Depression Measure Merit-based Incentive Payment System (MIPS): Measure Information Form

Depression Measure

Merit-based Incentive Payment System (MIPS):

Measure Information Form

2024 Performance Period

Depression Measure Information Form (MIF) 2

Table of Contents

1.0 Introduction 3

1.1 Measure Name 3

1.2 M

easure Description 3

1.3 Measure Rationale 3

1.4 Measure Numerator 5

1.5 Measure Denominator 5

1.6 Data Sources 5

1.7 Care Settings 5

1.8 Cohort 5

2.0 Methodology Steps 6

3.0 Measure Specifications Quick Reference 8

4.0 Detailed Measure Methodology 10

4.1 Identify Patients Receiving Care 10

4.2 Identify the Total Length of Care Between a Patient and a Clinician Group 11

4.3 Define an Episode 12

4.4 Attribute the Episode to a Clinician Group or a Clinician 13

4.5 Assign Costs to an Episode and Calculate Episode Scaled Observed Costs 14

4.6 Exclude Episodes 16

4.7 Estimate Scaled Expected Costs for Risk Adjustment 16

4.8 Calculate Measure Score 18

Appendix A. Example Illustrations of Scenarios for Episode Construction and

Assignment of Days 20

A.1. Simple Example of Defining an Episode 20

A.2. Episode Construction Examples 20

Appendix B. Illustration of Attribution to Individual Clinicians (TIN-NPI) 24

Appendix C. Measure Flowchart for Depression Measure 25

Appendix D. Measure Calculation Example 26

........................................................................................................................

...........................................................................................................

..................................................................................................

.....................................................................................................

...................................................................................................

................................................................................................

.............................................................................................................

........................................................................................................................

............................................................................................................

........................................................................

......................................................................................

..............................................................................

............

....................................................................................................

.........................................

.........

.....................................................................................................

.............................................

.........................................................................................

.........................................................................................................

..................................................................

..............................................................................

..........................

..................................................

..........................................................................

Depression Measure Information Form (MIF) 3

1.0 Introduction

This document details the methodology for the Depression measure and should be reviewed

along with the Depression Measure Codes List file, which contains the medical codes used in

constructing the measure.

1.1 Measure Name

Depression episode-based cost measure

1.2 Measure Description

Episode-based cost measures represent the cost to Medicare for the items and services

provided to a patient during an episode of care (“episode”). In all supplemental documentation,

the term “cost” generally means the standardized

Medicare allowed amount,

and claims data

from Medicare Parts A, B, and D

are used to construct this episode-based cost measure.

The Depression episode-based cost measure evaluates a clinician’s or clinician group’s risk-

adjusted an

d specialty-adjusted cost to Medicare for patients receiving medical care to manage

and treat depression. This chronic condition measure includes the costs of services that are

clinically related to the attributed clinician’s role in managing care during a Depression episode.

1.3 Measure Rationale

An estimated 13% of all Medicare patients age 65 or older experience depression,

and

prevalence of diagnosed major depressive disorder among older adults ranges from 5-10% in

primary care settings and up to 10-42% in inpatient and long-term care settings,

with highest

overall prevalence of major depressive disorder being among the oldest age group of patients

Claim payments are standardized to account for differences in Medicare payments for the same

service(s) across Medicare providers. Payment standardized costs remove the effect of differences in

Medicare payment among health care providers that are the result of differences in regional health care

provider expenses measured by hospital wage indexes and geographic price cost indexes or other

payment adjustments such as those for teaching hospitals. For more information, please refer to the

“CMS Part A and Part B Price (Payment) Standardization - Basics" and “CMS Part A and Part B Price

(Payment) Standardization - Detailed Methods” documents posted on the

CMS Price (Payment)

Standardization Overview

page (https://www.resdac.org/articles/cms-price-payment-standardization-

overview

Claim payments from Part D are payment standardized to allow resource use comparisons for providers

who prescribe the same drug, even if the drug products are covered under varying Part D plans,

produced by different manufacturers, or dispensed by separate pharmacies. For more information, please

refer to the “CMS Part D Price (Payment) Standardization” document posted on the

CMS Price (Payment)

Standardization Overview page

. (https://www.resdac.org/articles/cms-price-payment-standardization-

overview

Cost is defined by allowed amounts on Medicare claims data, which include both Medicare trust fund

payments and any applicable beneficiary deductible and coinsurance amounts.

Part D branded drug costs are also adjusted to account for post-point of sale drug rebates; more

information can be found in the Methodology for Incorporation of Rebates in Part D Standardized

Amounts on the QPP Cost Measure Information "About Cost Measures" page

(

https://www.cms.gov/medicare/quality-payment-program/cost-measures/about).

Ewald E, Loganathan S, Hache J, Lochner K. Access to Care Among Medicare Beneficiaries With and

Without Depression. Medicare Current Beneficiary Survey. 2017. https://www.cms.gov/Research-

Statistics-D

ta-and-Systems/Research/MCBS/Downloads/ATC_Depression_2017.pdf

McQuaid, JR, Lin EH, Barber JP, et al. 2019. Clinical Practice Guideline for the Treatment of Depression

Across Three Age Cohorts. American Psychological Association, Guideline Development Panel for the

Treatment of Depressive Disorders.

Depression Measure Information Form (MIF) 4

90 years or older.

Beyond considering the impacts and costs of treatment for depression itself,

findings from the literature report that patients with depression are more likely to utilize health

care services and resources for other types of medical illness beyond just mental health

disorders compared to patients without depression,

and the average total health care costs for

patients with depression aged 60 or older are greater compared to those without depression.

Researc

h has also indicated that the integration of primary care and mental health care could

reduce spendi

ng and lead to improvements in the management, treatment, and quality of care

for patients with mental health disorders and chronic conditions,

and specifically that ongoing

depression disease management for patients with major depression can increase clinical

improvement and be less costly overtime.

In addition to improving care integration, research

indicates another opportunity for improvement in effective case management to improve

medication and treatment adherence. A 2020 paper indicated that that over half of patients with

major depressive disorder do not adhere to prescribed medications (i.e., antidepressants), both

in the primary care and psychiatric settings.

which may be due both to patient-related factors

(i.e., due to concerns about side-effects, cultural issues, costs

12,13

), as well as factors that

clinicians can influence, such as inadequate patient education, lack of shared decision-making,

and lack of follow-up.

