EBM to PM - the Big Data Driven VBC - Phenix Qin

从循证到精准：大数据驱动下的价值医疗体系Phenix Qin 覃璞SAP首席产业专家，医疗卫生与生命科学

© 2016 SAP SE or an SAP affiliate company. All rights reserved. 2

主讲人：Phenix Qin 覃璞

担任SAP首席医疗卫生与生命科学行业专家，兼任其亚太区地区私募投资业务伙伴、中国医药信息学会北京分会副理事长等职务。他在健康产业管理、卫生政策和投资咨询方面拥有14年产、官、学相结合的国内外行业经验。曾在北京大学医学部从事心血管学相关

的基因及蛋白质信息大数据门户建设，后来在卫生部医院管理研究所担任信息标准化部副主任，并在卫生部电子病历委员会等多个标准化机构和行业协会中工作，参与工作包括国家电子健康档案标准草案、交换技术和用例、平台建设指南等。

之后数年间，他在澳大利亚加入私营医疗集团 HealtheCare工作（现已收购在绿叶医疗旗下），参与创立了其个人健康增值业务Healthe，并建立了其在健检、养老和乐活产业的渠道伙伴。他在运动与康复医学、协同与慢病照护、母婴照护、医疗健康大数据分析领域有着独到的见解。在此期间，他也代表 Healthe参与了ISO 个人健康信息工作组 (PHI) 的早期筹建及个人健康记录 (PHR)规范的建立工作。

他曾多年在北京大学光华管理学院、国家发展研究院、卫生部高级行政官员进修项目及其他一些海外MBA/EMBA项目中客座讲授医疗健康产业管理与投资等相关课程模块。他曾作为独立顾问为多家跨国IT企业、保险公司和投资机构提供关于中国医改和相关地方政策、资本进入机会、业务转型与产业链扩张等方面的咨询业务。

在战略投资方面，他在医疗健康类 (HCM)和科技类 (TMT) 产业基金的融资治理、投资标的、投后管理、退出设计等各个阶段均有丰富经验。特别是在产业链上下游业态之间开展协同业务模式设计和资本合作，从而推动价值链整合。

此类的投资和战略转型案例有：

• 为国内某知名诊所医疗集团设计其非核心供应链业务的剥离，战略转向其核心业务并进行资本化同业并购。同时以定向战略采购形式将原有供应链业务注入其三大主要供应商之一，以使该供应商在其业态市场迅速壮大，以此作为资本要约的重要部分完成对该供应商的资本注入。现两家企业已经在投后开始

供应链上下游整合。

• 为某大型国际中药集团设计五年生态发展规划，将重资产、高成本的药材基地和采购部分扩展为电子市场交易平台模式，融入了大宗商品交易和供应链金融模式，将松散型药材采收通过可追溯电子交易形式转型为订单农业生产。该规划直接支撑了对重要药材交易市场的投资并购，以及投后多个子公司的板

块性整合，并利好其海外产品投放项目。现基础平台已在逐期上线中。

• 为某知名专科医疗集团设计其协同医疗照护网络的商业模式，将原有粗放型医院全资并购扩展为同业协同网络，建立对基础康复和干预服务机构提供专科执业支持和管理输出的平台和流程，将面向医疗基础设施的重资产并购转型为面向品牌和服务网络的轻资产并购和合作。此项目已促成该医疗集团与上游

国际品牌的多点合资诊所项目，并将另一原潜在合资技术伙伴成功转为跨业并购标的。

www.linkedin.com/in/qinpu


SAP Footprint in Health Industry Chain


SAP Footprint in Health Industry ChainShowing only partial customers

Medicine RetailMedicine WholesalePharma Mfg. Healthcare Medical Insurance

Medical DeviceHealth

Administration

Medical Research

Wellness Products Wellness ServicesMedical School

TaslyPharm

Tasly Health

Tasly Deepur

BOHShanghai

Wuhan Asia Group

ManufactureGMP

CommercialGSP

Medical GLP

ServicesGCP, JCI and payer required compliance

HealthcareLife SciencesPublic Sector

InsuranceWholesaleHigher Edu RshConsumer Prod

Industry CodeColor Legend


价值医疗体系：医疗产业的“一句话解读”VBC: 1 formula to cover the healthcare industries

价值 Value = 适用度 Adaptivity x 疗效 Outcome

成本 Cost /时间 Time

关注V的筹资模式 à 推动了健康维系组织模式(HMO)关注O本身 à 推动了现代科学技术应用于医学关注O/C à 推动了药品福利管理 (PBM)关注O/(C/T) à 推动了慢病管理和长期协同照护关注A à 推动了病理学、组学基础研究关注整个体系 à 推动了ObamaCare, ACA, ACO


