Category Archives: Insights

How to fight crime with Machine Learning

Cyber Criminal

Businesses have to defend their environments from attacks, and security professionals are asked to accomplish impossible feats in the modern era of cyber defense: they have to protect users and critical information from unauthorized access. This is asymmetric warfare where the bad guys have to get it right only once, but the good guys have to get it right every time. Criminals can monetize stolen data fairly easily, and the criminal success rate has steadily improved over the past decade. As we’ve seen with political and cyber military attacks, money is not the only incentive.

To combat these problems, companies have armed themselves with a plethora of new security tools. As a result, those responsible for an organization’s security posture can be inundated with thousands of alerts — prioritizing and acting on these is a daunting task. A skilled security professional can do a great job when focusing on a specific investigation, but when that process requires stitching together the relevant pieces of information, humans need help extracting insights from an ocean of alerts and raw data coalesced across multiple security systems.

No company is immune to cyber criminal activity. In 2013, Target was hacked despite receiving as many as 10,000 security alerts per day. While Target is a Fortune 100 retailer, even medium-sized companies have to sift through hundreds of thousands of alerts each year. Alerts are investigated before being categorized as false positives and ultimately ignored, but most alerts are idiosyncratic to a product or application with little context of the overall business impact. To prevent financial and reputational loss, security teams are driven to find the most critical needles in an ever-growing haystack of security information.

The techno-elite companies like Facebook, Amazon, Netflix, Google, Apple and Microsoft (a.k.a. FANGAM) have successfully leveraged machine learning algorithms across their businesses that include security systems to protect their users, applications and the overall infrastructure. There are many definitions of machine learning (ML) but some would describe it as a type of artificial intelligence (AI) that provides computers with the ability to learn without being explicitly programmed.

Machine learning focuses on the development of computer programs that can teach themselves and develop innovative solutions when exposed to large quantities of data. Today, we are seeing machine learning based software tools exceed human intelligence in specific tasks within narrow disciplines.

Machine learning algorithms often start out as supervised by engineers and taught with labeled training sets. An example of supervised ML could be a training set with 100 critical security alerts where 20 are labeled as malicious activity and 80 are labeled as non-malicious. Based on the training set provided, the algorithms would attempt to determine malicious activity on new alerts that have not been investigated. However, one of the most intriguing classes of machine learning is “deep learning” where the software is unsupervised and must function with a self-learning approach to develop its own answers.

One example of deep learning would be to provide 10,000 critical alerts with a finite set of outcomes and no training set. The deep learning algorithm would determine its own grouping of the data. The more data the deep learning algorithm processes, the more accurate the algorithm could become at determining malicious and non-malicious activity.

Threat Hunting Starts with Your Data 
A business getting a cyber attack is a bit like a person getting sick. Everyone will eventually get sick, and when this happens, you want a quick and accurate diagnosis. You want access to the best medical care possible so that the sickness does not linger and lead to more serious problems. For a speedy recovery, you want to go to a doctor who is thorough and knowledgeable of the latest treatments, no matter how experimental. With some life threatening diseases, an experimental treatment may work better than the typical standard of care.

Machine learning for security is more like an experimental treatment because these algorithms aren’t deployed as standard practice in the industry yet. However, security teams need to care for their information systems in a manner similar to how we care for our health to limit and in the best case, prevent the damage that a criminal can do. Once a security breach is successfully executed, the challenge of discovery and incident response will occur along with the time-consuming and expensive task of cleanup and forensics analysis to understand what exactly had been compromised.

Threat discovery should always start with the data and being able to discern what pieces of information will lead you on the path to tracking down cyber criminal activity. Network and endpoint security tools like firewalls and antivirus programs generate scads of alerts and logs that describe when access to a protected system was blocked, allowed or flagged as a potential threat. Each event describes anomalous activity that does not conform to any normal or expected practice. If an alert is directly tied to a critical breach or an ex-filtration of sensitive data, then the security team becomes activated as Emergency 911 responders to that alert.

Very rarely will one alert illustrate a complete story around a major security attack. Generally, you need to assess dozens of alerts from several different systems across weeks or even months to triangulate a sophisticated attack. To add to the complexity of the process, security professionals need to review data from multiple systems that are stored in separate repositories.

Security professionals have to conduct what someone once described as “swivel chair analytics” and jump from console to dashboard to report to the command line before being able to determine that a cyber crime was committed. Reducing the need for “swivel chair analytics” is just one potential benefit of machine learning.

