Executive Summary
What’s changing
Telehealth utilization, which surged during 2021-2022, has been receding from its peak. The decline is uneven: mental health teleconsultations remain elevated relative to pre-surge levels, while telehealth use for acute care and routine primary care has fallen sharply, suggesting a reversion toward in-person settings for those categories.
Why it matters
Healthcare organizations, payers, and digital health investors built infrastructure and financial models on the assumption that virtual care adoption would continue expanding uniformly across care types. A category-specific pullback changes unit economics, capacity planning, and the credibility of growth projections tied to telehealth as a single undifferentiated market.
Who is affected
Health systems, payers, telehealth platform operators, employer-sponsored health benefit programs, behavioral health providers, and primary/urgent care networks that expanded virtual offerings during the pandemic-era surge.
Expected evolution
If this bifurcation persists, telehealth is likely to consolidate into a durable channel primarily for behavioral and mental health, with acute and routine primary care reverting largely to in-person delivery except where access barriers are severe. Confirmation from additional data sources over the coming months would strengthen this reading considerably.
Key Takeaways
- —Telehealth visit volumes peaked in 2021-2022 and have since declined from that high point.
- —Mental health telehealth use has remained elevated even as overall telehealth volumes fell.
- —Acute care telehealth visits have fallen sharply, indicating a return to in-person channels for time-sensitive conditions.
- —Routine primary care telehealth visits have also declined sharply, suggesting patients and providers favor in-person interaction for ongoing management.
- —The pattern implies telehealth adoption is not uniform across care types but is settling into a narrower, behavioral-health-centric niche.
- —This is currently a single-source, single-evidence observation and has not yet been corroborated by independent data points.
Behavioural Analysis
Previous behaviour
During 2021-2022, patients and providers broadly adopted telehealth across care categories, including acute, primary, and behavioral health, largely driven by pandemic-era access constraints and temporary regulatory and reimbursement flexibilities.
↓
Emerging behaviour
Utilization has since diverged by care type: mental health telehealth visits remain elevated, while acute care and routine primary care telehealth visits have fallen sharply, pointing to a return toward in-person care for conditions requiring physical examination or continuity-based relationships.
↓
What is driving the change
Plausible drivers include the retreat of pandemic-era necessity as in-person access normalized, patient and clinician preference for physical exams in acute and routine primary care, persistent stigma or scheduling friction that makes virtual formats particularly suited to behavioral health, and possible shifts in reimbursement policy or payer coverage that differentially favor or restrict telehealth by care category.
↓
Evidence supporting the change
The reading rests on a single reported observation (evidence_count: 1) drawn from one source (source_count: 1), with no supporting signal cluster (signal_count: null) and no time gap between creation and update to demonstrate persistence. The directional claim is internally coherent — a peak followed by category-specific decline — but currently lacks independent replication.
Source Overview
Evidence points
5
Independent sources
2
Per-source attribution (platform, publication) is not yet captured at the observation level — the figures above are the real aggregate counts detected for this item.
Geographic Distribution
Geographic attribution is not yet captured in the data pipeline for this item.
Evolution Timeline
First observed
July 23, 2026
Last reinforced
July 26, 2026
Published
July 23, 2026
Confidence Assessment
53
/ 100 overall confidence
Evidence consistency
45
The single evidence point presents an internally coherent narrative (peak, then differentiated decline), but with only one recorded observation there is no way to cross-check internal consistency against other data.
Source diversity
15
Source_count of 1 relative to evidence_count of 1 indicates no independent source diversity at all; the observation reflects a single vantage point.
Time consistency
10
Created_at and updated_at are identical, meaning there is no observed persistence over time yet for this specific signal entry.
Independent confirmation
10
Signal_count is null, indicating this is a standalone signal with no supporting pattern or independent corroboration; confidence here is scored conservatively low as instructed.
Strategic Implications
For CEOs
Executives leading health systems or digital health portfolios should reassess whether telehealth revenue and capacity assumptions still hold uniformly across service lines, given that the growth story now appears concentrated in behavioral health rather than acute or primary care.
For Founders
Founders building virtual-first care companies should reconsider whether their core wedge into acute or routine primary care remains defensible, or whether the more durable opportunity lies in behavioral health-specific virtual delivery models.
For Investors
Investors evaluating digital health assets should stress-test valuations that assume continued broad-based telehealth growth, since the underlying utilization data suggests the addressable market may be narrower and more segment-specific than originally modeled.
For Product Teams
Product teams should examine usage funnels by care category rather than aggregate telehealth metrics, since blended dashboards may mask a sharp divergence between resilient behavioral health demand and declining acute/primary care demand.
For Marketing
Marketing strategies built around telehealth convenience messaging for acute or routine care may be losing resonance with patients who are reverting to in-person preferences, while messaging around accessible, stigma-reducing mental health care may retain stronger traction.
For Innovation
Innovation teams should treat this as an early signal to prioritize R&D investment toward behavioral health-specific virtual care features and workflows rather than continuing to generalize telehealth capability across all care types equally.
For Strategy
Strategy leads should flag this as a single, unconfirmed data point requiring monitoring rather than immediate resource reallocation, while beginning scenario planning for a bifurcated telehealth market structure segmented by care acuity and modality fit.
