Business logic
The words behind the numbers
106 metrics and 0 glossary terms drawn from Cascadia Health Plan's own language. The semantics — not the SQL — are where the board dispute lives: 23 of 106 metrics are contested (defined more than one way and never reconciled), and 3 is undefined — a name the team uses that no query has ever computed.
contested Medical Loss Ratio (MLR) is defined more than one way and never reconciled.
contested Claims PMPM is defined more than one way and never reconciled.
contested Net Claims Cost is defined more than one way and never reconciled.
contested Days in Claims Payable (DCP) is defined more than one way and never reconciled.
contested Member Months is defined more than one way and never reconciled.
contested Active Membership is defined more than one way and never reconciled.
contested Total Premium Revenue is defined more than one way and never reconciled.
contested Risk Adjustment Factor (RAF) is defined more than one way and never reconciled.
contested Star Rating is defined more than one way and never reconciled.
contested Claims Denial Rate is defined more than one way and never reconciled.
contested Provider Directory Accuracy is defined more than one way and never reconciled.
contested Average Contracted Rate vs Medicare is defined more than one way and never reconciled.
contested Pre-Service Denial Rate is defined more than one way and never reconciled.
contested Denial Overturn Rate is defined more than one way and never reconciled.
contested 30-Day Readmission Rate is defined more than one way and never reconciled.
contested MLR Rebate Accrual is defined more than one way and never reconciled.
contested Recovery Rate is defined more than one way and never reconciled.
contested Cost Avoidance / Savings is defined more than one way and never reconciled.
contested Marketing Cost Per Acquisition (CPA) is defined more than one way and never reconciled.
contested Voluntary Turnover Rate is defined more than one way and never reconciled.
contested Overtime Hours Ratio is defined more than one way and never reconciled.
contested Mandatory Training Compliance is defined more than one way and never reconciled.
contested Average Base Salary is defined more than one way and never reconciled.
undefined Eligibility Verification Match Rate is named but never defined or computed here — a gap, not a number.
undefined Accumulator Reconciliation Variance is named but never defined or computed here — a gap, not a number.
undefined Unapplied Cash Ratio is named but never defined or computed here — a gap, not a number.
Metrics & KPIs
106 metrics · 23 contested · 3 undefinedMedical Loss Ratio (MLR)
Computed 5 different ways across the estate (claims: Paid-claims basis: SUM(claim_line.paid_amount) net of claim_line.cob_savings_amount, divided by earned premium for the period. Quality-improvement adds are excluded — this is the raw claims-side MLR the adjudication shop computes, not the regulatory MLR. = 86.4%; pharmacy: Pharmacy claims contribute to the enterprise MLR numerator: net Rx incurred cost (paid less rebates) is fed into the consolidated MLR calc. THIS app reports only the pharmacy numerator component, not the blended ratio — hence it contests the enterprise MLR figure. = 14.8% of premium; finance: GL/INCURRED basis: incurred_claims (paid_claims + IBNR reserve change posted to GL) divided by GL earned premium, per mlr_calc_worksheet. Pulls from the financial close, so it nets reinsurance and includes the quality-improvement adjustment to the numerator. This is the number the board sees and it diverges from Claims (paid/billed) and Actuarial (completion-factor) bases. = 86.9%; actuarial: Actuarial/incurred basis: (paid claims + IBNR reserve - net reinsurance/stop-loss recoverable) / earned premium, valued from the latest locked reserve_run and loss_ratio_summary. Reserves feed the numerator, so this differs from the finance paid-only and statutory federal MLR. = 86.9%; edw: Incurred claims (paid + IBNR from mart_ibnr_reserve) plus quality improvement expense, divided by earned premium from mart_revenue_monthly — assembled in mart_mlr_summary then hand-adjusted in Excel for the board deck = 86.4%). No single ratified definition.
