Section 3 - Disease Management

Requirements for a care management methodology to be valid

1. Familiarity- the purchaser must be familiar with the methodology, or at least be able to grasp it readily 2. Ease of replication and auditability- the methodology must be documented in sufficient detail for another practitioner to replicate the analysis 3. The results upon applying the methodology must be consistent with the client's savings expectations, and must be plausible 4. Results should be stable over time and between clients 5. The methodology must be practical (possible to implement it cost-effectively) 6. Inherent validity - lack of obvious bias 7. Scientific rigor 8. Market acceptance- how the method is perceived in the market 9. Application- how the methodology is applied in practice

Considerations when using claims data for evaluating disease management programs

1. Fixed time periods- a one-year time period to be too short for outcomes evaluation 2. Enrollment issues / eligibility- the timeliness of enrollment and disenrollment should be factored into the study 3. Claims run-out- due to claims lag, program results may not be known for up to two years after the program begins 4. Outlier claims- these may distort the study's results 5. Special problems with claims data- when using claims data to identify chronic members, some members are miscategorized (false positives and false negatives)

External sources of clinical identification algorithms

1. HEDIS (from the NCQA) has algorithms for identifying some conditions (eg, asthma, high blood pressure, and diabetes) 2. Disease Management Association of America developed algorithms for identifying chronic diseases 3. Grouper models - commercially-available models that identify member conditions and score them for relative risk and cost 4. Literature- articles will sometimes report the codes that are used for analysis

Considerations when evaluating results of disease management studies

1. Has the measurement been performed according to a valid methodology? 2. How has that methodology been applied in practice? 3. Are the results arithmetically correct?

Steps of the data warehousing process

1. Identify which patients to include in the dataset 2. Identify which data elements to merge with the patient list 3. Identify what the data says about the patient (eg, create flags that describe the patient's health and risk status) 4. Attach the derived variables and flags to the patient identifiers to create a picture of the patient history

Information needed in order to pay claims

1. Identity of policyholder and claimant- to detem1ine eligibility 2. Proof of loss/claim- according to the contract definition of loss 3. Date of loss- to determine that the claim occurred during the contract period 4. Amount of loss- according to the contract 5. Beneficiary or assignee 6. Information regarding other coverage- for coordination of benefits

Risk factors that indicate whether a person may have high claims

1. Inherent risk factors, such as age, sex, and race 2. Medical condition-related factors, such as diabetes or cancer 3. Family history (for conditions that are inheritable) 4. Lifestyle risk factors, such as smoking, lack of exercise, and poor nutrition 5. External risk factors, such as industry, location, and education

Types of predictive models that are not based on medical conditions

(These are the models actuaries have traditionally used, and are referred to as "non-condition risk based" models) 1. Age/sex - rates are established for a group based on the average age/sex factor of the members in the group 2. Prior cost - the prior year's claims are used to project future costs (isn't very accurate for smaller groups) 3. Combination of age/sex and prior cost- often used for rating smaller groups

Risk factors identified by a health risk assessment

(referred to as "variables collected in HRAs" in chapter 3) 1. Personal disease history 2. Family disease history 3. Health screenings and immunizations 4. Alcohol consumption 5. Injury prevention behavior 6. Nutrition 7. Physical activity 8. Skin protection 9. Stress and well-being 10. Tobacco use 11. Weight management 12. Women's health (eg, pregnancy status) 13. General health assessment 14. Functional health status 15. Mental health status (the last three items came from a similar list in Duncan (risk adj) chapter 3)

Re-sampling methods for validating a model

(these approaches help test the model's predictive power) 1. Bootstrap - the sampling distribution of an estimator is estimated by sampling with replacement from an original sample 2. Jackknife- the estimate of a statistic is systematically re-computed, leaving out one observation at a time from the sample set 3. Cross-validation- subsets of data are held out for use as validating sets 4. Permutation test- a reference distribution is obtained by calculating all possible values of the test statistic under rearrangement of the labels on the observed data points

Components of the Risk Management Economic Model

(these are the factors that contribute to the financial outcomes of the program) 1. Prevalence of different chronic diseases 2. The cost of the chronic disease 3. Payer risk- the most savings for the plan will come when the plan is at financial risk for all of the patient's costs 4. Targeting and risk-members should be prioritized based on the probability of experiencing the targeted event. Those with the highest risk rallies will be selected for the program. 5. Estimated cost of the targeted event 6. Contact rate- the rate at which the company is able to make contact with targeted members 7. Engagement (or enrollment) rate , 8. Member re-stratification rates- the initial1isk rank of the member will be re-stratified after the nurse interacts with the member and assesses the member's risk

