Personalized Medicine and Public Good: 3 Critical Issues (Part 2 by Johnathan Ng)

This article is the second part in a two-part series by Dr Johnathan Ng of Epibone, on personalized medicine and public good.


Personalized Medicine and Public Good: 3 Issues that Must Be Tackled

In my first article, I gave an overview of personalized medicine: personalized disease-modifying drugs, autologous cell therapies and stem cell therapies.

It is easy to be swept away by the promises of precision medicine. In his speech on the Precision Medicine Initiative, then-President Obama made uplifting points, “And that’s the promise of precision medicine – delivering the right treatments, at the right time, every time to the right person… we want to have a nation in which the accidents and circumstances of our birth aren’t determining our fate, and therefore born with a particular disease or a particular genetic makeup that makes us more vulnerable to something; that that’s not our destiny, that’s not our fate – that we can remake it.” 

Indeed, the potential benefits are tremendous, but so are the risks: in the form of escalating medical bills, with unproven – or worse still – harmful treatments. In this article, I give Governments and Healthcare providers three areas to pay attention to when it comes to personalized medicine: Regulation,  Healthcare Finance, and democratizing the benefits of personalized medicine.

  1. New Pressures on Regulation

As new personalized treatment modalities emerge, regulators are facing increasing pressure to green-light interventions, even if clinical benefits are not clear – to provide patients with a chance to live.

A watershed case was the US Food and Drug Administration (FDA)’s ruling against a scientific advisory panel, in favor of patient advocacy groups to approve Exondys 51 marketed by Sarepta for treating Duchenne Muscular Dystrophy. Despite a majority (7 to 6) of the experts citing inadequate convincing clinical evidence, the FDA director greenlit the approval of Exondys 51 due to a lack of clinical alternatives. Many commentators felt that this case set a precedent for the approval of personalized medicine products based on surrogate endpoints without clinical benefits.

There are also many grey areas when it comes to regulating clinical trials. As with any emerging technology, the benefits come at a risk, which people desperate for a cure may be willing to take. Due to the exploratory nature of trials involving new treatment modalities, patient safety is often left in the hands of researchers: a simple search on shows nearly 6000 clinical studies involving stem cells, some of which have not been approved.

Some of these trials have resulted in debilitating consequences. For example, severe adverse effects, some resulting in death, turned the spotlight on Juno Therapeutics’ lead CAR-T cell therapy for treating adults with late stage acute lymphoblastic leukemia. The risks of stem cell therapy are also not well understood. Although treatment with autologous fat-derived stem cells has been used for various indications, a poorly administered trial recently led to permanent eye damage in three elderly patients with macular degeneration (damage to parts of the retina in one’s eye).

Without proper regulation and well-controlled clinical trials, the safety and efficacy of stem cell treatments cannot be determined.

Some researchers show more caution than others. In the first human trial that uses an induced pluripotent stem cells (iPSCs) derivative, investigators from Japan successfully treated macular degeneration  by administering retinal pigment epithelial cells grown from the patients’ own stem cells. Yet, after identifying a few mutations in the second patient’s cells, the RIKEN group decided to suspend the trial in September 2015 before obtaining clearance from health authorities in Japan to resume in February 2017. Perhaps all scientists and clinicians would do well to hold themselves to a similar standard.

Regulatory bodies such as the FDA must continually engage and balance the needs of the scientific, patients, and clinical communities in meeting these new regulatory challenges – unfortunately, there are no easy answers.

2. New Pressures on Healthcare Finance

With the flood of new interventions, another issue to consider is cost. If all interventions are fully reimbursed (i.e. paid for) by state and private payers, the healthcare system will soon become bankrupt. Yet, if no help is given, the cost to patients of living longer is bankruptcy. The American Society for Clinical Oncology (ASCO) wrote in a brief that a patient living with cancer is now three times more likely to file for bankruptcy than a healthy person.

Policymakers must strike a fine balance of curbing the rapid rise in healthcare spending without disincentivizing innovation and depriving patients of access to life saving treatments.

When weighing the clinical benefits of a new drug product with the cost, healthcare economists typically apply a measure called the incremental cost effectiveness ratio (ICER) which takes the difference in cost between the new drug and existing alternatives and divides it by the change in quality adjusted life years (QALY). The National Institute for Health and Clinical Excellence (NICE) of the U.K., for example, sets an ICER limit of £30,000 per QALY gained for new drugs including targeted therapies. Most policymakers in the U.S. generally apply an ICER limit of USD$50,000 per QALY gained.

Early evidence suggests that personalized medicine tests are generally cost effective, with 20% of them resulting in cost saving and more than half achieving ICER of less than $50,000 per QALY gained. However, measures of cost effectiveness apply a single threshold to a heterogeneous population. If reimbursement was based on this alone, some people would receive more healthcare than they would choose, and others less. As such, commentators have noted that “reimbursement mechanisms for targeted therapies are still very blunt in an era of personalized medicine”.

Policymakers must leverage data and work with other stakeholders to improve reimbursement policies, especially taking into consideration the underserved population. Yet, the onus does not belong to the policymakers alone. Drugmakers, payers and clinicians are very much involved in the determining how drugs are priced and reimbursed. Recently, there have been exhortations by clinicians for more value-based pricing whereby reimbursement is contingent upon patient outcomes. The focus on outcomes could ensure that personalized medicine realizes its full clinical value. To achieve that, drugmakers could enter risk-sharing agreements with payers for partial reimbursement prior to demonstrating clinical effectiveness.

