Reducing unnecessary refer to medical director (RMD) responses represents one of the most impactful ways to improve automated underwriting (AU) performance. A new paper from RGA explains how, by refining impairment criteria, adding targeted reflexives, and grounding decisions in medical guidance, insurers can increase straight-through processing rates and reduce cost while preserving sound risk selection.
Go deeper: Read the full study today.
Technology is full of innovations that promise to make life easier – until users discover they require refinement to deliver real value.
Think of early GPS navigation: Drivers were thrilled to get turn-by-turn directions, but first-generation systems often sent cars into dead-ends or traffic unless the maps were updated, the routes recalibrated, and the logic refined. The tool itself held enormous potential, but without accurate inputs and modernized rules, it could not deliver on the promise of seamless travel.
Automated underwriting (AU) is in a similar place today. It was designed to accelerate decisions, reduce friction, and expand straight-through processing (STP). Yet many AU programs still route a significant volume of cases to “refer to medical director” (RMD) because the underlying rules are outdated or overly cautious. These unnecessary referrals slow cycle times, add cost, and undermine the speed and efficiency AU was built to provide.
But there is a solution: Insurers can safely reduce 15%–40% of MD referrals by converting predictable manual triggers into refined, rules-based decisions supported by reflexive questions and medical guidance.
The result is faster decisions, lower cost, and preserved mortality outcomes.
Many RMD triggers are remnants of manual underwriting eras when limited data forced underwriters to default to caution. Today, AU engines have access to prescription histories, lab data, electronic health records, and dynamic reflexive questioning. Still, many rules operate on outdated “better-safe-than-sorry” logic.
This misalignment creates several consequences:
Just as early GPS systems became reliable only once routing logic and map data improved, AU becomes powerful only when rules evolve to reflect the available data and clinical realities.
Examples abound in which updated rules and reflexive questions meaningfully reduce RMDs without compromising mortality. The three below show how the right approach can embed nuanced decisioning within automation.
Legacy manuals often flag asthma for MD review after two or more episodes, treating all recurrences as equal. Yet data shows that 95% of individuals with well-controlled asthma qualify for standard or near-standard rates.
A more precise approach would:
This additional nuance allows AU systems to confidently auto-approve low-risk asthma applicants while still routing severe cases to MD review.
OSA is another area where outdated triggers cause unnecessary RMDs. Historically, disclosure of OSA – especially alongside a high BMI – sent cases directly to MDs. Yet treatment adherence dramatically alters mortality risk. In fact, consistent CPAP usage (≥4 hours per night) significantly reduces adverse cardiovascular outcomes.
The refined rule framework includes:
This converts what was once a near-automatic RMD into a condition suited for rule-based decisioning.
Lab-driven conditions, especially diabetes and liver enzyme elevations, have traditionally triggered RMDs for expert physician interpretation. Yet abundant population risk data exists tying specific hemoglobin A1c levels and liver enzyme patterns to mortality outcomes.
Modern AU engines can integrate:
This shifts many diabetes cases into automated decision pathways, reserving MD review for extreme or inconsistent lab patterns.
Many manual triggers express concern vaguely, with phrases such as “multiple episodes,” “abnormal labs,” or “recent relapse.” Modern AU requires translating these into specific, structured variables such as counts, recency windows, and numeric thresholds. These inputs allow the rules engine to mimic MD reasoning more faithfully.
Dynamic interrogation capabilities, such as those in RGA’s Aura Next underwriting engine, capture the nuance MDs typically request, but only when necessary. This preserves a short and simple application experience for healthy individuals while collecting depth where risk signals arise.
Refined rules must be vetted collaboratively. MDs provide clinical judgment on rating boundaries, rule thresholds, and cases where automation should remain conservative. RGA’s Global Underwriting Manual supports this process with research-backed guidance.
Retrospective testing validates whether new rules replicate MD decisions. Controlled pilots confirm operational gains and ensure no unintended mortality impact. Continuous improvement maintains alignment with evolving medical knowledge.
Reducing unnecessary RMDs generates immediate value:
Guardrails remain essential. Rules must be carefully specified, data must be complete, and audits must catch edge cases. Automation should extend medical judgment, not replace it.
Ready to explore a partnership with RGA to bring these insights to life in your business? Contact us today.
