Stanford Medicine 2018 Health Trends Report The .
Stanford Medicine 2018 Health Trends ReportThe Democratizationof Health CareDecember 2018
Foreword fromDean MinorData and thedemocratizationof health care.In last year’s inaugural Health TrendsReport, we discussed the explosion ofdata in medicine and what it meant forpatients, researchers, and physicians. Thisyear, we drill down into operation andimplementation, exploring how this wealthof information is changing traditionalhealth care roles, the experience ofpatient care, and access to services in thedigital age. What we’ve found has vastimplications for medicine writ large, andespecially for patients: all early signs pointtoward the democratization of health care.For our second annual report, weagain interviewed industry experts andconducted a comprehensive review ofpublicly available data from a variety ofmedia, analyst, and academic sources.We also spoke to our own faculty to gaininsight into the trends influencing theirwork and the issues they see driving thefuture of health care.In short, more individuals andorganizations now hold more data thanever before, and the pace at which data ismoving between them is accelerating. Atthe same time, new tools are now availablethat can rapidly and accurately interpretmedical data—from radiology imagingto genomics—and push insights directlyto the point of care, which is less and lessdefined by physical location.Together, these trends have thepotential to bring about a future whereStanford Medicinesophisticated medical knowledge is nolonger confined to the clinic. One day,perhaps soon, this expertise will live inour smart devices—readily accessible,whenever and wherever it’s needed. Thiskind of access could have an enormouslypositive impact on global health, especiallyfor patients who lack high quality careclose to home.Today, we’re starting to see walls comedown in the health care industry, allowingdata to flow more freely to where it cando the most good. Realms of historicallysiloed expertise are opening up tomore and more people. Patients cannow get access to their personal healthinformation. And digital tools are givingrise to new health platforms that areincreasingly useful to physicians andpatients alike.With all of this said, we haven’t yet reachedthe promised land of digitally enabledhealth care and open access to data.Health care lags other industries when itcomes to data sharing and interoperability.Case in point: We can manage our moneyfrom mobile devices anywhere in theworld, but we still can’t manage our healthrecords in this way. It’s clear that we havework to do when fewer than one in threehospitals can electronically find, send,receive, and integrate patient informationfrom another provider.We must also address the very real issuesaround privacy and ethics with regardto patient data. The opportunities westand to realize are huge, but the risksare similarly significant. Maximizing thepotential of patient data while ensuringits privacy and security at every stepwill be one of the defining challenges ofthis new era.But we are undoubtedly on our way toa future of care that is more predictive,preventive, and personalized. This is ourvision for Precision Health at StanfordMedicine, one in which we build a healthsystem that goes beyond curing diseaseand moves toward preventing it beforeit strikes. It’s a future that is good forpatients, physicians, and the long-termviability of the American health caresystem, and it’s one that we are earnestlyworking to achieve.This report is not intended to offerfirm conclusions, but rather to begin aconversation about how we can bestrealize the promise information gleanedfrom data offers our patients and thosewho care for them. I hope you find thereport valuable, and I look forward tohearing your thoughts in the coming year.Lloyd B. Minor, MDDeanStanford University School of Medicine
Table ofContentsIntroductionStanford Medicine11. Data is Growing and Flowing12. New Technologies and Industry Players1Three Pillars of Health Care Democratization3Intelligent Computing3Sharing8Security, Privacy, and Safety12The Road Ahead16References17
