire-30393575

A £200m online patient-record system has been "fraught with problems" and medics' concerns "seemingly overlooked", senior hospital consultants have claimed.

A letter seen by the BBC reveals management at Addenbrooke's and Rosie hospitals in Cambridge were told of "serious" issues last month.  It came after the hospitals transferred 2.1 million records in October.

The trust said "unanticipated" issues led to "more than teething problems". 

The hospital is the first in the UK to use Epic's eHospital system, which is used in hospitals in the US.

To the CEO, these problems are just "hiccups":

... Chief executive Dr Keith McNeil admitted there had been "more than teething problems" and "some of it was anticipated and some of it was unanticipated". The "unanticipated" problems included problems with blood tests and "one of the busiest periods in the hospital's history", he said. He added: "We're profoundly sorry about that... people will understand that you can't do an information technology implementation of this size without some hiccups.

"Hiccups" are a euphemism for incompetence in system design, implementation and testing before it is used on live patients, Your Majesty.  I also note that a close relative of mine, and numerous other patients I know of are severely injured or dead due to these "hiccups."  

And now this:

Addenbrooke's and Rosie hospitals' patients 'put at risk'
BBC News
22 September 2015
http://www.bbc.com/news/uk-england-cambridgeshire-34317265

One of the UK's biggest NHS trusts has been placed in special measures after inspectors found it was "inadequate".

Cambridge University Hospitals Trust, which runs Addenbrooke's and the Rosie Birth Centre, was inspected by the Care Quality Commission in April and May.

Inspectors expressed concerns about staffing levels, delays in outpatient treatment and governance failings.

... Prof Sir Mike Richards, the Care Quality Commission's (CQC) chief inspector of hospitals, said while hospital staff were "extremely caring and extremely skilled", senior management had "lost their grip on some of the basics".

"[Patients] are being put at risk," he said. "It is not that we necessarily saw actual unsafe practice but we did see they would be put at risk if you don't, for example, have sufficient numbers of midwives for women in labour."

The trust, which is said to be predicting a £64m deficit this year, has apologised to patients.

I note that these hospitals had been the beta site for the first implementation of U.S. EHR maker EPIC company's product of the same name.  That £64m deficit looks a bit suspicious for IT overspend; for example see this U.S. hospital's experience of going in the red over fixing 10,000 "issues" (problems) with EPIC, in my post of June 2, 2014:  "In Fixing Those 9,553 EHR "Issues", Southern Arizona’s Largest Health Network is $28.5 Million In The Red" at http://hcrenewal.blogspot.com/2014/06/in-fixing-those-9553-ehr-issues.html.

... Perhaps the most worrying aspect of the Addenbrooke's story is not that such a world-renowned hospital has ended up in a predicament like this, but rather that it happened so quickly.

A year ago the trust which runs the hospital - Cambridge University Hospitals NHS Foundation Trust - wasn't even on the Care Quality Commission's radar in terms of being a failing centre.

I suggest a deep connection between this rapid fall, and the rapid rise of an EHR - an antiquated term for what is now an enterprise command-and-control system for hospitals.

... In fact, two years ago - as the regulator was embarking on its new inspection regime - it was among the band of hospitals considered to be the safest, according to the risk-rating system at the time.

But now a hospital which can boast to being a centre of excellence for major trauma, transplants, cancer, neurosurgery, genetics and paediatrics, has been judged to be a basket case and will join the 12 other failing hospitals already placed in special measures.

In my view, a major disruptive technology such as a new EHR is the Number One suspect in such a fall.

... Certainly it seems to have made mistakes - as the troubles with its £200m computerised patient records programme illustrates - but it's hard to escape the feeling that this is just the tip of the iceberg.

The "troubles with its £200m computerised patient records programme" is likely the iceberg, not just its tip.

The Care Quality Commission ("The independent regulator of health and social care in England", http://www.cqc.org.uk/) investigated these hospitals and issued a report, located at http://www.cqc.org.uk/location/RGT01/reports.

Among their key findings were:

Introducing the new EPIC IT system for clinical records had affected the trust’s ability to report, highlight and take action on data collected on the system. 

Excuse me?   Spend £200m on a computer system, and the result is impaired ability to report, highlight and take action on data collected?  Something is very wrong here.

 ... Although it was beginning to be embedded into practice, it was still having an impact on patient care and relationships with external professionals.

Clearly, the CQC does not mean a positive impact.

... Medicines were not always prescribed correctly due to limitations of EPIC, although we were assured this was being remedied.

