Episodes Report Full

7/30/2019 Episodes Report Full

1/50

NQFNATIONALQUALITYFORUM

PATIENTFOCUSEDEPISODESOF CARE

Measurement Framework:Evaluating Efficiency AcrossPatient-Focused Episodes of Care


2/50


3/50

Measurement Framework: Evaluating Efficiency AcrossPatient-Focused Episodes of Care

Foreword

AMERICANS DESERVE VALUE from their healthcare experiences. Valueencompassingquality, cost, and outcomes and ideally driven by patient preferences of careis sorelylacking in the U.S. healthcare system today.

Per capita spending on healthcare in the United States is more than double that of otherindustrialized nations, yet the United States ranks poorly compared with other countries onkey indicators of the quality of care patients receive and their health status. Furthermore,approximately 30 percent of healthcare spending is devoted to services that provide nohealth benefit to patients. These are the hallmarks of an inefficient system.

Efficiencyso important that the Institute of Medicine has deemed it one of the sixaims of a quality healthcare systemis notoriously difficult to measure. The road map tohealthcare quality improvement must include measures of efficiency that not only accordwith patients preferences but also reflect national priorities and goals for quality improve-

ment. Thus, the National Quality Forum (NQF), which was established in 1999 to facilitatewidespread healthcare quality improvement, sought to endorse a workable and effectiveframework for evaluating the efficiency of care over time.

This report represents the culmination of that project. The framework contained hereinwas vetted through NQFs public comment and voting process, granting it NQF-endorsed

status. It is viewed as a living document that will continue to evolve as evidence andpractice continue to inform its key components.

NQF thanks the Evaluating Efficiency Across Patient-Focused Episodes of Care SteeringCommittee, the Committees co-chairs Elliott Fisher, MD, MPH, and Kevin Weiss, MD,MPH, and NQF Members for their stewardship of this project and their commitment to thecreation of a high-performing, high-value healthcare system.

National Quality ForumNational Quality Forum ii

Janet M. Corrigan, PhD, MBAPresident and Chief Executive Officer


4/50

National Quality Forumii

National Quality Forum

The mission of the National Quality Forum is to improve the quality of Americanhealthcare by setting national priorities and goals for performance improvement,endorsing national consensus standards for measuring and publicly reporting onperformance, and promoting the attainment of national goals through educationand outreach programs.

Primary support for this project was provided by the Robert Wood Johnson Foundation(www.rwjf.org). Additional funding was provided by the Commonwealth Fund(www.commonwealthfund.org).

Recommended Citation: National Quality Forum (NQF). Measurement Framework: EvaluatingEfficiency Across Patient-Focused Episodes of Care. Washington, DC: NQF; 2009.

2009. National Quality ForumAll rights reserved

ISBN: 978-1-933875-42-2

No part of this report may be reproduced, stored in a retrieval system, or transmitted, inany form or by any means electronic, mechanical, photocopying, recording, or otherwise,without prior permission of the National Quality Forum. Requests for permission to reprintor make copies should be directed to:

Permissions

National Quality Forum601 13th Street NWSuite 500 NorthWashington, DC 20005Fax 202-783-3434www.qualityforum.org
http://www.rwjf.org/http://www.rwjf.org/http://www.commonwealthfund.org/http://www.qualityforum.org/http://www.commonwealthfund.org/http://www.commonwealthfund.org/http://www.rwjf.org/http:///reader/full/www.rwjf.orghttp://www.commonwealthfund.org/http://www.qualityforum.org/


5/50

National Quality Forum iii


Table of Contents

Executive Summary .................................................................................................... v

Introduction and Overview......................................................................................... 1

Statement of the Problem ........................................................................................ 1

Focus of This Report .................................................................................................. 5

Measurement Framework ........................................................................................... 5

Key Terms and Definitions....................................................................................... 5

Purpose of the Healthcare Delivery System and Supporting Role of the Performance Measurement System........................................................................... 6

Episodes of Care .................................................................................................. 7

Generic Episode of Care...................................................................................... 10

Domains............................................................................................................. 12

Guiding Principles ............................................................................................... 14

Path Toward a Comprehensive Measurement System .................................................. 16

Appendix A Case Studies .................................................................................... A-1

Appendix B Measuring Efficiency Across Patient-Focused Episodes of Care Steering Committee ........................................................................... B -1

Appendix C Application of Key Terms and Definitions: Case Scenario ..................... C-1

Appendix DContext for Considering an AMI Episode ............................................ D-1

Appendix E Context for Considering a Low Back Pain Episode................................ E-1

Notes.................................................................................................................... F-1


6/50


7/50

National Quality Forum v


Executive Summary

ALTHOUGH HEALTHCARE SPENDING per capita in the United States is more thandouble that of other industrialized nations, the United States ranks comparatively low onkey indicators of the quality of care and population health status.1 Inefficiencies such as

duplicate tests and widespread regional practice variations plague the system. In one study,more than 4 in 10 Americans reported experiencing inefficient, poorly coordinated, orunsafe care.2 This combination of high cost and low quality indicates a system that is ofpoor value, and Americans clearly deserve better.

Performance measurement is essential to system transformation.3 Substantial progress hasbeen made in developing and implementing reliable measures of healthcare quality. Andalthough there are several notable exceptions, most quality measurement efforts are poorlycoordinated and do not focus on areas with the greatest potential to improve outcomes orcontrol costs. Thus, we have yet to achieve the healthcare system we desire that embracesthe Institute of Medicines aims for safe, timely, effective, efficient, equitable, and patient-

centered care.To provide guidance to key stakeholder groups in accelerating toward a high-performing,

high-value healthcare system, the National Quality Forum (NQF) convened a SteeringCommittee to develop a framework for evaluating the efficiency of care over time, includingclear definitions and a shared vision of what can be achieved around quality, cost, andvalue, serving as a foundation for the work of larger performance improvement efforts.This report presents the NQF-endorsed measurement framework for assessing efficiency,and ultimately value, associated with the care over the course of an episode of illness andsets forth a vision to guide ongoing and future efforts.

This framework consists of the following: key terms and definitions;

an explanation of the patient-focused episode of care approach;

domains for performance measurement for evaluating efficiency; and

guiding principles.


8/50

National Quality Forumvi


Measurement Framework:Evaluating EfficiencyAcross Patient-FocusedEpisodes of Care

Key Terms and Definitions4

Quality of care is a measure of performanceon the Institute of Medicines (IOM) sixaims for healthcare: safety, timeliness,effectiveness, efficiency, equity, and patientcenteredness.

Cost of care is a measure of the totalhealthcare spending, including totalresource use and unit price(s), by payoror consumer, for a healthcare service orgroup of healthcare services associated witha specified patient population, time period,and unit(s) of clinical accountability.

Efficiency of care is a measure of cost ofcare associated with a specified level ofquality of care. Efficiency of care is ameasure of the relationship of the cost ofcare associated with a specific level ofperformance measured with respect to theother five IOM aims of quality.

Value of care is a measure of a specifiedstakeholders (such as an individualpatients, consumer organizations, payors,providers, governments, or societys)preference-weighted assessment of aparticular combination of quality andcost of care performance.

Purpose of the HealthcareDelivery SystemThe purpose of the healthcare delivery systemis to improve health, reduce the burden ofillness, and maximize the value of individualand societal resources allocated to healthcareand is fundamentally rooted in the needs ofthe patient, and, more broadly, society. Aneffective measurement framework shouldcontribute to that purpose by supporting

judgments about the degree to which thehealthcare delivery system and its componentparts contribute to achieving this purpose.

Theoretical Construct:Episodes of Care ApproachAn episode of care is defined as a seriesof temporally contiguous healthcare servicesrelated to the treatment of a given spell of

illness or provided in response to a specificrequest by the patient or other relevant entity. 5

The Committee developed a generic episodeof care model, which can be used to track thecore componentspopulation at risk, evaluation and initial management, and follow-upcarethat must be measured and evaluatedover the course of an episode of care. Thesecomponents are foundational to any assessment of efficiency. This model is adaptable to

multiple types of episodes, and the constructis designed to be applied to a broad set ofhealth conditions; this report has applied itto two different types of conditionsacutemyocardial infarction and low back paintoallow for examination of an acute conditionand transition between providers and settings,as well as a chronic, preference-sensitive


9/50

National Quality Forum vii

Measurement Framework: Evaluating Efficiency Across Patient-Focused Episodes of Care

Generic Episode of Care

Population at Risk

PHASE 1

Clinical episode begins

End of Episode

Risk-adjusted health outcomes(i.e., mortality & functional status)

Risk-adjusted total cost of care

PHASE 2

PHASE 3

Evaluation & Initial

Management Follow-up Care

Appropriate Times Throughout Episode

Determination of key patient attributes forrisk-adjustment

Assessment of informed patient preferencesand the degree of alignment of careprocesses with these preferences

Assessment of symptom, functional, andemotional status

Time

condition in which shared decision makingplays a significant role. Subsequent work hasbeen completed on breast and colorectalcancers, diabetes, and substance use illness.6

Domains for PerformanceMeasurement

The following domains represent the essentialcomponents and subcomponents for measuringefficiency as it relates to an episode of care.

