Comorbidity: From Bedside to Bench

1
Comorbidity From Bedside to Bench

• Summary of the NIA/AGS R13 Conference

ASG Annual Meetings, May 13, 2005, Orlando
2
Comorbidity, Multi-Morbidity

• any distinct clinical entity that has existed or
  may occur during the clinical course of a patient
  who has an index disease or condition under
  study. (Feinstein, 1970)
• distinct clinical entities coexisting or likely
  to co-occur during a patients clinical course

ASG Annual Meetings, May 13, 2005, Orlando
3
Symposium Presenters

• Rebecca Silliman, MD The NIA Comorbidity
  Taskforce
• Alison Moore, MD Comorbidity in Relation to the
  Study and Treatment of Index Conditions
• Christine Ritchie, MD The Health and Social
  Burden of Multiple Morbidity
• Stephanie Studenski, MD The Research Agenda

ASG Annual Meetings, May 13, 2005, Orlando
4
Multimorbidity Concepts and Research
Recommendations

• Thanks to Linda Fried for the use of some of her
  presentation and to the members of the
  preclinical break out session
• Stephanie Studenski MD MPH
• Professor, Department of Medicine (geriatrics)
  Staff Physician,
• VA Pittsburgh GRECC
• 3471 Fifth Avenue Suite 500
• Pittsburgh Pa 15213
• office 412 692 2360
• fax [phone]
• email studenskis_at_msx.dept-med.pitt.edu

ASG Annual Meetings, May 13, 2005, Orlando
5
Outline

• Multimorbidity the burden of illness
• Clusters of diseases and conditions causes and
  consequences

Definitions
Comorbidity additional diseases beyond the index
disease Multimorbidity co-occurrence of diseases
ASG Annual Meetings, May 13, 2005, Orlando
6
Multimorbidity

• Often no index condition.
• Systems serve as reserve capacity for each
  others losses.
• Multimorbidity reflects total burden of illness
  and has implications for reserve and tolerance
  to stress.

ASG Annual Meetings, May 13, 2005, Orlando
7
Measuring the Burden of Illness Challenges

• When burden is assessed by diagnoses, factors
  that influence the process of clinical diagnosis
  affect reports.
• Eg Clinical thresholds for the diagnosis of
  disease vary by provider recognition, shifts over
  time in definitions eg DM, hyperlipidemia, HBP
• Eg Severity measures may be affected by
  coexisting conditions eg treadmill testing and
  CAD. Subspecialists may ignore the effect of
  coexisting conditions.

ASG Annual Meetings, May 13, 2005, Orlando
8
Physiological system indicators may eliminate
variability due to clinical thresholds

• When burden is assessed by diagnoses, factors
  that influence the process of clinical diagnosis
  affect reports.
• Eg Clinical thresholds for the diagnosis of
  disease vary by provider recognition, shifts over
  time in definitions eg DM, hyperlipidemia, HBP
• Eg Severity measures may be affected by
  coexisting conditions eg treadmill testing and
  CAD. Subspecialists may ignore the effect of
  coexisting conditions.

ASG Annual Meetings, May 13, 2005, Orlando
9
Physiological system indicators may eliminate
variability due to clinical thresholds
Physiological System
Dx
Severity
10
Physiological system indicators may eliminate
variability due to clinical thresholds
Physiological System
Dx
Severity
11
Opportunities

• Create a system of basic markers of physiological
  functions across key systems (a battery like
  APACHE)
• Basic indicators by system (e.g., Hb, Creat).
• Develop and evaluate using existing data.
• Applications
• Compare to measures based on diagnoses.
• Might help ease barriers to including research on
  elders with comorbidity.
• Use battery to examine interactions and demands
  between physiological systems.
• Trials could look at subclinical adverse effects
  across subgroups

12
The Physiologic Battery as an indicator of burden
of illness/multimorbidity

• Expand battery to include axes within
  physiological systems/disease
• Duration and pattern over time
• Treatment effects
• Adds detail but increases complexity and demand
  of measure

13
Next Steps

• Modeling that accounts for time and patterns the
  NIA longitudinal analysis RFA
• Novel analytic methods
• Training (K awards), methodological publications
• Data sets (core data) with physiological
  indicators
• Health ABC, InChianti, BLSA

14
Other Measures of Burden of Illness/Multimorbidity

• Physical Performance measures can be thought of
  as summary measures of preclinical clinical
  conditions they are composites and are
  non-specific.
• One kind of indicator an integrative summary of
  multiple morbidity
• Do not attribute symptoms and function only to
  index condition

15
Clusters what can they tell us?

• Cluster system abnormalities that co-occur at a
  rate that is higher than expected by chance
  alone.
• Types of clusters
• single underlying common cause
• secondary consequences of index
  condition
• Clusters can provide insights into common causes
  and into combined effects on consequences like
  disability.

