EMR in a hurry: snafu ahead

Obama proposes a fast start on electronic medical records within the stimulus package, a bad idea. A case study from England on spine records.

Barack Obama has invited Paul Krugman to come up with better ideas for the stimulus package. Can anybody play? None of us ‘umble bloggers here has a Nobel, Fields or – much the best money – a godly Templeton. But between us we have already come up not only with an early broad list of initiatives but with nifty specific ideas for a few billions in instant spending:

* designer pylons for the interstate electricity grid that don’t make you think immediately of the Welsh for Semtex – these can be ordered in advance of fixing the routes

* 100 million cat’s eyes for safer roads (bonus feature: very labour-intensive to instal)

* best of all, Jonathan’s proposal for a malaria eradication campaign.

Obama’s planned stimulus package includes the electricity grid and road upgrades so we may be partly in luck. However one plan the transition does have is premature. They have not been playing close enough attention on how to fulfil Obama’s campaign promise for electronic medical records (here, here, and here).

From Obama’s Fairfax speech on 8 January – even the URL is on message – my italics:

To improve the quality of our health care while lowering its cost, we will make the immediate investments necessary to ensure that within five years, all of America’s medical records are computerized.

Er, what immediate investments? FWIW, I did suggest that two pieces can reasonably be started right away: a secure e-mail network (VPN) for American health care workers, and regional data warehouses for the results of X-rays and other scans. The data warehouses don’t depend much on the rest of the system; scan files only come in a limited number of formats, and scans are so expensive that writing conversion filters will generally be worthwhile.

But the main course is the actual detailed records of care. Here subsidising doctors and hospitals to rush out and buy lots of incompatible state-of-the-art kit is a bad idea. It’s roughly what happened in England in the 1990s, and they had to start all over again. Down the road you will have to spend even more to put it all together. You have to think out first what you are trying to achieve: a coordinated national EMR network that balances privacy with efficient care. And you have to get the medical practitioners on board, not just IT firms, management gurus and consultants. (Remember Ira Magaziner?)

Let’s walk through a small example from the English NHS project.

(England is for once the right term; Scotland, Wales and Northern Ireland now run their own NHS clones). At £12.4 bn for 50m patients, it’s a mammoth undertaking – with lots of problems to learn from. It was started in 2002, but won’t be complete before 2014; Obama’s five-year horizon is the kind of wishful thinking that guarantees trouble. Anyway, grinding forward they have just – six years in – announced progress on one small piece of the puzzle:

Adastra, which supplies almost all of the systems supporting out-of-hours care in England, have provided a link between Bury & Rochdale Doctors on Call (BARDOC) and the Spine, which will mean clinicians will no longer need to move between the two applications in order to access the SCR [Summary Care Record].

One in five Brits use out-of-hours NHS services in a given six-month period (National Audit Office research in 2006), so it’s a non-trivial issue. The reason why Bury and Rochdale – neighbouring small industrial towns in Lancashire, with a combined population of 180,000 – are highlighted is that this was the first area where summary care records were uploaded to the national NHS Spine record. So what happens in Bury today will happen everywhere in England tomorrow.

So you. an English resident, fall ill in Bury out of office hours and go to an out-of-hours provider – A&E department in a hospital, replacement GP, community nurse, dentist, etc. They can now look up your administrative details and selected medical data: minimally, current medications, allergies, and adverse reactions. In most cases the facility won’t make much immediate difference, but in a few it could be lifesaving, and hopefully will improve continuity of care with your regular GP and reduce duplicate tests. Not a revolution, but worthwhile. Remember: this is only one piece in the whole puzzle.

Now here’s the nub. For this to work at all, the spine record has to be completely standardized and integrated. Separate systems won’t do, even if they can communicate: that would work for transferring records between hospitals, but fiddly conversions and multiple access régimes can’t be handled by a community nurse from her laptop.

The administrative part of the English NHS spine record has 29 fields, which I’ve copied at the end. This is not optional for any patient, though you can object to the upload of the additional medical summary care record. (There’s a big argument over how much medical data should go into the Spine. The managers would like more than just the medications, allergies, and reactions, and privacy groups don’t want anything. The great British English compromise is to see how it works out in Bury. In the absence of real incentives for either GPs or patients to upload a lot of data, I can’t see this happening.)

