salzman patient selection for care management

Selecting Patients for Selecting Patients for ggComprehensive Care Comprehensive Care

M n nt PM n nt PManagement ProgramsManagement Programs

Brooke Brooke SalzmanSalzmanAssistant Professor Assistant Professor

Division of Geriatric MedicineDivision of Geriatric MedicineDepartment of Family & Community MedicineDepartment of Family & Community Medicine

Thomas Jefferson UniversityThomas Jefferson [email protected]@jefferson.edu

3/23/20123/23/2012

Obj tiObj tiObjectivesObjectives

•• BackgroundBackgroundBackgroundBackground•• Compare/contrast evidenceCompare/contrast evidence--based care based care

management programs in the U Smanagement programs in the U Smanagement programs in the U.S.management programs in the U.S.•• Discuss approaches, including predictive Discuss approaches, including predictive

d li f l i i fd li f l i i fmodeling, for selecting patients for care modeling, for selecting patients for care management programsmanagement programs

•• Discuss utilization of an officeDiscuss utilization of an office--based tool based tool

DisclosuresDisclosuresDisclosuresDisclosures

•• No conflicts of interestNo conflicts of interestNo conflicts of interestNo conflicts of interest•• HRSA GACA award, 2010HRSA GACA award, 2010--20152015

•• “Geriatrics may be thought of as the epitome of“Geriatrics may be thought of as the epitome ofGeriatrics may be thought of as the epitome of Geriatrics may be thought of as the epitome of chronic disease care. In many ways, it has chronic disease care. In many ways, it has pioneered salient approaches but much remainspioneered salient approaches but much remainspioneered salient approaches, but much remains pioneered salient approaches, but much remains to be done.”to be done.”–– Robert Kane MD What Can Improve Chronic DiseaseRobert Kane MD What Can Improve Chronic DiseaseRobert Kane, MD, What Can Improve Chronic Disease Robert Kane, MD, What Can Improve Chronic Disease

Care? J Am Geriatr Soc 2009Care? J Am Geriatr Soc 2009

Chr ni DiChr ni DiChronic DiseaseChronic Disease

•• Chronic disease is the leading cause of deathChronic disease is the leading cause of deathChronic disease is the leading cause of death Chronic disease is the leading cause of death and disability in the USand disability in the US

•• Acco nt for 70% of all deathsAcco nt for 70% of all deaths•• Account for 70% of all deathsAccount for 70% of all deaths•• 84% of total health care spending is on people 84% of total health care spending is on people

i h h i di ii h h i di iwith chronic conditionswith chronic conditions•• 145 million Americans have at least 1 CD, 2009 145 million Americans have at least 1 CD, 2009

•• Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care. 2010Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care. 2010

Chronic Disease in the ElderlyChronic Disease in the Elderly

•• Over 80% of older adults have at least 1 chronic Over 80% of older adults have at least 1 chronic conditionsconditions

•• Over 50% of older adults have at least 3 or more Over 50% of older adults have at least 3 or more chronic conditionschronic conditions

–– Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care, 2010Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care, 2010–– Guiding Principles for the Care of Older Adults with Multimorbidity: An Approach for Guiding Principles for the Care of Older Adults with Multimorbidity: An Approach for

Cli i i A i G i i S i E P l h C f Old Ad l i hCli i i A i G i i S i E P l h C f Old Ad l i hClinicians, American Geriatrics Society Expert Panel on the Care of Older Adults with Clinicians, American Geriatrics Society Expert Panel on the Care of Older Adults with Multimorbidity, draft, 2012Multimorbidity, draft, 2012

Challenges of MultimorbidityChallenges of Multimorbidity

•• Higher rates of death disability andHigher rates of death disability andHigher rates of death, disability, and Higher rates of death, disability, and institutionalization institutionalization

•• Increased utilization of health care resourcesIncreased utilization of health care resources•• Increased utilization of health care resourcesIncreased utilization of health care resources•• Decreased quality of life Decreased quality of life •• Higher rates of adverse effects of treatments and Higher rates of adverse effects of treatments and

interventionsinterventions–– Guiding Principles for the Care of Older Adults with Guiding Principles for the Care of Older Adults with MultimorbidityMultimorbidity: An Approach for : An Approach for

Clinicians, American Geriatrics Society Expert Panel on the Care of Older Adults with Clinicians, American Geriatrics Society Expert Panel on the Care of Older Adults with MultimorbidityMultimorbidity, draft, 2012, draft, 2012MultimorbidityMultimorbidity, draft, 2012, draft, 2012

C t f C rC t f C rCosts of CareCosts of Care

•• Cost of providing care for older Americans is:Cost of providing care for older Americans is:Cost of providing care for older Americans is:Cost of providing care for older Americans is:–– 33--5 times greater than for someone < 655 times greater than for someone < 65

17 times greater for someone with 5 or more17 times greater for someone with 5 or more–– 17 times greater for someone with 5 or more 17 times greater for someone with 5 or more conditionsconditions

•• By 2030 health care spending is expected toBy 2030 health care spending is expected to•• By 2030, health care spending is expected to By 2030, health care spending is expected to increase by 25% just based on the aging increase by 25% just based on the aging populationpopulationpopulationpopulation

–– Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care. Robert Wood Johnson Foundation. Chronic Care: Making the Case for Ongoing Care. 20102010

•• N d l fN d l f•• New models of careNew models of care–– Focus on chronic careFocus on chronic care–– Ensuring quality, evidence Ensuring quality, evidence

based care and safetybased care and safetyU ili h l hU ili h l h–– Utilize health care teams Utilize health care teams

–– Emphasize care Emphasize care di tidi ticoordination coordination

–– Increase engagement of Increase engagement of patients and caregiverspatients and caregiverspatients and caregiverspatients and caregivers

–– Payment for added value Payment for added value

Challenges of MultimorbidityChallenges of Multimorbidity

•• Limited evidenceLimited evidence--based resources for managingbased resources for managingLimited evidenceLimited evidence based resources for managing based resources for managing this populationthis population