Keep Your Friends Close and Your Enemies Closer
While Sony Pictures had several defensive measures around their crown jewels of unreleased films and scripts, many other vectors were vulnerable for attack. The cyber criminals working with and for the North Korean government were undetected before whipping out Sony Picture’s IT infrastructure and releasing sensitive internal company emails that ended up ostracizing the top executives from their own industry.

This wasn’t a smash and grab, and these sophisticated criminals weren’t after money or the most valuable assets of the business — the screenplays and unreleased movies. North Korea’s primary objective was to embarrass and intimidate an enterprise. Mission accomplished. As a result, many information technology and business leaders reassessed their strategic security plans. This class of cyber crime warrants a new approach in detection and response that we’re starting to see with machine learning.

Information security can often be broken down into three broad categories: defense, detection and response. Companies can invest and deploy all the leading infosec tools available to create many layers of defense, but the kicker is that no matter how much money is invested in blocking attacks, the probability of never getting compromised is slim. This modern reality has forced chief information security officers (CISOs) to shift their investment balance toward improving their detection and incident response capabilities.

Today, companies need to defend themselves against advanced persistent threats (APTs) like what we saw with the Sony Pictures attack, which are often associated with a nation state actor that’s well funded by a military or government entity. An APT organization will often gain unauthorized access to their target through unexpected ways and remain undetected for long periods of time. It’s like a lurking alligator waiting to steal data rather than cause immediate damage to the network or organization.

APTs are carefully planned and rehearsed in advance to avoid detection. They’re able to stay in a system for months, if not years, by waiting in the shadows until they became a normal part of the environment. They slowly increase activity and then one day, the intent of the enemy is revealed. Yet, their stealthy movements appear hidden in the shadows of the data they leave behind. Machine learning has the ability to shine a light on the criminal footprints hidden from human sight.

Having The Right Staff Isn’t Enough
Hiring enough security professionals has become an industry-wide challenge for businesses of all sizes. In 2016, several reports have cited the number of unfilled security jobs in the U.S. at about 209,000 and globally, at about one million jobs. There is a real talent gap within security that continues to widen. For the lucky few companies that have ample staffing in their security ranks, finding cyber threats with previous-generation tools is like finding needles in an enormous haystack.

To compound the challenge, cyber threats that bypass the traditional layers of defense are not black or white signals, but rather low-grade grey signals that are difficult to make sense of. What machine learning can do is find the disparate needles in the haystack and thread them together. There can be dozens of different needles along a single attack thread created over a five-month period that tells you a bigger threat has taken hold within your environment. Machine learning is well suited to flag anomalous behaviors that span across users, partners, networks and infrastructure systems. This level of insight is worthy of a security professional’s time, knowledge, and skills.

These new machine learning algorithm techniques have already reduced the cost of security operations and threat-hunting investigations between $500,000 and $1 million each year for mid-size Global 2000 enterprises. Once these machine learning algorithms find the important needles in the haystack, the next evolution will be to employ AI assistants to take corrective action within a narrow set of tasks to help bridge the talent gap in security.

Raise Your Security IQ — Fight Smarter, Not Harder
Security teams are always on alert because a cyber criminal can take advantage of one minor mistake to gain an edge. Machine learning can be a powerful countermeasure provided there’s plenty of useful data to feed the algorithms. Machines never get tired, and these types of algorithms become more accurate as they process more data to refine their capabilities.

Self-learning machines that become smarter than humans in specific tasks represent the promise of reversing the decade-long negative trends in cyber defense. Business leaders and security teams need to start leveraging machine learning to stay one step ahead of adversaries that are constantly innovating on how to commit crimes.

This article first appeared on geek.ly

The post How to fight crime with Machine Learning appeared first on Doug Dooley.

Source: http://www.dougdooley.com

Running Through Walls: Technology Enables it and Culture Demands It

YouNow CEO and founder Adi Sideman talks with Venrock partner David Pakman about his dream of an interconnected and livecasted world, and how he made it a reality with the founding of YouNow. He shares his views on the one pure form of rich media creation, what it’s like to compete with Twitter and Facebook, and how his game development experience helps him to deliver a magical experience to customers. They also discuss how YouNow avoids common safety pitfalls of other social media sites and why New York is the perfect home for the company.

Adi Sideman is a pioneer in participatory media, with more than 20 years of experience creating apps and companies in the user-generated content space. Founded in 2011, YouNow’s mission is to create an interactive platform where anyone can participate and express themselves live.