IntroductionIn last year’s Health Trends Report,we took a global look at theemergence of big data in the clinicaland experimental fields of medicine.The explosion of data sources,data types, and analysis toolsforeshadowed a sea change comingfor health care: in the laboratory, inthe clinic, in the living room.This year, we continue to track theexponential growth in the volumeand utility of that data, but wealso begin to take a deeper lookat how data sharing and its usewill transform research, medicalpractice, and the role patients playin their own health care. We haveseen this year a rapid increase inthe flow of data and informationacross an increasingly wide range ofstakeholders. More people hold moredata than ever before. As a result,for the first time in history, there isan opportunity to truly democratizehealth care.This health care democratizationis characterized by two majorfactors: the distribution of dataand the ability to generate andapply insights at scale. It promisesa world in which patients—armedwith data, technology, and access toexpertise—can take charge of theirown well-being and manage theirown health.Democratization will mean thatproviders focus less attention onroutine tasks and more on the areaswhere they provide the most valueand find the most satisfaction. Andindividuals managing their healthcare concerns will put less strain onthe health care system, lower costs,and improve public health overall.1. Data is growing—and flowing—acrossour health caresystem faster thanever before.Historically, health care hasoperated as a closed ecosystemof siloed institutions, with theresearch hospital as the hub withinthis environment, physicians haveserved as the primary gatekeepersof medical information. The flowof information has gone in onedirection: from expert to patient.Largely due to the digitizationof patient health records—rarea decade ago, but now nearlyubiquitous—data is flowing morefreely. Patients are now engagedwith the wider health care system inmore complex forms of informationsharing. More people are using digitaldevices, from smart watches tointernet-connected insulin pumps,to monitor and manage theirhealth. DNA testing is increasinglypopular and convenient—one in 25American adults has access to theirpersonal genetic data.1 Patients nolonger get all of their health careStanford Medicineinformation from the physician. Thisone-on-one relationship betweenexpert and patient is giving way to amultiplicity of information-sharingrelationships—one-to-many, manyto-one, and many-to-any.This transformation puts thepatient at the center, encouragingthe spread of medical knowledgein unprecedented ways. It is alsochallenging the health care sectorto adapt. The net result is that thepublic now has access to medicaldata, both personal and general,in ways it never has before: 93% ofhospitals and health systems enablepatients to access their health data,interact with health data, and usethat data to obtain health services.22. New technologiesand industry playersare taking medicalknowledge from ahuman scale to adigital scale.The growth and spread of data hasgenerated more information than anyone person could possibly interpret.However, recent breakthroughs indata science and artificial intelligence(AI) are quickly overcoming thischallenge. Medical experts andscientists are now training algorithmsthat can analyze vast quantities ofdata to extract insights.1
Stanford MedicineThese developments, though recent,have already attracted tech firms tothe health care sector in a big way.Tech companies see opportunity inthe need to bring health care intothe digital age, and they areinvesting heavily to address thesenew markets.Advances in DigitalHealth Taking Shapeat Stanford Medicine:Many roadblocks and inefficienciesthat previously encumbered thisdigital transition are gradually beingphased out and eliminated. Althoughwe have not yet reached an idealstate of democratization, early signsof where we’re headed are loud andclear. We must now take steps toaddress the challenges that stand inthe way of that ideal state.1. A Stanford-developed AIalgorithm for radiology canreliably screen chest x-raysfor more than a dozen typesof disease, and does so in lesstime than it takes to readthis sentence.3. Stanford and Apple teamedup on the Apple Heart Study,recruiting 400,000 participantsin a year with the goal ofdetecting atrial fibrillation inApple Watch wearers.2. Stanford Medicine’s SecondOpinion platform, powered byGrand Rounds, allows patientsto get a second opinion froma Stanford Medicine doctorwithout leaving home.4. Stanford Medicine’s Centerfor Clinical Excellence(CERC) is collaborating withStanford’s Artificial IntelligenceLab via the Partnershipin AI-Assisted Care, withcurrent pilot tests focusing oncomputer vision technology.Key Questions Going Forward What are the key opportunitiespresented by democratization? How do we realize this vision? How do we ensure that the benefitsplay out equitably for all people? What safeguards are necessary?The 2018 Health Trends Reportexamines these questions acrossthree domains that represent someof most important trends takingshape in health care today.1. Intelligent Computing2. Sharing3. Security, Privacy, and Safety2