Alternatively, clinicians can also exert pricing pressure on drugmakers indirectly. In 2012, researchers from Memorial Sloan Kettering Cancer Center (MSK) evaluated the drugs Zaltrap and Avastin for treating colorectal cancer. Although Zaltrap cost twice as much as Avastin, the MSK researchers found no differences in efficacy between the drugs. Consequently, MSK decided to not recommend Zaltrap to patients and this led the drug’s co-marketer Sanofi to drop the price.

Together, stakeholders can work to ensure that personalized medicine is conscionable and cost effective.

3. Democratizing the impacts of Personalised Medicine

Finally, perhaps personalized medicine should be about more than just the diagnosis and the cure. Personalized medicine could go a long way towards disease prevention and mitigation by engaging the laymen and teaching them to monitor and manage their own health. In a recent trip with my parents to their dental appointment at a polyclinic in Singapore, I could not help but notice that the Health Promotion Board set up a booth that encouraged senior citizens to get screened for colorectal cancer. Participants were instructed to fill out their information, collect samples of their stool at home in the kits provided, and send the kits back for analyses. Though seemingly mundane, campaigns like this are probably the most effective way of bringing personalized medicine to the masses.

Low cost point-of-care diagnostics can also help to bridge the divide between first world medicine and third world need for solutions. After all, if the goal of personalized medicine is to understand and improve lives, esoteric treatments will hardly do a majority of the public any good.

Finally, it is a positive development that countries are thinking about how to democratize the benefits of personalized medicine. For example, the U.S. National Institute of Health is collecting data from underserved populations that are historically underrepresented in biomedical research, so that they too can benefit from personalized medicine.

In conclusion

We have already seen the good that personalized medicine can do. Yet, if we want the broader public to benefit from personalized medicine while minimizing both the financial and clinical risks to society and patients, there is still so much more that we must do. Stakeholders must continue working together to advance the personalization of medicine, not for fame or fortune, but for the greater good.

Johnathan Ng
Thanks, Johnny!


Thanks for reading this two-part series on personalized medicine and public good. When I started this blog, one objective was to use it as platform for issue-experts in technology fields to give us mini crash-courses, and to sketch out the implications for society. I am sure Johnathan will be more than happy to discuss these issues further. Let me know if you’d like to be connected!


A Crash Course on Personalized Medicine: (Part 1 by Johnathan Ng)

This week, I have my dear friend and brilliant scientist, Dr Johnathan Ng, PhD in Biomedical Engineering at Columbia University, give us a crash course on personalized medicine – what is it? what does it mean for the field of medicine and society? Johnathan now works at a start up in NYC, which grows bone for facial reconstruction. This is a two-part series, with the first part focusing on an overview of personalized medicine, and the the second part focusing on implications for healthcare systems and Governments. A super educational read for anyone with interest in the healthcare space. I certainly learned a lot while editing. Enjoy!


It’s getting personal

Personalized medicine has been hailed as the future of healthcare. At the forefront of clinical and scientific debate lie questions that could transform our healthcare landscape. Can medicine truly be personalized? Will the “personalized medicine” of today simply be medicine in the future? How can we leverage the personalization of medicine for the betterment of humanity?

A Brief History of Personalized Medicine

First, what is personalized medicine? In contrast to conventional medicine, which applies statistical information taken of the general population to the individual, personalized medicine uses information about a person’s genes, proteins and environment to prevent, diagnose and treat diseases.

As the prescient Hippocrates once said, “It’s far more important to know what person the disease has than what disease the person has.” The roots of personalized medicine predate our understanding of the human genome. For example, blood type matching for transfusion between the donor and the recipient to prevent hemolysis due to incompatibility was first reported more than century ago.

However, it is only with recent advances in genome sequencing technology that we can map the human genome and study it at an unprecedented scale. This has ushered in a new era of personalized medicine. In this overview, I cover three aspects of personalized medicine: 1. Personalized disease modifying drugs; 2. Autologous cell therapies; and 3. Stem cell therapies.

  1. Personalized Disease-Modifying Drugs

Some of the earliest breakthroughs in personalized medicine came in the form of personalized disease-modifying drugs, including:

  1. Breast cancer. In 1998, researchers found that a particular type of protein, HER2, was overexpressed in aggressive breast cancer cases. Consequently, Herceptin, an antibody therapy which suppresses HER2 activity, and a companion diagnostic test for HER2 expression in breast cancer cells were approved. These have become standard treatments today.
  2. Cystic Fibrosis. In 2012, the U.S. Food and Drug Administration (FDA) approved Kalydeco, a drug for treating cystic fibrosis by restoring the function of a protein misfolded due to mutation of the G551D gene. Restoring this protein’s function abolishes mucus buildup that leads to life-threatening respiratory and digestive problems. With that approval, Kalydeco also became the first drug that treats the underlying cause of the disease and not the symptoms.
  3. Immunotherapy. Over the last two years, a new class of antibody called checkpoint inhibitors was approved for treating some cancers. Opdivo and Keytruda are antibodies that disable checkpoints in immune cells by neutralizing the programmed death receptor (PD-1). Thus, immune cells bypassing these checkpoints are able to kill cancer cells more effectively. In a recent pivotal study, Merck showed that Keytruda reduced the risk of death by 40% among patients expressing PD-L1 levels greater than 50%.

2. Giving our cells superpowers: Autologous Cell Therapies

Besides harnessing information encoded in our genes to improve treatment response, personalized medicine is also about helping to unleash the immense capacity of our body to repair and mend itself.

