what workslast verified 2026-09-24

how to find the right therapist

you type “how to find the right therapist” and the first thing offered is a quiz. ten questions, a spinner, a name. it feels like being matched. it is not: nobody has published a randomized test of any consumer quiz, and the one matching method that has been tested matched people to therapists’ measured results, which the quizzes do not have. here is what the research says a therapist is worth, what a quiz can and cannot do with it, and the way to find the right one that does not depend on a spinner.

how much the therapist matters

how important is it to find the right therapist?[1][2]

more than the type of therapy, less than the quiz pages imply. across 20 studies, who the therapist is explains about 5 percent of how people end up, 8 percent in randomized trials, and more for people who come in worse. five percent sounds small until you see the spread it hides: in 10,786 NHS patients, the recovery rate of the best of 119 therapists was 95.6 percent and the worst was 23.5 percent. same service, same referral pool, same forms.

does the type of therapy matter more?[12][25]

for depression, no. in a network meta-analysis of 198 trials and 15,118 adults, seven kinds of therapy all beat a waiting list and the differences between them ran from d 0.01 to 0.30, absent to small. the evidence is deeper for CBT, interpersonal and problem-solving therapy than for the rest, and social anxiety, PTSD and OCD did better with higher-intensity work in the biggest routine data set. the therapist you get varies more than the brand of therapy they do.

do credentials or years of experience predict a good therapist?[3][4][18]

barely. in 6,146 managed-care patients, therapist age, gender, experience and professional degree “accounted for little” of the difference between therapists. in the only large longitudinal study, 170 therapists followed for an average of 4.7 years got very slightly worse per year, not better, though they lost fewer clients early. the letters after the name tell you what they are allowed to do, not how their clients turn out.

so what does predict it?[5][6]

the therapist’s measured results with people like you. in 6,960 patients across 696 therapists, effectiveness was domain-specific: being good with depression said little about being good with substance use or with relationship problems. in a second data set, a therapist’s results on their first 30 cases predicted their next 30. that is the number that would let anyone match you to the right person, and it is the number no directory shows.

is it the relationship?[8][26]

partly, and you cannot measure it before you meet. across 295 studies and more than 30,000 patients, the working alliance correlates with outcome at r 0.28, the same online as in the room, and the APA task force rated nine relationship elements demonstrably effective regardless of therapy type. but the alliance is rated after a few sessions, by you. it is a thing to check, not a thing a form can predict.

what a therapist matching quiz can and cannot do

the position, plainly. no consumer matching quiz, directory filter or app intake questionnaire has published a randomized trial showing that the people it matches do better than people assigned by chance. we looked for one on PubMed and did not find it; if a platform has one, it has not put it in a journal. one algorithmic matching approach has been randomized, once: 218 adults at six community clinics were assigned either to a therapist whose measured track record across twelve problem domains fitted their three worst problems, or to whoever was next. the matched group improved more on general impairment (d 0.75), global distress (d 0.50) and their own worst problem (d 0.60).[7] that worked because each therapist had outcome data from fifteen or more previous clients in every domain, scored on a routine measure and risk-adjusted.[6] no directory collects that, and no platform publishes it. a quiz that asks you twelve questions and knows nothing measured about the therapist is doing half of a matching problem, the half that was never the hard part.

the other direction is real too. algorithms are not magic when they do have data: a treatment-selection algorithm for 996 adolescents choosing between two brief interventions matched no better than chance,[23] and the one that did work in adults, a machine-learning triage in 951 NHS patients, raised recovery from 45.1 to 52.3 percent by sending more people straight to full therapy, a seven-point gain for £104 a head, not a transformation.[22] and what a quiz can do is not nothing: getting the treatment you preferred cuts dropout (OR 1.79) with a small gain in outcome (d 0.28), across 53 studies and 16,000 clients.[10] same-race matching, the most studied identity preference, moves the preference itself (d 0.63) and your view of the therapist (d 0.32) far more than the outcome (d 0.09).[11] so use the quiz for what it can see. do not let it tell you it has seen the rest.