Three Pillars ofDemocratizationin Health Care1. IntelligentComputingAlgorithms are fueled by data; the morethey have, the further they can go—andthe more we can learn. An increase in datais fueling better algorithms, creating avirtuous cycle for the industry. As analyticmethods continue to improve, theseapproaches will have a compoundingeffect on the industry, in both thenear- and long-term: increasing efficiency,improving predictive capabilities, enablinggreater personalization, and democratizingaccess to enhanced care. As health careleaders adapt to this technology, theymust take into account the practical andethical implications it poses.Stanford MedicineScope and Scale ofthe AI MarketThe AI and machine learning market isimmense and fast-growing, especiallyin health care, where it promises steadyimprovements in cost, access, and quality.The size of the AI health market is expectedto reach 6.6 billion by 2021—that’s acompound annual growth rate of 40%.That means that from 2014 to 2021, aperiod of just seven years, the health AIHealth AI Market Size 2014 - 2021 600 M 6.6 B- Dr. Paul Tang, VP, ChiefHealth Transformation Officer,IBM Watson Health2014The health care industry appears readyto embrace this development. In a surveyof senior health care decision-makersregarding AI maturity levels by industry,health care ranks higher than retail andfinancial services, with pharmaceuticalsand life sciences sectors consideredthe most AI-ready.4Near-term andLong-term Impacts11x[In] the next 2 to 5years there will be a lotof attention on dataanalytics and artificialintelligence that willallow us to learn fromlarge observationaldatasets. It will teach uswhat we do today whichwe don’t understand,how varied our practicesare, and what outcomeswe get with them.conditions, or images of tumors and reacha diagnosis in a matter of minutes. And asthey continue to do this, they will becomemore accurate over time. In some cases,these algorithms are already capable ofproviding diagnoses more accurately thanspecialists working alone.2021Acquisitions of AI startups are rapidlyincreasing while the health market isset to register an explosive CAGR of40% through 2021.Source: Frost & Sullivan (January 2016).Artificial Intelligence & Cognitive ComputingSystems in Healthcare.market will have grown more than 10x.3This surge of health data has led to thedevelopment of highly sophisticatedalgorithms that are able to parseinformation at an unprecedented scale.Image recognition algorithms, for example,can now process hundreds of thousandsof images, such as x-rays, photos of skinAI has significant implications for thehealth care sector, perhaps more so thanother industries. The McKinsey GlobalInstitute estimates that 15-20% of thehealth care market has the potential to beimpacted by AI, making it one of the mostaffected sectors. 5Intelligent computing has the potentialto add a significant amount of valueby improving quality of care, specificallyby allowing for quicker and more accuratediagnoses, higher quality treatmentplans, and new ways of managingprocesses. Some of the most anticipatedadvances include:Prediction:Intelligent algorithms have already shownpromise in predicting and identifyingpublic health threats as well as outcomes3
Stanford Medicinefor at-risk patients in the hospital. Asthey continue to improve, medicalprofessionals will have a powerful toolto provide patient care that is moreprecise, immediate, and preventive. TheSloan Kettering Institute estimates thatphysicians use only one-fifth of availabletrial-based information when diagnosingand treating cancer patients.6 Wheneffectively utilized, algorithms couldprocess vast amounts of data and providea larger picture of medical evidence to adoctor within seconds.Personalization:Physicians will soon be able to customizetreatment plans, and even drugs, toindividual patients based on a complexinterplay of factors such as theirgenetic makeup, habits, and digitallymonitored biometrics, to name a few. Asthis continues to evolve, standardizedhealth care will increasingly give way topersonalization. Research shows thattailored treatments could reduce healthexpenditures by up to 9% and add up to1.3 years to average life expectancy.7Access:Virtual and mobile care now serve as aprimary health resource for many patients.In fact, over the next five years the globalmobile health market is expected to havea compound annual growth rate of 29%.8When asked to name the top advantageof AI in health care, over a quarter ofconsumers cite having their own healthcare specialist available at any time,on any device.9Cost and Health Outcomes:Today, 90% of the U.S.’s 3.3 trillion annualhealth care spend is on people withchronic health conditions. 