Risk factors for care management studies

(these should be reported with the study, to ensure reproducibility of results) 1. Demographic variables 2. Exclusionary conditions that exclude certain members- such as conditions that imply the member is not a good candidate for care management 3. Exclusionary conditions that exclude certain claims- exclude claims for conditions that disease management does not try to affect (eg, maternity) 4. Persistency- understand the terms under which a member may enter or leave the group 5. Chronic prevalence and risk classification - chronic prevalence is defined as the percentage of individuals in a population with the condition 6. Severity of illness- severity affects claims cost, and therefore the potential for savings 7. Contactability- this measures whether the manager is able to reach and engage the member 8. Operational issues- such as the number of eligible members; the number of chronic patients identified, contacted, and enrolled; the graduation rates; and the methodologies used

Reasons for using commercially-available grouper models

(why these models may be preferred over building your own clinical identification algorithm) 1. Building algorithms from scratch requires a considerable amount of work 2. Models must be maintained to accommodate new codes, which requires even more work 3. Commercially-available models are accessible to many users. Providers and plans often require that payments be based on a model that is available for review and validation.

Common features of Medicare prospective payment systems

1. A system of averages -providers cannot expect to make a profit on each case, but efficient providers can make a reasonable return on average 2. Increased complexity- DRGs are more complicated than a system based on per diem payments 3. Relative weights - associated with each patient group to reflect the average resources used by efficient providers 4. Conversion factor (base price)- the dollar amount for a unit of services is multiplied by the relative weight to determine payment. 5. Outliers - unusual cases that require above-average resources and receive extra payments 6. Updates- the conversion factor and relative weights are adjusted annually to reflect new technologies and changing practice patterns 7. Access and quality-policymakers monitor PPSs and survey patients to ensure that beneficiaries have adequate access to high quality care and that providers are compensated adequately

Legislation that limits the ability to manage care

1. Any willing provider laws- these prevent MCOs from selectively contracting with a limited group of providers 2. Physician antitrust exemptions -legislative efforts have attempted to exempt physicians from antitrust laws, allowing them to jointly negotiate with health insurance plans 3. Expanding health plan legal liability - some states have enacted laws exposing health plans to liability for their coverage determinations and utilization management activities 4. Mandated benefits -these laws mandate coverage for specific conditions. Mandates cause plans to raise premiums and sometimes have harmful, unintended consequences.

Claim process for health insurance

1. Benefit eligibility and proof of loss a) Need bills with dates and illnesses to verify the loss b) To determine eligibility, consider the coverage limits, which expenses are eligible, and any pre-existing conditions 2. Determine eligible charges- usually the billed charges subject to usual and customary limits or a managed care schedule 3. Determine the gross benefit level- apply the deductible, coinsurance, out-of-pocket limits, and max1mums 4. Determine the net payment- consider coordination of benefits (COB), subrogation, and assignment of benefits

Types of medical management interventions

1. Care coordination (focuses on the system) - includes case management, discharge planning, and in-hospital care coordination 2. Condition management (focuses on the patient) - includes disease management and risk factor management 3. Provider management (focuses on the provider) - includes provider profiling, pay-for-performance, and accountable care organizations

Uses of claim information

1. Claim adjudication 2. Financial reporting 3. Pricing 4. Reserving 5. Provider monitoring 6. Fraud and abuse control

Actions the insurer may take after evaluating a claim

1. Claim is approved and payment is made 2. Claim is disapproved- the company will generally provide a reason for denial and the claimant will then have a chance to appeal 3. Claim examiner requires more infonnation in order to make a determination 4. Contract is rescinded because of misrepresentation- this occurs when the insurer discovers intentional misinformation on the application and when policy provisions allow for rescission

Sources of data for developing risk factors

1. Claims data- for medical condition-related risk factors such as diabetes or cancer 2. Self-reported data- for lifestyle-related risk factors such as smoking, stress, lack of exercise, poor nutrition, etc. (see separate list of risk factors identified by a health risk assessment) 3. External data- for lifestyle-related risk factors such as industry, geography, education, and income level

Areas reviewed in market conduct exams

1. Company operations and management 2. Complaint handling 3. Marketing and sales 4. Producer licensing 5. Policyholder service 6. Underwriting 7. Claims

