December

2021 Computer-Based Training Series

Risk Adjustment
Methodology
 Are you or your
organization interested in
learning more about the foundations
and principles of risk adjustment?
Do you need to calculate
or validate a beneficiary’s
payment risk score?

If you’ve answered yes to these questions, then great

news! You are in the right place; this elearning course
was designed with you in mind.

Welcome to the Risk Adjustment Methodology course.

i-2
 The Modules Click on Module
Icon to Review
Review the modules in order by continuing to the next
page OR click on the Module 1, Module 2, or Module 3
buttons below to proceed to a specific module.

This course is broken down into three distinct modules:

Module 1
An Overview of Risk Adjustment
An Overview of Risk Adjustment—provides an intro-
duction to the foundation of risk adjustment, the legis-
lative history, and how CMS Hierarchical Condition
Category (or HCC) and Prescription Drug Hierarchical
Condition Category (or RxHCC) models are used for risk
adjustment.

Module 2
Risk Score Calculation
Risk Score Calculation—reviews the risk score calcula-
tion process and the resources you’ll need to compute
(or validate) a beneficiary’s payment risk score.

Module 3
Applying the Risk Score Formula
Applying the Risk Score Formula—uses scenario-based
learning to walk you through the four-step process of
calculating risk scores for certain Medicare Part C and
Part D beneficiaries.

i-3
Module
1 Risk Adjustment Methodology
 Table of Contents
Introduction ............................................................................................... 1-3

History of Risk Adjustment .......................................................................... 1-3

1997 – Balanced Budget Act (BBA) of 1997 ............................................................................. 1-3

2000 – Benefits Improvement and Protection Act (BIPA) of 2000 ............................................1-4
2003 – Medicare Prescription Drug, Improvement, and Modernization Act (MMA) of 2003 .....1-4
2004 – First CMS-Hierarchical Condition Category (HCC) Model Implemented .......................1-4
2005 – End-Stage Renal Disease (ESRD) Model Implemented ...............................................1-5
2006 – RxHCC Risk Adjustment Model Implemented .............................................................1-5
2010 – Patient Protection and Affordable Care Act (ACA) of 2010 ..........................................1-5
2016 – The 21st Century Cures Act of 2016 ..............................................................................1-6

Foundations of Risk Adjustment ..................................................................1-6

How Models Work .......................................................................................1-6

Risk Adjustment Models .............................................................................. 1-7

Part C (CMS-HCC NON-ESRD) ............................................................................................... 1-7

Part C (CMS-HCC ESRD) ........................................................................................................1-8
PACE .....................................................................................................................................1-8
Part D (RxHCC) .......................................................................................................................1-9

Comparing Part C & D Models ......................................................................1-9

Source of Diagnosis ................................................................................................................1-9

Predicting Expenditures .........................................................................................................1-9
Additive and Hierarchical Model ........................................................................................... 1-10
Adjusting Payment Amounts................................................................................................ 1-10
Expected Costs ..................................................................................................................... 1-10
Predicted Expenditures ........................................................................................................ 1-11

Conclusion ............................................................................................... 1-11

1-2
 Introduction History of Risk Adjustment

Risk adjustment is used to adjust payments to Medi- Here is a timeline from 1997 to 2016. Select each circle
care Advantage Organizations (MAOs), Program of All to learn more about the history of risk adjustment.
Inclusive Care for the Elderly (PACE), certain demon-
strations, and Part D sponsors for the expected health- 1997 – Balanced Budget Act (BBA) of 1997
care costs of their enrollees based on disease factors
and demographic characteristics.

This course covers the foundation and history of risk

adjustment, the risk adjustment models, and how
the CMS Hierarchical Condition Category (HCC) and
Prescription Drug Hierarchical Condition Category
(RxHCC) models are used to calculate risk scores.

Select the Next arrow to continue.

Risk Adjustment for Medicare Advantage (then, Medi-

care+Choice) was first required by the Balanced Budget
Act in 1997. The Act mandated that the risk adjustment
methodology account for variations in per capita costs
based on health status and other demographic factors
for payments.

1-3
 2000 – Benefits Improvement and 2003 – Medicare Prescription Drug,
Protection Act (BIPA) of 2000 Improvement, and Modernization Act
(MMA) of 2003

The Benefits Improvement and Protection Act of 2000

established an implementation schedule to achieve The Medicare Prescription Drug, Improvement, and
100% in risk adjustment payments by 2007, and it Modernization Act of 2003 created the Medicare Advan-
required risk adjustment for Medicare Advantage tage program to replace the Medicare+Choice program
enrollees in End-Stage Renal Disease (ESRD) status. and introduce the bidding process. It also created the
Medicare Part D prescription drug benefit to begin in
Risk adjustment was gradually phased in, starting
2006 and included risk adjusted payments.
with principal inpatient diagnoses and demographic
factors (age, sex, Medicaid status, and original reason 2004 – First CMS-Hierarchical Condition
for Medicare entitlement). PACE and dual demonstra- Category (HCC) Model Implemented
tion programs lagged one year behind the Medicare
Advantage plans on the implementation schedule and
completed their phase-in in 2008.

The first CMS-HCC model was implemented in 2004.

This model produces different risk scores for beneficia-
ries who reside in either the community or an institu-
tional setting, or who are new enrollees.

1-4
 2005 – End-Stage Renal Disease (ESRD) 2010 – Patient Protection and Affordable
Model Implemented Care Act (ACA) of 2010

CMS implemented a separate risk adjustment model The Patient Protection and Affordable Care Act of 2010
for beneficiaries with End-Stage Renal Disease in 2005. refined the Medicare Advantage risk adjustment meth-
This model produces different risk scores for beneficia- odology for new enrollee risk scores for Chronic Condi-
ries in dialysis status, beneficiaries who have a kidney tion Special Needs Plans (SNPs). This methodology
transplant, and beneficiaries who have a functioning takes into account that enrollees in these types of SNPs
graft. must have certain conditions as a prerequisite to being
enrolled.
2006 – RxHCC Risk Adjustment Model
Implemented

In 2006, the RxHCC risk adjustment model was imple-

mented. Similar to the CMS-HCC risk adjustment
model, the RxHCC model uses demographic character-
istics and disease variables to predict costs. The RxHCC
model predicts Medicare Part D plan liability costs for
prescription drugs under the Part D program.

1-5
 2016 – The 21st Century Cures Act of 2016 The CMS-HCC based Medicare risk adjustment models
are prospective – diagnoses in one year are used to
predict costs in the following year.

CMS has multiple models to address differences in the

beneficiary population (for example, the ESRD popu-
lation) and program costs (for example, Part C versus
Part D).

How Models Work

The 21st Century Cures Act of 2016 amended the Social

Security Act by, in part, requiring CMS to make changes
to the CMS-HCC risk adjustment model for 2019 and
subsequent years. The changes include the evalua-
tion of additional diagnosis codes for mental health,
substance use disorder, and chronic kidney disease into
the model. These changes take into account a Medi-
care Advantage enrollee’s total number of conditions
by making an additional adjustment as the number of
diseases or conditions of an individual increases. These Risk adjustment uses statistical models that measure
changes will be phased in by 2022. incremental predictive costs of a beneficiary’s demo-
graphic and disease characteristics, and age, sex, and
Foundations of Risk certain statuses such as Medicaid eligibility. The total
predictive costs are determined by adding the coef-
Adjustment ficients of a beneficiary’s demographics and health
status factors.

Diagnoses are grouped into condition categories with

similar Medicare predicted costs and those that are
clinically related. Hierarchies are applied to these
condition categories based on disease severity so that
risk scores reflect the most severe and costly category
of a condition.

Risk adjustment is used to adjust plan bids as well as

payments to plans based on their enrollee’s expected
health care costs.

1-6
 Risk Adjustment Models Part C (CMS-HCC NON-ESRD)

CMS uses the Part C CMS-HCC model to calculate risk

CMS has developed separate risk adjustment models
scores for beneficiaries who are not in dialysis, trans-
for the Parts A and B benefits offered by plans under
plant, or post-graft status. There are multiple segments
Part C and for the Part D benefits offered by Medicare
associated with this model including the communi-
Advantage Part D and prescription drug plans. Within
ty, institutional, and new enrollee segments. For the
each benefit, CMS also segments each model, creating
Part C models implemented in 2017 and beyond, there
subpopulations with distinct cost patterns.
are six segments for community-residing beneficia-
Take some time to explore the models applicable to ries – one each depending on a beneficiary’s aged and
you and your work. disabled status and their dual eligible status – whether
they are full dual, partial dual, or non-dual. The institu-
When you’re done exploring, select the Next button to
tional segment is used to calculate risk scores for indi-
learn about the similarities and differences between
viduals who are in long-term care institutions, including
the Part C and Part D models.
Skilled Nursing Facilities (SNFs), for 90 days or more.
New enrollee model segments are used for aged and
disabled beneficiaries who are new to Medicare or do
not otherwise have enough diagnoses to calculate a
risk score – operationalized as those who do not have
12 months of Part B in the data collection period. These
new enrollee risk scores do not use diagnoses to predict
costs and use demographic factors only. There is also a
segment for aged and disabled new enrollees in Chron-
ic Special Needs Plans, known as Chronic SNPs.

Select the Back button to continue reviewing the

models.

1-7
 Part C (CMS-HCC ESRD) PACE

The ESRD model has similar characteristics as the CMS pays PACE organizations similar to MAOs using
CMS-HCC model, but it differs in that it predicts medical risk adjustment that reflects beneficiary health status.
(Parts A and B) costs for the population in ESRD status. We typically use the same models to calculate risk
The ESRD model is a suite of models that produce scores for beneficiaries enrolled in PACE organizations
risk scores for beneficiaries in Dialysis, Transplant, as we use to pay Medicare Advantage and, starting in
and Functioning Graft statuses. The Dialysis model 2006, for Part D costs.
is segmented by continuing enrollee – those with 12
CMS utilized the 2017 CMS-HCC model for PACE
months of Part B in the data collection period – and
payment starting in payment year 2020.
new enrollee. The Transplant model covers the month
of transplant and the next two months. The Function- Similar to the 2020 CMS-HCC model used in the risk
ing Graft model includes segments for beneficiaries score calculations of non-PACE Part C plans, there are
with a functioning kidney graft based on the number of six segments for community-residing beneficiaries.
months post-transplant, whether it be 4 to 9 months or Each segment depends on a beneficiary’s aged and
10 or more months. disabled status and their dual eligible status – whether
they are full dual, partial dual, or non-dual.
Select the Back button to continue reviewing the
models. The institutional segment is used to calculate risk scores
for individuals who are in long-term care institutions,
including Skilled Nursing Facilities (SNFs), for 90 days
or more. New enrollee model segments are used for
aged and disabled beneficiaries who are new to Medi-
care and do not otherwise have enough diagnoses to
calculate a risk score – operationalized as those who do
not have 12 months of Part B in the data collection peri-
od. These new enrollee risk scores do not use diagnoses
to predict costs and use demographic factors only.

Select the Back button to continue reviewing the

models.