no defining term
Claims PMPM
Computed 4 different ways across the estate (claims: SUM(claim_line.paid_amount minus claim_line.cob_savings_amount) for the period divided by member months from capitation_roster member counts. Paid-claims view, not incurred — excludes IBNR. = $412.70; provider: Provider360 estimate: allowed-amount logic applied to projected utilization via allowed_amount_rule and fee_schedule_line, divided by member months — a contracting-side PMPM, not the paid-claims PMPM = $412.80; actuarial: Incurred claims PMPM: (paid_claims + IBNR) / member months on an incurred basis from pmpm_actual and reserve_run_detail, not the paid-only figure used elsewhere. = $472.18; edw: Total paid claims from fact_claim divided by member_months from fact_premium, rolled up in mart_claims_pmpm — excludes pharmacy unless the analyst manually adds it = $452.10). No single ratified definition.
no defining term
Net Claims Cost
Computed 6 different ways across the estate (claims: SUM(claim_line.paid_amount) minus SUM(claim_line.cob_savings_amount) minus overpayment_recovery.recovered_amount. The COB fork is SUBTRACTED here, where the recovery metric instead treats the same cob_savings_amount as its target. = $2.41B; provider: Provider360 contracted-allowed view: sum of allowed amounts derived from contract_rate_schedule and allowed_amount_rule for in-network utilization, before member cost-share — a pre-adjudication estimate = $2.71B; pharmacy: Gross pharmacy paid amount minus accrued manufacturer rebates and DIR fees; THIS app counts only the pharmacy slice, so it differs from the medical-claims app's enterprise Net Claims Cost. Rebate accruals from rebate_accrual feed the offset. = $41.2M/mo; finance: GL basis: incurred_claims accrual less reinsurance recoverable, per gl_claims_incurred_accrual and gl_reinsurance_recoverable. Net of ceded losses, so lower than the Claims-system gross paid figure. = $2.79B FYTD; fraud: Sentinel computes net claims cost as gross paid claims LESS confirmed FWA recoveries and prepay denials attributed to SIU activity — i.e. claims cost net of the unit's recovered/avoided dollars. Diverges from Finance's GL-based figure because Sentinel books recoveries on identification date, not GL posting date, and pulls from its own recovery ledger fork. = $1.71B; edw: Paid claims from fact_claim plus pharmacy net cost (mart_pharmacy_summary) less COB recoveries and fraud recoveries — assembled in EDW but rarely reconciled to finance's GL number = $3.32B). No single ratified definition.
no defining term
Days in Claims Payable (DCP)
Computed 3 different ways across the estate (claims: Claims-app view: outstanding pended_claim + accrued-not-paid balance divided by average daily paid claims (claim_line.paid_amount). Excludes IBNR the actuarial team layers in. = 38.6; finance: GL balance basis: (claims payable + IBNR reserve GL balance) / (incurred claims per day). Uses gl_balances reserve liability accounts, so it reflects booked reserves not the claims operational backlog. = 47.2 days; edw: Claims payable balance from fact_dcp_calc over claims paid in period times days — denominator basis differs between finance and actuarial versions of the working calc = 41.2 days). No single ratified definition.
no defining term
Member Months
Computed 5 different ways across the estate (enrollment: Sum of fractional eligible months from member_month_summary, computed on an ELIGIBILITY-SPAN basis: a member contributes their member_month_fraction for any month where an eligibility row overlaps the month (prorated by covered days / days-in-month), not on a paid-premium or enrolled-on-1st basis. = 7,184,200; finance: Revenue-recognition basis: sum of member_months underlying earned premium in gl_premium_revenue_recognition. Reflects retro-adjusted membership as posted to GL, so it lags the Membership system's live count. = 7.04M FYTD; actuarial: De-duplicated exposure sum of member_month_actuarial.exposure_fraction over the period; differs from raw enrollment counts because overlapping spans are pro-rated to avoid double-counting. = 7.18M; edw: Sum of fractional coverage months from fact_premium.member_months, reconciled against fact_enrollment_span — retro adjustments cause it to differ from billing's count = 7,210,800; quality: Sum of measure-eligible member months contributed to denominators (member_months_contributed); derived from Membership eligibility spans the quality engine ingests, NOT from Finance enrollment counts. = 7.05M). No single ratified definition.
