This programme is a treatment manual written for mental health professionals. It assumes clinical training, the ability to make a diagnosis, and access to supervision. It replaces neither your clinical judgement nor your professional responsibility. The diagnostic criteria here are reformulated in our own words and never reproduced: consult the original manuals for the letter of the text.
1. The programme at a glance
Indication. Screen use that has become uncontrollable and disabling in adults: video games, social media, streamed video, pornography, news, or mixed phone use. The programme suits gaming disorder as ICD-11 defines it, and the neighbouring presentations that are not diagnoses.
A warning that is not a stylistic precaution. "Screen dependence" does not exist as a diagnosis. Only one presentation is recognised — gaming disorder, in ICD-11. Everything else is a real clinical problem with no established category, with contested instruments and prevalence figures that are heavily inflated. Section 2 draws the practical consequences, and section 8 draws the most important one: in most patients who consult for this, the screen is the symptom of something else.
Reference model. A structured cognitive behavioural therapy, whose evaluated reference is the German protocol of Wölfling and colleagues (2019), the only good quality multicentre controlled trial in this field. It is complemented by motivational interviewing for ambivalence, which is here the rule and not the exception, by behavioural activation for the time freed up, and by stimulus control, which is the most powerful and least used lever.
Format. Fourteen individual sessions of 50 minutes, weekly, over about four months, followed by two booster sessions at one month and three months. The evaluated protocol had fifteen sessions over four months; this format takes up its dose.
Target mechanism. Not the reduction of a number of hours, which is a poor objective, but four changes: what the use fills in for, the environment that triggers it, what the patient does with the time freed up, and their capacity to bear boredom and emptiness.
| Session |
Focus |
What the session produces |
| 1 |
Assessment, differential, real measurement |
Objective record in place |
| 2 |
Formulation: what the use is for |
Diagram written with the patient |
| 3 |
Ambivalence, and life goals |
Decisional balance, three goals |
| 4 |
The triggers, made visible |
Map of the triggers |
| 5 |
The environment |
Five modifications made |
| 6 |
The internal triggers |
Three states identified and tested |
| 7 |
Replacing: the time freed up |
Programme of activities begun |
| 8 |
Sleep, and the mid-point review |
Bedtime rule, measures |
| 9 |
Attention and tolerance of boredom |
Two exercises kept up |
| 10 |
The thoughts that give permission |
Permissions listed and tested |
| 11 |
High-risk situations |
A plan for three of them |
| 12 |
What the screen was replacing |
One real contact taken up |
| 13 |
Digital obligations, work, residues |
Final inventory |
| 14 |
Stocktake, plan, relapse prevention |
Summary sheet written by the patient |
What the patient takes away. Seven printable worksheets, listed in section 32 and downloadable from this page: my real use, what it does for me, my balance, my environment, what I do instead, what I tell myself, my plan.
What sets this programme apart from a digital detox. The aim is almost never abstinence, and never a quota of hours. A patient who goes from six hours to three while staring out of the window has solved nothing. A patient who has taken back their sport, their sleep and two friendships, and who still plays four hours on a Saturday, is doing well. That difference governs the whole treatment.
2. Before you start
Who this programme is for
This text is addressed to psychologists, psychiatrists, psychotherapists and mental health nurses trained in cognitive behavioural therapy. It assumes that you can conduct a diagnostic interview, spot an untreated attention deficit disorder, and recognise a depression underneath a behaviour.
It is not written for patients. It contains criteria for non-response, guidance on what must not be said, and a diagnostic discussion that a patient would read as a judgement on their case.
The prior question: what are we talking about?
This is the most important decision in the manual, and it is taken in session 1.
Only one diagnosis exists. ICD-11 recognises gaming disorder, and it alone, plus a category of hazardous use. DSM-5-TR places internet gaming disorder in its section of conditions for further study: that is not a usable diagnosis.
There is no disorder of social media, phone, video or news dependence. These presentations exist clinically, they cause suffering, and they have neither consensus criteria nor solid instruments. The available scales take over the criteria for substance dependence by analogy, which produces absurd rates of "dependence" in the general population — up to one adult in three according to some questionnaires. Do not rely on those figures, and do not quote them to the patient.
Do not tell a patient they are dependent. The word does two kinds of damage: it gives them an identity they did not ask for, and it installs a fatalistic explanation — "it is stronger than me, it is chemical" — that discourages action. Describe the behaviour, its consequences, and what it replaces. The word "dependence" appears in the title of this manual because that is how patients frame their request; it must not appear in your formulation.
Treat when functioning is impaired: work or studies, sleep, health, relationships, money. And treat when the patient has tried to cut down and has not managed it, several times, in a documented way.