Health outcomes important to patients

Health status/health-related quality of life Patient experience with care

Cost and resource use

Processes of care

Guiding Principles

The following principles are intended toguide development and implementation of themeasurement framework as applied acrossepisodes of care.

1. Efficiency measurement is multidimensional.

2. The choice of measures to inform judgmentson efficiency should include considerationof potential leverage.


10/50

National Quality Forumviii


3. Measures used to inform judgments on effi

ciency should promote shared accountabilityacross providers and should be assignedto the smallest unit of accountability astechnically feasible.

4. Measures used to inform judgments onefficiency should respond to the need toharmonize measurement across settingsof care.

5. Measures to inform judgments on efficiencyshould be used for benchmarking.

6. Public reporting of measures of efficiencyshould be meaningful and understandableto consumers and entities accountable fortheir care.

7. Inappropriate care cannot be efficient.

8. The measurement framework shouldachieve its intended purpose and should bemonitored for unintended consequences.

9. Measures to inform judgments on efficiencyshould be an integral part of a continuous

learning system.

Notes1 The Commonwealth Fund,Mirror, Mirror on the Wall: AnInternational Update on the Comparative Performance ofAmerican Healthcare. New York: The Commonwealth Fund;2007. Available at www.commonwealthfund.org/usr_doc/1027_Davis_mirror_mirror_international_update_final.pdf. Last accessed January 2009, p. 1.

2 Ibid.

3 Institute of Medicine.Performance Measurement:Accelerating Improvement. Washington, DC: National

Academies Press; 2006.4 These terms are adopted fromAQA Principles of Efficiency

Measures. Available at www.aqaalliance.org/files/PrinciplesofEfficiencyMeasurementApril 2006.doc.Last accessed January 2009.

5 Hornbrook MC, Hurtado AV, Johnson RE, Health careepisodes: definition, measurement and use, Med Care Rev,1985;42(2):163-218, p. 171.

6 Additional information on subsequent work completed can befound at www.qualityforum.org/Projects/Episodes_

of_Care_Framework.aspx.
http://www.commonwealthfund.org/usr_doc/1027_Davis_mirror_mirror_international_update_final.pdfhttp://www.aqaalliance.org/files/PrinciplesofEfficiencyMeasurementApril2006.dochttp://www.qualityforum.org/Projects/Episodes_of_Care_Framework.aspxhttp://www.qualityforum.org/Projects/Episodes_of_Care_Framework.aspxhttp://www.aqaalliance.org/files/PrinciplesofEfficiencyMeasurementApril2006.dochttp://www.commonwealthfund.org/usr_doc/1027_Davis_mirror_mirror_international_update_final.pdfhttp://www.commonwealthfund.org/http://www.aqaalliance.org/fileshttp://www.qualityforum.org/Projects/Episodes


11/50

National Quality Forum 1


Introduction and Overview

Statement of the Problem

Most Americans will not be able to afford healthcare if expenditures for healthcare servicescontinue to grow at their current pace. A recent survey of U.S. adults found that 50 percent

of middle- and lower-income families reported serious problems paying for care, and anequal proportion are worried about the affordability of healthcare for themselves and theirfamilies in the near future.1 Although per capita spending on healthcare in the United Statesis more than double that of other industrialized nations, the United States ranks poorly compared with other countries on key indicators of the quality of medical care and the healthstatus of the population.2 For example, in a comparative study of the United States and fiveother industrialized nations, the United States ranked last in safe care and had the highestinfant mortality rates.3 In light of these findings, it is not surprising that 42 percent of adultsin this country report experiencing inefficient, poorly coordinated, or unsafe care over thepast two years.4

Inefficiency and waste are pervasive. Patients often receive duplicate tests or do not havetheir medical records available when they visit a doctor. More money is spent in the UnitedStates on administrative functions related to insurance than in other countries7.3 percentof national expenditures on health. If the United States were in-line with other countries whohave mixed private-public insurance systems, it is estimated that $32 to $46 billion a yearcould be saved.5 Clearly, Americans deserve better value for their healthcare dollars.

Widespread variation in spending also occurs across the country. This variation often isunrelated (or, at times, is inversely related) to the quality of care. For example, Medicarebeneficiaries in higher-spending regions of the United States do not experience higherquality of care than those in lower-spending regions. In some cases the quality of care isworse, as indicated by health outcomes and patient satisfaction. Differences in spendingappear to be due to differences in physician practice patterns that are driven in part by thegreater per-capita supply of hospitals and specialists: Patients in higher-spending regionsare much more likely to be treated as inpatients and by multiple specialists compared withsimilar patients in lower-spending regionsand they receive more tests, imaging services,and minor procedures.6


12/50

National Quality Forum2


However, there are reasons for optimism.

Performance measurement is widely acceptedas essential to improvementby identifyingopportunities for improvement, motivatingproviders to improve, and providing the basisfor aligning incentives with better perform-anceand substantial progress has been maderecently in the development and implementationof reliable measures. Promising examples existacross the country, in which organizations areimproving the quality of care while shedding

waste and controlling costs (see Appendix A).Finally, there is a growing recognition that thepotential savings from reducing waste andimproving efficiency are substantialwithan estimated 30 percent of U.S. healthcarespending devoted to services that provide nohealth benefit.7

Still, although significant resources arebeing invested in the development of performance measures by many organizations, current

efforts are poorly coordinated and often failto focus on high-leverage areas that havethe greatest potential to improve quality andcontrol costs. These efforts are thus unlikely tolead to fundamental change in the healthcaredelivery system.8 Notwithstanding certainexceptions, widespread adoption and diffusionof best practices have been slow to occur.

Multiple stakeholders must align in order toaccelerate improvement by:

creating a road map for healthcarequality improvement that includespriorities and performance improvementgoals to unify and build upon existingefforts, and set a more deliberate courseof action for the nation;

developing a workable and effective

framework for evaluating the efficiencyof care over time, including clear definitionsof terms and a shared vision of what canbe achieved around quality, cost, and valuethat promotes better care coordination anda sense of shared accountability among themultiple providers involved in a patientscare;

developing performance measures andefficient data collection and reportingstrategies that will inform our efforts to

improve key areas such as enhancing carecoordination, aligning care with patientspreferences, and controlling the cost of care;and

creating an integrated national,regional, and local leadership modelthat will guide and enable efforts to drivecare improvement.

Strategic Direction

To create a road map for healthcare

quality improvement, the National QualityForum (NQF) convened the National PrioritiesPartnership (NPP), a collaborative effort of32 major national organizations (Partners) thatcollectively influence every part of the health-care system. The Partners have identified a setof National Priorities and Goals for nationalaction that aggressively targets eliminatingharm, eradicating disparities, reducing diseaseburden, and removing waste from the health-

care system.9

The Partners also have agreed towork collaboratively and with policymakers,healthcare leaders, and other stakeholders todevelop action plans around the Priorities andto align the drivers of change (e.g., paymentreform, performance measurement, andaccreditation and certification) aroundcommon goals.10
http:///reader/full/goals.10http:///reader/full/goals.10http:///reader/full/goals.10


13/50



To develop a workable and effective

framework for evaluating the efficiencyof care over time, NQF sought to developa comprehensive measurement framework inorder to evaluate efficiency, and ultimatelyvalue, across patient-focused episodes ofcarethat is, the care of people over thecourse of an episode of illness. As with otherprojects, NQF convened a multistakeholderSteering Committee (Appendix B) to shepherdthis work.

In an attempt to operationalize the measurement framework, the Committee targeted twovery different types of conditionsacutemyocardial infarction and low back paintodetermine the applicability of the framework tothese conditions, thus making the frameworkmore likely to be generalizable.