Spend £200m on a computer system and the result is medicine prescription impairment (with the risks to patients that entails)?  Excuse me?

If those "limitations" affect these British hospitals, what "limitations" on getting prescriptions correct exist in all the U.S.-based hospitals that use this EHR, I ask?

... There was a significant shortfall of staff in a number of areas, including critical care services and those caring for unwell patients. This often resulted in staff being moved from one area of a service to another to make up staff numbers. Although gaps left by staff moving were back-filled with bank or agency staff, this meant that services often had staff with an inappropriate skills mix and patients were being cared for by staff without training relating to their health needs.

I suspect many staff were so unhappy with the EHR that they left, and recommended others not come.

Despite this patients received excellent care.

Odd how patient care and safety is never affected by bad health IT, as in the myriad stories at this site under the indexing key "patient care has not been compromised" (http://hcrenewal.blogspot.com/search/label/Patient%20care%20has%20not%20been%20compromised).

... Clinical staff were not always able to access the information they required – for example, diagnostic tests such as electrocardiographs (ECGs) to assess and provide care for patients. This was because ECGs had to be sent to a central scanning service to be scanned into the electronic recording system [a.k.a. EHR] once the patient had been discharged. This meant their ECGs would not be available for comparison purposes if a patient was re-admitted soon after discharge.

Very, very bad IT planning, potentially putting unstable patients at risk.  Cybernetic miracles always have "fine print" that needs be read by skeptical managers BEFORE implementation.

Where agency staff were used, they were not always able to access information about patients they were supporting. 

 Ditto.

... Some staff told us there were no care plans on the new IT system.  Some staff told us the doctors’ orders had replaced care plans on the new EPIC IT system. These orders were task-orientated and did not always reflect the holistic needs of the patients.

This defective arrangement sounds like it was designed by non-clinicians.   The hubris and arrogance of non-clinicians sticking their heads into clinical issues - especially those of an IT-management background - must be witnessed to be fully comprehended.  It is my belief that such individuals should be subject to the liability as are the clinicians whose work increasingly depends on these IT systems.   If you dare to stick your neck into clinical affairs regarding systems upon which clinicians depend, you should be subject to the same liabilities as a clinician.  Unfortunately, this rarely if ever occurs.

 ... Whilst there were up-to-date evidence-based guidelines in place, we were concerned that these were not always being followed in maternity. This included FHR monitoring, VTE and early warning score guidelines. Staff were competent and understood the guidelines they were required to follow, however, lack of staffing and familiarity with the computer system (EPIC) made this difficult.

The point being missed here is that paper records required no massive multi-hundred page training manual in order to to perform basic functions such as the above.  The complexity of EHRs is costly, unnecessary, impairs clinicians and the solution is a massive scale back and simplification of these systems' complexity and scope.  Unfortunately, that, too is unlike to happen until the negative impacts become increasingly visible and intolerable - a meltdown I predict will occur, eventually.

... Since the introduction of EPIC, outcomes of people’s care and treatment was not robustly collected or monitored. For example, there was no maternity dashboard available since December 2014.

Again, spend £200m and have this result?  Something is seriously wrong here.  I suspect it is that personnel no longer had the time to perform monitoring, as they were likely distracted and struggling to keep afloat with more fundamental medical issues (like keeping major mishaps from occurring) using a complex and buggy EHR system.

That theory is likely confirmed by the following:

... At unit level we observed examples of excellent leadership principles; however, leadership of the directorate overall required improvement. This was because senior managers had not responded appropriately or in a timely way to known and serious safety risks, there was a general lack of service planning, and because key performance data was not being collected robustly and therefore not being analysed. We recognised that EPIC was the root cause of the problems with data collection, and that prior to its introduction in October 2014 many of the data collection issues were not apparent, however, improving this issue was not seen as a priority.

Management, I suspect, became complacent due to their infatuation with cybernetics and a belief that with a big-name EHR in place, operational ills were accounted for and they could relax.  (I've written of this phenomenon as the "syndrome of inappropriate overconfidence in computing.")  Management complacency, bad health IT and struggling clinicians is a very, very bad combination.

... Staff understood their responsibilities for safeguarding children, and acted to protect them from the risk of avoidable harm or abuse. There were enough medical staff but there were nursing shortages in some areas, such as in the day unit and in the neonatal unit. The new ‘EPIC’ (a records management system) computer system added to pressures on staff but effective temporary solutions helped to protect patients.

In other words, workarounds were used to get around the work-impeding EHR.  Workarounds introduce yet more risk.