•• Clinical Practice Guidelines focus on singleClinical Practice Guidelines focus on single•• Clinical Practice Guidelines focus on single Clinical Practice Guidelines focus on single diseasesdiseasesAd l i hAd l i h l i bidil i bidi f l d df l d d•• Adults with Adults with multimorbiditymultimorbidity are often excluded are often excluded from clinical trialsfrom clinical trials

Chronic Care for EldersChronic Care for Elders

G id d CG id d C•• Guided CareGuided Care•• GRACE (Geriatric Resources for Assessment GRACE (Geriatric Resources for Assessment

and Care of Elders)and Care of Elders)•• PACE (Program of AllPACE (Program of All--inclusive Care for the inclusive Care for the ( g( g

Elderly)Elderly)

6 common elements6 common elements

•• 1 Comprehensive assessment1 Comprehensive assessment1. Comprehensive assessment1. Comprehensive assessment•• 2. Patient2. Patient--centered, evidencecentered, evidence--based care planbased care plan

3 P i i i3 P i i i•• 3. Proactive monitoring3. Proactive monitoring•• 4. Communication and coordination across 4. Communication and coordination across

settings and with all care providers settings and with all care providers •• 5. Promotion of the patient’s (and caregiver’s) 5. Promotion of the patient’s (and caregiver’s) p ( g )p ( g )

active engagement in his or her healthactive engagement in his or her health•• 6 Facilitating access to resources6 Facilitating access to resources6. Facilitating access to resources6. Facilitating access to resources

Models of Comprehensive Primary Care for Older Patients With Multiple Chronic Conditions

Boult, C. et al. JAMA 2010;304:1936-1943

Copyright restrictions may apply.

M GACAM GACAMy GACAMy GACA

•• Replicate “Guided Care” in an urban academicReplicate “Guided Care” in an urban academicReplicate Guided Care in an urban, academic Replicate Guided Care in an urban, academic primaryprimary--care practice for older adults with care practice for older adults with multimordiditymultimordiditymultimordiditymultimordidity

•• Evaluate GC as an Evaluate GC as an interprofessionalinterprofessional teaching teaching model in geriatric educationmodel in geriatric educationmodel in geriatric educationmodel in geriatric education

G id d C rG id d C rGuided CareGuided Care

•• What is Guided Care?What is Guided Care?What is Guided Care?What is Guided Care?–– Intensive care management program that Intensive care management program that

targets older ad lts ith m ltiple chronictargets older ad lts ith m ltiple chronictargets older adults with multiple chronic targets older adults with multiple chronic conditions who are at the highest risk for conditions who are at the highest risk for health care utilizationhealth care utilizationhealth care utilizationhealth care utilization

–– TeamTeam--based approach with an embedded RN based approach with an embedded RN h ll b i h PCP’ bh ll b i h PCP’ bwho collaborates with PCP’s to manage about who collaborates with PCP’s to manage about

5050--60 high risk patients60 high risk patients


•• InIn--home assessmenthome assessmentInIn home assessmenthome assessment•• Care plan and Action PlanCare plan and Action Plan

M i i i ’ di i hlM i i i ’ di i hl•• Monitoring patient’s conditions monthlyMonitoring patient’s conditions monthly•• Actively promotes selfActively promotes self--managementmanagement•• Coordinates care among all providers and sitesCoordinates care among all providers and sites•• Smoothes transitions between sites of careSmoothes transitions between sites of careSmoothes transitions between sites of careSmoothes transitions between sites of care•• Educates and supports caregiversEducates and supports caregivers

F ili iF ili i•• Facilitates access to community resourcesFacilitates access to community resources


•• Benefits of GC:Benefits of GC:Benefits of GC:Benefits of GC:–– Improve patient perception of quality of careImprove patient perception of quality of care–– Improve caregiver satisfaction with careImprove caregiver satisfaction with care–– Improve caregiver satisfaction with careImprove caregiver satisfaction with care–– Improve physician satisfaction with chronic careImprove physician satisfaction with chronic care–– Produces high job satisfaction among GC nursesProduces high job satisfaction among GC nursesProduces high job satisfaction among GC nursesProduces high job satisfaction among GC nurses–– Reduced costs by reducing use of expensive servicesReduced costs by reducing use of expensive services––

•• 21 to 49 % fewer hospital readmissions21 to 49 % fewer hospital readmissionspp•• 17% fewer emergency department visits17% fewer emergency department visits•• 30% fewer home health care episodes30% fewer home health care episodes

R r h q ti nR r h q ti nResearch questions….Research questions….

•• How should we choose the patients for GC orHow should we choose the patients for GC orHow should we choose the patients for GC or How should we choose the patients for GC or other care management programs?other care management programs?

•• How do we direct the delivery of resourceHow do we direct the delivery of resource•• How do we direct the delivery of resourceHow do we direct the delivery of resource--intensive interventions involved in GC?intensive interventions involved in GC?H d i hH d i h ff i fff i f•• How do we increase the costHow do we increase the cost--effectiveness of effectiveness of the program?the program?

