“We’ll get the information a lot sooner and for a lot less money by just sending a person.” I was dumbfounded. Dr Ellen Stofan, outgoing Chief Scientist of NASA, was talking to Jim Al-Khalili on the Life Scientific about no less a task than finding life on Mars.
It’s a fascinating interview from the start, or jump to the quote at 19 minutes in. So that’s why they want to send a person, not for the ultimate ego trip, but the simple purpose of finding life. Jim pushes her on the reason, doubtless at a cost which is telephone numbers cubed, and it’s simple: humans are creative, flexible and mobile.
All they have to do is break open rocks and look for fossilised microbes. With NASA’s vast resources and access to the world’s best brains, their secret weapon isn’t AI and robotics, but human intelligence.
I’m an engineer and a technophile. I read Wired online and I have three bicycles, one of them all carbon. But I’m ever so wary of the claims made for AI chatbots revolutionising healthcare any minute now.
At Best Practice show last week half a dozen companies were offering some clever algorithm to make your patients go away. They are so seductive, even plausible. But when I ask, “How many patients have actually used them?” I get stonewalled.
Substantial wedges of venture capital say “It must work”. Rather different from asking, “What works?”
What humans are good at:
- searching huge databases in milliseconds
- communicating instantly and securely
- organising and analysing information
What computers are good at:
- solving tricky problems through experience
- building relationships of trust
- caring for people in need
Our philosophy with askmyGP is very clear: we get the computers to do the boring easy bits they do so well, so the humans can get on with the real hard work of looking after patients.
We’re proud to call it HI and we’re on the same track as NASA.
PS Did I get that the wrong way round? Doh! Must get a new proof reader.
The humour of David Walliams’ character Carol Beer is all human, of course. “Computer says no” would never work if an actual computer said it. At least the hapless customer can try to reason, to get that look from Carol.
So when an actual computer says no to your patient in their time of need, it’s a punch in the face.
Let’s say you have pain in your hand (arm,, shoulder, hip, knee, ankle, foot, it matters not). It’s been there over six months, came on gradually, still mild but you decide you are going to do something about it and see the GP.
You put all this into webGP/eConsult, as you can’t seem to get through by phone. Then it asks:
Did the pain come on after an injury?
Yes No Not sure.
Honestly after six months it’s not at all clear, so “Not sure”
Computer says no.
Up comes the big red box listing six things you should do, from “seek urgent medical advice” right up to “go to A&E”. The one thing you can’t do is carry on with the eConsult. It stops right there.
Now I’m not suggesting that A&Es are filling up with cases of mild hand pain of uncertain origin over six months old. Most patients have far more sense. But who do we have to thank for this sage advice? Drs Clare Gerada, Arvind Madan, Murray Ellender and co at Hurley.
At the last count some 212 of what they call “red flags” were embedded in eConsult, sometimes as subtle as the difference between scoring pain at 5 or 6 out of 10. It can be a superhuman challenge to navigate your way through a questionnaire to reach the submit button.
“Red flags” are touted as a safety feature, but of course there is logically no way to cover all red flags for all patients in all circumstances. Any reassurance we may feel is false. Yet thinking you are wearing a safety belt, when it’s made of paper, is itself dangerous.
The one thing patients don’t have is patience, and this is the killer with red flag thinking. When the computer keeps saying no, patients won’t bother again, they go back to pleading with a human, however stressful. This is at the root of the study finding, “Online consultations don’t save time or money” where 36 Bristol practice running eConsult moved just 0.16% or 1 in 600 demands online, and this most commonly for admin issues.
As you know we have an interest in this through askmyGP, but our thinking is driven by the evidence of “what works?” We have to enable GPs to be much more efficient, and while online consultations can be a part of this (GPs tell me they save 3 minutes with each one), they only work if lots and lots of patients use them.
We’re up to 30% so far (see Concord case study), and we’re working on 50%. That will only happen if we welcome all patients, all problems, we give them a great service, and we find that patients want to help.
Computer NEVER says no.
PS You can try askmyGP as a patient on Bramley Demo Surgery. It’s simple because we tried clever and complex and it sort of worked, just not well enough. If only we’d realsed sooner! Anyway patients and practices love the new simpler version, and it’s focussed all our efforts on GP productivity.