Who can look at the Spine record? They will need a smartcard and access to the up-and-running VPN. The English NHS has issued 400,000 smartcards so far; the final total is expected to be 1.2 million. Scaling up even the current number, that would mean 2.8 million users in the USA, within the 14 million workers in the health sector. There are 236,000 doctors currently registered in the UK and almost 700,000 nurses. This includes the non-England bits, retirees and the otherwise inactive. It’s a reasonable assumption that although access can technically be limited, most of the 400,000 smartcard users can see a spine record. At this scale of access, the data are effectively public. The NHS has already lost 4,000 cards. The security is like putting a £13 shed lock on your front door: it doesn’t keep a thief out but ensures that he has to commit burglary to do it.

lock&smartcard.jpg

Now consider: would Americans stand for a database like this being accessible to say 2 million healthcare workers? I have little feel for the preference map, but would guess not. Even so, you can still get most of the benefits of continuity from much more local coordination. Most of the payoffs come at the city level, with diminishing returns as you go up to the state, region, and nation. US policy can then get most of the way working with the grain of federalism. The best here is very much the enemy of the good: a second-best solution that’s acceptable to medics and patients, and can be set up quickly, is far better than an ideal model that will be bogged down in protests, lawsuits and partisan politics.

Solving the conundrum is well above my pension grade. But what the hell, I doodled a map of health regions to show it could be done, so in the same spirit here’s my rash suggestion for US spine health records:

1. A national database (n ≈ 350m) limited to names, health numbers (Social Security numbers for most, but extended for aliens and children), dates and places of of birth to avoid ambiguity, and the location of:

2. Administrative records (10m < n < 40m) held at the state or regional level, similar to the English ones with extra data on insurers, linked to

3. Summary care records (200,000 < n <10m) held by cities, counties or metropolitan areas.

The Wonk Room has listed and demolished 11 myths about Democratic plans for health care reform. Surprisingly there is no 12:

The evil Gummint is going to put all your health data into a giant database in Washington so that your need for Viagra will be an open book to nosy neighbours, journalists, teenage hackers and UN agents in black helicopters!

Given the lack of detail in current plans, it’s a far more legitimate worry than the others.

********************************************************************************************

Appendix: administrative patient data (PDS) held in an English NHS national spine record

Source here. Explanations edited.

1 NHS Number

The unique patient identifier

2 Patient Name

Including any previous names, aliases and preferred name, e.g. Chris rather than Christopher

3 Date of birth

4 Place of Birth

5 Additional birth information

The delivery time and birth order for multiple births

6 Date of death

7 Death notification status

Indicates a formal death certificate has been issued for the patient and the death has been registered

8 Gender

Administrative gender

9 Address

Includes main, temporary and correspondence addresses

10 Alternative contacts

The patient’s legal guardian, proxy, family/close contact

11 Telecommunication contact details

12 Preferred contact times

13 Preferred contact method

The patient’s preferred contact method, e.g. telephone contact by proxy, no telephone contact, sign language required in face to face contact or minicom

14 Preferred written communication format

Specialised patient contact requirements, e.g. large print, Braille, audio tape

15 Preferred language

16 Interpreter required

17 NHS Care Record consent to share status

Indicates that the patient has agreed to share their health record

18 Nominated dispensing contractor

The patient’s nominated dispensing contractor that could include a community pharmacy, dispensing appliance contractor and a dispensing doctor

19 Reason for Removal

Indicates that the patient is not registered with a GP

20 Previous NHS contact indicator

Indicates that the patient confirms they have had previous NHS treatment. This allows the PDS National Back Office to check for a duplicate of a record

21 Patient call-back consent status

Indicates that the patient is willing to be called back from a Choose & Book call-centre

22 Shared secret

An encrypted password used to validate a patient’s identity when contacted from a Choose & Book call-centre

23 Sensitive record indicator

Indicates that either the record is not accessible to PDS users or that the content of the record is being reviewed to ensure the data is correct

24 Primary Care

The GP Practice with whom the patient is registered

25 Date of Registration

The date the patient registered with the GP Practice

26 Back office location

The back office location associated with the GP Practice where the patient is registered, used for the administration of primary care

27 Serial change number (tech stuff)

28 HealthSpace status

Indicates that the patient is registered to use HealthSpace

29 NHAIS Information (tech stuff)

Author: James Wimberley

James Wimberley (b. 1946, an Englishman raised in the Channel Islands. three adult children) is a former career international bureaucrat with the Council of Europe in Strasbourg. His main achievements there were the Lisbon Convention on recognition of qualifications and the Kosovo law on school education. He retired in 2006 to a little white house in Andalucia, His first wife Patricia Morris died in 2009 after a long illness. He remarried in 2011. to the former Brazilian TV actress Lu Mendonça. The cat overlords are now three. I suppose I've been invited to join real scholars on the list because my skills, acquired in a decade of technical assistance work in eastern Europe, include being able to ask faux-naïf questions like the exotic Persians and Chinese of eighteenth-century philosophical fiction. So I'm quite comfortable in the role of country-cousin blogger with a European perspective. The other specialised skill I learnt was making toasts with a moral in the course of drunken Caucasian banquets. I'm open to expenses-paid offers to retell Noah the great Armenian and Columbus, the orange, and university reform in Georgia. James Wimberley's occasional publications on the web