价值医疗体系的发展历程

n 健康维系组织HMO, Managed Care

n 按服务付费

Fee for Service(左栏)

n 诊断相关组

DRG（中栏）

n 责任共担医疗机构

Accountable Care Organization(ACA法案，ObamaCare，右栏）


医疗体系为实现“价值医疗”所做的努力

n 经验医学时代（最早的“个性化”治疗）n 始于2500年前的希波克拉底临床医学起源时代，在20世纪初达到鼎盛n 由于对疾病本质的了解还处于初级阶段，大量科学发现并未运用于临床，传统个性化治疗的理念没有根本改变，仍凭借医师个人的直觉及经验，局限于调整剂量、频繁换药的手法，经历着尝试及错误的循环，临床诊断失误率高达15%

n 循证医学时代

n 循证方法与20世纪80年代由肿瘤学家发明和利用，利用大量证据支持的临床指南作为基础，缩小治疗选择。临床指南最初作为一种标准的护理方法，主要的目标是提高患者的护理质量

n 基于科学研究的成果制定专家共识、指导临床实践，成为20世纪末的医学主旋律n 将聚类荟萃分析、RCT、队列研究、病例对照研究、个案报告、动物研究、体外实验等研究方法视为金标准n 最大局限：入组病人的疾病本质不同而造成研究结果的可靠性偏差，主要的根本原因

n 疾病的描述所依赖的ICD编码主要依赖于2000年前所使用的疾病表征加部位的组合方式，缺乏发病机理的刻画和分类n RCT受药物注册的商业动因影响，如研究时间有限、统计手段和边界条件的僵化


医疗体系为实现“价值医疗”所做的努力（续）

n 循证医学的改良

n 低级别但有价值的个案，颠覆证据层级的“金字塔”，更结合临床观察的医学循证（Medicine-Based Evidence）n 例子：动物试验发现胰岛素的研究

n 例子：关于心血管病药物万络特异人群风险的病例对比研究

n 例子：中草药复方丹参滴丸的在美国注册的临床试验

n 精准医疗的机会

n 病理解析研究，更深入地理解疾病的驱动因子

n 重新定义价值医疗体系的价值杠杆：“适用度” Adaptivityn 例子：乳腺癌的10类驱动因子研究，提示临床实用级别的10大类诊断分形n 例子：两种昂贵新药的截然不同医学临床接受度

n 结肠癌治疗药物阿柏西普（对照统计学意义成立，双倍治疗成本，生存期从12个月延长到13.5个月）n 治疗囊细胞纤维征(CF)的药物Kalydeco，虽然也高达30万美元每年，但针对Cf已知1900种基因变异因子中的一种

G551D可以精确进行靶向治疗，使得全球7万CF患者中的这一亚群3000例患者100%获益


改善循证医疗和精准医疗的共同基础：改善临床观察本身医疗集团的临床数据中心与“数据银行”模式

MedicationED

Pharmacy

Future

Operations Quality Improvement &

Innovation

Treatment Optimization & Real

World Evidence

Population Health

Health Economics Outcomes Research

Clinical Trials

Value Analysis

Personalized Medicine

SAP HANAAnalytics Predictive Insights

HRHospital

Billing

Lab

Registration

Pathology

Radiology

Clinic

ORReported

PatientOutcomes

Visualization

Validation Normalization Security

Fragmented Sources

Scalable Core

Actionable Insights

Longitudinal Patient Data

OUR SOLUTION

CONFIDENTIAL 8

CONFIDENTIAL+


医学观察的大数据分析举例：疗效预测

Predictive Outcomes: Exploring Patient Data

Predictive Analysis Data Visualization CONFIDENTIAL+


医学观察的大数据分析举例：未知强关联因子发现 Predictive Modeling: Exploring the Individual Factors

Live births are correlated with survival, though having 3 or more children becomes a factor in the mortality rate rising again.