Selecting PatientsSelecting PatientsSelecting PatientsSelecting Patients

•• Targeting the appropriateTargeting the appropriateTargeting the appropriate Targeting the appropriate patient is key to a patient is key to a program’s effectivenessprogram’s effectiveness

•• Comprehensive Geriatric Comprehensive Geriatric Assessments (CGAs)Assessments (CGAs)–– No benefit shown in No benefit shown in

studies that included studies that included elders who were too wellelders who were too wellelders who were too well elders who were too well or too illor too ill

–– Elders at risk for Elders at risk for progressive disabilityprogressive disability

Selection Methods for CareSelection Methods for CareSelection Methods for Care Selection Methods for Care managementmanagement

•• Predictive modelingPredictive modeling•• PCP selectionPCP selection•• PCP selectionPCP selection

Pr di ti M d linPr di ti M d linPredictive ModelingPredictive Modeling

•• HCC (Hierarchical Conditions Categories)HCC (Hierarchical Conditions Categories)HCC (Hierarchical Conditions Categories)HCC (Hierarchical Conditions Categories)•• Charlson Comorbidity indexCharlson Comorbidity index•• Probability of Repeated AdmissionProbability of Repeated Admission•• Probability of Repeated AdmissionProbability of Repeated Admission•• Ambulatory pharmacy dataAmbulatory pharmacy data

VESVES 13 (V l bl Eld S )13 (V l bl Eld S )•• VESVES--13 (Vulnerable Elders Survey)13 (Vulnerable Elders Survey)•• Lee/Schonberg Mortality Index Lee/Schonberg Mortality Index

•• Differ in types of variables, types of data sources, Differ in types of variables, types of data sources, timing of data collection, what they’re predictingtiming of data collection, what they’re predicting

Predictive ModelingPredictive ModelingPredictive ModelingPredictive Modeling

•• What should we be trying to predict to targetWhat should we be trying to predict to targetWhat should we be trying to predict to target What should we be trying to predict to target the right patients who will benefit the most?the right patients who will benefit the most?

Hospital admission/readmissionHospital admission/readmission–– Hospital admission/readmissionHospital admission/readmission–– MortalityMortality

C l iC l i–– ComplexityComplexity–– CostCost–– Functional decline/frailtyFunctional decline/frailty–– Placement in longPlacement in long--term careterm care

Predictive ModelingPredictive ModelingPredictive ModelingPredictive Modeling

•• AccuracyAccuracyyy–– How well are we able to How well are we able to

predict which patients are predict which patients are likely to be readmitted?likely to be readmitted?likely to be readmitted?likely to be readmitted?

•• ReversibilityReversibility–– Are we predicting an Are we predicting an p gp g

outcome that is reversible, outcome that is reversible, or or iis it too late?s it too late?

•• PriorityPriority•• PriorityPriority–– Patients, insurers, Patients, insurers,

hospitals, providershospitals, providers

Pr di ti M d lin : V ri blPr di ti M d lin : V ri blPredictive Modeling: VariablesPredictive Modeling: Variables

M di l bidiM di l bidi•• Medical comorbidityMedical comorbidity•• Mental health comorbidityMental health comorbidity•• Prior use of medical services Prior use of medical services •• Demographics: age sex race/ethnicityDemographics: age sex race/ethnicityDemographics: age, sex, race/ethnicityDemographics: age, sex, race/ethnicity•• Illness severity: classification, lab findings or Illness severity: classification, lab findings or

clinical values (Cr Na Hgb BP)clinical values (Cr Na Hgb BP)clinical values (Cr, Na, Hgb, BP)clinical values (Cr, Na, Hgb, BP)


•• Overall health and functionOverall health and functionOverall health and functionOverall health and function–– SelfSelf--rated health, performance in ADL’s, mobilityrated health, performance in ADL’s, mobility–– Cognitive impairment, visual or hearing impairmentCognitive impairment, visual or hearing impairmentg p , g pg p , g p

•• Social determinants of healthSocial determinants of health–– Socioeconomic status insurance education marital statusSocioeconomic status insurance education marital statusSocioeconomic status, insurance, education, marital statusSocioeconomic status, insurance, education, marital status–– Caregiver availability, social supportCaregiver availability, social support–– Access to care, health literacyAccess to care, health literacyyy–– Substance and/or alcohol abuse, smoking statusSubstance and/or alcohol abuse, smoking status


•• H it l h lth t f tH it l h lth t f t•• Hospital or health system factorsHospital or health system factors–– Transitional care issues: PostTransitional care issues: Post--discharge followdischarge follow--up, up,

coordination of care medication reconciliationcoordination of care medication reconciliationcoordination of care, medication reconciliationcoordination of care, medication reconciliation–– Quality of inpatient careQuality of inpatient care–– Number of different providersNumber of different providerspp

C nfli tin D tC nfli tin D tConflicting DataConflicting Data

•• Retrospective cohort of 28 430 persons > 64 yrsRetrospective cohort of 28 430 persons > 64 yrsRetrospective cohort of 28,430 persons > 64 yrsRetrospective cohort of 28,430 persons > 64 yrs–– Unplanned Unplanned admissions in the previous year***admissions in the previous year***

•• LopezLopez--AguilaAguila et al Am J Manag Careet al Am J Manag Care 20112011LopezLopez--Aguila Aguila et al, Am J Manag Care, et al, Am J Manag Care, 20112011

•• Retrospective cohort evaluating Evercare for Retrospective cohort evaluating Evercare for persons > 65 with 2 or more hospitalizationspersons > 65 with 2 or more hospitalizationspersons > 65 with 2 or more hospitalizations persons > 65 with 2 or more hospitalizations –– Care management did not reduce admissionsCare management did not reduce admissions

G l d i d i i f 2 6/ 0 6/G l d i d i i f 2 6/ 0 6/–– General drop in admissions from 2.6/yr to 0.6/yr General drop in admissions from 2.6/yr to 0.6/yr with no interventionswith no interventions

•• Roland et al BMJ 2005Roland et al BMJ 2005•• Roland et al, BMJ, 2005Roland et al, BMJ, 2005

Charlson Comorbidity IndexCharlson Comorbidity IndexCharlson Comorbidity IndexCharlson Comorbidity Index

Probability of Repeated Hospital Probability of Repeated Hospital Admission (Pra)Admission (Pra)

1.1. In general, would you say your health is:In general, would you say your health is:1.1. Excellent, very good, good, fair, poorExcellent, very good, good, fair, poor

2.2. In the previous 12 months, have you stayed overnight as a patient in a In the previous 12 months, have you stayed overnight as a patient in a hospital?hospital?11 Not at all 1 time 2Not at all 1 time 2--3 times more than 3 times3 times more than 3 times1.1. Not at all, 1 time, 2Not at all, 1 time, 2 3 times, more than 3 times3 times, more than 3 times

3.3. In the previous 12 months, how many times did you visit a physician In the previous 12 months, how many times did you visit a physician or clinic?or clinic?1.1. Not at all, 1 time, 2Not at all, 1 time, 2--3 times, 43 times, 4--6 times, more than 6 times6 times, more than 6 times

4.4. In the previous 12 months, did you have diabetes? Yes, noIn the previous 12 months, did you have diabetes? Yes, no5.5. Have you every had?Have you every had?