Running Through Walls: Olympic Gold Medalist Turned CEO on Building a Business

Brent Lang, an Olympic Gold Medalist and CEO/President of Vocera, speaks with Venrock partner Brian Ascher about his journey from winning an Olympic gold in swimming during the 1988 Seoul Games to leading a public company. Brent shares lessons learned from competing in the Olympics and discusses his insights on becoming an effective leader and managing a board. He also reflects on maintaining Vocera’s strong company culture during times of transition, and what he’s learned about establishing a healthy work/life “harmony.”

Clara Lending: A Big Swing

Screen Shot 2016-08-17 at 7.12.02 AM

There are few markets larger or more important to the US economy than the consumer mortgage market, which consists of $1.5 trillion in annual originations. Or more emotionally important to consumers, for whom homes represent an opportunity to build stability, a family and a better life.

Or more structurally broken. As was made clear in 2008, the mortgage market is fragmented into tens of thousands of companies in many different layers — brokers, originators, servicers, securitizers, government sponsored enterprises — whose complex interactions add costs, skew incentives and obscure risks, sometimes with devastating results.  

If one were seeking to reimagine this industry from scratch, the core problem to solve is much simpler than all this complexity suggests. On one side of the market you have consumers seeking low-cost financing for their homes. On the other side, you have the U.S. government, which finances more than 70% of consumer mortgages through Fannie Mae, Freddie Mac and the Federal Housing Administration and sets clear variables for the qualified mortgages it will subsidize.  

Why can’t one build an online platform to sit between these two sides of the marketplace, bringing transparency, lower costs, integrated data and a delightful consumer experience? That is is the vision of Clara Lending, a recent investment we’ve made that represents a big swing by its founders in one of the most important consumer markets there is. Clara is not simply reimagining the front end of the consumer mortgage experience. It is reimagining the entire mortgage bank from the ground up with software and data.  

The founders know this market unusually well and are as motivated as much by the social good the company can do as they are by the economic opportunity it represents. Jeff Foster, Clara’s cofounder and CEO, served as a senior policy advisor at the US Treasury during the first term of the Obama Administration to help fix the mortgage market and understand where the core data and incentive problems were. Lukasz Strozek, Clara’s cofounder and Head of Product and Technology, was previously a senior technologist at Bridgewater Associates, the world’s largest hedge fund, where he focused on translating complex processes and risk analyses into software.

If Clara is successful, it will lower mortgage financing costs for consumers and bring transparency and trust to an industry that tends to lack both. It will also bring transparency and integrated data to the mortgage supply chain, reducing macroeconomic risk and providing regulators with a clearer view of the market. It is a company we believe can create enormous value and bring enormous social benefit, the kind of investment we are most eager to make.

Source: http://www.nickbeim.com

Running Through Walls: The Yoda of Genomics

John Stuelpnagel speaks with Venrock partner Bryan Roberts about his career transition from large animal veterinarian to entrepreneur, the keys to Illumina’s dominance in the industry, and the future of genomics. As the founder and board member of many successful genomics companies including Illumina, Ariosa Diagnostics, 10X Genomics and more, John shares his advice for the next generation of entrepreneurs and offers his perspective on the next frontier of innovations in genomics.

Running Through Walls: Dollar Shave Club CEO Michael Dubin on the Unilever acquisition

As an early investor in Dollar Shave Club, I spoke with CEO Michael Dubin about the company’s acquisition by Unilever and the journey to this point on the latest episode of Venrock’s podcast, Running Through Walls.

Many people remember the humorous viral video that first launched the company. “I was freaked out that maybe we wouldn’t recover from our success,” says Dubin, after the site crashed thanks to the video’s unanticipated popularity.

The company did recover, and Dubin went on to build one of the most recognizable brands in men’s grooming. He knew he had a hit when he visited a distribution center and saw the volume of packages on the conveyer belt, bringing to life how many people interact with the company daily. He says, “Three percent of Americans wake up and engage with Dollar Shave Club.”

Dubin took improv classes early in his career, and humor has long played a role in the culture of DSC. He’s even adapted some lessons from improv to the role of CEO: “When you’re on stage with no script, it teaches you to live in a scary moment and still perform, which is great training for a young company.”

Following the acquisition, Dubin will stay on as CEO, and it will be “business as usual” at DSC with a focus on launching new products and expanding internationally. Dubin cites “Unilever’s position as one of the most progressive and innovative CPG companies in the world” as the reason the company is a good fit for DSC.

