Practice-Based Insights into Adult Genetics: High Diagnostic Yield, Demographic Determinants, and Patterns of Test Utilization across 10,000 Patient Encounters

This study analyzes eight years of data from nearly 10,000 adult patient encounters to demonstrate that genetic testing in adults offers a high diagnostic yield, particularly with whole-exome sequencing, while revealing significant variations in test utilization and outcomes based on patient age, referral indication, and operational factors.

The Gist

Imagine the human body as a massive, intricate library. For a long time, doctors thought this library only needed serious "book repairs" (genetic testing) when a child was born with a broken shelf. But this new study suggests that the library keeps getting messy and confusing as we get older, and we need to check the books in the adult section, too.

Here is a breakdown of what this study found, using simple analogies:

1. The Big Picture: A Busy Library

The researchers looked at 10,000 visits to a major adult genetics clinic over eight years. Think of this as watching a busy library for eight years to see who comes in, why they come, and what happens when they ask a librarian for help.

• Who comes? Mostly adults under 50, but people of all ages show up. There are more women than men, and most are White, which highlights that some groups (like Black and Hispanic communities) still face barriers to getting through the front door.
• Why do they come? People came for all sorts of reasons: family history of cancer, joints that are too loose, heart issues, or just feeling generally unwell.

2. The "Book Search": How Often Do They Find the Answer?

When a patient asks for a genetic test, it's like asking a librarian to find a specific, missing page in a book.

• The Success Rate: About 1 in 4 (24%) of these searches actually found the missing page (a diagnosis). That's a pretty good success rate for a complex search!
• The Age Factor: You might think older people have "older books" that are harder to read, but the study found that even in older adults, the search still worked about 17% of the time. The library never stops having answers to find.

3. The Tools: Flashlights vs. Drones

The study compared different ways of searching for the answer:

• The Flashlight (NGS Panels): This is the most common tool. It's like shining a flashlight on just a few specific shelves. It's cheap and fast, but it only finds the answer 16% of the time.
• The Drone (Whole Exome/Genome Sequencing): This is a high-tech drone that scans the entire library at once. It's more expensive and harder to get permission to use, but when they used it, they found the answer 40% of the time.
• The Lesson: Sometimes, using a flashlight on the wrong shelf takes forever and finds nothing. If the librarian suspects a big problem, sending the drone immediately might save time and money in the long run.

4. The "Why" Matters: Different Problems Need Different Searches

The study found that the type of problem a patient has changes how likely they are to get an answer:

• High Success: If someone came in with "multisystem" issues (trouble in many parts of the body) or specific kidney/vascular problems, the "drone" was very likely to find the answer.
• Low Success: If someone came in just because their joints were very loose (hypermobility), the search was rarely successful. It's like looking for a missing page in a book that doesn't actually have one.
• The "Maybe" Zone: Some searches came back with "Variants of Uncertain Significance" (VUS). This is like the librarian saying, "I found a page that might be the right one, but I'm not sure." This happened often with connective tissue issues, which can be frustrating for patients.

5. The Bureaucracy: Who Pays for the Search?

The study showed that what gets tested isn't just about what's medically best; it's often about insurance and money.

• If a lab offered a "capped cost" (a set price), the clinic suddenly ordered more tests from them.
• If a new insurance plan covered a specific test, the volume of those tests skyrocketed.
• The Metaphor: It's like the library having to choose which books to buy based on which publisher offers a discount, rather than which books are actually the most important for the readers.

6. The Future: A New Way to Run the Library

The authors suggest a few changes to make adult genetics better:

• Smart Triage: Don't use the flashlight for every problem. If the symptoms look complex, send the drone (exome sequencing) right away.
• Manage Expectations: Tell patients upfront, "If you have joint pain, we might not find a genetic cause, so let's not get our hopes up." This prevents disappointment.
• Telehealth: The study saw a huge jump in video visits during the pandemic. This is great! It's like having a librarian visit you at home to explain the results, which is much easier than driving across town for a follow-up.

The Bottom Line

This study is a wake-up call. Genetic testing isn't just for kids anymore. It works well for adults, too, but we need to be smarter about when we use it and how we pay for it. By matching the right tool to the right problem, we can help more adults get answers to their health mysteries, regardless of their age.

Technical Summary

1. Problem Statement

Precision medicine relies heavily on genetic testing, yet the landscape of adult medical genetics remains poorly characterized compared to pediatrics. Key gaps include:

• Lack of Guidelines: Evidence-based guidelines for adult genetic evaluation and testing are sparse.
• Workforce Mismatch: Primary care and internal medicine clinicians report lower preparedness for interpreting genetic results, while the adult-focused genetics workforce is limited.
• Data Deficit: There is a lack of systematic, large-scale data linking referral patterns, testing modalities, and diagnostic outcomes in routine adult practice. Most existing studies are small, pediatric-focused, or limited to specific high-risk cohorts (e.g., ICU patients).
• Operational Blind Spots: It is unclear how payer policies, laboratory partnerships, and operational shifts (e.g., telehealth) influence testing utilization and diagnostic yield in a real-world setting.

2. Methodology

The study utilized a retrospective, single-center analysis of electronic health record (EHR) data from the University of Pennsylvania Health System (UPHS) Adult Genetics Clinic.