Our cells contain information, latent or potent, that can be manifested into cure. Autologous cell therapy involves harvesting cells from a patient’s body, enriching the cell population outside of the body, and re-infusing the cells into the body.

The New York Times documented the miraculous journey of Celine Ryan who enrolled in a revolutionary clinical trial for her advanced colon cancer. Inherent in our immune system is the ability of lymphocytes[1] to locate, infiltrate and kill tumors. However, some tumors grow to counteract our immune response by damping or evading it. To help Ms Ryan overcome her tumors, the doctors mined her lymphocytes from the tumors, enriched and re-infused them into her body. These enriched lymphocytes intensified their attack on the tumors and after 9 months, she gradually recovered and entered full remission.

The success of Ms Ryan’s clinical trial provided scientists with a new strategy: engineering and enhancing patients’ T-cells to target and destroy tumor cells with distinct markers. These engineered cells are known as chimeric antigen receptor (CAR) T-cells, and they have both the ability to locate and destroy their targets. Emma Whitehead, 6, suffered from acute lymphoblastic leukemia (ALL) and twice relapsed from chemotherapy treatment. Without any other resort, her parents turned to an experimental treatment which used CAR T-cells to target CD-19, a marker expressed by both her healthy and malignant B-cells. The doctors rescued Emma from the brink of death and she is now cancer free.

3. Stem Cell Therapies

Stem cells are unspecialized cells with the ability to renew and differentiate into specialized cell types that make up our entire body during development. Even in adulthood, stem cells exist in multiple places in the body such as the bone marrow and fat tissues. Not surprisingly, they have also been heralded as a frontier for personalized medicine. At the biotechnology startup where I work, we engineer bone from a patient’s fat-derived stem cells to replace bone where it is needed. We successfully engineered bone from fat-derived stem cells and used it to regenerate a pig’s missing facial bone. Our next goal is to get the product into the clinic to help patients suffering from bone defect. There is reason to be optimistic: skin, trachea and bladder engineered from patients’ cells have already been successfully implanted.

Some key limitations remain in stem cell therapy as adult stem cells have a limited range of differentiation. Although embryonic stem cells are pluripotent (meaning that they can differentiate into any cell type), there are ethical limitations to using them as they require the sacrifice of embryos.

To overcome these limitations, Dr. Shinya Yamanaka and colleagues discovered a method to induce adult somatic cells into a pluripotent state. These cells, termed induced pluripotent stem cells (iPSCs), have ignited the imagination of scientists and clinicians as they could enable the treatment of diseases caused by the failure of specialized cells such Parkinson’s disease and heart failure. In a recent interview, Dr. Yamanaka (now a Nobel laureate) confirmed that clinical trials for iPSCs therapy will be underway over the next decade. However, he also cautioned against overstating the benefits of targeted stem cell therapies as they can only address a small subset of all human diseases.

What does personalized medicine mean for society?

Personalized medicine is improving the precision and efficacy of treatments by enabling the clinicians to make more well-informed decisions. Advances in pharmacogenomics have helped to reduce wastage of drugs and their incurred cost due to non-responders, and tailor the dosage according to the patient’s metabolism.

However, these efforts are not without cost. The cost of developing targeted therapies in an era of precision medicine is almost $2.6 billion. These treatments also bring about new regulatory risks for hospitals and Governments, who are facing increasing pressure to green light advances that give people unprecedented (but perhaps unproven) hope. My next article elaborates on three areas that Governments and healthcare systems need to pay attention to when it comes to personalized medicine, to maximise its benefit to public good.

Stay tuned!


[1] Tumour-infiltrating Lymphocytes

Tech for Health – Too Little Too Late (Alam Kasenally)

This week, I feel lucky to have a veteran in healthcare technology and data science from the Silicon Valley, Alam Kasenally, give us an overview on how technology has already transformed healthcare, and the gaping hole which has yet to be filled: patient experience.

Alam recently moved to Mauritius with his wife, Min Xuan (one of Singapore’s brightest entrepreneurs), where they manage a hospital. They’re also busy inspiring youth toward entrepreneurship, and building an innovation hub in Mauritius. Prior to his relocation, Alam worked in Cancer Commons in the Bay Area, which provides patients and their physicians with the knowledge needed to select the best available therapies and trials, and to continuously update that knowledge based on each patient’s response. He also worked in Oracle, Yahoo and Crowdcast prior.

Alam and Min are two people who will inspire you with their commitment to using tech and innovation for public good, how deeply their invest in others, and their entrepreneurial experience. For our entrepreneur readers: If you are a founder trying to gain quick access to real users and customers to pilot quickly, their hospital in Mauritius provides an immediate incubator for medical technologies, while they also have trusted partners especially in the agriculture and tourism verticals that can move quickly. Let me know if you’d like to be in touch!


What’s the opportunity for Tech in Health?

Early Sunday morning, and that ridiculously healthy neighbor of mine is already lunging and squatting on my, I mean our lawn, ready for her half-marathon practice. My overwhelming instinct is to grab my fitbit and try and compete, but really, I should be (from an economist’s point of view) happy that I have an additional neighbor in my community that is healthy. For a start, the workforce is larger by one (and perhaps more than one: serious illnesses affect entire families of people who care for the patient). I enjoy a larger share of tax dollars deployed to Leslie Knope rather than, uh Gregory House. Finally, my neighbor is probably not lunging with a dripping nose. My neighborhood is safer.

So, now we’ve established that Health is a Public Good (as well as being “in the Public Good”), is there therefore a role for technology in Health? Well, there already is, and let’s take a tour of the landscape.