The formative work of this Committee andthe ongoing work of NPP cumulatively hope tolead to the development of performancemeasures and efficient data collection

and reporting strategies and to the encouragement of the creation of an integratednational, regional, and local leadership

model that will guide and enable efforts todrive care improvement. These efforts willrequire ongoing engagement of the manystakeholder groups that already have advancedperformance measurement. Ultimately, theseefforts will facilitate better alignment of meas

urement development and reporting activitieswith the National Priorities and Goals; addressgaps in the quality measurement agenda;and begin to define longitudinal performancemetrics of patient-level outcomes, resource use,and key processes of care. Furthermore, theseefforts aim to simplify the measurement process

so that it can motivate and support healthcare

professionals, provider organizations, patients,and communities to ensure that patients receivethe most efficient, high-quality healthcarepossible.

The framework contained in this documentproposes a patient-centered approach tomeasurement that focuses on patient-leveloutcomes over timesoliciting feedback onpatient and family experiences; assessingfunctional status and quality of life; ensuring

treatment options are aligned with informedpatient preferences; and using resourceswisely. It will require fundamental change inthe healthcare delivery system.

The framework presented here is viewed asa living document that will continue to evolveas we learn more about how to best addressindividuals with complex chronic illnesses andbetter integrate public health and personalhealth systems. Nonetheless, this frameworkprovides a starting point for identifyingmeasurement gaps and for examining modelsof shared accountability to help move us closertoward attaining a high-performing, high-valuehealthcare system.


14/50


15/50

5National Quality Forum 555


Focus of this Report

IN THIS REPORT, the groundwork is laid for a measurement framework that evaluatesefficiency, and ultimately value, across patient-focused episodes of care. This frameworkwill help key stakeholders move toward a high-performing healthcare system that is patient-

centered, focused on quality, mindful of costs, and vigilant against waste.

Measurement FrameworkThe Committees recommended measurement framework is presented below. First, key termsare defined to establish a common understanding of what is meant by efficiency andrelated constructs (such as quality and cost). Next, the purpose of the healthcare deliverysystem and the role that performance measurement should play in achieving that purposeare clarified. The definitions and the purpose provide the foundation for the Committeesdecision to emphasize the importance of focusing on health outcomes and total costs over

episodes of care when measuring performance. The advantages and disadvantages of theepisode of care approach to performance measurement are provided, and a model of ageneric episode of care is explained. Finally, the recommended domains for performancemeasurement and a set of principles to guide future work are presented.

Key Terms and Definitions

The Committee recognized the importance of agreeing upon a common vocabulary aroundefficiency measurement, and it capitalized on the many efforts already underway in thefield. The Committee looked to the work of the AQA, which had already come to consensus

on straightforward definitions for the constructs that are inherent to measuring and evaluatingefficiency, and which gained approval from the AQA membership.11 The Committee agreedwith AQAs approach, and it recommended the adoption of the definitions in Box 1 as ameans to promote a common understanding among the many stakeholders committed tothis work and to better align existing and future undertakings in this area. Appendix Cdiscusses these terms in greater detail and provides a real-life example of how they canbe applied.
http:///reader/full/membership.11http:///reader/full/membership.11http:///reader/full/membership.11


16/50



Box 1Key Terms and Definitions

The four terms and accompanying definitions presented below are distinct but interrelated constructs. The Committee recommended that measurement within these constructs not be pursuedindividually or in isolation, but rather as an essential subcomponent of a larger set of measuresneeded to adequately assess efficiency overall.

Quality of careis a measure of performance on the six Institute of Medicine (IOM) specifiedhealthcare aims: safety, timeliness, effectiveness, efficiency, equity, and patient-centeredness.

Cost of careis a measure of the total healthcare spending, including total resource useand unit price(s), by payor or consumer, for a healthcare service or group of healthcare

services, associated with a specified patient population, time period, and unit(s) of clinicalaccountability.

Efficiency of careis a measure of cost of care associated with a specified level of quality ofcare. Efficiency of care is a measure of the relationship of the cost of care associated witha specific level of performance measured with respect to the other five IOM aims of quality.

Value of careis a measure of a specified stakeholders (such as an individual patients,consumer organizations, payors, providers, governments, or societys) preference-weightedassessment of a particular combination of quality and cost of care performance.

Purpose of the Healthcare DeliverySystem and Supporting Role of thePerformance Measurement System

Ultimately, a measurement framework willbe deemed successful to the degree that itcontributes to the success of the healthcaredelivery system at achieving its purpose.Drawing on earlier work by the Institute of

Medicine (IOM)12 and others, the Committeeconcluded that the purpose of the healthcaredelivery system is rooted fundamentally inthe needs of the patient, and more broadly,society. The purpose of the healthcaredelivery system is

to improve health, reduce the burdenof illness, and maximize the value ofindividual and societal resources allocatedto healthcare.

Accordingly, the measurement frameworkshould support judgments about the degree towhich the healthcare delivery system and itscomponent parts (e.g., providers, health plans,payers, and government agencies) contributeto achieving this purpose. Even more impor

tantly, a performance measurement systemshould both motivate and support continualimprovement in the healthcare delivery systemand its demonstrated capacity to achieve thegoal of improving health and reducing boththe cost and burden of illness.


17/50



The Committee recognized that having a

clear purpose will not eliminate disagreementsor the need to make difficult decisions. Differentstakeholders will have different perspectives onefficiency. For example, a patient may considerbeing seen for an appointment in a timelymanner as a relevant barometer of efficiency,considering the opportunity costs of runninglate (e.g., missed wages), whereas a providermay consider it more efficient to overbookpatients to ensure that all appointments are

filled and the practice sustained. However, theCommittee concluded that its work should proceed primarily from the patients perspective,which offers a clear path toward redesigningpayment and care models to reduce the burdenof illness, while eliminating waste and maximizing value. A more difficult set of issuesinvolves decisions about how to best allocateresources to the healthcare system itself(compared to other societal investments) and

among the potential competing priorities withinhealthcare. In this case as well, however,judgments and prioritization will be fairerand better aligned with the purpose of thehealthcare system to the extent that they canbe based on reliable information about theimpact of different clinical interventions andapproaches to care delivery on patients andpopulations health, the burden of illness, andthe overall costs and value to patients of thealternative approaches to providing care.13

Episodes of Care

Rationale for using episodes of careto characterize performance

A measurement framework that can informstakeholders judgments of the degree to which

the delivery system is improving health and

reducing the burden of illness, at an appropriate level of investment, should parallel thenatural trajectory of the clinical conditions(i.e., injuries, diseases, and disabilities) to beassessed.14 Therefore, the Committee found thetheoretical construct of an episode of care as auseful approach to characterizing performance.Specifically, an episode of care is defined as

a series of temporally contiguous healthcareservices related to the treatment of a given

spell of illness or provided in response toa specific request by the patient or otherrelevant entity (p. 171).15

The Committee concluded that an episodeperspective is required to determine if thedelivery system is indeed achieving its intendedpurpose, because this approach allows forcare to be analyzed over time and offers abetter assessment of the patients resultanthealth status.

Types of Episodes of Care

An episode of care may be acute, such as afractured arm (for which onset is easily definedand the period of recovery is relatively predictable), or more chronic, such as diabetes(for which onset is gradual, treatment requiresongoing efforts, and the goals are to preventdisease progression and to minimize complications over a prolonged period of time). A single

episode may also include both acute andchronic care. In the case of heart attacks,for example, the initial management requireseffective coordination of acute care resources(e.g., emergency services, hospital emergencyrooms, multiple healthcare professionals,cardiac catheterization suites, and intensivecare units). However, maximal recovery
http:///reader/full/assessed.14http:///reader/full/assessed.14http:///reader/full/assessed.14


18/50



requires marshalling post-acute care services

(e.g., rehabilitation) and implementing effective,secondary strategies (e.g., smoking cessationcounseling, lipid-lowering medications) toprevent further progression of disease andperhaps another heart attack. These examplesunderscore the need to pay careful attention tothe varying durations of different healthcareepisodes and to the important contributionsof multiple healthcare professionals (e.g.,nurses, pharmacists, allied health professionals,

physicians, and others) to the delivery ofhigh-quality care.