1-8
 Part D (RxHCC) Source of Diagnosis

CMS uses a separate risk adjustment model to predict Each beneficiary is assigned their own risk score for
plan liability for prescription drugs under the Part D Part C and for Part D. Since these models predict differ-
program. The Part D risk adjustment model (RxHCC ent costs, a beneficiary’s Part C and Part D risk scores
model) shares many of the characteristics of the may differ.
CMS-HCC model. Note: we will discuss similarities
The International Classification of Diseases Codes are
and differences in the next section. There are sepa-
used for both the calibration of the models and to
rate segments based on aged versus disabled status,
calculate risk scores for payment. CMS uses diagnoses
and low income versus non-low-income status, and
from both MAOs and from Fee-For-Service providers.
continuing versus new enrollee.

Select the Back button to continue reviewing the Predicting Expenditures

models.

Comparing Part C & D Models

Both models predict benefit costs for which the plans

are responsible for covering. The CMS-HCC model
predicts medical expenditures covered under Parts A
and B Medicare. The RxHCC model predicts the drug
Select each tab to learn more about the Part C and D costs for which the Part D plan is liable – in other words,
similarities and differences. it does not include cost sharing amounts for which the
enrollee or costs for which the government is responsi-
ble for paying.
1-9
 Additive and Hierarchical Model Expected Costs

The two models generate enrollee risk scores by adding Each beneficiary’s risk score is calculated to estimate
relative risk factors for each HCC or RxHCC associated that specific beneficiary’s expected costs relative to
with each beneficiary. Both models use hierarchies. the average beneficiary. For each model, a risk score
Hierarchies use the most severe manifestation of a of 1.0 reflects the Medicare-incurred expenditures of
condition to be considered for risk scores. an average beneficiary. An RxHCC risk score of 1.0 indi-
cates the beneficiary is expected to incur the average
Adjusting Payment Amounts liability amount for prescription drugs when covered
by the standard Part D Medicare benefit. A CMS-HCC
risk score of 1.0 indicates the beneficiary is expected to
incur the average Medicare program expenditure for
Parts A and B services.

Risk adjustment is intended to adjust capitated

payment amounts to pay plans accurately for groups of
beneficiaries, thereby increasing incentives for health
plans to enroll all beneficiaries, regardless of health
status. Both models adjust payments for the expected
costs, based on underlying health status, of the benefi-
ciaries enrolled in the plan.

The CMS-HCC model adjusts Part C monthly payments

to Medicare Advantage plans, PACE organizations, and
certain demonstrations. The RxHCC model adjusts the
monthly Part D direct subsidy.
1-10
 Predicted Expenditures Conclusion

While both the CMS-HCC and the RxHCC models use

The risk score formula is equal to the sum of the demo-
health status or diagnoses and demographic charac-
graphic factors and the disease factors. The sum of
teristics to predict expenditures, the total expendi-
those factors equals the raw risk score. CMS then
tures that each model is predicting are quite different
applies several methodological adjustments to the raw
(medical versus prescription drug) and, therefore, may
risk score. These adjustments are the normalization
result in different weights on similar HCCs and RxHCCs,
factor, used to keep the average risk score at 1.0, and
as well as different Part C and Part D risk scores for a
the Medicare Advantage coding pattern adjustment,
beneficiary.
to account for differential coding patterns between
Risk adjustment attempts to account for the differences MAOs and Fee-For-Service providers. The result is the
in expenditures incurred by a plan due to differences Payment Risk Score.
in the health status of the beneficiaries enrolled in the
We will discuss these adjustments more in Module 2,
plan. Since the impact of health status factors and the
where we will introduce how risk scores are calculated,
benefit design are different between Parts C and D, the
including the steps in the process to access the infor-
two risk adjustment models have been designed to
mation that you will need to calculate the Risk Score.
predict the relevant costs.
A future module will provide detailed calculations for
each of the risk score models.

If you have questions, you can submit them to

[email protected].

1-11
Module
2 Risk Adjustment Methodology

Risk Score
Calculation
 Table of Contents
Introduction ..............................................................................................2- 3

Resources.................................................................................................. 2-3

Plan Communication User Guide (PCUG) .............................................................................. 2-3

Annual Rate Announcements ...............................................................................................2-4
Risk Adjustment Model Software ..........................................................................................2-4
Monthly Membership Detail Report (MMR) ..........................................................................2-4
Model Output Report (MOR) ................................................................................................2-4

Risk Score Calculation Checklist ..................................................................2-5

Step 1 ....................................................................................................................................2-5
Step 2 ...................................................................................................................................2-5
Step 3 ...................................................................................................................................2-5
Step 4....................................................................................................................................2-6

Step 1: Access and Download Reports ..........................................................2-6

Step 2: Identify Demographic and Disease Factors .......................................2-7

Demographic and Other Payment Information ..................................................................... 2-7

Disease Information ............................................................................................................ 2-14
What’s Next ........................................................................................................................ 2-19

Step 3: Identify Relative Factors and Adjustment Factors ............................. 2-19

Step 4: Calculate the Risk Score ................................................................. 2-22

Step A: Raw Risk Score ........................................................................................................ 2-22

Step B: Normalized Risk Score ............................................................................................ 2-22
Step C: MA Coding Pattern Adjusted Risk Score .................................................................. 2-22
Blended Final Risk Score ..................................................................................................... 2-23

Conclusion ............................................................................................... 2-23

2-2
 Introduction Resources

The risk adjustment models are used to calculate risk There are several resources needed to calculate and
scores, which predict an individual beneficiary’s health validate the risk scores.
care expenditures or plan liability for Part D drugs rela-
This section links you to this information and will be
tive to the average beneficiary.
available as a resource throughout the training.
While the Centers for Medicare and Medicaid Services,
or CMS, uses multiple risk adjustment models in Plan Communication User Guide (PCUG)
payment to address differences in the beneficiary
population, the same general process is used to calcu-
late risk scores.

This module will introduce how to calculate and vali-

date risk scores, including the steps in the process, as
well as the beneficiary information used to calculate
risk scores.

The Plan Communication User Guide for Medicare

Advantage Prescription Drug Plans, also referred
to as the PCUG, is on the Research, Statistics, Data
& Systems, MAPD Help Desk page of the cms.gov
website.

2-3
 Annual Rate Announcements Monthly Membership Detail Report (MMR)

Annual Rate Announcements for the applicable The Monthly Membership Detail Report, referred to
payment years are on the Medicare Advantage Rates as the MMR, is the basic accounting file of beneficiary
& Statistics, Announcements and Documents page of level payments and adjustments for Medicare Advan-
the cms.gov website. tage Organizations and Prescription Drug Plans. There
are several fields used in the risk score calculations
Risk Adjustment Model Software which we will describe in this training.

For field layout information, please see the PCUG.

Model Output Report (MOR)

The Medicare Advantage Rates & Statistics, Risk

Adjustment page of the cms.gov website has addi-
tional risk adjustment information including the Risk
Adjustment model software (HCC, RxHCC, ESRD) and
The Model Output Report (MOR) is a monthly report
the diagnosis to HCC (or RxHCC) mappings.
that lists the diseases applicable to each beneficiary.

For field layout information, please see the PCUG.

2-4
 Risk Score There are separate MORs for the Part C, ESRD, and
Part D models.
Calculation Checklist
The MMR and MOR layout field descriptions are avail-
able in the PCUG.

Step 2

Risk Scores are validated in essentially four steps.

Select each circle to learn more.

Step 1
You will use the MMR and MOR to identify demographic
and disease information used for risk score calculation.

You will also look for beneficiary demographic and

payment indicators that affect risk score calculation
in the MMR and the Hierarchical Condition Categories
(HCCs) for Part C and RxHCCs for Part D in the respec-
tive MOR.

Step 3

The Monthly Membership Detail Report, or MMR, and

the Model Output Report, or MOR, include the data
used to calculate the risk score for a beneficiary in your
contract.

CMS sends the reports to Medicare Advantage Orga-

nizations and other submitters every month, but if
you do not have them available and are an active
contract, you may download them from the Medicare
Advantage Prescription Drug User Interface, also
After you have identified the demographic factors and
known as the MARx UI.
the HCCs or RxHCCs, you will use the Rate Announce-
The MMR includes Part C, ESRD, and Part D content. ment for the applicable payment year, or the risk

2-5
 adjustment model software that is posted on the CMS Step 1: Access and
risk adjustment website, to identify the relative factors
for each demographic factor and HCC or RxHCC. Download Reports
You will also find adjustment factors used in the risk
score formula in the Rate Announcement. These include
the normalization factors and the Medicare Advantage
(MA) coding pattern adjustment factor.

Step 4

Medicare Advantage Organizations or third-party

submitters collect and submit to CMS the diagnostic
information and information for items and services
provided to beneficiaries.

Diagnoses and additional variables are submitted

into the Front-End Risk Adjustment System which
Once you have identified all relative factors and adjust- forwards the data to the Risk Adjustment Processing
ment factors, you can calculate a beneficiary’s risk System, RAPS.
score.
Encounter data goes through a similar but separate
process. Encounter Data is submitted to the Encoun-
ter Data Front-End System which forwards the data to
the Encounter Data Processing System, also known as
the back-end system, where it resides in the Encounter
Operational Datastore.

Diagnoses from RAPS, encounter data, and fee-for-

service claims, and demographic information from the
Common Medicare Environment are utilized for risk
score calculation. In addition, information from nursing
home records housed in the Minimum Data Set is used
for long term institutional status.
Select each Learn More button to continue exploring The risk scores are then sent to MARx for payment
each step. At any time, you can select the Main Menu calculations.
button to return to this page.

2-6
 Demographic and Other
Payment Information

Finally, CMS sends the MMRs and MORs to the plans.

The disease and demographic information CMS uses
to calculate risk scores is provided in the MMRs and There are several demographic indicators on the MMRs
MORs. Plans may therefore use the data in the MMR that identify the demographic factors used when calcu-
and MOR to validate risk scores. lating risk scores.

The demographic information includes beneficiary

Step 2: Identify Demographic facts, such as age and sex, and includes certain benefi-
ciary status indicators that enable grouping into similar
and Disease Factors segments, such as those living in the community versus
an institution.

Select each factor to learn more.

At any time, you may click the MMR Fields button to

review the fields, indicators, and descriptions which
are also outlined in the PCUG. When you are finished
exploring, select the Next button to return to the Step
2 menu.

Calculating risk scores requires specific demographic

information, other payment information, and disease
information.

Select each button to learn more.

2-7
 Demographic Information Long Term Institutional (LTI) Status

Age and Sex

In most cases, the risk adjustment models have sepa-

rate segments to calculate risk scores for community
The models include variables for age and sex groupings.
and long-term institutional residents. Long-term insti-
Please note, the terms gender and sex are used inter-
tutionalized (LTI) MA enrollees are individuals residing
changeably. The age and sex variables in the model
in an institution for at least 90 days as identified using
have associated relative factors published in the Rate
90-day assessments in the Minimum Data Set (MDS).
Announcements. Age, for risk adjustment purposes, is
LTI beneficiaries have risk scores calculated using the
the beneficiary’s age as of February 1st of the payment
LTI segment of the model based on their monthly
year.
status during the payment year. Short term institution-
Original Reason for Entitlement (OREC) Code alized MA beneficiaries have risk scores (i.e., less than
90 days) calculated based on the community segment
of the model.