PPS If you can’t get enough of Carol’s “Computer says no” you can see her with German subtitles for extra giggles.
OK, you’re forgiven for thinking I wrote DNA rate, and you were expecting to read about how to reduce Did-Not-Attends, perhaps a mix of clever texting apps and a patient blame-fest of warnings, charges and excommunication. No matter, although we have shown that DNAs are a system problem, not a patient problem, and if you change the system, DNAs disappear.
I wrote DNH meaning Did-Not-Help, the patients who are told by reception, “Sorry, nothing left, call back tomorrow. And make it on the dot of 8, I should, we’re very busy.”
No clinical systems measure this, NHS England would rather ignore it, no political party understands it or has the method or means to deal with it. The various prescriptions of “waiting time targets” or “8-8 x 7 day opening hours” aren’t based on examination of the evidence, so the diagnosis is wrong and the treatment will fail.
We collect the evidence through our Datalog audit, real time, patient by patient, we have n > 300,000 from over 200 practices, and the average rate is… 12%. Yes, it’s roughly twice as high as the average patient DNA rate, and it’s not caused by patients but by GP practices and systems. The highest we’ve measured in any practice is a staggering 31% of demand turned away.
I’ve worked with enough GPs to know that the vast majority joined the profession to help patients, so the data may come as something of a shock, but something which should move us to action. True, there’s a small but powerful minority who seem to take a different view, maximising profit by minimising service to the point where they just escape being caught. Sometimes they do get caught.
But we are concerned with the vast majority, those working hard to provide excellent patient service. Getting GPs to work harder is a sure fire vote loser, I think you know that. So the answer is giving them the method and means so that they can work much more efficiently, dealing appropriately with all demand and feeling in control.
Our task in fulfilling the vision “to transform access to medical care” is working out ever better ways to make this happen. We know that dreaded phrase, “Call back tomorrow… ” is uttered around 100,000 times per working day in the UK and our manifesto is to eliminate it.
You will vote many times in your life. Make your vote count.
You will live only once. Make your life count.
Founder, Chief Executive, GP Access Ltd
PS Exciting to report that the next little piece of that road is now in place, two way secure messaging between GPs, their staff and patients. It enables more efficient use of GP time by not always having to find phone numbers and hope for an answer. Spoke to one user yesterday who already loves it after sending 3 messages! It’s built into askmyGP Transform and Improve.
PPS as election guide I defer to the inestimable @jtweeterson
You told parliament yesterday that it was time for “an honest conversation with the public about their use of A&E“. Are you up for one too, as Secretary of State?
Unfortunately there is no evidence of honest conversations, cajoling, beating or any other persuasive endeavour having the slightest effect on patient behaviour in seeking healthcare. You may even be stoking demand simply by talking about it (as the Behavioural Insights Team has shown.) Dr Taj Hassan is spot on when he says it’s about the system. As Deming said, 95% of the problem is the system, and that’s where leaders spend their efforts.
You’re fiddling with four hour targets, and the language of targets, but Deming said 35 years ago, “numerical targets must be eliminated”, tellingly calling his book “Out of the Crisis.” It wasn’t until 1992 and after mental torment over what this meant that I personally got it. Well, the next principle is to eliminate exhortations…
It’s crucial to understand the difference between targets and measures, and to Deming (a statistician) measures were absolutely critical. It matters to patients how soon they are seen Mr Hunt, and therefore it must matter to you and all NHS staff: measure the times, the demands and the flow in A&E and strive continually to improve them.
You’ll be asking me how in a minute, but look, read Simon Dodds on how they did it in Luton and Dunstablehospital. There’s a very simple parallel to what they did and our work in primary care: put the senior clinician at the front of the house where the demand comes in. It saves time for the senior clinicians and saves time for everyone else.
I spent yesterday in Riverside Practice, Portadown, NI, on the day they launched their new system. Patients reported an average 39 minutes to be in touch with their GP. 81% said the new system was better. The GPs said it had gone much better than expected, and when they’d cleared the decks, most of them went home early. Someone please pass this on to Helen Stokes-Lampard who is threatening us with 4 week waits.