Dubin hopes that five years down the road, “We’ve been able to meaningfully change the way people think about shopping on the internet.”

DOLLAR SHAVE CLUB: HOW MICHAEL DUBIN CREATED A MASSIVELY SUCCESSFUL COMPANY AND RE-DEFINED CPG

Success has many fathers,
but in this case, there is only one.

Slide 3 of Dollar Shave Club’s Series A Pitch Deck

 

When I first met Michael in June of 2012, I had already seen the DSC launch video like everyone else. I saw his site melt down and watched the video explode and actually go viral. I asked Peter Pham for an intro, which he graciously provided, as Science had seeded and helped incubate the company.

What Michael was creating fit a number of my investment theses. I believed (still do!), in the age of social media, brands must become direct-t0-consumer in order to know their own customers. Having run eMusic and a few other subscription businesses, I knew that subscription is a business model that only actually works for a select few product categories, and that churn rates must be very low (well under 5% monthly) in order for subscription businesses to succeed at scale. I believed it was possible to use asymmetric marketing to injure existing incumbents who overly depend on broadcast advertising and distribute only through retailers. When I saw DSC’s early numbers, I immediately knew they were on to something big. There were many well-known seed investors and large VC funds already on the cap table — I was sure there would be a fight. Strangely, none showed interest. With other investors circling, and with Michael’s blessing, I led Dollar Shave Club’s Series A round and we became partners.

Michael knew all these things too. On top of it, it was clear he had enormous ambition and intended to build a dominant men’s lifestyle brand that went far beyond razors. He intuitively understood how to use content and conversation as marketing at a time when legacy brands were still shouting at their customers with TV ads, purchased without actually knowing their customers. He believed in transparency, making great products, and putting convenience and value first. And he knew it was crucial to build a trusted and beloved brand, albeit one that is entirely direct-to-consumer. His plan was grand, but his formula was simple…

Slide 19 in Dollar Shave Club’s Series A Pitch Deck

 

Michael built an incredible team. Hiring Kevin Datoo as COO very early on was a brilliant move. The team hit (or beat) every number they ever put in front of the Board. Following Javier Hall’s creative direction led to enormously high conversion rates coupled with elegant design. Adam Weber’s marketing strategy helped propel DSC to be the very clear number two razor company in the U.S., second only to Gillette and light-years ahead of the many followers who entered the market later. Janet Song’s obsession with high-quality customer service became a hallmark of the brand.

Success was not always obvious. Despite growing from $7M in revenue in year one to $20M in revenue in year two, no new investor was willing to lead the Series B round. At Venrock, we had such conviction in the team and the formula, we happily led that round too, fortunately increasing our ownership.

From that point on, it was like riding on a rocket ship. We grew to $60M in revenue in year three, and more than $150M in year four. As new competitors entered, we outmaneuvered them. We were eating marketshare, quickly reaching more than 15% U.S. men’s razor cartridges share last year, and getting smarter. As the team introduced new products they designed themselves, our millions of customers happily adopted them. Our software and systems performed admirably, and we ingested a large amount of data every day about our customers and their usage, refining the service and our products. It was a pleasure to watch the company transition from a razor subscription service to a trusted men’s lifestyle brand, increasing margins each step of the way, and serving more than 3 million people.

Many investors shy away from commerce companies. The multiples tend to be low, Amazon is ever-present, and lots of capital can be required to scale. To us, we didn’t see DSC as an “e-commerce” company, but instead as a model for new full-stack consumer products companies. Our investment criteria for this space is as follows:

  • Offer highly-differentiated products with high product margins (In DSC’s case, value and convenience were the differentiators and their product margins are very high. Avoid product categories that can be Amazoned.)
  • Invest only in zero-sum markets (A customer buying your product means they stop buying your competitor’s products. This is clear for DSC, but often lacking in apparel categories, for instance.)
  • Choose categories where incumbents sell only through retailers and have no direct relationship with their actual customers
  • Choose categories where incumbents overly depend on broadcast advertising
  • Choose categories where the CEOs of the incumbents are professional CEOs, not founders (thus are far less-likely to cannibalize existing businesses and adopt new business models)
  • Look for products and services which gather usage data and utilize machine learning to improve over time

Seeing Unilever recognize the importance of Dollar Shave Club’s incredible success and the brilliance of their team is a wonderful outcome for all of us. DSC will now have enormous resources to compete globally and to attain Michael’s original vision. It was an incredible privilege to work with and learn from this team. In addition, the co-investors Michael assembled around the table were a joy to work with. Thank you to Mike Jones and Peter Pham at Science, Kirsten Green at Forerunner, Woody Marshall at TCV, Rick Prostko at Comcast, Marc Stad at Dragoneer and the many others (like Mark Levine) who pitched in to help.

