In this episode of the McKinsey Podcast, Simon London speaks with McKinsey senior partners Penny Dash and Shubham Singhal about the common challenges facing health systems around the world, and the technologies that could make clinical care more efficient and affordable.
Simon London: Hello, and welcome to this episode of the McKinsey Podcast, with me, Simon London. How to organize healthcare is one of the hottest political issues of our time. Every country is struggling to find the right balance between social provision and market forces, to find the optimal number of payers and providers, and to figure out how all the different pieces should fit together.
Stepping back, we see that health systems in every country, at least in developed economies, face common challenges. Populations are aging, the burden of disease is shifting from infectious diseases to long-term chronic conditions, and then there’s the onward march of technology. Not only the eye-catching clinical technologies but also information technology that could make health systems orders of magnitude more efficient, if only they could be deployed at scale. The upside is enormous; the degree of difficulty is very high.
To discuss these thorny issues, I sat down in Chicago with Penny Dash, who is a former practicing physician turned McKinsey senior partner, based in London, and Shubham Singhal. Shubham is a senior partner based in Detroit and the global leader of McKinsey’s Health Systems and Services Practice.
Penny and Shubham, welcome to the podcast. Thank you for doing this.
Penny Dash: Thank you for having us.
Shubham Singhal: Pleasure to be here.
Simon London: Without going too deep into the challenges facing any one country, if you look across the OECD [Organisation of Economic Co-operation and Development], look across developed economies, Penny, what are some of the common challenges facing health systems?
Penny Dash: We tend to say that the challenges are actually very common across all those different healthcare systems. Despite having very different regulatory systems, despite having different payer–provider systems, the challenges are remarkably consistent.
The challenges are, first of all, an aging population—in many ways that’s a great thing, we’re all living longer. But we are having a different burden of disease than the burden of disease that we would’ve seen 20 years ago and certainly 50 years ago and 100 years ago. What we are now seeing is that people don’t tend to die as children of infectious diseases. They don’t tend to die during their adult years of infectious diseases, they don’t tend to have accidents as much. If they have a heart attack, we can usually treat it and cure it and people survive. As a result, the diseases that we’re seeing are chronic diseases associated with aging populations, but also with poor lifestyle behaviors.
Simon London: I had to look this up as a piece of terminology to make sure I had it, a chronic condition is basically defined as something long-term, right?
Penny Dash: It’s something that you live with, but it may be something that you die with, so it’s a condition that you have for a long time. As long as we effectively manage it—so you either take the right drugs or you modify your diet or your behavior—then it’s something you could have for ten years, 20 years, 30 years, or indeed, longer.
Simon London: Diabetes, cardiovascular, depression, certain things like that.
Penny Dash: Absolutely. As you just raised, mental health is increasingly one of the big challenges that sits within that. As people get older, it starts to become frailty, so it may be arthritis, it may be chronic back pain, and of course, dementia is an increasing challenge.
The key thing is that all of those require a very different type of healthcare than the healthcare model we would’ve seen 20, 30, 40 years ago, to deal with the things that one might call more acute illnesses. The more acute illnesses of having an accident, having a major infection, having a heart attack, are all things that we deal with in a hospital setting.
This different set of diseases need a continuum of care. They need care typically organized around you and around your life, they need care that focuses on your lifestyle, focuses on your health behaviors, focuses on your broader well-being, but also, crucially, requires healthcare that is proactive rather than reactive. That is a completely different mind-set both for caregivers as well as for the different infrastructures and the systems and the mechanisms that deliver healthcare. That is a major challenge for all developed countries: how to shift from this reactive, hospital-based model, to a more proactive, community-based model.
Simon London: It’s not just that we need more care, it’s that we need different types of care as well, so there’s a sort of structural adjustment required.
Penny Dash: Absolutely. The third challenge is that the costs of healthcare are escalating. They’re escalating partly because of an aging population—particularly an aging population with more of these chronic diseases—but they’re also escalating because we are developing more and more technology, which is very exciting.