the claimwhat a quiz can usewhat it would needdoes it exist
takes your insurance, fits your budgetyour plan, their panel, their feenothing moreyes; this is what a filter is for
available this week, evenings, onlinetheir calendarnothing moreyes
shares your language, faith, identityyour stated preference, their profilenothing more for the preference itself. accommodated preferences cut dropout (OR 1.79) with a small outcome gain (d 0.28); same-race matching moves preference (d 0.63) far more than outcome (d 0.09)yes; a real, small effect, mostly on staying
specialises in anxietythe therapist’s self-description and trainingthat therapist’s measured results with anxious clients; results in one problem area predict results in another poorlynot on any consumer directory or app we could find published
matches you to the right therapist for youyour questionnaire answerseach therapist’s outcomes across 12 problem domains from 15 or more prior cases, then a randomized test of the matchonce, in one trial of 48 therapists; not on a consumer platform
predicts you will clicknothing measuredthe alliance, which is rated by you after a few sessionsno; alliance predicts outcome (r 0.28) but only once it exists

sources for the table: preferences[10], identity matching[11], domain-specific results[5][6], the matching trial[7], the alliance[8]. the directory-by-directory detail is on the psychology today alternatives page.

how to actually find the right therapist

six steps. the quiz does step two. you do the rest, and the rest is where the outcome is decided.

1. write the problem in one sentence, and pick the number that would show it changing[13]

not “anxiety” but “i have not slept through the night since march” or “i cancel plans three times a week.” then choose the measure: the PHQ-9 for low mood, the GAD-7 for worry, or a count of the thing itself. early change in the first four weeks is the strongest known predictor of how therapy ends (g 0.87), and you cannot see early change without a baseline. take the measure this week, before you have a therapist.

2. set the logistics filters, and only the logistics filters[19][20]

insurance or fee, in-network or sliding scale, days and times, online or in person, language, and any identity preference you hold. use any directory for this; it is what they are good at. do not use “specialises in” as a filter of results; it is a filter of training. online versus in person is a preference, not an efficacy decision: across 43 intervention studies video therapy was largely equivalent to in-person, and across 47 between-group studies the difference was negligible.

3. shortlist three, not one[18][9]

one in five clients drops out of therapy, and getting the treatment you wanted is one of the few things that halves that. three names lets you compare answers instead of grading one person against your imagination. if you would rather not do this by yourself, the free peer support in the veisund app is the place people compare notes on what they were told.

4. book 15-minute consultations and ask the six questions[15][16][17]

most therapists offer a free call. the six questions and the reason each works are on the therapy-length page: how long their average client stays; whether they measure progress with a form and will show you the scores; what changes if you have not moved by session eight; whether two sessions a week for a month beats one a week for two months; what an ending looks like; and whether the therapy they do is the therapy you asked for. the second question is the one that sorts the field. therapists who measure and act on the scores get slightly better results overall and much better results for the people not improving, and in a randomized trial of 77 therapists the not-on-track patients did better when the therapist was shown the numbers.

5. treat sessions one to three as the test[14][8]

take the measure again after session three or four. in 3,321 patients, a drop of six or more PHQ-9 points by session four made recovery six times more likely; a quarter of eventual recoverers had no early response, so no movement is a prompt to talk, not a verdict. rate the relationship honestly at the same point: do you feel understood, do you agree on what you are working on, do you agree on how. that is the alliance, and it is the most replicated predictor in the field.

6. the switch rule[25][24][18]

if the number has not moved by session four, say so in the room and ask what changes. if it has not moved by session eight, change something: the plan, the frequency, or the therapist. by session eight most people who are going to respond have started to; in 102,206 patients, 95 percent of responders to full CBT had responded by session fourteen, and the routine-care optimum is 4 to 26 sessions. switching is not failure. one in five people leave anyway; leaving on a plan is the version you get to choose.

the six consultation questions, each with its evidence, are on how long should therapy take.

how do i know if my therapist is good for me

how do i know if my therapist is good for me?[13][8][16]

two checks, both yours to run. the number you chose in step one has moved by session four, or the therapist has noticed it has not and changed something. and you would rate the relationship well on three counts: you feel understood, you agree on the goals, you agree on the method. a therapist who measures nothing and never asks how the work is going for you is missing the two things the evidence says predict how this ends.

