Voice Biomarker Pilot — Case Study

55% of Patients Had
Unmet Clinical Needs.
None of Them Knew It.

Amplifier Health's Sona-2 analyzed recorded patient interactions across a network of Texas-based primary care clinic locations — revealing that 55% of patients carry undetected conditions that align with existing clinical programs. Closing this care gap also unlocks significant new revenue. No new workflows. No patient burden. Just the voice.

OrganizationTexas-based primary care clinics
Pilot TypeVoice Biomarker Analysis
DateMarch 2026
Powered byAmplifier Health
374
Patient appointments processed for voice biomarker analysis
Across 19 Texas-based primary care clinic locations
55.2%
Of analyzed patients flagged for at least one new program
107 of 194 patients
4
High-margin service lines identified and mapped
Sleep · Cardio · Nutrition · Allergy
32.5%
Patients flagged for 3 or more programs simultaneously
The highest-value multi-condition cohort

The Problem

Patients Were Leaving
With Conditions Undetected.

The network of Texas-based primary care clinics operates one of the most comprehensive men's health networks in the country. Yet like most multi-location systems, a significant share of patients with diagnosable, treatable conditions were cycling through appointments without those conditions ever being identified — not through any failure of care, but through the structural limits of a standard clinical encounter.

🩹

Conditions Were Going Undiagnosed

Sleep apnea, cardiovascular risk, metabolic dysfunction, and allergic disease are highly prevalent yet severely underdiagnosed in primary care. Patients carrying these conditions were completing visits without any indication that a relevant program or workup was warranted.

A 15-Minute Visit Can't Screen for Everything

Clinicians cannot be expected to proactively screen every patient for every applicable condition during a standard appointment. The care gap is not a reflection of clinical quality — it is a structural impossibility baked into how primary care operates at scale.

🔊

The Health Signals Were Already Present

Every patient interaction contains acoustic biomarker information — vocal patterns that correlate with sleep disorders, cardiovascular stress, metabolic strain, and more. Without a system to interpret them, these signals passed through unheard and unacted upon.

The Solution

The Signal Was in the Voice.

Amplifier Health deployed its Sona-2 — the most advanced foundation model for human health acoustics — to passively analyze patient voices during routine clinical interactions — appointment calls, scribe recordings, or any captured conversation. No new workflows. No patient surveys. No additional clinical burden.

Sona-2 analyzes only acoustic features of the voice and does not use any personally identifiable information (PII) to perform its analysis.

Phase 1
01

Integrate & Activate

Amplifier Health integrated with the practice's existing recording infrastructure — capturing patient voice from routine clinical interactions already taking place, with no change to clinical workflows or patient experience.

Phase 2
02

Analyze 20+ Biomarkers

As patients responded, Sona-2 extracted hundreds of acoustic features — pitch, rhythm, speech rate, vocal quality, breathing patterns — and ran them through 20+ condition-specific models simultaneously, generating probability scores for each health condition.

Phase 3
03

Prioritize Patients for Clinical Outreach

Detected conditions were weighted through a clinical scoring framework to produce tiered care recommendations — Strong, Moderate, or Consider — for each patient across four service lines. The result: an actionable map of which patients need clinical attention and why, delivered at the clinic level.

🔬

20+ Condition-Specific Models. One Routine Phone Call.

Each voice sample was processed through separate models for overweight/obesity, elevated blood pressure, sleep disorders, depression, anxiety, ADHD, anemia, and 14 additional conditions — simultaneously, in the background, from recorded patient interactions already taking place across the clinic network.

Unmet Clinical Need by Service Line

Four Programs. A Consistent Signal:
Patients Are Underserved.

Across four clinical service lines, Sona-2 identified patients carrying risk profiles that warranted evaluation — patients who had not yet been referred. Strong indicates multiple converging risk factors requiring prompt clinical outreach. Moderate indicates a clear risk profile with likely clinical benefit. Consider indicates meaningful signals warranting a screening conversation. Enrollment in these programs also represents significant incremental revenue.

Sleep Apnea Program
Obstructive sleep apnea is severely underdiagnosed in men and, left untreated, compounds cardiovascular disease, metabolic dysfunction, and cognitive decline. Sona-2 identified converging risk signatures in over a third of analyzed patients — most of whom had no prior sleep disorder referral.
35.6%
Strong &
Moderate
Recommendation Distribution — 194 Analyzed Patients
22.7%
12.9%
14.4%
Strong — 44 patients (22.7%)
Moderate — 25 (12.9%)
Consider — 28 (14.4%)
None — 97 (50.0%)
Cardiovascular Testing
Elevated blood pressure was detected acoustically in 39.7% of the patient panel — a clinically significant finding in a population where hypertension often goes unmanaged between visits. Combined with co-occurring metabolic risk factors, over a third of patients showed a profile warranting cardiovascular evaluation.
35.6%
Strong &
Moderate
Recommendation Distribution — 194 Analyzed Patients
14.4%
21.1%
18%
Strong — 28 patients (14.4%)
Moderate — 41 (21.1%)
Consider — 35 (18.0%)
None — 90 (46.4%)
Health & Nutrition Coaching
Obesity, hypertension, dehydration, and anemia are among the most clinically modifiable conditions in primary care — yet they are chronically under-addressed without proactive identification. Nearly half of analyzed patients showed signals aligned with metabolic risk, representing the broadest unmet clinical need across the panel.
32.5%
Strong &
Moderate
Recommendation Distribution — 194 Analyzed Patients
10.8%
21.6%
16%
Strong — 21 patients (10.8%)
Moderate — 42 (21.6%)
Consider — 31 (16.0%)
None — 100 (51.5%)
Allergy Testing
Allergic rhinitis is frequently undiagnosed in primary care, contributing to chronic fatigue, sleep disruption, and reduced quality of life. Over a quarter of analyzed patients showed vocal indicators consistent with allergic disease — identified passively, without any additional patient burden.
27.3%
Strong &
Moderate
Recommendation Distribution — 194 Analyzed Patients
5.7%
21.6%
Strong — 11 patients (5.7%)
Moderate — 42 (21.6%)
None — 141 (72.7%)