no defining term
Active Membership
Computed 6 different ways across the estate (enrollment: Distinct member_id with an open eligibility span (eligibility_status='ACTIVE' and eligibility_end_date null or >= as_of_date) on the snapshot date — eligibility-span basis, NOT enrollment_span or billed basis. Includes retro-active spans posted by as_of. = 601,840; care: CareTrak's basis: distinct member_id with an active member_eligibility_snapshot span as seen by UM at decision time; lags the enrollment system by the eligibility-feed delay and double-counts members with overlapping spans. = 598,210; billing: Distinct member_id in billing_member_coverage with billing_status='ACTIVE' and a billed coverage span overlapping the as-of month (BILLED-MEMBER basis — a member counts only if premium was billed, distinct from the Enrollment system's eligibility-based count). = 589,400 members; crm: Distinct member_id in sold_membership_snapshot where is_active_sold=true for the current snapshot_month. Counts on a SOLD/book-of-business basis: includes applications in 'pending_effectuation' status (broker has sold the policy but enrollment has not yet effectuated it), and continues counting a member until the broker-of-record is terminated. This inflates the count vs the enrollment system, which only counts effectuated, premium-paid members. = 619,400; appeals: Count of appeal_case + grievance_case received in period divided by Active Membership, annualized to per-1000; A&G's local member denominator differs from the enrollment system of record. = 4.7; edw: Distinct member_id with active_flag on the last day of month from snapshot_membership_daily — point-in-time, differs from average-enrollment basis used elsewhere = 601,240). No single ratified definition.
no defining term
Total Premium Revenue
Computed 5 different ways across the estate (enrollment: Sum of member_premium.net_premium across coverage months for active eligibility spans — enrollment-side view (billed-expected from rate tables), differs from Finance's GL-recognized premium. = $2.94B; billing: Sum of billed_amount across invoice_line for the coverage period (BILLED basis — gross premium invoiced, before cash receipt or earned/unearned adjustment). Includes APTC portion; nets premium holidays and credit adjustments posted in-period. = $3.94B annual; finance: EARNED/GL basis: sum of earned_premium recognized in gl_premium_revenue_recognition for the period, net of unearned premium and inclusive of retro membership restatements posted to closed periods. Differs from Billing (billed/invoiced) because GL recognizes earned, not billed. = $3.21B FYTD; crm: Sum of earned_premium_basis across commission_line for the period, rolled up from the sold book of business. Derived from the commission engine's premium feed (not the billing ledger), so it reflects expected/quoted premium on sold policies and diverges ~2% from billing's collected-premium figure. = $3.71B; edw: Earned premium from mart_revenue_monthly (billed less unearned), plus capitation revenue — but the board deck often substitutes collected premium, producing a different number = $3.91B). No single ratified definition.
no defining term
Risk Adjustment Factor (RAF)
Computed 4 different ways across the estate (care: CareTrak's basis: member-level RAF estimated from member_condition / HRA chronic-condition capture surfaced for care-gap targeting — a clinical-suspecting view, NOT the actuarial encounter-submitted RAF. = 1.07; actuarial: Membership-weighted mean normalized RAF from risk_score.normalized_raf across active members for the payment year, blending demographic + disease (HCC) scores per CMS-HCC/HHS-HCC model version; reconciled to CMS return files in raf_reconciliation. = 1.043; edw: Member-month-weighted average raf_score from fact_risk_score in mart_risk_adjustment — blends V24 and V28 model versions, so the value depends on the version mix = 1.084; quality: Quality app consumes member RAF from member_risk_segment.risk_score_id purely to stratify measure rates (e.g. dual/LIS adjustment, Categorical Adjustment Index); NOT the actuarial bid RAF. Distinct stratification basis. = 1.04). No single ratified definition.
no defining term
Star Rating
Computed 2 different ways across the estate (edw: Weighted measure-level star values from mart_stars_summary rolled to contract level — an internal estimate that frequently diverges from CMS's official published rating = 4.0; quality: Weighted average of measure-level star values (HEDIS + CAHPS + HOS + Part D admin) rolled to contract via measure_star_assignment -> contract_star_rating, with reward factor and CAI adjustment applied. Quality app's clinical-measure view of overall stars. = 3.5). No single ratified definition.
no defining term
COB Recovery Rate
SUM(cob_recovery.recovered_amount) divided by SUM(claim_line.cob_savings_amount). Treats cob_savings_amount as the recovery TARGET — the same field that MLR and Net Claims Cost SUBTRACT. Ratio can exceed 100% on retro primary payments.
