3 - Claims and Disease Management

Question 1

Types of care management models

Answer 1

1. Pre-authorization
2. Concurrent review - monitor care while still in care setting
3. Case management
4. Demand management - passive intervention like nurse advice line
5. Disease management - chronic conditions
6. Specialty case management - expertise in area
7. Population health management - statistical tools for full pop
8. Patient centered medical home - phys responsible for patient care
9. Accountable care organization: network of docs+hospitals share responsibility for patient care
10. Non-traditional provider interventions and care settings - pharmacists and diff types of clinics
11. Gaps in care and quality improvement programs
12. Telehealth, telemedicine, automated monitoring
13. bundled payment initiatives

Question 2

Characteristics of chronic conditions that make them suitable for disease management programs

Answer 2

1. Once contracted, remains with patient for rest of life
2. Manageable with combination of pharmaceutical therapy and lifestyle changes
3. Patients can take responsibility for their own conditions
4. Average annual cost is sufficiently high, warrants resources to manage
5. Expected cost of non-adherent patient is high

Question 3

Principles for establishing a patient centered medical home

Answer 3

1. Personal physician - each patient has personal phys trained to provide comprehensive care
2. Phys-directed medical practice
3. Whole person orientation - arranging care with other qualified professionals
4. Care coordinated and integrated across all elements of the health care system and patient’s community
5. Quality and safety
6. Enhanced access through open scheduling, expanded hours, e-visits
7. Reimbursement structure to support and encourage this model of care

Question 4

Types of interventions conducted by pharmacists

Answer 4

1. Drug utilization review: sub low cost alternatives, prior auth for certain drugs
2. Medication therapy management: Part D required to have. Aim: improve use, reduce adverse events for benes with multiple chronic conditions, taking multiple drugs, expected to incur 4K+ cost
3. Pharmacist-delivered care management programs. Often focus on adherence
   –medication possession ratio = days supply in possession / days in period could have had drug
   –proportion of days covered = days of coverage / total days in period

Question 5

Components of an MTM program

Answer 5

Part D
1. Performing or obtaining necessary assessments of the patient’s health status
2. Formulating a medication treatment plan
3. Selecting, initiating, modifying, or administering medication therapy
4. Monitoring and evaluating the patient’s response to therapy
5. Performing a comprehensive medication review to identify, resolve, prevent medication related problems
6. Documenting care delivered and communicating essential info to patient’s other primary care providers
7. Providing verbal education and training designed to enhance patient understanding and appropriate use of meds
8. Providing info, support services, and resources to enhance patient adherence to drug regimens
9. coordinating and integrating MTM services with other health care management services

Question 6

types of clinics that can be used to provide basic health care

Answer 6

1. retail convenient care clinics
2. employer worksite clinics
3. urgent care clinics
4. federally qualified health centers

Question 7

benefits of being designated an FQHC

Answer 7

1. reimbursement for services provided under Medicare and Medicaid
2. medical malpractice coverage
3. eligibility to purchase medications for outpatients at a reduced cost
4. access to National Health Service Corps
5. access to the Vaccine for Children Program
6. eligibility for various other federal grants and programs

Question 8

possible reasons why DM studies show improved clinical outcomes but not cost savings

Answer 8

1. the measurement of financial outcomes is not stable enough, or measurement techniques are not sensitive enough, to detect positive financial outcomes
2. programs are either not focused on financial outcomes or not structured to optimize financial outcomes
3. program sponsors do not understand the economics of DM programs and therefore do not optimize the programs for financial return
4. improvements in quality of care do not always lead to lower costs. some improvements may actually increase costs, but still be worth the investment

Question 9

financial measures for disease management programs

Answer 9

1. return on investment
   – net roi = (gross savings - cost) / cost
   – gross roi = gross savings / cost
   – costs include direct, indirect, set up, etc
   – gross savings come from decreased utilization as a result of DM program or intervention
2. total savings
   – avg savings = total savings net of program cost / total population
   – marginal savings per chronic member = increase in savings (net of costs) due to intervention on the marginal population, divided by number of mems in marginal population

Question 10

key metrics in the design of disease management programs

Answer 10

1. the number and risk-intensity of members to be targeted - large enough to produce net savings but not so large that marginal cost exceed marginal savings
2. types of interventions to be used in the program - such as mail or automated outbound dialing
3. the number of nurses and other staff needed for the program and program costs
4. methodology for contacting and enrolling members
5. rules for integrating the program with the rest of the care management system
6. timing and numbers of contacts, enrollments, and interventions
7. the predicted behavior of the target population if there were no intervention, and predicted effectiveness of the intervention at modifying that behavior

