Validation and Reconciliation of MCP Data

©️ 2024 Kaiser Foundation Health Plan, Inc.

This resource is part of the Data Quality & Reporting Implementation Guide, offering steps and activities to ensure your practice is capable of reporting valid and reliable data for selected population health measures. It is the first in the “Building the Foundation” series of implementation guides.

A critical step of data validation is reconciliation with rates produced through other sources, such as managed care plan (MCP) rates, and is critical to the PHMI process for the following reasons:

• MCPs are producing similar rates through their pay-for-performance (P4P) programs for an overlapping subset of measures that offer an important comparison opportunity, and
• MCP P4P programs hold financial value for community health centers (CHCs).
  • The reconciliation of these rates ensures the MCP P4P rates are an accurate reflection of the care provided and outcomes achieved by CHCs.
  • CHCs have accurate and reliable rates for the pursuit of quality improvement.

Through the Population Health Management Initiative (PHMI), CHCs should have a validation process in place to ensure the accuracy of measurement and reporting for each quarterly submission of core HEDIS measures for PHMI.

This includes:

• Initial validation process to compare health center-produced PHMI/HEDIS measures with MCP P4P reports where PHMI/HEDIS measures are used.
• Development of a process for ongoing reconciliation of health center and managed care plan data.

This document defines an approach for the data validation and reconciliation of MCP data related to the seven core HEDIS measures for PHMI. It includes methods to identify and explain variances and discrepancies between CHC-reported metrics compared to those same metrics calculated by the MCPs. This process also includes steps for reconciliation of the CHC and MCP rates for each measure. It should be used in alignment with the Data Quality and Reporting Resource 7: Data Validation Process.

This document is limited to comparison of the CHC and MCP measurement rates. As the MCP rates are tied to payment, CHCs have a vested interest in ensuring the rates are accurate and high performing. Reconciliation improves the overall quality of the CHC rates, reduces the likelihood of under/overpayments and ensures compliance with PHMI goals.

MCP P4P programs are unique and may contain different measures than the core HEDIS measures for PHMI, however there is overlap between almost all measures and each MCP. For the measures that overlap, CHC rates are an opportunity to cross-check the validity of MCP rates and ensure optimal performance.

Working with practice coaches/subject matter experts (SMEs), follow this general process for validation and reconciliation of MCP data.

Steps for Validation:

Step 1: Assess overlap in MCP and core HEDIS measures for PHMI.

Determine which measures are used by MCPs and to what extent they overlap with core HEDIS measures for PHMI. The remaining steps can be completed for all overlapping measures.

Step 2: Create MCP-specific rate.

CHCs will have to separate their core HEDIS measures for PHMI rates into a separate rate for each MCP with which they contract to conduct the analyses and processes listed below independently for each MCP.

Step 3: Determine how MCPs calculate their rates.

To compare and analyze PHMI and MCP measure rates, it is important to know the steps, sources and criteria that the MCPs used to calculate their rates. The CHC can use this information to narrow the areas of focus, thereby making the comparison more efficient.

• Obtain information on the specifications used to determine the MCP rates (e.g., in P4P manuals).
• Identify any variation between MCP and PHMI calculation methodologies, for example:
  • Differences in measure specifications (e.g., HEDIS vs. Center for Medicare and Medicaid Services (CMS) quality measure specifications; HEDIS administrative vs. HEDIS Electronic Clinical Data Systems (ECDS) specifications).
  • Time frames for the reporting period (e.g., whether quarterly data are yearto-date or based on a rolling year) and for the data calculation period (e.g., quarterly data provided one month after the close of the quarter versus three months after the close of the quarter would be different due to claims/encounter lag).
  • Differences in methodology for determining the eligible population (e.g., if the MCP uses continuous enrollment criteria, or if the MCP uses an algorithm for the P4P-attributed population that differs from the member population attributed to the CHC).

Step 4: Review PHMI and MCP rates (including eligible populations and by race and ethnicity sub-populations, if available) and analyze variances.

Begin with a side-by-side comparison of HEDIS measure for PHMI rates with those of the MCPs. Note those where there are variances or differences in the rates and investigate potential causes.

• Obtain data from MCP to facilitate reconciliation such as patient-level detail or gaps in care reports. If the MCP does not already provide these (e.g., through a provider portal or other application), the CHC should engage the MCP to identify if patient-level detail can be provided.
• CHCs should use the reconciliation processes and the checklist below to reconcile each measure.

Step 5: Analyze how gaps can be remediated and establish ongoing processes to ensure alignment.

Based on the gaps identified, CHCs will work with their practice coaches and the data quality and reporting SMEs to develop an action plan to remediate gaps and establish ongoing processes.

This action plan could include a process for:

• Sending documentation that was provided but not captured in the MCP rate to the MCP capturing the care.
• Obtaining data for care provided by other providers from the MCP (or from the primary source).
• Resolving discrepancies or missing data related to race and ethnicity.
• Resolving coding issues that may have contributed to the MCP not being able to capture care that was provided (see Data Quality and Reporting Resource 5: Documentation and Coding Playbook).
• Ensuring fidelity to PHMI/HEDIS specifications to better align with MCP rates (see the Data Quality and Reporting Resource 3: Measure Calculation and Reporting for PHMI).
• Ensuring ongoing sustainability: Policies and procedures to ensure ongoing reconciliation of PHMI and MCP rates, including thresholds, cadence, process and remediation.

To validate and reconcile PHMI rates against MCP rates, consider population-level checks that are applicable to all measures, as well as measure-specific steps that are individual to each measure.

