CfH to fund NHS clinician trips to India and USA

Tags: A CfH England Fujitsu GP GPs Information iS iSoft London Lorenzo LSP Millennium South

10 Aug 2005

NHS Connecting for Health is urgently looking for experienced NHS clinicians, interested in ensuring that NHS IT systems are fit for purpose, to be flown out and work on assignment with clinical software developers in India or the United States.

Both of the brand new clinical systems that Connecting for Health has contracted to be delivered, Lorenzo from iSOFT and Carecast from IDX, are being written from scratch overseas by software developers with little or no immediate experience of the English NHS.

The hope is that the inevitable problems that arise in remotely developing these systems to meet the particular requirements of the NHS can be more quickly overcome if NHS clinicians are on hand to advise. CfH says that it plans to fund a permanent presence of NHS clinicians at the two software development sites on a rotating basis.

In a note last week to staff involved in the project Dr Gillian Braunold, CfH clinical lead, says: "There is an immediate need by NHS Connecting for Health to find experienced NHS clinicians (open to all Clinicians, not just Doctors) to travel to Hyderabad in India and Seattle in the USA for several weeks."

Dr Braunold told EHI that she did not have a number of clinicians in mind, but it would be "less than 100". Only a few clinicians would be taken over at a time, she added, and the requirement was mostly for clinicians "other than GPs".

She stressed that it was a requirement in the contracts that suppliers "make sure their product is fit for purpose". Dr Braunold added that the contracts also say "CfH will supply NHS experience. It’s a contractual requirement".

She said the note, seen by EHI, was sent out to 600 clinical staff who had expressed an interest in becoming involved in the project. In it Dr Braunold explains that the intention is that clinicians will be able to advise software developers in Seattle, home of IDX; and Hyderabad, home of iSOFT's development team, "on the workings of the NHS and to resolve misunderstandings if they arise".

Additional guidance provided says that the aim is to shorten the time taken for development. "Such a presence will enable the Authority [CfH] to provide NHS knowledge and understanding and help inform the software development and defect resolution and speed up information flows back to CfH."

Initial commitments of two weeks, followed by another two weeks within a few months are being sought. The CfH guidance says that it is expected that clinicians will work with the software provider their local service provider has contracted with, though individual preferences will be catered for.

There is no mention of volunteers being sought from the South of England to guide software development in Kansas City, home of Cerner, which is expected to provide its Millennium software to LSP Fujitsu for the South of England.

Dr Braunold told EHI that she had recently returned from a visit to iSOFT's development centre in Chennai, India. Asked how the GP component of iSOFT's Lorenzo product was shaping up she said: "It’s a big product that we've got to get right".

Pressed to expand on how close Lorenzo was to being ready, she added: "I've seen a product in development that has a lot of promise, but it's in development and lots of work will be needed before its ready for GPs desktops."

The CfH guidance says that one of the jobs the volunteers will perform, will be "to gain an understanding of progress in each of the locations and during their assignment provide CfH with an objective assessment of progress, problems and issues".

Problems with remotely trying to develop or rewrite clinical software to meet the needs of the NHS have often been experienced in previous NHS IT projects. Anglicising IDX's Carecast system from Seattle was identified by University College London Hospitals NHS Foundation Trust (UCLH) as contributing to delays in its deployment of the system. UCLH successfully went live with the first phase of Carecast in June.

Readers Comments

Life cycle

10 Aug 05 11:25

don't the companies in question have their own analysts, architects, designers and testers that have a full understanding of the requirements? At this rate the NHS might as well have hired 500 hundred of their own developers directly and got on with the job. It just doesnt seem right.

Lifecycle

10 Aug 05 13:22

This isn't looking for developers but clinical staff to help the software engineers better develop the solution. The article makes this clear. No matter how detailed a design there's always questions about the actual implementation.

Another example of CFH engaging with clinicians is as test witnesses to verify that the LSP tests are a fair and accurate representation of how the system will be used. The clinicians get to help raise software quality, participate in the process and get a better idea of what they will be receiving in the software.