1.1. Coronary heart disease, angina pectoris, an MI, any other heart attackCoronary heart disease, angina pectoris, an MI, any other heart attack66 I h f i d l i i hb h ld k f fI h f i d l i i hb h ld k f f6.6. Is there a friend, relative or neighbor who would take care of you for a Is there a friend, relative or neighbor who would take care of you for a

few days, if necessary?few days, if necessary? Yes, noYes, no7.7. Are you?Are you? male, femalemale, female88 What is your date of birth?What is your date of birth? M th dM th d8.8. What is your date of birth?What is your date of birth? Month, day, yearMonth, day, year

VESVES--1313VESVES 1313

•• 1. Age1. Agegg•• 2. In general, compared to other people your age, would you say 2. In general, compared to other people your age, would you say

that your health is:that your health is:–– Poor fair Good Very Good ExcellentPoor fair Good Very Good ExcellentPoor, fair, Good, Very Good, ExcellentPoor, fair, Good, Very Good, Excellent

•• 3. How much difficulty, on average, do you have with the 3. How much difficulty, on average, do you have with the following physical activities:following physical activities:

S i hi k li ?S i hi k li ?–– Stooping, crouching or kneeling?Stooping, crouching or kneeling?–– Lifting, or carrying objects as heavy as 10 pounds?Lifting, or carrying objects as heavy as 10 pounds?–– Reaching or extending arms above shoulder level?Reaching or extending arms above shoulder level?

W i i h dli d i ll bj ?W i i h dli d i ll bj ?–– Writing, or handling and grasping small objects?Writing, or handling and grasping small objects?–– Walking a quarter of a mile?Walking a quarter of a mile?–– Heavy housework such as scrubbing floors or washing windows?Heavy housework such as scrubbing floors or washing windows?

VESVES--1313VESVES 1313

•• 4. Because of your health or a physical4. Because of your health or a physical4. Because of your health or a physical 4. Because of your health or a physical condition, do you have any difficulty:condition, do you have any difficulty:–– Shopping for personal items (like toilet items or medicines)?Shopping for personal items (like toilet items or medicines)?–– Managing money (like keeping track of expenses or paying Managing money (like keeping track of expenses or paying

bills)?bills)?–– Walking across the room? Use of cane or walker OKWalking across the room? Use of cane or walker OK–– Walking across the room? Use of cane or walker OKWalking across the room? Use of cane or walker OK–– Doing light housework (like washing dishes, straightening up, Doing light housework (like washing dishes, straightening up,

or light cleaning)?or light cleaning)?–– Bathing or showering?Bathing or showering?

Lee 4Lee 4--yr mortality Indexyr mortality IndexLee 4Lee 4 yr mortality Indexyr mortality Index

1.1. AgeAge22 SS2.2. SexSex3.3. BMI < 25BMI < 254.4. Has a doctor ever told you that you have Has a doctor ever told you that you have diabetes diabetes or high blood sugar?or high blood sugar?5.5. Has a doctor told you that you have Has a doctor told you that you have cancercancer or a malignant tumor, excluding minor or a malignant tumor, excluding minor y yy y g , gg , g

skin cancers?skin cancers?6.6. Do you have a Do you have a chronic lung diseasechronic lung disease that limits your usual activities or makes you need that limits your usual activities or makes you need

oxygen at home?oxygen at home?7.7. Has a doctor told you that you have Has a doctor told you that you have congestive heart failurecongestive heart failure??y yy y gg8.8. Have you smoked cigarettes in the past week?Have you smoked cigarettes in the past week?9.9. Because of a health or memory problem, do you have any Because of a health or memory problem, do you have any difficulty with bathing or difficulty with bathing or

showering?showering?1010 Because of a health or memory problem do you have anyBecause of a health or memory problem do you have any difficulty with managingdifficulty with managing10.10. Because of a health or memory problem, do you have any Because of a health or memory problem, do you have any difficulty with managing difficulty with managing

your moneyyour money——such as paying your bills and keeping track of expenses?such as paying your bills and keeping track of expenses?11.11. Because of a health or memory problem, do you have any Because of a health or memory problem, do you have any difficulty with walking difficulty with walking

several blocks?several blocks?1212 Because of a health or memory problem do you have any difficulty withBecause of a health or memory problem do you have any difficulty with pulling orpulling or12.12. Because of a health or memory problem, do you have any difficulty with Because of a health or memory problem, do you have any difficulty with pulling or pulling or

pushing large objectspushing large objects like a living room chair?like a living room chair?