Technology, and I’ll focus on tech as the Valley knows it, minus traditional medical technology (prosthetics, diagnostic and treatment infrastructure, etc) has made a serious impact in the last 10 years. Smart entrepreneurs everywhere have caught on that tech can:

  1. Lower overall costs through automation and efficiency (Epic, the Goliath of the industry, now faces an impressive challenge led by lean startups)
  2. Lower overall costs through the finding of patterns in hospital big data
  3. Avoid adverse selection and moral hazard through finding of patterns in insurance data and monitoring patient behavior (though I have to yet to see these cost-savings trickle down to patients)
  4. Provide a variety of “quantified self” (steps, sleep, calorie, breath) in an effort to influence behavior change and lower their healthcare system’s cost
  5. Lower overall costs through remote monitoring of patients (FBS, SPO2, EKG) and we’re even seeing these devices cross into the consumer space.

So is there any scope for the use of technology left?

There’s scope for a combination of technology, process, regulation and people. Healthcare is not only an expensive good, but a remarkably complex one. Sleep, Dieting, Breathing are just fine (though they have attracted the most VC dollars, as they are the easiest to do). The patient experience, on the other hand, is broken, in a million pieces. The complexity of choosing the institution and doctor that will lead to the best (and most consistent) outcome is daunting. The complexity of referrals is mind-boggling and reimbursement is ludicrous. The simple knowledge of viable treatment options and associated outcomes is not available to patients, their families and even doctors.

Is this the limit of the use of technology for Health? No, it’s only the beginning. It’s time for a true Uber of Healthcare to emerge.

“Huber” re-invents the patient experience just like Uber successfully re-invented the taxi experience. This company (and maybe government) will successfully join different partners and datasets, to create an experience that is to the patient’s (and her family’s) satisfaction, safety and in her interests. Datasets that only get smarter, as Healthcare outcomes, treatment models and patient preference filter back into the system. From my experience, certain countries remain crippled in regulation that thwart such efforts, often with the reasonable but ironic pretext of patient privacy. But others (Singapore comes to mind) have an honest broker, trustworthy IT custodian of the data and could write regulation and create necessary conditions that could well be in the patient’s interest.

Soon, participants will begin to realize that it isn’t just Health that’s the Public Good. But Data. Now, excuse me while I grab my fitbit.

You’re Probably a Cyberchondriac, But Google will Help You.

Have you worried that your headaches are the result of a brain tumour, or that your child’s leg pain is caused by cancer? You’re not alone. You may well be a cyberchondriac: “a person who compulsively searches the Internet for information on real or imagined symptoms of illness.”  If this sound familiar, you are in good company.


If you search “child leg pain”, google will auto-complete your search with “leukemia” – not because it is the most likely cause of your child’s leg pain, but because people who have searched “child leg pain” in the past were most likely to have clicked on links correlating this phrase with leukemia (probably because they wanted to understand the worst-case scenario). That’s how machine learning works – it pushes up the article that was most popular among other readers.

It makes sense to push up an article that most previous users clicked on – this is one of the best proxies for relevance to new users. However, the engineers behind search engines realise this isn’t necessarily beneficial for google users:

  • It’s scary – the average reader may assume cancer is the most common cause of child leg pain, or brain tumours are a common reason for headaches. Cyberchondriacs get even more paranoid.
  • It can encourage harmful behaviour – imagine if you search “best way to kill myself” and the top hits documented in detail the most painless way to die. Will the information push you over the edge in your decision?

Engineers behind search engines have to make a choice on what information to present to users – what people want (the traditional way) versus what they may need. 

The Making of “Dr Google” 

It was my pleasure to have Evgeniy Gabrilovich, Senior Staff Research Scientist working on health-related searches at Google, shed light on how Google thinks about it’s responsibilities to users. Evgeniy is addressing a sizeable group of Google’s customers. 5% of all google searches are health-related, 20% of which are people who type in a symptom hoping to find a cause.

Evgeniy’s team works on The Health Knowledge Graph, which aims to give users the best facts when they enter their symptoms.  The Health Knowledge Graph does not replace traditional web search, it complements it. Try it out: Type in “chest pain”, “depressed” or “child leg pain” and you will get a side bar on the right which covers the ranked list of likely conditions, how common or critical the condition is, incidence by age group, etc. The center section still presents traditional web-search results.

Screen Shot 2017-03-15 at 12.41.02 pm.png


When you type in a symptom you’re experiencing “child leg pain“, Evegeniy’s team aims to give you the most accurate diagnosis while minimising cyberchondria “Growing pains”.

Google realised that they didn’t have the expertise to do this on their own. It’s a huge technical challenge because of the large number of conditions and symptoms, and the overlaps between them. Furthermore, people use colloquial language to describe their symptoms, which the machine needs to decipher. Finally, user intent is often unclear. For example, if someone types in “weight loss” – are they trying to lose weight? Are they describing a side effect of medication?

Together with doctors from Harvard Medical School and the Mayo Clinic, they used machine learning to establish correlations between symptoms, conditions and treatments such that when you type in your symptom, you will get information that closely mirrors what a doctor might tell you (although it doesn’t go so far as to diagnose you… yet). Just to make sure, every result is evaluated by real doctors, who are asked “would you be comfortable with google showing these results”? 

What does this mean for the medical profession? 

Fifteen years ago, very few would have trusted medical advice that wasn’t from a doctor. Ten years ago, people started turning to the search engines for advice it wasn’t ready to give. Now, search engines are training themselves to give professional medical advice. They will only get better.