how do i know if my therapist is not right for me?[5][8]

no measured change by session eight with no change of plan; a therapy you did not ask for and were not told about; a relationship you would rate poorly on understanding, goals or method at session three. none of these is a moral judgment about the person. results are domain-specific: a therapist who is excellent with someone else’s problem can be ineffective with yours, and in the largest data set, depending on the problem domain, 33 to 65 percent of therapists were unclassifiable, ineffective or harmful.

when should i switch therapists?[14][25]

when the switch rule fires: no movement by session eight after you raised it at session four. earlier if the therapist will not measure, will not discuss the plan, or does a therapy you declined. later if you came in severe and your own therapist can show you the trend line bending, because the people who change slowest are the people who came in worst, and early non-response is a prompt, not a verdict.

how long does it take to find the right therapist?[9][10]

about three weeks of admin and four to eight sessions of test. the admin is filters, three calls, one booking. the test is the measure at session zero and session four. people who type this question are usually asking whether it is normal to try more than one; it is, and the research on preferences says getting what you asked for halves your chance of quitting.

therapist vs psychologist vs psychiatrist vs counselor

four of the ten things google completes after “therapist vs psychologist vs psychiatrist” add “counselor” or “psychotherapist” to the list, so here is the short answer. the titles describe training and what the licence permits. in the outcome data, degree and years in practice account for little of the difference between one clinician’s results and another’s.[3][4]

psychiatrist[21]

a physician (MD or DO) with residency training in psychiatry. can prescribe. some do psychotherapy; many do assessment and medication management in short visits. in a US survey, 55.3 percent accepted new privately insured patients against 88.7 percent of other physicians, so the insurance filter bites hardest here (2009 to 2010 data, the most recent in a journal we could verify).

psychologist

a doctorate (PhD or PsyD) in psychology, licensed by a state board. does therapy and psychological testing. cannot prescribe in most states.

therapist, counselor, psychotherapist

in the US, usually a master’s-level licence: LCSW, LPC, LMHC, LMFT and their state variants. does therapy. “therapist” and “psychotherapist” are not protected titles everywhere, so check the licence number on the state board, which is public. this is most of the people in any directory.

nothing on this page is advice about starting, stopping or changing a medication. if a prescriber is part of your care, that conversation is with them.

how much does therapy cost, and what if you cannot afford it

the cost question is the biggest therapist-related search that reaches this site, and the honest answer is that it depends on the route: in-network copay, out-of-network reimbursement, a sliding-scale community clinic, a university training clinic, a low-fee network. no dollar figure on this page comes from a journal, so none is given. the therapy cost calculator works out your number, and the free-therapist hub lists every route that costs nothing, with a page for each of 51 cities. if the answer for now is that you cannot afford a therapist at all, what works that costs nothing has eight things with trial evidence behind them, each with its dose and its catch, and the veisund app is free peer support, by text or video, under a name you pick. it is not therapy and it does not pretend to be. it is where people compare notes on the six questions, and on what they were told.

two things from the research that help when money is the constraint. weekly beats fortnightly and twice weekly beats weekly for the same total number of sessions, so ten sessions spent in ten weeks do more than ten spread over five months.[24][27] and video therapy is not a downgrade; across two meta-analyses the difference from in-person was negligible, which widens the pool of who you can afford to the whole state.[19][20]

questions people ask

How do I find the right therapist for me?

Write the problem in one sentence and pick a measure for it (the PHQ-9, the GAD-7, or a count of the thing itself). Use a directory only for logistics: insurance, fee, times, online or in person, language, identity preference. Shortlist three. Book 15-minute consultations and ask the six questions on our therapy-length page, above all whether they measure progress with a form and will show you the scores. Take the measure again after session four. If nothing has moved by session eight, change the plan, the frequency, or the therapist.

Is there a quiz to find the right therapist?

There are many, and none of them has published a randomized test showing its matches do better than assignment by chance. The one matching method with a randomized trial (JAMA Psychiatry 2021, 218 patients, 48 therapists) matched people to therapists’ measured track records across 12 problem domains, built from 15 or more prior cases each, and improved outcomes by d 0.50 to 0.75. Consumer quizzes and directory filters do not have that data. They can handle logistics and stated preferences, which have a real but small effect, mostly on dropout. They cannot predict fit or outcome.

How important is it to find the right therapist?