Schonberg 5Schonberg 5--year Life expectancyyear Life expectancySchonberg 5Schonberg 5 year Life expectancyyear Life expectancy

1.1. AgeAge2.2. SexSex3.3. BMI < 25BMI < 254.4. Would you say Would you say your health your health in general is: Excellent, very good, good, fair, poorin general is: Excellent, very good, good, fair, poor55 H b t ld b d t h lth f i l th t h dH b t ld b d t h lth f i l th t h d5.5. Have you ever been told by a doctor or health professional that you had: Have you ever been told by a doctor or health professional that you had:

a. a. Emphysema/chronic bronchitis, cancer, diabetesEmphysema/chronic bronchitis, cancer, diabetes6.6. Because of a physical, mental, or emotional problem, do you need the help of Because of a physical, mental, or emotional problem, do you need the help of

other persons in other persons in handling routine needshandling routine needs such as everyday household chores, such as everyday household chores, pp gg y y ,y y ,doing necessary business, shopping, or getting around for other purposes?doing necessary business, shopping, or getting around for other purposes?a. Yes, noa. Yes, no

7.7. By yourself, and without using any special equipment, how By yourself, and without using any special equipment, how difficult is it for difficult is it for t lkt lk q rt r f milq rt r f mil b t 3 it bl k ?b t 3 it bl k ?you to walkyou to walk a quarter of a milea quarter of a mile--about 3 city blocks?about 3 city blocks?

a. Not at all difficult, a little difficult to very difficult, can’t do at all/do not a. Not at all difficult, a little difficult to very difficult, can’t do at all/do not dodo

8.8. Which best describes your Which best describes your cigarette usecigarette use? Never, former, current? Never, former, currentyy gg , ,, ,9.9. During the past 12 months, how many times were you During the past 12 months, how many times were you hospitalized overnighthospitalized overnight??

a. None, once, twice or morea. None, once, twice or more

Predictive Modeling: Predictive Modeling: P rf rm nP rf rm nPerformancePerformance

•• Perform poorlyPerform poorlyPerform poorlyPerform poorly•• Review of 26 Review of 26 readmission risk prediction modelsreadmission risk prediction models

which demonstrate poor predictive abilitywhich demonstrate poor predictive abilitywhich demonstrate poor predictive ability which demonstrate poor predictive ability (Kansagara et al, JAMA, 2011)(Kansagara et al, JAMA, 2011)

–– Poorly understood and complex phenomenon Poorly understood and complex phenomenon y d s d d c p py d s d d c p pbeyond typical factors (medical comorbidities, basic beyond typical factors (medical comorbidities, basic demographic data, clinical variables) included in PMdemographic data, clinical variables) included in PM

–– If we can’t accurately predict which patients will be If we can’t accurately predict which patients will be readmitted, can we really prevent readmissions?readmitted, can we really prevent readmissions?

Pr di ti F t rPr di ti F t rPredictive FactorsPredictive Factors

•• Predicting complicated postPredicting complicated post--hospital carehospital carePredicting complicated postPredicting complicated post hospital care hospital care patternspatterns

•• Variables that impro ed acc rac of predictionsVariables that impro ed acc rac of predictions•• Variables that improved accuracy of predictionsVariables that improved accuracy of predictions–– SelfSelf--rated health statusrated health status

Vi l i iVi l i i–– Visual impairmentVisual impairment–– Mean scores on physical functional tasksMean scores on physical functional tasks–– Assistance with ADLs scoreAssistance with ADLs score

•• Better discrimination than PM only using administrative dataBetter discrimination than PM only using administrative data(Coleman et al Health Services Research 2004)(Coleman et al Health Services Research 2004)–– (Coleman et al, Health Services Research, 2004)(Coleman et al, Health Services Research, 2004)

Pr di ti M d linPr di ti M d linPredictive ModelingPredictive Modeling

•• Automated ambulatory pharmacy data used toAutomated ambulatory pharmacy data used toAutomated ambulatory pharmacy data used to Automated ambulatory pharmacy data used to predict future health care costpredict future health care cost–– Compared with complex PM like HCCs and ACGsCompared with complex PM like HCCs and ACGs–– Compared with complex PM like HCCs and ACGs Compared with complex PM like HCCs and ACGs

(ambulatory clinical groups)(ambulatory clinical groups)–– Less accurate than HCCs but similar to ACGsLess accurate than HCCs but similar to ACGs–– Data often more reliable and complete than ICDData often more reliable and complete than ICD--9 9

codescodes–– Reasonable alternativeReasonable alternative

•• (Fishman et al, Medical Care, 2003)(Fishman et al, Medical Care, 2003)

Ph i i n S l ti nPh i i n S l ti nPhysician SelectionPhysician Selection

•• Ability to predict which hospitalized patientsAbility to predict which hospitalized patientsAbility to predict which hospitalized patients Ability to predict which hospitalized patients were most likely to be readmitted were most likely to be readmitted

•• Not necessarily any better but not worse thanNot necessarily any better but not worse than•• Not necessarily any better, but not worse than Not necessarily any better, but not worse than PMPMC d PM (P ) i h h i iC d PM (P ) i h h i i•• Compared PM (Pra) with physicians, case Compared PM (Pra) with physicians, case managers and nursesmanagers and nurses

•• Poor for all providers and the Pra (0.50Poor for all providers and the Pra (0.50--0.59)0.59)(Allaudeen et al, J Gen Intern Med, 2011)(Allaudeen et al, J Gen Intern Med, 2011)

Comparing PM and PCP Comparing PM and PCP l ti n m th dl ti n m th dselection methodsselection methods

•• PCP selection:PCP selection:PCP selection: PCP selection: –– 14 PCPs from 10 primary care practices 14 PCPs from 10 primary care practices

Screen a list of their patients and select up to 30 toScreen a list of their patients and select up to 30 to–– Screen a list of their patients and select up to 30 to Screen a list of their patients and select up to 30 to participate in a care management program aimed at participate in a care management program aimed at reducing avoidable hospitalizationsreducing avoidable hospitalizationsd c g v d b sp sd c g v d b sp s

•• PM:PM:Software package Case Smart Suite GermanySoftware package Case Smart Suite Germany–– Software package Case Smart Suite GermanySoftware package Case Smart Suite Germany

–– Diagnostic codes, prior costs, hospital admissions, Diagnostic codes, prior costs, hospital admissions, demographic datademographic datademographic datademographic data

–– (Freund et al, Am J Manag Care, 2011)(Freund et al, Am J Manag Care, 2011)