"Sleep apnea and cardiovascular risk each surfaced in over a third of analyzed patients. These are not incidental findings — they represent a systematic gap between the clinical need present in the practice's patient panel and the care those patients are currently receiving. Closing that gap improves outcomes. It also happens to represent significant untapped revenue."

— Amplifier Health Clinical Analysis, February 2026

What the Voice Revealed

A Patient Population Carrying
More Than the Chart Shows.

Beyond program recommendations, Sona-2's condition-level analysis produced a population health portrait spanning 20+ conditions. Most findings aligned with expected ranges for a male primary care population. One finding did not.

Condition Prevalence — 194 Analyzed Patients vs. Expected U.S. Adult Range
0% 10% 20% 30% 40% 50% Overweight / Obesity 40.2% Within range Elevated Blood Pressure 39.7% Within range Fatigue / Malaise 21.6% Within range Sleep Disorders 21.1% Within range Stress / Nervousness 20.6% Within range Anxiety 14.9% Within range Substance Use 14.9% Within range Depression 14.4% Within range ADHD 6.7% Within range Alcohol Use 10.3% Within range PTSD 6.2% Within range Dehydration 9.8% Within range ★ Anemia 11.9% ▲ ABOVE EXPECTED RANGE Expected U.S. adult range Observed rate ★ Above expected range
🚨

Clinical Alert: Anemia Flagged at 1.5–4× the Expected Rate

Sona-2 flagged anemia indicators in 11.9% of analyzed patients — well above the U.S. adult expected range of 3–8%. This was not a targeted objective of the pilot. It surfaced organically from the voice data. Anemia is associated with chronic fatigue, cognitive impairment, and reduced treatment responsiveness — conditions that directly degrade patient outcomes and program efficacy across every service line. This single finding warrants a dedicated clinical review and may indicate a systemic screening gap within the patient panel.

The Most Underserved Patients

The Same Risk Factors.
Multiple Unmet Clinical Needs.

One of the pilot's most significant structural findings: clinical risk doesn't distribute randomly. The same metabolic and behavioral factors that drive sleep apnea also compound cardiovascular risk and nutritional need. When a patient carries one undetected condition, they typically carry several — and none were being addressed.

The Multi-Program Cohort

Of the 194 analyzed patients, 32.5% carried risk profiles warranting three or more clinical programs simultaneously. These are the practice's most clinically complex patients — already in the system, attending regular appointments, yet leaving without the care pathways their conditions warrant.

Proactively reaching these patients improves health outcomes and reduces the long-term burden of untreated chronic disease. It also represents a compounding revenue opportunity: each appropriate referral generates incremental program enrollment across multiple service lines — surfaced from a single voice biomarker screen.

52
patients flagged for
exactly 3 programs
11
patients flagged for
all 4 programs
Program Qualification Distribution — 194 Patients
0 programs
87 patients
44.8%
1 program
34
17.5%
2 programs
10
5.2%
3 programs ★
52
26.8%
All 4 programs ★
11
5.7%
★ 32.5% of patients carry risk profiles warranting 3 or 4 clinical programs simultaneously — unaddressed conditions that reactive identification will never surface at scale.

The Scale of the Opportunity

19 Clinics.
A Network of 74.

The pilot spanned 374 appointments across 19 Texas-based primary care locations — and delivered a consistent signal: 55% of analyzed patients carry undetected clinical conditions that align with existing programs. The pattern held across diverse clinic locations and patient populations. At scale across 74+ locations, the clinical and revenue impact is substantial.

74+
Texas-based primary care clinic locations eligible for a nationwide rollout of voice-powered program screening
2.7
Average conditions flagged per analyzed patient — reflecting the co-occurring metabolic and behavioral risk burden in this population
55%
Of analyzed patients flagged for at least one program opportunity — a consistent signal across diverse clinic locations

The Revenue Signal Is
Already in the Voice Data.

Every patient appointment call carries acoustic intelligence that maps directly to high-margin programs. Amplifier Health transforms that signal into proactive program identification, improved patient outcomes, and measurable revenue growth.

Schedule a Partnership Review