Electronic connectivity at the mercy of humans Andrew Dalley D Pub Hlth, MB, BS, Adv Dip Bus Mgt, DRACOG CEO, Illawarra Division of General Practice, Hon.

Electronic connectivity at Electronic connectivity at the mercy of humansthe mercy of humans

Andrew DalleyAndrew DalleyD Pub Hlth, MB, BS, Adv Dip Bus Mgt, DRACOGD Pub Hlth, MB, BS, Adv Dip Bus Mgt, DRACOG

CEO, Illawarra Division of General Practice, CEO, Illawarra Division of General Practice,

Hon Principal Fellow, Faculty of Informatics, UoWHon Principal Fellow, Faculty of Informatics, UoW

Euphoric grandiosity of 2008Euphoric grandiosity of 2008

““Doctors should have all of the information Doctors should have all of the information about all of their patients all the time” about all of their patients all the time”

Kaiser’s CEO and Chair, George Halvorson, Kaiser’s CEO and Chair, George Halvorson,

quoted in the Washington Business Journal quoted in the Washington Business Journal

(Rauber, 2008).(Rauber, 2008).

Parochial pragmatism of 1998

“I’m a firm believer that the concept of general practice and computerisation is being held back by finance, underdevelopment and political decisions by government.That doesn’t stop me computerising my practice.”

The late Harold Shipman at the inquest into the death of Kathleen Grundy

1998

I DIANA PRINCESS OF WALES of Kensington Palace London W8 HEREBY REVOKE all former Wills and testamentary dispositions made by me AND DECLARE this to be my last Will which I make this First day Of June One thousand nine hundred and ninety three

1 I APPOINT my mother THE HONOURABLE MRS FRANCES RUTH SHAND KYDD of Callinesh Isle of Seil Oban Scotland and COMMANDER PATRICK DESMOND CHRISTIAN JEREMY JEPHSON of St James's Palace London SW1 to be the Executors and Trustees of this my Will

2 I WISH to be buried

3 SHOULD any child of mine be under age at the date of the death of the survivor of myself and my husband I APPOINT my mother and my brother EARL SPENCER to be the guardians of that child and I

IM - auditing

The truth soon surfaced, when practice receptionist Marjorie Walker stumbled upon some disturbing entries in a druggist’s controlled narcotics ledger. The records showed how Shipman had been prescribing large and frequent amounts of pethidine in the names of several patients.

Ted Ottley

IM - benchmarking

Local undertaker Alan Massey began noticing a strange pattern: not only did Shipman’s patients seem to be dying at an unusually high rate; their dead bodies had a similarity when he called to collect them. “Anybody can die in a chair,” he observed, “But there’s no set pattern, and Dr. Shipman’s always seem to be the same, or very similar. There was never anything in the house that I saw that indicated the person had been ill. It just seems the person, where they were, had died.

The PC: the basis of an EHRThe PC: the basis of an EHR

For 14 out 17 quality For 14 out 17 quality indicators, there was indicators, there was no significant no significant difference in outcomes difference in outcomes between consultations between consultations using an electronic using an electronic record and those not record and those not using one. using one.

(Linder et al, 2007).(Linder et al, 2007).

EHREHR

History History SummarySummary Fragmented detailFragmented detail ComprehensiveComprehensive

Multiple user Multiple user organisationsorganisations

Why use an EHR?Why use an EHR?

I don’t know, I don’t know,

but somebody else may have known, but somebody else may have known,

bits of it,bits of it,

or thought they did, or thought they did,

at some other time, at some other time,

in some other place, and in some other place, and

its worth my while to find out.its worth my while to find out.

Is it worth my while to find out?Is it worth my while to find out?

Relevance* ValidityRelevance* Validity

WorkWorkIM =

Slawson and Shaughnessy, 1994Slawson and Shaughnessy, 1994

Information mastery

Relevance

PACS

Validity

Validity

Cancer surgery databases Omission rate

Upper GIT cancers 27.6%Breast cancer 19.6%Colorectal cancers 32.7%

Clinical data associated with high omission rate

Demographic data associated with low omission rate

Warsi et al 2002 Euro J Surg Oncol 28(8): 850-856

Validity

Orthopaedic database V clinical records62% completeness but96% accuracy.

Researchers’ comments: “Compliance by users was poor.

Completeness of data capture can be improved by providing feedback to users”.

Barrie and Marsh,1992, BMJ 304: 159-162

----- Original Message -----From: <[email protected]>To: <[email protected]>Sent: Wednesday, February 12, 2003 5:35 PMSubject: anony mous; Cas Admit; FRACTURED PELVIS******** MESSAGE FOR DR F AKEID **********> MRN 0X1-4X-XX> Name ANON YMOUS> Address C-MAYFLOWER RETIREMENT VILLAGE, GERRINGONG, 2534> DoB XX/XX/19XX Age: XX Sex: M> Hospital KIAMA HOSPITAL> AMO DR A COLLINS> Pres. Prob FRACTURED PELVIS> Cas Admit 31/01/2003 11:10 MED> Discharged 11/02/2003 05:57> Dis Status Died with no autopsy performed> DocMail No. B45499

Work

Sometimes there are just too Sometimes there are just too many competing interestsmany competing interests

Where does it work OS?

Where does it work?Where does it work?

Kaiser PermanenteKaiser Permanente““Customers”: 8.6 millionCustomers”: 8.6 million

Dollars:Dollars: 4 billion4 billion

($465.11 each)($465.11 each)Washington Business Journal, Washington Business Journal, 6.05.20086.05.2008

Veterans Health Information Systems and Veterans Health Information Systems and Technology Architecture (VistA)Technology Architecture (VistA)

Lesson one

Be prepared to spend money up front for (dubious) downstream benefit.