... the electronic records system (EPIC) created significant numbers of delayed discharges that impacted on patients receiving end-of-life care.  ... Many staff said they had struggled with EPIC and it was time consuming. The specialist palliative care team found patients dropped off the system, so kept two lists to avoid losing patients.

One does not struggle with paper records.  (My current colleagues tell me the EHR struggle is non-ending.)  I further note that a computer system's rights, it appears, took precedence over patients' dying with dignity.

... While introducing EPIC, processes to deal with remaining paper records were unclear. For example, staff documented follow-up appointment requests on notepads. Paper records which were not stored in EPIC were inconsistently stored within the outpatients department. Inaccurate discharge summaries led to a risk that patients would not receive appropriate follow up care.

A fetish to totally eliminate paper, even where paper is the best medium for a purpose (e.g., as here:  http://cci.drexel.edu/faculty/ssilverstein/cases/?loc=cases&sloc=Cardiology%20story), creates major chaos and increases risk.

In conclusion, Your Highness, it might benefit your citizens (and those of the U.S.) if a national re-education programme were instituted to de-condition your leaders from unfettered belief in cybernetic miracles in medicine, a mental state they attain in large part due to mass EHR vendor and pundit propaganda.

A more sober mindset is recommended by your subject Shaun Goldfinch in "Pessimism, Computer Failure, and Information Systems Development in the Public Sector" (Public Administration Review 67;5:917-929, Sept/Oct. 2007, then at the University of Otago, New Zealand): 

The majority of information systems developments are unsuccessful. The larger the development, the more likely it will be unsuccessful. Despite the persistence of this problem for decades and the expenditure of vast sums of money, computer failure has received surprisingly little attention in the public administration literature. This article outlines the problems of enthusiasm and the problems of control, as well as the overwhelming complexity, that make the failure of large developments almost inevitable. Rather than the positive view found in much of the public administration literature, the author suggests a pessimism when it comes to information systems development. Aims for information technology should be modest ones, and in many cases, the risks, uncertainties, and probability of failure mean that new investments in technology are not justified. The author argues for a public official as a recalcitrant, suspicious, and skeptical adopter of IT.

Such a mindset would be helpful in preventing massive wastes of healthcare Pounds, Euros and Dollars better spent on patient care than on cybernetic pipe dreams.

Sincerely,

S. Silverstein, MD
Drexel University
Philadelphia, PA

------------------

Addendum:

I would like to hear from those in the know if my suspicions are correct.  Please leave comments.

-- SS

    8 comments:

    Anonymous said...

    Bravo

    Matthew Holt said...

    I'm sure the Queen is a regular HCRenewal reader Scott. And if she it's she will be after this!

    Anonymous said...

    The suffering of doctors, nurses, and other staff at Addenbrooke's after EPR deployment is widespread in the USX and UK.

    But, the vendors have a way of quashing any public comment, and use the hospital administration to punish whistleblowers.

    Even at Addenbrooke's, the likely backbreaking culprit, Epic EPR, is not getting appropriate scrutiny.

    Anonymous said...

    @matthew_holt

    At least, the @queen does not need to log in w twitter to issue a comment. Btw, your blog has been diminished by your sign in policy.

    I am not even a queen and I no longer rad your blog, cause the comments are vanilla.

    InformaticsMD said...

    Matthew Holt said...

    I'm sure the Queen is a regular HCRenewal reader Scott. And if she it's she will be after this!


    Although my title is deliberately hyperbolic, one never knows.

    The patients and clinicians of Cambridge Trust deserve far better than being unwilling test subjects of health IT experiments.

    -- SS

    InformaticsMD said...

    I am not even a queen and I no longer rad your blog, cause the comments are vanilla.

    I am not sure who wrote this, but I can assure you we at HC Renewal are not mourning the loss.

    Only an intellectually limited individual would boycott a blog based on the comments being "vanilla" - ignoring the actual content of the blog.

    It's like ignoring Harrison's Textbook of Internal Medicine because one does not like the font of the Acknowledgements page.

    -- SS

    Anonymous said...

    That vanilla comment was addressed to @matthew_holt 's the health care blog requiring tweets and facebooks to comment. This blog is a frontrunner.

    Anonymous said...

    Holt circa 2015: "I'm sure the Queen is a regular HCRenewal reader Scott. And if she it's she will be after this!

    Does the Queen use Facebook? Does the Queen tweet?

    If not, whe will never be able to comment on Holt's blog.

    I agreeeeeeee with anonymous that the Holt blog has evolved to bland, at best. Do not take this personally, Matthew.