By David Lareau, President and CEO, Medicomp Systems
LinkedIn: David Lareau
LinkedIn: Medicomp Systems
Medicare Advantage oversight has entered a more data-driven phase, with regulators focusing closely on how diagnoses are identified, documented, and submitted for risk adjustment. At the center of that scrutiny is the practice of “suspecting,” which involves searching patients’ medical records to identify potential undiagnosed conditions by piecing together available information, rather than through direct evaluation during a clinical encounter.
An analysis published in 2025 reveals that suspecting is highly common, but the reasons are likely more related to financial gain than clinical quality. The review of 2022 Medicare Advantage encounter data found that more than 60% of enrollees had at least one chart review conducted by their insurer, and that diagnoses added through those reviews increased federal payments for 17% of enrollees.
In total, chart reviews accounted for an estimated $24 billion in additional Medicare Advantage payments in a single year, making them the largest contributor to higher spending tied to coding intensity. Many of the added diagnoses were not present in the encounter data submitted by providers, indicating that they were identified outside the point of care.
Diagnoses were far more likely to be added than removed through these reviews, reinforcing concerns that retrospective reviews likely prioritize revenue impact over clinical validation. As enforcement actions accelerate, understanding how suspecting operates within Medicare Advantage has become essential for organizations seeking to reduce compliance risk while preserving clinical integrity.
Conflating Correlation
A common source of error in suspecting arises from conflating clinical correlation with diagnostic relevance. Many conditions co-occur frequently within patient populations. That co-occurrence does not establish a causal or clinically managed relationship.
For example, chronic conditions such as diabetes often appear alongside age-related findings such as cataracts. Some suspecting approaches infer a complication relationship based on co-occurrence alone, elevating the diagnosis to a higher acuity category. If the provider did not assess, evaluate, or manage the condition as a complication during the encounter, that inference lacks clinical support.
The distinction has become more important as natural language processing and large language model technologies are introduced into documentation workflows. Automated summaries may combine conditions in ways that appear clinically plausible yet do not reflect physician intent or care delivered. Without a mechanism to validate diagnostic relevance in real time, these errors can result in diagnosis upcoding.
Retrospective Risks
Another high-risk pattern involves retrospective suspecting performed after the encounter has concluded. Coders or reviewers may identify symptoms, exam findings, or historical data that suggest a condition associated with higher reimbursement. Providers are then asked to amend documentation to include diagnoses that were not addressed during the visit.
This practice carries significant compliance concerns. Medicare Advantage risk adjustment requires that diagnoses be supported by treatment, evaluation, assessment, and management during the encounter. Adding diagnoses after the fact, without corresponding clinical activity, creates exposure under the False Claims Act. Recent enforcement actions have underscored regulators’ focus on this behavior.
Importantly, the issue is not intent. Many organizations adopt retrospective suspecting in good faith, believing they are improving documentation completeness. The regulatory standard, however, centers on what occurred during the encounter rather than what could have been documented.
Why Suspecting Persists
Retrospective reviews persist because, in theory, they fill a perceived operational gap. Physicians operate under significant time constraints, coding teams are evaluated against accuracy benchmarks, and health plans manage financial performance tied to risk-adjustment outcomes. Amid these pressures, suspecting can offer efficiency by surfacing diagnoses that might otherwise be missed.
The problem is that efficiency achieved outside the point of care trades short-term gains for long-term risk. When suspecting operates independently of clinical validation, it becomes a mechanism for amplifying documentation bias rather than improving care accuracy.
Shifting the Timeline
Reducing Medicare Advantage fraud risk requires shifting suspecting to a clinical support function rather than a revenue-optimization strategy. More specifically, suspecting should occur before or during the encounter, supporting providers in confirming whether conditions are present, relevant, and managed.
Within the encounter workflow, physicians could access a diagnostic natural language processing and a clinical data engine to connect symptoms, findings, ICD-10 codes, and Hierarchical Condition Categories. This decision-support tool could prompt physicians in real time to confirm that each diagnosis is supported by documented evaluation, assessment, and management before the visit is closed.
By reinforcing complete and specific documentation at the point of care, and by filtering clinically relevant concepts within the existing electronic health record workflow, the solution would strengthen coding accuracy and reduce reliance on retrospective chart reviews that add unsupported diagnoses.
Organizations that adopt this model position themselves to withstand regulatory scrutiny. They demonstrate that diagnoses submitted for risk adjustment are grounded in documented clinical activity, supported by evidence, and validated at the time of care delivery. Most importantly, suspecting during the encounter supports patients directly by prompting real-time confirmation of active conditions, strengthening diagnostic accuracy, and ensuring that care plans reflect clinical relevance at that moment.
Re-Centering Suspecting
Heightened enforcement and increased data transparency are reshaping expectations for how Medicare Advantage diagnoses are identified, documented, and submitted.
The question facing stakeholders is straightforward: Will suspecting remain a retrospective exercise driven by financial incentives, or evolve into a clinically grounded practice that strengthens compliance? The answer will shape the outcomes of these audits and penalties in the years ahead.