We have more technology, we have more complicated treatments for more people, but we have also seen quite a significant increase in the unit cost of many of the components of healthcare, whether that’s the unit cost of a stent, or the unit cost of a doctor or a nurse, or indeed of a building. Healthcare is quite a capital-intensive industry.
Simon London: Shubham, let me bring you in here. I am somebody who is a long way from the sector in terms of really understanding the intricacies, but it feels like a classic productivity problem in some ways, at least.
We need more healthcare, better healthcare, different types of healthcare, but we need more healthcare for less, because that ultimately is going to be the only way to sort of get a grip on the cost issues, the spending issues. Would you agree with that? Is that a sensible and appropriate way to think about it?
Shubham Singhal: There’s a little bit of nuance that exists underneath that. I think if you pick up from what Penny said, you’ve got three kinds of issues. The first issue is you have built up a high fixed-cost infrastructure that is no longer entirely relevant. It’s relevant in parts, we still need some of it, but we built a hospital-based infrastructure, which required people that got sick, we brought them in, they got well, we sent them out. If we now have to do continuous care, how are we going to do that?
Normally, in other industries, what would happen is the old models would go away—through bankruptcy or other ways—and new ones would emerge.
That is not how it works in healthcare. These are very important pillars of the community, these are places where people go for care, lots of people are employed. It’s not the easiest of things to do. Even in the United States, which is often nimble at restructuring industries, the bankruptcy code that exists in the United States for every company doesn’t apply to healthcare.
Simon London: The gales of creative destruction are not going to apply to healthcare in the same way that they apply to the widget industry.
Shubham Singhal: When you think about productivity, the first part of this is restructuring. We don’t have a mechanism to restructure. There’s a second part of it, which Penny also mentioned, around technology. It is fascinating, in almost every other industry, the application of technology over time, not instantly, but over time, leads to more being delivered for less.
Not so in healthcare. Some amount of it is because healthcare is complicated. New treatments that come out, the R&D that is required, that has to be recovered, that is definitely true. But some of it is also partly because of the previous point in the creative destruction, and the lack of a lot of disruptive change—you don’t get step-change improvement and productive business models that would be that much better than the previous one.
There is something about how do we, while delivering the right quality of care and the safety and the like, create the incentives to drive productivity, labor productivity, so we can do more?
Penny Dash: I think what sits behind this in pretty much all countries is that it’s an emotive subject, and it’s an emotive subject for individuals, it’s an emotive subject for families, it’s an emotive subject for communities, and it’s certainly an emotive subject for politicians.
While there are absolutely opportunities for a disruptor to come in, and there are examples of disruptors globally, so there are examples of organizations or individuals who have said, “We can deliver this same treatment at higher quality and for lower cost.” The implication of that would be that they would essentially win, and that the old would lose.
It’s not anything that most people are willing to do. On an individual level, you may feel completely comfortable buying a new laptop that gives you better computing power at lower cost, and you’ll absolutely make that transaction. Most people equate the highest cost, the most traditional form of healthcare delivery, with the better quality. Often erroneously, but that is what they do.
Shubham Singhal: Picking up on that, it’s interesting because it gets in the way of technology-driven productivity improvement. We all know technology. We buy a new phone, it’s going to have a glitch or two, the patch will come out, we’re going to fix it. Now, how many of us are going to be comfortable having somebody get care where it glitched while their loved one was getting that care and say, “Oops, we’re going to fix it.” It just is not something that we can all wrap our heads around.
Simon London: Yeah, and that is legitimate.
Shubham Singhal: That is extremely legitimate from a standpoint of safety and quality, so there are good reasons that get in the way of saying we can apply the same. But it does create, under the legitimate umbrella, a general sense of not wanting to change or disrupt as well. Even when disruptors come out, it’s hard for them to scale.