But most importantly, thank you to Michael Dubin. You are one of the greatest CEOs I have ever seen operate. You deserve all the spoils of great success. As you always said, “Great things happen when your ass smells fantastic.”

Michael Dubin at his offices in LA on 7/15/16. michaellewisphotography.com

Michael Dubin at his offices in LA on 7/15/16.
michaellewisphotography.com

Source: http://www.pakman.com

How to Make Our Crazy, Expensive, Amazing, and Uneven Health Care System Better Faster

This post first appeared in NEJM Catalyst.

I am often asked by health care entrepreneurs, policymakers, health system leaders, and clinicians to “explain” how health care in America works — and to offer ideas for how to improve it. Here, within a whirlwind synthesis, is a provocative thought starter for how to make U.S. health care better faster.

The State of U.S. Health Care

The U.S. health system is bigger than the entire economy of France.

The United States spends about $3.2 trillion on health care. This equates to $8,500 per capita, which is twice what other developed countries spend per capita. While health care spending growth has slowed since the 2008 recession, it still exceeds GDP growth. Unique to the U.S. economy, health care experiences negative labor productivity. It is also the largest source of new jobs since the recession, despite flat demand.

We waste the equivalent of the entire health care system of Spain annually.

Many economists believe that about 30% of spending, or about $900 billion per year, is wasted in the U.S. health system. The largest primary driver of excessive spending is high prices rather than high utilization. Prices in the United States are about 60% higher than in other OECD countries. High prices are a result of a fee-for-service reimbursement system based on paying “cost plus,” coupled with the market power of providers and suppliers to raise prices at a 5% CAGR over the past 20 years. Supplier market power is further augmented by a lack of transparency of prices and quality, complexity, insurance benefits with limited cost sharing, supply-induced demand, structural local provider market power, and consolidation. Commercial insurance prices vary by 50%–400% for all services, in all markets, and are seldom correlated with quality.

We compete with Cuba and Costa Rica on quality.

The United States comes in 37th on international comparisons of quality, just behind Cuba and ahead of Costa Rica. Our life expectancy lags behind other OECD countries and is improving less quickly than others.

We only perform better on cancers and preterm babies. We diagnose cancer earlier thanks to more widespread screening programs that are very profitable to providers. We do not do better on chronic diseases. The United States is “less sick” than other OECD countries only because we are younger and have lower prevalence of smoking. We are the only OECD country without national insurance; in 2013, 50 million Americans were uninsured. Moreover, 18% of Medicare patients are readmitted to hospitals for the same condition within 30 days. About half of these readmissions are believed to be avoidable.

Just 10% of patients account for 65% of U.S. health care spending.

A cohort of 35 million Americans who suffer from multiple chronic diseases and spend $90,000 per year, on average, drives most U.S. health care spending. Most of these people are unable to work, poor, and covered by Medicare and Medicaid (referred to as dual eligibles). End-of-life spending accounts for only about 10% of spending. So while people may spend large amounts in hospitals in their last year of life, it is a relatively small contributor to overall spending since only about 2.6 million people die annually. Expensive biologic drugs, cancer drugs, high-tech medical devices, and new technologies collectively account for only 15% of U.S. spending.

Our health care market is massively fragmented.

Health care in America is subscale. Care is organized around roughly 500 disease categories and delivered by loosely organized collections of local providers and hospital systems. The largest hospital system in the country, HCA, accounts for only 5% of the market. While a bit more than half of the roughly 850,000 doctors are employed by local hospitals, those in private practice are in groups of 4–8 physicians on average.

As a result, most doctors care for several types of patients within their specialty, and virtually all hospitals are general hospitals with low volumes of many types of patients. The typical Medicare patient under treatment has 16 doctors involved in his or her care. Nevertheless, scale appears to be one of the most important predictors of quality, with doctors and hospitals performing higher volumes of similar cases delivering better outcomes. Other countries have fewer, larger, and more specialized hospitals, and specialty providers with narrower practice areas.

The Affordable Care Act is a big step forward.