I don't think it's fair to criticise CFH for lack of clinical engagement and then when there's a real sign of it happening disparage that work as well.

Life cycle

bill.douglas@whis.nhs.uk

10 Aug 05 14:13

True - can't knock the principle of having NHS clinicians vetting the development work at this stage. Still see the point of the first poster above, though. Shouldn't the LSPs' development teams have been advised "on the workings of the NHS" long before now? It's the initial overseas trip angle to this which grabs the attention, although I'm sure the reality will be serious work for those who take part. Hope that those selected to travel will themselves be vetted too. Important to get the right people involved.

Testing of Software

rk@rkhanna.co.uk

10 Aug 05 15:06

I, similar to Bill completely agree that the Clinicians should be involved in identifying the needs of a clinical software system. I am not sure what we mean by Clinicians -- is it just doctors? If it is then it is imperative that other healthcare professionals are also involved as they too will need to use the system.

I appreciate Bill's comments that should this not have taken place two years ago -- absolutely. However as it did not, it is better later than never as in the past.

I do however, believe that an initial workshop sessions involving a range of healthcare professional should take place in the UK - with both parties (to India and UK) 'singing from the same hymn sheet'.

The individuals that eventually go the respective countries are there as representatives and not as individuals- to explain the concept of NHS working to the software developing teams. It is essential that they deliver similar messages.

Any changes to any requirements must be discussed as a group in the UK- rather than those in India and US making separate decisions.

In summary we have three teams -- one in UK (in a co-ordinating role) and two overseas. This will also ensure that we do not have 100 people away at the same time.

IDX Carecast is 25 years old

11 Aug 05 13:13

"Both of the brand new clinical systems that Connecting for Health has contracted to be delivered, Lorenzo from iSOFT and Carecast from IDX, are being written from scratch overseas by software developers with little or no immediate experience of the English NHS. "

According to the British Journal of Healthcare Computing, http://www.bjhc.co.uk/news/1/2005/n507005.htm , IDX Carecast is 25 years old:-

"It (Carecast) is a development of an architecture that is 25 years old and only runs on Tandem mainframes. There are also few programmers left who can write in its development languages, COBOL and SCOBOL."

Is this the right path?

11 Aug 05 14:33

Reading this story makes me feel very uneasy. My concern is that the development programme may be guided down the wrong path by a relatively limited number of clinicians visiting developers who are based overseas and who have limited/no experience of how NHS organisations actually operate.

During system specification, it is not uncommon for mistakes to be made in the design stages, particularly if too much focus is given to a particular role. I've attended plenty of meetings with developers where the senior clinician has provided a detailed run-through of how the proposed system should work, only for their comments to be derided by the more junior clinical and clerical staff responsible for much of the management and treatment of the patient. This isn't too damaging if it's picked up early in the development, but if it's not discovered until later in the implementation, it inevitably leads to project slippage.

Whilst I categorically support comprehensive clinical involvement, I feel that so much more would be gained (and probably at much lower cost) if the LSP-employed analysts/developers spent more time on site, working in out-patient clinics, GP practices, attending ward-rounds etc. The feeling may be that this has already been done, but having been closely involved with the programme from the start, I haven't witnessed much evidence of this.

From acorns to mighty oaks

11 Aug 05 14:34

The current version of IDX is indeed based upon COBOL. iSoft uses Win32 C++. However, in both cases the manufacturers committed to write their software from the ground up as part of their ongoing product development before the national programme started. In line with industry trends, a lot of this work is being done offshore as it's cheaper.

Now that the software organisations have contracted (indirectly) with CFH the NHS has a better chance to influence that development. The early work has been the adoption of open standards from the start (such as HL7 v3). It looks like this has now moved on to the detailed clinical design.

Suspect it's too late, better save some money...

11 Aug 05 16:27

IDX Carecast/Lastword has a long history. Its technology is hopelessly old at this stage, but Its definitions of clinical workflow have been tested by a large number of hospitals.

iSoft Lorenzo got J2EE and all the fancy tools, but remmber it will be used for healthcare, let's hold our breath for NHS as it puts itself as the trial user.