The Original Reason for Entitlement can be due to age,

disability, ESRD, or a combination of disability with
ESRD. Because CMS calculates initial and mid-year risk scores
before it has complete data on beneficiaries’ LTI status
There is an additional relative factor applied to the risk
in the payment year, they use the presence of a 90-day
scores of beneficiaries who were originally entitled to
assessment reported for any one month during the
Medicare due to disability.
12-month data collection period as a proxy for LTI
status in the payment year. At the final payment recon-

2-8
 ciliation that takes place post-contract year, CMS uses Other Payment Information
each beneficiary’s actual month-by-month LTI status
in the payment year to determine which risk score to
apply.

Medicaid/Dual Status

Select each factor to review other payment information.

End-Stage Renal Disease (ESRD) Status Codes

For the Part C model, Medicaid status determines

which community segment risk score is used in
payment. The MMR includes the Medicaid status infor-
mation that was used to determine the appropriate
community risk score for a payment month. You can
use this field to determine if the beneficiary received
a full dual or partial dual risk score for the month. For
non-community beneficiaries, such as those with an
LTI status and new enrollees, this field is informational
only.

The Medicaid dual status code that is in effect for the End-Stage Renal Disease (ESRD) is a medical condi-
month is used to determine the appropriate commu- tion in which a person’s kidneys cease functioning on
nity risk score for a non-ESRD, full-risk, non-PACE a permanent basis, leading to the need for a regular
beneficiary. CMS identifies full-risk beneficiaries as course of long-term dialysis or a kidney transplant to
those who are entitled to Part A and have at least 12 maintain life. Beneficiaries may become entitled to
months of enrollment in Part B. Medicare based on ESRD.

2-9
 contract or plan-level frailty of those in the community.

There is a separate risk adjustment model for benefi-

ciaries who are in ESRD status. Frailty adjustments are only added to the risk scores
for PACE participants and community-based beneficia-
The Medicaid Indicator, field 21, is used for the
ries enrolled in qualifying Fully Integrated Dual Eligible
CMS-HCC ESRD risk adjustment calculation model. It
Special Needs Plans, known as FIDE SNPs. For FIDE
will indicate whether the ESRD beneficiary is eligible
SNPs to qualify for a frailty adjustment, they must have
for Medicaid. For ESRD beneficiaries, this field has the
a similar level of frailty as PACE organizations. There is
values that will impact the beneficiary risk score.
no frailty adjustment applied to the risk scores of long-
Any value in the Medicaid Dual Status Code, field 84, term institutionalized beneficiaries.
for ESRD beneficiaries is only informational.

Frailty Score

Field 47 on the MMR, Frailty Indicator, indicates if a

frailty score is used in payment.

The purpose of the frailty adjustment is to account Field 80 on the MMR, The Part C Frailty Score Factor,
for Medicare expenditures of community populations provides the frailty factor that was used in payment.
with functional impairments that are unexplained by
risk adjustment. The frailty score is derived from the
number of functional limitations represented by the
Activities of Daily Living (ADL) scale to calibrate the
frailty model and then used to determine the relative

2-10
 Part C Risk Adjustment Factor (RAFT) Codes then the RAFT code CF will appear in the MMR and indi-
cate that the risk score was calculated using factors for
Community Full Dual.

Part D Risk Adjustment Factor (RAFT) Codes

Risk Adjustment Factor Type (RAFT) codes indicate

which status was used to select the type of risk score
used in calculating the monthly payment for an indi-
vidual beneficiary. RAFT codes with ESRD or PACE
in the factor description field indicate those models The Part D RAFT code also indicates which Part D
were used to calculate the beneficiary’s risk score. For risk adjustment model segment was used to calcu-
all other codes, the CMS-HCC (non-ESRD) model was late a beneficiary’s risk score. The segment of the Part
used to calculate a beneficiary’s risk score. D model that is used to calculate risk scores is based
on a combination of the following factors: whether
The RAFT codes will also indicate which segment of or not the beneficiary is continuing or new enrollee,
each model was used to calculate the risk score. The low income, ESRD or non-ESRD, and residing in the
segment of the CMS-HCC (non-ESRD) model that is community or an institution.
used to calculate risk scores is based on a combination
of the following factors: whether or not the beneficiary
is continuing or new enrollee, dual status, and residing
in the community or an institution.

For example, if the beneficiary is a new enrollee, low

income, non-ESRD, and residing in the community,
then the RAFT code D6 will appear in the MMR and
indicate that the risk score was calculated using factors
For example, if the beneficiary resides in the community that meet these conditions.
and is eligible for both Part C Medicare and Medicaid,

2-11
 Default Risk Factor Code Select each field to review the indicators.

When you’re done, select the Back button to return to

the demographic factors menu.

Field 7

The RAFT codes indicate which status was used

to select the type of risk score used in calculating
the monthly payment for an individual beneficiary.
However, beneficiaries are given a Default Risk Factor
Code when they do not have a risk score produced Field 15
from RAS, the risk adjustment system.

For example, when a newly entitled beneficiary is in

the CMS systems at the time CMS conducted the risk
score run, they will receive a new enrollee risk score
that was assigned by RAS. When a beneficiary is new
to Medicare and was not in the CMS systems at the
time a risk score run was conducted, a default risk score
may be used.

MMR Data Fields and Indicators

Field 20

The demographic components applicable to the risk

adjustment models are listed.

2-12
 Field 21 Field 40

Field 23 Field 46

Field 39 Field 47

2-13
 Field 48 Disease Information

Field 84 The disease portion of the risk score is based on HCCs

or RxHCCs and any disease and disabled interactions.

There are also post graft add-on factors applicable

to beneficiaries with ESRD who have received a trans-
plant and are in post graft status.

Select each button to learn more.

At any time, you may click the MOR button to review

the MOR report fields, and descriptions which are also
outlined in the PCUG.

When you are finished exploring, select the Next button

to return to the Step 2 menu or select Home to return
to the main menu.
Field 86
HCCs, RxHCCs, and Hierarchies

The Part C and ESRD MORs provide the HCCs that

CMS uses to calculate risk scores for each beneficiary.

2-14
 There is a separate Part D MOR that provides the RxHCC that was incorporated superseded the one that
RxHCCs that CMS uses to calculate Part D risk scores. was not included.
The MOR already accounts for hierarchies, and only
indicates HCCs or RxHCCs that are applicable to the
risk score used in payment.

Alternatively, you may use your internal database of

accepted diagnoses for a beneficiary, map the diagno-
ses to the payment HCCs and RxHCCs in the appropri-
Disease hierarchies address situations when multiple ate model, and then use this information to calculate
levels of severity for a disease, with varying levels of the beneficiary’s risk score. When using this method,
associated costs, have been reported for a benefi- be sure to verify if a hierarchy exists and then include
ciary. The hierarchies prioritize the inclusion in a risk the relative factor only for the most severe manifesta-
score of multiple HCCs or RxHCCs where diagnoses are tion of the condition in the hierarchy in the risk score
clinically related and ranked by costs. calculation. Remember, hierarchies are published in
the Rate Announcement, along with the relative
factors, when a model is updated or recalibrated.

Disease and Disabled Interactions

If you are calculating risk scores and wondering why a

specific HCC or RxHCC was not incorporated into the
calculation, check the disease hierarchy list in the rele-
vant Rate Announcement to determine whether that
The models include relative factors for interactions
HCC or RxHCC is in a hierarchy. If so, check where the
with certain conditions when the presence of another
missing HCC or RxHCC is in relation to the one incorpo-
disease or demographic status, such as disabled status,
rated in the payment risk score. Most likely, the HCC or
is indicative of higher costs.

2-15
 Post Graft Add-On and Transplant
Factors (ESRD only)

Disease Interactions – Certain combinations of coex-

isting diagnoses for an individual can increase medical
costs more than the additive nature of the CMS-HCC If the beneficiary’s risk score is calculated using the
model predicts. ESRD functioning graft model, then they receive an
additional factor based on a combination of his or her
aged and disabled status and the duration since the
transplant was performed. This factor is added to the
other relative factors.

Disabled Interactions – Interactions between certain

diseases and disabled status for an enrollee can also
increase medical costs more than the additive nature
of the CMS-HCC model predicts.
You may determine which functioning graft factor
For risk score calculation, the disease and disabled applies to the beneficiary using the RAFT code. RAFT
interaction relative factors are added to the demo- codes with Post Graft 4-9 in the field name description
graphic and disease relative factors. signify that it has been between 4–9 months since the
transplant, and functioning graft RAFT codes ending in
2 (C2 and I2) signify that it has been 10 or more months
since the transplant.

2-16
 Model Output Report (MOR)

For Part C and ESRD, plans can determine the HCCs

and interactions used for risk score calculation using
the Part C and ESRD MORs, respectively.

For Part D, plans can determine the RxHCCs and inter-

actions used for risk score calculation using the Part D
MOR.

CMS populates the MOR with payment HCCs or

RxHCCs, and it does not include any HCCs or RxHCCs
that are dropped due to the application of the hierar-
chies. So if you are calculating the risk score based on
information from the MOR, it is not necessary to assess
If the beneficiary risk score is calculated using the trans-
whether to drop HCCs based on hierarchies.
plant segment of the ESRD model, the RAFT code will
include Graft I or Graft II in the field name description. Select each button to explore each report.
Transplant payments are calculated by multiplying the
When you are done, select the Back button to return to
transplant factor by the ESRD dialysis state rate.
the disease factors menu.

2-17
 Part C/ESRD MOR RxHCC MOR

Select each button to highlight key information. Select each button to highlight key information.

2-18
 What’s Next Step 3: Identify Relative
Factors and
Adjustment Factors

The demographic factors, including beneficiary status

indicators, and disease factors provide all the infor-
mation necessary to identify the model and model
segment that should be used to calculate the risk score.
You just learned where to find information in the MMR
Now you know how to identify pertinent risk score and MOR. Now you are going to learn how to use the
information on the MMR, including determining the Rate Announcement to gather relative factor values.
model and segment that applies to your beneficiary.

In Step 3, you will be taking these demographic

and disease factors and using the applicable Rate
Announcement to look up the specific relative factor
values.

You start by determining which payment year applies.

The MMR discussed in Step 2 also specifies the appro-
priate payment month and year in field 3, Payment
Date. You can find the Rate Announcement applica-
ble to that payment year on the Medicare Advantage
Rates and Statistics page on the cms.gov website and
as mentioned in the Resources section of this training.

Next, you need to determine what model to use.

2-19
 In the applicable year’s Rate Announcement, there will Once you have identified the appropriate risk adjust-
be a section that highlights the key changes for each ment model relative factor table, you will look up the
model. Read this section carefully and refer to it as you relative factor values that correspond to a beneficia-
proceed. ry’s model segment and demographic indicators. For
example, you need to use your beneficiary’s age and sex
from the left column, and model segment from the top
row (as shown in the table columns which indicate the
beneficiary is either institutionalized or in a community
and whether Medicaid eligible, and aged or disabled) to
find the demographic relative factor values.

You already used the MMR to determine all the demo-

graphic factors that apply to a particular beneficiary,
which will indicate the model and segment that applies.

Now you will use those indicators and refer to the

appropriate model and segment relative factor table in
the Rate Announcement.
For example, a 77-year-old female beneficiary living in
Note that relative factor tables that pertain to a model the community who is not Medicaid eligible and was
are published in the Rate Announcement for the year in originally entitled to Medicare due to her age would
which the model was first implemented. have a value of 0.451 for the age-sex relative factor.

If there are Medicaid or Originally Disabled Interac-

tions, you will need those relative factor values as well.