Given the prevalence of major depressive disorder in the Medicare population, and the high

costs as

sociated with the management of the disease and its complications, the Depression

cost measure represents an opportunity for improvement on overall cost performance. The

Depression episode-based cost measure was selected for development because of its high

impact in terms of patient population, clinician coverage, and Medicare spending, and the

opportunity build a complex, yet feasible, chronic condition measure that would address a

condition not captured by other cost measures. Following initial feedback gathered during the

Bashyal R, Du H, Wang L, Yuce H, Baser O. PMH17 – Mortality and Prevalence of Major Depressive

Disorder in the US Medicare Population from 2008-2013. Value in Health. 2016; 19(3): A184.

Zivin K, Wharton T, Rostant O. The Economic, Public Health, and Caregiver Burden of Late-Life

Depression. Psychiatric Clinics of North America. 2013; 36(4): 631-649.

Katon WJ, Lin E, Russo J, et al. Increased Medical Costs of a Population-Based Sample of Depressed

Elderly Patients. JAMA Psychiatry. 2003; 60(9):897-903.

Bao Y, Casalino LP, Pincus HA. Behavioral Health and Health Care Reform Models: Patient-Centered

Medical Home, Health Home, and Accountable Care Organization. Journal of Behavioral Health Services

and Research. 2013; 40(1):121-132.

Rost K, Pyne J, Dickinson LM, LoSasso A. Cost-Effectiveness of Enhancing Primary Care Depression

Management on an Ongoing Basis. Annals of Family Medicine. 2005; 3(1): 7-14.

Dell’Osso B, Albert U, Carra G, et al. How to Improve Adherence to Antidepressant Treatments in

Patients with Major Depression: A Psychoeducational Consensus Checklist. Annals of General

Psychiatry. 2020; 19(61).

Piette JD, Heisler M, Wagner TH. Cost-Related Medication Underuse Among Chronically Ill Adults: The

Treatments People Forgo, How Often, And Who Is At Risk. American Journal of Public Health.

2004;941782-1787.

Bambauer KZ, Safran DG, Ross-Degnan D, et al. Depression and Cost-Related Medication

Nonadherence in Medicare Beneficiaries. JAMA Psychiatry. 2007; 64(5):602-608.

Dell’Osso B, Albert U, Carra G, et al. How to Improve Adherence to Antidepressant Treatments in

Patients with Major Depression: A Psychoeducational Consensus Checklist. Annals of General

Psychiatry. 2020; 19(61).

Depression Measure Information Form (MIF) 5

Wave 4 public comment period,

the subsequent measure-specific clinician expert workgroup

provided extensive, detailed input on this measure.

1.4 Measure Numerator

The measure numerator is the weighted average ratio of the winsorized

scaled standardized

observed cost to the scaled expected

cost for all Depression episodes attributed to a clinician,

where each ratio is weighted by each episode’s number of days assigned to a clinician. This

sum is then multiplied by the national average winsorized scaled observed episode cost to

generate a dollar figure.

1.5 Measure Denominator

The measure denominator is the total number of days from Depression episodes assigned to

the clinician across all patients.

1.6 Data Sources

The Depression measure uses the following data sources:

• Medicare Part A, B, and D claims data from the Common Working File (CWF)

• Enrollment Database (EDB)

• Long Term Care Minimum Data Set (LTC MDS)

1.7 Care Settings

The Depression measure focuses on the care provided by clinicians practicing in non-inpatient

hospital settings for patients with depression. The most frequent settings in which a Depression

episode is triggered include: office, nursing facility, skilled nursing facility (SNF), and outpatient

hospital.

1.8 Cohort

The cohort for this cost measure consists of patients who are Medicare beneficiaries enrolled in

Medicare fee-for-service that receive care for depression.

The cohort for this cost measure is also further refined by the definition of the episode group

and measure-

specific exclusions (refer to Section 4).

“Wave 4 Public Comment Summary,” QPP Cost Measure Information, Prior cost measure development

and input, (https://www.cms.gov/medicare/quality-payment-pro

gram/cost-measures/prior

)

For information on how costs are winsorized, please refer to Section 4.7.

Expected costs refer to costs predicted by the risk adjustment model. For more information on

expected costs and risk adjustment, please refer to Section 4.7.

For information on how LTC MDS data are used in risk adjustment, please refer to Section 4.7.

Depression Measure Information Form (MIF) 6

2.0 Methodology Steps

There are 2 overarching processes in calculating chronic condition episode-based cost measure

scores: episode construction (Steps 1-5) and measure calculation (Steps 6-8). This section

provides a brief summary of these processes for the Depression measure. Section 4 describes

the pr

ocesses in detail and further defines the related concepts, and Appendix C contains a

sual flowchart depicting these steps.

1. Identify patients receiving care: A trigger event identifies the start or continuation of a

clinician group’s management of a patient’s chronic condition. A trigger event is identified by

the occurrence of 2 Part B Physician/Supplier (Carrier) claims billed by the same clinician

group practice within 180 days of one another. The pair of services must include a trigger

claim and a confirming claim. The trigger claim is any code from a set of CPT/HCPCS codes

for clinically relevant outpatient services when accompanied by an ICD-10 diagnosis code

indicating depression. The confirming claim is the instance of another trigger code. Once a

trigger event is identified, this opens a one-year attribution window from the point of the

trigger claim, in which the patient’s chronic condition care will be monitored by a clinician

group.

2. Identify the total length of care between a patient and a clinician group: O

nce an

attribution window is opened, it continues for 1 year unless there is a service that

demonstrates a continuing care relationship, also known as a reaffirming claim. This service

is billed during an open attribution window (from Step 1) by the same clinician group that

billed the trigger event, and reaffirms and extends a clinician group’s responsibility for

managing a patient’s chronic condition. A reaffirming claim is another instance any

confirming code.

After a reaffirming claim is identified, the attribution window is extended

by 1 year from the point of each reaffirming claim billed during an open attribution window.

The total attribution window begins with the trigger claim and concludes 1 year after the final

reaffirming claim. Therefore, the total attribution window can span multiple years and vary in

length for different patients. This requires that the total attribution window is measured

incrementally and periodically across multiple measurement periods.

3. Define an episode: E

pisodes are segments of the total attribution window that are counted

in a particular measurement period, allowing clinicians to have their costs for Depression

episodes assessed for that year. Episodes are assigned to a clinician group (identified by

Tax Identification Number [TIN]) or individual clinicians (identified by combination of TIN and

National Provider Identifier [TIN-NPI]), and can vary in length. Episodes are assessed in the

measurement period in which they conclude and only attribute days not previously

measured in preceding measurement periods, so there is no double counting of episode

costs.

