Ghana has roughly 38 psychiatrists for the entire country — about one for every 800,000 people, according to figures reported by Harvard's Global Health Institute. Ghana's population is projected at about 34.4 million for 2026 by the Ghana Statistical Service, so the ratio today is, if anything, worse. If you have ever tried to book a mental health appointment in Accra and been told to come back in weeks, that is the reason.
So it makes sense that people reach for an app instead. A phone is already in your hand, it costs nothing extra, and nobody has to know. The question is whether these apps do anything.
Here is what the published research actually shows, and what it does not.
Start here: an app is not treatment
Every honest study on this subject says the same thing in its limitations section. These apps are tested as additions to care, or as something to use while waiting for care. None of them has been shown to replace a clinician.
That matters more in Ghana than in the countries where the research was done. If an app is the only thing standing between you and no support at all, you should know exactly how thin the evidence is.
Woebot: the most-cited study, and the app no longer exists
Woebot is the app you will still see recommended in most listicles. It is worth explaining carefully, because it shows how this kind of writing goes wrong.
The study people cite is real. Fitzpatrick, Darcy and Vierhile published it in JMIR Mental Health in 2017. Seventy people aged 18 to 28 were recruited from a university community and randomly assigned either to use Woebot for two weeks, or to read an information ebook from the US National Institute of Mental Health.
The Woebot group's average score on the PHQ-9 depression questionnaire moved from 14.3 to 11.1. The control group's barely moved. That difference was statistically significant.
Now the parts that usually get left out:
- Both scores sit in the same band. On the PHQ-9, anything from 10 to 14 is "moderate" depression. The average participant started moderately depressed and finished moderately depressed.
- Seventy people, two weeks, one university. That is a small, short, early-stage trial, not proof of lasting benefit.
- The trial was unblinded. Participants knew which group they were in.
- The second author founded the company. The paper discloses that Alison Darcy is the founder of Woebot Labs and has a financial interest in it, and that the company covered participant incentives.
None of that makes the study worthless. It makes it a promising small trial by a team with a stake in the result, which is a very different claim from "clinically proven".
Founder Alison Darcy told STAT News the shutdown came down to the cost and difficulty of meeting the US Food and Drug Administration's requirements, alongside the company's wish to use large language models that regulators have not yet worked out how to approve. The company now sells to employers and insurers instead of to you.
Wysa: more studies, and a result that did not go its way
Wysa is still available and has considerably more research behind it. Its own clinical evidence page lists more than 36 peer-reviewed publications with academic partners, and the company received an FDA Breakthrough Device Designation in 2022 for its use in chronic pain with associated depression and anxiety.
A Breakthrough Device Designation is worth understanding properly, because it is often reported as if it were approval. It is not. It is a programme that speeds up the review of a device that might address a serious condition. It means the FDA agreed to look faster. It does not mean the FDA said the device works.
The more interesting result is a randomised controlled trial run with patients in England waiting for NHS Talking Therapies, measuring change in depression severity over 12 weeks. Depression scores changed about the same amount in both groups. The authors state plainly that this study found no evidence Wysa treats depression, while noting that participants described it as a useful supplement to clinical services.
That is a longer trial than the Woebot one, with a harder outcome, and it did not produce the headline anybody wanted. It is also the single most useful thing to know before you rely on the app.
A separate study of 1,205 users found that within five days people reported a sense of "therapeutic alliance" — the feeling of working with something that is on your side — comparable to scores reported for in-person and internet-based CBT. Feeling supported and getting better are not the same measurement, but the first is not nothing when the alternative is nobody.
What the evidence base looks like overall
| What you will read | What the studies actually say |
|---|---|
| "Clinically proven to reduce depression" | Small short trials show modest movement, often within the same severity band |
| "As good as therapy" | Almost no trial compares an app against actual therapy. The Woebot trial's comparison was an ebook |
| "FDA approved" | Usually a Breakthrough Device Designation, which is faster review, not approval |
| "Backed by dozens of studies" | Often true, but study counts include pilots and surveys, and negative results are less likely to be published |
Two structural weaknesses run through this whole field. Most trials last two to eight weeks, so long-term benefit is largely unmeasured. And trials that fail tend not to get written up, which means the published record looks better than reality.
Questions worth asking before you install anything
You do not need to read clinical papers to judge an app. Five questions get you most of the way:
- Is it still running? Check the app store listing date and recent reviews. Woebot is the warning.
- What happens in a crisis? A serious app routes you to a real helpline. If it does not, it is not built for the worst night.
- Does it work offline, and what does it cost in data? An app that needs a live connection for every exchange is expensive to use on a Ghanaian bundle.
- Where does what you type end up? Read the privacy policy for whether conversations are stored, and whether they are used to train models.
- Is the price in cedis or dollars? A subscription billed in dollars moves with the exchange rate, and many require a card that works for international payments.
What actually exists in Ghana
Alongside any app, these are real and local:
- The national helpline, 0800678678 — free, and listed as verified on the international helpline directory Find A Helpline.
- Emergency services, 193 — for immediate danger.
- NHIS-covered care. Ghana Health Service has trained community nurses to diagnose and prescribe for mental illness, specifically to work around the psychiatrist shortage. Your nearest clinic may be able to help more than you expect.
The honest summary is this. The research on mental health apps is thinner than the marketing, one of the most-recommended apps has been switched off, and the best-designed trial of another found no effect on depression. They may still help you feel less alone at 2am, and that has value. Just do not let an app be the reason you never call anyone.
This article describes services and published research. It is not medical advice. If you are struggling, speak to a qualified health professional, or call 0800678678.
Sources
- Fitzpatrick KK, Darcy A, Vierhile M. "Delivering Cognitive Behavior Therapy to Young Adults With Symptoms of Depression and Anxiety Using a Fully Automated Conversational Agent (Woebot): A Randomized Controlled Trial." JMIR Mental Health, 2017. mental.jmir.org/2017/2/e19
- STAT News, 2 July 2025 — Woebot shutdown and the founder's stated reasons. statnews.com
- Wysa clinical evidence page, including the FDA Breakthrough Device Designation. wysa.com/clinical-evidence
- University of Plymouth research portal — randomised controlled trial of Wysa with patients waiting for NHS Talking Therapies. researchportal.plymouth.ac.uk
- Harvard Global Health Institute — mental health service provision in Ghana. globalhealth.harvard.edu
- Find A Helpline — verified helpline listing for Ghana. findahelpline.com/countries/gh
- Ghana Statistical Service population projection for 2026.
Common questions
Is Woebot still available?
No. Woebot Health retired the consumer app on 30 June 2025 and moved to selling through employers and insurers. Articles that still recommend downloading it are out of date.
Do mental health apps actually work?
The evidence is mixed and weaker than advertising suggests. The 2017 Woebot trial showed a small improvement over two weeks in 70 young adults. A 12-week trial of Wysa with NHS patients found no difference in depression scores between app users and the comparison group. Most trials are short, and negative results are under-published.
Can an app replace seeing a therapist?
No study supports that. These apps are tested as additions to care or as something to use while waiting for it. In Ghana, with roughly one psychiatrist per 800,000 people, an app may be what is available — but it is a stopgap, not treatment.
What does "FDA Breakthrough Device Designation" mean?
It means the FDA agreed to review the device faster because it might address a serious condition. It is not approval and not evidence of effectiveness.
Which mental health number can I call in Ghana?
The National Mental Health and Suicide Prevention Helpline is 0800678678 and is free. For an immediate emergency, call 193.