You will use the demographic relative factor values that

2-20
 you found here in the next step when you calculate the If there are any disease or disabled and disease inter-
Raw Risk Score. actions, you will retrieve those relative factor values as
well.

Next, you will use the information you gathered from

the MOR to determine the disease portion of a bene- So far, you have used the MOR, MMR, and Rate
ficiary’s risk score. You will reference the same relative Announcement to gather all the demographic and
factor table and model segment used for the demo- disease relative factors. Finally, you will need to gather
graphic factors to look up the beneficiary’s HCCs or the adjustment factors, including the normalization
RxHCCs. factor and, if applicable, the MA coding pattern adjust-
ment factor, which are in the Rate Announcement for
For example, if a beneficiary has chronic hepatitis listed
the relevant payment year.
in the MOR, then you will retrieve the relative factor
value for that HCC.

Now you know how to translate the pertinent risk score

information on the MMR and MOR to the specific rela-
Let’s return to our previous example of the 77-year-old
tive factor values. You have also gathered the adjust-
female living in the community, who is not Medicaid
ment factors that will be applied.
eligible and was originally entitled to Medicare due
to age, and who is diagnosed with Chronic Hepatitis, In Step 4, you will be using these relative factor values
HCC29. The relative factor value for this diagnosis and adjustment factors to calculate the beneficiary’s
would be 0.147 in the same relative factor table that risk score.
you referenced for the age-sex relative factor.

2-21
 Step 4: Calculate the Step B: Normalized Risk Score
Risk Score

Next, a normalization factor is applied to keep the

average risk score for all beneficiaries at 1.0 in the
To calculate the Payment Risk Score, follow steps A, B, payment year.
and C.
Divide the Raw Risk Score by the applicable model’s
Select each to learn more. normalization factor, which is in the appropriate
payment year’s Rate Announcement.
Step A: Raw Risk Score
Round the number calculated to the third decimal.

Step C: MA Coding Pattern

Adjusted Risk Score

To calculate the Payment Risk Score, you must first

calculate the Raw Risk Score. This requires adding all
the demographic relative factor values to the disease
relative factor values you looked up in Step 3. This will
Next, you will use the Rate Announcement to deter-
give you your Raw Risk Score.
mine the MA coding pattern adjustment factor for
the appropriate payment year. You will then multiply
the Rounded Normalized Risk Score by 1 minus the
MA coding pattern adjustment factor. The result is the
Normalized, MA Coding Pattern Adjusted Risk Score.

2-22
 Conclusion

Finally, round the Normalized, MA Coding Pattern

Adjusted Risk Score to the third decimal point to deter-
mine the Final Payment Risk Score. Now that you are familiar with the 4 steps necessary to
calculate a risk score, future Risk Adjustment modules
Blended Final Risk Score will further explore the risk score calculations for the
Part C and Part D models for a variety of payment years.

The scenarios you will go through describe a bene-

ficiary’s specific demographic, other payment, and
disease information. Next, you will follow a step-by-
step demonstration of the risk score calculation using
the appropriate Rate Announcement information for
that model, model segment, and payment year for the
example beneficiary.

If you have questions or suggestions regarding the

series of Risk Adjustment computer-based trainings,
you can submit them to RiskAdjustmentPolicy@cms.
For payment years where a blended final risk score is hhs.gov.
being calculated, the blending percentages will be
found in the applicable Rate Announcement in the
section that discusses each risk adjustment model.

Beginning in 2022, the Medicare Advantage non-PACE

risk scores will be calculated using diagnoses solely
from encounter data.

2-23
Module
3 Risk Adjustment Methodology

Applying the
Risk Score Formula
 Table of Contents
Introduction................................................................................................ 3-4

Scenario Selection....................................................................................... 3-4

Click on Scenario Icon to Review................................................................... 3-5

Scenario 1: Part C (NON-ESRD).....................................................................3-6

Step 1: Gather Reports & Resources........................................................................................3-6

Step 2: Identify Demographics & Disease Information.............................................................3-8
Note #1: Scenario 1.................................................................................................................3-9
Step 3: Identify Relative Factors & Adjustment Factors............................................................3-9
Note #2: Scenario 1............................................................................................................... 3-13
Step 4: Using the Risk Score Formula.................................................................................... 3-13
Scenario 1 Conclusion............................................................................................................ 3-14

Scenario 2: Part C (ESRD)............................................................................ 3-15

Step 1: Gather Reports & Resources...................................................................................... 3-15

Step 2: Identify Demographics & Disease Information........................................................... 3-17
Note #1: Scenario 2............................................................................................................... 3-18
Step 3: Identify Relative Factors & Adjustment Factors.......................................................... 3-18
Note #2: Scenario 2.............................................................................................................. 3-22
Step 4: Using the Risk Score Formula.................................................................................... 3-22
Scenario 2 Conclusion...........................................................................................................3-24

Scenario 3: Part C (PACE)............................................................................ 3-25

Step 1: Gather Reports & Resources...................................................................................... 3-25

Step 2: Identify Demographics & Disease Information........................................................... 3-27
Note #1: Scenario 3...............................................................................................................3-28
Step 3: Identify Relative Factors & Adjustment Factors..........................................................3-28
Note #2: Scenario 3............................................................................................................... 3-33
Step 4: Using the Risk Score Formula.................................................................................... 3-33
Scenario 3 Conclusion............................................................................................................ 3-35
3-2
 Scenario 4: Part D (RxHCC)......................................................................... 3-36

Step 1: Gather Reports & Resources...................................................................................... 3-36

Step 2: Identify Demographics & Disease Information........................................................... 3-38
Note #1: Scenario 4............................................................................................................... 3-39
Step 3: Identify Relative Factors & Adjustment Factors.......................................................... 3-39
Note #2: Scenario 4..............................................................................................................3-42
Step 4: Using the Risk Score Formula.................................................................................... 3-43
Scenario 4 Conclusion...........................................................................................................3-44

3-3
 Introduction Scenario Selection

Welcome to Module 3 of the Risk Adjustment Method- It’s time to pick your scenario. Select the name of each
ology series. beneficiary to learn more.

In this module, we will walk you through the four-step

process of calculating a beneficiary’s risk score for
Payment Year 2022.

You’ll be presented with four scenarios, each scenario

representing one of the four models covered in Module
2: Risk Score Calculation.

For each scenario, we will show you how to identify

which risk adjustment payment model to use, how to
find your beneficiary’s relative factors and adjustment
factors, and how to calculate a beneficiary’s risk score
using the risk adjustment formula.

Please note that if you have little to no experience

calculating risk scores, we recommend walking through
all four scenarios.

Select the NEXT button to continue.

3-4
 Click on Scenario Icon to Review
Review the scenarios in order by continuing to the next At the end of each scenario, select the Choose Scenario
page OR click on the Start Scenario button below each button to return to this page.
example beneficiary to proceed to a specific scenario.

3-5
 Scenario 1: Step 1: Gather Reports & Resources
Part C (NON-ESRD)

To accurately calculate Mr. Carter’s risk score, the

first step in the process is to gather all the reports and
Meet Mr. Carter, who is age 81 and lives at home. resources you will need which you can now access by
selecting the RESOURCES button at the bottom of
This scenario demonstrates how to calculate a risk
your screen.
score for a Part C beneficiary who is not in End-Stage
Renal Disease (or ESRD) status. These reports and resources include the Monthly
Membership Detail Report (or MMR), the Model Output
To begin the process of calculating Mr. Carter’s risk
Report (or MOR), the Plan Communication User Guide
score, select the START SCENARIO button. To choose
(or PCUG), and the Annual Rate Announcement.
another scenario, select the BACK button at the top
right of your screen. Please select the NEXT button to continue.

Monthly Membership Detail Report

Ready to walk through the process of calculating Mr.

Carter’s risk score?
The Monthly Membership Detail Report provides Plans
Select the STEP 1 button to begin. with beneficiary demographic and status information that
was used for risk score calculation for a specific payment
month. You will need the data found in this report to
complete Step 3 of the risk score calculation process.

3-6
 Model Output Report Annual Rate Announcement

Recall from Module 2 that the Model Output Report is To ensure that you are entering correct data into the
a monthly report that lists the diseases applicable to risk score calculation formula, first you must verify
each beneficiary. You will need the data from this report which payment year your risk score calculation should
to complete Step 3 in the risk score calculation process. be based on. For this scenario, Mr. Carter’s risk score
calculations will be based on Payment Year 2022.
Plan Communication User Guide
You will therefore need to refer to the 2022 Rate
Announcement to determine which risk adjustment
payment model to use.

Select the SHOW ME button to learn how.

The PCUG provides layout and field description infor-

mation for the MMR and MOR.

To verify which risk adjustment payment model you’ll

need to use for 2022 risk score calculation, open up the
2022 Rate Announcement and scroll to the section that
describes the CMS-HCC Risk Adjustment Model. The
information in this section states that CMS completed
the phasing in of the model implemented in 2020. There-
fore, 100 percent of the risk score will be calculated using
the 2020 CMS-HCC model for non-PACE organizations.
3-7
 Since we will be using the 2020 CMS-HCC model to Demographic Factors
calculate Mr. Carter’s risk score, we will also need to
refer to the 2020 Rate Announcement to look up his
demographic and disease relative factors. This infor-
mation is required to complete Step 3 of the risk score
calculation process.

Select the NEXT button to continue.

Step 2: Identify Demographics

& Disease Information

To gather Mr. Carter’s demographic information, you

will need to reference the Monthly Membership Detail
Report.

The highlighted data shown on this MMR table are

required demographic and model indicators that you
will need to compute the sum of Mr. Carter’s demo-
graphic relative factors during Step 3 of this process.

The “M” value in field 7 indicates that this beneficiary

The next step in the risk score calculation process is a male.
involves gathering Mr. Carter’s demographics and The <SPACE> value in field 20 indicates that this benefi-
disease information. ciary does not reside in a long-term institution.
Please select each button below to learn more. When The <SPACE> value in field 23 indicates that this bene-
you have finished reviewing each section, select the ficiary did not receive a default score.
NEXT button to continue.
The number “1” value in field 39 indicates that this bene-
ficiary has Medicaid and is either full or partial dual.

The value in field 40 indicates that this beneficiary is in

the 80–84 age group.

The “CF” RAFT Code value in field 46 indicates that

this beneficiary is categorized as Community Full Dual,
which means he lives in the community, is enrolled in
Medicare, and receives full Medicaid benefits.

The “N” value in field 47 indicates that this beneficiary

does not have Frailty status.

The number “0” value in field 48 indicates that this

beneficiary is Medicare eligible due to age.