The ACA was designed to expand coverage largely by redistributing about 5% of current Medicare spending to subsidize coverage for up to 30 million uninsured. Newly insured people would be covered by private health plans sold on exchanges or Medicaid (in states that expanded their programs). Most of the uninsured are poor, young, male, and comparatively healthy. The ACA has improved insurance market competitiveness for individuals by standardizing policies (for example, gold, silver, or bronze) and eliminating underwriting, so that all people are able to purchase insurance at prices that vary only by age.

The ACA has made minimal changes to employer-sponsored insurance plans; most Americans will continue to receive health insurance from their employers for the next decade. The ACA has also introduced new payment models containing incentives of varying strength for providers to improve quality and reduce costs. The transition from fee-for-service to alternative payment models is going to take at least a decade, as 90% of payments made in the United States today are still fee-for-service. The Congressional Budget Office expects the net effect of the ACA to be a small slowing of federal health care spending, about $300 billion of deficit reduction, and an expansion of health coverage from 85% of Americans to 95%.

Employers are best positioned to improve the health care market.

Nearly all of the margin for hospitals and doctors comes from patients with commercial insurance. For a hospital, a 1% decrease in commercial patient volume equates to a 10% reduction in EBITDA or surplus. This is a byproduct of commercially insured patients paying prices that are far higher (10%–400%) than Medicare (set payments that are plus or minus 3% net margin for providers) and Medicaid (money-losing for providers). If employers are willing to consolidate their purchasing, they can negotiate substantial discounts from hospital systems and, in some cases, warranties for quality. Tools like reference pricing, direct contracting with specific facilities, bundled payments, second opinions and referrals to higher value clinicians, and benefit designs that reward value-consciousness are promising approaches being used by some employers to reduce spending trends and improve quality.

Some care organizations are succeeding.

There are many examples of higher-quality and lower-cost care delivery models in the United States. Kaiser Permanente, HealthCare Partners, andCareMore, all in California, are held up as national models.

All tend to focus on caring for people with chronic disease to keep them out of the hospital. They are able to achieve 10%–40% lower rates of hospital utilization, fewer diagnostics tests, better medication adherence, and greater likelihood of following clinical guidelines, and much more use of information technology. They also attain high patient satisfaction and, in the case of Kaiser, lower churn than other health plans. All of these systems have in common two elements: capitated reimbursement models, and salaried doctors where margins are maximized by spending less money. In this way they have closed the payment and provision of care loop within one income statement. There is no evidence that these systems deny people needed care — rather, they coordinate care better to avert hospital use, duplicative tests, and unnecessary specialist visits.

What Needs to Be Done?

While the political climate makes additional legislative changes politically unlikely, there is broad agreement that the following changes would be smart:

  • Accelerate the rate at which the payment system shifts from fee-for-service to risk-based approaches, to incentivize lower-cost, more efficient approaches to care
  • Standardize care approaches to reduce unnecessary variation
  • Expand the scope of practice of non-doctor clinicians
  • Utilize technology to reduce duplicative care, increase the use of evidence-based care, exchange data, and coordinate care
  • Release data on drugs, devices, hospitals, and doctor performance at the patient and disease level to reduce information asymmetry
  • Reform the medical malpractice system to eliminate the incentive to perform unnecessary defensive care and reward adherence to evidence-based medicine
  • Reduce administrative cost and complexity of health care transactions
  • Incentivize patients to engage in their care and providers to engage in shared decision-making over treatment plans and goals

Fortunately, all of these changes can be initiated, tested, and proven by states or the private sector without new federal legislation. This is a call for hospitals, health systems, physician organizations, payers, self-insured employers, and state regulators to move ahead with changes to the health care system that will improve quality and reduce costs. There are mutual benefits to be captured, but only if all these parties work in tandem — and to the benefit of patients.

How can technology help?

Information technology is critical for improving the fidelity and specificity of care processes. It is also critical for catalyzing a much more competitive and efficient market.

To lower overall health care costs, payers, providers, and suppliers must employ technology to improve labor productivity similar to other manufacturing businesses. It should be possible to integrate all medical data to deliver patient-specific care plans optimized for simplicity, cost, and quality of life. Moreover, these same systems should be capable of generating services that instruct caregivers on how to achieve these goals. Finally, technology holds promise for engaging patients in the planning, purchasing, and the ongoing monitoring of their care to assure that treatments are working and course corrections are made seamlessly to avert complications. Done well, it may even be possible to make preventive and chronic disease care both safer and cost-efficient. I believe that IT remains the key to making U.S. health care better faster, just as it has most every other sector of our economy.