Steps for Reconciliation:

Step 1: Validating reasonability.

Apply certain basic validation checks to each measure to assess for variation beyond what is reasonable or easily explainable. When identifying the reasonability of differences in populations, sub-populations, numerators and denominators, CHCs should consider what they know. What have they learned about the variation in the HEDIS measure for PHMI rate and the MCP measure rate that might explain variances in the rates?

Validation checks should include examination of the following:

1) Denominator: PHMI denominators should be based on the attributed member population from the MCP.

• There should be very limited variation.
• Variations should be explained by acceptable causes such as:
  • Continuous enrollment.
  • The specific “anchor” date of attribution.
  • Methodology for P4P attribution vs. attributed member population.

2) Numerator: CHCs might anticipate more variability in numerators based on the sophistication and completeness of data capture at the PHMI and MCP. For example:

• Measures that are solely primary care-based (e.g., well child visits) should have significant alignment with MCP data. The data needed to calculate the measure are reasonably expected to be internal to the CHC. If data are not aligned, variations could be due to causes needing remediation, such as:
  • Provided services are not properly coded.
  • MCP attribution and/or PCP assignment is inaccurate, and the patient is seeking care elsewhere.
• Measures that rely on specialists or providers outside the CHC (e.g., colorectal cancer screening) may show higher degrees of variation based on a CHC’s sophistication in capturing referral information, medical history, etc. Variations could be due to causes needing remediation such as:
  • External care data is not consistently or accurately captured.

Step 2: Conducting primary source verification.

Primary Source Verification (PSV), a common best practice process to validate the accuracy of data and identify gaps, is described in detail in PHMI DQR Tool 6A: Data Validation Process.

PSV involves:

• Using the patient-level data file for a measure such as a file that identifies all patients in the denominator.
• Including a flag indicating whether they were numerator-compliant.
• Tracing that patient back to their primary source of data (e.g., the medical record) to ensure that the documentation supports the patient having been included in the measure and whether they are compliant.

When specific to the reconciliation of PHMI and MCP rates PSV should include:

1) Use member/patient-level files to identify the variation in MCP and PHMI rates, including:

• Individuals in compliance and/or in the denominator of the MCP rate, but not the PHMI rate.
• Individuals in compliance and/or in the denominator of the PHMI rate, but not the MCP rate.
• Individuals with missing or discrepant race and ethnicity data.

2) Review each discrepancy using the primary source of the data:

• Trace back to the CHC’s primary source to identify appropriate capture of all data.
• Work with the MCP to identify the primary source for appropriate capture of all data.

Note: If conducting PSV on all discrepancies is not feasible due to the number of discrepancies detected, CHCs could consider a smaller random sample. It is advisable to continue PSV at least until strong patterns are detected in the discrepancies (e.g., if a high percentage of discrepancies found in a reconciliation are caused by the same variation reason(s), the CHC would have strong evidence for pursuing a remediation of that variation). If many reasons for variation are found, the CHC likely needs more PSV to determine the universe of variation causes.

Step 3: Determining need for ongoing reconciliation in future reporting cycles.

Reconciliation and validation should occur before the initial submission of core HEDIS measure for PHMI rates, as well as any time there is a change to the specifications, methodologies or data sources. The initial analysis of validation/reconciliation with MCP rates can be used to determine how much data discrepancy is expected.

Through PSV, the CHC can determine the percentage of the variance that was based on differences in the way the measure was calculated versus the percentage variance that was based on inconsistent documentation of diagnoses or services for a patient in the initial validation. The percentage based on measure calculation differences is a reasonable variation. Going forward, the CHC could use this “reasonable” variation as a threshold to determine the need for ongoing reconciliation in future reporting cycles.

Reconciliation Example

The CHC detects a 15% difference in their HEDIS measure for PHMI for controlling high blood pressure and the MCP rate for the same measure.

The CHC analyzes the discrepancy and discovers that:

• 3% is attributable to claims/encounter lag.
• 2% is attributable to differences in the eligible population based on the MCP applying continuous enrollment criteria.
• 10% is based on missing data for the MCP due to inadequate usage of Current Procedural Terminology Category II (CPT II) codes (supplemental performance tracking codes) and/or lack of medical record documentation to support the care provided coming from the CHC.

The 5% (3% + 2%) is a reasonable difference based on valid measure specification differences and a lag in data that would have been reconciled, once received.

The 10% based on missing data is not reasonable, as it reflects an inaccurate rate and missed opportunity to achieve a high-performing measure.

In this case, the CHC could assume the 5% variation in rates is a reasonable threshold for determining if reconciliation is needed. In future reporting cycles, if the PHMI rate and MCP rate have less than or equal to 5% variance, the CHC would not need to reconcile the rate for that reporting cycle.

MCP Measure Reconciliation Checklist

With assistance from their practice coach, reporting SME and data analytics, CHCs should determine a process to calculate the MCP rates and flag instances where MCPs are not calculating measures according to the same specifications as CHCs. When differences are detected, CHCs should explore the impacts of those differences on each measure. For example, if MCPs are using a different measure specification, the CHC should examine how similar the diagnosis codes used to identify a population for a measure are.

This process, combined with assessing the MCP and HEDIS measure for PHMI rates for reasonable alignment and conducting PSV where discrepancies are detected, provides a basis for validating and reconciling each measure. Use the below table as a checklist to reconcile against MCP rates for the same measure.

FIGURE 8.1: MCP MEASURE RECONCILIATION CHECKLIST