3-8
 Finally, the number “2” value in field 84 confirms both did not provide the Medicare Advantage Prescription
the values in fields 39 and 46, that this beneficiary is in Drug (or MARx) System a risk score. In this case, MARx
fact full dual. assigns a default score based on the demographics of
the beneficiary.
Disease Factors
Step 3: Identify Relative Factors &
Adjustment Factors

To gather Mr. Carter’s disease information, or HCCs,

you will need to reference the Model Output Report.
The next step is to gather Mr. Carter’s relative factors
Take a moment to review Mr. Carter’s information
and adjustment factors.
shown on your screen. Notice that the payment HCCs
identified for this beneficiary are HCC19, Diabetes Please select each button below to learn more. When
Without Complications, and HCC111, Chronic Obstruc- you have finished reviewing each section, select the
tive Pulmonary Disease. You will need this information NEXT button to continue.
when computing the sum of Mr. Carter’s disease rela-
tive factors in Step 3. Relative Factors

Note #1: Scenario 1

To gather Mr. Carter’s demographic relative factors and

disease relative factors, open the 2020 Rate Announce-
If Mr. Carter was a default enrollee, then field 23, Risk ment and navigate to the CMS-HCC, ESRD, and RxHCC
Adjustment Factor Code would have a number “1” Risk Adjustment Factors section.
through “7” indicating the Risk Adjustment system

3-9
 Since we know that Mr. Carter is a Part C Non-ESRD
beneficiary, Tables VI-1 through Table VI-3 may apply.
We know from the RAFT code that Mr. Carter is a
continuing enrollee. Therefore, the most appropri-
ate table to use when looking up Mr. Carter’s relative
factors is Table VI-1: 2020 Alternative Payment Condi-
tion Count Model Relative Factors for Continuing
Enrollees.

Click on the name of this table to continue.

To calculate Mr. Carter’s risk score, you must first

compute the sum of his demographic relative factors
as well as the sum of his disease relative factors.

Select each button below to learn more.

Demographic Factors

Now that we have selected the correct table, we need

to determine which column to use.

Take a few moments to review the headings for each

of the seven columns highlighted on your screen. Given
what you know about Mr. Carter, which one best fits his
status?

Go ahead; select the name of the column you think we

should use.
To calculate the sum of Mr. Carter’s demographic rela-
Well done! tive factors, you must first identify the correct factor for
each relevant demographic variable.
In Step 2, we learned that Mr. Carter is a full dual Medi-
care beneficiary, originally eligible due to age. There- Since we know that Mr. Carter is a male aged 80 to 84,
fore, we should reference the “Community, Full Bene- we need to capture the factor displayed in the cell that
fit Dual, Aged” column to gather his demographic and intersects row “80 to 84 Years” and the “Community,
disease relative factors. Full Benefit Dual, Aged” column, which is 0.803.

Select the button on your screen to continue. The only other value we need to consider when
computing the sum of Mr. Carter’s demographic rela-
tive factors is his Medicaid Interaction factor. Since Mr.
Carter resides in the community, this factor does not
apply. Therefore, Mr. Carter’s demographic relative

3-10
 factor, using the 2020 CMS-HCC model, is 0.803. This makes the sum of Mr. Carter’s disease relative
factors, using the 2020 CMS-HCC model, 0.537.
Select the NEXT button to continue.
Select the NEXT button to continue.
Disease Factors
Adjustment Factors

Using the same table (Table VI-1) and the “Community,

Full Benefit Dual, Aged” column that we used to gath- It is now time to identify the adjustment factors you will
er Mr. Carter’s demographic relative factors, the next need to apply when calculating Mr. Carter’s risk score.
step in the process is to locate the values for each of Mr.
Carter’s disease factors (or HCCs) in order to compute Select the name of each adjustment factor to learn
the sum of his disease relative factors. more. When you are finished, select the NEXT button
to continue.
During our MOR review in step 2, we confirmed that
Mr. Carter has two HCCs: Normalization Factor

• HCC19, Diabetes without complication, and

• HCC111, Chronic obstructive pulmonary disease.

To identify the relative factor for HCC19, we must

capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC19,” which is 0.107.

To identify the relative factor for HCC111, we must

capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC111,” which is 0.430.
The normalization factor adjusts for population and
Since there are no disease or disabled/disease interac- coding changes between the denominator year and the
tions for this beneficiary, to compute the sum of Mr. payment year. To locate the normalization factor infor-
Carter’s disease relative factors, simply add the relative mation for calculating Mr. Carter’s 2022 risk score, you
factor for HCC19, which is 0.107, to the relative factor will need to reference the 2022 Rate Announcement.
for HCC111, which is 0.430.
Select the SHOW ME button on your screen to learn more.

3-11
 Open the 2022 Rate Announcement by selecting the Open the 2022 Rate Announcement and navigate to
RESOURCES button below and navigate to the “Final the “Adjustment for MA Coding Pattern Differences”
2022 Part C and ESRD Normalization Factors” section. section.

Notice that the Rate Announcement includes separate Based on the information in this Announcement, we
normalization factors for each model. must use the MA coding pattern adjustment factor
of 5.90 percent, or 0.059, to calculate Mr. Carter’s risk
Since we are using the 2020 CMS-HCC model to calcu-
score for the 2022 Payment Year.
late Mr. Carter’s risk score, the normalization factor we
need to use is 1.118.

Select the NEXT button to continue.

MA Coding Pattern Adjustment Factor

Select the NEXT button to continue.

Great job! You have gathered all the values needed

to complete the fourth and final step of the risk score
calculation process.
The Medicare Advantage coding pattern adjustment
Select the NEXT button to continue.
factor we will need to compute Mr. Carter’s 2022 risk
score can also be found in the 2022 Rate Announce-
ment.

Select the SHOW ME button on your screen to learn

more.

3-12
 Note #2: Scenario 1 Select the SHOW ME button on your screen to learn
more.

When using the CMS-HCC model for a beneficiary that

resides in the community, the Medicaid factor will When using the risk score formula, we must perform
always be zero because in this model, the Medicaid several steps to accurately compute Mr. Carter’s final
factor only applies to beneficiaries who reside in insti- payment risk score. Select each button below to learn
tutions. For beneficiaries residing in the community, more.
Medicaid status and eligibility are used to determine
which segment of the model to use for risk score calcu- When you are finished, select the NEXT button to
lation. continue.

Step 4: Using the Risk Score Formula 1. Calculate the Raw Risk Score

It is finally time to calculate Mr. Carter’s risk score! The first step is to compute Mr. Carter’s raw risk score
by adding the sum of his demographic factors, which is
In this step, we will enter the values generated in Step 0.803, to the sum of disease factors, which is 0.537.
3—the sum of demographic relative factors, sum of
disease relative factors, normalization factor, and the Mr. Carter’s raw risk score is 1.340.
MA coding pattern adjustment factor—into the risk
score formula to produce Mr. Carter’s final risk score for
Payment Year 2022.

3-13
 2. Calculate the Rounded adjusted risk score of 1.1283. You must round this
Normalized Risk Score number to the third decimal point to get a rounded MA
coding pattern adjusted risk score of 1.128. This is Mr.
Carter’s risk score for Payment Year 2022.

Scenario 1 Conclusion

Next, you must divide the raw risk score, 1.340, by

the 2022 normalization factor for the 2020 CMS-HCC
model, which is 1.118, to get a normalized score of
1.19857. You must round this number to the third deci-
mal place to get a rounded normalized score of 1.199. Congratulations! You have just completed the four-step
Please note that if you skip rounding it may affect the process of calculating Mr. Carter’s risk score.
risk score at the third decimal place, which will impact
From gathering the necessary reports and resources
the final calculated payment risk score.
and finding beneficiary demographic and disease infor-
3. Calculate the MA Coding mation to identifying relative and adjustment factors
Pattern Adjusted Risk Score and using the risk score formula, you now know how to
compute the 2022 payment risk score for a Part C bene-
ficiary who is not in End-Stage Renal Disease status.

There are three additional scenarios in this module. To

choose another scenario, simply select the CHOOSE
SCENARIO button displayed on your screen. Other-
wise, select the NEXT button to continue.

Since we are using the Part C CMS-HCC model to calcu-

late Mr. Carter’s risk score, we also need to apply the
MA coding pattern adjustment factor. To do this, simply
multiply the rounded normalized risk score, which is
1.199, by 1 minus the MA coding pattern adjustment
factor, which is 0.941, to get the MA coding pattern

3-14
 Scenario 2: Part C (ESRD) Step 1: Gather Reports & Resources

To accurately calculate Ms. Ellis’ risk score, the first step

Meet Ms. Ellis, a 76-year-old beneficiary with kidney
in the process is to gather all the reports and resources
disease. Ms. Ellis received a transplant 5 months ago.
you will need which you can now access by selecting
She is currently residing in a long-term care facility. Ms.
the RESOURCES button at the bottom of your screen.
Ellis is eligible for both Medicare and Medicaid.
These reports and resources include the Monthly
This scenario demonstrates how to calculate a risk
Membership Detail Report (or MMR), the Model Output
score for a Part C beneficiary who is in End-Stage Renal
Report (or MOR), the Plan Communication User Guide
Disease status.
(or PCUG), and the Annual Rate Announcement.
To begin the process of calculating Ms. Ellis’ risk score,
Please select the NEXT button to continue.
select the START SCENARIO button. To choose another
scenario, select the BACK button at the top right of Monthly Membership Detail Report
your screen.

The Monthly Membership Detail Report provides Plans

Ready to walk through the process of calculating Ms. with beneficiary demographic and status information.
Ellis’ risk score? You will need the data found in this report to complete
Step 3 of the risk score calculation process.
Select the STEP 1 button to begin.

3-15
 Model Output Report Annual Rate Announcement

Recall from Module 2 that the Model Output Report is To ensure that you are entering correct data into the
a monthly report that lists the diseases applicable to risk score calculation formula, first you must verify
each beneficiary. You will need the data from this report which payment year your risk score calculation should
to complete Step 3 in the risk score calculation process. be based on. For this scenario, Ms. Ellis’ risk score calcu-
lations will be based on Payment Year 2022.
Plan Communication User Guide
You will therefore need to refer to the 2022 Rate
Announcement to determine which risk adjustment
payment model to use.

Select the SHOW ME button to learn more.

The PCUG provides layout and field description infor-

mation for the MMR and MOR.

To verify which risk adjustment payment model you’ll

need to use for 2022 risk score calculation, open up the
2022 Rate Announcement and navigate to the section
that describes the ESRD Risk Adjustment Model. The
information in this section states that CMS complet-
ed the phasing in of the model implemented in 2020.
Therefore, 100 percent of the risk score will be calculat-
ed using the 2020 ESRD model.
3-16
 Since we will be using the 2020 ESRD Functioning Graft Demographic Factors
model to calculate Ms. Ellis’ risk score, we will also need
to refer to the 2020 Rate Announcement to look up her
demographic and disease relative factors. This infor-
mation is required to complete Step 3 of the risk score
calculation process.

Select the NEXT button to continue.

Step 2: Identify Demographics

& Disease Information

To gather Ms. Ellis’ demographic information, you

will need to reference the Monthly Membership
Detail Report.

The highlighted data shown on this MMR table are

required demographic and model indicators that you
will need to compute the sum of Ms. Ellis’ demographic
relative factors during Step 3 of this process.

The “F” value in field 7 indicates that this beneficiary is

The next step in the risk score calculation process a female.
involves gathering Ms. Ellis’ demographics and disease The “Y” value in field 15 indicates this beneficiary has
information. ESRD status.
Please select each button below to learn more. When The “Y” value in field 20 indicates that this beneficiary
you have finished reviewing each section, select the resides in a long-term institution.
NEXT button to continue.
The “Y” value in field 21 indicates that a Medicaid
add-on factor is used in the payment.

The <SPACE> value in field 23 indicates that this bene-

ficiary did not receive a default score.

The value in field 40 indicates that this beneficiary is in

the 75-79 age group.