Types of methodologies for estimating care management savings

1. Control group methods- these attempt to match the study subjects with other subjects that are not part of the study (see separate list) 2. Non-control group methods- population methods that do not use control groups (see separate list) 3. Statistical methods- these use purely statistical techniques, rather than constructing an explicit reference population (see separate list)

Factors for choosing the right model

1. Correlation structure- more complicated models may be needed for data containing correlated variables 2. Purposes of the analysis 3. The nature of the available data 4. Characteristics of the outcome variable (eg, quantitative vs. qualitative, unrestricted vs. truncated, binary choice vs. unrestricted choice) 5. Distribution of the outcome variable (eg, normal vs. skewed) 6. Functional relationship (eg, linear vs. non-linear)- when the equation cannot be transformed into a linear form, iterative processes or a maximum likelihood procedure may be used instead of ordinary regression methods 7. Complex decision model- whether a single equation model is sufficient or a simultaneous equation model is needed (if there is more than one dependent variable)

Components of the disease management value chain process

1. Data warehousing- integrate membership and claims data, and identify member conditions 2. Predictive modeling- apply models to determine members to target for interventions 3. Intervention development- develop campaigns to deliver interventions to target populations 4. Outreach and enrollment- contact members and enroll them in the program. Includes follow-up. 5. Member coaching and assessment-including maintaining enrollment and graduating members from the program 6. Outcomes assessment- including clinical, financial, and operational outcomes

Factors complicating the comparison of financial savings for care management programs

1. Different research designs are used by different studies 2. The basis of savings calculations varies (some report an ROI, which is difficult to compare across studies, and many do not provide information on the cost of the program) 3. The timing of the studies distorts numbers, since health care costs increase over time 4. Different studies use different population sizes and durations 5. Many of the published studies focus on clinical, rather than financial, outcomes

Conditions that would exclude a member from a disease management program

1. End-stage renal disease (ESRD)- this condition is excluded because management of the condition may delay cost, but it cannot ultimately reduce or postpone those costs 2. Transplants- claims are high up to a period shortly after the transplant, at which point the claims are reduced and stabilized 3. HIV, AIDS, mental health- privacy issues make it difficult or impossible for a vendor to receive complete data feeds, or manage the member 4. Members who are institutionalized- these members may not be reachable, or may not benefit from disease management interventions 5. Members with catastrophic claims- these members are not manageable by the DM program, and are often subject to management by another program 6. Members who are eligible for other management programs

Central features of Massachusetts health care reform

1. Establishment of an exchange (purchasing pool) 2. A requirement that all employers establish Section 125 accounts (so employees could pay premiums on a pre-tax basis) 3. Large subsidies for families living below 300% of FPL 4. For those above 300% of FPL, availability of a more limited plan (so insurance would be affordable even outside the subsidy range) 5. A mandate that all individuals must purchase health insurance coverage 6. Funding through use of federal funds previously paid to safety net hospitals or paid for uncompensated care

Tools used for validating that a claimant is disabled

1. Examination by company physicians - the insurer may require that a physician of its choice examine the claimant, to validate the findings of the claimant's physician 2. Telephone validations- call the claimant and discuss the claim 3. Field examinations - a field investigator tries to discover if the claimant is performing tasks that he or she is supposedly unable to perform

Claim process for disability insurance

1. Initial determination of the potential liability (assess the claimant's condition relative to the contractual definition of disability) 2. Establish the disability status of the claimant - May involve an assessment by the claimant's physician, independent medical examination, or even field investigations (surveillance) 3. The insurer establishes a plan for managing the disability (timeline for recovery or rehabilitation program) 4. Determine the level of periodic benefit payments (look at pre-disability salary and offsets) 5. Ongoing review of the disability to assure payments are made only if the claimant remains disabled and to manage long-term disabilities (may try to re-train the claimant for another activity or consider a lump sum payment to settle the claim)

Common chronic diseases addressed by disease management programs

1. Ischemic heart disease 2. Heart failure 3. Chronic obstructive pulmonary disease 4. Asthma 5. Diabetes

Challenges with patient classification systems based on coding systems

1. Need for new DRGs- due to new diseases and new procedures 2. ICD coding- some codes may not be sufficiently precise as diseases and procedures are refined 3. Upcoding- providers may be tempted to exaggerate a patient's secondary diagnoses to get paid more 4. New coding systems- adopting the new ICD-10 systems will be a major challenge for hospitals and CMS