Do not treat heavy use with no consequences. An adult who plays twenty hours a week, who sleeps, who works and who has friends does not have a disorder, whatever the concern of the people around them. It is a frequent consultation, and the right answer is to say so.
What this programme is not
It is not a treatment for attention deficit hyperactivity disorder, which is the first diagnosis to rule out and the most frequent explanation for use that gets out of hand.
It is not a treatment for depression, social anxiety or insomnia, which are the three other causes to look for before treating the screen in its own right.
It is not a treatment for compulsive sexual behaviour disorder, which ICD-11 recognises and which calls for its own frame when pornography sits within a broader picture.
It is not a treatment for gambling, including online: that is a long recognised behavioural addiction, with its own protocols and its own financial and legal stakes. Do not confuse them because the screen is the same.
It is not a digital detox. Total-break formats of a few days or weeks have virtually no data in their favour, and they fail for the simplest of reasons: they replace nothing.
How to use it
Read the whole thing before the first session. Sections 3 to 13 lay down the frame; sections 14 to 27 are worth rereading the day before each session.
Every session is described on the same frame: the objective, a step-by-step run-through, what you say, the common mistakes, and the criterion for moving on. That last point sets the pace: the programme advances by acquisition, not by calendar.
One warning specific to this presentation. You will be tempted to moralise, and the patient expects it — they have already heard it from their partner, their parents, their doctor. One disapproving remark about the content of what they watch or play, and you lose the honest description you need. Neutrality here is not a posture, it is a measuring instrument.
3. The clinical picture
What the patient describes
They rarely come of their own accord, and rarely for this. They come for tiredness, insomnia, falling results, a threat at work, a conflict with a partner. Or they come pushed by someone — the most frequent case, and it governs session 3.
Asked, they describe a sequence they know by heart and cannot interrupt. They settle down for ten minutes and get up three hours later. They do not know exactly how much time they spend, and they underestimate it by half. They have tried to stop, several times, with blocking apps they have uninstalled.
Three sentences recur from one patient to the next: "I do not know where my evening goes", "it is not even that I enjoy it", and "I am angry with myself every night".
What makes up the picture
Loss of control over duration, more than over starting. Most of these patients have no trouble not opening the app; they have considerable trouble closing it. That is an important difference from substances, and it points the treatment towards the environment rather than towards willpower.
Displacement. The screen does not add to life, it replaces it. Look for what has disappeared: sport, shared meals, reading, sleep, friends, sex, plans. That list is the inventory of the treatment.
Night-time use. Almost constant, and it is often the route by which everything deteriorates. Bedtime slips later, sleep shortens, the day becomes hard going, and the evening becomes the only pleasant part.
Failed attempts to stop, with their train of consequences: blocking apps circumvented, accounts deleted then recreated, the console put away in a cupboard then taken out again.
Shame. It is major and rarely voiced. These patients judge themselves with a severity they would apply to nobody else, and many lie about durations — first to those around them, then to you.
What keeps the use going
Function. This is the heart of the formulation, and it is the question that has to be answered first: what is this use for? The usual answers are few. To stop feeling anything. Not to be alone. Not to be bored. Not to think about tomorrow. To feel competent somewhere. To be recognised. To fall asleep. To avoid a conversation.
A use with no function does not last. A use whose function has not been identified cannot be treated.
The environment. It is the most decisive factor and the most neglected. The phone within reach, the notifications, the automatic play of the next video, the endless scroll, the app on the home screen, the console plugged in facing the sofa, the permanent connection. These devices are designed to prolong use, and they succeed better than the patient's willpower succeeds at interrupting it.
Say so to the patient, once, because it reduces their shame and increases their engagement: this is not a character flaw, it is an uneven contest, and we are going to change the ground.
The void. The time freed up by cutting down does not fill itself. A patient from whom three hours of evening are taken with nothing put in their place will take them back within ten days. Session 7 exists for that reason.
Intolerance of boredom. It is built up: the more the empty moments are filled, the more unbearable they become. Many of these patients can no longer wait for a lift, or queue, or sit doing nothing.
Rumination and the evening self-reproach. They increase the distress, and the distress brings them back to the screen. It is a loop, and the patient knows it.
What you need to know about the figures
The prevalence of gaming disorder is estimated very variably depending on the studies and the instruments, from under one to a few per cent of players. The highest estimates come from the least demanding questionnaires.
For everything else — social media, phone, video — there is no usable prevalence, because there is no definition. The work of Orben and Przybylski on large cohorts showed that the link between screen time and wellbeing, at population level, is very weak. That says nothing about your patients, who are clinical cases and not an average — but it should guard you against two errors: believing in an epidemic, and treating a screen time rather than a suffering.