The “I1” RAFT Code value in field 46 indicates that this

beneficiary is categorized as Institutional Post Graft 4
to 9 months, which means she lives in an institution and
is 4 to 9 months post-transplant.

And lastly, the number “0” value in field 48 indicates

that this beneficiary is Medicare eligible due to age.

3-17
 Disease Factors Step 3: Identify Relative Factors &
Adjustment Factors

To gather Ms. Ellis’ disease information, or HCCs, you

will need to reference the Model Output Report. The next step is to gather Ms. Ellis’ relative factors and
adjustment factors.
Take a moment to review Ms. Ellis’ information shown
on your screen. Notice that the payment HCCs iden- Please select each button below to learn more. When
tified for this beneficiary are HCC12, Breast, Prostate, you have finished reviewing each section, select the
and Other Cancer and Tumors, and HCC18, Diabetes NEXT button to continue.
with Chronic Complications. You will need this informa-
tion when computing the sum of Ms. Ellis’ disease rela- Relative Factors
tive factors in Step 3.

Note #1: Scenario 2

To gather Ms. Ellis’ demographic relative factors and

disease relative factors, open the 2020 Rate Announce-
ment and navigate to the CMS-HCC, ESRD, and RxHCC
You learned earlier that a beneficiary’s ESRD status can Risk Adjustment Factors section.
be found in the MMR. However, did you know that you
Since we know that Ms. Ellis is a Part C ESRD beneficia-
can also confirm a beneficiary’s ESRD status by refer-
ry, Tables VI-5 through Table VI-11 may apply. We also
encing the MOR? If there is a “Y” value under the ESRD
know that Ms. Ellis is living in an institution and has
column (to the far right), your beneficiary’s status is
a functioning graft. Therefore, the most appropriate
ESRD.
table to use when looking up Ms. Ellis’ relative factors

3-18
 is Table VI-9: ESRD Model Functioning Graft Relative Demographic Factors
Factors for Institutionalized Population.

Click on the name of this table to continue.

To calculate the sum of Ms. Ellis’ demographic relative

factors, you must first identify the correct factor for
each relevant demographic variable.
Take a few moments to review the table highlighted on
your screen. Notice the relative factors column to the Since we know that Ms. Ellis is a female aged 75 to 79,
right and the functioning graft factors section at the we need to capture the factor displayed in row “75 to 79
top of the table. Years,” which is 1.182.

You will need to reference this information later in the Since we know that Ms. Ellis was originally eligible for
process. Medicare due to age, the “Originally Disabled” factor
does not apply.
Select the NEXT button to continue.
Since we know that Ms. Ellis is Medicaid eligible, we
need to capture the factor displayed in the Medicaid
row, which is 0.089.

Select the NEXT button to continue.

To calculate Ms. Ellis’ risk score, you must first compute

the sum of her demographic relative factors as well as
the sum of her disease relative factors.

Select each button below to learn more.

3-19
 Disease Factors

The only other value we must account for to calculate

the sum of Ms. Ellis’ demographic relative factors is Using the same table (Table VI-9) that we used to gath-
her functioning graft factor. er Ms. Ellis’ demographic relative factors, the next step
in the process is to locate the values for each of Ms.
We know that Ms. Ellis is over age 65 and that her
Ellis’ disease factors (or HCCs) in order to compute the
transplant surgery was five months ago. Therefore, we
sum of her disease relative factors.
must capture the value displayed for row “Ages 65+,
with duration since transplant of 4-9 months,” which is During our MOR review in step 2, we confirmed that
3.064. Ms. Ellis has two HCCs:

Now, to compute the sum of Ms. Ellis’ demographic • HCC12, Breast, Prostate, and Other Cancers and
relative factors, simply add up her age and sex factor Tumors, and
from the previous screen, which is 1.182, the Medicaid
factor, which is 0.089, and her functioning graft factor, • HCC18, Diabetes with Chronic Complications.
which is 3.064. To identify the relative factor for HCC12, we must
This makes the sum of Ms. Ellis’ demographic relative capture the value for row “HCC12,” which is 0.232.
factors, using the 2020 ESRD model, 4.335. To identify the relative factor for HCC18, we must
Select the NEXT button to continue. capture the value displayed for row “HCC18,” which is
0.446.

Since there are no disease or disabled/disease interac-

tions for this beneficiary, to compute the sum of Ms.
Ellis’ disease relative factors, simply add the relative
factor for HCC12, which is 0.232, to the relative factor
for HCC18, which is 0.446.

This makes the sum of Ms. Ellis’ disease relative factors,

using the 2020 ESRD model, 0.678.

Select the NEXT button to continue.

3-20
 Adjustment Factors

Open the 2022 Rate Announcement by selecting the

It is now time to identify the adjustment factors you will RESOURCES button below and navigate to the “Final
need to apply when calculating Ms. Ellis’ risk score. 2022 Part C and ESRD Normalization Factors” section.

Select the name of each adjustment factor to learn Notice that the Rate Announcement includes separate
more. When you are finished, select the NEXT button normalization factors for each model.
to continue.
Since we are using the 2020 ESRD Functioning Graft
Normalization Factor model to calculate Ms. Ellis’ risk score, the normaliza-
tion factor we need to use is 1.126.

Select the NEXT button to continue.

MA Coding Pattern Adjustment Factor

The normalization factor adjusts for population and

coding changes between the denominator year and the
payment year.

To locate the normalization factor information for

The Medicare Advantage coding pattern adjustment
calculating Ms. Ellis’ 2022 risk score, you will need to
factor we will need to compute Ms. Ellis’ 2022 risk score
reference the 2022 Rate Announcement.
can also be found in the 2022 Rate Announcement.
Select the SHOW ME button on your screen to learn
Select the SHOW ME button on your screen to learn
more.
more.

3-21
 Note #2: Scenario 2

Open the 2022 Rate Announcement and navigate to

the “Adjustment for MA Coding Pattern Differences” Ms. Ellis’ risk score used the post-transplant or function-
section. ing graft component of the ESRD model. This model
segment includes additional factors in the risk score
Based on the information in this Announcement, we
that account for the extra costs of immunosuppressive
must use the MA coding pattern adjustment factor of
drugs and higher intensity of care for this group. The
5.90 percent, or 0.059, to calculate Ms. Ellis’ risk score
additional factors differ for months 4-9 after a trans-
for the 2022 Payment Year.
plant and for month 10 onward.
Select the NEXT button to continue.
Step 4: Using the Risk Score Formula

Great job! You have gathered all the values needed

to complete the fourth and final step of the risk score It is finally time to calculate Ms. Ellis’ risk score!
calculation process. In this step, we will enter the values generated in Step 3—
Select the NEXT button to continue. the sum of demographic relative factors, sum of disease
relative factors, normalization factor, and the MA coding
pattern adjustment factor—into the risk score formula to
produce Ms. Ellis’ final risk score for Payment Year 2022.

Select the SHOW ME button on your screen to learn

more.

3-22
 2. Calculate the Rounded
Normalized Risk Score

When using the risk score formula, we must perform

several steps to accurately compute Ms. Ellis’ final
payment risk score. Select each button below to learn Next, you must divide the raw risk score, 5.013, by the
more. 2022 normalization factor for the 2020 ESRD Function-
ing Graft model, which is 1.126, to get a normalized
When you are finished, select the NEXT button to
score of 4.45204. You must round this number to the
continue.
third decimal place to get a rounded normalized score
1. Calculate the Raw Risk Score of 4.452. Please note that if you skip rounding, it may
affect the risk score at the third decimal place, which
will impact the final calculated payment risk score.

3. Calculate the MA Coding

Pattern Adjusted Risk Score

The first step is to compute Ms. Ellis’ raw risk score by

adding the sum of her demographic factors, which is
4.335, to the sum of disease factors, which is 0.678.

Ms. Ellis’ raw risk score is 5.013.

Since we are using the Part C ESRD model to calcu-
late Ms. Ellis’ risk score, we also need to apply the MA
coding pattern adjustment factor. To do this, simply
multiply the rounded normalized risk score, which is
4.452, by 1 minus the MA coding pattern adjustment
factor, which is 0.941, to get the MA coding pattern

3-23
 adjusted risk score of 4.18933.

You must round this number to the third decimal point

to get a rounded MA coding pattern adjusted risk score
of 4.189.

This is Ms. Ellis’ risk score for Payment Year 2022.

Scenario 2 Conclusion

Congratulations! You have just completed the four-

step process of calculating Ms. Ellis’ risk score.

From gathering the necessary reports and resources

and finding beneficiary demographic and disease infor-
mation to identifying relative and adjustment factors
and using the risk score formula, you now know how to
compute the 2022 payment risk score for a Part C bene-
ficiary who is in End-Stage Renal Disease status.

There are three additional scenarios in this module. To

choose another scenario, simply select the CHOOSE
SCENARIO button displayed on your screen. Other-
wise, select the NEXT button to continue.

3-24
 Scenario 3: Part C (PACE) Step 1: Gather Reports & Resources

To accurately calculate Ms. Price’s risk score, the first

Meet Ms. Price, a 72-year-old who lives at home.
step in the process is to gather all the reports and
This scenario demonstrates how to calculate a risk resources you will need which you can now access by
score for a Part C beneficiary who is not in End-Stage selecting the RESOURCES button at the bottom of
Renal Disease status and is enrolled in a Program of All your screen.
Inclusive Care for the Elderly (or PACE) plan.
These reports and resources include the Monthly
To begin the process of calculating Ms. Price’s risk Membership Detail Report (or MMR), the Model Output
score, select the START SCENARIO button. To choose Report (or MOR), the Plan Communication User Guide
another scenario, select the BACK button at the top (or PCUG), and the Annual Rate Announcement.
right of your screen.
Please select the NEXT button to continue.

Monthly Membership Detail Report

Ready to walk through the process of calculating Ms.

Price’s risk score?
The Monthly Membership Detail Report provides Plans
Select the STEP 1 button to begin.
with beneficiary demographic and status information.
You will need the data found in this report to complete
Step 3 of the risk score calculation process.

3-25
 Model Output Report Annual Rate Announcement

Recall from Module 2 that the Model Output Report is To ensure that you are entering correct data into the
a monthly report that lists the diseases applicable to risk score calculation formula, first you must verify
each beneficiary. You will need the data from this report which payment year your risk score calculation should
to complete Step 3 in the risk score calculation process. be based on. For this scenario, Ms. Price’s risk score
calculations will be based on Payment Year 2022.
Plan Communication User Guide
You will therefore need to refer to the 2022 Rate
Announcement to determine which risk adjustment
payment model to use.

Select the SHOW ME button to learn how.

The PCUG provides layout and field description infor-

mation for the MMR and MOR.

To verify which risk adjustment payment model you’ll

need to use for 2022 risk score calculation, reference
the 2022 Rate Announcement and navigate to the
section that describes the CMS-HCC Risk Adjustment
Model used for PACE organizations. The information
in this section states that the risk scores for non-ES-
RD PACE participants will be calculated using the 2017
CMS-HCC model.
3-26
 Since we will be using the 2017 CMS-HCC model to Demographic Factors
calculate Ms. Price’s risk score, we will also need to
refer to the 2017 Rate Announcement to look up her
demographic and disease relative factors. This infor-
mation is required to complete Step 3 of the risk score
calculation process.

Select the NEXT button to continue.