The therapist explains about 5 percent of outcome variation across studies, 8 percent in randomized trials and more for people with severe symptoms. In one NHS data set of 119 therapists, individual recovery rates ranged from 23.5 to 95.6 percent. That matters more than the type of therapy for depression, where seven therapies differ by d 0.01 to 0.30.

How do I know if my therapist is good for me?

Two checks. Your chosen measure has moved by session four, or the therapist has noticed that it has not and changed something. And you would rate the relationship well on feeling understood, agreeing on goals, and agreeing on method. Early response predicts final outcome with g 0.87 across 25 studies, and the alliance correlates with outcome at r 0.28 across 295 studies.

When should I switch therapists?

If your measure has not moved by session four, raise it. If it has not moved by session eight and nothing has changed in the plan, switch something: the plan, the frequency, or the therapist. Earlier if the therapist will not measure or does a therapy you declined. A quarter of people who eventually recover show no early response, so no movement at session four is a prompt to talk, not a verdict.

What is the difference between a therapist, a psychologist, a psychiatrist and a counselor?

A psychiatrist is a physician who can prescribe. A psychologist has a doctorate and does therapy and testing. A therapist or counselor in the US is usually master’s-level and licensed by a state board (LCSW, LPC, LMHC, LMFT). In the outcome data, professional degree and years of experience account for little of the difference between therapists.

Is online therapy as good as in-person?

Across 43 intervention studies and 4,336 clients, video-delivered therapy produced effects largely equivalent to in-person; a second meta-analysis of 47 between-group studies found a negligible difference. The alliance-outcome link is the same online (r 0.275) as face to face (r 0.278). Both reviews note the limited number of randomized trials.

How much does therapy cost without insurance?

It depends on the route: an in-network copay, an out-of-network reimbursement, a sliding-scale community clinic, a university training clinic, or a low-fee network. Our therapy cost calculator works out your number, and the free-therapist hub lists every route that costs nothing, city by city. No dollar figure on this page comes from a journal, so none is given here.

Does this page say anything about medication?

No. It is about choosing a psychological therapist only. Nothing here is a reason to change any treatment on your own.