16
Clusters in late life implications for causation

• 24 year old woman with rash, arthritis and kidney
  disease
• 84 year old woman with rash, arthritis and kidney
  disease
• Since conditions are more rare in younger adult,
  a cluster is unlikely to be due to chance, and is
  likely to have a common cause.
• Conversely, since conditions are more common in
  older adults, clusters are more likely to be due
  to chance and may not have a common cause.

17
Late Life Clusters

• Unrecognized underlying process or condition
  precipitates multiple abnormalities
  inflammation as cause of atherosclerosis,
  malnutrition, frailty, neurodegeneration creates
  new target for intervention. (Ferrucci L et al A
  flame burning within. Aging Clin Exp Res. 2004)
• A known condition precipitates others eg
  diabetes, atherosclerosis, renal failure target
  precipitating condition for intervention (Volpato
  et al Diabetes Care 2002)

18
Primary clusters
Clusters with no recognized underlying common
cause are an opportunity for research into
prevention and treatment of late life
multimorbidity
Disease B
Potential underlying cause
Disease C
Disease D
19
Secondary ClustersDiabetes and complications

• Duration of diabetes associated with presence of
  CHD, CHF, PAD, HTN, Depression
• Diabetes associated with
• Peripheral neuropathy, CVD, visual impairment,
  obesity
• Disability mobility, ADL, IADL
• Volpato,Diabetes Care 2002

20
Consequences combinations of diseases
synergistically associated with disability
21
Two Diseases Present Concurrently have Joint
Effects

• Risk of Mobility Disability
• Heart Disease Only OR 2.3
• Arthritis Only OR 4.3
• Both Heart Disease
• and Arthritis OR 13.6
• NHANES III
• Ettinger et al
  
  

22
Clusters and Consequences

• Much of the action is in the interaction
• The interactions between diseases contribute to
  disability, over and above the independent
  contribution of each disease.
• Research questions Interactions between specific
  disease pairs might have effects specific to
  different types of function.
• Clinical implications target preservation of
  specific functions by minimizing specific
  interactions?

23
Comorbidity in relation to study and treatment of
index disease

• Alison A. Moore, MD, MPH
• David Geffen School of Medicine at UCLA
• Division of Geriatric Medicine

24
HIV/AIDS as a Chronic Disease the Veterans
Aging Cohort Study

• Amy C. Justice, MD, PhD
• PI, Veterans Aging Cohort Study
• GIM Section Chief, West Haven VAMC
• Yale University

25
Why Study HIV and Comorbidity?

• Clinical Reasons
• Prevalence People with HIV are living long
  enough to age
• Incidence As more people with HIV are aging,
  more older individuals will contract HIV
• Toxicity Difficult to determine what is due to
  treatment if we dont understand underlying risk
  of comorbid disease
• Research Reasons
• Bench effect modification may lead to
  pathophysiologic insights
• Outcomes due to implications for survival
  optimal management of HIV may differ by age
  optimal management of comorbidity may differ by
  HIV status

26
Life Expectancy after HIV diagnosis with and
without HAART
Years
Age 50

without
Age 40
with
without
with
without
Age 30
with
CD4 750
CD4 500
CD4 200
27
Non-AIDS Deaths with and without HAART (Virtual
Cohort)

Age 50
Age 40
with
without
without

without
with
Age 30
with
CD4 750
CD4 500
CD4 200
28
Conclusions

• More HIV pts will die from non-HIV causes
• Nearly half of patients with agegt40 years
• If mean age at HIV diagnosis remains 38,
• Mean survival will approach 19.6 years
• Mean age at death will approach 58 years
• Guidelines for management of diseases occurring
  with complex chronic disease must account for
• Shortened life expectancy
• Increased risk due to primary disease and its
  treatment