4. Attribute the episode to the clinician group and clinician(s): The episode is attributed to

the clinician group that bills the trigger and confirming claims for the total attribution window.

To attribute the episode to an individual clinician, any clinician within the attributed clinician

group who plays a substantial role in the care for the patient (i.e., billing at least 30% of

trigger or confirming codes on Part B Physician/Supplier claim lines during the episode) is

While a trigger event requires two claims, a single reaffirming claim is needed to extend a clinician

group’s responsibility for managing a patient’s chronic condition. This is because workgroups who have

developed chronic condition measures to-date have favored a less strict reaffirming algorithm, indicating

that once a clinician-patient relationship was established, a single reaffirming claim would be sufficient to

extend the attribution window.

Depression Measure Information Form (MIF) 7

attributed the episode. There are also additional checks to ensure that clinicians are not

attributed to an episode before they have their first encounter with the patient.

5. Assign cost

s to the episode and calculate the episode scaled observed cost: Services

that are clinically related to the care and management of a patient’s chronic condition that

occur during the episode are included in the measure. The standardized cost of the

assigned services is summed and averaged across the number of days in an episode. This

average daily cost is then multiplied by 365 to determine each episode’s scaled (i.e.,

annualized) standardized observed cost.

6. Exclude episodes

: Exclusions remove unique groups of patients or episodes from cost

measure calculation in cases where it may be impractical or unfair to compare the costs of

caring for these patients to the costs of caring for the cohort at large.

7. Calculate the scaled expected cost for risk adjustment: Risk adjustment predicts the

expected costs by adjusting for factors outside of the clinician’s or clinician group’s

reasonable influence (e.g., patient age, comorbidities, dual Medicare and Medicaid eligibility

status, clinician specialty, and other factors). The episode group’s scaled standardized

observed costs are winsorized at the 98

percentile for each model to handle extreme

observations. A regression is then run using the risk adjustment variables as covariates to

estimate the expected cost of each episode. Further statistical techniques are applied to

reduce the effects of extreme outliers on measure scores.

8. Calculate the measure score: For each episode, the ratio of winsorized scaled

standar

dized observed cost to scaled expected cost (both of which are from Step 7) is

calculated. The measure is calculated as a weighted average of these ratios across all of a

clinician’s or clinician group’s attributed episodes, where the weighting is each episode’s

number of assigned days. The weighted average episode cost ratio is then multiplied by the

national average winsorized scaled observed episode cost to generate a dollar figure for the

cost measure score.

Depression Measure Information Form (MIF) 8

3.0 Measure Specifications Quick Reference

This page provides a quick, at-a-glance reference for the Depression measure specifications.

More details on each component can be found in Section 4, and the full list of codes and logic

used to defi

ne each component can be found within the Depression Measure Codes List file.

Episode Window: During what time period are costs measured?

An episode is a segment of time during which clinicians or clinician groups are assessed for the

car

e that they provide to a patient with depression.

• The episode window length for the Depression measure is between 1 year (365 days) and 2

years minus 1 day (729 days), and can vary in length across patients.

Triggers: How does the measure identify the patient cohort and start of care?

• Patients receiving medical care for treatment of their depression are included in the

measure.

• The start or continuation of a clinician group’s management of a patient’s depression is

identified by the appearance of a pair of services within 180 days of one another: a trigger

code followed by a confirming code. For the Depression measure:

• A trigger code i

s any code from a set of CPT/HCPCS codes for clinically relevant

outpatient services (outpatient E/Ms, behavioral health visits, collaborative care

management, brain stimulation therapy, chronic care management, complex chronic care

E/M, telehealth visits, in-home visits) when accompanied by an ICD-10 diagnosis code

indicating depression.

• A confirming code is any code from the same trigger set of CPT/HCPCS codes for

clinically relevant outpatient services when accompanied by an ICD-10 diagnosis code

indicating depression.

Sub-Groups: Is the measure stratified into smaller patient cohorts?

1. Depression w

ith Psychotic Features

2. Depression w

ithout Psychotic Features

Service Assignment: Which clinically related costs are included in the measure?

Assigned services generally fall within the following clinical themes:

• Outpatient Services and Psychotherapy

• Outpatient Infusions and Diagnostics (Laboratory, Imaging, EEG)

• (Re)admissions, ED Visits, Other

• Transportation

• Post-Acute Care and Ancilla

ry Services

Risk Adjustors: Which risk factors are accounted for in the risk adjustment model?

• Standard risk adjustors, including comorbidities captured by 86 Hierarchical Condition

Category (HCC) codes that map with thousands of ICD-10-CM diagnosis codes, count of

Cs, interaction variables accounting for a range of comorbidities, patient age category,

patient disability status, patient end-stage renal disease (ESRD) status, patient dual

eligibility status, number and types of clinician specialties from which the patient has

received care, and recent use of institutional long-term care.

• Risk adjustors for factors specific to this condition:

two or more hospitalizations related to

depression in the prior year, memory loss, eating disorders, chronic pain, suicide ideation,

recent suicide attempt, esketamine within the last year, transcranial magnetic stimulation

(TMS) within the last year, electroconvulsive therapy (ECT) within the last year, and prior

observation stays that may indicate treatment-resistant depression (TRD). For the full list of

Depression Measure Information Form (MIF) 9

standard and measure-specific risk adjustment variables, please reference the “RA” and

“RA_Details” tabs of the Measure Codes List file.

• A separate linear regression is run for each sub-group and Medicare Part D enrol

lment

status combination to ensure fair comparison. The episode group’s scaled (i.e., annualized)

observed costs are winsorized at the 98

percentile prior to the regression for each model to

handle extreme observations.

Exclusions: Which populations are excluded from the measure?

• Measure-specific exclusions including bipolar disorder, schizophrenia, and drug or alcohol

psy

chosis. For the full list of measure-specific exclusions, please reference the “Exclusions”

and “Exclusions_Details” tabs of the Measure Codes List file.

• Standard exclusions to ensure data completeness:

• The patient has a primary payer other than Medicare for any time overlapping the

epis

ode window or 120-day lookback period prior to the episode window.

• The patient was not enrolled in Medicare Parts A and B for the entirety of the 120-day

look

back period plus episode window, or was enrolled in Part C for any part of the

lookback plus episode window.