I really don't know how much the clicnial visits to India and US can help. It's too late. And we don't know why the project manager and architects can't visit UK instead, on their own expenses.

'Other than GPs'

jonathan@bayly.org

11 Aug 05 19:34

As a clinician with a primary care background approached to be involved in this initiative it is disturbing to note that while the overwhelming majority of patient-clinician interactions occur in primary care, this is seen as marginal by Dr. Braunold. I have to ask is this because we have got it right, that secondary and community care has the most need of clinician input or because we are not seen, in primary care, as important to CfH? I will leave readers to judge.

Not a Sound Development Methodology

jc@e-konsult.com

11 Aug 05 20:04

Sending clinicians to talk to software developers seems like a recipe for disaster to me (think XML meets Frontal Lobotomy). It's the kind of thing I'd expect to see in a Dilbert strip! If you want to develop usable software you need to be a bit more structured in your approach - effective user requirements capture, user interface modelling, prototyping, user testing, usability testing - then you can write the code!

I can't help feeling that the whole project is unravelling at the edges...

Where are the LSPs?

12 Aug 05 09:22

CfH will have to be both skillful and lucky to find people with the aptitude and knowledge to fly out to Seattle and Chennai, let alone make them a balanced representative sample.

This situation also raises questions about the control of contracts and specifications. NHS CfH do not have contracts with the suppliers, the LSP's do! I have worked either for or with LSP's for the past 18 months. They have been gathering requirements frantically for the duration of the project. What are the reasons for bypassing LSP's now? Who negotiates and pays for the change requests?

The last major project to be signed off before NPfIT went live successfuly based on a direct contractual relationship between the University College Group Hospitals and the supplier. This of course was the '25 year old' IDX solution. Age before beauty, or has this more to do with philosophy and process than the technology?

There is no substitute for systems analysis in wards, labs and clinics. This is not just bodies on the ground from suppliers; it requires significant ring-fenced time from front line clinical workers. Cue protests of 'we can't spare the people', but here lies a large part of the gap between the headline and projected actual cost of NPfIT. Without that commitment all one has is a shiny XML white elephant.

Offshore development just doesn't work

andy.mak@virgin.net

12 Aug 05 09:25

Is this another example (in a long list of examples) that proves off-shore development just doesn't work? It might save in developer costs in the short term, but how much is being wasted by extra testing and re-writing the code when it doesn't work?

Add to that the time delays between bug found and bug fixed and the time zone differences it just doesn't make sense.

I am thinking there must be hundereds of developers right here in the UK that have developed for the NHS for decades and understand it inside out... and would be itching to get involved in the delivery of the new systems.

When are development companies going to wake up realise that the off-shore development trend just doesn't work!

Less than 100 clinicians

12 Aug 05 11:30

Less than 100 clinicians will have some input in the development? And what happens if they all disagree with the project, all have their own opinion and bring back negative reports about development.

Is design by committee such a good move?

Why always clinicians?

12 Aug 05 13:15

Having only worked in the NHS for 2 years and having experience of commercial IT projects, I cannot understand why all the emphases is always on clinicians.

In reality, it is more often the administrative staff surrounding clinicians who are using the system and who have a greater understanding of how the whole operational processes interlink.

I wholeheartedly agree with user involvement in this size of project, but we seem to always involve one particular type of user and it is often not these users who are using the systems themselves. Either that or they only use certain parts of a system, so their knowledge is limited.

Why aren't our well qualitifed NHS developpers who understand whole systems from start to finish advising our developpers? They have the best knowledge to do so, and to then liaise with clinicians and other user groups.

Having also used off-shore developpers myself in the past, I learnt the hard lesson that although technically brilliant, their developpers cut corners, fail to provide documentation, are rubbish at change control and often write very shoddy code in a rush to please their customer ... I'd be interested to know what QA is in place to prevent this as it inevitably leads to a nice system with performance holes all over it.

