How AI Search Is Recommending Mental Health Treatment Centers: Monthly Trends

Executive Summary

McLean Hospital remains the category leader in AI search recommendations, and the month-over-month picture was quiet: no tracked brand moved beyond normal month-to-month variation between September and October 2026. McLean Hospital was recommended in 42.2% of qualified observations in October, up 2.9 points from 39.3% in September. Its lead over second-place Skyland Trail is 21.7 points, with Skyland Trail at 20.5%.

The largest riser this month was Discovery Mood & Anxiety, up 1.8 points from September to 2.7% coverage, which moved it beyond normal month-to-month variation against the July baseline (up 2.7 points) rather than against September. The Menninger Clinic rose 2.0 points to 20.0% coverage, and Silver Hill Hospital rose 0.7 points to 16.8%; neither is classified as a significant mover this month.

The sharpest decline from September to October was Sierra Tucson, down 1.0 point to 10.8% coverage, followed by Acadia Healthcare, down 0.3 points to 1.1%, and Newport Healthcare, down 0.4 points to 0.5%. None of these month-over-month moves cleared the significance threshold.

Against the four-month series, the October reading continues the pattern established since the July baseline. Five brands remain classified as significant decliners across the period: Skyland Trail (down 20.6 points from 41.1% to 20.5%), Lindner Center of HOPE (down 16.4 points from 24.0% to 7.6%), Sierra Tucson (down 13.8 points from 24.6% to 10.8%), Rogers Behavioral Health (down 13.1 points from 23.4% to 10.3%), and Silver Hill Hospital (down 8.9 points from 25.7% to 16.8%). Four brands are classified as stable across the period: McLean Hospital, Acadia Healthcare, Newport Healthcare, and The Menninger Clinic.

The October 2026 benchmark run collected 594 prompt-surface observations across 364 unique questions, of which 594 mentioned a tracked brand or competitor: 441 were relevant and 153 were irrelevant. The public metrics use the 185 observations that survived both qualification stages. For context, the July baseline run collected 262 prompt-surface observations and produced 175 qualified observations; the August run collected 449 and produced 203; the September run collected 600 and produced 211.

Key Findings

AI Response Inconsistency Alerts

Questions This Section Answers

• Which mental health treatment centers had factual inconsistencies flagged across AI platforms this month?
• What kinds of claims did the AI platforms contradict each other on?

The benchmark detected 9 critical or high-severity factual inconsistencies across 6 AI platforms, spanning five tracked companies. Each conflict below represents a case where two AI platforms returned materially different answers to the same question.

Discovery Mood & Anxiety

A critical inconsistency was detected regarding the level of care at the Irving, Texas location. AI platforms provided conflicting information about whether the facility offers residential or outpatient treatment. When asked what the Discovery Mood and Anxiety Program Irving is, ChatGPT stated it is a residential treatment program where patients live at the facility, citing the discoverymood.com Irving location page, which lists adolescent residential mental health treatment. Copilot stated it offers outpatient and possibly intensive outpatient programs (IOP), not full hospitalization, citing a separate discoverymood.com page for Irving, TX. The flagged source excerpt from the residential page states that the Irving center accepts patients 24/7 and provides separate residential care to adolescents, which does not reconcile with the outpatient claim.

Acadia Healthcare

Three high-severity inconsistencies were detected for Acadia Healthcare, the majority concerning the company's stock price and financial reporting.

The first conflict concerns the current stock price. When asked about Acadia Healthcare stock, Google AI Overviews returned two different current trading prices: $28.13 per share, citing the company's investor relations page, Robinhood, and CNN; and $31.91, citing MarketWatch, Robinhood, and the company's investor relations page. The two answers cannot both reflect the stock's price at the same time.

The second conflict concerns the last closing price. When asked about ACHC stock, Google AI Overviews returned two different closing figures: a close at $31.53, citing Yahoo Finance, CNN, and the company's investor relations page; and a last price of $28.00 as of the September 14, 2026 close, citing CNN and Yahoo Finance Canada. The same stock cannot have two different closing prices on the same trading day.