• The patient was not found in the Medicare Enrollment Database (EDB).

• The patient’s death date

occurred before the episode end date.

• The patient has an episode window shorter than one year.

• The patient has extremely low treatment costs.

• The patient resided outside the United States or its territories during the episode

ndow.

Depression Measure Information Form (MIF) 10

4.0 Detailed Measure Methodology

This section contains the technical details for the 2 overarching processes in calculating the

Depression cost measure in more detail: Sections 4.1 through 4.5 describe episode

cons

truction, and Sections 4.6 through 4.8 describe measure calculation.

4.1 Identify Patients Receiving Care

A trigger event is used to indicate the start of a clinician group’s management of a patient’s

depression and is identified by the occurrence of 2 Part B Physician/Supplier (Carrier) claims

billed by the same clinician group practice. To identify a trigger event, the following 2 claims

must be billed within the trigger window (within 180 days of one another): a trigger claim,

followed by a confirming claim.

• A trigger claim i

s a Part B Physician/Supplier claim that contains a trigger code. For the

Depression measure, a trigger code is:

• Any code from a set of CPT/HCPCS codes for clinically relevant outpatient

ser

vices when accompanied by an ICD-10 diagnosis code indicating depression.

These outpatient services can be summarized as:

• Outpatient ev

aluation and management (E/M) codes

• Behavioral health visits

• Collaborative care management

• Brain stimulation therapy

• Chronic care management

• Complex chronic care E/Ms

• Telehealth visits

• In

-home visits

• A confirming claim is a second Part B Physician/Supplier claim billed by the same

cli

nician group practice as the trigger claim, which contains a confirming code. For the

Depression measure, a confirming code is:

• Any code from the same set of CPT/HCPCS codes for clinically relevant

outpatient s

ervices when accompanied by an ICD-10 diagnosis code indicating

depression, as listed above in trigger codes.

For the ful

l list of trigger and confirming codes, as well as the requisite diagnosis codes, please

refer to the “Trigger_Confirming” and “Trigger_DGN” tabs of the Depression Measure Codes

List file.

Once the trigger event is identified, the trigger event opens an attribution window, which is a

year-long time period that begins on the date of the trigger claim. The attribution window defines

a time period during which the patient’s depression care will be monitored by a clinician group.

Depression Measure Information Form (MIF) 11

Figure 1. Trigger Event and Attribution Window

4.2 Identify the Total Length of Care Between a Patient and a

Clinician Group

When the beginning of the clinician-patient relationship is identified, there might be evidence of

a continuation of this relationship, as identified by reaffirming claims. A reaffirming claim is a

service billed during an open attribution window by the same clinician group that billed the

trigger event, and it reaffirms and extends a clinician group’s responsibility for managing a

patient’s depression. A reaffirming claim has the same definition as a confirming claim as

defined in Section 4.1, meaning that a reaffirming claim is:

• Any code from the set of trigger CPT/HCPCS codes for clinically relevant outpatient

ervices when accompanied by an ICD-10 diagnosis code indicating depression. These

outpatient services can be summarized as:

• Outpatient E/Ms

• Behavioral health visits

• Collaborative care management

• Brain stimulation therapy

• Chronic care management

• Complex chronic care E/M

• Telehealth visits

• In-home visits

Each time a reaffirming claim is identified during an open attribution window, the attribution

window will be extended by 1 year from the point of the reaffirming claim. The resulting overall

time period of responsibility is defined as the total attribution window, which begins with the

trigger claim and concludes 1 year after the final reaffirming claim. Therefore, the total

attribution window can span multiple years and vary in length for different patients. Appendix A

ontains an illustration of the relationship between a trigger event, reaffirming claims, and a total

attribution window.

Figure 2 below contains an example illustration of the relationship between a trigger event,

reaffirming claims, and a total attribution window. In this hypothetical example, reaffirming claim

1 occurs 6 months into attribution window 1 and extends that attribution window by 1 year (until

the end of attribution window 2), and then reaffirming claim 2 occurs 9 months into attribution

window 2, extending that attribution window by another year (until the end of attribution window

3). Once all reaffirming claims are identified, the total period of time of the clinician-patient

relationship is defined as the period covered by all attribution windows, beginning with the

Depression Measure Information Form (MIF) 12

trigger claim and concluding 1 year after the final reaffirming claim. For this example, the total

attribution window is 27 months long.

Figure 2. Example of Reaffirming Claims and Total Attribution Window

4.3 Define an Episode

Once the total attribution window has been constructed, it is divided into segments of time, also

known as episodes. Episodes allow the measure to be calculated for a given measurement

period, which is a static year-long period (i.e., calendar year) in which a clinician or clinician

group will be measured.

An episode is defined, at a minimum, as a one-year segment of the total attribution window.

pisodes are assessed in the measurement period in which they end and only include days not

previously measured in preceding measurement periods. Clinicians or clinician groups are

measured on a patient at the end of the calendar year if there are at least 365 days’ worth of

claims data that has not previously been assessed or when the total attribution window ends,

ensuring that costs are only assessed once. The episode window lengths may vary depending

on the length of the total attribution window and the number of days that have not been

assessed in preceding measurement periods.

After the episode windows are constructed, the number of assigned days for each episode is

determined and used as a weighting factor in the measure score calculation step. This

weighting is done to ensure fair comparison across episodes, where cost is effectively scaled

respective to the episode length to allow like comparisons between episodes of similar length.

Appendix A contains a simplified example of episode construction, as well as a more detailed

llustration of episode construction and assignment of days.

1. 365-day episode window, w

here there are no reaffirming claims during the year-long

total attribution window

• The epis

ode start date is set as the start date of the total attribution window.

• The epis

ode end date is set as 365 days after the episode start date.

• Assign the total num

ber of days that have not been previously measured in the

preceding episodes. In this case, the number of assigned days equals the

number of days in the episode.

2. 366- to 729-

day episode window, where reaffirming claims extend the total attribution

window to greater than one year

• The epis

ode start date is set as the start date of the total attribution window.

• The epis

ode end date is set as either:

Depression Measure Information Form (MIF) 13

• The end of the total attribution window (which is 366 to 729 days after the

episode start date), if the total attribution window ends by December 31 of

the next calendar year (i.e., the measurement year);

• December 31 of the next full calendar year (which is 366 to 729 days after

the episode start date), if the total attribution window extends beyond

December 31 of the next calendar year (i.e., the measurement year).