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

1. Once contracted, the disease remains with the patient for the rest of the patient's life 2. The disease is often manageable with a combination of pharmaceutical therapy and lifestyle change 3. Patients can take responsibility for their own conditions 4. The average annual cost is sufficiently high to warrant spending resources to manage the condition

Characteristics for assessing the quality of a model

1. Parsimony- should introduce as few variables as are necessary to produce the desired results 2. Identifiability- if there are more dependent variables than independent equations, then issues such as bias will result 3. Goodness of fit- va1iations in the outcomes variable should be explained to a high degree by the explanatory variables (measured by R2 and other statistics) 4. Theoretical consistency- results should be consistent with the analyst's prior knowledge of the relationships between variables 5. Predictive power- should predict well when applied to data that was not used in building the model

Principles for establishing a patient-centered medical home

1. Personal physician- each patient has a personal physician trained to provide comprehensive care 2. Physician-directed medical practice- consists of a team of individuals taking responsibility for the patient's ongoing care 3. Whole person orientation- appropriately arranging care with other qualified professionals 4. Care coordinated and integrated across all elements of the health care system and the patient's community 5. Quality and safety- includes patient-centered outcomes, evidence-based medicine, and continuous quality improvement 6. Enhanced access through open scheduling, expanded hours, and E-visits 7. Reimbursement structure to support and encourage this model of care

Types of data sources for predictive modeling

1. Physician referral/chart (high reliability, low practicality) - medical charts provide the most information, but have serious drawbacks (see separate list) 2. Enrollment (high reliability, high practicality) -can be used to convert claims data into PMPM amounts 3. Claims (medium reliability, high practicality) -usually available to health plans and continually refreshed as events occur. Data quality varies greatly (must check for accuracy). Lots of information is provided in claim forms for hospital (UB04) and professional (CMS 1500) claims. 4. Pharmacy (medium reliability, high practicality) - high quality data that completes quickly. But there is no diagnosis on the claims and prescriptions that aren't filled won't generate claims. 5. Laboratory values (high reliability, low practicality)- can be difficult to obtain, and vendors do not use a standard format 6. Self-reported (low/medium reliability, low practicality) - will become important since members can report information that isn't available elsewhere, but there are drawbacks (see separate list)

Types of care management methods

1. Pre-authorization- requires a provider to obtain approval before performing a service 2. Concurrent review- involves monitoring a health plan member's care while the member is still receiving care in a hospital or nursing home 3. Case management- typically involves a health care professional who coordinates the care of a patient with a serious disease or illness (such as stroke, AIDS, or cancer) 4. Demand management- refers to certain passive forms of informational intervention, often provided over the telephone. Includes nurse advice lines and shared decision making. 5. Disease management- focuses on chronic conditions with certain characteristics that make them suitable for clinical intervention (see separate list for these characteristics) 6. Specialty case management- a care manager who has expertise in a particular area coordinates care for patients in that area 7. Population health management- the entire membership of a health plan is evaluated, using statistical tools to identify potential high-cost patients who can benefit from some type of voluntary intervention program 8. Medical Home- this model returns to the physician the responsibility for coordinating all of the patient's care

Areas where condition-based models are used in healthcare financial applications

1. Program management- identifying and high-risk individuals, financial modeling and resource allocation, and program evaluation (eg, calculating savings) 2. Provider or health plan reimbursement-normalizing populations to pay providers or plans for the risks they accept and to evaluate provider effectiveness. Profiling providers to assess quality and efficiency. 3. Actuarial and underwriting functions- pricing health plans, underwriting groups, and projecting future claim costs

Statistics for determining whether a model is good

1. R2 - measures how much of the variation in the dependent variable is explained by the variation in the independent variables. A more valid measure may be Adjusted R2 = 1 - (1 - R2 * (N- 1) /(N - k - 1)), where N =number of observations and k =number of parameters. 2. Regression coefficients - examine the signs of the parameter estimates to ensure they make sense, then determine whether the value of the parameter estimate is statistically significant 3. F-Test- ratio of variance explained by the model divided by unexplained or error variance 4. Statistics used for logistic models: a) Hosmer-Lemeshow statistic b) Somers' D statistic c) C-statistic 5. Multicollinearity- occurs when a linear relationship exists between the independent variables. May be addressed by removing one of the collinear variables. 6. Heteroscedasticity- occurs when the error terms do not have a constant variance 7. Autocorrelation- occurs when there is correlation in the error term in the regression function