Simon London: Thinking about this through a layperson’s lens while accepting that there are legitimate reasons why we might not want the gales of creative destruction to blow through the healthcare sector in the same way as the other parts of the economy, you’d step back and say, “Is one of the issues here transparency?” If people had the information, and they knew that what they took to be the Rolls-Royce or Cadillac version of care didn’t deliver better outcomes than something less expensive, they might choose it. As a healthcare consumer, those things are not always apparent.
Penny Dash: It’s been about 30 years since people started to call for greater transparency in healthcare. There has been a move towards greater transparency, and you can now, in many countries, find a website that will give you some sort of ratings of performance of hospitals and doctors and so on and so forth. It is still not actually scientifically, technically, quite as robust as it could be.
There are some reasons as to why it perhaps hasn’t quite taken off in the way that it could. One of the issues is how you adjust for case mix. Often the best surgeons will take on the most complicated cases, and de facto then they might have a slightly higher mortality rate than the less good surgeon who takes on the easy cases. You need to take in to account that case mix adjustment. While you can do that, it’s quite clunky and it’s quite complicated, so that is one reason.
The other reasons, though, are quite interesting. One is an awful lot of doctors, and indeed, hospital managers, administrators, as well as politicians, have been very resistant to greater transparency.
The third thing, though, which is fascinating, is how few patients actually look at this stuff. Some research shows people spend more time choosing where they go for breakfast or where they go for dinner than they spend looking at “Is this surgeon actually any good?” It’s quite fascinating the degree to which people don’t do that. I think that’s transparency on the quality side. Then the other bit is transparency on the cost side. In some ways, transparency on the cost side, one could argue, would be a much more rapid driver of change, but the reality is most consumers do not see the cost.
It’s usually coming at them in a proxy way. Either it’s coming through their health insurer or it’s coming through government and so on. As a result, if you took some of these examples, most consumers do not know whether their hip prosthesis, for example, cost $3,000 or whether it cost $300. Therefore, the decision about which one of those hip prostheses is used is down to the individual surgeon, the individual institution, and very occasionally, the payer.
Shubham Singhal: I’d add a couple of things. One is on the cost side. Transparency has to be contextual. My mother has just had a knee replacement. There was absolutely no way, when picking the surgeon and the institution, to be able to know what the total cost would be. Even now, as the bills arrive, they arrive from the anesthesiologist and the physician and the facility and the walker and the pharmacy that gave the drugs, et cetera.
There is no obvious way in which we can figure it out today, because of the way these institutions are organized. If you were to even bring it back to, in the chronic disease, where we need this to be kind of continuous care, we don’t have a way to be able to engage the person continuously around their lives.
Even the best of chronic disease management today is, “Well, we’ll call you up. We’ll set up every three months.” Maybe if you had multiple chronics, somebody shows up to your house, but we haven’t quite knit together the caregivers, the technology, and the monitoring to be able to make something easy for the person. To say, “We see that certain metrics are deviating from what we would expect them to be. This means you should either show up, you should go to the ER, or you should go somewhere else, and by the way, these are the three we picked for you, given your location that are closest and cost X,” and so on. We don’t have an ecosystem that is intuitive, personalized, and available around the person.
Simon London: I mean, are we basically saying that one of the reasons that the healthcare system doesn’t work very well or doesn’t operate very transparently as a market for consumers is it’s just so darn complicated? There are so many piece parts of it. It is just hard to figure out. Is that the gist?
Shubham Singhal: It is the gist, and just to add to it, given in most developed economies, it is a primarily third-party paid system. It is designed for the professionals and the industry more so than it is designed for the consumer.
Simon London: Let’s segue back to technology because we talked about very exciting clinical technology, which is true, and we read about them, we heard about them. I would have thought that there’s a lot of nonclinical technology—electronic health records being the obvious example—data ecosystems, cloud computing, a lot of quite granular robotic process automation, things that you hear about in other aspects of the service industries, that could be applied here if there was a will and if the investment was there, to, if not immediately fix, but over time solve a lot of these issues.