Advantages of an Episode of Care Approach

The Committee identified several advantagesto using the episode of care approach toassess performance. First and foremost, thisapproach offers a more patient-centered wayto evaluate health system performance, and itmay, therefore, help to address many of the

gaps in our current performance measurementsystem that were identified by the IOM16 andothers. Because the natural trajectory of manyepisodes extends over a long time period(e.g., one year), evaluation can provide insightinto how effectively services are coordinatedacross multiple settings and during criticaltransition points, such as discharge from thehospital to the nursing home, where the evidence shows that errors and miscommunications

are likely to occur.17,18 A longitudinal, episode-based approach contrasts sharply with currentapproaches to performance assessment, whichusually focus on a specific setting or provider(e.g., hospital or nursing home) and on a singlepoint in time. A longitudinal approach tomeasurement can help to shift the focus awayfrom how individual providers act to how

multiple providers can more effectively work

together to improve the quality, cost, andoutcomes of care. The IOM called formeasurement approaches that foster sharedaccountabilitywhere all members of theteam are held accountable for high-qualitycare and for the warranty many are callingon the delivery system to provide to patients.19

Second, the Committee considered theepisode of care approach to be a way to shiftperformance measurement toward assessments

that allow judgments to be made about valueby providing measures of quality, cost of care,and outcomes that can only be interpreted inlight of concordance with patients well-informedpreferences. While they may serve as indicatorsof over-utilization if appropriately benchmarked,traditional measures of resource use that focuson the volume of services received by a definedpopulation (e.g., practitioner office visits,hospital admissions, and surgical procedures)

provide an incomplete picture of how medicalservices relate to one another, and they provideno insight into the relationship between thedelivery of service(s) and the outcome achievedonce the decision to provide particular treatment(s) has been made. Conversely, if theepisode of care is the unit of analysis, theentire set of interrelated services involved inthe delivery of medical care to treat a specificproblem over time can be captured, as well

as the results achieved through the delivery ofthose services. Therefore, focusing on episodesof care allows for a more direct assessment ofthe linkage between the provision of specificservices (and their costs) and the outcomes ofthose services.

Third, it was believed that episodes canfoster and enable new strategies for financing
http:///reader/full/patients.19http:///reader/full/patients.19http:///reader/full/patients.19


19/50



healthcare that may eliminate current incentives

to overuse certain services (e.g., imaging forlow back pain) and underuse others (e.g.,preventive care such as colon cancer screening).An episode approach can also facilitate thedevelopment of alternate payment models thatcompensate processes of care that have beenshown to contribute to better patient outcomes(e.g., patient self-management support andmedication reconciliation),20,21 as well asequitably reward all healthcare professionals

who deliver care across the episode.Finally, at least in theory, an episode

approach based on prolonged episodes (oneyear or more) can provide more generalizableinsights into the overall performance ofdelivery systems. Patients with diabetes, heartattacks, cancer, depression, or other seriouschronic conditions will tend to experience otheracute or chronic conditions during their periodof follow-up. Whether the outcomes and cost

of care over time for different conditions arehighly correlated will thus be an importantempirical question. If overall performance onone condition predicts longitudinal performanceon another, then a measurement frameworkthat focuses on important, high-prevalencetracer conditions might be generalized toprovide meaningful comparisons acrossdelivery systems, communities, and regions.

Limitations of an Episode of Care ApproachDespite the advantages enumerated above, theCommittee recognized the limitations associatedwith evaluating efficiency across episodes.These stem mainly from challenges entailed in:1) addressing appropriateness of care; 2) risk-adjusting for different populations; 3) sorting

out patients with multiple chronic conditions;

and 4) facilitating comparisons amongorganizations.

Determining the appropriateness of care iscritical from two standpoints. First, it is importantto ensure that patients receive evidence-basedinterventions for which they are eligible (evenif they cost more). For example, screening forbreast or colorectal cancer may incur someearly costs, but the potential benefits from earlyintervention with regard to patient outcomes

and cost savings are often not realized untilyears downstream. Conversely, it would bewrong to label a provider as efficient forperforming a procedure at low cost and witha good outcome if the procedure should nothave been performed in the first placeeitherbecause it was not clinically indicated or, ifclinically indicated, the patient would not havechosen to receive it (based on personal values)if fully informed of the risks and benefits. An

example of the latter would include manyinstances of surgery for low back pain.

A recent analysis by the Medicare AdvisoryPayment Commission (MedPAC) of twoepisode grouper tools provides an example ofthe challenges inherent in measuring efficiencywith current commercial measures. MedPACfound that Miami, Florida, appeared to bemore efficient than Minneapolis, Minnesota, interms of relative resource use per episode forcoronary artery disease (CAD).22 However,upon further investigation it was discoveredthat Medicare beneficiaries in Miami werediagnosed and treated much more frequentlyfor CAD and other heart-related episodes thanthose in Minneapolis. Thus, Miamis CADexpenditures alone were spread across more


20/50



episodes, giving the appearance of lower

costs and greater provider efficiency butmasking the possibility of overdiagnosis andovertreatment.

An important lesson from the examplesabove is that episode groupers do not necessarily distinguish the appropriateness of clinicalservices and patient preferences for the clinicalservices rendered, and therefore efficiencymeasurement based purely on episodes mustbe balanced with population-based, per capita

resource use measures.23 Other options wouldbe to measure the number of episodes percapita of a given type of service or to measurethe degree to which care is aligned with well-informed patients preferences.

Two other limitations also deserve mention.One is that episodes of care traditionally havebeen constructed on a condition-by-conditionbasis. This is not a patient-centric approach,

because the majority of patients cope withmore than one chronic condition. Therefore,capturing quality and cost of care for patientsthat account for multiple comorbidities andthus overlapping episodes is a methodologicalchallenge that still needs to be resolved inorder to create a patient-centric measurementframework. Another limitation is that theepisode of care strategy does not facilitatecomparisons of relative efficiencies of one

organization versus anotherefficiencies thatmay indeed transcend diseases. This limitationemphasizes the need for at least some institution-specific or site-specific measurement to supportquality improvement and patient choice.

The limitations of an episode of careapproach clearly point to the need for acomprehensive measurement system that not

only can accurately and reliably assess the

efficiency of care delivered but also canmonitor for appropriateness of care; accountfor patient preferences; address diversepopulations and those with multiple chronicconditions; and allow for meaningful comparisons across organizations to support qualityimprovement and patient values. Indeed, oneof the major obstacles to adopting an episodeof care model is the translation of theory intopracticeas the Committee recommends the

ability to measure outcomes not only in theshort term (e.g., 30-day mortality) but alsoover extended intervals (e.g., initially at one

year, and then three to five years and beyond),which has yet to be successfully mastered ona wide scale basis.

The following discussion presents theCommittees conceptualization of an episodeof care model that can be applied broadly toacute and chronic conditions. This generic

episode follows a patient through the experienceof care, followed by discussion of the specificdomains of measurement proposed by theCommittee.

Generic Episode of Care

Figure 1 illustrates a generic model for trackinga patient with a health problem and formeasuring and evaluating the core components,

or phases, of the episode of care. This modelis adaptable to multiple types of episodes, andit provides a foundation for working out casescenarios. It incorporates the measurementdomainspatient-level outcomes, cost andresource use, and processes of carethat areessential for evaluating the efficiency of care,while recognizing that judgments of value
http:///reader/full/measures.23http:///reader/full/measures.23http:///reader/full/measures.23


21/50



Figure 1: Generic Episode of Care

Population at Risk

PHASE 1

Clinical episode begins

End of Episode

Risk-adjusted health outcomes(i.e., mortality & functional status)

Risk-adjusted total cost of care

PHASE 2

PHASE 3

Evaluation & Initial

Management Follow-up Care

Appropriate Times Throughout Episode

Determination of key patient attributes forrisk-adjustment

Assessment of informed patient preferencesand the degree of alignment of careprocesses with these preferences

Assessment of symptom, functional, andemotional status

Time

require an understanding of patient prefer-encesthus highlighting the importance ofactively engaging patients in their healthcaredecisions. Appendices D and E discuss theapplication of this model to two commonconditions, acute myocardial infarction (AMI)and low back pain, respectively.

Because a growing body of evidence

alerts us to the potentially harmful, unintendedconsequences of measurement approaches thatfocus too narrowly on guideline adherence forindividuals with multiple chronic conditions,24,25

this model should be viewed as the first stageof an evolutionary strategy that will be informedby emerging evidence. The complexity of illness,the clustering of illnesses (e.g., a diabetic

patient with heart disease, back pain, anddepression), and other health risks (e.g., riskfor falls, delirium, or pressure ulcers) willrequire over time a more sophisticated modelthat allows for the measurement of how wellclinical services are aligned with patientneeds, preferences, and social supports.