What’s next? I recently met a start-up, Mendel Health, which automates matching cancer patients to clinical trials through personal medical history and genetic analysis. Founder Karim Galil was previously a medical doctor. He was motivated by the fact that a single doctor’s brain cannot capture all information about diseases, possible treatments and clinical trials. He had patients die because he, as their doctor, was not aware of a clinical trial that could have saved their life.

Let’s take Karim’s idea a step further – suppose all my genetic, medical information and daily physical conditions (heart rate, glucose levels…) are constantly updated in a database that is linked to all potential interventions, treatments and medications.

  • While I am healthy, I can be alerted to risk factors and preventative actions (for example – you have a 50% chance of becoming diabetic in the next year. If you do X, Y and Z, the probability drops to 20%).
  • When I am ill, I can understand all my treatment options and the probability of success.

When a machine can diagnose me and recommend potential treatments, what will be the role of my doctor?

  • Much of what a primary care doctor does – assessing my condition, referring me to other specialists or recommending basic medications – can be encoded in software and search engines. Will they simply be a ‘stamp of approval’ – a safety blanket of sorts – before I take my next steps to get treatment?
  • Perhaps new roles for doctors will open up – for example, in training and verifying Dr Google as more and more people rely on it.
  • Complex surgical procedures will likely still require human attention. However, with robotic technologies like Verb Surgical, which enable top surgical expertise to be propagated across many doctors, will the average level of surgical skill required by each doctor be lower than before?

Why does this matter?

I honestly can’t envision a world with no doctors. Health is so close to our hearts that it requires a personal and emotional touch. However, it is important to understand how technology will change the role of the doctor:

This this will have large impacts on how countries train doctors (e.g. how long? what skills?),  allocate resources (e.g. primary care vs specialists), and design incentives in their healthcare system (e.g. if patients have access to so much information, will there be a trend towards over-consumption of medical services? Do co-payments have to change?). 

I am certainly not an expert in the field of medicine or medical technology, but would like to continue exploring this topic – especially from the perspective of what countries need to know, and how they should respond. Ping me if you are a doctor / work in healthcare and medical technology – I would love to hear your thoughts.


Better Consumer Access AND System-level Sustainability: Can Cities Have Both?


This week I read three parallel articles: one on healthcare, two on transport, all with the same theme: how the introduction of disruptive technology in traditional ‘public services’ led to a flood of new demand, calling sustainability into question.

I’ve thus far painted a positive picture of how new technologies can democratize access to services: Riding in the comfort of a private vehicle is no longer restricted to those who have money to own a car. Tele-health, where patients can consult their doctors online rather than face-to-face, is cheaper and more accessible than a traditional doctor’s visit, cutting down unnecessary waiting and travelling time (issues that disproportionately affect the poor and elderly!).

But improving access often leads to a surge in demand, creating new problems for society. These articles point towards an important trade-off between consumer access and system-level health that I haven’t quite addressed. [Spoiler alert: we should care about both because they are ultimately about the consumer!]


“The Downside of Ride-Hailing: New York City Gridlock” empirically shows how ride sharing has worsened congestion in NYC because many have replaced their subway rides with an Uber or Lyft. “Average travel speeds in the heart of Manhattan dropped to about 8.1 miles per hour last year, down about 12 percent from 2010”. New Yorkers have famously pushed back against their Mayor’s attempts to restrict the number of Uber cars.

“Autonomous Vehicles: Hype and Potential” shows how autonomous vehicles can also exacerbate traffic congestion, slowing down the movement of people and goods around the city.

  • One of the promises of autonomy is that the car can be re-imagined. IDEO imagined how cars might become work-spaces in the picture below. Once the car becomes a comfortable place to work or relax, many of us might not mind spending more time on the roads. I might opt for an Uberpool even if takes twice the time of a train journey because it’s such a comfortable, productive ride.
  • If these autonomous vehicles are privately owned, people might send their cars on trips they would normally take. For example, sending their car to the McDonald’s drive-through, or far out of the city center to find cheap parking.
  • We will also take some time to get to roads where vehicles are 100% autonomous. In the interim, human drivers are likely to “bully” autonomous vehicles because they know that these autonomous vehicles are programmed to be risk-averse (an autonomous vehicle killing a person is perceived as a greater travesty than a distracted driver killing a person). In such a scenario, we will see autonomous vehicles driving at slower-than optimal speeds, creating more congestion.

Source: IDEO

Autonomous work spaces


The parallel in the healthcare system is a study by RAND Corporation, showing how only 12% of tele-health visits have replaced visits to the doctor, while 88% represented new use of medical services. Unsurprisingly, this finding suggests that doctors’ visits are highly price-elastic – by halving the cost, we see a surge in new demand. Net annual spending on healthcare among patients with respiratory illnesses increased by US$45 per tele-health user.

This is a bigger problem if the new users actually didn’t need to see a doctor and a smaller one if they would have deteriorated if not for the medical treatment. The answer is likely somewhere in between – I believe closer to the former – 88% is huge (But a more in-depth study correlating the new use of medical services with health outcomes is needed). There is potential for tremendous waste in our already-stretched healthcare systems if we massively lower the cost of healthcare services without creating disincentives for unnecessary usage.

How can we get the best of both worlds: access and sustainability?

Technologies have amazing potential to help us use scarce resources like doctors’ time and road space more efficiently, creating greater supply. By lowering cost, they also ensure that this greater supply is spread out more evenly across the population, regardless of income.