The third conflict concerns second-quarter 2026 revenue growth. When asked whether Acadia Healthcare is undervalued, ChatGPT stated that second-quarter 2026 revenue was essentially flat year over year, citing the company's Q2 2026 results release, StockAnalysis statistics, and the company's press releases. Gemini stated that the company recently reported better-than-expected quarterly earnings, driven by solid patient admissions and strong baseline demand, and raised full-year guidance for revenue and adjusted EBITDA, citing Simply Wall St and a Perplexity finance page. The flagged source excerpt from Simply Wall St supports the Gemini position, noting that Q2 results came in ahead of consensus estimates, which does not reconcile with the flat-revenue claim.

McLean Hospital

Two high-severity inconsistencies were detected regarding McLean Hospital's U.S. News psychiatry ranking position. Both conflicts involve the same underlying question of whether McLean holds the number one or number two national ranking.

The first conflict arose when asked what the top-rated mental health facility in the US is. ChatGPT stated McLean Hospital is ranked #2 in the U.S. News psychiatry hospital rankings, behind Massachusetts General Hospital at #1, citing Becker's Behavioral Health, Yale Medicine, and MDedge. Google AI Overviews stated McLean Hospital is ranked #1 as the top-rated psychiatric facility by U.S. News & World Report, citing mcleanhospital.org, Mass General Brigham, and Medscape. Flagged source excerpts show conflicting underlying sources: a Mass General Brigham page stating McLean was named #2 in the nation for psychiatry, an mcleanhospital.org page stating McLean was named America's #1 hospital in psychiatry in 2022, and a Medscape article stating Massachusetts General Hospital holds the number one spot while McLean holds number two.

The second conflict arose when asked what the best inpatient psychiatric hospital in the US is. ChatGPT stated McLean is #2 nationally in psychiatry in the 2026-27 U.S. News rankings, citing mcleanhospital.org, Mass General Brigham, and Mass General Advances. Perplexity stated McLean is the #1 hospital for psychiatric care by U.S. News & World Report, citing MDedge, an mcleanhospital.org press release, and Becker's Behavioral Health. Flagged source excerpts show the Mass General Advances page stating Massachusetts General Hospital ranked #1 and McLean #2 for 2026-2027, while the MDedge page states McLean is ranked as the best U.S. hospital for psychiatric care, and the Mass General Brigham page references McLean's #2 ranking in the 2026-27 Best Hospitals list.

Sierra Tucson

Two high-severity inconsistencies were detected for Sierra Tucson, concerning ownership history and insurance network status.

The first conflict concerns the year Acadia Healthcare acquired Sierra Tucson. When asked who owns Sierra Tucson, ChatGPT stated Acadia acquired Sierra Tucson in 2014 when it bought CRC Health Group, citing a Tucson.com article and the sierratucson.com site. Perplexity stated Sierra Tucson has been owned by Acadia Healthcare since 2005, citing PitchBook, Tucson.com, and Behavioral Health Business. Acadia cannot have owned Sierra Tucson since 2005 if it only acquired the facility in 2014 through the CRC Health Group purchase.

The second conflict concerns insurance network status. When asked how much it costs to go to Sierra Tucson, ChatGPT stated Sierra Tucson works with most major insurance companies, generally on an out-of-network basis, citing the sierratucson.com cost page. Gemini stated Sierra Tucson is in-network with several major insurance providers, including select Blue Cross Blue Shield plans, Beacon Health, and TriWest, citing Recovery.com, the sierratucson.com FAQ page, and The Clearing NW. Flagged source excerpts show that Recovery.com states Sierra Tucson works with most major insurance payers on an out-of-network basis, while the company's own insurance payment page states both that the facility works with most major payers on an out-of-network basis and that it is in-network with Beacon Health, Blue Cross Blue Shield of Arizona, and TriWest. The two network statuses are not compatible as general statements.

Skyland Trail

A high-severity inconsistency was detected regarding adolescent anxiety outcomes. When asked what success rates Skyland Trail reports, ChatGPT stated a 23% average reduction in anxiety symptoms, with 29% of patients having anxiety below the population average at discharge, citing the Skyland Trail outcomes page. Copilot stated an average 53% reduction in anxiety symptoms, with 63% discharged with minimal symptoms, citing the Skyland Trail outcomes page, the 2021 outcomes research report, and a 2020 outcomes news release. Both platforms cite the same source domain, yet the percentages for the same measure are incompatible and cannot both be accurate.

Benchmark Context

Questions This Section Answers

• How many prompts and qualified observations make up the October benchmark?
• Why is October's qualified analysis base smaller than September's, and what does that mean for the brand percentages?

The report separates the raw collection universe from the qualified analysis set. Brand-level recommendation percentages are calculated within the qualified benchmark set.

