Streamlining oversight in Medicare advantage: measuring coding intensity via contract rankings

Risk-adjusted payments in Medicare Advantage (MA) incentivize insurers to maximize reported diagnoses, resulting in high coding intensity, a major source of Medicare overpayments. However, distinguishing justifiable coding intensity from misreporting is challenging: some approaches risk conflating coding intensity due to extraneous diagnoses with that due to variation in clinical practice or beneficiary health, while others assume specific overreporting mechanisms.

We propose a data-driven method for ranking MA contracts by coding intensity using comparisons between geographically co-located contracts based on adjusted diagnosis rates among switchers from Traditional Medicare to MA. The approach is applied to a 20% sample of 2019 MA claims.

Health Maintenance Organization (HMO)-dominated contracts tend to rank higher than preferred provider option (PPO)-dominated contracts, echoing previous findings. Furthermore, highly ranked contracts administer procedures and prescriptions relevant to beneficiaries' diagnoses less often than lower-ranked contracts, supporting the validity of the ranking. A simulation study suggests even an imperfect ranking can recover overpayments faster than a random audit, if mean diagnosis rates are low (10%) and vary widely across contracts.

As MA looks to expand auditing strategies, MA enrollment continues to grow. The proposed approach can facilitate payer-specific audits even as mechanisms driving coding intensity evolve.