Why did it work?Why did it work?

Autonomous organisationAutonomous organisationService linked to patient entitlementService linked to patient entitlementAudit quality service deliveryAudit quality service deliveryKnown protocols of care including Known protocols of care including

medicationsmedicationsAutomatic patient enrolmentAutomatic patient enrolment

Where does it not work?Where does it not work?

Trouble in ParadiseTrouble in Paradise

KP in ParadiseKP in Paradise ““Sad story of the failed implementation” of an Sad story of the failed implementation” of an

ehrehr Decision not made with cliniciansDecision not made with clinicians Clinicians not involved in designClinicians not involved in design Clinical productivity declinedClinical productivity declined Poor leadership (“Culture eats strategy for Poor leadership (“Culture eats strategy for

breakfast”)breakfast”)Teasdale, S., Commentary: Trouble in paradise – learning from Hawaii, Teasdale, S., Commentary: Trouble in paradise – learning from Hawaii,

BMJBMJ, Vol 331: 03.12.2005, Vol 331: 03.12.2005

KP in Hawaii

“We had a 12 month preparation period. The product wasn’t delivered for another 14 months. That affects your culture”

The delay … “Lots of things happen in people’s lives. And my internist need to go do other things. And my paediatrician also needed to move. So the implementation date is shifting, and my players are (too)”.

“Several respondents doubted whether they would be able to achieve pre-CIS productivity levels”.

Scott et al, 2005, Kaiser Permanente’s experience of implementing an electronic medical record,

BMJ, Vol 331, p1313 ff

Lesson 2

Clinician engagement at design and implementation phases

Failure to improve clinical productivity results in increased clinician resistance

The ehr should support existing clinical work patterns not require new ones.

Trade-offs are an important element of ehr implementation.

Trouble in ParadiseTrouble in Paradise

Two major solutionsTwo major solutions Clinician championsClinician champions Achieve ownership by cliniciansAchieve ownership by clinicians

Teasdale, S., Commentary: Trouble in paradise – learning from Hawaii, Teasdale, S., Commentary: Trouble in paradise – learning from Hawaii, BMJBMJ, Vol 331: 03.12.2005, Vol 331: 03.12.2005

6 by 4 bygone6 by 4 bygone

Limpopo

Where did it not work?Where did it not work?

Limpopo, SA, 1999Limpopo, SA, 1999

IBM 134m Rand projectIBM 134m Rand project Hospital basedHospital based Poor implementation, poor change Poor implementation, poor change

managementmanagement Culture phaged strategy againCulture phaged strategy again Clinician workload increasedClinician workload increased No improvement in outcomesNo improvement in outcomes

6 by 4 bygone6 by 4 bygone

6 by 4 bygone6 by 4 bygone

Why does it not work?Why does it not work?

Poor design, implementationPoor design, implementationRelevance, validity and workload of Relevance, validity and workload of

clinicianscliniciansCritical mass of information, patients and Critical mass of information, patients and

clinicianscliniciansPatient enrolment (opt in v opt out)Patient enrolment (opt in v opt out)

Are we any better?Are we any better?

Health Connect in Oz

Inherent pilot approach ehealthNT Brisbane (GP Partners) Health Record eXchange Ballarat and Hobart SA Northern Rivers Maitland/Westmead Barwon Health (Geelong) Each state “is progressing with its own independent E-

health program without co-ordination or governance at a national level.” (Booz and Co, 2008, p28)

Medicolegal risk

Bounds of knowledge risk (ie how far do I have to go to find out about this patient)

How do I know there is information about this patient?

Who should access the information?Authorship is irrefutable

Irrefutability

Police Officer: I’ll just remind you of the date of this lady’s death – 11th May ’98. After 3 o’clock that afternoon, you have endorsed the computer with the date of 1st October ’97 which is 10 months prior, ‘chest pains’.

Dr Shipman: I have no recollection of me putting that on the machine.

Officer: It’s your passcode; it’s your nameTed Ottley

http://www.trutv.com/library/crime/serial_killers/notorious/shipman/death_7.html

Increased accountability

“Its almost like they didn’t really care what they wrote on paper, but now its electronic and people can read everything”

Scott et al, 2005, Kaiser Permanente’s experience of implementing an electronic medical record, BMJ, Vol 331, p1315

Barriers

Lack of strategic direction Poor execution of initiatives Insufficient clinician engagement (Coiera, MJA, 2007)

Time overruns for sceptical clinicians Poor acceptance of mooted benefits Benefits generated by one clinician are utilised

by another Clinician acceptance determined by existing

culture of trust and cooperation

Barriers

Patient consent (opt in) High hype to delivery ratio (Booz & Co, 2008)

High rate of burnout from early adopters eg Ballarat

Changes to work patterns Access to computers

Works best for hospital clerical staff (Laerum, Karlsan, Faxvaag,

2004, Use and attitudes to a hospital information system by medical secretaries, nurses and

physicians, BMC Medical informatics and Decision Making; 4:18)

Quick wins approach

ADE ($400M - $2B pa, improved patient care) Prescribing (point to point v centralised v distributed) Limited functionality EHRs eg Diabetes,CVD Need assurance changes will improve existing practice Solution and commercially focused initiatives are the

most successful eg PACS reporting, pathology, prescribing, e-booking

Booz & Co, 2008, E-health: Enabler for Australia’s health reform, National Health and Hospitals Reform Commission (Authored by Christopher Bartlett and Klaus Boehncke in conjunction with Dr Mukesh Haikerwal)