• Cohort: 9,867 unique patient visits representing 7,766 unique patients between October 1, 2016, and September 30, 2024.
• Data Integration: EHR data (demographics, visit type, provider) was joined with internal clinical databases (referral indications, test orders, results). Missing fields (4,264 visits) were resolved via manual chart review.
• Classification:
  • Indications: Collapsed into 21 broad clinical indication groups (e.g., Cancer Genetics, Connective Tissue Disorders, ID/DD/ASD).
  • Test Results: Categorized as Pathogenic/Likely Pathogenic (Diagnostic), Variant of Uncertain Significance (VUS), Carrier, Candidate Gene, or Negative.
  • Demographics: Race and ethnicity were harmonized from EHR data, combining self-reported and clinician-assigned identities into standardized categories.
• Statistical Analysis: Performed using R (v4.2.2). Analyses included logistic regression (adjusting for age, sex, race, insurance, and visit type) to determine odds ratios for testing utilization and diagnostic yield. Significance was determined using Bonferroni corrections for multiple comparisons.

3. Key Contributions

This study provides the first large-scale, practice-level mapping of the adult genetics ecosystem, offering:

• Diagnostic Yield Benchmarks: Establishes baseline diagnostic yields across the entire adult lifespan and various clinical indications.
• Modality Comparison: Directly compares the diagnostic efficiency of Next-Generation Sequencing (NGS) panels vs. Whole Exome/Genome Sequencing (WES/WGS) in a routine clinical setting.
• Operational Insights: Quantifies how external factors (payer policies, lab integrations, telehealth adoption) drive testing behavior.
• Demographic Stratification: Reveals disparities in referral patterns, testing rates, and outcomes based on age, sex, and race.

4. Key Results

A. Demographics and Visit Patterns

• Patient Mix: 60.4% female; 77.1% White. The age distribution skewed younger, with peaks at ~30 and ~60 years.
• Visit Types: 79.3% were new consultations; 20.7% were follow-ups. Telehealth usage rose sharply in 2020 (13.8% of total visits) and stabilized for follow-ups.
• Indications: Top referrals were Cancer Genetics (22%), Connective Tissue Disorders (16.6%), and Multisystem Disorders (16.5%).
• Temporal Trends: Hypermobility referrals declined significantly (due to new triage criteria), while referrals for cancer, multisystem, and renal disorders increased.

B. Testing Utilization

• Ordering Rates: Genetic testing was ordered for 52% of new patients and 8% of follow-ups.
• Demographic Drivers:
  • Age: Older patients were more likely to be tested (OR 1.01 per year), suggesting a higher threshold for suspicion or cascade testing.
  • Sex: Males were more likely to be tested than females (OR 1.17).
  • Indication: Testing was most likely for ID/DD/ASD (80.9%) and Neuromuscular disorders; least likely for Hypermobility and Chromosomal disorders (often already diagnosed).
• Modality Shifts: NGS panels were the most common test (62.8%), followed by Single Gene (15.2%). Over time, NGS panels and WGS increased, while CMA and Single Site testing decreased.
• Lab Utilization: Three labs (Invitae, Ambry, GeneDx) performed 83% of tests. Shifts in lab usage closely tracked payer policies (e.g., capped-cost programs, in-network status) and EMR integrations.

C. Diagnostic Yield

• Overall Yield: 24.3% of all tests returned a pathogenic/diagnostic result.
• Modality Yield:
  • WES/WGS: Highest yield at 39.7%, despite comprising only 3.7% of tests.
  • Single Site/Karyotype: High yield (~43%), but low volume.
  • NGS Panels: Lowest yield among high-volume tests at 16.2%.
  • Metabolic Testing: Lowest yield at 6.2%.
• Age Impact: Diagnostic yield declined with age (from 37.2% in ages 18–25 to ~17% in >75), but WES/WGS yield remained >30% across all age groups, including the elderly.
• Indication Impact:
  • High Yield (>40%): ID/DD/ASD (52.3%), Chromosomal Disorders (50.8%), Vascular Malformations (48.8%).
  • Low Yield (<15%): Hypermobility (9.0%), Immune/Hematologic (9.7%), Lung Disorders (13.8%).
• VUS Rates: Higher in Connective Tissue (22.9%) and Immune/Hematologic (38.1%) disorders. Males had a significantly higher rate of VUS results than females.

D. WES/WGS Specifics

• Patients undergoing WES/WGS were significantly younger (median 31 vs. 43.6 years) and more likely to have Medicaid insurance.
• WES/WGS showed no attenuation of diagnostic yield with age, maintaining high utility (>40%) even in patients >55 years old.
• ID/DD/ASD referrals had a significantly higher likelihood of a pathogenic finding via WES/WGS compared to other indications.

5. Significance and Implications

• Adult Genetics is High-Yield: Contrary to the perception that genetic diagnosis is primarily a pediatric enterprise, this study confirms that adult genetics offers high diagnostic value, particularly when utilizing broad genomic approaches.
• Testing Strategy Optimization: The mismatch between the high volume of NGS panels and their lower yield (16%) compared to WES/WGS (40%) suggests a need for "Exome/Genome-First" strategies, especially for complex, multisystem, or undiagnosed adult cases.
• Triage and Guidelines: The data supports the development of indication-aware triage pathways:
  • Low Threshold for Panels: For high-yield indications (e.g., Renal, Vascular).
  • Rapid Escalation to WES/WGS: For low-yield indications where panels fail, or for high-complexity presentations (ID/DD, Multisystem).
  • Counseling: Pre-test counseling must manage expectations for indications with high VUS rates (e.g., Connective Tissue).
• Operational Policy: Testing behavior is heavily driven by reimbursement and infrastructure (e.g., EMR integration, capped costs). Aligning payer policies with clinical value (e.g., covering WES for high-yield indications) is critical for equitable care.
• Workforce and Access: Telehealth is a viable, efficient tool for follow-ups and cascade testing, potentially expanding access to rural areas. However, disparities in race and insurance persist, highlighting the need for targeted outreach and policy reform.

Conclusion: The study provides actionable evidence to shift adult genetics from a reactive, panel-driven model to a proactive, indication-stratified, and often exome-first approach, optimizing diagnostic yield and resource allocation across the adult lifespan.