Shubham Singhal: I think technology exists. I might beg to argue about whether there’s enough money available. Just take the United States, the stimulus bill put out $20 billion for EHRs and technology, and we did electronify everything and spend all kinds of money.
Simon London: EHRs being electronic health records.
Shubham Singhal: Correct. But back to the industry structures that exist, while the data, by law, is supposed to be the consumer’s, good luck trying to get that data for yourself, let alone aggregate it across different sites of care. If you showed up at doctor A, hospital B, and pharmacy C, you have no way to aggregate that around yourself, and in effect, the custodians of the data, which are the technology systems working on behalf of the providers, have made these very siloed. It’s gone from being in their files in their basement to being on servers that are dedicated to them.
The clinical data is electronic in most countries, and so it’s a matter of do we have the right regulatory construct in place that makes it a patient-centric, longitudinal, complete data set versus institution-centric? That’s where I think there is immense promise. Because two or three things happened. One, it moves the power away from the industry to the consumer, because they have the data, their ability to use that to delegate to whoever, or people on behalf of them.
The ability for us to apply artificial intelligence (AI) and being able to look at that data and sort out what the interventions may be, some of which may not be clinical at all. It may be around the person needs food, the person needs to have a place to call home.
It could be any number of things in which we could intervene that could prevent adverse health events from happening. There’s a significant opportunity to refashion healthcare to make it work around the person with the data that we have, all of these technologies that exist today. Then the means for engagement that exist; in developed countries, almost everyone has a smartphone and the ability to be able to interface with people across a full continuum, from virtual in their home to a community-based site to, if needed, bringing them into a hospital, is all possible to do.
Simon London: Are there countries that are further ahead in this? I mean, if you look across the world, which are the countries that seem to be furthest forward?
Penny Dash: If you go to Sweden, you go to Finland, it’s now completely normal for an older person with dementia to maybe wear some sort of tag around their neck, or a bracelet, or to put it into their watch, or whatever it may be, that allows you to set off an alarm that says, “This older person with dementia is now leaving their home or leaving their premises.”
Now, that is a ridiculously simple intervention, which, prior to that would’ve required 24 hours a day, seven days a week monitoring of that person with dementia to make sure they don’t go off wandering and put themselves in danger. A simple sort of tagging device gives you quite a significant productivity opportunity.
Probably most countries in Europe now, as well as here in the US, are now starting to offer some sort of online booking, whether it’s to see your primary-care physician or to make a hospital appointment and so on. That is a significant productivity opportunity. Even in the UK, which has significantly fewer administrative staff than we have here in the US, a typical hospital in the UK would still employ 300 people just to make outpatient bookings and bookings for elective surgery.
Another good example that is starting to happen is things like linking up availability in imaging centers for X-rays. For example, this is now starting to happen in a couple of cities in the UK, so rather than you waiting to try to get an appointment, to have an X-ray, or a CT scan, or an MRI at one hospital, you can actually say, “I’ll be available for any hospital in this city.”
Then you can have an alert system that says, “We got a cancellation for tomorrow.” Now, that is both great for me as a user, I don’t have to wait three weeks to get this test, which is good, but also it means that we increase the utilization of those slots. If you look utilization of any procedure in healthcare anywhere in the world, utilization is woeful.
If you actually looked at how most doctors in particular spend their days, an enormous amount of their day is capturing information that has already been captured somewhere else. It is around sharing information that you could’ve shared electronically. If you mapped out the typical way in which many doctors in particular spend their working hours, which are a very precious commodity, there’s a significant opportunity to really help them to become more efficient and more effective.
Simon London: There’s a lot of productivity hidden within the daily work routines or schedules of clinicians, but also nonclinical professionals that could be unlocked, and they might even welcome being unlocked.