29
Late Life Depression and Medical Comorbidity

• Ira R. Katz, MD, PhD
• Professor of Psychiatry
• University of Pennsylvania
• Director, MIRECC
• Philadelphia VA Medical Center

30
Depression amplifies morbidity

• Disability
• Cognitive impairment
• Pain (and other symptoms)
• Subnutrition
• Decreased treatment adherence
• Increased use of health services
• Increased mortality
• Suicide and non-Suicide

31
Associations between Depression and Frailty
Proportion with CES-D gt 10 by Frailty Status
From Fried et al, J Gerontol Med Sci 56A
M146-M156, 2001 CHS data
32
Depressive Symptoms Confer VulnerabilityGlaser
et al, Arch Gen Psych 60 1009-1014, 2003
Changes in IL-6 after influenza vaccination in
normal older individuals
33
Conclusions

• Depression is a manifestation of morbidity and a
  source of vulnerability
• that arises from multiple comorbidities and
  paths
• and leads to multiple adverse health effects
• Therefore, it can be considered a frailty

34
Cardiovascular DiseaseThe 1 Comorbidity in
Aging Patients

• Anne B. Newman, MD, MPH
• Professor of Epidemiology and Medicine
• University of Pittsburgh

35
Comorbidity - CVD and other diseases

• CVD and Osteoarthritis
• Most common combination
• CVD and Depression
• Numerous studies show depression increases risk
  of CVD
• Also possible that there is a vascular
  depression
• CVD and Dementia
• Vascular dementia vs. AD with vascular disease?
• CVD and Cancer
• CVD and Chronic Lung Disease

36
Multivariate Analysis of Subclinical
Cardiovascular Disease for 1st MI CHS n4,946
follow-up 4.8 yrs.
Adjusted for age, race, gender, SBP, glucose,
AAI, ICA-IMT, and EF. Variables that did not
make into final model LV mass by ECG, FVC,
HDL-C, smoking, and fibrinogen.
Psaty BM, Furberg CD, Kuller LH, Bild DE,
Rautaharju PM, Polak JF, Bovill E Gottniener JS.
Traditional risk factors and subclinical disease
measures as predictors of first myocardial
infarction in older adults The Cardiovascular
Health Study. Arch Intern Med. 1999
1591339-1347.
37
Probability of Successful Aging by Age, Gender,
and Subclinical Cardiovascular Disease
65-69
70-74
75-79
80
65-69
70-74
75-79
80
Men
Women
Newman AB, Arnold AM, Naydeck BL, Fried LP, Burke
GL, Enright P, Gottdiener J, Hirsch C, OLeary D,
Tracy R. Successful Aging Effects of Subclinical
Cardiovascular Disease. Arch Intern Med.
20031632315-2322.
38
Summary

• CVD is so common that it will - more often than
  not - be comorbid with something else
• Clinically diagnosed CVD represents less than
  half of the total burden of CVD
• An equal proportion have subclinical CVD
• Subclinical CVD is related to
• Physical performance
• Frailty
• Cognitive decline
• Dementia

39
Diabetes and Comorbidity in Older Adults

• Caroline S. Blaum
• University of Michigan
• Ann Arbor VA Medical Center
• March, 2005

40
Research questions and hypotheses

• In type 2 diabetes, do frailty and disability
  result from accumulating comorbidities or is it
  the underlying pathophysiological disruption that
  causes comorbidity accumulation, frailty and
  disability development?
• Is there a stepwise relationship between the MS,
  Diabetes, and Diabetescomorbidities, and frailty
  and disability?

41
Percent change in mobility score associated with
metabolic syndrome group
42
Summary

• Comorbidity prevalent in older adults with
  diabetes
• Increases with age
• Stepwise progression from MS to new diabetes to
  longstanding diabetes
• MS related to worsening in mobility disability
  but diabetes has much stronger association
• Obesity and diabetes are independently related to
  prevalent frailty.
• MS is related to incident frailty and may
  maintain association in the presence of incident
  diabetes
• Diabetes and many comorbidities are related to
  incident frailty

43
Clinical Epidemiology of Comorbidity in Aging
Patients Findings and Insights from Geriatric
Oncology

• William A. Satariano, Ph.D, MPH
• School of Public Health
• University of California, Berkeley

44
Reasons for Research on Comorbidity and Cancer

• There are age-associated patterns of cancer
  incidence, stage, treatment, and survival (both
  duration and quality of life).
• It is hypothesized that age-associated patterns
  of comorbidity may help to account for those
  age-associated differences in cancer outcomes.

