Philip Scott Head of IT Projects & Development Portsmouth Hospitals NHS Trust Co-chair, HL7 UK NHS Implementers group HL7 UK Conference – October 2006.

Philip ScottHead of IT Projects & DevelopmentPortsmouth Hospitals NHS Trust

Co-chair, HL7 UK NHS Implementers group

HL7 UK Conference – October 2006

When is a standard not a standard?The case for unifying LSP approaches to local interfacing

Why discuss this?

There is an opportunity now to mitigate migration/maintenance problems down the line as CRS deployments progress

There is an opportunity to win over clinical enthusiasm for “yet another system”

If an Enterprise Architecture is being adopted, the “Enterprise” must be complete and not have holes or gaps round the edges

Aim: generate discussion not give answers

Topics

Are standards a good idea?But we’ve already got standards!But what about …?Are there workable solutions?Questions

Are standards a good idea?

Yes“Enable clinical semantic interoperability”Jolly good!Well that’s alright thenThe end(nearly…)

But we’ve already got standards!

Clinical information content: Snomed CTElectronic messaging: HL7 v3But – what about now?

What about…

Existing non-HL7 message flows supporting operational needsPMIP and things hanging off it or disguised

inside itDischarge summaries, clinic letters, OOH…

“Legacy” system interoperabilityInformation flows outside MIM scope

1 - NASP Interface – HL7 V3 2 - Existing Systems Interface – HL7 2UK (VA.2) 3 & 4 - Departmental Systems Interface – HL7 V2.3/V2.4

CSC P1R1 Data Centre

CSC P1R1 Interfacing

P1R1 NASP

EBS ETP PDS GPtoGP

NASP MHS

Trust A

Existing System 1 (ES -A1)



Trust Interfacing Engine - A (TIE-A)


Trust B

Existing System 3 (ES -B3)




Trust Interfacing Engine -B (TIE-B)

1b

Data Centre Hosted Departmental System

TheatresMaternity

P1R1 PAS

i.IE

Lorenzo

i.CM i.PM

NASP Interfacing Engine (NASPIE)

Interfacing Engine (EBIE)

Existing System Interfacing Engine (ESIE)

1a

3a

2a

2b

2b

2b

2b

2b

2b

4a

4b

4a

4b

4b

LRS SSB SDS

Maternity IE

Theatres IE

CSC Alliance

Fujitsu

Fujitsu Data Centre

Trust System

Trust System

Trust System

Trust System

Trust System

Trust System

Trust TIEH

L7 2

.3

HL7

2.3

HL7

2.3

HL7

2.3

Hosted TIE

DIE

HL7 V2.3OpenEngine

Millennium

HL7 V2.3

BT

Have market forces simplified this already?

HL7 v2 alone does not offer semantic consistency (to say the least)

When you’ve seen one… [laugh now]“Rampant optionality”Given Benson’s eq.1: Therefore:

)(

1

LSPsni

iiii ncp

ncp

Information flows outside MIM scope


Other issues of scope:Realm (home countries, Europe, world)Service (NHS, social care, police, YOT,

housing, education…)And more immediately for England:

Provider type (NHS, private sector, military)LSP region (cross-cluster flows)LSP sub-units (deployment groups ≠ clinical

networks, initial “external” visibility = 0)


Current approach is pragmatic commercial reality rather than “ruthless standardisation”

CSC: UK vA.2, v2.3, v2.4 BT: UK A.2? Fujitsu: v2.3 Which v2.3? Which v2.4? Which UK A.2? Issues:

Maintenance: no re-use possible, upgrades complex Risk of varying semantics or data quality workarounds Cross-boundary flows: specialty systems, cancer networks,

ISTCs, tertiary referrals, lab to lab

Are there workable solutions?

Use IHE profilesDevelop HL7 v2.5 UK (?)Change to HL7 v3 (UK?)Develop Logical ModelsHL7 UK NHS Implementers subgroup

favoured some sort of logical model to constrain v2

Certifiable testing (≈ sandpit etc)

And what about…?

“Legacy” systems interoperabilityNeeded until children grow up (at least)Eventually true EPR, not for some yearsSpecialty systems will have to co-existRepositories for unsolicited results will

have to co-exist (not to mention EDM)Can we make it easier for clinicians?

Legacy system interoperability

HL7 CCOWFew applications have it out of the boxSome add-on products offer CCOW-like

behaviour/functionalityFront-end integration may be easier in

some cases than a messaging interfaceClinicians will bite your hand off for

patient context synchronization


GOSH is procuring a solution via OJEUPHT is currently procuring via CatalistLSPs apparently take varying stances

(flavours of No)Is CFH interested…?Could be low-hanging fruit!










Conclusion

There is an opportunity NOW to mitigate migration/maintenance problems down the line as CRS deployments progress

There is an opportunity to win over HUGE clinical enthusiasm for “yet another system”

If an Enterprise Architecture is being adopted, the “Enterprise” must be COMPLETE and not have holes or gaps round the edges

Discuss…

Philip Scott Head of IT Projects & Development Portsmouth Hospitals NHS Trust Co-chair, HL7 UK NHS Implementers group HL7 UK Conference – October 2006.

Documents