Comparing PM and PCP Comparing PM and PCP l ti nl ti nselectionselection

•• PaperPaper--based questionnaire:based questionnaire:PaperPaper based questionnaire:based questionnaire:–– European Quality of life indexEuropean Quality of life index

Medication adherence report scaleMedication adherence report scale–– Medication adherence report scaleMedication adherence report scale–– Sociodemographic variablesSociodemographic variables


•• Selected versus nonSelected versus non--selected patients (regardlessselected patients (regardlessSelected versus nonSelected versus non selected patients (regardless selected patients (regardless of method)of method)–– Older higher likelihood of future hospitalization higherOlder higher likelihood of future hospitalization higher–– Older, higher likelihood of future hospitalization, higher Older, higher likelihood of future hospitalization, higher

morbidity burden, higher predicted costsmorbidity burden, higher predicted costs

•• PM (301 patients)PM (301 patients)PM (301 patients)PM (301 patients)–– More prior hospitalizations, higher likelihood of future More prior hospitalizations, higher likelihood of future

hospitalization, higher morbidity burden, higher hospitalization, higher morbidity burden, higher p , g y , gp , g y , gpredicted costs, 1 of every 2 pts had CHFpredicted costs, 1 of every 2 pts had CHF

Comparing PM and PCP selectionComparing PM and PCP selectionComparing PM and PCP selectionComparing PM and PCP selection

•• PCP selection (203 patients)PCP selection (203 patients)PCP selection (203 patients)PCP selection (203 patients)–– Lower prevalence of CHF, higher rates of Lower prevalence of CHF, higher rates of

prior participation care managementprior participation care managementprior participation care management prior participation care management programsprograms

B h (32 iB h (32 i ll l i )ll l i )•• Both (32 patientsBoth (32 patients--mutually exclusive)mutually exclusive)–– 1 of every 2 pts with CHF1 of every 2 pts with CHF–– Higher rates of prior participation in care Higher rates of prior participation in care

management programsmanagement programsg p gg p g


•• Patient survey: 40 7% of total respondedPatient survey: 40 7% of total respondedPatient survey: 40.7% of total respondedPatient survey: 40.7% of total responded•• Patients selected by PCP responded more Patients selected by PCP responded more

freq entl (58 1 % s 22 7%)freq entl (58 1 % s 22 7%)frequently (58.1 % vs 22.7%)frequently (58.1 % vs 22.7%)•• NonNon--respondents: more hospital admissions, respondents: more hospital admissions,

bidi b d hi h di dbidi b d hi h di dgreater morbidity burden, higher predicted costs greater morbidity burden, higher predicted costs and hospitalizationsand hospitalizations


•• PMPMPMPM–– Higher risk for future healthcare utilization and high Higher risk for future healthcare utilization and high

predicted annual costspredicted annual costspp

•• PCPPCP–– Higher rates of prior participation in intensified careHigher rates of prior participation in intensified care–– Higher rates of prior participation in intensified care Higher rates of prior participation in intensified care

programs and higher response rateprograms and higher response rate

•• BothBothBothBoth–– Had both Had both

C r S n iti itC r S n iti itCare SensitivityCare Sensitivity•• WWillingness may be a key characteristic toillingness may be a key characteristic toWWillingness may be a key characteristic to illingness may be a key characteristic to

identifyidentify–– Engage in and benefit from the programEngage in and benefit from the programg g p gg g p g–– Too healthy to benefit?Too healthy to benefit?Or, on the other hand….Or, on the other hand….

•• Lack of willingness may be a key characteristicLack of willingness may be a key characteristic–– Highest risk patients less willingHighest risk patients less willingg p gg p g–– Too ill or frail to benefit?Too ill or frail to benefit?–– May be due to complex social/economic factorsMay be due to complex social/economic factors

Physician SelectionPhysician SelectionPhysician SelectionPhysician Selection

•• Physician selection:Physician selection:Physician selection: Physician selection: –– Choose patients who we feel are more willing or Choose patients who we feel are more willing or

approachable?approachable?–– Choose patients who are more complex or difficult Choose patients who are more complex or difficult

to manage?to manage?–– What shapes the perception of patients seen as What shapes the perception of patients seen as

“most likely to benefit?”“most likely to benefit?”

S h t?S h t?So what?So what?

•• Need for more clinically based PM that doesn’tNeed for more clinically based PM that doesn’tNeed for more clinically based PM that doesn t Need for more clinically based PM that doesn t rely on complex administrative datarely on complex administrative data

•• Incl de ariables for care sensiti itIncl de ariables for care sensiti it•• Include variables for care sensitivityInclude variables for care sensitivity•• Practicable concepts for case findingPracticable concepts for case finding•• Standardized assessment tools should be Standardized assessment tools should be

developed and implemented to assist clinicians developed and implemented to assist clinicians in refining PMin refining PM--based selectionbased selection

Pr j t IdPr j t IdProject IdeaProject Idea

•• PracticePractice--based realbased real--time clinical tool to helptime clinical tool to helpPracticePractice based, realbased, real time clinical tool to help time clinical tool to help identify patients who are:identify patients who are:

At highest risk for health care utilization ANDAt highest risk for health care utilization AND–– At highest risk for health care utilization ANDAt highest risk for health care utilization AND–– Most likely to benefit from intensive resources (GC)Most likely to benefit from intensive resources (GC)

Combine PM PCP selection and patient illingnessCombine PM PCP selection and patient illingness–– Combine PM, PCP selection, and patient willingness Combine PM, PCP selection, and patient willingness into 1 single, simplified approach in a practice settinginto 1 single, simplified approach in a practice setting

S r T lS r T lSurvey ToolSurvey Tool

•• Tool attached to encounter of all patientsTool attached to encounter of all patients•• Tool attached to encounter of all patients Tool attached to encounter of all patients 65 years of age or older at Jefferson Family 65 years of age or older at Jefferson Family M di i i h PSCM di i i h PSCMedicine in the PSCMedicine in the PSC