10By ensuring that the right care takes placeat the right time, and in the right setting,advances in data science and AI have thepotential to significantly slow the onset ofchronic disease at a population level andprevent some forms of disease altogether.Even small improvements—powered byintelligent computing—could result in largeeconomic savings for all stakeholders inthe health care system. This incentive islikely to encourage partnerships and morevalue-based purchasing models in thefuture, further bolstering the industry’sshift to preventive care.11By one estimate, AI could help reducehealth care costs by 150 billion by 2026in the U.S. alone.12 From a businessperspective, intelligent computing createsan opportunity for industry players tooptimize savings and profitability whilestill taking advantage of growth. VariousAI applications are expected to bringbillions of dollars in near-term value tothe economy, the most valuable beingrobot-assisted surgery ( 40 billion),virtual nursing assistants ( 20 billion),and administrative workflow assistance( 18 billion).13 These applications willonly become more accurate and efficientover time, implying even greater valuedown the line.Top Perceived Advantages of Using AI for Health CareHealth care would be easier andquicker for more people to accessFaster and moreaccurate diagnoses34%31%Will make better treatmentrecommendations27%Like having your own health carespecialist, available any timeand on any device27%Source: PwC (November 2016). Survey: The new imperatives for health.4
Stanford Faculty on the Benefitsof Intelligent Computing acrossStanford MedicineSpecialty Areas:Estimated Annual Benefits of AI Applications by 2026 (Billions) 40Robot-Assisted Surgery 20 18Virtual Nursing AssistanceAdministrative Workflow Assistance 17 16 14 13Fraud DetectionDosage Error ReductionConnected MachinesClinical Trial Participant IdentiferPreliminary DiagnosisAutomated Image DiagnosisCybersecurity 5 3 2Source: Accenture (December 2017). Artificial Intelligence in Healthcare.New advances in intelligent computingwill reduce the cost of treatments andprocedures that were previously seenas inaccessible. The price of genomesequencing continues to fall, while at thesame time, the actionable informationavailable from combining the results ofgenomic tests with other key determinantsof health is growing. Treatments forcomplex diseases such as cancer arebeing made increasingly more specificas a result of information obtained frompopulations of patients with comparablecharacteristics. Greater specificitytypically means less trial and error aswell as better outcomes.Additionally, AI holds the potential toaddress a significant challenge in modernmedicine: physician burnout. A recentStanford Medicine study found that amajority (55%) of physiciansreport symptoms of burnout—or,exhaustion, cynicism, and feelings ofreduced effectiveness. The study alsonotes that burnout influences quality ofcare, patient safety, turnover rates, andpatient satisfaction. 14Specifically, AI can provide relief tophysicians’ already hectic work days, sixhours (on average) of which are currentlyspent on electronic health record (EHR)data entry—a task that can be easilyhandled via automated methods. 15 Bymaking clerical processes faster and moreefficient, AI stands to play a meaningfulrole in reducing physician fatigue andallowing health care professionals to focusmore time on their patients.You have these large geneticassociations between aphenotype and a certaingenotype, and it is very hardfor the human eye to figureout which ones are real andwhat they’re telling us. Butthere have been machinelearning approaches appliedto genomic data sets that arenow suggesting candidatedisease teams.- Stanford FacultyGeneticsOn the more clinical side,there’s been a proliferationrecently of National CancerRegistry based studies thatanalyze data from tens orhundreds of thousands ofpatients. That also has let usask some questions that wecould never have asked withprospective studies. So, Ithink that big data has beena huge impact.- Stanford FacultyOncologyAI in the ClinicVarious specialty areas within healthcare have already experienced thebenefits of intelligent computing.For example, AI can be used to assistradiologists in identifying abnormalitiessuch as heart disease and cancer, andin conducting image guided proceduresand biopsies. The proper deploymentof algorithms to such processes has thepotential to greatly improve the accuracyof screening and diagnostics.When you look at cells orneurons under the microscope,there are a lot of featuresgoing on in those cells and thehuman eye and brain can onlyreally detect a few of them.There have been excitingdevelopments where machinelearning algorithms can detectthings and patterns that wecan’t really see with our eyes.- Stanford FacultyNeuroscience5