sources

  1. Johns RG, Barkham M, Kellett S, Saxon D. A systematic review of therapist effects: a critical narrative update and refinement to review. Clinical Psychology Review 2019;67:78-93. 20 studies (3 RCTs, 17 practice-based). Therapist effects “found in 19 studies”; range 0.2% to 29%, “weighted average = 5%”; 8.2% in RCTs; “patient severity appeared related to TE size”; therapist effects “are a robust phenomenon although considerable heterogeneity exists.” PMID 30442478. https://doi.org/10.1016/j.cpr.2018.08.004
  2. Saxon D, Barkham M. Patterns of therapist variability: therapist effects and the contribution of patient severity and risk. Journal of Consulting and Clinical Psychology 2012;80:535-546. 119 therapists, 10,786 NHS primary-care patients with planned endings. Therapist effect “6.6% for average patient severity,” ranging “from 1% to 10%” as severity rose; “recovery rates for individual therapists ranged from 23.5% to 95.6%.” PMID 22663902. https://doi.org/10.1037/a0028898
  3. Wampold BE, Brown GS. Estimating variability in outcomes attributable to therapists: a naturalistic study of outcomes in managed care. Journal of Consulting and Clinical Psychology 2005;73:914-923. 6,146 patients, about 581 therapists. “About 5% of the variation in outcomes was due to therapists.” “Therapist age, gender, experience, and professional degree accounted for little of the variability in outcomes among therapists.” PMID 16287391. https://doi.org/10.1037/0022-006X.73.5.914
  4. Goldberg SB, Rousmaniere T, Miller SD, Whipple J, Nielsen SL, Hoyt WT. Do psychotherapists improve with time and experience? A longitudinal analysis of outcomes in a clinical setting. Journal of Counseling Psychology 2016;63:1-11. 6,591 patients, 170 therapists, on average 4.73 years of data each. “A very small but statistically significant change in outcome was detected indicating that on the whole, therapists’ patient prepost d tended to diminish as experience (time or cases) increases”; early termination fell with experience. PMID 26751152. https://doi.org/10.1037/cou0000131
  5. Kraus DR, Castonguay L, Boswell JF, Nordberg SS, Hayes JA. Therapist effectiveness: implications for accountability and patient care. Psychotherapy Research 2011;21:267-276. 6,960 patients, 696 therapists, outcomes across multiple symptom domains. Harmful therapists (average client reliably worsened) showed “large, negative treatment effect sizes (d = −0.91 to −1.49)” and effective ones “d = 1.00 to 1.52”; the share unclassifiable, ineffective or harmful ran “from 33 to 65%” by domain. “Therapist domain-specific effectiveness correlated poorly across domains, suggesting that therapist competencies may be domain or disorder specific.” PMID 21623550. https://doi.org/10.1080/10503307.2011.563249
  6. Kraus DR, Bentley JH, Alexander PC, Boswell JF, Constantino MJ, Baxter EE, Castonguay LG. Predicting therapist effectiveness from their own practice-based evidence. Journal of Consulting and Clinical Psychology 2016;84:473-483. 3,540 clients, 59 therapists; each therapist’s first 30 risk-adjusted cases classified across 12 domains, then compared to the next 30. “Therapist effectiveness was relatively stable, although somewhat domain specific”; therapists classed “exceptional” were “significantly more likely to remain above average with future cases.” PMID 26881446. https://doi.org/10.1037/ccp0000083
  7. Constantino MJ, Boswell JF, Coyne AE, Swales TP, Kraus DR. Effect of matching therapists to patients vs assignment as usual on adult psychotherapy outcomes: a randomized clinical trial. JAMA Psychiatry 2021;78:960-969. Six community clinics in Cleveland; 288 randomized, 218 patients and 48 therapists analysed. Therapist performance was scored pretrial “across 15 or more historical cases” on “12 problem domains” of a routine outcome measure, each therapist classed effective, neutral or ineffective per domain. Matched vs case assignment as usual: general symptomatic and functional impairment d = 0.75, global distress d = 0.50, domain-specific impairment d = 0.60, “with no adverse events.” 88.5% of patients were White. “To our knowledge, no prior studies have tested the causal efficacy of prospectively matching patients to therapists with empirically derived strengths.” NCT02990000. PMID 34106240. https://doi.org/10.1001/jamapsychiatry.2021.1221
  8. Flückiger C, Del Re AC, Wampold BE, Horvath AO. The alliance in adult psychotherapy: a meta-analytic synthesis. Psychotherapy 2018;55:316-340. 295 studies, more than 30,000 patients, 1978 to 2017. Alliance-outcome correlation for face-to-face therapy r = .278 (95% CI .256 to .299; d = .579); for internet-based therapy r = .275 (k = 23); 2% of effect sizes negative; “consistent across assessor perspectives, alliance and outcome measures, treatment approaches, patient characteristics, and countries.” PMID 29792475. https://doi.org/10.1037/pst0000172