Shubham Singhal: The other interesting aspect of this that we’re getting to now is the ability to expand our access to care. Take rural populations, for example. It is quite difficult to ensure that the right clinicians are available. If you look at it with all the technology that we have now, you could have a less credentialed person—could be a nurse, could be, in some cases, just a primary-care physician—you could have either the physician, in the case of the nurse, or the specialist, in the case of the primary-care physician, be available remotely. That’s another way in which we could gain productivity while extending access and being able to reach and touch and be able to provide care to more individuals.
There is some interesting productivity to be had, as we think about what the makeup of the spending the time that the physicians spend, and then what the makeup of the caregiver is that we need to bring together to be able to deliver to the patient.
Simon London: Penny, I know you do a lot of work with hospitals in particular. What’s the hospital of the future going to look like, and how long is it going to take us to get there?
Penny Dash: We’re starting to see some of it evolve now, we’re starting to see some development. I think for all the reasons that we’ve said before, we probably need fewer hospitals despite an aging population. We need fewer beds, we need fewer operating theaters and so on. We definitely need a lot more technology. We are, interestingly, over the last 20 to 30 years increasing sub-specialization, creating some centers of excellence. For the reasons that Shubham’s just said, I think actually some of that may change, and we will see the ability to provide more care at a local level, networked into the more specialist centers.
I think some of that trend towards centralization is starting to be reversed slightly. I think we could see some radical differences in what happens to you when you’re in a hospital. The first is, I think, absolutely it will become much more tech-enabled. Everything is RFID [radio-frequency identification] tracked, all of your vital signs and investigations go directly online.
I think we’ll absolutely see monitoring devices. The concept of the nurse coming around and taking your blood pressure every four hours I think will be gone in the next five years, and everyone will just be connected to a monitor and it will all be automatically connected. We will start to see things like pills and patches on your body that will, again, start to do a lot of that in an automatic way. One of the things I think would be the much more radical change in the future will be around robotics.
One of my more radical assertions is that in ten years’ time, 15 years’ time, maybe 20 years’ time, that the concept of putting hands inside a body will be seen as remarkably weird. Large chunks of surgery now—cholecystectomy, prostatectomy, and so on—many of those things are now done through noninvasive surgery. I think that’s just going to grow and grow and grow.
Shubham Singhal: I think I just wanted to add a couple of things. One is, with all this information available inside a hospital, already a number of them are beginning to apply artificial intelligence.
Even being able to predict that the indicators are showing, and some tests have shown, four hours earlier before the patient begins to decline, and being able to intervene, is a significant opportunity that exists.
Simon London: Artificial intelligence, it’s interesting. We’re just talking about very good pattern recognition, right?
Shubham Singhal: Indeed.
Simon London: Across very large patient populations, and the ability to spot patterns that are probably beyond the ken of the most educated human physician.
Shubham Singhal: Indeed.
Penny Dash: That’s one of my big hopes, is how do we harness without compromising confidentiality, the fact that we have eight billion people on the planet with largely similar physiology, but with some subtle differences. How we start to really use big data and analytics to spot trends, to spot that this combination of, you know, so Penny Dash with whatever genetic makeup, if Penny Dash doesn’t get enough sleep each night and goes and does this, this, and this, will put herself at higher risk of this particular disease.
If she did sleep an extra hour at night, and she did eat more green vegetables and so on, she may then put herself at much lower risk of that disease. That is all feasible now, today.
Simon London: Just. It’s a huge just. It requires the data to be available in formats that can be combined in databases to which one can apply AI.
Penny Dash: It also requires an economic model, and that’s one of the quite interesting challenges in some of this, is who is going to do that research. Who has the economic interest to want to know it benefits me to take on this pattern of activities and so on? That’s quite an interesting thing about the whole big data and analytics in healthcare. We haven’t quite worked out the economic model: Who’s doing it for what benefit?