The generic model for evaluating the

efficiency of care over time consists of threephases: 1) population at risk, 2) evaluationand initial management, and 3) follow-upcare. These three phases are consideredfoundational to any assessment of efficiencyregardless of the type of health problempresentedacute, chronic, or a combinationthereof.


22/50



Phase 1: Population at Risk

The Committee acknowledged that primaryprevention of a health problem, both in ageneric sense for this model and more specifically for an acute condition with long-termchronic implications such as heart attack, isthe most efficient approach to carefrom thepatients perspective and with respect to societalresources. Thus, the Committee believed thata strategy for evaluating health promotionand primary prevention may require sampling

from a population of patients other than thosewho have started in the clinical phase of theepisode of care.

Phase 2: Evaluation and Initial Management

This phase begins at the onset of clinical illness,and it should include an initial assessment ofan informed patients preferences with regardto the available treatment options and, if

warranted, palliative care. An assessment ofthe key, evidence-based processes of care anda determination of how well the processes alignwith the patients preferences should occur atappropriate times during this phase. For somemore complex episodes, it may be informativeto ascertain relevant measures such as qualityof life, functional status, morbidity/ mortality,patient experiences with care, and costs.

Phase 3: Follow-Up CareMeasurement during this phase should focuson seamless care coordination, targetingcare transitions and medication reconciliation,particularly during hand-offs between providersand across different settings. Initiation ofappropriate secondary prevention also figuresprominently in this phase. Patient preferences

should continue to inform which processes

of care are executed. Ongoing evaluation ofrelevant patient outcomes may be appropriate.

Assessment at the End of the Episode of Care

At the end of the episode of carethe lengthof which will vary depending on the type ofhealth problemtwo key outcomes shouldbe assessed: 1) patient-level outcomes and2) overall resource use. Patient-level outcomesideally should include risk-adjusted morbidity

and mortality and domains that encompasshealth-related quality-of-life measures such aspatient self-reported functional status. Resourceuse comprises the risk-adjusted total cost ofcare across the entire episode. Risk adjustmentshould use a well-tested and validated statisticalmodel to account for the key patient demographic and clinical factors that affect outcomes. The Committee recommended that forchronic conditions the initial standard should

be to evaluate patient-level outcomes and totalcost of care at one year, and, when feasible,at three to five years or beyond.

Domains

The following domains represent the essentialcomponents and subcomponents for measuringefficiency as it relates to an episode of care.All of these domains are important, because

there is a need for a complement of measuresthat will pull providers together across the fulltrajectory of a particular episode, which willvary. A focus on longitudinal efficiency impliesoverall assessment of both quality and cost.Clear insights will require measuring the keycomponents needed by stakeholders so theycan judge quality, cost, and value.


23/50



In selecting the domains, the Committee

strove for balancehoping to ensure that theywere comprehensive enough to accurately andfairly evaluate performance both at the systemand provider level without imposing an undueburden for data collection. In keeping with theoriginal statement of purpose, it was also criticalthat the domains capture patient-level outcomesand that they be capable of detecting waste inthe system or of exposing unjustifiable costs.Therefore, the Committee recommended three

overarching domainspatient-level outcomes,cost and resource use, and processes ofcarefor assessing efficiency, and ultimatelyvalue, across healthcare episodes.

Domain 1: Patient-Level Outcomes

Drawing on the earlier work of Donabedian,the Committee recognized the utility ofunderstanding measurement in terms of theclassical structure-process-outcome triad.26 The

Committee also acknowledged that the currentrepertoire ofperformance measures consistspredominately of process measures (e.g.,administration ofaspirin after AMI); far fewerstructural measures (e.g., nurse staffing levels);and only a sprinkling of outcome measures(e.g., 30-day mortality for AMI). Nonetheless,the Committee advocated for collecting andreporting patient-level outcome measures, andit viewed doing so as an essential component

of any efficiency assessmentwhether at theoverall healthcare system level or at the individual provider level. Hence, there is an urgentneed for additional outcome measures to bedeveloped, tested, and more fully implemented.In the meantime, the selection of existingprocess/structure measures for purposes ofaccountability and quality improvement should

be guided by an evidence base that sufficiently

links the measures to desirable patient outcomesand that results in as parsimonious a set ofmeasures as possible to ensure adequatebreadth and high compliance.

Health Status/Health-Related Quality of Life

In addition to evaluating outcomes suchas morbidity and mortality for a particularcondition or treatment intervention, otherimportant constructs to be measured in this

domain include patient self-perception ofhealth status, functional status, and physicaland psychological health. These variablesrepresent key dimensions of health-relatedquality of life (HRQoL), which is a more relevant and appropriate construct for chronicallyill populations because of its focus on theaspects of an individuals overall well-beingthat are affected by progressive changes inhealth status and the quality of healthcare.

The ability to cope with functional deficitsand views about the meaning of ones life isprominent among the factors that can significantly affect an individuals perceptions ofhealth status and quality of life. These perceptions also can be influenced by the quality ofhealthcare services.27,28,29 HRQoL is consistentwith the current emphasis on person-centeredcare (healthcare responsive to the personswants, needs, and preferences), a principle

stressed in recent IOM reports on quality.30

Data on the multiple domains of HRQoL caninform decisions about the use of innovativeclinical practices or technologies for this population. Recent studies have demonstrated thesignificance of HRQoL measures in selectingamong alternative interventions and in guidingdecisionmaking when there is a real tradeoff
http:///reader/full/triad.26http:///reader/full/triad.26http:///reader/full/quality.30http:///reader/full/quality.30http:///reader/full/triad.26http:///reader/full/quality.30


24/50



between length and quality of life.31 Many

standardized survey instruments have beendeveloped to measure these dimensions ofoutcome.32

Patient Experience with Care

Patient experience with care over an episode-based encounter is an equally important outcome for determining the overall efficiency ofcare delivery. Therefore, feedback should besolicited from patients and their families in a

formal and systematic fashion and then actedupon. One such mechanism for doing so isthe administration of a survey tool such as theCAHPS family ofinstruments.33

Domain 2: Cost and Resource Use

The most important measure of resource usewould be one that captures the total cost ofcare across the episodeboth the quantity ofservices provided to patients and the true costspaid for each service. Because prices paid foridentical services can vary across geographicareas and among payers, and because individual providers may not have control over allprices, measures of overall resource use basedon both the actual prices paid and standardized prices are important. Other measures thatcan help provide insight into differences inresource use include the volume of services,such as the quantity of physician visits and

the number of hospital or nursing home daysprovided to patients during the episode.Another measure of resource whose usefulnessis supported by a growing evidence base arenursing intensity weights, which are relativevalues that reflect the quantity and types ofnursing services provided to patients in each

diagnosis-related group.34 It is also important

to consider opportunity costs to patients resulting from inefficiencies in the healthcare system(e.g., time wasted waiting for an appointment)as well as out-of-pocket costs for care.

Domain 3: Processes of Care

This domain includes measures of process,such as administration of an evidence-basedbundle of appropriate medications at admission

and discharge for patients who have had aheart attack. To minimize the data collectionburden, process measures should be stronglylinked to desired intermediate and finaloutcomes.

Also within this domain is the importantprocess of engaging patients proactively inshared decisionmaking. This collaborativeprocess between patients and their providersis specifically designed to assist them in

making informed choicesaligned with theirpreferences and valuesregarding potentialtreatment options. Measuring decision qualitybecomes particularly relevant in determiningthe appropriateness of care, as discussedearlier, because many patients when givenbalanced information will choose not to havea given procedure or diagnostic test, even ifthey meet eligibility and appropriatenessrequirements.35

Guiding Principles

The following principles are intended to guidethe development and implementation of theCommittees recommended measurementframework as applied across episodes of care.
http:///reader/full/outcome.32http:///reader/full/outcome.32http:///reader/full/instruments.33http:///reader/full/instruments.33http:///reader/full/group.34http:///reader/full/group.34http:///reader/full/requirements.35http:///reader/full/requirements.35http:///reader/full/outcome.32http:///reader/full/instruments.33http:///reader/full/group.34http:///reader/full/requirements.35


25/50



Principle 1: Efficiency measurement is

multidimensional.Judgments about efficiency should be basedon a comprehensive set of measures thatadequately portray performance in threedomains: patient-level outcomes, cost andresource use, and processes of care. Explicitly,these domains should drive toward outcomesrelevant to all the IOM aims for the healthcaresystem in addition to efficiency includingsafety, timeliness, effectiveness, equity, andpatient-centeredness. Therefore, efficiency

measurement is by definition multidimensionaland as such should be evaluated accordingly.