•• Filled out by PCP and patient togetherFilled out by PCP and patient togethery p gy p g•• Validated instrument to identify highValidated instrument to identify high--risk risk

older patientsolder patientsolder patientsolder patients–– Pra, VESPra, VES--13, Schonberg mortality index13, Schonberg mortality index

S r T lS r T lSurvey ToolSurvey Tool•• If the patient is identified as “highIf the patient is identified as “high--risk” by toolrisk” by toolIf the patient is identified as highIf the patient is identified as high risk by tool risk by tool •• Do you [provider] think this patient would benefit from Do you [provider] think this patient would benefit from

a care management program like Guided Care?a care management program like Guided Care?g p gg p g•• If no, why?If no, why?•• If yes, is the patient willing to be contacted to hearIf yes, is the patient willing to be contacted to hearIf yes, is the patient willing to be contacted to hear If yes, is the patient willing to be contacted to hear

more information and consider receiving additional more information and consider receiving additional services through Guided Care?services through Guided Care?

•• If no, why?If no, why?

EvaluationEvaluation

•• Patients who meet criteria who are and are notPatients who meet criteria who are and are notPatients who meet criteria who are and are not Patients who meet criteria who are and are not recommended by PCP recommended by PCP

•• Patients who meet criteria who are and aren’tPatients who meet criteria who are and aren’t•• Patients who meet criteria who are and aren t Patients who meet criteria who are and aren t willing to participatewilling to participateD illi i i hD illi i i h•• Does willingness to participate change over Does willingness to participate change over time? time?


•• Compare tool to PM + PCP selectionCompare tool to PM + PCP selectionpp•• List of high risk older patientsList of high risk older patients•• PCP circles who he/she thinks would benefit from GCPCP circles who he/she thinks would benefit from GC•• Answers 2 openAnswers 2 open--ended questions:ended questions:

–– What are the major characteristics of patients that made you What are the major characteristics of patients that made you l h f hi ?l h f hi ?select them to refer to this program?select them to refer to this program?

–– What are the major characteristics of patients that made you What are the major characteristics of patients that made you not select them for this program?not select them for this program?p gp g

•• Selected patients are called and asked if willing to Selected patients are called and asked if willing to participateparticipate


•• Compare patients willing to participate in GCCompare patients willing to participate in GCCompare patients willing to participate in GC Compare patients willing to participate in GC identified by each methodidentified by each method

Does patient willingness to enroll increase with tool?Does patient willingness to enroll increase with tool?–– Does patient willingness to enroll increase with tool?Does patient willingness to enroll increase with tool?

Additional factors?Additional factors?Additional factors?Additional factors?

•• Insurance typeInsurance typeInsurance typeInsurance type•• Zip codeZip code

S i lS i l•• Social supportSocial support

ConclusionConclusionConclusionConclusion

•• New models of care must address the complexNew models of care must address the complexNew models of care must address the complex New models of care must address the complex health care needs of multimorbid older patientshealth care needs of multimorbid older patients

•• Guided Care GRACE PACE are 3 evidenceGuided Care GRACE PACE are 3 evidence•• Guided Care, GRACE, PACE are 3 evidenceGuided Care, GRACE, PACE are 3 evidence--based models of comprehensive, coordinated based models of comprehensive, coordinated carecarecarecare

•• Evaluation of these models must consider Evaluation of these models must consider i l i h ill b fii l i h ill b fitargeting populations who will benefit mosttargeting populations who will benefit most

•• Approaches to patient selection can include a Approaches to patient selection can include a variety of measures and methods variety of measures and methods

R f r nR f r nReferencesReferences•• Kansagara D, Englander H, Salanitro A, Kansagara D, Englander H, Salanitro A, KagenKagen D, Theobald C, Freeman M, D, Theobald C, Freeman M, KripalaniKripalani S, Risk S, Risk

P di i M d l f H i l R d i i A S i R i JAMA 2011 306(15) 1688P di i M d l f H i l R d i i A S i R i JAMA 2011 306(15) 1688Predictive Models for Hospital Readmission: A Systematic Review, JAMA, 2011; 306(15): 1688Predictive Models for Hospital Readmission: A Systematic Review, JAMA, 2011; 306(15): 1688--1698.1698.

•• Grant RW, Grant RW, AshburnerAshburner JM, Hong CC, Chang Y, Barry MJ, Atlas SJ, Defining Patient Complexity JM, Hong CC, Chang Y, Barry MJ, Atlas SJ, Defining Patient Complexity from the Primary Care Physician’s Perspective, Ann Intern Med, 2011; 155: 797from the Primary Care Physician’s Perspective, Ann Intern Med, 2011; 155: 797--804.804.

•• Freund T, Mahler C, Freund T, Mahler C, ErierErier A, A, GensichenGensichen J, J, OseOse D, D, SzecsenyiSzecsenyi J, PetersJ, Peters--KlimmKlimm F, Identification of F, Identification of Patients Likely to Benefit From Care Management Programs, Am J Manag Care. 2011; 17 (5): Patients Likely to Benefit From Care Management Programs, Am J Manag Care. 2011; 17 (5): 345345--352.352.