Question 11

components of the risk management economic model [contributors to financial outcomes of the program]

Answer 11

1. prevalence of different chronic diseases
2. cost of the chronic disease
3. payer risk - most savings for plan when plan at financial risk for all patient costs
4. targeting and risk - member prioritization based on probability of experiencing the targeted event. highest ranks selected for program
5. estimated cost of the targeted event
6. contact rate
7. engagement/enrollment rate
8. member restratification rates

Question 12

common chronic diseases addressed by disease management programs

Answer 12

1. chronic obstructive pulmonary disease
2. heart failure
3. ischemic heart disease
4. diabetes
5. asthma

Question 13

description of opportunity analysis for care management programs

Answer 13

1. definition: a data driven analytical process that extends traditional predictive modeling by matching opportunities within a population to care management programs and services
2. required components to perform analysis:
   –knowledge of member benefit design
   –information on evidence based care management programs currently in place or that could be reasonably introduced
   –eligibility and claims data for past 2-3 years
3. retrospective. use past data to identify opportunities
4. applied prospectively

Question 14

models typically used to stratify members in a care management program

Answer 14

1. stratify members according to the predictive risk score
   –doesn’t necessarily correspond with opportunity
2. condition specific model
   –miss comorbidities
3. rules based approach
   –clinicians not good at identifying patients

opportunity analysis intended to address shortcomings

Question 15

components for designing a care management program using opportunity analysis

Answer 15

1. analytics: segment members, review utilization data vs benchmarks
2. searching evidence base
   –literature review
   –three step approach: search for relevant publications, assess quality of evidence, determine generalizability
3. weigh economics
   –risk rank population using predictive model (expected cost)
   –compare cost against cost without intervention to determine savings
   –savings is compared to cost of intervention to determine economic feasibility

Question 16

steps for implementing a care management program using opportunity analysis

Answer 16

1. develop a predictive model to populate the risk distribution
2. establish a production analysis and reporting unit. develop reports and reporting application
3. determine likely number of care managers required
4. develop a budget for the program, account for all required resources
5. hire and train care managers to conduct interventions and manage patients
6. develop a plan, including estimates of the numbers of patients identified and engaged
7. roll out the intervention and enroll patients
8. operate the program, track outcomes, modify as necessary

Question 17

reasons for using opportunity analysis for identifying patients for care management interventions

Answer 17

1. studies have shown that clinicians are not particularly good at identifying high risk patients
2. the economics of program planning cannot be ignored in a system with limited resources
3. this structured approach is important for understanding which subpopulations are amenable to intervention and the likely value of that intervention
4. the structured financial model provides a framework against which actual outcomes may be compared, identifying areas where the program needs to be corrected or improved

Question 18

description of propensity score matching

Answer 18

1. a technique used for making a participant group comparable to a non participant group. controls for observable variables but not unobservable
2. each member in participant group is matched with a member of nonparticipant group based on propensity scores
3. propensity score = probability that member will be in participant group
   –calculated with logistic regression
   –reduces many variables to single score
   –mems with similar scores matched, even if not matched exactly on independent variables
   –there should still be relatively close matches on those other variables

Question 19

steps for applying propensity score matching to a study

Answer 19

1. run logistic regression. include any independent variables that may influence a persons decision to participate in the program
   – ln(p/(1-p))= a + Bx+e
   – p = exp(a+Bx) / (1 + exp(a + Bx))
2. match participants to non participants
3. test model for appropriateness and bias. difficult since only observable variables. models hould be parsimonious (min # vars necessary to achieve stable model)

Question 20

types of propensity score matching

Answer 20

1. nearest neighbor: closest score, with or without replacement
2. caliper matching: fixed distance
3. mahalanobis metric: measures dissimilarity between two vectors
4. stratification matching: observations stratified then matched by stratum

Question 21

comparison of propensity scoring and risk adjustment

Answer 21

similarities:
1. both reduce effect of multiple risk factors to single score using multiple regression

dissimilarities
1. prop score wider range of independent variables, risk score more detailed diagnosis variables
2. risk adjustment uses entire population, prop score may discard many mems