Principle 2: The choice of measures toinform judgments on efficiency shouldinclude consideration of potentialleverage.

When choosing among potential measures,consideration should be given to those thathave the highest likelihood of positivelyinfluencing desirable patient outcomes at

reasonable costs and that offer the greatestopportunity to spur system-level improvement.Process measures should have strong evidenceof their link to outcomes.

Principle 3: Measures used to informjudgments on efficiency shouldpromote shared accountability acrossproviders and should be assigned tothe smallest unit of accountability astechnically feasible.

To promote shared accountability for patientoutcomes and total costs across episodes, theframework for efficiency measurement shouldaddress all levels within the healthcare system,including individual patients, independenthealthcare professionals, provider organizations, and communities. When feasible, andwhen data can support measures that arevalid, accurate, and reliable, the smallest unit

of accountability should be measured and

reported.

Principle 4: Measures used to informjudgments on efficiency shouldrespond to the need to harmonizemeasurement across settings of care.

For existing measures, efforts should be madeto reconcile measurement specifications/definitions among healthcare professionals(e.g., physicians, nurses) and across multiplesettings (e.g., ambulatory, hospital, nursing

home, home health, community, populations).The development of new measures shouldstrive for harmonization and should includemeasures that cross settings of care.

Principle 5: Measures to informjudgments on efficiency should beused for benchmarking.

When assessing efficiency of care either atthe individual healthcare professional, providerorganization, or system level, performance

should be compared to, or indexed against, anappropriate benchmark. Whenever possible,benchmarks should reflect the current assessmentof best attainable care (based on both qualityand cost)not simply average performanceand should be tracked over time.

Principle 6: Public reporting ofmeasures of efficiency should bemeaningful and understandable toconsumers and entities accountable

for their care.Publicly reported data on efficiencyqualityand cost of careshould be meaningful anduseful to consumers, accountable care entitiesbeing measured, and other relevant stakeholders. Data should be presented in a format thatis understandable to consumers and otherend users so they can easily make informedjudgments about both providers and treatments.


26/50



Principle 7: Inappropriate care cannot

be efficient.Measures to inform judgments on efficiencyshould be capable of detecting misuse, overuse,and underuse of care within the episode time-frame. Inappropriate care, including failing toprovide an evidence-based intervention to aneligible patient or administering an interventionthat is unwarranted, cannot be efficient.

Principle 8: The measurementframework should achieve its intended

purpose and should be monitored forunintended consequences.

A measurement framework that is designed toinform judgments on efficiency should facilitateimproving health and reducing the cost andburden of illness. As such, the frameworkshould be periodically (every three to five

years) evaluated to ensure its effectiveness,and it should be continuously monitored tosafeguard against unintended consequences.

Principle 9: Measures to informjudgments on efficiency should be anintegral part of a continuous learningsystem.

In addition to assessing individual healthcareprofessionals, provider organizations, and system performance, efficiency measurement alsoshould be designed for continuous learning toinform clinical practice, measure development,policy, and the research agenda.

Path Toward aComprehensiveMeasurement SystemMany stakeholder groups have spent a considerable amount of time identifying the bestavailable measures, and they have engaged insubstantial and well-intentioned efforts to collectthe data needed to populate the databases for

these measures and to generate public reportson clinical aspects of the quality of care and,in some instances, patient perspectives on care.Until recently, with the exception of a handfulof innovators in the field, too few attemptshave been made to define and measureefficiency in ways that combine measuresof resource use with measures of other important dimensions of quality including safety,timeliness, effectiveness, equity, and patientcenteredness. The work of this Committee ismeant to help advance the field by highlightingthe need for a measurement framework thatrecognizes the longitudinal nature of health-care (i.e., episodes of care) and placesemphasis on both quality and cost.

The Committee recognized that there aremany hurdles to measuring efficiency acrossextended episodes of care. First, a morecomprehensive measurement system needs to

evolveone that is capable of assessing theclinical efficiency of care as well as monitoringfor unintended consequences, supportingcontinuous quality improvement, adequatelyadjusting for risk, and determining the appropriateness of care. However, the lack of suchan ideal system should not prevent us frommoving forward. Rather, the framework can


27/50



serve as a blueprint for what we are trying

to achieve. The Committee recognized thatachieving the goal of a comprehensivemeasurement system will require adopting aphased approach to implementation.

Second, the Committee recognized thatmany technical issues will need to be resolvedincluding:

how to ensure data integrity;

how data should be aggregated and at

what level (i.e., national versus local); how best to adjust for varying case/severity

mix among different providers;

how to attribute care across multipleproviders; and

how to develop new measures to fill gapsin the episode of care framework.

Third, the Committee realized that changingperformance measurement to a patient-focused,episode of care approach, as described, willbe difficult, particularly because currentaccountability and payment systems tend tofocus on individual providers and distinctsettings of care.

Although a full discussion of these issues isbeyond the scope of this report, the Committeediscussed these and other challenges, and italso held a workshop to solicit input on theseissues from additional content experts in the

field. Based on these discussions, theCommittee agreed that the path toward implementation will require efforts to integrate andcoordinate local providers and their data, andit will necessitate the development of practicalapproaches to support the collaborative andintegrative work required to improve careacross episodes that span multiple providers

and diverse settings of care. To achieve these

goals an organizational context for the workmust be established. The Committee believedthat this is likely to require the fostering ofshared accountability through the developmentof accountable care entities to provide (oreffectively manage) the continuum of care as areal or virtually integrated local delivery system.The Committee recommended that such anorganizational structure have the followingcharacteristics:

organizational commitment to patient-centered focus for measurement includingroutinely collecting and acting upon patientand family caregiver feedback;

strong organizational structure with a locusof accountability with clear authority, and areceptive environment with financial incentives, collaboration, and communicationwithin and across providers and settingsof care;

information management systems thatinclude seamless information exchangeacross providers and settings; and

learning systems with a quality improvementinfrastructure that can engage providersand drive improvement with mechanisms forinnovation and learning within and acrossorganizations.

This organizational structure could takemany forms. For example, it could be an

existing integrated delivery system such asthe Geisinger Health System, described inAppendix A. Other alternative structuresmight include multispecialty group practices,physician-hospital organizations/practicenetworks, regional collaboratives, or healthplans serving as the locus of accountabilityfor a network of providers, or conceivably an


28/50



advanced medical home with a designated

health professional playing a coordinating rolefor his or her patients.36 All of these potentialstructures have strengths and weaknesses andundoubtedly would face formidable barriers incurrent markets, but the Committee believedthat some level of organizational accountabilitywill almost certainly be required and thatlearning which organizational forms aremost successful (in what contexts) will requireboth pilot testing and careful evaluation of

alternative approaches.

In closing, the measurement frameworkpresented here is viewed as a living documentthat will continue to evolve as we learn more

about how to adapt this model to individuals

with complex chronic illnesses, including frailelders, and to population-based approaches.The Committee recognized that there are manyother determinants of health that necessitatebetter integration between the public healthand personal health systems, as well as compounding issues around access and disparities,that warrant further work on fully incorporatingthese areas into a farther-reaching performancemeasurement system. Nonetheless, this frame

work provides a starting point for identifyingmeasurement gaps and for examining modelsof shared accountability to help move us closerto attaining a high-performing, high-valuehealthcare system.
http:///reader/full/patients.36http:///reader/full/patients.36http:///reader/full/patients.36


29/50

National Quality Forum A-1


Appendix ACase Studies

The 100,000 Lives CampaignTHE 100,000 LIVES CAMPAIGNspearheaded by the Institute for HealthcareImprovementengaged more than 3,000 hospitals in an 18-month project during whichan estimated 122,000 deaths were avoided by improving the quality and efficiency ofcare. Healthcare organizations were encouraged to implement up to six interventionsincluding 1) deploying rapid response teams in inpatient settings at the first sign of patientdecline; 2) delivering a core, evidence-based bundle of clinical services to patients whoexperienced a heart attack; 3) reconciling medications throughout care to avert adversedrug events; 4) preventing central line infections; 5) preventing surgical site infections; and

6) preventing ventilator-associated pneumonia (VAP).37

Virginia Mason Medical Center (VMMC) employs tenets derived from Toyotas LeanProduction System, which maps out processes of care in step-by-step detail to look foropportunities to eliminate waste. VMMC chose the reduction of the incidence of VAP and itsrelated complications from the 1000,000 Lives Campaign interventions as one target, andit implemented VAP care bundles, which consisted of four relatively simple interventions,such as elevating the head of the bed. As a result of these actions, VAP decreased from40 patients per year in 2000 to 5 per year in 2006, with an estimated savings to theinstitution of $1.7 million.38

Geisinger Health SystemTHE GEISINGER HEALTH SYSTEM (GHS) in central Pennsylvania has a pilot programnamed ProvenCare that offers a 90-day guarantee on the heart bypass surgeries thatits physicians perform.39 The program consists of 40 steps that are designed to ensurethat best practices are followed during both pre- and post-operative care and to eliminate
http:///reader/full/million.38http:///reader/full/million.38http:///reader/full/perform.39http:///reader/full/perform.39http:///reader/full/million.38http:///reader/full/perform.39


30/50

National Quality ForumA-2


variation across physicians and facilities.