  9. Swift JK, Callahan JL. The impact of client treatment preferences on outcome: a meta-analysis. Journal of Clinical Psychology 2009;65:368-381. 26 studies, over 2,300 clients. Clients who received a preferred treatment did slightly better (r = .15, 95% CI .09 to .21; “a 58% chance of showing greater improvement”) and were “about half as likely to drop-out.” PMID 19226606. https://doi.org/10.1002/jclp.20553
  10. Swift JK, Callahan JL, Cooper M, Parkin SR. The impact of accommodating client preference in psychotherapy: a meta-analysis. Journal of Clinical Psychology 2018;74:1924-1937. 53 studies, over 16,000 clients. Preference accommodation: “fewer treatment dropouts (OR = 1.79) and more positive treatment outcomes (d = 0.28)”; the effect “was not moderated by… preference type, treatment options, client age, client gender, client ethnicity.” PMID 30091140. https://doi.org/10.1002/jclp.22680
  11. Cabral RR, Smith TB. Racial/ethnic matching of clients and therapists in mental health services: a meta-analytic review of preferences, perceptions, and outcomes. Journal of Counseling Psychology 2011;58:537-554. Preference for a therapist of one’s own race/ethnicity d = 0.63 (52 studies); perceptions of matched therapists d = 0.32 (81 studies); treatment outcomes d = 0.09 across 53 studies, “indicating almost no benefit to treatment outcomes”; effects “highly variable” and highest among African American participants. PMID 21875181. https://doi.org/10.1037/a0025266
  12. Barth J, Munder T, Gerger H, et al. Comparative efficacy of seven psychotherapeutic interventions for patients with depression: a network meta-analysis. PLoS Medicine 2013;10:e1001454. 198 studies, 15,118 adults. Each of seven therapies beat waitlist (d −0.62 to −0.92); differences between therapies “absent to small (range d = 0.01 to d = −0.30)”; robust large-study evidence for CBT, interpersonal and problem-solving therapy, “unavailable or limited” for the others. PMID 23723742. https://doi.org/10.1371/journal.pmed.1001454
  13. Beard JIL, Delgadillo J. Early response to psychological therapy as a predictor of depression and anxiety treatment outcomes: a systematic review and meta-analysis. Depression and Anxiety 2019;36:866-878. 25 studies, 11,091 patients; early response “typically observed during the first 4 weeks.” Early responders had better post-treatment outcomes, g = 0.87 (95% CI 0.63 to 1.10); larger for anxiety (g 1.37) than depression (g 0.76); “robust and replicated.” PMID 31233263. https://doi.org/10.1002/da.22931
  14. Duffy KEM, Simmonds-Buckley M, Saxon D, Delgadillo J, Barkham M. Early response as a prognostic indicator in person-centered experiential therapy for depression. Journal of Counseling Psychology 2022;69:803-811. 3,321 NHS patients; early response defined as a PHQ-9 drop of 6 or more by session 4. 38.7% met it and were “six times more likely to recover at the end of treatment”; but “a quarter of patients displayed a pattern of eventual response, reaching recovery at end of treatment despite not experiencing an initial improvement.” PMID 36037492. https://doi.org/10.1037/cou0000633
  15. de Jong K, Conijn JM, Gallagher RAV, Reshetnikova AS, Heij M, Lutz MC. Using progress feedback to improve outcomes and reduce drop-out, treatment duration, and deterioration: a multilevel meta-analysis. Clinical Psychology Review 2021;85:102002. 58 studies, 21,699 patients. Progress feedback: symptom effect d = 0.15 (95% CI 0.10 to 0.20), d = 0.17 for not-on-track cases, dropout OR 1.19. PMID 33721605. https://doi.org/10.1016/j.cpr.2021.102002
  16. Lambert MJ, Whipple JL, Kleinstäuber M. Collecting and delivering progress feedback: a meta-analysis of routine outcome monitoring. Psychotherapy 2018;55:520-537. 24 studies. Two-thirds found feedback-assisted therapy beat the same practitioners’ usual care; feedback “reduced deterioration rates and nearly doubled clinically significant/reliable change rates in clients who were predicted to have a poor outcome.” PMID 30335463. https://doi.org/10.1037/pst0000167
  17. Delgadillo J, de Jong K, Lucock M, Lutz W, Rubel J, Gilbody S, et al. Feedback-informed treatment versus usual psychological treatment for depression and anxiety: a multisite, open-label, cluster randomised controlled trial. Lancet Psychiatry 2018;5:564-572. Eight NHS trusts, 77 therapists, 2,233 patients. Among patients flagged not on track, the feedback arm ended with less severe symptoms: PHQ-9 d = 0.23, GAD-7 d = 0.19. PMID 29937396. https://doi.org/10.1016/S2215-0366(18)30162-7