45
Reasons for Research on Comorbidity and Cancer

• There is an extensive network of hospital-based
  and, more important, population-based cancer
  registries and surveillance systems.
• Assessment of large number of cancer cases by
  cancer site, stage, histology, first-course of
  treatment.
• System of linkage with other sources of health
  data that include records of diagnosis and
  treatment for other health conditions.
• Affords opportunity to conduct detail analysis of
  cancer outcomes.

46
Reasons for Research on Comorbidity and Cancer

• There is a significant area of clinical and
  epidemiological research on multiple primary
  cancers, a history of two or more primary cancers
  dx in a single individual.

47
Benefits and Risks of Alcohol Use among Older
Persons

• Alison A. Moore, MD, MPH
• Division of Geriatric Medicine

48
Conditions which may be prevented by light to
moderate alcohol use

• All-cause mortality
• Coronary heart disease
• Congestive heart failure
• Cerebrovascular disease
• Ischemic stroke

• Diabetes
• Cholelithiasis
• Dementia
• ?Falls

49
Conditions that may be caused or worsened by
alcohol use

• Female breast cancer
• Epilepsy
• Hypertension
• Cardiac arrhythmias
• Hemorrhagic stroke
• Psoriasis
• Depression
• Gout
• Alcohol use disorders

• Lip and oropharyngeal cancer
• Esophageal varices and cancer
• Laryngeal cancer
• Liver cirrhosis and cancer
• Gastro-esophageal hemorrhage
• Acute and chronic pancreatitis

50
What is the effect of moderate drinking if you
have comorbidities for which alcohol is
beneficial?

• Evidence that moderate alcohol use is beneficial
  among those persons having
• CHD
• Stroke
• Diabetes

51
What about the effects of drinking and multiple
comorbidity?

• No data!
• Studies have included comorbidity as covariates
  rather than considering the combination of
  alcohol use and selected comorbidity on outcomes

52
Drinking Patterns in Older Persons
53
Mortality risks among at-risk drinkers and
abstainers as compared to not at-risk drinkers
N3726 persons aged 60 participating in NHANES I
(1971-75) and NHEFS 1992
54
Conclusions

• 40-60 of older persons drink alcohol and many
  have comorbidities
• Alcohol has benefits or risks in regard to CHD
  and CHD-related outcomes depending on amount of
  alcohol use
• Alcohol is a risk for many other adverse outcomes
  
• It is unknown whether the CHD-related benefits of
  light to moderate alcohol use persist in the face
  of multiple comorbidity

55
The Health and Societal Burden of Multiple
Morbidity

• Christine S. Ritchie, MD, MSPH
• Associate Professor of Medicine
• University of Alabama at Birmingham

56
Multimorbidity (Comorbidity)

• The co-occurrence of multiple diseases in an
  individual person
• The total burden of all concurrently occurring
  biological processes (clinical and sub-clinical )
  that are intrinsic to the individual
• Explicitly excludes socioeconomic factors,
  lifestyle factors, and access to health care
• In the Nagi pathway terminology, impairment is
  included, while disability is excluded, since
  disability is environment-dependent.

Adapted from Karlamangla A, NIA Comorbidity
Conference, 2005
57
(No Transcript)
58
Prevalence of Multimorbidity

• Using 24 major diagnostic categories
• 82 percent of people 65 and older had one or more
  chronic conditions
• 65 percent two or more
• 43 percent two or more
• 24 percent four or more.
• On average there are 2.3 chronic conditions
  reported by people 65 and older

Wolff JL, Starfield B, Anderson G. Arch Intern
Med. 20021622269-2276
59
Prevalence of Multimorbidity
Wolff JL, Starfield B, Anderson G. Arch Intern
Med. 20021622269-2276
60
Multimorbidity identifies those at risk for more
diseases

