How have changes in program integrity funding historically affected the annual number and targeting of CDRs?
Executive summary
Changes in program integrity funding in federal health programs have historically expanded the scope and number of review activities—adding contractors and new postpayment and investigative tools—while persistent measurement and data-use shortfalls have limited the ability to translate dollars into sharply targeted case-level reviews (sources: GAO, CMS, OIG) [1] [2] [3].
1. Funding increases translated into more activities and contractors, not just more CDRs
When appropriations rose for program integrity, agencies used the funds to stand up new activities and contractors—examples include the addition and evolution of contractors such as CERT, ZPICs, RAs, MACs, and other program safeguard contractors rather than a single uniform review type—so increased money produced a larger integrity apparatus rather than a simple, proportional rise in a specific category of reviews [1] [2] [4].
2. The rise in postpayment review capacity altered annual counts but measurement problems obscure the link
Expanded funding supported more postpayment mechanisms—Recovery Auditors and zone integrity contractors that perform postpayment claims reviews—which increases the total annual reviews and recoveries, yet auditors and oversight reports found that CMS did not consistently use CERT program data to prioritize error-prone providers, undercutting confidence that more funding led to smarter targeting of reviews [2] [3].
3. Guidance and manuals give agencies levers to target high‑risk providers, but practice lags guidance
The Medicare Program Integrity Manual instructs Medicare Administrative Contractors to target providers with high denial rates or atypical billing patterns, giving a clear framework for focused reviews when resources permit; however, oversight work shows CMS and contractors sometimes failed to apply CERT findings to identify and apply additional integrity tools to providers with elevated risk, revealing a gap between intended targeting and operational reality [4] [3].
4. Program interruptions and statutory timelines change annual review volumes independently of funding
Operational pauses and legal reporting timelines also influence yearly counts: CERT activities were paused during COVID and resumed in August 2020, and the cyclical nature of CERT plus statutory reporting requirements means annual measurement and therefore the number of formally counted reviews can fluctuate for reasons unrelated to funding levels [5].
5. Accountability, ROI measurement, and evidence needs shape how funding affects targeting
GAO found that CMS’s allocation decisions consider potential effects on improper payments but that officials did not consistently connect program integrity activities to agency goals and that data used to calculate return on investment were flawed—meaning that even when funding increased, the absence of reliable performance measurement hindered evidence-based decisions about where to direct reviewers and whether to expand or contract specific review types [1].
6. What oversight bodies recommend and what that implies about historical effects
Oversight recommendations have been consistent: use annual CERT data to identify individual error‑prone providers and take specific actions like prior authorization or targeted prepayment reviews; this implies that historically, funding without improved data use produced more activity but not necessarily better-targeted reviews, and that improving data-to-action pipelines could change that relationship [3].
7. Bottom line and reporting limits
Available federal reports show a clear pattern: increases in program integrity funding historically expanded contractor capacity and the universe of postpayment and investigative activities, which tended to raise the aggregate number of reviews, but chronic measurement, data-use, and implementation gaps meant those funds did not always sharpen the targeting of case reviews to the highest‑risk providers; the sources reviewed do not define or specifically quantify "CDRs," so this analysis cannot assert precise historical counts for that undefined term [1] [2] [5] [3] [4].