no defining term
Auto-Adjudication Rate
COUNT(adjudication_result WHERE auto_adjudicated_flag) divided by total adjudicated lines. Lines that hit any claim_edit with manual disposition are excluded from the numerator.
no defining term
First-Pass Resolution Rate
Share of claims reaching a finalized claim_status in claim_status_history without ever entering pended_claim.
no defining term
Claims Denial Rate
COUNT(denied_claim lines) divided by total adjudicated claim_line count for the period. Counts line-level denials, not claim-level.
no defining term
Pended Claim Aging
Average pended_claim.pend_age_days across open pends, bucketed by pend_category. Note negative-age rows from clock skew are clamped to zero.
no defining term
Clean Claim Turnaround Time
Median claim_sla_tracking.turnaround_days for claims that never pended. Regulatory prompt-pay clock basis.
no defining term
Average Paid per Claim
SUM(claim_line.paid_amount) divided by COUNT(distinct claim_id). Does NOT net cob_savings_amount — diverges from Net Claims Cost on COB-heavy populations.
no defining term
Edit Override Rate
COUNT(edit_override) divided by COUNT(claim_edit fired). High override rate signals edit-config noise; free-text reasons make root-cause hard.
no defining term
Provider Payment Accuracy
1 minus (SUM(overpayment_recovery.overpaid_amount) / SUM(provider_payment.payment_amount)). Post-pay accuracy measured by recoveries found, so it lags.
no defining term
Capitation as % of Medical Spend
SUM(capitation_payment.total_cap_amount) divided by (capitation + SUM(claim_line.paid_amount)). Measures shift toward value-based arrangements.
no defining term
Retro Enrollment Rate
Count of retro_enrollment rows in period / total enrollment_transaction rows in period; flags backdated adjustments straining downstream claims reprocessing.
no defining term
834 Auto-Adjudication Rate
Share of enrollment_834_inbound records processed to ACTIVE without manual touch (process_status='PROCESSED' and no enrollment_834_error row).
no defining term
Member Months Growth (YoY)
Current period Member Months (eligibility-span basis) divided by same period prior year minus 1.
no defining term
Dependent Coverage Ratio
Distinct dependent member_id / distinct subscriber_id from active eligibility spans.
no defining term
Termination Rate (Annualized)
Annualized count of termination rows / average Active Membership over the period.
no defining term
Eligibility Verification Match Rate
Share of eligibility_verification_log inquiries matched to a member_id with a response_code of active coverage.
no defining term
Accumulator Reconciliation Variance
Mean absolute difference between benefit_accumulator.accumulated_amount and the claims-engine accumulator for the same member/type, expressed as % of members out of tolerance.
no defining term
Average Coverage Tenure (months)
Mean span length in months across closed enrollment_span rows, eligibility-span derived.
no defining term
Active Network Providers
Count of distinct provider_id with at least one active network_affiliation span (par_status='PAR', termination_date null) as of report date
no defining term
Credentialing Cycle Time
Mean days from provider_credentialing.initial_cred_date to committee_decision_date for completed cycles in period
no defining term
Recredentialing Compliance Rate
Share of providers with recred_due_date in period whose committee_decision_date <= due date
no defining term
Provider Directory Accuracy
Share of provider_directory rows where display_address matches active provider_location and last_verified_date within 90 days
no defining term
Network Adequacy Pass Rate
Share of provider_network_adequacy rows where meets_standard=true across county-specialty cells in current measurement cycle
no defining term
Average Contracted Rate vs Medicare
Volume-weighted contract_rate_schedule.contracted_rate divided by medicare_fee_reference.medicare_allowed across active contracts
no defining term
Fee Schedule Load Completeness
Share of fee_schedule_line rows with allowed_amount > 0 over total expected procedure codes per active fee schedule
no defining term
PAR Provider Ratio
PAR-status providers in network_affiliation divided by all affiliated providers (PAR + non-PAR) as of period end
no defining term
Provider Sanction Exposure
Count of active provider_sanction rows (is_active=true) linked to providers with active network_affiliation
no defining term
Authorization Turnaround Time (TAT)
CareTrak's basis: mean (and 95th pct) elapsed business-clock hours from auth_request.received_datetime to auth_determination.decided_datetime, by urgency tier, excluding sla_event pause_minutes. Standard/non-urgent measured in calendar days.