• People with multimorbidity at higher risk of
  getting 2 or more new diseases than those with no
  disease, people gt18 years
• (Netherlands van den Akker 1998)

From Fried L, NIA Comorbidity Conference 2005
61
Multimorbidity Joint Effects of Two Diseases

• Risk of Mobility Disability
• Heart Disease Only OR 2.3
• Arthritis Only OR 4.3
• Both Heart Disease
• and Arthritis OR 13.6
• NHANES III
• Ettinger et al
  
  

From Fried L, NIA Comorbidity Conference 2005
62
Impact of multi-morbidity on physical limitations
Kriegsman et al. Disability Rehabilitation
19971971-83
63
Impact of multimorbidity on 3-year decline in
physical functioning
Kriegsman et al. J Clin Epidemiol 20045755-65
64
Impact of Multimorbidity on Quality of Life

• In a systematic review
• Inverse relationship between the number of
  medical conditions and QOL related to physical
  domains.
• For social and psychological dimensions of QOL,
  studies reveal a similar inverse relationship in
  patients with 4 or more diagnoses

Fortin et al. Health Qual Life Outcomes. 2004 2
51
65
Multimorbidity and depressive symptoms
Penninx et al. J Psychosom Res 199640521-534 Ad
apted from Kriegsman D, NIA Comorbidity
Conference, 2005
66
Impact of multimorbidity on coping resources

• Negatively influences coping resources
• Self-esteem
• Mastery
• Self-efficacy

Kriegsman D, NIA Comorbidity Conference, 2005
67
Impact of Multimorbidity on Hospitalization
Hospitalization for Ambulatory Care Sensitive
Condition by Number of Chronic Conditions
Wolff, J. NIA Comorbidity Conference, 2005
68
Multimorbidity and Clinical Outcomes in the VA
Adjusted Clinical Groups (ACGs) Diagnostic
Cost Groups (DCGs)
Petersen et al. Med Care 20054361 from
Berlowitz D, NIA Comorbidity Conference 2005
69
Impact of Multimorbidity on Medicare Expenditures
63 95
Wolff JL, Starfield B, Anderson G. Arch Intern
Med. 20021622269-2276
70
Impact of Multimorbidity on Medicare Expenditures
Wolff JL, Starfield B, Anderson G. Arch Intern
Med. 20021622269-2276
71
Impact of multimorbidity on 3-year mortality
Kriegsman Deeg. In Autonomy and well-being in
the aging population 2 (1997)
72
The Problem of Single Disease Focus vs
Multimorbidity Focus

• Evaluation of severity of disease and impact on
  function
• Evaluation of patient experiences and preferences
• Disease management and health system practice

73
Index Diseases vs Multimorbidity Evaluation of
severity of disease and impact on function

• Aggregate effects of Multimorbidity on function
• Not known dose response effects of disease
  prevention (decreasing by 1, 2, 3)
• Not known Whether additive and synergistic have
  joint mechanisms to be targeted

74
Index Diseases vs Multimorbidity Evaluation of
severity of disease and impact on function

• Need to develop systems to measure severity of
  individual diseases, designed to be used both
  within and across diseases, in patients with
  multimorbidity
• categorization of severity perhaps based on
  similar domains across diseases

From Boyd C, NIA Comorbidity Conference 2005
75
Impact of Multimorbidity on Patient Experiences

• Poor functioning
• Negative psychological reactions
• Negative effects on relationships and
  interference with work or leisure
• Concerns about polypharmacy
• Problematic interactions with providers and the
  health care system including incidents in which
  providers had ignored concerns or provided
  conflicting advice
• Knowledge and skills deficits interfered with
  self-management

Noel et al. Health Expectations. 20058 54-63. 
76
Multimorbidity and Problems with Quality Chronic
Care in the Medicare Program

• Orientation toward acute care, including coverage
  criteria
• Exclusion of catastrophic coverage
• Lack of incentives to provide state-of-the art
  chronic care. In particular
• Reliance on physician orders
• Reimbursement for visits of short duration
• No impetus to coordinate care
• Absence of information technology infrastructure
• Inadequate training of health professionals

From Wolff J Comorbidity Conference 2005
77
Multimorbidity and Chronic Disease Management