• Assign the total number of days that have not been previously measured in the

preceding episodes. In this case, the number of assigned days equals the

number of days in the episode.

3. 365-day episode window, where reaffirming events have resulted in a total attribution

window that is at least two years in length that can be split into 365-day segments

across multiple measurement periods

• The episode start date is set as the beginning of a new calendar year (January

1) if it is a subsequent episode with at least 365 days’ worth of claims data not

captured in a preceding measurement period.

• The episode end date is set as 365 days after the episode start date, at the end

of that calendar year (December 31).

• Assign the total number of days that have not been previously measured in the

preceding episodes. In this case, the number of assigned days equals the

number of days in the episode.

4. 365-day episode window, where the total attribution window concludes after a segment

was measured in the previous measurement period

• The episode start date is set as 365 days prior to the total attribution window

end date if the remaining number of assigned days in the total attribution window

is less than 365 days.

• The episode end date is set as the end date of the total attribution window.

• Assign the total number of days that have not been previously measured in the

preceding episodes. In this case, the number of assigned days is smaller than

the number of days in the episode, since the episode window would partially

overlap with the preceding episode window. Only days not previously measured

are assigned to the episode. This is done to ensure there is no double counting

of episode costs.

Once a Depression episode window is defined, the episode is placed into one of the episode

sub-groups to enable meaningful clinical comparisons. Sub-groups represent more granular,

mutually exclusive and exhaustive patient populations defined by clinical criteria. Sub-groups

are useful in ensuring clinical comparability so that the corresponding cost measure fairly

compares clinicians with a similar patient case-mix.

Codes used to define the sub-groups can be found in the “Sub_Groups” and

“

Sub_Groups_Details” tabs of the Depression Measure Codes List file. This cost measure has 2

sub-groups:

• Depression with Psychotic Features

• Depression without Psychotic Features

4.4 Attribute the Episode to a Clinician Group or a Clinician

Once an episode has been defined, it is attributed to one or more clinicians of a specialty that is

eligible for MIPS. The episodes are attributed to clinician groups, who are identified by their

unique TIN, and individual clinicians, who are identified by their TIN and NPI pair (TIN-NPI). For

codes relevant to this section, please refer to the “Attribution” tab of the Depression Measure

Codes List file.

Depression Measure Information Form (MIF) 14

TIN level attribution: An episode is attributed to the clinician group that billed the trigger event

(trigger and confirming claims) for the total attribution window. The clinically related costs from

the total number of assigned days are attributed to that clinician group.

TIN-NPI level attribution: An episode is attributed to any clinician within the attributed clinician

group that billed at least 30% of the trigger or confirming codes on Part B Physician/Supplier

claim lines during the episode.

The measure’s attribution methodology also imposes additional

checks to ensure that TIN-NPIs are appropriately attributed. Specifically, TIN-NPIs that meet the

30% threshold must have billed at least one trigger or confirming code within 1 year prior to or

on the episode start date.

Future attribution rules may benefit from the implementation of patient relationship categories

and codes.

As required by section 101(f) of the Medicare Access and CHIP Reauthorization

Act of 2015 (MACRA), the Centers for Medicare & Medicaid Services (CMS) will consider how

to incorporate the patient relationship categories into episode-based cost measurement

methodology as clinicians and billing experts gain experience with them.

4.5 Assign Costs to an Episode and Calculate Episode Scaled

Observed Costs

Medicare Parts A, B, and D services, and their costs, are assigned to an episode only when

clinically related to the management and treatment of the patient’s depression during the

episode. Assigned services may include treatment and diagnostic services, ancillary items,

services directly related to treatment, and those furnished as a consequence of care (e.g.,

complications, readmissions, unplanned care, and emergency department visits). Unrelated

services are not assigned to the episode. For example, the cost of care for a procedure that

occurs during the episode that is not clinically related to the management and treatment of the

patient’s depression (i.e., a knee arthroplasty) would not be assigned to the episode.

To ensure that only clinically related services are included, services during the episode window

are assigned to the episode based on a series of service assignment rules, which are listed in

the “Service_Assignment_AB” and “Service_Assignment_D” tabs of the Depression Measure

Codes List file.

For the Depression episode group, services performed in the following service categories are

consider

ed for assignment to the episode:

• Outpatient (OP) Facility and Clinician Services

• Emergency Department (ED)

• Inpatient (IP) – Medic

For a diagram illustrating an example of attribution to a TIN and TIN-NPI, please refer to Appendix B.

The MACRA Patient Relationship Categories aim to distinguish the relationship and responsibility of a

clinician with a patient at the time of furnishing an item or service, thereby facilitating the attribution of

patients and episodes to one or more clinicians for purposes of measure score calculations. For more

information on Patient Relationship Categories, please refer to the Patient Relationship Categories and

codes operational list. (

https://www.cms.gov/Medicare/Quality-In

itiatives-Patient-Assessment-

Instruments/Value-Bas

ed-Programs/MACRA-MIPS-and-APMs/CMS-Patient-Relationship-Categories-

and-Codes.pdf

)

The MACRA Patient Relationship Codes are HCPCS Level II modifier codes that clinicians report on

claims to identify their patient relationship category. For the Patient Relationship Codes, please refer to

Table 27 of the CY 2018 Physician Fee Schedule final rule. (

https://www.federalregister.gov/d/2017-

23953/p-2203

)

For more information on the Patient Relationship Categories and Codes, please download the Patient

Relationship Categories and Codes FAQ. (https://qpp-cm-prod-

ontent.s3.amazonaws.com/uploads/236/Patient-Rel

ationship-Categories-and-Codes-webinar-FAQ.pdf)

Depression Measure Information Form (MIF) 15

• IP – Surgical

• Inpatient Rehabilitation Facility (IRF),

Long Term Care Hospital (LTCH), SNF

• Durable Medical Equipment, Prosthetics, Orthotics, and Supplies (DME)

• Home Health (HH)

• Part D drugs

In addition to service category, s

ervice assignment rules may be modified based on the service

category in which the service is performed, as listed above. Service assignment rules can also

be defined based on specific service information alone or service information combined with

diagnosis information. Services may be assigned to the episode based on the following

combinations:

• High level service code alone

• High level service code combined with first 3 di

gits of the ICD-10 diagnosis code

• High level service code combined with more specific service code

• High level service code combined with more specific service code and with 3-digit ICD-

10 diagnos

is code

The steps for assigning costs are as follows:

• Identify al

l services on claims with positive standardized payment that occur within the

episode window.