no defining term
Prior Auth Approval Rate
approved auth_determinations (outcome in approved/partial) divided by all finalized determinations in period; auto_approved counted as approved.
no defining term
Pre-Service Denial Rate
count of denial rows (denial_type=pre-service) over total finalized auth_requests in period; medical-necessity denials only.
no defining term
Denial Overturn Rate
denials with overturned_on_appeal=true over all denials that reached an appeal (via appeal_link); UM-internal view, may differ from Appeals app.
no defining term
SLA Compliance Rate
sla_event rows with met_flag=true over all completed sla_event rows by regulatory_standard (CMS/state UM turnaround standards).
no defining term
Case Management Engagement Rate
program_enrollment rows with engagement_level in (engaged/active) over all opened enrollments in period; duplicate active enrollments inflate denominator (~8k pairs).
no defining term
30-Day Readmission Rate
transition_of_care rows with readmission_30d_flag=true over index discharges (inpatient_stay) in period; CareTrak's UM view, not the claims-derived HEDIS rate.
no defining term
Average Length of Stay (ALOS)
mean inpatient_stay.observed_los_days for stays with non-null discharge_datetime in period; observed, not benchmark-adjusted.
no defining term
Care Plan Currency Rate
open care_plan rows with next_review_due_date not past due over all open care plans; flags the 31% overdue backlog.
no defining term
Rx PMPM
Net pharmacy spend (sum of pharmacy_claim_line.total_amount_paid less rebate_accrual.accrual_amount and dir_fee.dir_amount) divided by member months from member_pharmacy_eligibility spans. Computed monthly per line of business.
no defining term
Generic Dispensing Rate (GDR)
Count of generic-dispensed claim lines (ndc_drug_master.brand_generic_indicator='G') divided by total dispensed claim lines, per period and line of business.
no defining term
Rebate Capture Rate
Realized rebate payments (rebate_payment.paid_amount) divided by accrued rebates (rebate_accrual.accrual_amount) for the period; measures collection effectiveness against accrual.
no defining term
Prior Auth Approval Rate
PA requests with decision_status='approved' divided by total adjudicated PA requests in prior_auth_request, per drug class.
no defining term
Medication Adherence (PDC)
Share of members with PDC >= 0.80 in medication_adherence_metric for STAR adherence measures (diabetes, hypertension, cholesterol).
no defining term
Specialty Spend Share
Sum of paid amounts for specialty drugs (gpi_therapeutic_class.is_specialty=true) divided by total pharmacy paid; tracked because specialty drives most trend.
no defining term
MAC Effectiveness (MAC %)
Percent of generic claim lines priced at MAC (pricing_adjudication_result.pricing_basis_used='MAC') vs AWP/U&C; measures generic cost control.
no defining term
Premium AR Aging % > 90 days
Sum(bucket_91_120 + bucket_over_120) / total_ar from ar_aging_snapshot at period close.
no defining term
Premium Collection Rate
Premium receipts applied (cash_application.applied_amount) divided by total billed premium (invoice_line.billed_amount) for the coverage period.
no defining term
Binder Payment Effectuation Rate
Count of binders with effectuation_status='EFFECTUATED' / count of binder_payment records due, for the enrollment cohort.
no defining term
Days Sales Outstanding (Premium DSO)
Average premium AR balance / average daily billed premium, over trailing 90 days.
no defining term
Write-off Rate
Sum(write_off.write_off_amount) / Total Premium Revenue (billed) for the period.
no defining term
Delinquency Rate
Count of billing accounts with days_past_due > 30 in delinquency_status / total active billing accounts, at as-of date.
no defining term
Unapplied Cash Ratio
Sum(premium_receipt.unapplied_amount) + suspense_account balance / total premium receipts in period.
no defining term
Administrative Expense Ratio
Sum of admin/operating GL accounts (non-claims, non-quality-improvement) divided by earned premium, sourced from gl_balances and gl_premium_revenue_recognition.