•• Freund T,Freund T, WensingWensing M, Mahler C,M, Mahler C, GensichenGensichen J,J, ErlerErler A, Beyer M,A, Beyer M, GerlachGerlach FM,FM, SzecsenyiSzecsenyi J,J,Freund T, Freund T, WensingWensing M, Mahler C, M, Mahler C, GensichenGensichen J, J, ErlerErler A, Beyer M, A, Beyer M, GerlachGerlach FM, FM, SzecsenyiSzecsenyi J, J, PetersPeters--KlimmKlimm F, Development of a primary careF, Development of a primary care--based complex care management intervention based complex care management intervention for chronically ill patients at high risk for hospitalization: a study protocol, Implementation for chronically ill patients at high risk for hospitalization: a study protocol, Implementation Science, 2010; 5: 70 at Science, 2010; 5: 70 at http://www.implementationscience.com/content/5/1/70http://www.implementationscience.com/content/5/1/70

•• LopezLopez Aguila SAguila S ContelContel JCJC FarreFarre JJ CampuzanoCampuzano JLJL RajmilRajmil L Predictive Model for EmergencyL Predictive Model for Emergency•• LopezLopez--Aguila S, Aguila S, ContelContel JC, JC, FarreFarre J, J, CampuzanoCampuzano JL, JL, RajmilRajmil L, Predictive Model for Emergency L, Predictive Model for Emergency Hospital Admission and 6Hospital Admission and 6--Month Readmission, Am J Manag Care, 2011; 17(9): e 348Month Readmission, Am J Manag Care, 2011; 17(9): e 348--357.357.

•• Coleman EA, Min SJ, Coleman EA, Min SJ, ChomiakChomiak A, Kramer AM, A, Kramer AM, PosthospitalPosthospital care transitions: patterns, care transitions: patterns, complications, and risk identification. Health complications, and risk identification. Health ServServ Res. 2004; 39(5): 1449Res. 2004; 39(5): 1449--1465.1465.

•• Fishman PA, Goodman MJ, Fishman PA, Goodman MJ, HornbrookHornbrook MC, MC, MeenanMeenan RT, Bachman DJ, O’Keeffe RT, Bachman DJ, O’Keeffe RosettiRosetti MC, MC, Risk Adjustment Using Automated Ambulatory Pharmacy Data: The Risk Adjustment Using Automated Ambulatory Pharmacy Data: The RxRiskRxRisk Model, Med Care. Model, Med Care. 2003; 41: 842003; 41: 84--99.99.

ReferencesReferencesReferencesReferences

•• Allaudeen N, Allaudeen N, SchipperSchipper JL, JL, OravOrav EJ, EJ, WachterWachter RM, RM, VidyarthiVidyarthi AR, Inability of Providers to Predict AR, Inability of Providers to Predict Unplanned Readmissions J Gen Intern Med 2011; (26) 7: 771Unplanned Readmissions J Gen Intern Med 2011; (26) 7: 771 66Unplanned Readmissions, J Gen Intern Med, 2011; (26) 7: 771Unplanned Readmissions, J Gen Intern Med, 2011; (26) 7: 771--6.6.

•• Schonberg MA, Davis RB, McCarthy EP, Schonberg MA, Davis RB, McCarthy EP, MarcantonioMarcantonio ER, Index to Predict 5ER, Index to Predict 5--Year Mortality of Year Mortality of CommunityCommunity--Dwelling Adults Aged 65 and Older Using Data from the National Health Interview Dwelling Adults Aged 65 and Older Using Data from the National Health Interview Survey, J Gen Intern Med, 2009, 24 (10): 1115Survey, J Gen Intern Med, 2009, 24 (10): 1115--22.22.

•• Schonberg MA Davis RB McCarthy EPSchonberg MA Davis RB McCarthy EP MarcantonioMarcantonio ER External Validation of an Index toER External Validation of an Index to•• Schonberg MA, Davis RB, McCarthy EP, Schonberg MA, Davis RB, McCarthy EP, MarcantonioMarcantonio ER, External Validation of an Index to ER, External Validation of an Index to Predict up to 9Predict up to 9--year mortality of Communityyear mortality of Community--dwelling adults aged 65 and older, J Am dwelling adults aged 65 and older, J Am GeriatrGeriatr SocSoc, , 2011, 59: 14442011, 59: 1444--1451.1451.

•• CharslsonCharslson ME, ME, PompeiPompei P, Ales KL, P, Ales KL, MacKenzieMacKenzie CR, A new method of classifying prognostic CR, A new method of classifying prognostic comorbidity in longitudinalcomorbidity in longitudinal sutdiessutdies: development and validation Journal of Chronic diseases: development and validation Journal of Chronic diseasescomorbidity in longitudinal comorbidity in longitudinal sutdiessutdies: development and validation, Journal of Chronic diseases, : development and validation, Journal of Chronic diseases, 1987, 40 (5): 3731987, 40 (5): 373--383.383.

•• Lee SJ, Lindquist L, Segal MR, Lee SJ, Lindquist L, Segal MR, CovinksyCovinksy KE, Development and Validation of a Prognostic Index KE, Development and Validation of a Prognostic Index for 4for 4--year mortality in older adults, JAMA, 2006, 295 (7): 801year mortality in older adults, JAMA, 2006, 295 (7): 801--808.808.

•• Walter LC,Walter LC, CovinksyCovinksy KE, Cancer Screening in Elderly Patients: A Framework for IndividualizedKE, Cancer Screening in Elderly Patients: A Framework for IndividualizedWalter LC, Walter LC, CovinksyCovinksy KE, Cancer Screening in Elderly Patients: A Framework for Individualized KE, Cancer Screening in Elderly Patients: A Framework for Individualized Decision Making, JAMA, 2001; 285 (21): 2750Decision Making, JAMA, 2001; 285 (21): 2750--2756.2756.

•• SalibaSaliba D, Elliot M, Rubenstein LZ, Solomon DH, Young RT, D, Elliot M, Rubenstein LZ, Solomon DH, Young RT, KambergKamberg CJ, Roth C, MacLean CJ, Roth C, MacLean CH, CH, ShekelleShekelle PG, PG, SlossSloss EM, Wenger NS, The Vulnerable Elders Survey: A Tool for Identifying EM, Wenger NS, The Vulnerable Elders Survey: A Tool for Identifying Vulnerable Older people in the Community, J Am Vulnerable Older people in the Community, J Am GeriatrGeriatr SocSoc, 2001; 49: 1691, 2001; 49: 1691--1699.1699.p p y Jp p y J

salzman patient selection for care management

Documents