Since the inception of the program in February2006, preliminary findings suggest thatpatients have spent fewer days in the hospital,have been less likely to be readmitted forsurgery-related complications, and havereturned to their homes after discharge asopposed to making an intermediary stop ata nursing home. To avoid the pitfalls of thecurrent fee-for-service financing system thatrewards more procedures and trips back to the

hospital, GHSs network of hospitals charge

insurers a flat fee for the surgery and half theamount of the cost of care related to the surgerythree months after discharge based on historicaldata. No charges are incurred for follow-uptreatment beyond this amount. Thus, the incentive exists to administer high-quality care asefficiently as possible to avoid preventablereadmissions and to stay within the capitatedallowance.


31/50

National Quality Forum B-1


Appendix BMeasuring Efficiency Across Patient-FocusedEpisodes of Care Steering Committee

Elliott S. Fisher, MD, MPH (Co-Chair)Director, Center for Health Policy Research,

Dartmouth Institute for Health PolicyProfessor of Medicine and Communityand Family Medicine, Dartmouth MedicalSchool and Senior Associate, VAOutcomes Group, White River Junction

Hanover, NH

Kevin Weiss, MD, MPH (Co-Chair)President and Chief Executive Officer,

American Board of Medical Specialties

Professor of Medicine and Director,Institute for Healthcare Studies,Northwestern University FeinbergSchool of Medicine

Chicago, IL

Lawrence BeckerDirector, BenefitsXerox CorporationRochester, NY

Robert O. Bonow, MD, FACCProfessor, Medicine and Chief, Division

of CardiologyNorthwestern University Feinberg School

of MedicineChicago, IL

Carolyn M. Clancy, MDDirectorAgency for Healthcare Research and QualityRockville, MD

Franois de Brantes, MS, MBAChief Executive OfficerBridges to Excellence and National

Coordinator, PROMETHEUS Payment

Newtown, CT

Joyce DubowAssociate Director, Public Policy InstituteAARPWashington, DC

William E. Golden, MD

Professor of Medicine and Public HealthUniversity of Arkansas for Medical SciencesLittle Rock, AR

Sam Ho, MDExecutive Vice President and Chief Medical

Officer, Pacific and Southwest RegionsUnitedHealthcareCypress, CA

David S. P. Hopkins, PhDDirector of Quality MeasurementPacific Business Group on Health

San Francisco, CAGeorge J. Isham, MD, MS

Chief Health Officer and Plan MedicalDirector

HealthPartners Inc.Bloomington, MN

Christine IzuiExecutive Director of Quality Initiatives,

Office of Clinical AffairsBlueCross BlueShield AssociationChicago, IL

Paul E. Jarris, MD, MBAExecutive DirectorAssociation of State and Territorial Health

OfficialsWashington, DC

Robert M. KrughoffPresident, Center for the Study of ServicesConsumers CHECKBOOKWashington, DC

Lindsay Martin, MSPHSenior Research AssociateInstitute for Healthcare ImprovementCambridge, MA

Mary D. Naylor, PhD, FAAN, RNMarian S. Ware Professor in GerontologyUniversity of Pennsylvania School of

NursingPhiladelphia, PA

Margaret E. OKane, MHSPresidentNational Committee for Quality AssuranceWashington, DC


32/50

National Quality ForumB-2


Rebecca M. Patton, MSN, RN, CNOR

PresidentAmerican Nurses AssociationSilver Spring, MD

Jonathan B. Perlin, MD, PhDPresident, Clinical Services and Chief

Medical OfficerHCA, Inc.Nashville, TN

Christopher J. Queram, MAPresident and Chief Executive OfficerWisconsin Collaborative for Healthcare

QualityMadison, WI

Mark C. Rattray, MDFounder and PresidentCareVariance, LLCEdmonds, WA

Cary Sennett, MD, PhDSenior Vice President, Strategy and

Clinical AnalyticsAmerican Board of Internal MedicinePhiladelphia, PA

Thomas Valuck, MD, JDMedical Officer and Senior Advisor, Center

for Medicare ManagementCenters for Medicare & Medicaid ServicesBaltimore, MD

Margaret VanAmringe, MHS

Vice President, Public Policy andGovernment RelationsThe Joint CommissionWashington, DC

James Weinstein, DO, MScDirector, Dartmouth Institute for Health

Policy and Clinical Practice, DartmouthCollege

Chair, Department of Orthopaedic Surgery,Dartmouth Medical School

Hanover, NH

Liaison MembersRebecca Hayes (Liaison to theSteering Committee for the AQA)

Senior Research AssociateAmericas Health Insurance PlansWashington, DC

Nancy Foster (Liaison to the SteeringCommittee for the Hospital QualityAlliance)

Vice President for QualityAmerican Hospital Association

Washington, DC

NQF STAFFJanet M. Corrigan, PhD, MBA

President and Chief Executive Officer

Karen Adams, PhDVice President, National Priorities

Anisha S. Dharshi, MPHProgram Director, National Priorities


33/50

National Quality Forum C-1


Appendix CApplication of Key Terms and Definitions: Case Scenario

THE FOLLOWING PRESENTS an exercise in which the concepts of cost and quality as

measures of efficiency can allow for judgments of the value of care delivered to patients.Table 1 shows that the average cost of care per diabetic episode for Physician 1 is$1,500. What this means, quite simply, is that the average reimbursable expense (forservices billed to an insurer) for diabetes care delivered to Physician 1s panel of patientsover some time interval was $1,500. One might argue that this amount does not trulycapture the total cost of care for those diabetics, but that argument is about the validity ofthe measure and not about the underlying construct. So, for the moment, let us proceed.

Table 1

PHYSICIAN AVERAGE COST PER DIABETIC EPISODE

1 $1,500

Table 2 shows that the cost of diabetes care for Physician 2 is $1,250significantly lessthan that for Physician 1. It would be accurate and meaningful to state that Physician 2 provides diabetes care (on average) at a lower cost than does Physician 1. But the inferencesone can make from that statement are limited; it is a matter of fact, but it provides insufficientbasis for a decision about which physician is more efficient.

Table 2

PHYSICIAN

1

AVERAGE COST PERDIABETIC EPISODE

$1,500

NOTES

2 $1,250 Physician 2 provides diabetes care at a lowercost than Physician 1.


34/50

National Quality ForumC-2


Table 3 provides critical additional information about the quality of care that Physicians 1 and 2

provide. They have very different rates of success with respect to achieving an outcome (HgbA1ctarget) relevant to (that is, an indicator for) a health benefit. It is accurate to suggest that Physician1 provides higher quality diabetes care40 than does Physician 2 (just as it was accurate to suggestthat Physician 2 provides lower cost care).

Table 3

PHYSICIAN

1


$1,500

PERCENTAGE OF PATIENTSAT HgbA1c TARGET

80

NOTES

2 $1,250 60 Physician 1 provides higherquality diabetes care thanPhysician 2.

Does that mean that Physician 1 is more efficient?Using the definition of efficiency adopted bythe Committee, the answer must be we dont know. Physicians 1 and 2 do not achieve the sameoutcome (so we cannot use their relative costs to make an inference about efficiency), and they donot achieve the different outcomes at the same cost (so we cannot use their relative outcomes tojudge efficiency).

Fortunately, we meet Physician 3 in Table 4. Physician 3 achieves the same outcome asPhysician 1 but at a lower cost. Physician 3 provides care at the same cost as Physician 2 butachieves a better outcome. So Physician 3 is clearly more efficient than Physicians 1 and 2. Whilewe can make no comment about Physicians 1 and 2, we have the data that we need to make animportant observation that is relevant to our understanding of the performance of all physicians inthe set.