• Assign i

dentified services to the episode based on the types of service assignment rules

described above.

• Assign al

l trigger and reaffirming Part B Physician/Supplier claims occurring during the

episode window.

• Assign al

l SNF stays based on the following criteria:

• Identify SNF stays where both (

i) the SNF stay’s qualifying IP stay is assigned to

episode and (ii) the SNF stay occurs during the episode window.

• For those identified SNF stays,

determine the number of days that overlap with

the episode window; if the overlap is greater than 30 days, cap claim amount

assigned to the episode at 30 days.

• Assign all IRF and LTCH stays based on the following criteria:

• Identify IRF and LTCH stays for

which (i) there is a preceding IP stay discharged

within 7 days prior to the stay’s start date, (ii) the preceding IP stay is assigned to

the episode, and (iii) the IRF and LTCH stays occur during the episode window.

• For those identified IRF and LTCH stays, determine the distribution of grouped

aim cost across episodes and cap claim amount assigned to the episode at the

percentile of each observed cost distribution.

• Assign al

l inpatient evaluation and management (E/M) claims during IP stays assigned

to episode.

• Sum the standar

dized Medicare allowed amounts for all claims assigned to each

episode to obtain the total standardized episode observed cost.

• Average the

total standardized episode observed cost over the number of days in the

episode to get the episode average daily standardized observed cost.

Services performed in the IRF, LTCH, and SNF settings are assigned to an episode based on their

association with the grouped IP stay.

Capping costs aims to limit the effects of extreme observed cost values on episode observed costs.

Capping involves limiting the amount of claim costs that a provider can be assigned during an episode.

For Major Depressive Disorder episodes with related LTCH and/or IRF costs, the value of the 90

percentile is assigned to all LTCH and IRF observed costs above the 90

percentile.

Depression Measure Information Form (MIF) 16

• Multiply the episode average daily standardized observed cost by 365 to get the

episode scaled (annualized) standardized observed cost.

4.6 Exclude Episodes

Before measure calculation, episode exclusions are applied to remove certain episodes from

measure score calculation. Certain exclusions are applied across all chronic condition episode

groups, and other exclusions are specific to this measure, based on consideration of the clinical

characteristics of a homogenous patient cohort. All measure-specific exclusions are listed in the

“Exclusions” and “Exclusions_Details” tabs in the Depression Measure Codes List file.

Episodes are excluded from the Depression measure if they meet any of the following cross-

epis

ode group conditions:

• The patient has

a primary payer other than Medicare for any time overlapping the

episode window or 120-day lookback period prior to the episode window.

• The patient w

as not enrolled in Medicare Parts A and B for the entirety of the 120-day

lookback period plus episode window, or was enrolled in Part C for any part of the 120-

day lookback period plus episode window.

• The patient is not found in the Medicare EDB.

• The patient has

an episode window shorter than 1 year.

• The patient’s death date occurred before the episode end date.

• The patient has extremely low treatment costs.

• The patient r

esided outside the United States or its territories during the episode

window.

4.7 Estimate Scaled Expected Costs for Risk Adjustment

Risk adjustment is used to estimate episode expected costs in recognition of the different levels

of care patients may require due to comorbidities, disability, age, specialty care, and other risk

factors. The risk adjustment model includes variables from the CMS Hierarchical Condition

Category Version 24 (CMS-HCC V24) 2021 Risk Adjustment Model,

as well as other standard

risk adjustors (e.g., patient age) and variables for clinical factors that may be outside the

attributed clinician’s reasonable influence. A full list of risk adjustment variables can be found in

the “RA” and the “RA_Details” tabs of the Depression Measure Codes List file.

Steps for defining risk adjustment variables and estimating the risk adjustment model are as

foll

ows:

CMS uses an HCC risk adjustment model to calculate risk scores. The HCC model ranks diagnoses

into categories that represent conditions with similar cost patterns. Higher categories represent higher

predicted healthcare costs, resulting in higher risk scores. The 86 HCC codes included in the CMS-HCC

V24 model are mapped to thousands of ICD-10-CM diagnosis codes.

Service Assignment Example

• Clinician Group A has been providing continuous care management for Patient K’s

depr

ession, and is attributed an episode with Patient K during the measurement period.

• Clinician Group A provides a psychotherapy visit for Patient K during the episode

window. Because the psychotherapy visit is considered a clinically related service, its

costs will be assigned to Clinician Group A’s Depression episode with Patient K.

Depression Measure Information Form (MIF) 17

• Define HCC, number and types of clinician specialties from which the patient has

received care,

and episode group-specific risk adjustors using service and diagnosis

information found on the patient’s Medicare claims history in the 120-day default

lookback period prior to the episode start date (or the timing specified in the

“RA_Details” tab of the Measure Codes List file) for certain billing codes that indicate the

presence of a procedure, condition, or characteristic. For clinician specialty information,

include information obtained on the episode start date.

• Create the fol

lowing categories to identify HCC frequency as a marker of patient

comorbidity: 0, 1, 2-3, 4-6, and 7+ HCCs.

• Define other

risk adjustors that rely upon Medicare beneficiary enrollment and

assessment data as follows:

• Identify beneficiaries who are originally “Disabled without ESRD” or “Disabled

th ESRD” using the original reason for joining Medicare field in the Medicare

beneficiary EDB.

• Identify beneficiaries with ESRD if their enrollment indicates ESRD coverage,

RD dialysis, or kidney transplant in the Medicare beneficiary EDB in the 120-

day lookback period.

• Identify beneficiaries who have spent at least 90 days in a long-term care

ins

titution (LTCI) without having been discharged to the community for 14 days,

using LTC MDS assessment data. Then, identify the beneficiaries whose

Depression episode start date overlaps with their stay in an LTCI.

• Identify benefi

ciaries who have partial or full dual Medicare and Medicaid

eligibility status as of the episode start date; adjust for dual eligibility status when

risk-adjusted costs are on average higher for dually enrolled beneficiaries (i.e.,

drop risk adjustor when coefficient is less than 0).

• Drop r

isk adjustors that are defined for less than 15 episodes nationally for each sub-

group and Part D enrollment status combination to avoid using very small samples.