Source: http://bobkocher.org

How to Make Our Crazy, Expensive, Amazing, and Uneven Health Care System Better Faster

This post first appeared in NEJM Catalyst.

I am often asked by health care entrepreneurs, policymakers, health system leaders, and clinicians to “explain” how health care in America works — and to offer ideas for how to improve it. Here, within a whirlwind synthesis, is a provocative thought starter for how to make U.S. health care better faster.

The State of U.S. Health Care

The U.S. health system is bigger than the entire economy of France.

The United States spends about $3.2 trillion on health care. This equates to $8,500 per capita, which is twice what other developed countries spend per capita. While health care spending growth has slowed since the 2008 recession, it still exceeds GDP growth. Unique to the U.S. economy, health care experiences negative labor productivity. It is also the largest source of new jobs since the recession, despite flat demand.

We waste the equivalent of the entire health care system of Spain annually.

Many economists believe that about 30% of spending, or about $900 billion per year, is wasted in the U.S. health system. The largest primary driver of excessive spending is high prices rather than high utilization. Prices in the United States are about 60% higher than in other OECD countries. High prices are a result of a fee-for-service reimbursement system based on paying “cost plus,” coupled with the market power of providers and suppliers to raise prices at a 5% CAGR over the past 20 years. Supplier market power is further augmented by a lack of transparency of prices and quality, complexity, insurance benefits with limited cost sharing, supply-induced demand, structural local provider market power, and consolidation. Commercial insurance prices vary by 50%–400% for all services, in all markets, and are seldom correlated with quality.

We compete with Cuba and Costa Rica on quality.

The United States comes in 37th on international comparisons of quality, just behind Cuba and ahead of Costa Rica. Our life expectancy lags behind other OECD countries and is improving less quickly than others.

We only perform better on cancers and preterm babies. We diagnose cancer earlier thanks to more widespread screening programs that are very profitable to providers. We do not do better on chronic diseases. The United States is “less sick” than other OECD countries only because we are younger and have lower prevalence of smoking. We are the only OECD country without national insurance; in 2013, 50 million Americans were uninsured. Moreover, 18% of Medicare patients are readmitted to hospitals for the same condition within 30 days. About half of these readmissions are believed to be avoidable.

Just 10% of patients account for 65% of U.S. health care spending.

A cohort of 35 million Americans who suffer from multiple chronic diseases and spend $90,000 per year, on average, drives most U.S. health care spending. Most of these people are unable to work, poor, and covered by Medicare and Medicaid (referred to as dual eligibles). End-of-life spending accounts for only about 10% of spending. So while people may spend large amounts in hospitals in their last year of life, it is a relatively small contributor to overall spending since only about 2.6 million people die annually. Expensive biologic drugs, cancer drugs, high-tech medical devices, and new technologies collectively account for only 15% of U.S. spending.

Our health care market is massively fragmented.

Health care in America is subscale. Care is organized around roughly 500 disease categories and delivered by loosely organized collections of local providers and hospital systems. The largest hospital system in the country, HCA, accounts for only 5% of the market. While a bit more than half of the roughly 850,000 doctors are employed by local hospitals, those in private practice are in groups of 4–8 physicians on average.

As a result, most doctors care for several types of patients within their specialty, and virtually all hospitals are general hospitals with low volumes of many types of patients. The typical Medicare patient under treatment has 16 doctors involved in his or her care. Nevertheless, scale appears to be one of the most important predictors of quality, with doctors and hospitals performing higher volumes of similar cases delivering better outcomes. Other countries have fewer, larger, and more specialized hospitals, and specialty providers with narrower practice areas.

The Affordable Care Act is a big step forward.

The ACA was designed to expand coverage largely by redistributing about 5% of current Medicare spending to subsidize coverage for up to 30 million uninsured. Newly insured people would be covered by private health plans sold on exchanges or Medicaid (in states that expanded their programs). Most of the uninsured are poor, young, male, and comparatively healthy. The ACA has improved insurance market competitiveness for individuals by standardizing policies (for example, gold, silver, or bronze) and eliminating underwriting, so that all people are able to purchase insurance at prices that vary only by age.