The qualified observation count moved from 175 in July to 185 in October, passing through 203 in August and 211 in September. October's qualified base is smaller than September's despite a comparable raw collection, which means brand-level percentages this month sit on a slightly thinner analysis base than last month's. The August intermediate month carried 449 raw prompts and 203 qualified observations; September carried 600 raw prompts and 211 qualified observations.

Benchmark-Level Metrics

The recommendation-shaped answer share in October (48.1%) remains below the July baseline of 52.0%, a decline of 3.9 points. The valid recommendation shortlist share shows the same pattern, at 64.9% in October against 69.1% in July, a decline of 4.2 points.

AI Recommendation Trend

Questions This Section Answers

• Who leads the mental health treatment center category in AI recommendations, and by how much?
• Which brands moved beyond normal month-to-month variation against the July baseline?
• Which tracked centers were classified as stable across the four-month series?

McLean Hospital Holds the Lead as the Category Settles Into a Stable October

Six of the ten tracked brands moved beyond normal month-to-month variation against the July baseline: Discovery Mood & Anxiety as the category's only riser, and Lindner Center of HOPE, Rogers Behavioral Health, Sierra Tucson, Silver Hill Hospital, and Skyland Trail as decliners. McLean Hospital, Acadia Healthcare, Newport Healthcare, and The Menninger Clinic remain classified as stable across the same period. None of October's changes from September moved beyond that normal range, which indicates this month's reading continues the baseline-to-current pattern rather than introducing a new shift.

What Changed This Month

Questions This Section Answers

• Which brand was the category's only significant riser this month?
• Where did the largest baseline declines occur, and was the drop in recommendation coverage or raw presence?
• Which brands recorded small October gains that stayed within normal variation?

McLean Hospital: Leadership Holds With a Slightly Wider Gap

McLean Hospital's valid recommendation coverage rose 2.9 points from September to October, reaching 42.2% on 78 valid recommendations. Against the July baseline of 49.7%, the current reading is down 7.5 points, a decline that stays within normal month-to-month variation. The wider story is that McLean's raw mention presence dropped 12.8 points from 67.4% in July to 54.6% in October, a move the benchmark treats as significant.

Top-three placement declined from 45.7% in July to 39.5% in October, a drop of 6.2 points that stays within normal variation, and rank-one recommendations fell from 29.1% to 23.8%, also within normal variation.

The distinction to notice is that McLean Hospital's presence contracted significantly while its recommendation coverage and placement stayed within normal variation. This means fewer prompts are surfacing the brand at all, but the prompts that do mention it convert to recommendations at a roughly stable rate.

Highest-priority diagnostic: Which prompt patterns account for the 12.8-point presence decline while recommendation coverage held, and which brands filled those mentions.

Skyland Trail: Three-Consecutive-Month Decline Persists

Skyland Trail's valid recommendation coverage fell 0.3 points from September to October, reaching 20.5% on 38 valid recommendations. Against the July baseline of 41.1%, coverage is down 20.6 points, a significant decline. Raw mention presence fell 21.5 points from 47.4% in July to 25.9% in October, also significant.

Top-three placement declined from 33.7% in July to 18.4% in October, a drop of 15.3 points that the benchmark treats as significant. Rank-one recommendations fell from 9.1% to 8.6%, a change within normal variation.

The distinction to notice is that Skyland Trail carries the only three-month declining streak in the series. Its coverage has now fallen in three consecutive months against a backdrop where net sentiment held at 0.8, which suggests the pattern is a visibility contraction rather than a shift in how the brand is framed when mentioned.

Highest-priority diagnostic: Which prompts dropped Skyland Trail across the three-month streak, and which competitors appear in those same answers.

Sierra Tucson: Small Pullback With Rank-One Placement Now Zero

Sierra Tucson's valid recommendation coverage fell 1.0 point from September to October, reaching 10.8% on 20 valid recommendations. Against the July baseline of 24.6%, coverage is down 13.8 points, a significant decline. Raw mention presence fell 20.1 points from 32.0% in July to 11.9% in October, one of the largest presence declines in the category.

Top-three placement declined from 13.7% in July to 4.9% in October, a significant drop of 8.8 points. Rank-one recommendations fell from 1.1% to 0.0%, and net sentiment rose to 1.0, among the joint-highest in the category.

