PCP Performance Analysis Case Study

Client question

A client wanted to evaluate primary care physician performance in a target market. The client had internal utilization and quality data, but its member counts for many PCPs were too small to support reliable provider-level conclusions.

The client needed a market-level view of PCP performance that could support network strategy, value-based contracting, provider engagement, and targeted quality improvement.

Claims and encounter data approach

FastHSR used Medicare claims and Medicare Advantage encounter files to construct PCP-level performance measures across a larger beneficiary base. The project required many years of large claims and encounter datasets, extensive programming, patient attribution, denominator construction, risk adjustment, cost measurement, utilization logic, and quality-measure specifications.

The result was a standardized PCP performance file that allowed the client to compare providers using larger, more stable measurement populations than the client could generate from its own members alone.

Population and attribution

The analytic file assigned beneficiaries to PCPs and calculated member months for medical and pharmacy coverage. Attribution and denominator logic were critical because PCP-level measures are sensitive to panel size, coverage months, eligibility, and the availability of Part A, Part B, and Part D data.

• Member months for Part A and Part B.
• Member months for Part A, Part B, and Part D.
• PCP panel construction across multiple years of claims and encounter files.
• Measure denominators designed to reduce noise from small samples and incomplete coverage.

Cost and risk measures

FastHSR produced PCP-level cost measures using Medicare payment fields and coverage-specific denominators. These measures supported comparisons of total cost, average cost, risk, and proxy financial performance.

• Part A Medicare payments.
• Part B Medicare payments.
• Part D Medicare payments.
• Total cost of care.
• Average cost of care.
• Average risk adjustment factor.
• Proxy medical loss ratio.

Utilization and care-management measures

Utilization measures helped the client understand where PCP panels differed in acute care use, emergency department use, post-acute care, readmissions, and preventive or transition-focused care.

• Annual wellness visits.
• Flu vaccine compliance.
• Transitional care management.
• Admissions per 1,000 beneficiaries.
• Inpatient days per stay.
• Emergency room visits per 1,000 beneficiaries.
• Avoidable emergency room visits per 1,000 beneficiaries.
• Skilled nursing facility stays per 1,000 beneficiaries.
• Skilled nursing facility days per stay.
• Readmission rate.
• Post-acute care spend per beneficiary.
• Skilled nursing facility, home health, and hospice spend.

NCQA quality measures

FastHSR also programmed NCQA-style quality measures at the PCP level. These measures required careful numerator, denominator, exclusion, timing, diagnosis, procedure, pharmacy, and post-discharge logic across large claims and encounter files.

• BCS Breast Cancer Screening.
• COL Colorectal Cancer Screening.
• OSW Osteoporosis Screening in Older Women.
• OMW Osteoporosis Management in Women who had a Fracture.
• EED Diabetes Care Eye Exam.
• MRP Medication Reconciliation Post-Discharge.
• PCR Plan All-Cause Readmissions.
• SPC Statin Therapy for Patients with Cardiovascular Disease.
• SPD Statin Therapy for Patients with Diabetes.
• Medication adherence for diabetes, hypertension, and cholesterol.

Why this required heavy programming

PCP performance analytics is not a simple claims pull. The work required processing many years of huge Medicare claims and MA encounter files, aligning beneficiary eligibility across coverage types, attributing members to PCPs, applying clinical and utilization logic, calculating risk and cost measures, and validating provider-level outputs for reliability.

The NCQA-style measures added another layer of complexity because each measure required its own denominator, numerator, exclusion, timing, and code-set logic. FastHSR's role was to turn raw claims and encounter data into a market-ready PCP performance dataset.

Deliverables

• PCP-level performance dataset for the target Medicare Advantage market.
• Member-month, cost, risk, proxy MLR, utilization, post-acute, readmission, and preventive-care measures.
• NCQA-style quality measures programmed from claims, encounter, and pharmacy data.
• Provider-level outputs designed to reduce small-sample instability.
• Documentation of measure definitions, denominator logic, and calculation rules.

Frequently asked questions

How can a client evaluate PCP performance when its own sample size is too small?

A client can supplement its internal view with Medicare claims and Medicare Advantage encounter files to build larger market-level PCP panels, then calculate standardized cost, utilization, risk, and quality measures.

What PCP performance measures can be created from Medicare claims and MA encounter data?

Measures can include member months, Part A payments, Part B payments, Part D payments, total cost of care, average cost of care, risk adjustment factors, proxy medical loss ratio, annual wellness visits, admissions, emergency room visits, readmissions, post-acute care spend, and NCQA quality measures.

Why does PCP-level Medicare Advantage performance analytics require heavy programming?

The work requires many years of large Medicare claims and MA encounter files, patient attribution, PCP panel construction, measure specifications, denominator logic, risk adjustment, quality-measure programming, and validation across cost, utilization, and quality outputs.

Which NCQA-style quality measures can support PCP performance analysis?

Measures can include breast cancer screening, colorectal cancer screening, osteoporosis screening and management, diabetes eye exam, medication reconciliation post-discharge, plan all-cause readmissions, statin therapy, and medication adherence for diabetes, hypertension, and cholesterol.

For Medicare Advantage PCP performance analytics, MA encounter data analysis, or claims-based provider benchmarking, please email us.

See more case studies