no defining term
Combined Ratio
MLR (GL) plus Administrative Expense Ratio; the underwriting profitability measure used in the STAT package.
no defining term
IBNR Reserve Balance
Period-end GL balance of IBNR/unpaid-claims reserve liability accounts, per gl_ibnr_reserve_posting. Carries a small reconciliation variance against the actuarial source figure due to rounding.
no defining term
Risk-Based Capital (RBC) Ratio
Total adjusted capital divided by authorized control level, per rbc_calculation; the solvency ratio filed with regulators.
no defining term
Days to Close
Business days from period-end to final ledger close, measured as max(days_to_close) across close_task_status for the period.
no defining term
MLR Rebate Accrual
Estimated ACA 80/85 rebate liability per LOB per benefit year, using the statutory MLR numerator (which itself differs from the board MLR), per gl_rebate_accrual_mlr.
no defining term
Unapplied Cash
Sum of unapplied_amount in ar_premium_receipt_application sitting in suspense; a working-capital and revenue-recognition risk indicator.
no defining term
IBNR Reserve Balance
Sum of reserve_run_detail.ibnr_amount for the latest locked reserve_run across all incurred months and service categories; ultimate_incurred minus paid_to_date driven by selected completion factors.
no defining term
Completion Factor
Selected cumulative development factor (selected_cdf) by LOB/service category/lag bucket; the inverse estimates the percent of ultimate claims paid to date for a given incurred month.
no defining term
Reserve Redundancy/Deficiency
reserve_development: (restated ultimate incurred - original estimate) / original estimate for prior valuations; negative = redundancy, positive = deficiency.
no defining term
Loss Ratio Trend
Composite annualized trend (unit cost x utilization) from trend_assumption.composite_trend applied to incurred PMPM; the pricing trend used in manual_rate build-up.
no defining term
Total Recoveries (FWA)
Sum of recovered_amount across recovery where recovery_status='collected', net of vendor contingency fees, attributed to the SIU case that identified it. Excludes cost-avoidance.
no defining term
Identified Overpayment Dollars
Sum of overpaid_amount in overpayment for the period regardless of collection status; the top-of-funnel exposure figure before recoupment.
no defining term
Recovery Rate
recovered_amount / overpaid_amount over identified overpayments aged 12+ months; measures collection effectiveness. Contested because cost-avoidance is sometimes folded into the numerator in exec decks.
no defining term
Cost Avoidance / Savings
Sum of realized_savings in savings_estimate for prepay edits and denied claims (dollars never paid). Known to double-count against recovery in some rollups.
no defining term
Aberrancy Hit Rate
Share of providers scored above model threshold that convert to a substantiated finding; aberrancy_score joined to investigation_finding where substantiated=true.
no defining term
Case Cycle Time
Median days from siu_case.opened_date to closed_date for cases closed in period; SLA-anchored operational metric.
no defining term
Net Promoter Score (NPS)
(% Promoters [score 9-10] minus % Detractors [score 0-6]) over nps_response in the trailing 90 days. Excludes the 8% of responses with out-of-range scores. Computed per product line and channel.
no defining term
Member Persistency Rate
Share of prior-period sold members still active on the SOLD basis this period (1 minus voluntary lapse). Computed from sold_membership_snapshot month-over-month; uses CRM active flag, so lapses lag the enrollment system by up to a cycle.
no defining term
Lead Conversion Rate
Applications submitted divided by qualified leads created in the same cohort window. Numerator from application; denominator from lead where lead_status reached 'qualified'. Affected by ~8% duplicate leads inflating the denominator.
no defining term
Broker Commission Expense Ratio
Net commission paid (net_commission across commission_statement, including chargebacks and overrides) divided by Total Premium Revenue on the sold-book basis. Used for distribution-cost management.
no defining term
Marketing Cost Per Acquisition (CPA)
Total campaign spend (campaign.budget actuals) divided by attributed new applications via campaign_attribution. Distorted by the mid-year attribution-model change, which is not reconciled between pre/post cohorts.
no defining term
Book-of-Business Retention Rate
Share of a broker's prior-year book still active under the same broker-of-record. Computed from book_of_business effective/term spans; sensitive to duplicate BOR rows created during AEP.