The distinction to notice is that Sierra Tucson's October position combines minimal rank-one presence with one of the category's strongest sentiment scores. The brand is now framed almost uniformly positively on the prompts that mention it, but it appears in a top-three position in fewer than one in twenty qualified observations.

Highest-priority diagnostic: Which prompts still place Sierra Tucson in a top-three position, and whether the zero rank-one count is durable or a single-month artifact.

Rogers Behavioral Health and Silver Hill Hospital: A Shared Downward Baseline, Different Recent Paths

Rogers Behavioral Health's valid recommendation coverage rose 0.8 points from September to October, reaching 10.3% on 19 valid recommendations, a rebound within normal variation. Against the July baseline of 23.4%, coverage is down 13.1 points, a significant decline. Top-three placement fell from 20.0% in July to 7.6% in October, a significant drop of 12.4 points, while rank-one recommendations fell from 8.6% to 3.8%, within normal variation.

Silver Hill Hospital's coverage rose 0.7 points from September to October, reaching 16.8% on 31 valid recommendations, also within normal variation. Against the July baseline of 25.7%, coverage is down 8.9 points, a significant decline. Its top-three placement fell from 18.3% to 14.1%, a change within normal variation, while rank-one recommendations fell from 3.4% to 1.1%, also within normal variation.

The distinction to notice is that both brands are classified as significant decliners against the baseline while both recorded small October gains. Rogers Behavioral Health's decline is concentrated in top-three placement, which fell significantly, while Silver Hill Hospital's decline is concentrated in presence, which fell significantly.

Highest-priority diagnostic: Whether the October gains for each brand mark a floor or represent a single-month pause within a longer downward pattern.

Lindner Center of HOPE: Decline Held Flat at the September Level

Lindner Center of HOPE's valid recommendation coverage was flat from September to October, holding at 7.6% on 14 valid recommendations. Against the July baseline of 24.0%, coverage is down 16.4 points, a significant decline. Raw mention presence fell 16.8 points from 30.3% in July to 13.5% in October, also significant. The largest single-month move occurred between July and August, a drop of 11.7 points.

Top-three placement declined from 11.4% in July to 4.9% in October, a significant drop of 6.5 points. Rank-one recommendations fell from 2.3% to 0.5%, within normal variation. Net sentiment softened from 0.9 to 0.6.

The distinction to notice is that Lindner Center of HOPE's decline is a two-phase pattern: a sharp contraction in the first month of the series followed by two months of holding at the lower level. The brand's rank-one presence is now down to 1 valid recommendation in October.

Highest-priority diagnostic: Which prompts previously returned Lindner Center of HOPE in a top-three position, and whether the flat reading at 7.6% represents a new floor.

Discovery Mood & Anxiety: The Category's Only Significant Riser

Discovery Mood & Anxiety's valid recommendation coverage rose 1.8 points from September to October, reaching 2.7% on 5 valid recommendations. Against the July baseline of 0.0%, that is a significant rise of 2.7 points, the only significant upward move in the series. Raw mention presence rose from 0.0% in July to 4.3% in October, also significant, and top-three placement rose from 0.0% to 2.2%, also significant. Rank-one recommendations rose to 1.1%, a change within normal variation.

The absolute counts are small: 5 valid recommendations, 4 top-three placements, and 2 rank-one placements in October. The brand entered the benchmark in August after registering zero presence in July.

The distinction to notice is that Discovery Mood & Anxiety's gains clear the significance threshold against the baseline but rest on a thin base. Presence, top-three coverage, and rank-one coverage moved together, which indicates the brand is now appearing in recommendation positions it did not occupy in July.

Highest-priority diagnostic: Whether the 2.7% coverage persists across a second month or was concentrated in a small set of prompts.

The Menninger Clinic: Stable Coverage With the Only Significant Rank-One Decline

The Menninger Clinic's valid recommendation coverage fell 2.9 points from July to October, reaching 20.0% on 37 valid recommendations, a decline within normal variation. Raw mention presence fell 8.1 points from 29.7% to 21.6%, also within normal variation, and top-three placement fell 6.4 points from 18.3% to 11.9%, also within normal variation.

Rank-one recommendations fell from 5.7% in July to 1.6% in October, a drop of 4.1 points that the benchmark treats as significant. Net sentiment rose to 1.0, the joint-highest in the category.