Step 2: Identify Demographics &

Disease Information

To gather Ms. Price’s demographic information, you

will need to reference the Monthly Membership Detail
Report. The highlighted data shown on this MMR table
are required demographic and model indicators that
you will need to compute the sum of Ms. Price’s demo-
graphic relative factors during Step 3 of this process.

The “F” value in field 7 indicates that this beneficiary is

a female.

The next step in the risk score calculation process The <SPACE> value in field 20 indicates that this benefi-
involves gathering Ms. Price’s demographics and ciary does not reside in a long-term institution.
disease information. The <SPACE> value in field 23 indicates that this bene-
Please select each button below to learn more. When ficiary did not receive a default score.
you have finished reviewing each section, select the The number “1” value in field 39 indicates that this bene-
NEXT button to continue. ficiary has Medicaid and is either full or partial dual.

The value in field 40 indicates that this beneficiary is in

the 70–74 age group.

The “PI” RAFT Code value in field 46 indicates that

this beneficiary is categorized as PACE Community
Full Dual, which means she lives in the community, is
enrolled in a PACE plan for Medicare, and receives full
Medicaid benefits.

The “Y” value in field 47 indicates that this beneficiary

has Frailty status.

The number “0” value in field 48 indicates that this

beneficiary is Medicare eligible due to age.

3-27
 The number “0.105” value in field 80 is the Part C Frailty Note #1: Scenario 3
Factor used in the risk score calculation.

Finally, the number “8” value in field 84 confirms the

value in field 46, that this beneficiary is in fact full dual.

Disease Factors

The Program of All Inclusive Care for the Elderly (or

PACE) is a set of programs that provide capitated
managed care plans for frail elderly and disabled bene-
ficiaries.

One of the main goals of PACE plans is to keep the

To gather Ms. Price’s disease information, or HCCs, you elderly in the community. To realize this goal, PACE
will need to reference the Model Output Report. plans offer the following services: drug coverage,
Take a moment to review Ms. Price’s information primary care, hospital care, nursing home services,
shown on your screen. Notice that the payment HCCs emergency services, dental services, social services,
identified for this beneficiary are HCC19, Diabetes and more.
Without Complications; HCC35, Inflammatory Bowel
Disease; HCC40, Rheumatoid Arthritis and Inflamma-
Step 3: Identify Relative Factors &
tory Connective Tissue Disease; and HCC111, Chron-
Adjustment Factors
ic Obstructive Pulmonary Disease. You will need this
information when computing the sum of Ms. Price’s
disease relative factors in Step 3.

The next step is to gather Ms. Price’s relative factors

and adjustment factors. Please select each button
below to learn more. When you have finished reviewing
each section, select the NEXT button to continue.

3-28
 Relative Factors Given what you know about Ms. Price, which one best
fits her status?

Go ahead; select the name of the column you think we

should use.

Well done!

In Step 2, we learned that Ms. Price is a full dual Medi-

care beneficiary, originally eligible due to age. There-
fore, we should reference the “Community, Full Benefit
Dual, Aged” column to gather her demographic and
disease relative factors.

Select the button on your screen to continue.

To gather Ms. Price’s demographic relative factors and
disease relative factors, open the 2017 Rate Announce-
ment and navigate to the CMS-HCC and RxHCC Risk
Adjustment Factors section.

Since we know that Ms. Price is a Part C beneficiary,

Tables VI-1 through Table VI-3 may apply. We also know
from her RAFT code that Ms. Price is a continuing enroll-
ee. Therefore, the most appropriate table to use when
looking up Ms. Price’s relative factors is Table VI-1: 2017
CMS-HCC Model Relative Factors for Community and
Institutional Beneficiaries.

Click on the name of this table to continue.

To calculate Ms. Price’s risk score, you must first
compute the sum of her demographic relative factors
as well as the sum of her disease relative factors.

Select each button below to learn more.

Now that we have selected the correct table, we need

to determine which column to use.

Take a few moments to review the headings for each

of the seven columns highlighted on your screen.

3-29
 Demographic Factors Disease Factors

To calculate the sum of Ms. Price’s demographic rela- Using the same table (Table VI-1) and the “Community,
tive factors, you must first identify the correct factor for Full Benefit Dual, Aged” column that we used to gather
each relevant demographic variable. Ms. Price’s demographic relative factors, the next step
in the process is to locate the values for each of Ms.
Since we know that Ms. Price is a female aged 70 to 74,
Price’s disease factors (or HCCs) in order to compute
we need to capture the factor displayed in the cell that
the sum of her disease relative factors.
intersects row “70 to 74 Years” and the “Community,
Full Benefit Dual, Aged” column, which is 0.511. During our MOR review in step 2, we confirmed that
Ms. Price has four HCCs:
The only other factor we need to consider when calcu-
lating Ms. Price’s sum of demographic relative factors is • HCC19, Diabetes without complication,
the Medicaid interaction factor. Since Ms. Price resides
• HCC35, Inflammatory Bowel Disease,
in the community, this factor does not apply.
• HCC40, Rheumatoid Arthritis and Inflammatory
Therefore, the sum of Ms. Price’s demographic relative
Connective Tissue Disease,
factors, using the 2017 CMS-HCC model, is 0.511.
• and HCC111, Chronic obstructive pulmonary
Select the NEXT button to continue.
disease.

To identify the relative factor for HCC19, we must

capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC19,” which is 0.097.

To identify the relative factor for HCC35, we must

capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC35,” which is 0.334.

To identify the relative factor for HCC40, we must

capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC40,” which is 0.370.

3-30
 To identify the relative factor for HCC111, we must Normalization Factor
capture the value displayed in the cell intersecting the
“Community, Full Benefit Dual, Aged” column and row
“HCC111,” which is 0.422.

Since there are no disease or disabled/disease interac-

tions for this beneficiary, to compute the sum of Ms.
Price’s disease relative factors, simply add the value for
each of her four HCC factors together.

This makes the sum of Ms. Price’s disease relative

factors, using the 2017 CMS-HCC model, 1.223.

Select the NEXT button to continue.

Adjustment Factors The normalization factor adjusts for population and

coding changes between the denominator year and the
payment year.

To locate the normalization factor information needed

to calculate Ms. Price’s 2022 risk score, you will need to
reference the 2022 Rate Announcement.

Select the SHOW ME button on your screen to learn

more.

It is now time to identify the adjustment factors you will

need to apply when calculating Ms. Price’s risk score.

Select the name of each adjustment factor to learn

more. When you are finished, select the NEXT button
to continue.

Open the 2022 Rate Announcement by selecting the

RESOURCES button below and navigate to the “Final
2022 Part C and ESRD Normalization Factors” section.
Notice that the Rate Announcement includes separate
normalization factors for each model.

Since we are using the 2017 CMS-HCC model to calcu-

late Ms. Price’s risk score, the normalization factor we
need to use is 1.128. Select the NEXT button to continue.

3-31
 MA Coding Pattern Adjustment Factor Frailty Score Factor

The Medicare Advantage coding pattern adjustment To calculate the risk score for a beneficiary enrolled in a
factor we will need to compute Ms. Price’s 2022 risk PACE plan, you will need their frailty score.
score can also be found in the 2022 Rate Announce-
Select the SHOW ME button on your screen to learn
ment.
more.
Select the SHOW ME button on your screen to learn
more.

You learned earlier that frailty score information can be

found on field 80 in the MMR.
Open the 2022 Rate Announcement and navigate to
Frailty scores are also posted to the Health Plan
the “Adjustment for MA Coding Pattern Differences”
Management System or HPMS. Plans are notified once
section.
frailty scores become available via an HPMS memo.
Based on the information in this Announcement, we
To locate your beneficiary’s frailty score information on
must use the MA coding pattern adjustment factor of
the HPMS once you are logged in, navigate to the Risk
5.90 percent, or 0.059, to calculate Ms. Price’s risk score
Adjustment module and select SURVEY INFORMA-
for the 2022 Payment Year.
TION from the left navigation pane.
Select the NEXT button to continue.
Under survey, from the SELECT FUNCTION dropdown,
choose “Survey Results for Frailty Adjustment.”

3-32
 Select the appropriate year before selecting the Per the statute, for 2019 and each subsequent year, the
SUBMIT button. minimum required coding adjustment factor is 5.90
percent.
Ms. Price’s frailty score is 0.105.

Select the NEXT button to continue. Step 4: Using the Risk Score Formula

It is finally time to calculate Ms. Price’s risk score!

Great job! You have gathered all the values needed
to complete the fourth and final step of the risk score In this step, we will enter the values generated in Step
calculation process. 3 — the sum of demographic relative factors, sum of
disease relative factors, the normalization factor, the
Select the NEXT button to continue.
MA coding pattern adjustment factor, and the frail-
ty score — into the risk score formula to produce Ms.
Note #2: Scenario 3
Price’s final risk score for Payment Year 2022.

Select the SHOW ME button on your screen to learn

more.

Because CMS calibrates the CMS-HCC model using FFS

or fee-for-service data, the relative factors in this model
reflect the fee-for-service pattern of diagnostic coding.

Since the Medicare Advantage coding patterns differ When using the risk score formula, we must perform
from FFS coding patterns, CMS created the MA coding several steps to accurately compute Ms. Price’s final
pattern adjustment factor to keep MA risk scores payment risk score.
comparable to FFS risk scores.
3-33
 Select each button below to learn more. When you are 3. Calculate the MA Coding
finished, select the NEXT button to continue. Pattern Adjusted Risk Score

1. Calculate the Raw Risk Score

Since we are using the Part C CMS-HCC model to calcu-

late Ms. Price’s risk score, we also need to apply the MA
The first step is to compute Ms. Price’s raw risk score coding pattern adjustment factor. To do this, simply
by adding the sum of her demographic factors, which multiply the rounded normalized risk score, which is
is 0.511, to the sum of disease factors, which is 1.223. 1.537, by 1 minus the MA coding pattern adjustment
factor, which is 0.941, to get the MA coding pattern
Ms. Price’s raw risk score is 1.734.
adjusted risk score of 1.44632.
2. Calculate the Rounded
You must round this value to the third decimal place to
Normalized Risk Score
get Ms. Price’s rounded MA coding pattern adjusted
risk score of 1.446.

4. Calculate the Frailty Adjusted Risk Score

Next, you must divide the raw risk score, 1.734, by

the 2022 normalization factor, which is 1.128, to get
a normalized score of 1.53723. You must round this
number to the third decimal place to get a round- Since Ms. Price is a PACE beneficiary, we need to incor-
ed normalized score of 1.537. Please note that if you porate her frailty score into the final calculation of her
skip rounding, it may affect the risk score at the third payment risk score.
decimal place, which will impact the final calculated
payment risk score.

3-34
 To do this, simply add Ms. Price’s rounded MA coding
pattern adjusted risk score, which is 1.446, to her frailty
score of 0.105, which gives you her frailty adjusted risk
score of 1.551.

This is Ms. Price’s final PACE risk score for Payment

Year 2022.

Scenario 3 Conclusion

Congratulations! You have just completed the four-step

process of calculating Ms. Price’s risk score.

From gathering the necessary reports and resources

and finding beneficiary demographic and disease infor-
mation to identifying relative and adjustment factors
and using the risk score formula, you now know how to
compute the 2022 payment risk score for a Part C bene-
ficiary who is not in End-Stage Renal Disease status and
is enrolled in a PACE plan.