• Categorize benefi

ciaries into age ranges using their date of birth information in the

Medicare beneficiary EDB. If an age range has a cell count less than 15, collapse this in

the next adjacent age range category towards the reference category (65-69).

The following steps are performed separately for each sub-group and Part D enrollment

status combination:

• Winsorize

the episode scaled observed cost as follows:

• Assign the v

alue of the 98

percentile to all episode scaled observed costs

above the 98

percentile.

Specialty groups include Allergy/Immunology, Anesthesiology, Behavioral Medicine, Cardiology, Chest

Surgery, Chiropractor, Critical Care, Dentistry, Dermatology, Diagnostic Imaging, Emergency Medicine,

Endocrinology, Facility, Gastroenterology, General Medicine/Family Practice, General Surgery and

Surgical Oncology, Hospice and Palliative Care, Infectious Disease, Interventional Radiology,

Nephrology, Neurology, Neurosurgery, NP/PA/Nurse Specialists.

Winsorization aims to limit the effects of extreme values on expected costs. Winsorization is a statistical

transformation that limits extreme values in data to reduce the effect of possible outliers. Winsorization of

the lower end of the distribution (i.e., bottom coding) involves setting extremely low predicted values

below a predetermined limit to be equal to that predetermined limit, and similarly for the higher end of the

distribution involves setting extremely high predicted values above a predetermined limit to be equal to

that predetermined limit.

Depression Measure Information Form (MIF) 18

• Run an ordinary least squares (OLS) regression model to estimate the relationship

between all the risk adjustment variables and the dependent variable, the episode

winsorized scaled observed cost calculated from the previous step, to obtain the episode

scaled expected cost.

• Winsorize the epis

ode scaled expected cost as follows:

• Assign the val

ue of the 0.5

percentile to all episode scaled expected costs

below the 0.5

percentile.

• Renormalize

values by multiplying each episode’s winsorized scaled expected

cost by the average scaled expected cost, and dividing the resultant value by the

average winsorized scaled expected cost.

• Exclude episodes with outliers as follows:

• Calculate each epi

sode's residual as the difference between the re-normalized,

winsorized scaled expected cost computed above and the winsorized scaled

observed cost.

• Exclude epis

odes with residuals below the 1

percentile or above the 99

percentile of the residual distribution.

• Renormalize the res

ultant scaled expected cost values by multiplying each

episode’s winsorized scaled expected costs by the average winsorized scaled

observed cost (after excluding outliers), and dividing by the average winsorized

scaled expected cost (after excluding outliers).

4.8 Calculate Measure Score

Measure scores are calculated for a clinician or clinician group practice as follows:

• Calculate the ratio of winsorized scaled standardized observed cost to scaled expected

episode cost for each episode attributed to the clinician or clinician group.

• Calculate the meas

ure as a weighted average of these ratios across all of a clinician’s

or a clinician group’s attributed episodes, where the weighting is the number of assigned

days for a clinician or a clinician group during the episode.

• Multiply the weighted average episode cost ratio by the national average winsorized

scaled observed episode cost to generate a dollar figure for the cost measure score.

The clinician-l

evel (or clinician group practice-level) risk-adjusted and specialty-adjusted cost for

any attributed clinician (or clinician group practice) “j” can be represented mathematically as:

Renormalization is performed after adjustments are made to the episode’s expected cost, such as

winsorization or residual outlier exclusion. This process multiplies the adjusted values by a scalar ratio to

ensure that the resulting average is equal to the average of the original value.

Depression Measure Information Form (MIF) 19

where:

A diagram demonstrating a visual depiction of an example measure calculation can be found in

Appendix D.

A lower measure score indicates that the observed episode costs are lower than or similar to

expected costs for the care for the particular patients and episodes included in the calculation.

A higher measure sc

ore indicates that the observed episode costs are higher than expected for the

care provided for the particular patients and episodes included in the calculation.

Depression Measure Information Form (MIF) 20

Appendix A. Example Illustrations of

Scenarios for Episode Construction and

Assignment of Days

This appendix provides additional details on how an episode is constructed and attributed to a

particular measurement period, and how days are assigned to an episode.

A.1. Simple Example of Defining an Episode

In Figure A-1 below:

• Episode 1 is a portion of the total attribution window that starts on the day of the trigger

claim and concludes at the end of the subsequent measurement period (December 31).

Since episode 1 ends in measurement period 1, the associated costs will be measured

in measurement period 1.

• Episode 2 is a one-year long portion of the total attribution window that starts at the

begi

nning of measurement period 2 (January 1) and ends at the end of the

measurement period (December 31). Since episode 2 ends in measurement period 2,

the associated costs will be measured in measurement period 2.

Figure A-1. Episode Windows

A.2. Episode Construction Examples

The figures below provide examples of how episodes are constructed and attributed to a

particular measurement period. Overall, an episode’s window is defined based on:

• whether the pati

ent-clinician relationship during the measurement period was

continuous, and

• the amount of c

laims data that has not been assessed in preceding measurement

periods.

These examples also show how days are assigned to episodes. In each of these examples, we

focus

on episodes assessed in measurement period 2, which are used in Appendix D to

dem

onstrate how the measure score is calculated in a given measurement period. Assigned

days are used as a weighting factor at the measure score calculation step, where the observed

to expected ratio of each episode is weighted by the number of assigned days to that episode

and then averaged over all episodes attributed to the clinician or clinician group. Therefore, to

Depression Measure Information Form (MIF) 21

ensure fair comparison, longer episodes are given more weight during measure calculation than

shorter episodes.

Episode Window 1. 365 Days; No Reaffirming Claims During the Total Attribution Window

Figure A-2 illustrates a Depression episode that is 365 days long. This episode begins during

the fi

rst measurement period with a pair of triggering services that opens a one-year long

attribution window that extends into the second measurement period. While a reaffirming

service would have extended the relationship between the patient and the attributed clinician,

the absence of a reaffirming claim ends this clinician-patient relationship after 365 days.

Therefore, in this example, the length of the total attribution window and the episode are the

same.

• Measurement Period 1: Costs will not be assessed during measurement period 1

bec

ause there was not a year’s worth of claims data to assess during this measurement

period.

• Measurement Period 2: Costs will be assessed during measurement period 2 because

the episode ended in measurement period 2 and contained a year’s worth of claims data

that have not been previously assessed.

• Since none of the days were previously assessed, all 365 days would be

signed to episode 1 and would be used as a weighting factor at the measure

score calculation step.