However, doctors’ time and road space are ultimately still scarce resources that need to be rationed somehow. Capitalist countries are happy to ration these services by income. Countries on the socialist end of the spectrum (think the UK National Health System) tend to ration by waiting time. Neither fully takes into account the most important consideration: need and urgency.

How can we incentivize people to only use these new, accessible services only when they really need it? Here are some ideas.

In transportation

In transportation, cities need to make mass people-mover systems (trains, buses) the core service used by most commuters: ride-sharing must complement, not replace trains and buses. The bulk of commuters should spend most of their journey in trains and buses where the road space per commuter is significantly lower. Ride-sharing can be a first-mile and last-mile solution (e.g. home to train station), but certainly not the default for the whole journey.

To achieve this, cities need to up their game in public transportation. It has to at least be reliable and predictable (which many, many aren’t). Examples of how Singapore has done this here and here. Taking a step further, payments and arrival/departure times should be integrated with ride-sharing platforms so that people can minimize waiting and inconvenience when transiting between ride-sharing and public transportation. Work-friendly design in public transportation (think flip-out work tables in public buses) will also help make these options less unattractive compared to IDEO’s self-driving pods.

When it comes to autonomy, cities also need to think about moving to 100% autonomous vehicles as quickly as possible, since the dynamics between human drivers and autonomous cars will likely increase congestion. A 100% autonomous vehicle scenario also creates the most gains in efficiency and safety – vehicles can travel bumper to bumper (more efficient use of roads) and provide information to each other about road and traffic conditions (safety and efficiency are both enhanced). I cover some strategies in this article though this is a topic worth exploring in greater depth.

Finally, slightly more “interventionist” policies may be needed, such as limiting private-use autonomous vehicles and rationing the total number of cars dedicated to ride sharing so that people are prodded towards mass people-mover systems like trains and buses.

Tech companies sometimes paint these suggestions as the Government acting against the consumer interest. I disagree: it is in the commuter and patients’ interest if we can manage the demands on our roads and doctors such that those who need it most can get the services in an affordable and timely manner.

In healthcare

In healthcare, raising co-payments is a commonly-used tool which helps people think twice before using a service. “Triaging” patients is another way – for example, having them first speak to a nurse practitioner and only passing them to the doctors if it is needed.

But let’s take the patient’s perspective for a minute. What’s motivating them to use a service they may not need? Anxiety that their condition may be more serious than they think, and lack of a place to clarify (short of calling up a doctor). Any new parent empathizes with this. I probably went to the doctor every week in the first month of my daughter’s birth for no good reason at all.

We need solutions that assuage a patients’ anxiety. I believe equipping home caregivers is going to be a big part of this. Home caregiving is currently an informal sector with minimal training, which is an incredible waste. Imagine if home caregivers could be the first line of defence – giving the patient assurance when they do not need a doctor, and quickly helping them access a doctors’ time when it is urgent.

If healthcare systems and healthcare insurance providers want to use tele-health to optimise their use of resources, the technology has to be complemented by human-centred solutions that assuage patients’ anxiety. If not, the technology won’t save them any money at all!


I hope that with the addition of this article, I’ve now painted a fuller picture of the impact of disruptive technologies on public services like transportation and healthcare. Indeed, they will make resources more abundant and accessible to people with lower-incomes. However, complementary policies and services are absolutely necessary to ensure that the system is not over-used – ultimately, so that those who really need the services can get it in both a timely and affordable manner.


Two (Game-Changing) Ways Cities can use Technology to Fight Inequality

The story of income inequality is not new – as lower and middle-class incomes stagnate while the highest income brackets race ahead, the wealthy have access to goods and services that are increasingly out of the average person’s reach.

But we now see its detrimental effects more clearly than ever. I live in the Silicon Valley, and when news of Donald Trump’s election broke, the overwhelming feeling was disbelief. It was unimaginable. Tears of anguish were shed, yet a large part of the country celebrated. To me, that moment captured the deeper impact of inequality – fragmentation of society. Our politics become polarized, we are unable to find middle ground in our interests, and we increasingly feel like a nation of enemies, not countrymen.

While the problem gets more serious, our typical approaches to tackling inequality are reaching their limits. Redistribution is a political hot potato that pits the interests of the “haves” and “have-nots” against each other. Investing heavily in educational opportunities has diminishing marginal returns on social mobility both in the absolute sense (because the future of jobs is increasingly uncertain) and in the relative sense (because wealthier parents give their children more and more advantages).

We are in desperate need of new paradigms to fight inequality in cities. Here are two ways I believe technology can be a powerful, game-changing force – if deployed thoughtfully by cities.



First, cities should use technology to make life experiences in the city more and more independent of incomes.

 It would be impossible to close the income gap completely, short of communism. A society where incomes are totally equal is also undesirable, as it erodes the motivation to work.

However, I believe that technology can make life in the city increasingly independent of income, which can go a long way towards mitigating the daily experience of inequality.

Let me start with explaining the notion of an aspirational good – things that people wish they had money to buy. In transport, most people aspire towards owning a car. In housing, it is a condominium or a private home (American friends: as opposed to a publicly-built Housing Development Board apartment, which 80% of Singaporeans live in). In healthcare, it is a private doctor or hospital bed – at your choice and convenience. In education, it is getting into top schools and universities.

There is an unsustainable dynamic behind aspirational goods. Because these goods are limited in supply, the more people can afford it, the more expensive they get, and the further out of reach of the average citizen. Aspirational goods are the sources of a huge amount of angst in the middle class.

Technology has the potential to overturn the entire notion of an aspirational good. By creating new forms of value, it can make the alternatives so attractive that even those who have money choose not to buy the aspirational good. 