Control group methods for estimating care management savings

1. Randomized- compares equivalent samples drawn randomly from the same population (the preferred method) 2. Geographic- compares equivalent populations in two different locations 3. Temporal- comparing equivalent samples drawn from the same population before and after the intervention program 4. The product control methodology- compares samples drawn from the same population at the same point in time, but differentiates between members who have different products 5. "Patient as their own control"- patients are used as their own control group 6. Participant vs. nonparticipant studies- the experience of those who voluntarily participate is compared to the experience of those who choose not to participate (has selection bias)

Components of the claim processing workflow

1. Receipt processing-once it is received, the claim is logged into the system and given a unique identifier 2. Optical character recognition (OCR) - paper claims are scanned and translated to electronic data using OCR technology 3. Repair- claims that do not contain all of the necessary information are rejected and require manual intervention 4. Auto-adjudication- if all of the necessary fields are present, the system will adjudicate the claim automatically 5. The payment process- some plans transmit funds electronically, while others send paper checks

Principles for measuring results of care management programs

1. Reference population- any outcome's measurement requires a reference population against which to evaluate the statistics of interest 2. Equivalence- the reference population should be equivalent to the intervention population 3. Consistent statistics -the same statistic should be measured in the same way in the reference and intervention populations 4. Appropriate measurement- avoid (if possible) extraneous, irrelevant, or confounding variables 5. Exposure- the exposure group must be clearly defined and all members who meet the definition should be included in the appropriate group 6. Reconcile the results- reconcile the outcomes of a small population with those of the entire health plan ("plausibility analysis")

Issues that affect disease management evaluations for chronic populations

1. Regression to the mean- a high percentage of high-cost patients in one period will not be high cost in the next period, simply because the high-cost event was a one-time event that is not likely to be repeated 2. Identifying patients- due to regression to the mean, it may not be appropriate to use patients' past data as the comparison group. A common alternative is to use the population approach (uses the entire population). 3. Establishing uniform risk measure for comparability- objective, consistent definitions should be used to identify candidates for the care management program (this will ensure equivalence) 4. Patient selection bias- this results when a study is based on those volunteering for a program 5. Patient drop outs- drop opts may also create a bias (eg, those feeling better may drop out) 6. General versus specific population- some interventions are performed on an extremely small population, making some methodologies inappropriate for measuring results

Financial measures for disease management programs

1. Retum on investment- this is the most common metric. DM programs typically use Gross ROI. a) Net ROI = (gross savings - cost) / cost b) Gross ROI = gross savings / cost 2. Total savings- this metric may be more useful, since it represents the dollar savings for the plan a) Average savings equals total savings net of program cost, divided by the total population b) Marginal savings per chronic member equals the increase in savings (net of costs) due to intervention on the marginal population, divided by the number of members in the marginal population

Formulas for calculating disease management program savings

1. Savings= [ChrUtilprior yr * (1 +trend) - ChrUtilactual] * Chronic members * Cost per service ChrUtil is the utilization rate per chronic member The trend rate comes from the non-chronic population of the health plan 2. Savings PMPM = Savings / Member months

Non-control group methods for estimating care management savings

1. Services avoided methods - savings are calculated as the estimated cost of a service requested through pre-authorization minus the actual cost after the intervention 2. Clinical improvement methods- the change in a clinical measure is observed and the resulting improved health and reduced utilization is estimated from outside studies

Regulators of MCOs

1. State insurance and health departments 2. US Department of Labor- regulates employer plans that fall under ERISA 3. US Department of Health and Human Services (through CMS)- oversees the Medicare and Medicaid programs, and the State Children's Health Insurance Program (SCRIP) 4. State Medicaid departments regulate Medicaid programs (in conjunction with CMS) 5. US Office of Personnel Management- regulates coverage offered to federal employees 6. US Department of Defense- regulates coverage delivered through its military health services system

Drawbacks of using survey data

1. Surveys must be commissioned, budgeted, and executed in order to generate the data 2. Data isn't updated as medical events occur, so it can become stale unless the survey is updated periodically 3. Response bias can make it dangerous to draw conclusions from survey responses 4. Respondents may submit untruthful answers