Will somebody take notice

rk@rkhanna.co.uk

12 Aug 05 14:53

WIthin a couple of days, a large number of very valid and sensible comments have been added on this topic. What I would like to know - from CfH perhaps - is what are we going to do about this. Are we going to have a rethink as to how we approach this in a structured and sensible manner or are we just going to dive in with both feet and get it wrong again?

Last minute... again!

EHealthInsider@RakPatel.com

12 Aug 05 19:46

As the previous commenter said "Will someone take notice"

"Why always clinicians?" - well the NHS is about health care isn't it? Most people (many clinicians included) who are involved in the (better late than never) IT revolution taking place in our business forget this. We are public sector workers being forced to work in a 'competitive' way to maximise efficiency. Unfortunately this is not easily measured when it comes to health care (quality vs quantity as always). The emphasis is not at all on clinicians! If it was, the software design would be completely different. There has been a lot of noise and money spent on involving (a proportionately extremely small number of) clinicians, but only listening to those who give the answers wanted. Unfortunately, getting the software properly right and useful for patient care (the reason for the NHS existing) as opposed to maximising profits/efficiency (the primary aim of people elevated enough in NHS management structure to not comprehend how thing work on the shop floor) takes a lot of time, investment and input from all sorts of backgrounds. IT professionals, quite understandably, place emphasis on things that can be easily counted and how the system works or can be made to work. What is needed is a system which primarily works ground-up for improved patient care, with other functions an important second. Clinicians are best placed to assess and contibute to this.

In my eyes the current plan to send clinicians to India/USA is not well thought out. Initially I responded as an interested party, but (for various home, hospital staffing and clinical reasons) have decided it is logistically impossible for me in the timescale envisaged. As an aside, I have never received any reply from my initial registration of interest. More importantly, why less than 100 clinicians? - there are more than this number of clinicians working in different jobs/specialties within a single hospital (all with different priorities and limited, if any, knowledge of others' experience). Who is important and who isn't? One nurse or Consultant is completely different to another in this respect, and all are different to physios, dieticians, OTs, midwives, etc.. and these too will have different requirement according to their subspeciality practices. More, not less, proper clinical involvement is necessary. Just as well as hospital IT staff involvement, patient administration/medical records involvement, catering services, cleaning/domesitc services, protering services, EBME, .... The system must work for the NHS as a whole. In my eyes, the only way of achieving this is to save the money spent on sending clinicians abroad, and invest more in involving many more shop floor workers 'at home'. Funding could be provided for appropriate work cover, and input be at regional workshops, teleconferencing, online workgroups, etc. as appropriate for the role required. It could involve discussion of proposed software requirements/fuctions, assessment of proposals, testing of modules, discussion of ways of working, etc. with software developers involved in the group (either in person, or more likely remotely). This would make varied input more likely, and (because of home/work constraints) more successful. I appreciate many regions have workgroups already set up, but these are limited in terms of scope and specialty input as well as difficult to attend for many because of staffing problems or geographical location. Also these groups do not have direct access to the software developers - I can easily see how much may get lost in transit/translation.

Rant over. Once again "Will somebody take notice"

Amazing

jpeart@btinternet.com

12 Aug 05 23:34

Are we really so inept in terms of IT in England that we cannot even write our own programmes and develop something as vital as our own NHS? If we need assistance in this particular case wouldn't it be more cost effective bring a couple of "expert" developers to this country rather than send x number of clinicians abroad for several weeks especially when we still do not really have enough of them doing the actual clinical jobs to cope with the demands here.

Development of CfH systems

14 Aug 05 18:39

The clinicians visiting the overseas development areas are not the 1st wave of people to have input into this software. The fact that the software companies have something to show to the clinicians shows that there has already been a lot of development work into these products.

The Business Analysists and Technical Architects for the iSOFT Lorenzo packages are all UK based, it is the programmers who are based in Chennai, it seems to me that getting clinicians involved in the detail stages of the development of these products can only be a good move.