新发现：妇女所生育子女数量与乳腺癌发病率之间的变化关系

Advocate Health Corporate, ACO美国伊利诺伊州AHC医疗集团探索程序基于SAPHANA内存数据库环境，由Convergence CT开发

CONFIDENTIAL+


医学观察的大数据分析举例：卫生经济学策略Predictive Findings: Tumor Detection in Women by Age Group Breast Cancer Tumor Discovery Rate Test of Independence:

40's 50's ALL

Tumor 545 653 1,198

Non-Tumor 478,645 477,639 956,283 Total 479,190 478,292 957,481

Tumor Rate 0.114% 0.137% 0.125%

Pearson Chi Square Stat 9.52 P-Value <0.01

!  Mammograms for women age 50-59 are significantly more likely to find tumors than for women age 40-49 (based on BI-RADS score of 5 or 6.)

!  Significance was shown for p=.01 (Pearson’s Chi-Squared test.)

Key++Results+

新发现：部署乳腺癌筛查的最有效年龄组

Advocate Health Corporate, ACO美国伊利诺伊州AHC医疗集团

探索程序基于SAPHANA内存数据库环境，由Convergence CT开发

CONFIDENTIAL+


精准医疗在大数据实时分析方面所需的性能和场景

基因测序数据

临床知识库患者病例

临床数据

药物诊疗

临床报告诊疗方案

其他诊断数据

Charité柏林大学医院

MKI生物医学分析机构大数据临床与科研分析SAP案例

搭建了比磁盘系统快40.8万倍的分析试验系统

DNA分析结果从2-3天缩短至20分钟，加速216倍

肿瘤数据分析提速1000倍，从数小时缩短至2-10秒

能够实时分析药物对特定人群的疗效

让更多癌症患者获得个性化诊疗实时大数据处理的行业实际意义

临床医学研究的大数据分析场景

l 医学观察的统计可行性实时模拟

l 临床试验和患者群体的实时级别互相匹配

l 高参与度的临床实践-临床科研协作比例

分析性能的提升导致分组迭代的速度大大加快，


基于SAP HANA的组学分析研究举例：蛋白质组特征数据库

新发现：

在大规模蛋白质组数据库上实现实时级别的分析建模交互，大大加快“假设-验证”周期甚至替代假设建模的传统方式www.proteomicsdb.org

发表在Nature杂志，2014.5

一作：慕尼黑工业大学二作：SAP三作：GSK Cellzome


SAP HANA平台上的医疗卫生项目

速度提升的场景

患者管理分析 (IS-H) 50x (55 秒à 800 毫秒)虚拟患者平台 5000x (4小时à 2-3 秒)处方明细分析 3600x (1小时à 1 秒)DNA序列排布 17x (85 小时à 5 小时)基于蛋白质组学的癌症诊断 22x (15 分钟à 40 秒)

全新的使用场景（传统方法难以做到的）

医学信息浏览器 Genome Analysis临床试验匹配 ProteomicsDB基因组浏览器 Biological Pathway Analysis大规模患者对照组分析 HANA Data Scientist

基因组学数据处理和分析流建模


SAP HANA 数据库技术

大规模加载可以对基因数据集和其他数据集进行高速

记录插入

Discovery Service

Read Event Repositories

Verification Services

SAP HANA

● ●

P A

up to 8.000 read event notifications

per second

up to 2.000 requests

per second

Discovery Service

Read Event Repositories

Verification Services

SAP HANA

● ●

P A

up to 8.000 read event notifications

per second

up to 2.000 requests

per second

T

文本抽取搜索医生

批注、诊断等等(非结构化数据)

SQL 行或列式接口便于连接其他工具(如R-Studio)

SQL

轻量级压缩在主内存中高效存储大量数据，同时允许

高速存取

多核并行计算跨多结点加速关联查询

在线扩展性适应新格式要求（例如改变VCF文件）而无

需系统停机离线

+++

与其他大数据技术的比较和关联

相比Hadoop体系n Hadoop体系强调大规模存储，以及大负

载问题的分解(M-R)，即使有SPARCn HANA体系关注大数据存储后的实时利用

（分析、流处理、可视化、内置算法等）n 二者结合部署的解决方案为HANA Vera

(与Intel合作的Hadoop联合套件)