Wonderful stories came throughout the day. The man who got the GP call in the library. The woman who broke the new rules, turning up in the surgery without calling, but with agonising back pain. She saw the doctor within minutes. Any healthcare system which doesn’t have compassion at its heart is worthless.
Riverside GP has gone from having one of the worst access records in NI to one of the best, overnight. OK, so the preparation took four weeks. OK, so it took a year for the GPs to overcome their fear of change. But they changed overnight.
Leave aside political shocks for a minute, there’s a medical earthquake happening in NI right now, standing primary care the right way up.
It’s such a dreadful shame that NHS England is missing out, but I’m sitting by the phone Mr Hunt.
Founder, Chief Executive, GP Access Ltd
PS You can change patient behaviour, not by exhorting them but by changing their perception of where to get help. Rapid response from their own GP does it, and don’t worry, it’s less work for GPs. But please get off their backs, stop the 7 day nonsense and as Deming also said, cease dependence on inspection to achieve quality.
It’s been a fun week after our official endorsement in the Daily Mail which meant two radio interviews before coffee on Monday. Huge support came from many GPs and patients who know the truth that a demand led GP telephone consulting system has transformed their lives and their access – thank you all so much.
The DM is right on the button (in its inverted fashion) as a big report has just been quietly snuck out out by NHS England. It’s the final evaluation of the GP Access Fund wave 1. (yes, it’s right at the bottom of this page, just published though it relates to September 2015).
I’ve read the ever-so-small print and have concerns about some of their arithmetic but the headline is: from a £60m investment they identify £1.9m savings. Nevertheless, one bright spot in the conclusions is “Telephone-based GP consultation models have proved most popular and successful.”
We were involved in several of these, the financials showing a positive investment return (p31) and increasing GP appointments by 7 – 16% within core hours. This is no surprise as we’ve been saying the same for years.
The contrast with the next conclusion is stark, “Other non-traditional modes of contact (for example video or e-consultations) have had fewer tangible benefits and have generally had low patient take-up to date”
Seven schemes offered e-consultations and sadly askmyGP wasn’t among them at that stage but the leading platform was webGP/eConsult.
Moving swiftly on, it’s time for NHS England Shared Planning Guidance. Don’t worry yourself about the evidence for telephone access, the word isn’t mentioned, but on page 50 there’s £45 million ring fenced for e-consultations. That means you can’t spend it on anything else, so your CCG might as well apply. Do hurry! Applications must be in by December 23rd.
Some say it’s no coincidence that the author of the GP Forward View and NHS England National Director for Primary Care, Dr Arvind Madan, is also a major shareholder in Hurley Group/webGP/eConsult, provider of such online e-consultation software. Of course, “I couldn’t possibly comment”.
But I will be reviewing and comparing the evidence on competing platforms over the next few weeks.
As we know, only the Daily Mail can get away with ignoring the evidence.
PS A number of readers have been commenting on this Analysis of access in large GP groups. It’s all from public domain data, so I’d be delighted if any researcher would reproduce or extend it. A close look at the names reveals some surprises.
The BMA guidance includes the key message:
- Doctors must use resources efficiently for the benefits of patients and the public. Difficult decisions about resource allocation are inevitable, but should be evidence based and made in consultation with other colleagues and patients.
It’s been a tumultuous week as big business interests threaten our way of life like never before. Thankfully the attempt by Toblerone to bury the bad news as they cut 10% off our chocolate has spectacularly backfired.
More important news struggling to get heard is the publication of the first of 44 STPs covering England. BMA complains “GPs shut out of STP talks”, so what do they mean for primary care? Well I’ve pored over them so you don’t have to (at least, the parts I want to, with my secret method, ctrl-f in the document and search for the words…).
So I started with “demand” and there are dozens of references. What’s funny is that about half of them bemoan the inexorably growing demand across all our services, usual stuff. And then without a trace of irony, the other half outline our plans to reduce demand. One quotes 30% of GP activity!
Compared with predicting election outcomes, forecasting healthcare demand is laughably easy. Add 1 – 2% to last year’s demand and you’ll be within a gnat’s.
Apart from a few niche areas, the reason that demand keeps going up is because no one has found a way to reduce it. Everyone including us has tried – are we the only ones to admit it isn’t working? So reducing “demand” in STPs is either wishful thinking or in reality reducing supply, more concisely called rationing. That word doesn’t occur even once as it very quickly gets you the sack.