no defining term
Appeals Timeliness Rate
Share of appeals where clock_stop_datetime <= clock_due_datetime in timeliness_clock, evaluated against regulatory_threshold_ref by LOB/level/expedited; net of valid tolling in timeliness_pause. CMS MA Part C standard reconsideration threshold = 30 days; expedited = 72 hours.
no defining term
Grievance Timeliness Rate
Share of grievances resolved within the regulatory window (standard 30 days, expedited 24 hours for quality-of-care) measured from grievance_case.received_date to resolution_date against regulatory_threshold_ref.
no defining term
Overturn Rate
determination.overturn_flag (plus partial_overturn_flag weighted 0.5) over total determinations, by appeal_level and LOB; level-1 reconsideration overturns are the STARS-relevant cut.
no defining term
Expedited Appeal Compliance Rate
Share of expedited_flag appeals meeting the 72-hour expedited threshold in regulatory_threshold_ref; pauses for member-requested extensions excluded from the numerator clock.
no defining term
Acknowledgement Letter Timeliness
Share of acknowledgement_letter rows where sent_date <= due_date (5-day ack standard for MA appeals); a leading indicator that feeds the broader timeliness clock.
no defining term
Total Headcount
Count of employee_master where employment_status = 'A' (active) as of report date; includes regular FT/PT, excludes contingent_worker
no defining term
Full-Time Equivalent (FTE)
Sum of position.fte_value for filled positions, reconciled against active employee_master rows; contractors excluded
no defining term
Voluntary Turnover Rate
Voluntary terminations (termination.termination_type='V') in trailing 12 months divided by average active headcount; numerator suspect due to ~6% mislabeled type
no defining term
Time to Fill
Mean days from requisition.open_date to offer.accepted_date for filled reqs in period; aged reqs with null hiring manager skew the denominator
no defining term
Total Labor Cost
Sum of payroll_detail.gross_pay plus employer_cost from benefit_enrollment and bonus_award.actual_amount, by cost_center and period
no defining term
Benefits Participation Rate
Distinct employees with an active benefit_enrollment (medical) divided by benefits-eligible active headcount
no defining term
Overtime Hours Ratio
Sum of timesheet.overtime_hours divided by total timesheet.total_hours in period; understated where clock-outs are missing
no defining term
Mandatory Training Compliance
Employees with all mandatory training_course completions current divided by active headcount; inflated because 15% null completion_date counted as complete
no defining term
Average Base Salary
Mean of compensation.base_salary for active employees, annualized; 7% of rows store hourly vs annual inconsistently, biasing the mean
no defining term
Offer Acceptance Rate
Count of offers with status='Accepted' divided by total offers extended in period
no defining term
Pharmacy Net Cost PMPM
Net pharmacy cost (gross less rebates) from mart_pharmacy_summary divided by member_months — rebate lag of 2 quarters makes recent months overstated
no defining term
Provider Network Adequacy %
Share of members with an in-network PCP within distance/time standard, derived from bridge_member_pcp and bridge_provider_network
no defining term
Appeal Overturn Rate
Overturned appeals over total resolved appeals from fact_appeal/mart_appeals_summary
no defining term
Fraud Recovery Rate
Recovered amount over flagged amount from mart_fraud_recovery
no defining term
Underwriting Margin %
Underwriting margin over premium revenue from mart_finance_pnl — pre manual journal adjustments
no defining term
HEDIS Measure Rate
measure_numerator compliant count / measure_denominator in-denominator count after exclusions, per measure per plan per MY. Denominator depends on Membership continuous-enrollment spans.
no defining term
Open Gaps in Care
Count of gap_in_care rows where gap_status='open' across all STARS-weighted measures, member-level rollup.
no defining term
Medication Adherence (PDC ≥80%)
Members with Proportion of Days Covered >=80% / eligible members, per adherence drug class (diabetes, hypertension, cholesterol). From pharmacy_measure_event PDC numerator days.
no defining term
CAHPS Composite Score
Case-mix-adjusted top-box percentage per CAHPS composite, converted to star value via CMS cut points.
no defining term