相比Oracle ExaData/Exanlytics等

n HANA-完整的内存级DBMS和内存开发环境，SAP为此已迁移和逐步优化其主要企业应用套件

n Oracle-关注在不更改企业级应用的前提下加大内存的缓冲机制


SAP 医疗与组学数据平台：Foundation for Health

+


SAP 医疗与组学数据平台：Foundation for Health

Presen

tatio

n

SAPHANA

SQL,SQLScript,JavaScript,HTML5,andSAPWebIDE

Replication,streaming,andETLintegrationservices

Search

Applicationfunctionlibrary

Datavirtualization Textanalysis andmining

Spatial Databaseservices

Storedprocedureanddatamodels

Predictiveandplanningengine Businessrules

ApplicationandUIservices

IS-H

VariantbrowserDashboards

Data

Service

s

Plug-inframeworkandlogicaldatamodel

Externalalgorithmsandtools(R-server content)

Extensionforhealthcareandlifesciences

Clinical data Genomicsdata

CollectionsStandardcontent

DataofSAPPatientManagement

Partnerdata

Partnerservices

Portal services

SAPMedicalResearchInsights ...Healthengagement Partnerapplications

Nativedatawarehouseservices

Replication

Patienttimelineuserinterface (UI)

IS-HElectronic data capture, clinical trial management, and lab

systems

Electronicmedicalrecorddata

Anonymization(partner)

Genomicspipeline

UsermanagementTextualintegration

&Natural LanguageProcessing(NLP)

DataqualityAdapters, replication,andExtract, Transform, Load(ETL)


SAP Medical Research Insights: R&D analysis and decision making

Analysis of Big Data in R&DAnalyze structured unstructured data, including genomics, proteomics, and other omics data in real time through a user-friendly interface.

Real-world data analysisCapture and explore patient longitudinal data with a real-world evidence component.

Ad hoc reportingEnable ad hoc reporting by harmonizing data from many sources and representing it in understandable visualization options.

Secure platform to understand, predict, and decideAnalyze data and run scenarios for product-portfolio decisions and forshaping preclinical, clinical, and postmarket studies.

Minutes For the National Center for Tumor (NCT) Diseases to analyze a patient with up to 1,200 unique data elements in his or her records – which used to take weeks without SAP Medical Research Insights

97%“With CancerLinQ, we can also learn from the care given to the 97% of adult patients who do not currently participate in clinical trials.”American Society of Clinical Oncology (ASCO) President Clifford A. Hudis, MD, FACP

Typical results1

Access and analyze diverse medical data

1Sources: NCT SAP Customer Journey, CancerLinQ Press Release

Search interface – slice, dice, and dive deep into research and clinical data


SAP Medical Research Insights: Omics analysis

Deep-dive analysisExplore the complete data set, such as a full genome, to base level within seconds in an interactive user interface.

Broad analysisIntegrate and access relevant data from public and clinical sources.

Generation and validation of findingsDraw reliable conclusions concerning variants to drive personalized medicine.

-99.9%Mitsui Knowledge Industry (MKI) uses SAP HANA to speed genome analysis and bring down cost for DNA extraction and analysis from up to US$1 million to below $1,000.

90%Coverage of human proteome in ProteomicsDB by Technical University Munich powered by SAP HANA, which can be used in research for discovering therapeutic targets and developing new drugs.

Typical results1

Explore omics data in real time

1Sources: MKI SAP Customer Journey; Press Release SAP on Technical University Munich

Genomic variant browser – get a visual impression of a human genome sequence


SAP Medical Research Insights: Patient cohort analysis

Up to 600x After 9 months, Stanford School of Medicine run computations in analyzing its genomics data between 17 and 600 times faster with SAP solutions.

Up to 100 GB For a single patient are processed with the help of the SAP HANA platform to identify treatment options for that person by Molecular Health Inc.

Clinical and postmarket studiesInvestigate drug effectiveness for different medical traits.

Insights for trial strategyIdentify disease root causes and validate research hypothesis for new trials.

Outcome-based approachesFacilitate evidence-based outcome discussions with regulators and payers for new drug submission and annual reimbursement.