Try another word, “digital”. Dozens and dozens of references, all positive, and generally linked with the word “enabler”. I like this way to describe digital, because it can be an enabler of change though I don’t see it working as a driver. We haven’t yet found a way to reduce demand but it certainly can enable us to deal with demand more efficiently.
There is no doubt that digital is going to be a major component of change in the NHS over the next few years. What is unknown is how, but I’m sure it will only make an impact if it works brilliantly both for patients and GPs.
STOP PRESS: this new Dutch study on the impact of their NHG health info website is seriously encouraging for demand reduction.
PS We put a lot of effort into measuring patient feedback, as unless it’s very positive patients won’t use digital channels. High usage is the only way to get a return on all that investment in technology. We’re putting these patient measures in the public domain and keeping them updated.
Not long ago a local NHS manager told me they were pursuing the Dave Brailsford model of lots of tiny little improvements – he the cycling supremo who let us bring home so much gold from the Olympics. I thought it a one off comment, but a blog of Roy Lilley’s this week called it an NHS fad. Like him, I’ve no doubt it’s a misguided one.
Cycling is a beautiful thing, highly sophisticated and at the Olympic level, a more streamlined cut of sideburn can make the difference between podium places. But the NHS is not a competitive sport, it’s an operational business and if everyone had a rough imitation of the current best model, we’d all be miles better off.
The problem with thinking about marginal gains when the system is broken is that you miss the biggest opportunities. Attention, that very limited resource, is in the wrong place.
“Improvement is not enough. We need transformation.” W Edwards Deming said it thirty years ago.
“If I had asked people what they wanted, they would have said a faster horse.” Henry Ford said it, a century ago.
If he had listened, we’d have carbon fibre wagon wheels by now.
Gee up, Dobbin.
PS. You’ll have noticed all the noise about STPs. They have transformation in the title, of course. But they are really about money, and I’m afraid that’s no way to achieve transformation. Cost cuttting, yes, restructuring, yes, noise, aplenty. What we need is different thinking, starting with patient demand.
PPS. “Marginal gains secured by new care models…” was the title of a King’s Fund blog this week. Marginal gains? Remember, the new care models were the flagship policy of the Five Year Forward View. Two years in, untold £millions spent, and they are bigging up “marginal gains”??? Please don’t get out of bed. If it was working they would have something more to shout about
I shared a lift to the Best Practice show this week with my very good friend Gareth McCague and crawling along the M6 he asks me, “what does your demand led system mean for premises?”
Well that got me thinking, because one of the first things we notice is the empty chairs on Mondaymorning – GPs are all hard at work, consulting with patients at home. We call it the “no waiting room.” I know of practices who have carved out another consulting room from the unused space – but of course in a new build, cutting the seats from 40 to 20 would make quite a difference to building costs.
Then of course there’s the car park – it’s always a sign of things working well when we can drive straight into one of the many empty spaces. “That’s it!” cries Gareth (who knows everything there is to know about about GP premises and pharmacies by the way)
You’ve all seen the new surgeries where most of the land is taken up by parking. With all the attention on premises in the ETTF & GPFV funds, you can bet lots more tarmac will be laid for patients to park their cars, so they can wait on the chairs, to see a GP they didn’t need to see – it could have all been done in minutes over the phone or online.
Premises, premises won’t solve the problem. Everything is connected – it’s the system.
PS. Michael McKenna runs a single hander practice off the Falls Road in Belfast. There is no parking. Until this April he would have 30 patients lined up of a morning, for his 11 seats. They’ve gone, and I particularly love the bit in the NI video where he looks out of his room on launch day and quietly says “Wow”.
PPS. I admit to writing this in Luton airport, plane delayed. Just noticed in the gents that the new Dyson Airblade dryers are built into the taps. Ah ha – no separate queue for the dryers, lower footprint, value in airport space £££ – everything is connected.
People are sending more and more parcels these days, if you’ve noticed the queues at your local post office. This has got me thinking. I’ve learnt from a focus group that people would love to be able to send parcels from home at a time convenient to them, evenings and weekends especially.