What if Physician 3s outcome was better than Physician 1s (e.g., 90 percent) or was achievedat a lower cost than Physician 2s (e.g., $1,000 per episode)?It is no longer strictly true thatPhysician 3 achieves a better outcome than Physician 1 at the same cost (or achieves the sameoutcome as Physician 2 at a lower cost). Are we unable to comment on efficiency because there is

no strict equality?


35/50

National Quality Forum C-3


Table 4

PHYSICIAN

1


$1,500

PERCENTAGE OF PATIENTSAT HgbA1c TARGET

80

NOTES

2 $1,250 60

3 $1,250 80 Physician 3 is more efficientthan Physician 1. (Physician 3provides same quality at lowercost) and Physician 3 is moreefficient than Physician 2

(Physician 3 provides higherquality at same cost).

Of course not. If one physician achieves a better outcome at a lower cost than another, thatphysician is clearly more efficient. This is the strategy behind most current efforts to assess efficiencyto array physicians on cost and quality axes and to look for those in the northwest corner(that is, those whose quality is above average and whose cost is below average).

Table 5 provides important additional information about Physicians 1, 2, and 3. We are remindedby the data in Table 5 that quality is multidimensional, that is, that consumers value the quality of

their experience as well as the physiologic outcome associated with clinical encounters.

Table 5

AVERAGE COST PER PERCENTAGE OF PATIENTS PERCENTAGE OF PATIENTS VERYPHYSICIAN DIABETIC EPISODE AT HgbA1c TARGET SATISFIED WITH COMMUNICATIONS

1 $1,500 80 90

2 $1,250 60 90

3 $1,250 80 60

Table 5 shows that Physicians 1, 2, and 3 achieve different outcomes with respect to theirpatients assessments of the quality of their communication skills. Inferences about efficiency are nolonger quite so clear. On the other hand, it may be possible to make some statements about value.Which physician offers care that represents the highest value?The Committees answer: Thatdepends. In particular, it depends on the rate at which the potential receiver of services trades offcost for one or the other outcome (clinical/physiologic or experiential) or the rate at which thereceiver trades off those outcomes themselves.


36/50

National Quality ForumC-4


These, in turn, depend on individual prefer

ences. The high-income patientor the patientwith deep insurancewho is committed to thecontrol of his diabetes may gladly trade off theadditional cost he faces for the better outcomethat Physician 1 achieves. To that patient, thiscost-quality combination represents high value.

Not so, perhaps, to the patient who is lessconcerned about his diabetes and/or who canless afford the cost differential. And not so,perhaps, to the employer, who may face the

prospect of cutting back on the benefits hercompany can offer if that company has toface the costs that attend Physician 1s care.From this employers perspective, Physician 3may represent the highest valueacceptableclinical outcomes at a lower cost. Or it may

be Physician 2, if that employer believes that

a level of satisfaction of 60 percent would beseen as unacceptable (or at least perceivedvery negatively) by employees and theirfamilies.

The point simply is this: The value eachphysician/each cost-quality combinationrepresents depends on the preferences of theindividual/stakeholder making the valuation.Different people in different settings (andconceivably at different times in their lives) will

value these combinations differently. Efficiencyis objective; when measured as the Committeesuggested, all observers would agree onrank order. The inputs of cost and quality tovalue are objective, but the determination itselfis not.


37/50

National Quality Forum D-1


Appendix DContext for Considering an AMI Episode

THE COMMITTEES CURRENT EFFORTS to define episodes of care and to establish a

performance measurement framework must be understood in the context of a relatively simple goal: to improve health and reduce the burden of illness while maximizing the value ofindividual and societal resources allocated to healthcare. This goal implies a responsibilityto consider the definition of an episode within a broader context. In the case of acutemyocardial infarction (AMI), for example, a responsible evaluation of a healthcare deliverysystem should consider the efficiency with which each patient with an AMI received careand the frequency with which AMI occurred in the community. Decisions about investmentsintended to improve health and reduce the burden of illness should take into account notonly improved clinical services but also strategies that could effectively prevent the acutecondition.

The Committees proposed conceptual framework and episode model for AMI (Figure D-1),however, is constrained by the realities of the current capabilities of performance measurement systems and our capacity to implement episode-based measurement systems. Thereforethe Committee decided to focus primarily on the relatively discreet definition of an AMIepisode that begins with the onset of chest pain and continues through the period that maybe required for recovery and stabilizationrecognizing the importance of the period preceding the AMI. As such, the Committee defined four distinct phases of the care of patientswith AMI that purposively correspond with the Foundation for Accountabilitys (FACCT)41

domains of consumer needs: the population at risk, acute care, post-acute care/rehabilitation,and secondary prevention. Thus, the Committee focused on the latter three phases of the

episode, because these represent the most direct, concrete, and easily measured componentsof AMI care. The population at risk is included in the episode model, because it remainsimportant to look upstream to understand and perhaps intervene to prevent AMI.


38/50

National Quality ForumD-2


Figure D-1: Context for Considering an AMI Episode

Population at Risk

1oPrevention

(no known CAD)

2oPrevention

(CAD no prior AMI)

2oPrevention

(CAD with prior AMI)Advanced Care Planning

AcutePhase

Post Acute/RehabilitationPhase

2o Prevention

PHASE 1

PHASE 2 PHASE 3 PHASE 4

Staying Healthy Getting Better Living w/ Illness/Disability (T1)

Coping w/ End of Life (T2)

Episode begins

onset of symptoms

Episode ends

1 year post AMI

Post AMI Trajectory 1 (T1)

Relatively healthy adult

Focus on:

Quality of Life

Functional Status

2o Prevention Strategies

Rehabilitation

Advanced Care Planning

Post AMI Trajectory 2 (T2)Adult with multiple co-morbidities

Focus on:

Quality of Life

Functional Status

2o Prevention Strategies

Advanced Care Planning

Advanced Directives

Palliative Care/Symptom Control

Assessment of

Preferences

The Committee first recognized that individuals with AMI tend to follow one of two trajectories that have different outcomes and differentideal patterns of carean acknowledgedsimplification but one that is useful for bothmeasurement and conceptual clarity.Individuals in Trajectory 1 (T1) are relativelyhealthy at the time of their initial MI and, if

care is effectively delivered, they should expectto return to active, productive lives followingrecovery from the MI. Those in Trajectory 2(T2) have their MI superimposed on seriousunderlying illness (i.e., multiple chronic conditions). For these patients, AMI represents anadditional (and perhaps final) assault in theirprogression toward increased frailty and

death. In either case, however, the Committeebelieved it important to identify distinct phasesof care and most importantly to assess andadhere to patients preferences.

Episode Phases

Phase 1: Population at RiskIdeally, in evaluating the performance of ahealthcare system in addressing the problemsof AMI, it would be important to consider thepopulation at risk of AMI and to capture theperiod preceding the event, when it is conceivable that the first heart attack could have been


39/50

National Quality Forum D-3


preventedeither through appropriate primary

prevention that occurs in populations with noprior evidence of heart disease or throughsecondary prevention for those with knowncoronary artery disease (CAD).

Phase 2: Acute Phase

The acute MI phase should begin with theonset of symptoms (although this will be difficultto measure in most current approaches) and

end at 30 days postindex hospital discharge.The advantage of focusing on symptom onsetlies in the opportunity it affords to address system-level interventions including the adequacyof the emergency medical response system andaccess to cardiac revascularization facilitiesin a community. Optimally, at the point ofentry, patients should be assessed as to whichtrajectoryT1 or T2they fall into (althoughat times this distinction may not be clear) so

that appropriate treatment protocols can befollowed. For patients who clearly enter underT2, it is imperative that their advance careplan be adhered to and that their preferencesbe respected. Regardless of trajectory, thisphase should capture the acute care providedto the patient from arrival at the emergencydepartment, through appropriate diagnosis,treatment, and stabilization. This would includeany initial revascularization and the appropriate

management of complications, and wouldextend through the transition to rehabilitationand post-AMI management.

Phase 3: Post-acute care/Rehabilitation PhaseThe Committee proposed that rehabilitation(Phase 3) be the focus of the episode of carefrom the end of the acute phase (Phase 2)through three months post-index hospitaldischarge (while acknowledging that post-acute care begins the day of admission andmay continue for an extended period). In T1,where the patient is relatively healthy at the

time of AMI, the focus should be on gainingmedical stability and returning to work andnormal activities of daily living. Additionally,advance care planning should be initiated.In T2, the focus also should be on achievingmedical stability through symptom control andreturning

Documents

Episodes Report Full