Clinical input

BrowningC@Doctors.org.uk

14 Aug 05 20:44

It sounds very late in the day for clinical input at this level. The last clinical IT project I worked on, arriving late in the day to provide guidance which should have been available from inception, went belly-up as a result of this inadequate planning.

Even given the late start here, it seems unlikely that a handful of clinicians having a couple of fortnights each will be anything like adequate. They need RESIDENT clinical input, with the residents reporting back to a wider committee of interested parties in person in UK every 4-6 weeks.

This has not been set up properly; all we can now hope is that it fails spectacularly enough for it to be unrecoverable and thus for us all to be spared.

Are we not in danger of reinventing EPR

15 Aug 05 12:03

With all this new NPfIT overseas development I do wonder what happened to all those leading edge EPR products we used to see (but could rarely afford) at Harrogate. They might not have been scaleable to cluster level, but at least the suppliers/developers understood the NHS and what was needed. Where has all this investment in systrems, knowledge and ability gone that we have to go overseas to reinvent what we already had?

Pseudo Tests

15 Aug 05 16:23

Can someone share his experience on use of pseudo (dummy) systems for capturing feedback over a wider spectrum of users?

e.g. deploy a dummy system across the NHS net for all clinicians to "play" with it and give their feedback and a central team managing the quality & priority of these inputs to the development team.

Why to India and USA?

andy.mak@virgin.net

15 Aug 05 23:00

I always thought it was the LSP's that are managing the software. Why are clinicians not consulting with LSP's and the LSP's sending out people to kick thier software developers. After all its the LSP's that will take the kick if the product fails !!!

Loose-leaf binder

paper.records@ntlworld.com

16 Aug 05 06:30

Isn't the current medical record (chart) in effect a loose-leaf binder. It's this design that *supports* the variance that is plaguing this debate, and the development of EPRs.

How many of those involved in the policy for and design of EPRs have sat down and analysed a large sample of existing records. Zero?

Prototyping

jc@e-konsult.com

16 Aug 05 08:34

A 'dummy' system sounds like a real system being trialled. It's too late at the trial stage to do much about the basic system architecture or even the user interface. I work with 'prototype' systems early in the design phase to walk users through various aspects of system functionality. These systems are often just created in PowerPoint with some limited user interactions but they are very powerful in helping to identify some key usability issues. They are also useful in developing improved stakeholder involvement.

Engagement thru Prototype

16 Aug 05 15:05

As I have gathered from many talks on NPfIT (now CfH), there has been adequate amount of clinical engagement to start with.

Also, the documentation available with NHS should also be adequate for sytem designers of the LSPs to evolve the right system. Besides, all LSP's have have been catering to the NHS for many years (if NOT 25).

Hence, the question at this stage is NOT of appropriate information to the developers, it is moe of increasing "user" engagement and buy-in.

Hence, it is advisable for dummy systems (pre-release version of functional software with dummy data) to be made available to as many stakeholders as possible.

Making less than 100 clinicians travel and "sit" besides the developers will end up with exactly the same situation i.e. 1) spend lot of time & money on part of NHS 2) delay development process as LSP / developers will have to "entertain" the visitors 3) majority of clinicians will still complain about "lack of adequate clinical engagement" !!!

LSP's Catering for the NHS for 25+ years?

george.brown4@btinternet.com

17 Aug 05 09:48

"...all LSP's have have been catering to the NHS for many years (if NOT 25).." It has to be remembered that a number of LSP's have made **experienced** development staff redundant and replaced them with (albeit more?) offshore development staff. Although these offshore developers may be very hard workers - they may not have any experience of medicine with in the UK. They may not even be seasoned developers. So in fact if the LSP in question had 25+ years developing for the NHS - and then makes the experienced developers redundant then they now have a whole lot less experience than they had before. At a time like this I would suggest that experinced staff have never been more valued. This is just my personal opinion.