Stanford MedicineInternational CompetitionThe United States is not the only marketlooking to increase capabilities in AI andintelligent computing. China has becomeprogressively competitive with the U.S.in terms of investment in AI applicationsand solutions. China leads the world in thenumber of research publications on AI.16The country’s intense focus on this areahas resulted in the rapid development ofnew health care diagnostic tools.The U.S. leads the global AI sector inmany other respects: in average years ofexperience of data scientists (majorityhave more than 10 years), number of AIpatent applications (15,317), number ofworkers in AI positions (850,000), andpercent of AI investment coming fromprivate sources (66%).17Ethics of AutomationDespite the many perceived benefits ofAI in health care, public opinion splits onsome sensitive questions about its role indelivering care. In a recent global study ofconsumers, just over half of respondentsindicate that they would be willing to let anAI “doctor” perform a
2. Stanford Medicine’s Second Opinion platform, powered by Grand Rounds, allows patients to get a second opinion from a Stanford Medicine doctor without leaving home. Advances in Digital Health Taking Shape at Stanford Medicine: 3.Stanford and Apple teamed up on the Apple Heart Study, in a year with the goal of detecting atrial fibrillation in
SEISMIC: A Self-Exciting Point Process Model for Predicting Tweet Popularity Qingyuan Zhao Stanford University qyzhao@stanford.edu Murat A. Erdogdu Stanford University erdogdu@stanford.edu Hera Y. He Stanford University yhe1@stanford.edu Anand Rajaraman Stanford University anand@cs.stanford.edu Jure Leskovec Stanford University jure@cs.stanford .
Stanford Medicine 2018 Health Trends Report . Stanford Medicine . new era. But we are undoubtedly on our way to a future of care that is more predictive, preventive, and personalized. This is our vision for Precision Health at Stanford Medicine, one in which we build a health . public now has access to medical data, both personal and .
Stanford Health Care Organizational Overview 3 Contract Administration is a Shared Service of Stanford Health Care to Eight Other Stanford Medicine Entities Stanford Health are ("SH")is the flagship academic medical center associated with the Stanford University School of Medicine. SHC has 15,232 employees and volunteers, 613 licensed
Domain Adversarial Training for QA Systems Stanford CS224N Default Project Mentor: Gita Krishna Danny Schwartz Brynne Hurst Grace Wang Stanford University Stanford University Stanford University deschwa2@stanford.edu brynnemh@stanford.edu gracenol@stanford.edu Abstract In this project, we exa
Computer Science Stanford University ymaniyar@stanford.edu Madhu Karra Computer Science Stanford University mkarra@stanford.edu Arvind Subramanian Computer Science Stanford University arvindvs@stanford.edu 1 Problem Description Most existing COVID-19 tests use nasal swabs and a polymerase chain reaction to detect the virus in a sample. We aim to
Stanford University Stanford, CA 94305 bowang@stanford.edu Min Liu Department of Statistics Stanford University Stanford, CA 94305 liumin@stanford.edu Abstract Sentiment analysis is an important task in natural language understanding and has a wide range of real-world applications. The typical sentiment analysis focus on
Trends in Care Delivery and Community Health State Public Health Leadership Webinar Deloitte Consulting LLP June 20, 2013. . Current state of Accountable Care Organizations (ACOs) and trends. Current state of Patient-Centered Medical Homes (PCMHs) and trends. Introduction.File Size: 2MBPage Count: 38Explore further2020 Healthcare Trends and How to Preparewww.healthcatalyst.comFive Health Care Trends For 2020 Health Affairswww.healthaffairs.orgTop 10 Emerging Trends in Health Care for 2021: The New .trustees.aha.orgRecommended to you b
Andreas Wagner Head of Building Science Group Karlsruhe Institute of Technology Department of Architecture. Background Occupant behaviour has a strong influence on building energy performance Reasons for occupants’ interventions: dissatisfaction with building automation interfaces are not designed/equipped for intended purpose designers / building managers do not fully consider –or .