Take transportation for example. Owning a car is so attractive today because public transportation is an inferior option on many counts – the low cost cannot make up for its lack of time efficiency (it takes about twice the amount of time as a car ride), comfort (especially in humid weather), and customization (as a car owner, I know I can get a ride whenever I want).

What if public transport can be faster, more comfortable, more customized and cheaper than owning a car? With technology, this need not be a pipe dream. Imagine a day when you can wake up in the morning and your phone already knows where you need to be. It recommends the top three ways to get there. You select one, and within a minute, your ride shows up at your door – perhaps a shared car, or an electric bike if it’s sunny. It gets you to the train station just as your train pulls in. When you get out of the train, your minibus has just arrived to take you to the office. After work, you can summon a sleek designer vehicle for your dinner date. On the weekend, an autonomous jeep shows up at your door-step to take your family around for a day of fun.

You don’t need to buy multiple tickets – everything is paid through your phone. Or, you can even pay for transport just like a Netflix or Amazon Prime Subscription: a flat fee for unlimited rides. You never need to worry about parking again. With alternatives like this, how many people would still want to own a personal car? Even the wealthy may reconsider, especially if we simultaneously put in policies to make driving more inconvenient, such as no-drive zones in the city.

Just as technology brings about new forms of value (e.g. customization, flexibility) for those who don’t own a car, how can it do the same for other sectors?

  • How can technology help to transform Singapore’s public housing estates such that they offer new forms of value which private estates cannot provide? For example, how can we help HDB dwellers feel like the entire estate – with all its facilities and open spaces – is their home, one much bigger and diverse than any private estate? Digital communities and intra-town transportation may be aspects of this.
  • How can technology make a face-to-face doctors’ appointment something that people no longer seek as the “premium option”, for example, by making tele-health so attractive and pervasive?

I believe if domain experts and technologists put their minds to this, they will be able to come up with much better ideas than these! In short, technology can help catapult currently “inferior” options to equal status as “aspirational” options by creating new forms of value.

2. Second, cities should use technology to distribute scarce land and human resources more equitably.

In most countries, there is a healthy debate on how progressive and equitable the tax and redistribution regime is. However, not as much attention is paid to how other scarce city resources – land and manpower – are used. These too, must be used equitably, and technology can help cities achieve this.


Reducing the land used on roads is a great example of how we can use land more equitably. Roads and parking lots tend to be utilized disproportionately by those who own cars, who – in Singapore – tend to be wealthier. Can we cut down on roads and parking, and reallocate this land to purposes such as community facilities and public housing, which benefit a wider proportion of the population?

Yes, and technology is critical to this. How much land we need for roads and parking is determined by the concept of “peak demand” – the maximum number of vehicles on the road, ever. We can cut down peak demand by encouraging people to use shared mobility options rather than drive a private car (I write about how tech enables this here), and by investing in autonomous freight and utility so that these activities can be done at night, when roads are far emptier.

Public Sector Manpower

Similarly, we can use public sector manpower more equitably by investing in technology. Technology can significantly reduce the manpower we commit to customer services. For example, Govtech rolled out MyInfo, which enables citizens to automatically fill in their administrative information for Government schemes with the click of a button. Chatbots on Government websites will increasingly be able to answer public queries; phone lines will no longer be needed. Public sector manpower can now be dedicated to functions which are in great need of resources. One such area is social work and education. Families in the bottom rung of society often face a cocktail of challenges – divorce, low-income, lack of stable employment, cycles of incarceration and so on. Giving them (or their children) a real chance of breaking out involves an extremely high level of hand-holding and investment by social workers and schools. Resources are sorely needed here.

Access to top quality healthcare

Let’s take another scarce resource – top surgeons. People who can pay for their services access better quality care, and stand a higher chance at recovery. Technology can change this dynamic. Companies like Verb Surgical are using machine learning to propagate top surgeons’ expertise more widely. This is how it works: every time the best surgeons perform a procedure, every single action is recorded in a common machine “brain”. The “brain” is trained to associate each action with the probability of a successful surgery. As the “brain” records more and more surgeries, it gets smarter and smarter. Now, the “brain” is made accessible to ALL surgeons. At each step of their surgery, they are told what successful surgeons did. Now, the best surgical expertise is within the reach of the average citizen.

Technology that enables our scarce resources (e.g. land, public sector manpower and top surgeons) to benefit the broad population and serve those in acute need are the types of technologies that cities should invest in, and quickly enable through regulations.


If you google the “Smart City” movement, you’ll find many broad and loose definitions. Generally, it refers to how cities deploy technology to improve city life and allocate resources more efficiently, whether it is helping transport systems run more efficiently, making interactions with various Government services easier, or to adding fun to the city experience.

Unfortunately, such broad and loose definitions give cities little guidance to on what to focus on in prioritising investments and regulatory reform, which is an incredibly important conversation given the limited resources at most cities’ disposal. It also does not paint a compelling vision for why being a Smart City matters, which disengages most of the population. Personally, before I worked in tech, I felt absolutely no connection to the idea of a ‘Smart City’. Tech was cool, but I never thought it was crucial.

I believe that using technology to tackle inequality and its effects should be a Smart City’s ambitious goal and guiding force, providing focus and rallying support from its constituents. This article spelled out two ways to do so.

Disrupting Elderly Caregiving (and why Uberisation won’t work)

I came across a really encouraging article about tech start-ups that are trying to fix the elderly caregiving sector. This is incredibly important work.