Figure A-2. Episode Window (365 Days; No Reaffirming Claims)

Episode Window 2. 366 to 729 days; Reaffirming Claims During the Total Attribution

Window

Figur

e A-3 illustrates a Depression episode that is longer than 365 days.

This episode begins

during measurement period 1, contains 1 reaffirming claim 135 days into the attribution window

that extends the initial attribution window by another 365 days, and ends 500 days after the

trigger claim during measurement period 2.

• Measurement Period 1: Costs will not be assessed during measurement period 1

bec

ause of the absence of a year’s worth of claims data to assess during this

measurement period.

• Measurement Period 2: Costs will be assessed during measurement period 2 because

the epi

sode ended in measurement period 2 and contained a year’s worth of claims data

that have not been previously assessed.

Episodes can be up to 729 days long. At 730 days, the patient’s episode would be split into 2 distinct

365-day long episodes because there would be a year’s worth of claims data available in each episode.

Depression Measure Information Form (MIF) 22

• Since none of the days were previously assessed, all 500 days would be

assigned to episode 1 and would be used as a weighting factor at the measure

score calculation step.

Figure A-3. Episode Window (366 to 729 Days; Reaffirming Claims)

Episode Window 3. 365 days; Multi-Year Total Attribution Window

Figure A-4 illustrates a long total attribution window that is at least two years in length with a

epression episode that is 365 days long, where sufficient claims data was assessed in the

preceding measurement period.

The total attribution window begins with a pair of trigger services billed 35 days before

measurement period 1, and ends approximately 38 months later, when the clinician-patient

relationship ends during measurement period 3.

• Measurement P

eriod 1: Episode 1 started on the day of the trigger claim and ended at

the end of measurement period 1 (on December 31).

• Costs will be assessed during measurement period 1 because episode 1 ended

n measurement period 1 and contained at least a year’s worth of claims data

that have not been previously assessed. Since none of the days were previously

assessed, all 400 days would be assigned to episode 1.

• Measurement Period 2: Episode 2 started on January 1 of measurement period 2 and

ended on D

ecember 31 of measurement period 2.

• Costs will be assessed during measurement period 2 because the episode

ended i

n measurement period 2 and contained a year’s worth of claims data that

have not been previously assessed. Since none of the days were previously

assessed, all 365 days would be assigned to episode 2.

• Measurement Period 3: Episode 3 started on January 1 of measurement period 3 and

ended on D

ecember 31 of measurement period 3.

• Costs will be assessed during measurement period 3 because the episode

ende

d in measurement period 3 and contained a year’s worth of claims data that

have not been previously assessed. Since none of the days were previously

assessed, all 365 days would be assigned to episode 3.

Depression Measure Information Form (MIF) 23

Figure A-4. Episode Window (365 days; Multi-Year Total Attribution Window)

Episode Window 4. 365 Days; Overlapping Episodes

Figure A-5 depicts how the remaining days of long total attribution windows are assessed when

ther

e are less than 365 days of claims data that has not been previously assessed.

In this example, the total attribution window begins with a pair of trigger services billed

appr

oximately 35 days before measurement period 1 and ends 670 days (approximately 22

months) later, when the clinician-patient relationship ends during measurement period 2.

• Measurement Period 1: For epi

sode 1, costs will be assessed during measurement

period 1 because episode 1 ended in measurement period 1 and contained at least a

year’s worth of claims data that have not been previously assessed. Since none of the

days were previously assessed, all 400 days would be assigned to episode 1.

• Measurement Period 2: For episode 2, there is not a year’s worth of claims data

betw

een the end of episode 1 and the end of the total attribution window. Therefore, the

start date of episode 2 is set as 365 days prior to the end of the total attribution window,

and falls during episode 1.

• Since the costs during the days where episodes 1 and 2 overlap have already

been as

sessed during measurement period 1, only the days occurring after the

episode 1 end date will be assigned to episode 2 (270 days). These 270 days will

be used as a weighting factor at the measure score calculation step.

Figure A-5. Episode Window (365 Days; Overlapping Episodes)

Depression Measure Information Form (MIF) 24

Appendix B. Illustration of Attribution to

Individual Clinicians (TIN-NPI)

This appendix provides a detailed illustration of the attribution methodology at the TIN and TIN-

NPI levels. Once a Depression episode has been defined, it is attributed to the:

• TIN that billed the trigger services (trigger claim and confirming claim) for the total

attr

ibution window, and to the

• TIN-N

PI(s) within the attributed TIN that billed at least 30% of trigger or confirming

codes on Part B Physician/Supplier claim lines during the episode.

The measure’s attribution methodology also imposes additional checks to ensure that TINs and

TIN-

NPIs are appropriately attributed. Specifically, TIN-NPIs that meet the 30% threshold must

have billed at least one trigger or confirming code within 1 year prior to or on the episode start

date.

Figure B-1. TIN-NPI Attribution

** Only services that occurred during the episode window are used to determine whether the clinician met the 30% threshold

Figure B-1 illustrates a scenario in which 3 clinicians (A, B, and C) within an attributed clinician

group (TIN 1) have billed services during a patient’s episode window. Within the episode

window, there are a total of 10 services billed across the 3 clinicians. Each of these services is

uniquely marked depending on the clinician that billed the service.

For TIN level attribution, TIN 1 is attributed the episode because it billed the trigger services

for

the patient. For TIN-NPI level attribution, Clinician A bills 5 qualifying services (5/10, 50%),

Clinician B bills 2 services (2/10, 20%), and Clinician C bills 3 services (3/10, 30%) during the

episode window. Clinicians A and C met the 30% threshold, so they are considered for

attribution. Clinician B did not meet the 30% threshold, so it is not considered for attribution.

• Check: Cl

inician A billed at least one trigger or confirming code within 1 year prior to or on

the episode start date, so it is considered for attribution. Clinician C did not bill any such

services, so Clinician C is not considered for attribution.

Since only Clinician A met the 30% threshold and the additional check, it is attributed this

episode.

Depression Measure Information Form (MIF) 25

Appendix C. Measure Flowchart for

Depression Measure

Depression Measure Information Form (MIF) 26

Appendix D. Measure Calculation Example

This sub-section shows how the measure score is calculated. Figure D-1 below provides an

illustrated example of measure calculation, using an example measure where the clinician group

has only 4 attributed episodes for demonstration purposes.

Figure D-1. Chronic Condition Episode-Based Cost Measure Calculation Steps