LSP's versus system vendors

17 Aug 05 11:49

There seems to be a major confusion in some of the above threads. Whereas iSoft and IDX have extensive experience in delivering NHS clinical healthcare software solutions, it would be fair to say the LSP's had rather less (at the outset).

The system vendors programmers to whom the NHS staff would be flown out to 'assist' are IDX (Seattle) and iSoft (India). The idea that suppliers are writing wholly bespoke software 'from scratch' for the NHS gives everyone a warm feeling but is misleading. All EMR suppliers operate in an International market and have to be able to tailor their systems for multiple markets. Likewise suppliers all have their existing systems that are being enhanced: this is not 'green field' software indeed it would be terrifying if it were!

Meanwhile the LSP's are Accenture, BT and CSC. These management consultancy companies have the contracts with NHS/NPfIT to deliver EMR systems NOT the system vendors iSoft and IDX.

Although system vendors were selected by NPfIT these are only 'recommended' via Memoranda of Understanding (along with indicative pricing of contracts). There is no obligation on the LSP's to use these suppliers.

LSP's can make (and break) their own arrangements as long as the functionality is delivered e.g. Fujitsu dropping IDX in favour of Cerner. Beyond iSoft and IDX there is another tier of suppliers of specialised software services with NPfIT memoranda of understanding but no guaranteed contracts from the outset.

The virtues of the model are debatable. What is not in question is that an intervention by NHS CfH to send clinicians to talk to system vendors is inconsistent with that model.

If you buy your Kiwi fruit from Tesco's and Waitrose, you do not fly out to New Zealand to haggle with individual farmers.

LSP's versus system vendors

17 Aug 05 13:46

"....The idea that suppliers are writing wholly bespoke software 'from scratch' for the NHS gives everyone a warm feeling but is misleading...." I would suggest that that is exactly what is happening. If nothing else then this clinician's trip could be viewed as a PR exercise?

Broad clinical input?

Matthew.Grove@northumbria-healthcare.nhs.uk

17 Aug 05 14:54

Surely to get a broad knowledge of how the NHS works and to get input from multiple "clinicians" into whether their software is up to the job it would be better for the software developers to come to the UK?

Sending a small bunch of interested consultants and GPs (and I'm sorry, but I really doubt any secretaries, ward clerks, physios, OTs and maybe only a handful of nurses will get to go) will just give the developers the benefit of the experience of a very limited number of NHS staff. Ones who can afford to take time out to visit another country (hence not secretaries and ward clerks who are rather too vital for smooth running for that).

Coming to the UK and doing travelling seminars whilst showing prototype interfaces, powerpoint etc would give them a broader clincal exposure in every way.

the bottom line

18 Aug 05 07:55

we used to have access to a number of state of the art level 6 EPRs, now we don't and have little hope of for several years to come.

Freebies

18 Aug 05 10:45

So the NHS is effectively giving free consultancy to two of the biggest companies in the IT market. Are they planning on offering the same amount of clinicians' time to anyone else?

no way out

18 Aug 05 22:04

'This has not been set up properly; all we can now hope is that it fails spectacularly enough for it to be unrecoverable and thus for us all to be spared. '

The one thing that NPfIT has achieved is binding NHS organisations without their knowledge or consent into long term, legally binding contracts which IMHO deliver poor value for the taxpayers money, and take some Trusts significantly backwards from EPR progress.

As Frank Burns highlighted with his 1998 strategy, we should have concentrated on the patient-carer interface, and core clinical process. Unfortunately it seems C&B has little to do with that, SUS [the Secondary Uses Service] even less. CRS [Care Records Service] is too big, and way late.

The choice of suppliers was dictated by which ones were deemed big enough to cope with aggregated clusters, rather than necessarily having the best systems. Now we have to drag them into 21C

Personally, I am convinced it will spectacularly fail to deliver good systems to support the clinical process, but this is partly because of CfH 'standardisation', and suppression of innovation and ability to react quickly to the perpetually changing political and health environments.

All very predictable, and heartbreaking for the committed NHS IT professional.

[post edited by EHI]