There are three additional scenarios in this module. To

choose another scenario, simply select the CHOOSE
SCENARIO button displayed on your screen. Other-
wise, select the NEXT button to continue.

3-35
 Scenario 4: Part D (RxHCC) Step 1: Gather Reports & Resources

To accurately calculate Mr. Daniels’ risk score, the

Meet Mr. Daniels, a 67-year-old who lives at home.
first step in the process is to gather all the reports and
This scenario demonstrates how to calculate a risk resources you will need which you can now access by
score for a Part D beneficiary. selecting the RESOURCES button at the bottom of
your screen.
To begin the process of calculating Mr. Daniels’ risk
score, select the START SCENARIO button. To choose These reports and resources include the Monthly
another scenario, select the BACK button at the top Membership Detail Report (or MMR), the Model Output
right of your screen. Report (or MOR), the Plan Communication User Guide
(or PCUG), and the Annual Rate Announcement.

Please select the NEXT button to continue.

Monthly Membership Detail Report

Ready to walk through the process of calculating Mr.

Daniels’ risk score?

Select the STEP 1 button to begin.

The Monthly Membership Detail Report provides Plans

with beneficiary demographic and status information.
You will need the data found in this report to complete
Step 3 of the risk score calculation process.

3-36
 Model Output Report Annual Rate Announcement

Recall from Module 2 that the Model Output Report is To ensure that you are entering correct data into the
a monthly report that lists the diseases applicable to risk score calculation formula, first you must verify
each beneficiary. You will need the data from this report which payment year your risk score calculation should
to complete Step 3 in the risk score calculation process. be based on. For this scenario, Mr. Daniels’ risk score
calculations will be based on Payment Year 2022.
Plan Communication User Guide
You will therefore need to refer to the 2022 Rate
Announcement to determine which risk adjustment
payment model to use.

Select the SHOW ME button to learn how.

The PCUG provides layout and field description infor-

mation for the MMR and MOR.

To verify which risk adjustment payment model you’ll

need to use for 2022 risk score calculation, reference
the 2022 Rate Announcement and scroll to the section
that describes the RxHCC risk adjustment model. The
information in this section states that CMS finalized the
updated version of the RxHCC risk adjustment model,
as proposed in the 2022 Advance Notice, that uses diag-
nosis data from 2017 “fee for service” claims and MA
3-37
 encounter data submissions, along with expenditure Demographic Factors
data from 2018 prescription drug events. Therefore,
we must use the 2022 RxHCC payment risk adjustment
model to calculate Mr. Daniels’ risk score.

The 2022 RxHCC Risk Adjustment model factors,

which are required to complete Step 3 of the risk score
calculation process, are also included in the 2022 Rate
Announcement.

Select the NEXT button to continue.

Step 2: Identify Demographics

& Disease Information
To gather Mr. Daniels’ demographic information, you
will need to reference the Monthly Membership Detail
Report.

The highlighted data shown on this MMR table are

required demographic and model indicators that you
will need to compute the sum of Mr. Daniels’ demo-
graphic relative factors during Step 3 of this process.

The “M” value in field 7 indicates that this beneficiary

is a male.

The value in field 40 indicates that this beneficiary is in

The next step in the risk score calculation process the 65–69 age group.
involves gathering Mr. Daniels’ demographics and The number “0” value in field 48 indicates that this
disease information. beneficiary is Medicare eligible due to age.
Please select each button below to learn more. When Finally, the “D1” Part D RAFT Code value in field 86
you have finished reviewing each section, select the indicates that this beneficiary is categorized as a
NEXT button to continue. non-low-income continuing enrollee residing in the
community.

3-38
 Disease Factors Step 3: Identify Relative Factors
& Adjustment Factors

To gather Mr. Daniels’ disease information, or RxHCCs,

you will need to reference the Model Output Report. The next step is to gather Mr. Daniels’ relative factors
and adjustment factors.
Take a moment to review Mr. Daniels’ information
shown on your screen. Notice that the payment Please select each button below to learn more. When
RxHCCs identified for this beneficiary are RxHCC19, you have finished reviewing each section, select the
Breast and Other Cancers and Tumors, and RxHCC97, NEXT button to continue.
Immune Disorders. You will need this information when
computing the sum of Mr. Daniels’ disease relative Relative Factors
factors in Step 3.

Note #1: Scenario 4

To gather Mr. Daniels’ demographic relative factors and

disease relative factors, open the 2022 Rate Announce-
ment and navigate to the RxHCC Risk Adjustment
The Part D Low-Income Subsidy (or LIS) provides extra Factors section.
help with prescription drug costs for eligible individu-
Since we know that Mr. Daniels’ is a Part D beneficia-
als whose income and resources are limited. The LIS
ry, Table VIII-1 and VIII-4 may apply. We know from
provides premium and cost-sharing (including deduct-
the Part D RAFT code that Mr. Daniels is a continuing
ibles and cost-sharing during the coverage gap) associ-
enrollee. Therefore, the most appropriate table to use
ated with the Part D benefit.
when looking up Mr. Daniels’ relative factors is Table

3-39
 VIII-1: RxHCC Model (2017/2018) Relative Factors for
Continuing Enrollees.

Click on the name of this table to continue.

To calculate Mr. Daniels’ risk score, you must first

compute the sum of his demographic relative factors
as well as the sum of his disease relative factors.

Select each button below to learn more.

Now that we have selected the correct table, we need
to determine which column to use. Demographic Factors
Take a few moments to review the headings for each
of the five columns highlighted on your screen. Given
what you know about Mr. Daniels, which one best fits
his status?

Go ahead; select the name of the column you think we

should use.

Well done!

In Step 2, we learned that Mr. Daniels is a 65 to 69 year

old, non-low-income enrollee residing in the commu-
nity. Therefore, we should reference the “Communi-
ty, Non-Low Income, Age ≥ 65” column to gather his To calculate the sum of Mr. Daniels’ demographic rela-
demographic and disease relative factors. tive factors, you must first identify the correct factor for
Select the button on your screen to continue. each relevant demographic variable.

Since we know that Mr. Daniels is a male aged 65 to 69,

we need to capture the value displayed in the cell that
intersects row “65 to 69 Years,” and the “Community,
Non-Low Income, Age ≥ 65” column, which is 0.196.

The only other factor we may need to account for in

order to determine Mr. Daniels’ demographic relative
factor, is his “Originally Disabled” factor.

Since Mr. Daniels was eligible for Medicare due to age,

3-40
 the “Originally Disabled” factor does not apply. There- tive factor for RxHCC19, which is 0.110, to the relative
fore, Mr. Daniels’ demographic relative factor, using factor for RxHCC97, which is 0.782.
the 2022 RxHCC payment model, is 0.196.
This makes the sum of Mr. Daniels’ disease relative
Select the NEXT button to continue. factors, using the 2022 RxHCC payment model, 0.892.

Disease Factors Select the NEXT button to continue.

Adjustment Factors

Using the same table (Table VIII-1) and the “Communi-

ty, Non-Low Income, Age ≥ 65” column that we used
to gather Mr. Daniels’ demographic relative factors, the It is now time to review the adjustment factor you will
next step in the process is to locate the values for each need to apply when calculating Mr. Daniels’ risk score.
of Mr. Daniels’ disease factors (or RxHCCs) in order to Please note that when calculating risk scores using a
compute the sum of his disease relative factors. Part D model, the MA coding pattern adjustment factor
During our MOR review in Step 2, we confirmed that should not be applied.
Mr. Daniels has two RxHCCs: Select the NORMALIZATION FACTOR button on your
• RxHCC19, Breast and Other Cancer and Tumors, screen to learn more. When you are finished, select the
and NEXT button to continue.

• RxHCC97, Immune Disorders.

To identify the relative factor for RxHCC19, we must

capture the value displayed in the cell intersecting the
“Community, Non-Low Income, Age ≥ 65” column and
row “RxHCC19,” which is 0.110.

To identify the relative factor for RxHCC97, we must

capture the value displayed in the cell intersecting the
“Community, Non-Low Income, Age ≥ 65” column and
row “RxHCC97,” which is 0.782.

Since there are no disease or disabled/disease interac-

tions for this beneficiary, to compute the sum of Mr.
Daniels’ disease relative factors, simply add the rela-

3-41
 Normalization Factor

Great job! You have gathered all the values needed

The normalization factor adjusts for population and to complete the fourth and final step of the risk score
coding changes between the denominator year and the calculation process.
payment year.
Select the NEXT button to continue.
To locate the normalization factor information for
calculating Mr. Daniels’ 2022 risk score, you will need to Note #2: Scenario 4
reference the 2022 Rate Announcement.

Select the SHOW ME button on your screen to learn

more.

While both the CMS-HCC and the RxHCC models use

health status or diagnoses to predict expenditures, the
total expenditures that each model is predicting are
quite different. The CMS-HCC model predicts medi-
Open the 2022 Rate Announcement by selecting the cal costs while the RxHCC model predicts prescription
RESOURCES button below and navigate to the “2022 drug costs.
RxHCC Normalization Factors” section. The normaliza-
Therefore, depending on the model you are using, you
tion factor listed is 1.043. This is the value we will use
may find different weights on similar HCCs, as well as
when computing Mr. Daniels’ risk score.
different risk scores for an individual beneficiary when
Select the NEXT button to continue. calculating their payment risk scores under different
models.

3-42
 Step 4: Using the Risk Score Formula 1. Calculate the Raw Risk Score

It is finally time to calculate Mr. Daniels’ risk score! The first step is to compute Mr. Daniels’ raw risk score
by adding the sum of his demographic factors, which
In this step, we will enter the values generated in Step
is 0.196, to the sum of disease factors, which is 0.892.
3 — the sum of demographic relative factors, sum of
disease relative factors, and the normalization factor — Mr. Daniels’ raw risk score is 1.088.
into the risk score formula to produce Mr. Daniels’ final
risk score for Payment Year 2022. 2. Calculate the Rounded
Normalized Risk Score
Select the SHOW ME button on your screen to learn
more.

Next, you must divide the raw risk score, 1.088, by

the 2022 normalization factor, which is 1.043, to get
When using the risk score formula, we must perform a normalized score of 1.04314. You must round this
several steps to accurately compute Mr. Daniels’ final number to the third decimal place to get a rounded
payment risk score. Select each button below to learn normalized score of 1.043.
more.
Since MA coding pattern adjustment is not applied to
When you are finished, select the NEXT button to Part D models, 1.043 is Mr. Daniels’ final Part D risk
continue. score for Payment Year 2022.

3-43
 Scenario 4 Conclusion

Congratulations! You have just completed the four-step

process of calculating Mr. Daniels’ Part D risk score.

From gathering the necessary reports and resources

and finding beneficiary demographic and disease infor-
mation to identifying relative and adjustment factors
and using the risk score formula, you now know how
to compute the 2022 payment risk score for a Part D
beneficiary.

There are three additional scenarios in this module. To

choose another scenario, simply select the CHOOSE
SCENARIO button displayed on your screen. Other-
wise, select the NEXT button to continue.

3-44
 Conclusion
In this course you learned about the history and process
of risk adjustment, the resources you’ll need, and how
to use the risk score formula to calculate or validate a
beneficiary’s risk score for payment year 2022.

If you have questions about the information covered in

this course, you may email us at
[email protected].

Congratulations!

You have just completed

the Risk Adjustment
Methodology course.