Question 22

description of actuarially adjusted historical control methodology

Answer 22

1. objective criteria are used to determine which members will be included in the baseline and intervention populations
   –open group method since populations not identical.
   –populations comparable, assumed equivalent, since same selection criteria used
2. savings not directly measured. derived as difference between estimated projection and actual
3. savings formula = [util_base * trend - util_actual] * chronic mems * cost per service

Question 23

issues related to determining and controlling exposure for a disease management study

Answer 23

1. managed vs measured population
2. eligible members
3. member months: mems move categories
4. chronic and non chronic (index) members
5. excluded members
6. measured and non-measured members: tests for inclusion may include continuous coverage test and claim free period
7. enrolled, targeted, and reachable members - avoid bias in results, outcomes should be measured for all targeted members

Question 24

reasons a member may be excluded from a disease management program

Answer 24

1. the member class is not receptive to disease management
2. member is a candidate for a program administered by another vendor (e.g. MH)
3. the pattern of claims that the member exhibits is subject to sharp discontinuity and can distort trend calc
4. members claims are significant and experience likely to dominate the group, or introduce noise to calc

Question 25

conditions that would exclude a member from a disease management program

Answer 25

1. ESRD - DM may delay but cannot reduce
2. transplants
3. HIV / AIDS / mental health - privacy issues make it difficult or impossible for a vendor to receive complete data feeds, or manage the member
4. members who are institutionalized - members may not be reachable or may not benefit from DM interventions
5. members with catastrophic claims
6. eligible for other management programs

Question 26

challenges when calculating disease management savings

Answer 26

1. applying the proper trend rate - non chronic pop typically used
2. demonstrating equivalence between baseline and measurement periods - account for mix of new, continuing, and terminating members and any changes in conditions and co-morbidities. [reweight]

Question 27

how vendors can impact medical cost

Answer 27

1. utilization management - medical necessity, appropriateness of care, medically redundancy
2. site of care - shift to less expensive venues
3. diagnosis or patient type - identify and manage patients with specific diagnosis. reduce ip admissions, er visits
4. severity/downcoding - identify and reverse upcoding/code creep

Question 28

matching models used to measure care management vendor savings

Answer 28

1. pre/post analysis: compare experience period to base period with adj for trend
2. par/non-participating: populations measured in same period, no trend needed
3. regression/trend line analysis: like pre/post, more complex
   –use control population to generate regression equation
4. matched cohort analysis: like par/non, more complex
5. propensity score matching
6. coarsened exact matching: ranges/bins allow greater degree of exact matches between test and control population

Question 29

adjustments/considerations to data or neutralizing factors used in savings calculations

Answer 29

1. scope
2. trend: unit cost, util
3. class of claims: billed/allowed/paid/combo
4. seasonality
5. episodic care
6. care shifting - care moving to other procs not impacted by care mangement program
7. risk adjustment / risk change over time
8. overlap: various initiatives
9. credibility
10. delay in claim impact

Question 30

methods of payment to care management vendors

Answer 30

1. PMPM fee per eligible member
2. capitation: full risk from payer to vendor under fixed PMPM
3. risk share - vendor awarded percentage of savings

Question 31

strengths and weaknesses of propensity score matched population-based cancer cohort study

Answer 31

strengths:
1. large study, consistent exposure and outcome definitions, long period
2. propensity scores matched patients, reduced selection bias
3. study controlled for previously unmeasured confounding variables
weaknesses:
1. study matches those who had similar propensity to have received early palliative care, may not represent entire population of cancer decedents
2. study does not directly measure patient preferences, which is a confounding variable of early palliative care

Question 32

key conclusions from the propensity based cancer cohort study

Answer 32

1. patients receiving early palliative care were more likely to receive supportive home care in their last month of life
2. patients receiving early palliative care were less likely to receive hospital care in last month of life
3. policies and educational strategies to support early palliative care may reduce risk of dying in hospital and receiving aggressive end of life care
4. policies that prohibit palliative care services, like forgoing curative treatments, are disincentives to earlier and concurrent access to palliative care

Question 33

goals of care for emergency department palliative care intervention during COVID 19

Answer 33

1. focus of GOC conversations:
   – conveying prognosis in clear and simple way
   – exploring patients goals and values
   – making care recommendations based on elicited goals
2. GOC outcomes
   – full code (default option): pursue all life sustaining treatments: intubation and CPR
   – do not resuscitate only: all life sustaining treatments but no CPR
   – DNR/do-not-intubate, continue medical treatment
   – comfort directed care - forgo sustaining treatments, deliver symptom focused treatment only