  18. Swift JK, Greenberg RP. Premature discontinuation in adult psychotherapy: a meta-analysis. Journal of Consulting and Clinical Psychology 2012;80:547-559. 669 studies, 83,834 clients: weighted dropout 19.7% (95% CI 18.7 to 20.7); moderated by diagnosis, age, “provider experience level” and setting, not by therapy orientation or format; higher with trainee clinicians. PMID 22506792. https://doi.org/10.1037/a0028226
  19. Batastini AB, Paprzycki P, Jones ACT, MacLean N. Are videoconferenced mental and behavioral health services just as good as in-person? A meta-analysis of a fast-growing practice. Clinical Psychology Review 2021;83:101944. 57 studies; 43 on intervention outcomes covering 281 outcomes and 4,336 clients. Video “consistently produced treatment effects that were largely equivalent to in-person delivered interventions”; the authors flag “the relatively limited number of randomized controlled trials.” PMID 33227560. https://doi.org/10.1016/j.cpr.2020.101944
  20. Fernandez E, Woldgabreal Y, Day A, Pham T, Gleich B, Aboujaoude E. Live psychotherapy by video versus in-person: a meta-analysis of efficacy and its relationship to types and targets of treatment. Clinical Psychology and Psychotherapy 2021;28:1535-1549. 47 between-group studies, 3,564 participants. Video therapy beat waitlist (g 0.77) and was “negligible in difference” from in-person; strongest for CBT and for anxiety, depression and PTSD. PMID 33826190. https://doi.org/10.1002/cpp.2594
  21. Bishop TF, Press MJ, Keyhani S, Pincus HA. Acceptance of insurance by psychiatrists and the implications for access to mental health care. JAMA Psychiatry 2014;71:176-181. US office-based physician survey, 2009 to 2010: psychiatrists accepting new patients with private insurance 55.3% vs 88.7% for other specialties; Medicare 54.8% vs 86.1%; Medicaid 43.1% vs 73.0%. PMID 24337499. https://doi.org/10.1001/jamapsychiatry.2013.2862
  22. Delgadillo J, Ali S, Fleck K, et al. Stratified care vs stepped care for depression: a cluster randomized clinical trial. JAMA Psychiatry 2022;79:101-108. 951 NHS patients, 30 clinicians. A machine-learning recommendation of low- vs high-intensity treatment at first assessment, against the usual step-up pathway: reliable and clinically significant improvement 52.3% vs 45.1% (OR 1.40, 95% CI 1.04 to 1.87), at £104.5 more per patient because more people went straight to high-intensity therapy. PMID 34878526. https://doi.org/10.1001/jamapsychiatry.2021.3539
  23. Ahuvia IL, Mullarkey MC, Sung JY, Fox KR, Schleider JL. Evaluating a treatment selection approach for online single-session interventions for adolescent depression. Journal of Child Psychology and Psychiatry 2023;64:1679-1688. 996 adolescents; a Personalized Advantage Index algorithm chose between two interventions. “No significant difference in 3-month depression outcomes between participants assigned to their matched intervention and those assigned to their nonmatched intervention”; predicted vs observed response r = .39 and .24. PMID 37183368. https://doi.org/10.1111/jcpp.13822
  24. Robinson L, Delgadillo J, Kellett S. The dose-response effect in routinely delivered psychological therapies: a systematic review. Psychotherapy Research 2020;30:79-96. 26 studies. “Optimal doses of psychotherapy in routine settings range between 4 and 26 sessions”; weekly sessions accelerate improvement. PMID 30661486. https://doi.org/10.1080/10503307.2019.1566676
  25. Robinson L, Kellett S, Delgadillo J. Dose-response patterns in low and high intensity cognitive behavioral therapy for common mental health problems. Depression and Anxiety 2020;37:285-294. 102,206 NHS patients. “Most responders (95%) attained [reliable and clinically significant improvement] within 7 sessions of LiCBT and 14 sessions of HiCBT”; social anxiety, PTSD and OCD needed high-intensity work and 6 to 16 sessions. PMID 32027435. https://doi.org/10.1002/da.22999
  26. Norcross JC, Lambert MJ. Psychotherapy relationships that work III. Psychotherapy 2018;55:303-315. Summary of the APA task force’s 16 meta-analyses on relationship elements: nine judged demonstrably effective, seven probably effective, one promising. “The psychotherapy relationship makes substantial and consistent contributions to outcome independent of the type of treatment.” PMID 30335448. https://doi.org/10.1037/pst0000193
  27. Cuijpers P, Huibers M, Ebert DD, Koole SL, Andersson G. How much psychotherapy is needed to treat depression? A metaregression analysis. Journal of Affective Disorders 2013;149:1-13. 70 studies, 5,403 patients. Number of sessions: “only a small association… no longer significant” after adjustment; “an increase from one to two sessions per week increased the effect size with g = 0.45, while keeping the total number of treatment sessions constant.” PMID 23528438. https://doi.org/10.1016/j.jad.2013.02.030

related: how long should therapy take · what works that costs nothing · how to find a free therapist · psychology today alternatives

adam