Areas where actuaries can be involved with care management programs

1. The economics of care management programs- help with understanding the relationship between care management program inputs and outputs 2. Risk adjustment and predictive modeling a) Predictive modeling is used to identify candidates for intervention programs b) Risk adjustment is used to assess outcomes 3. Financial outcomes evaluation- help in achieving comparability between the reference and the intervention population

Principles for designing care management programs

1. The intervention population should be chosen carefully 2. The economics of the program should be analyzed carefully (since interventions can be costly) 3. The objectives of a program should be clearly defined, and the program should be designed to achieve those goals 4. Interventions require the active participation of both providers and patients 5. Financial savings may take a long time to emerge

Challenges when constructing a condition-based model

1. The large number of procedure and drug codes 2. Deciding the severity level at which to recognize the condition 3. The impact of co-morbidities for conditions that are often found together 4. The degree of certainty with which the diagnosis has been identified 5. The extent of the data (claims data will cover all members, but self-reported data will not) 6. The type of benefit design that underlies the data

Key metrics in the design of disease management programs

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

Drawbacks of using data from medical charts

1. They do not cover out-of-network services or drugs prescribed by out-of-network providers 2. They do not record the patient's compliance with physician orders (such as prescription filling) 3. Transcribing the data and transferring it to a uniform format is time consuming and requires highly-trained staff 4. There is not uniformity in how physicians code conditions and their severity 5. Charts are typically unavailable to the health plan or the actuary

Statistical methods for estimating care management savings

1. Time-series methods- a curve is fit to the data over time and a divergence from this best-fit line can be observed once the intervention is applied 2. Regression discontinuity- a line is fitted to data that relates pre- and post-intervention experience. A dummy variable is used to test whether the intervention produced a change. For example: Yi = B0 + B1Xi + B2Zi + ei, where: a) X= independent variable (year 1 cost) b) Y = dependent variable (year 2 cost) c) Z = dummy variable for the intervention in question d) B = regression coefficients and e = random error 3. Benchmark methods - the values of certain key statistics are compared between the population being managed and some benchmark population

Claim types commonly excluded from disease management program evaluations

1. Trauma and accident 2. Behavioral and substance abuse 3. Malignant neoplasms 4. Maternity and childbirth claims 5. Pharmaceutical drugs

Questions to answer when building a clinical identification algorithm

A clinical identification algorithm is a set of rules that is applied to a claims data set to identify the conditions present in the population 1. Where are the diagnoses? 2. What is the source of the diagnosis (claims, medical charts, etc.)? 3. If the source is claims, what claims should be considered (inpatient, outpatient, laboratory, etc.)? 4. If the claim contains more than one diagnosis, how many diagnoses will be considered for identification? 5. Over what time span, and how often, will a diagnosis have to appear in claims for that diagnosis to be incorporated? 6. What procedures may be useful for determining severity of a diagnosis? 7. What prescription drugs may be used to identify conditions?

Advantages and disadvantages of using diagnosis codes for identifying member conditions

Advantages 1. Codes are almost always present on medical claims 2. A uniform format exists 3. Usefulness for identifying conditions Disadvantages 1. Usually only the primary and secondary codes are populated in the claims data 2. Coding errors may occur 3. Codes may sometimes be selected to drive maximum reimbursement 4. Different physicians may follow different coding practices

Factors used in developing risk scores in the CMS-HCC risk model

HCC = hierarchical condition category 1. Demographics- age and gender factors are the starting point. Higher risk scores are assigned to beneficiaries who are eligible for both Medicaid and Medicare. 2. Disabled indicators- a separate set of age and gender factors are used for beneficiaries under age 65 who are eligible for Medicare due to disability 3. Separate models are used for beneficiaries who: a) Reside in a long-term care institution, or b) Suffer from end-stage renal disease 4. New enrollees- since no claim history exists, only age and gender factors are used. Separate factors are developed specifically for new enrollees. 5. A prospective risk adjustment methodology is used to risk-adjust future payments based on actual historical medical experience 6. Calibration- every two years, CMS re-calibrates by updating the model weights to reflect new prescription drugs and changes in medical technologies, practice patterns, and provider coding practices 7. Health status risk factors are developed based on the beneficiary's diseases (using ICD-9 diagnosis codes and grouping into HCCs)

Definitions of sensitivity and specificity

When building clinical identification algorithms, the proper balance between sensitivity and specificity must be found 1. Sensitivity- the percentage of members correctly identified as having a condition ("true positives") 2. Specificity- the percentage of members correctly identified as not having a condition ("true negatives")