Typical results1

Gain insights for new personalized medicines

Search according to indications and patient pattern analysis – detect efficacy levels for specific patient populations

1Sources: Article MedCityNews, SAP Article “Molecular Health Wins SAP HANA Innovation Award”


Red Dot Award SAP Medical Research Insights Wins Red Dot Award for Communication Design in 2015.

Germany SAP Medical Research Insights recognized in “Germany –Land of Ideas” competition in 2015.

SAP Medical Research Insights: Validated data and real-time analysis

Validated data captureGet trusted results through full transparency into how data is collected.

Real-time analysisExplore medical statistics and clinical studies within seconds.

FlexibilityChange perspectives on how to slice and dice the data easily through a user-friendly interface.

Recognitions1

Understand diseases faster

Kaplan-Meier analysis – explore survival probability of different patient populations per cancer type or therapy

1Sources: Website Red Dot Award; SAP News


技术案例：韩国首尔盆塘医院实时级别临床数据中心架构

-3-

EMR/OS/TxRd

HIS*

ODS

TxDS

TxDP

TxDP

ETLETL

OLPSv

Sd

Ud

Dbd

RSd

DW

D

Pdv

I-MmD

Md

O

I

N

D

GL

HmIMdU

O

R

Md

M

Lb

TxSEMm

3

TPdSmwSP

TdmDWf)TxDP,)RSd3)D,)Pdvd5)Dbd

5


技术案例：半结构化到结构化的实时级文本挖掘和标记

--

Db

3S-TxDP

UdxvbdvbvdddDx

ddDW

TxP

RSd

Pdv

3 Dbd

5


技术案例：更高临床用户参与度的临床数据实时分析看板

-8-

7S5-Dbd

Dbddfzdfmffv I

dOI/O/ERPIfDExmOTmPR

N-I

PRF/SHRFMdD

UG

Ifm

IdDbdvdmfmfmddm

R

D

Mm

MdSff

R-m

P

FL

EmbdddR

SPHN

NSmm

mzd

SDM

R-m

P

FL

EmbdddR

HIS(OS,PS,EMR)

/HR

S

RM

---

ddDW

TxP

RSd

Pdv

3 Dbd

5



-3-

dx3RSdbdS

Md

Sex Age

Blood Type Region

Date

Age

Diagnosis

Narcotics

Result

Search Save Open Down Load Help

T58

Male

Female

A

B

C

D

RH+

RH-

VA

CA

TX

AZ

MA

NY

Research Search

> Statistics

Age

Sex

Blood Type

Region

> Date

> Diagnosis Date

> Treatment

Treatment Type

¾Treatment Process

Process Name

> Record

Entire Record

Exam Record

Pathology Record

> Exam

Research Exam

Exam Result

CI Monitor

User Setting

EMR Analysis

S

SR

H



--

Smoking/Drinking Medical Exam Medicine

Insurance Fee Operation Diagnosis

Scenario Based Search

Research Search

CI Monitor

User Setting

EMR Analysis

Treatment Scenario

Operation Scenario

RdS

DS

Basic Scenario

Department Research

Stomach Cancer

Colorectal Cancer

Breast Cancer

Final Result

Registered Numbers

Total 120 Records

Preview

Click

Drugs Codes

Drugs Codes Drugs Codes

Search Result Selection Result

Application

Smoking within one month Smoking within 6 months Smoking (Years) Drinking within one day Drinking within 7 days Drinking Period (Years) z

Exam Date Exam Code Exam Name Exam Report date Exam Result

Drug Prescription Date Drug Code Drug Name (Product) Drug Name (Component) Drug Name (General) Drug Daily Dosage

Diagnosis Date Department Diagnosis Code Diagnosis Name Age ICD 10 Diagnosis Code

Operation Date Operation Code Operation Name ICD 9CM Code Department Operating Surgeon

Insurance Prescription Date Insurance Code Insurance Name Insurance Categorization Related Medical Department

And And Or Sex/Age Dep. Doctor Diagnosis Exam OP Drug

RSbjP

SG

Sbd

Close

Save Refresh Preview Final Result H

dx3RSdbdSG


SNUBH: Transforming Patient Care and Data Access with SAP HANA®

CompanySeoul National University Bundang Hospital (SNUBH)