So I’ve invented a new service called Mailidioto where we collect from your own home at any time you choose for one simple fixed fee. We take it towards the destination address and if you’re lucky, your fee will get it all the way. Obviously, if the fee runs out before we get there we’ll have to chuck it in the hedge, but that’s business.
How do you rate my chances with Mailidioto? Have I stumbled upon the difference between a contract for inputs and a contract for accountability?
There’s a wonderful thing in the NHS called the GMS/PMS contract for essential services, which balances unlimited accountability for managing the registered patient who is ill (in hours), with unlimited discretion on how to manage them (for the qualified professional performer).
So when we hear talk of salaried models, or “GP hubs”, WICs etc, these are all about inputs. They undermine accountability for the registered list, losing continuity of care and increasing costs. They are about hours worked, or appointments supplied (most APMS contracts) . We know the tenuous connection between “inputs” and “outcomes” (hello USA), and yet we see NHS England promoting these input models.
“£6 per patient to extend GP access from 2019” runs the headline. “Well actually…I’m just going to locum for as long as I can because I cannot see why anyone would want to commit to losing their (social) life” runs the comment from one anguished reader. You all know how hard it is to recruit GP partners. The more sessions are available for transactional GP, both in and out of practices, the harder it is to recruit for long term relational GP – they are the same people.
Extending hours makes no difference to capacity, and therefore has no impact on the three week waits to see a GP. Worse, it cuts capacity as utilisation is lower out of hours. Worse, it takes GP time away from high demand core hours. Worse, the unscrupulous will simply work the same time but claim the extra pay. The unscrupulous near to hubs are simply fending their patients off to them. It’s happening now – bad money drives out good money.
It doesn’t have to be like this. Scotland has abolished the abomination of QOF, and put the money into core funding. Northern Ireland is helping its GMS practices cut the wait for patients to minutes, and cut GP workload at the same time for a tiny fraction of £6.
The BMA and RCGP need to be absolutely clear against this undermining of the GMS contract and GP professional accountability. Put the £6 into core GP, making partnership more attractive. Put it disproportionately into deprived areas, bringing equity in place of the inverse care law.
GPs, where were you when they stole your profession?
Are you going to stand up?
Before it’s too late?
PS We’ll be at the RCGP conference next Thursday/Friday, on stand 33. See you in Harrogate!
PPS “Meet the digital GPs” in Pulse includes an interview with Dr Rupert Bankart about his practice and askmyGP.
How many times have you heard the promise that patients will self care with health apps and websites instead of bothering their GP? And then compared the promise with the daily reality of demand slowly, steadily ratcheting up?
I wrote The Diversion Myth over a year ago and I’m glad to say we can revisit it with our own real data. We’ve never claimed that our askmyGP online access platform would divert demand away from GPs, but we’ve said there may be potential and we’ve been trying very hard to prove it. Every welcome page links directly to the excellent NHS Choices site, and we go further: having entered their symptoms, patients are invited with one click to go straight to the relevant NHS Choices page. Perfectly tailored patient education!
Since adding the tailored links, the results are:
16,710 total patient submissions
4,458 links presented
519 links visited
10 patients decided not to consult.
One of them was me, with a throbbing earache, discovering to my delight that antibiotics were not advised and it should clear up within 3 days. It did.
Friends, this isn’t going to work. I was hoping for perhaps 5% diversion. 2% would have disappointed. 0.06% is so tiny that even with brilliant optimisation multiplying the effect by a factor of 10, it would reach a barely perceptible 0.6%.
What we have shown is that patients are keen to seek help online. Fears of a deluge of trivia are unfounded, you need to beg them to use it, but so long as they get the message from a trusted GP and they get a rapid response, 20% will shift from phone to online. It’s then easier and less frustrating for patients and practice to manage demand. I don’t think that is impressive, compared with 80% online say for travel bookings, but it’s promising and we are working on how to build that to over 50%.
I’ve explored the challenges and the reality for digital primary care in a poster at last week’s EFPC conference:
Interested in your views on this.
PS Our competitors webGP are still claiming 18% of patients self managed an issue for which they “planned” to consult, indeed that “60% of patients were able to resolve their health concern without a visit”, implying they did so themselves. Strangely, we never see any numbers. Does anyone independent out there using webGP have any evidence of overall demand reduction?