The double whammy is here 

The graphic below, from the UN, shows how our global population is aging. It will happen faster in developed countries. In Japan people above 65 already constitute more than a quarter of the population. Singapore will get there in the 2030s, and the U.S. in the 2040s.


This will be a double whammy on societies. Countries will have a shrinking tax base to support the growing number of elderly who need care. There won’t be abundant resources to build new nursing homes and hospital beds. At the same time, it will be more difficult for the elderly to continue living at home – with shrinking family sizes, the responsibility for caregiving will fall on one or two children, instead of three or four.

Advances in biotechnology, personalised medicine and genomics will go a long way towards mitigating these challenges. For example, if companies like Calico and Unity Biotechnology manage to reverse aging, people can have longer, healthier, independent lives. The periods of time when they need caregiving will plummet. This is referred to as a longer “healthspan” (as opposed to just “lifespan”).

Unfortunately, many developed countries are already starting to face the double whammy and need a more immediate solution. From a policy perspective, it does not make sense to build new infrastructure for the elderly because these will become redundant when subsequent elderly cohorts are smaller.

The best solution is to help people receive care in their own homes as they age. It is also what a large majority (some statistics suggest 90%) of people prefer.

Why is Home Caregiving such a tough nut to crack? 

A fragmented home care sector with low standards and revolving door caregivers is the norm in many cities. I don’t see a level of innovation and investment flowing into this sector that is commensurate to its importance. Three likely reasons why this is such a tough nut to crack.

  • Home caregiving is highly personal. When someone takes an Uber ride, they don’t have to talk to the driver if they don’t want to. In contrast, a caregiver gets deep into your personal space if they bathe, feed and change your diaper. You’re obviously going to be very picky about who you choose. Pickiness is not a one-way street – caregivers have their preferences too!
  •  The job is just not a lot of fun. In fact, that’s why many people out-source care to paid contractors, even care for those they love most. It’s not glamorous, it’s repetitive, it can be highly physical, and in some cases, you interact with suffering on a daily basis. For such a difficult job, you mostly work alone without a network of support and accountability. Benefits, training and wages are patchy. Motivation is low. It is more difficult to recruit a good part-time Caregiver than a good part-time Uber driver.

What are the implications for tech companies and smart cities?

Disrupting the flailing home care sector is a classic problem that the tech, business, Government and non-profit sectors need to go at together. Here are three things that anyone working in this space needs to consider.

  1. Tackle the root problem – motivation.  

The FT article features Josh Bruno,who left Bain Capital to set up Hometeam. He started with a tech solution in mind – perhaps the equivalent of Uber for in-home care, he thought – a platform that would match caregivers to elderly and take a cut.

After volunteering in about 40 elderly care organisations, he quickly realised that the deepest issue in the caregiving sector was not matching, but motivation – “People blame them as lazy, but they are the worst working conditions — low pay, no training.” Defying all industry standards, he decided to hire the caregivers, give them good benefits  of paid holidays, maternity leave, training and retirement.

I think Josh is getting to the heart of the issue – how can you expect people to deliver quality care if you don’t care for them? The first step is turn caregiving into a worthy career. Business can’t tackle this alone, which leads me to point two.

2. Partner to make home care an economically viable business 

Josh’s company was turned away by many VCs because hiring the caregivers was not deemed to be maximally profitable.

I believe many  have not yet realized the economic benefits of a strong home care sector. When elderly are able to continue living in their own homes, a whole ecosystem of services such as tele-health, food delivery and transport has a larger market. For large healthcare providers like Kaiser Permanente, a strong home care sector combined with tele-health can help manage its elderly clients outside the expensive hospital setting. This optimizes the use of hospital facilities and doctors’ time, making the business more profitable. For the same reasons, countries where healthcare systems are operated by the Government should consider investing public funds in the home care space, for example, by extending subsidies in nursing homes to the home care sector.

Furthermore, to help the home care sector grow, the public and non-profit sectors should play a role in helping to overcome inherent”diseconomies of scale”. It does not make sense for small home care companies to have their own training programmes and qualifications system. One idea is for Governments or non-profits to offer basic, modularised caregiver training as a “shared resource” for home care companies to utilise.

3. Use technology to build a care ecosystem around the elderly 

Although it may not currently feel like it, home caregivers are really part of a team of family members, healthcare professionals and community members who help make living at home comfortable and enjoyable.

Technology can enable better communication and support between different members of this team. The home caregiver should be able to ask the healthcare professional whether a situation requires medical attention, and share information that would help the healthcare professional with diagnosis and prescription. Community members should be able to volunteer in bite-sized chunks of time, a topic I wrote about in my article “Three ways to build a transportation system that services the vulnerable”. Ultimately, when members of the team communicate, they can collectively give better care and receive support from each other. This can help to tackle the motivation problem for caregivers.

Technology can be part of the care team too! It can enable an elderly person to live independently as long as possible, only activating their caregiving team when necessary. One of the greatest fears of an elderly person is that something happens (a fall, a stroke) when they are alone. The solution, 24-hour care, is expensive and intrusive.

The Singapore Management University has been piloting sensor-enabled homes in Singapore through their SHINESeniors initiative. These sensors observe and analyze the elderly’s living patterns and immediately activate his care team when a change in pattern suggests a deterioration. Besides enabling independent living, this also allows companies and family members to allocate their time and resources efficiently.


Disrupting the flailing home care sector is essential to the quality of life of older folk, and the sustainability of healthcare systems when populations age. It’s a tough nut to crack, but I can’t think of a more important issue to work on right now, because our parents – and then us – are going to be the beneficiaries of positive disruption!