Shubham Singhal: I think that will depend on different systems, and how care is paid for and constructed. If you follow all of this through, the interesting thing where it leads you is the care part of healthcare. If and when we’re able to do all of these things, we will be able to have every one of the healthcare providers spend more time on care.
The physician can be unhurried, and the nurse can spend more time not trying to read 15 monitors but instead helping the patient with empathy-based care, and other care extenders, whether it’s a nutritionist who’s needed, et cetera, can be available to help people.
The interesting thing, in some sense, is people often think about productivity as a negative term, that we somehow are going to lose jobs. In fact, I would argue, it has the potential to enrich the jobs that exist and allow for us to fund additional jobs that are needed to be able to help the patient, which is what healthcare is about. That’s the other part of the optimism that I have, which is it becomes more personal and it becomes more care than things that get done to the human body.
Simon London: Fifteen percent of health outcomes are determined by clinical care, and the standard and quality of clinical care that you have. The rest of it is genetics, behaviors, and the socioeconomic factors. Do we as a society spend too much time and energy on the clinical side of it, and actually not enough time on those other factors?
Shubham Singhal: I think we do, and in some sense, healthcare is designed as sick care, essentially, and it was very clinical in nature and we’ve got a specialized profession, of course.
I think what we are learning more and more is that we need to not separate these out and put them together, because when you follow the continuum these factors are indeed very important.
An old person that has a couple of chronic diseases could be just fine if they have a set of friends, or family around, because they’re feeling good. They get food, and they go out and eat at the right times. They get sleep. They are mentally doing well. Or, they could easily—a factor that we may not monitor—be lonely. That may lead them to have depression. That may lead them to not go to the grocery store. That may lead them to not have food. That may lead their blood sugar to go awry. That would lead them to the ER. Four other complications later, it could be a $70,000 event.
Simon London: It’s the social determinants of health.
Shubham Singhal: That is extremely important. But as the society ages, it also is simple things like loneliness. A whole lot of people do not have the family infrastructure that perhaps might have once existed. If they are alone, if their friend passed away and their friend circle now is smaller, all of those things can create a very predictable pattern of how that person’s health would deteriorate. How do we help solve for that? It could be as simple as somebody, in this case, an unpaid volunteer worker who comes and spends some time and plays games with them or reads poetry with them.
You could imagine interventions as simple as that, that would avoid the progression that we might see over time. Again, I come back a little bit to our ability to have the data and understand what’s going on is quite significant.
Penny Dash: I think European countries have a longer history of recognizing some of the wider socioeconomic determinants, and hence, perhaps a greater focus on less income inequality, a greater focus on education, a greater focus on social housing and so on and so forth, and also have, I think, gone a bit further than probably the US has in looking at things like social prescribing. Social prescribing is now quite a common thing in many countries in Europe, whereby a primary-care physician would prescribe a set of classes at the local gym, might prescribe for you to go on a walking holiday, for example.
Simon London: Before we run out of time, a big takeaway, again, as a layperson, from this conversation is there are potentially some amazing clinical-technology breakthroughs of which genetics is one. The application of artificial intelligence in clinical settings is another. I mean, all of those things are amazing and will help. But there is an awful lot of basic technology blocking and tackling around transparency, clinical efficiency, scheduling, a lot of these things which are much less glamorous and deserve an awful lot of attention. Again, this is almost irrespective of how you choose, as a country, to organize your health system. It doesn’t matter how many payers and providers you’ve got, a lot of these are common challenges. Again, the optimistic view may be there’s a lot of upside.
Penny Dash: Oh, enormous amount of upside. It could be so much better.
Shubham Singhal: Definitely.
Simon London: Well, Penny, Shubham, thank you so much.
Shubham Singhal: Thank you, sir.
Penny Dash: Thank you.
Simon London: And thanks, as always, to you, our listeners for tuning in to this episode of the McKinsey Podcast. If you want to learn more about health systems, services, medical technologies, and other aspects of this complicated puzzle, please do visit us at McKinsey.com.