Full Research
Overview
The rapid expansion of telehealth during 2020-2022 was widely treated as a structural, permanent shift in how care is delivered. This signal complicates that narrative. It indicates that telehealth visit volumes peaked in the 2021-2022 window and have since declined, but not uniformly. Mental health telehealth utilization has remained elevated relative to pre-surge baselines, while acute care and routine primary care telehealth visits have fallen sharply. This suggests the pandemic-era telehealth boom is resolving into a more differentiated, care-type-specific equilibrium rather than a uniform new normal.
From Uniform Surge to Differentiated Retrenchment
The initial telehealth surge was driven by a mix of necessity — reduced in-person access during acute public health constraints — and temporary regulatory flexibilities that made virtual visits reimbursable and operationally viable across nearly all care categories. In that environment, acute care, routine primary care, and behavioral health telehealth visits likely rose together, driven by the same access constraint rather than by category-specific preference.
What this signal describes is the unwinding of that uniform surge. As in-person access normalized, the categories retreated at different rates. Acute care — conditions requiring rapid diagnosis, physical examination, or urgent intervention — appears to have reverted most sharply to in-person settings, which is consistent with the inherent limitations of remote assessment for time-sensitive or physically diagnosable conditions. Routine primary care, which often involves preventive screening, physical exams, and longitudinal relationship-building, also shows a sharp decline, suggesting that patients and providers alike may have concluded that in-person visits better serve these functions once the option was fully available again.
Mental health telehealth, by contrast, has remained elevated. This is a meaningfully different trajectory and points to structural rather than merely pandemic-driven demand. Behavioral health care is less dependent on physical examination, more dependent on consistent access and reduced friction, and historically has faced greater barriers related to stigma, geographic availability of specialists, and scheduling constraints. Virtual delivery directly addresses several of these historical frictions in a way that may not apply as strongly to acute or routine primary care.
Why the Distinction Matters
Much of the early digital health narrative treated telehealth as a single undifferentiated category experiencing secular growth. Investment theses, health system capacity planning, and payer reimbursement policy were often built on this aggregate view. If the underlying reality is instead a bifurcated market — durable growth in behavioral health telehealth alongside reversion to in-person care for acute and routine primary care — then strategies calibrated to an undifferentiated telehealth growth curve are likely miscalibrated.
This has direct implications for capacity allocation. Health systems that built out generalized virtual care infrastructure across all specialties may find utilization concentrating disproportionately in behavioral health, while acute and primary care virtual capacity goes underused. Reimbursement policy design, similarly, may need to differentiate more explicitly by care type rather than applying blanket telehealth coverage rules, if utilization patterns themselves are differentiating.
For digital health companies, the implication is more existential. Platforms whose core value proposition was virtual-first acute or primary care may be swimming against a behavioral tide, while companies focused specifically on behavioral health telehealth may be riding a more durable wave. This reframes competitive positioning within the digital health sector: the addressable market is not "telehealth" as a monolith, but a set of distinct care-type markets with very different retention dynamics.
Plausible Drivers
Several factors plausibly explain the divergence, though it is worth noting that this signal is based on a single observation and does not itself specify causal mechanisms in granular detail. First, the return of normalized in-person access removes the primary structural driver of the original acute and primary care telehealth surge — necessity. Once patients can again see a physician in person without friction, many appear to prefer doing so for conditions where a physical exam adds diagnostic value or where continuity of an in-person relationship is valued.
Second, behavioral health has structural characteristics that make virtual delivery persistently attractive: reduced stigma associated with attending sessions from a private location, easier scheduling around work and personal obligations, and — in many regions — a shortage of in-person behavioral health specialists that virtual delivery can partially offset regardless of pandemic conditions. These are not necessarily pandemic artifacts; they are pre-existing frictions that telehealth addresses independent of acute public health circumstances.
Third, reimbursement and regulatory policy may be evolving in ways that differentially support behavioral telehealth relative to acute or primary care virtual visits, though the specific policy mechanisms are not detailed in the available evidence and should not be assumed without further confirmation.
Evidence Base and Its Limits
This observation currently rests on one recorded data point from one source. There is no accompanying signal cluster, and the created and updated timestamps are identical, meaning there is no track record yet of this pattern persisting or recurring across additional observations or time periods. The directional claim — peak, followed by uneven decline — is internally plausible and consistent with widely discussed shifts in care delivery, but it has not yet been independently corroborated within this evidence base.
This matters for how the signal should be used. It is reasonable to treat this as an early, directionally coherent observation worth monitoring, but not yet as a validated trend on which to base major resource reallocation. The absence of independent source corroboration and the lack of a supporting signal cluster mean the confidence assigned to this observation should remain moderate, reflecting genuine uncertainty about whether the pattern will hold, deepen, or reverse as more data becomes available.
Trajectory
Looking forward, the most plausible scenario is a continued bifurcation: telehealth consolidating as a durable, perhaps even growing, channel for behavioral health care, while acute and routine primary care telehealth settles at a lower baseline utilization than its 2021-2022 peak, used primarily for triage, follow-up, or access-constrained populations rather than as a substitute for in-person primary care generally. Confirmation of this trajectory would require additional independent evidence — ideally from multiple sources and repeated observations over time — showing that the mental health telehealth elevation is sustained rather than transitional, and that the acute/primary care decline is not merely a temporary correction that could partially reverse.
Organizations planning multi-year digital health strategy should treat this signal as an early flag to disaggregate their own utilization data by care type immediately, rather than waiting for broader confirmation, since the cost of continuing to plan around an undifferentiated telehealth growth assumption may be higher than the cost of premature specialization.
