Precision medicine’s cutting edge had promised targeted benefit, yet care too often proceeded without the timely, covered tests that determined who would actually gain and who might be harmed by treatment. A new analysis from UCSF in Science set the stakes plainly: the United States had built express lanes for high-cost drugs while leaving diagnostics to jostle through slow traffic, with regulatory, payment, and workflow barriers compounding delays at the bedside. The result showed up in everyday practice—a reliance on empiric prescribing, repeat visits to chase authorization, and mounting uncertainty when a simple, validated assay could have clarified the decision. This gap was not academic. It shaped who received disease-modifying therapy, when monitoring began, and whether safety risks were anticipated or discovered too late.
The Diagnostic Gap Is Real
Clinicians increasingly confronted targeted therapies without matching, accessible tests, a mismatch that pushed decisions into the gray zone of “treat and see.” The UCSF authors detailed how test availability, coverage, and logistics lagged even when assays existed, turning what should be a single clinical pathway into a multi-stop obstacle course. Blood-based biomarkers for neurodegeneration, pharmacogenomic panels for adverse event risk, and next-generation sequencing for rare disorders were proven in principle yet patchily available in practice. When a test required an out-of-network lab, a separate appointment, or a prior authorization with unclear criteria, clinicians defaulted to empiric treatment, exposing patients to avoidable side effects and systems to misallocated spending.
The bottleneck at the point of care was practical as much as scientific. Consider a routine memory clinic visit: cognitive testing flagged concern, but confirmatory blood biomarkers awaited scheduling at a distant site, while imaging and cerebrospinal fluid assessment required separate approvals, each with different documentation rules. A similar pattern unfolded in metabolic clinics, where eligibility for GLP-1 therapy hinged on body mass index and comorbidities rather than biomarkers that could forecast response, durability, or adverse effects. In pediatrics, families navigating a rare disease evaluation still cycled through specialty referrals before comprehensive sequencing, even as gene therapies appeared on formularies. The science was no longer the limiting factor; the system was.
Misaligned Incentives and Coverage Paradoxes
Payment policy deepened the contradiction: expensive therapies sailed through coverage while the lower-cost tests meant to guide their use faced denials, caps, or opaque criteria. In Alzheimer’s disease, a blood test priced around $1,000 could help identify candidates for disease-modifying antibodies and prompt safety monitoring, yet coverage remained inconsistent or absent. Meanwhile, therapies costing about $30,000 per year were commonly authorized once a diagnosis was inferred by clinical criteria alone. Similar incoherence surfaced with pharmacogenomics, where labs documented reductions in adverse drug events but payers questioned utility, even as hospitalizations from preventable interactions remained far costlier than the underlying test.
Population-level cost anxiety often drove these policies. If everyone over 50 sought Alzheimer’s screening, the aggregate testing bill looked daunting, so insurers imposed prior authorization, narrow indications, or site-of-service limits. However, those controls risked pushing care into costlier territory—paying for ineffective treatment, managing preventable side effects, or repeating inconclusive evaluations. In metabolic medicine, surging demand for GLP-1 agonists collided with the absence of validated predictors of response, inviting broad prescribing and stop-start coverage cycles that frustrated patients and clinicians alike. Equity suffered most. Patients with fewer resources were least able to pay out of pocket for tests or secure time off for multiple appointments, entrenching disparities in who benefited from “precision.”
Structural Drivers We Can Fix
The drivers were structural and therefore solvable. Regulatory silos treated therapeutics and diagnostics as separate lanes: drugs frequently received expedited review, while companion or enabling tests navigated a slower path, especially when offered as laboratory-developed tests rather than cleared in vitro diagnostics. That disconnect let a therapy reach the market before its guiding assay achieved broad access, leaving clinics to improvise with provisional criteria. Reimbursement models compounded the drag. Payers often categorized diagnostics as cost centers rather than value enablers, dampening investment in clinical utility studies and the real-world data that regulators and insurers said they wanted. As federal leaders emphasized, powerful tools stayed “on the shelf” when the evidence pipeline stalled.
Correctives were within reach if stakeholders chose alignment over fragmentation. Joint evaluation of drugs and their guiding tests would reduce launch-to-utility gaps; linked coverage policies could condition therapy payment on use of validated diagnostics, while also funding evidence generation to refine indications. Payment models that recognized downstream savings—avoided adverse events, fewer futile treatments, and more precise monitoring—could raise reimbursement to sustainable levels for high-value assays. Health systems could embed testing into standard workflows by triggering pre-visit labs, co-locating phlebotomy with specialty clinics, and automating documentation for prior authorization. For GLP-1s, pragmatic trials paired with biobanking could yield predictive panels; in rare disease, first-line exome or genome sequencing could shorten odysseys and guide therapy earlier.
What This Means for Patients and Systems
For patients, the stakes included safety, speed, and confidence. Without appropriate testing, candidates for Alzheimer’s therapies might receive infusions without biomarker confirmation, increasing the risk of edema or microhemorrhage that monitoring protocols were designed to catch. People with cardiometabolic disease might rotate through GLP-1s without insight into who would sustain weight loss or glycemic control, exhausting time-limited coverage windows before identifying a durable regimen. Families facing suspected rare disorders—often juggling travel, time off, and childcare—encountered repeated evaluations that a single comprehensive test could have streamlined. Each missed or delayed diagnostic step added friction to a journey already defined by uncertainty.
Health systems absorbed parallel strain. Pharmacy budgets swelled under broad utilization, yet savings from precise targeting went unrealized when tests were treated as optional add-ons. Clinicians spent hours on documentation and appeals rather than care, an opportunity cost that burned out teams and delayed treatment starts. Data fragmentation—results parked in external portals, mismatched patient identifiers, inconsistent coding—hampered quality measurement and obscured the true value of diagnostics in routine care. Crucially, inequity widened. When coverage faltered, access followed income and geography, so the patients most likely to benefit from structured, test-informed decisions were least likely to receive them. Precision stalled without its prerequisite: reliable, timely measurement.
The Path Forward: From Promise to Practice
Actionable reforms flowed from the evidence. Regulators could pilot synchronized reviews, pairing therapy approvals with clear pathways for associated diagnostics, including conditional coverage that funded post-market studies. Payers could adopt value-based arrangements that tied reimbursement to verified test use and outcome benchmarks, reducing uncertainty while rewarding real-world utility. Health systems could redesign clinics so testing occurred ahead of decision points—pre-visit draws, standing orders keyed to problem lists, and embedded navigators for prior authorization. Specialty societies could publish living guidance that specified which tests were essential, which were optional, and how to interpret borderline results, cutting through ambiguity at the bedside.
Clinicians, meanwhile, had practical levers. Ordering panels that bundled necessary markers, documenting clinical utility in notes that mirrored payer language, and enrolling patients in registries or pragmatic trials strengthened the evidence base. For GLP-1 therapy, teams could track early digital phenotypes—wearable-derived activity, meal timing, and glycemic variability—to build local prediction rules while broader biomarkers matured. Memory clinics could standardize blood-based biomarker pathways with reflex imaging only when needed, accelerating safe therapy starts. The imperative had been clear: the system would treat diagnostics as integral infrastructure, not ancillary services. The next steps were to align approvals, coverage, and workflows so that tests unlocked, guided, and safeguarded the therapies funded at scale.