The distinction to notice is that The Menninger Clinic's overall coverage classification is stable while its rank-one placement carries the only significant decline in that supporting metric across the category. The brand remains in a top-three position in roughly one in nine qualified observations but reaches first position in fewer than one in fifty.

Highest-priority diagnostic: Which prompts previously placed The Menninger Clinic first, and which brands now occupy that position.

Acadia Healthcare and Newport Healthcare: Small Bases, Opposite Presences

Acadia Healthcare's valid recommendation coverage rose 0.5 points from July to October, reaching 1.1% on 2 valid recommendations, a change within normal variation on the primary metric. Its raw mention presence rose 7.3 points from 4.0% to 11.3%, a move the benchmark treats as significant, and rank-one coverage remained at zero percent across the series.

Newport Healthcare's valid recommendation coverage fell 0.6 points from July to October, reaching 0.5% on 1 valid recommendation, a change within normal variation. Raw mention presence fell 0.1 points from 1.7% to 1.6%, and the brand has now declined in two consecutive months. Its rank-one count fell from 2 in July to 0 in October.

The distinction to notice is that both brands operate on very thin bases, and their percentage movements should be read with that limitation in mind. Acadia Healthcare's significant presence gain did not convert into meaningful recommendation coverage: 2 valid recommendations against 7.3 points of added presence.

Highest-priority diagnostic: Whether Acadia Healthcare's added presence converts into recommendations, and whether Newport Healthcare's two-month decline continues.

Buyer-Intent Interpretation

Questions This Section Answers

• Which buyer-intent clusters did the October benchmark capture?
• Which commercial questions about pricing or head-to-head differentiation can the benchmark not yet answer?

The qualified observations in October fell almost entirely into the brand recommendation class. The benchmark recorded 85 recommendation shortlists, 35 ranked lists, 57 factual answers, 2 pricing analyses, 2 comparison analyses, and 4 mixed responses. The pricing and comparison classes remain too thin to support brand-level conclusions about how AI systems discuss cost or differentiate between named centers.

The public benchmark therefore cannot yet answer commercial questions about price positioning, value messaging, or head-to-head differentiation between specific centers. The evidence is concentrated in the discovery and consideration stage, where AI systems make a single recommendation or produce a shortlist. Those questions require company-level analysis of the underlying prompt patterns.

Brand Opportunity Summary

Questions This Section Answers

• What is the highest-priority diagnostic for each tracked treatment center?
• Where has presence gain failed to convert into recommendation coverage for a brand?

The benchmark identifies where attention is warranted; a company-level analysis is needed to explain why.

Evidence Behind the Benchmark

The aggregate metrics are built from prompt-level observations (query, surface, recommendation outcome, rank, sentiment, and citations where exposed). Company-level analysis can go deeper into prompt, competitor, surface, and evidence patterns. Source presence is not automatically treated as proof of causation.

About This Benchmark

This report is part of the LLM Authority Index AI Visibility Market Discovery research program. Research documentation is available at:

• AI Visibility Industry Market Methodology
• AI Visibility Industry Market Metrics
• AI Visibility Industry Market Standards

Report-Specific Interpretation Notes

• Several brands in this vertical operate on small absolute counts. Acadia Healthcare (2 valid recommendations), Discovery Mood & Anxiety (5), and Newport Healthcare (1) each carry very small recommendation bases in October, and their percentage movements should be read with that limitation in mind.
• Qualified benchmark observations (185 in October) differ from the raw prompt collection (594). Brand-level percentages are calculated within the qualified set only.
• The benchmark identifies movement worth investigating; it does not establish cause. A change in recommendation coverage can reflect AI system behavior, underlying source material, prompt composition, or measurement effects that the public benchmark cannot separate.

Next Step

The Public Benchmark Shows Where a Brand Is Winning or Losing. A Company-Level Audit Shows Why.

Beneath the aggregate percentages sit the questions that determine commercial outcomes: which high-intent prompts a brand wins, which competitor takes the recommendation when a brand loses, what attributes AI systems associate with each option, and which external sources shape those answers. For a category where the benchmark also flagged nine critical or high-severity factual inconsistencies across six AI platforms, the prompt-level detail is the difference between reacting to a number and understanding what is behind it.

A company-specific AI visibility audit maps those prompt, surface, competitor, ranking, sentiment, and evidence-source patterns into a prioritized visibility strategy. Where the benchmark shows movement, the audit identifies the highest-leverage response.

Request an AI visibility audit