• Applicability of evidence-based guidelines
• Focus has been on single disease despite high
  prevalence of multi-morbidity
• Patient preferences/sx often not included in
  outcomes
• Translation of education self-management
  techniques
• Often does not account for polypharmacy
  accompanying multimorbidity
• Does not account for fragmented, single disease
  focused care
• Ability to engage physicians
• Physicians predominantly fee-for-service with
  time constraints
• Large numbers of physicians, not a restricted
  network

78
Multimorbidity and Questions that Remain

• What are realistic assessments/interventions?
• Taxonomy of goals
• What is the effectiveness of following
  disease-specific guidelines in multi-morbidity?
• What are the tensions between prevention,
  treatment, palliation?
• How do we change provider behavior?
• How can current care be better integrated/coordina
  ted?
• Are specialists really better than generalists
  for outcomes that matter in the multi-morbidity
  patient population?

79
Multimorbidity and Untapped Health and Societal
Outcomes

• Pts goals of care/Shared decision making
  tools/Goal attainment scaling
• Pain/Symptom burden
• Trajectories of decline that incorporate multiple
  outcomes (transition probability)
• Self management/caregiver management
• Advance care planning
• Medication review
• Patient experience (AHRQ survey currently being
  investigated by Medicare and mentioned in the
  MedPAC report)

80
Multimorbidity and Health and Societal Outcomes

• X axis Quantitative measures/Qualitative
  measures/Safety/Medical errors
• Y axis Process/Outcomes
• Z axis Time (short vs long term)

81
Report from the NIA Task Force on Comorbidity

• Rebecca A. Silliman, MD, PhD

82
NCI Cancer in the Elderly Initiative

• Outgrowth of an NCI/NIA working conference in
  1981
• 1983 RFA Patterns of Care for Elderly Cancer
  Patients Implications for Cancer Control
• Rosemary Yancik, PhD, Project Officer

83
Subsequent NCI Aging Initiatives

• 1991 NCI RFA The goal of this project is to
  decrease morbidity and enhance survival from
  breast cancer in women 65 years of age and
  older.
• Program Announcements Breast and Prostate -
  1996
• P20 RFA - 2003

84
Geriatric Oncology

• AGS Geriatric Oncology Interest Group - 1991
• John A. Hartford Foundation Geriatric Education
  Retreat (Oncology) - 1997
• International Society of Geriatric Oncology
  (SIOG) - 2000
• ASCO/Hartford Geriatric Oncology Fellowship
  Programs - 2002

85
Comorbidity and Cancer in Older Adults

• Workshop - Comorbidity Assessment of Older Cancer
  Patients, July 29-30, 1999, National Institute on
  Aging and National Cancer Institute Workshop
• Convened by Rosemary Yancik, PhD

86
Parallel Developments

• Case-mix adjustment in health services research
• Understanding the relationships among aging,
  comorbidity, functional status, and frailty
• Improving chronic illness care

87
NIA Comorbidity Task Force

• Geriatric Oncology Harvey Cohen, William
  Ershler, Martine Extermann, Carrie Klabunde,
  Jeanne Mandelblatt, Vincent Mor, William
  Satariano, Rebecca Silliman
• Geriatrics/Gerontology Luigi Ferrucci, Linda
  Fried, Jack Guralnik, Jerry Gurwitz, Jeffrey
  Halter, William Hazzard, Marco Pahor, Stephanie
  Studenski, Mary Tinetti, Terrie Wetle, Darryl
  Wieland
• Convened by Rosemary Yancik with participation
  from key NIA staff

88
Task Force Objectives

• Identify research opportunities
• interactive health issues affecting older adults
• impacts of comorbidity on treatment efficacy and
  tolerance
• diagnostic, prognostic, treatment, and prevention
  strategies in the presence of comorbidity

89
Thinking about Comorbidity

• What is comorbidity?
• The extent to which comorbidity affects treatment
  for an index condition
• The extent to which the management of an index
  condition affects ongoing treatment of
  pre-existing or concurrent comorbidity
• The interaction of specific conditions
• Overall comorbidity burden

90
Thinking about Comorbidity

• Complicating factors
• Severity of diseases
• Contributions of treatment as well as disease
• Functional status as comorbidity versus an
  outcome
• Influence of behavioral/lifestyle issues