The ACA has made minimal changes to employer-sponsored insurance plans; most Americans will continue to receive health insurance from their employers for the next decade. The ACA has also introduced new payment models containing incentives of varying strength for providers to improve quality and reduce costs. The transition from fee-for-service to alternative payment models is going to take at least a decade, as 90% of payments made in the United States today are still fee-for-service. The Congressional Budget Office expects the net effect of the ACA to be a small slowing of federal health care spending, about $300 billion of deficit reduction, and an expansion of health coverage from 85% of Americans to 95%.

Employers are best positioned to improve the health care market.

Nearly all of the margin for hospitals and doctors comes from patients with commercial insurance. For a hospital, a 1% decrease in commercial patient volume equates to a 10% reduction in EBITDA or surplus. This is a byproduct of commercially insured patients paying prices that are far higher (10%–400%) than Medicare (set payments that are plus or minus 3% net margin for providers) and Medicaid (money-losing for providers). If employers are willing to consolidate their purchasing, they can negotiate substantial discounts from hospital systems and, in some cases, warranties for quality. Tools like reference pricing, direct contracting with specific facilities, bundled payments, second opinions and referrals to higher value clinicians, and benefit designs that reward value-consciousness are promising approaches being used by some employers to reduce spending trends and improve quality.

Some care organizations are succeeding.

There are many examples of higher-quality and lower-cost care delivery models in the United States. Kaiser Permanente, HealthCare Partners, andCareMore, all in California, are held up as national models.

All tend to focus on caring for people with chronic disease to keep them out of the hospital. They are able to achieve 10%–40% lower rates of hospital utilization, fewer diagnostics tests, better medication adherence, and greater likelihood of following clinical guidelines, and much more use of information technology. They also attain high patient satisfaction and, in the case of Kaiser, lower churn than other health plans. All of these systems have in common two elements: capitated reimbursement models, and salaried doctors where margins are maximized by spending less money. In this way they have closed the payment and provision of care loop within one income statement. There is no evidence that these systems deny people needed care — rather, they coordinate care better to avert hospital use, duplicative tests, and unnecessary specialist visits.

What Needs to Be Done?

While the political climate makes additional legislative changes politically unlikely, there is broad agreement that the following changes would be smart:

  • Accelerate the rate at which the payment system shifts from fee-for-service to risk-based approaches, to incentivize lower-cost, more efficient approaches to care
  • Standardize care approaches to reduce unnecessary variation
  • Expand the scope of practice of non-doctor clinicians
  • Utilize technology to reduce duplicative care, increase the use of evidence-based care, exchange data, and coordinate care
  • Release data on drugs, devices, hospitals, and doctor performance at the patient and disease level to reduce information asymmetry
  • Reform the medical malpractice system to eliminate the incentive to perform unnecessary defensive care and reward adherence to evidence-based medicine
  • Reduce administrative cost and complexity of health care transactions
  • Incentivize patients to engage in their care and providers to engage in shared decision-making over treatment plans and goals

Fortunately, all of these changes can be initiated, tested, and proven by states or the private sector without new federal legislation. This is a call for hospitals, health systems, physician organizations, payers, self-insured employers, and state regulators to move ahead with changes to the health care system that will improve quality and reduce costs. There are mutual benefits to be captured, but only if all these parties work in tandem — and to the benefit of patients.

How can technology help?

Information technology is critical for improving the fidelity and specificity of care processes. It is also critical for catalyzing a much more competitive and efficient market.

To lower overall health care costs, payers, providers, and suppliers must employ technology to improve labor productivity similar to other manufacturing businesses. It should be possible to integrate all medical data to deliver patient-specific care plans optimized for simplicity, cost, and quality of life. Moreover, these same systems should be capable of generating services that instruct caregivers on how to achieve these goals. Finally, technology holds promise for engaging patients in the planning, purchasing, and the ongoing monitoring of their care to assure that treatments are working and course corrections are made seamlessly to avert complications. Done well, it may even be possible to make preventive and chronic disease care both safer and cost-efficient. I believe that IT remains the key to making U.S. health care better faster, just as it has most every other sector of our economy.


Source: http://bobkocher.org

The Coming Battle over Shared Savings — Primary Care Physicians versus Specialists

This paper was co-authored with Anuraag Chigurupati.

Abstract: New health care payment models will create economic winners and losers. As primary care physicians are rewarded for contributing to savings, specialists who fail to develop the capabilities to drive value improvement will face a threat to their incomes and practices.

Read the full article in The New England Journal of Medicine.


Source: http://bobkocher.org