HeadquartersGyeonggi-do, South Korea

IndustryHealthcare

Products and ServicesOutpatient and inpatient services, medical examination, billing and insurance, issuance services, emergency services, and home care

Employees>1,900 employees

RevenueUS$275 million

Web Site www.snubh.org/dh/en

Business challenges• Improve management and monitoring of 320 clinical indicators • Provide physicians with relevant information in real time• Ensure researchers have fast, easy access to historical clinical

data, including critical patient records

Technical implementation• Ran proof-of-concept tests with numerous providers and selected

SAP for its superior in-memory computing technology• Built a clinical in-memory data warehouse on the SAP HANA®

platform to support Big Data and deliver real-time performance• Created a separate data mart to catalogue clinical indicators

Key benefits• Cut access time to key medical and research data from days to

seconds• Improved patient care and shortened inpatient hospital stays• Reduced the staff necessary to manage clinical indicators• Decreased researchers’ reliance on IT staff for data retrieval

“Managing clinical indicators and pulling relevant information out of a huge data warehouse is critical. With that data, doctors can ensure the correct treatment is prescribed and monitor results.”Soo Young Yoo, PhD, Assistant Research Professor, Center for Medical Informatics, Seoul National University Bundang Hospital

<2 secondsTo analyze quarterly data, compared to 1 to 2 months

700xFaster retrieval of up to 10 years of research data

147%Return on investment within 5 years according to PricewaterhouseCoopers


临床实践与组学分析协作的未来图景


总结

n 价值医疗体系

n 关注在既定的成本时间效益(C/T)上，产出最大化成效(O)，但要首先考虑医疗干预方式的适用性(A)n 在老龄化和疾病谱变迁的背景下，全球医疗支付保障体系的一致发展方向

n 价值医疗体系的动因驱动医疗实践的不断改良：经验医疗à循证医疗à精准医疗

n 从循证医疗到精准医疗

n 补充传统循证医疗中金标准的缺陷

n 大大改善适用性(A)在小样本对照组甚至个人级别的有效性n 共同基础：价值医疗体系中的临床观察能力

n 获取商业模式：“数据银行”联合数据使访问模式——临床机构与药企等进行联合临床试验研究的基础n 技术基础：一致的临床数据仓库(CDR) 参考信息模型，以及实时内存计算分析能力处理大量以“数据银行”模式大规模的

n 价值医疗体系衍生出的其他领域也有机会参与组学转化的成果

n 人群特征分组与个性化的药品福利管理PBM、慢病管理CDM等n 访问但非复制型的“数据银行”模式依然有效


© 2016 SAP SE or an SAP affiliate company. All rights reserved.

No part of this publication may be reproduced or transmitted in any form or for any purpose without the express permission of SAP SE or an SAP affiliate company.

SAP and other SAP products and services mentioned herein as well as their respective logos are trademarks or registered trademarks of SAP SE (or an SAP affiliate company) in Germany and other countries. Please see http://global12.sap.com/corporate-en/legal/copyright/index.epx for additional trademark information and notices.

Some software products marketed by SAP SE and its distributors contain proprietary software components of other software vendors.

National product specifications may vary.

These materials are provided by SAP SE or an SAP affiliate company for informational purposes only, without representation or warranty of any kind, and SAP SE or its affiliated companies shall not be liable for errors or omissions with respect to the materials. The only warranties for SAP SE or SAP affiliate company products and services are those that are set forth in the express warranty statements accompanying such products and services, if any. Nothing herein should be construed as constituting an additional warranty.

In particular, SAP SE or its affiliated companies have no obligation to pursue any course of business outlined in this document or any related presentation, or to develop or release any functionality mentioned therein. This document, or any related presentation, and SAP SE’s or its affiliated companies’ strategy and possible future developments, products, and/or platform directions and functionality are all subject to change and may be changed by SAP SE or its affiliated companies at any time for any reason without notice. The information in this document is not a commitment, promise, or legal obligation to deliver any material, code, or functionality. All forward-looking statements are subject to various risks and uncertainties that could cause actual results to differ materially from expectations. Readers are cautioned not to place undue reliance on these forward-looking statements, which speak only as of their dates, and they should not be relied upon in making purchasing decisions.

Documents

EBM to PM - the Big Data Driven VBC - Phenix Qin