91
Commissioned Papers

• The Nosology of Impairments, Diseases, and
  Conditions
• Severity of Disease Classification Systems The
  Continuum of Conditions, Impairments and Diseases
  
• Methodology, Design, and Analytic Techniques
• Data Sources Relevant to Comorbidity and Aging
  Research

92
I. Nosology of Impairments, Diseases, and
Conditions

• Organizing Principles
• Classified by organ/physiologic/psychological
  systems
• Decrements in health start before onset of
  symptoms
• Accommodates both positive (protective) and
  negative (deleterious) changes
• Avoids arbitrary diagnostic thresholds

93
Includes Thirteen Systems

• mental respiratory
• sensory digestive
• voice/speech metabolic
• cardiovascular endocrine
• hematologic immunologic
• neuromuscular genitourinary
• skin

94
Domains within Individual Systems Streams

• Within each system, there will be one or more
  domains, one for each physiological/
  psychological/functioning measure
• Each domain will be conceptualized as a stream,
  from protective to sub-clinical to overt disease
• Examples Glycosylated hemoglobin
  Blood pressure

95
Using the Nosology

• Comorbidity indices can be created by
• assigning monotonically increasing points within
  each stream
• combining streams using weights
• creating interactions between streams
• Interactions between comorbidity and
  lifestyle/social factors may be important

96
II. Severity of Disease Classification Systems

• Conceptual Framework
• Severity approaches have been developed for
  different purposes
• Screening
• Prognosis
• Defining impact of disease on well-being
• Making treatment decisions
• Determining if treatment alters severity
• Answering specific research questions

97
Points on Causal Pathway for Disease Severity
Assessment Conceptual Framework
Other/Secondary Diseases
Experiential
Outcomes
Disease Process
Pathology Physiology
Symptoms Impairments
Exercise Tolerance Physical Performance
Physical Function Quality of Life Treatment
Required for Control
Increasing Etiologic Specificity
Increasing Relevance to Older Patients
98
II. Severity of Disease Classification Systems

• A three-stage system
• Goal of measurement prognosis, treatment
• Domain classification symptoms, treatments
  required to control symptoms, function, quality
  of life
• Sources of data patient-report, laboratory
  tests, functional tests, health care utilization

99
CHF Goals of Classification Systems
100
CHF Domains for Classification
101
CHF Sources of Information
102
Discussion

• The single disease focus of severity
  classification systems has led to a chaotic array
  of systems not readily amenable to use in the
  study of the import of disease severity.

103
III. Methodology, Design, and Analytic Techniques

• Data Sources
• Inverse relationship between dataset size and
  data quality and quantity
• Missingess varies as a function of data source
• Cost, privacy, feasibility of collection

104
III. Methodology, Design, and Analytic Techniques

• Measurement issues
• Lack of equivalency in relation to outcome e.g.,
  metastatic cancer and diabetes
• Lack of independence e.g., hypertension and
  diastolic dysfunction
• Double counting e.g., when organ
  dysfunction/disease is a severity indicator

105
III. Methodology, Design, and Analytic Techniques

• Analytic Techniques
• Sensitivity analysis estimates uncertainty due
  to non-random error
• Multiple informants uses all available measures
  simultaneously

106
Data Sources Relevant to Comorbidity and Aging
Research

• Review of
• Large comorbidity databases (5000 patients)
• Studies (gt500 patients) containing functional
  status information with either comorbidity
  information, or a potential to retrieve it

107
Epidemiological Studies

• Aging cohorts EPESE, WHAS, LSOA
• Insurance Medicare
• Oncologic SEER, NCCN
• Adult cohorts NHS PHS, WHI

108
Observations

• Studies that have functional information
  frequently do not have detailed comorbidity
  information and vice versa
• Retrospective retrieval of either is limited
  and/or not feasible.

109
Cooperative Clinical Trials in Oncology

• Selection factors (exclusions volunteers) are
  challenging
• Most have reliable and systematic functional
  information
• Comorbidity data usually are poor
• Excellent opportunity for retrospective retrieval
  of comorbidity information

110
Why Should We Care about Comorbidity?

• Etiology
• Prevention
• Treatment Decisions and Benefits/Risks
• Prognosis
• In Short All That We Do