4. The uses, and what they change
The word "screen" covers behaviours that have neither the same mechanism nor the same treatment. Identify the dominant use in session 1; it determines half the programme.
Video games
The only presentation that corresponds to a diagnosis. Three particular features.
Games are structurally designed to retain: progression, deferred rewards, daily appointments, seasons, team commitment. A patient who plays in a team does not merely have a habit, they have obligations towards other people.
Games are often the patient's only place of competence and recognition. This is decisive: taking that away without building something elsewhere is a real loss, not progress.
Games involve a genuine social life. Do not treat it as an illusion. Many of these patients have their only relationships in the game, and the question is not to abolish them but to add others.
Social media and scrolling
The mechanism is different: it is not competence, it is comparison and the expectation of a social signal. Endless scroll adds an absence of any natural stopping point — the use does not end, it is interrupted.
Two targets of its own: social comparison, which feeds mood and self-esteem, and use as an immediate emotion regulator, to be dealt with in session 6.
Streamed video
Often underestimated because it looks harmless. The dominant mechanism is the avoidance of boredom and delayed sleep onset. Automatic play of the next episode is one of the most effective devices in existence against going to bed, and turning it off is one of the most profitable interventions in the whole programme.
Pornography
A case apart, and the most delicate. Three points.
Volume of use predicts suffering poorly. The work of Grubbs and colleagues has shown that the sense of being dependent is strongly linked to moral incongruence — the gap between what the person does and what they believe they should do — more than to quantity. Two patients with the same use may be, one unharmed, the other devastated.
The practical consequence is important: you must first establish whether you are treating a behaviour that has got out of hand, or a suffering born of a conflict of values. Both can be treated, and not in the same way.
When the picture goes beyond pornography — multiple sexual behaviours, general loss of control, risk-taking — ICD-11 recognises a compulsive sexual behaviour disorder, and the frame changes.
News and information
Under-recognised. The mechanism is anxious: compulsive checking looks for a certainty it does not find, exactly like a checking ritual. Treat it as such, with the tools for certainty seeking, and not as a need for information.
Work use and messaging
The patient will say it is work, and sometimes they will be right. Two distinctions to make: what the employer requires, what the patient imposes on themselves, and what serves to avoid doing something else. The third category is the largest.
5. The model that guides this programme
Why this one
The field is young and the data are thin. That has to be said, and one should not lean on a model as though it were more assured than it is.
The evaluated protocol. Wölfling and colleagues (2019) conducted the only good quality multicentre controlled trial: a manualised cognitive behavioural therapy of fifteen sessions over four months, compared with a waiting list, in adult men presenting with gaming or internet use disorder. Remission was markedly more frequent in the active arm. That is the reference for the dose and structure of this programme.
Functional analysis is its core. The behaviour is understood as a behaviour maintained by its immediate consequences: a relief, a stimulation, a recognition. You do not treat a duration, you treat a function.
Stimulus control supplies the most powerful lever. It is an old and solid principle in behaviour therapy: when a behaviour is triggered by an environment, it is far more effective to modify the environment than to reinforce resistance. Applied here, that is session 5, and it is often the session that produces the most change in one week.
Behavioural activation treats the void. It is borrowed from the treatment of depression, where it is solidly established, and its principle applies directly: you do not remove an activity, you substitute another, and desire comes after the action.
Motivational interviewing treats ambivalence, which is here the rule. A majority of these patients come pushed by someone and do not really want to change. Starting a behavioural protocol with a patient in that state is the surest way to lose them within three sessions.
The five levers
Function. What the use is for, in this particular patient.
The environment. The ground, concretely modified.
Replacement. What fills the time freed up, chosen and planned.
Tolerance of boredom and emptiness. It can be trained.
The thoughts that give permission. "I have earned this", "just five minutes", "I will stop tomorrow".
What the model implies you should not do
Do not aim for a number of hours. It is the first reflex and the first error. Screen time is a convenient measure and a poor objective: it says nothing about the function, it ignores what the use replaces, and it sets the patient up to fail against a figure.
Do not begin with restriction. An ambivalent patient who is made to cut down in session 2 does not come back in session 4.
Do not explain the behaviour by dopamine. The explanation is popular, heavily oversimplified, and clinically harmful: it gives the patient a reason to believe themselves powerless. If the patient brings it up, do not humiliate them, but shift ground: "what will be of use to us is not what happens in your neurons, it is what your evening contains."
Do not judge the content. Neither the game, nor what they watch, nor the networks they frequent.
Do not aim for abstinence by default. It has precise indications, set out in section 12, and it is not the usual objective.
Frequently asked questions
The patient asks whether they are really dependent. What do I answer?
Answer the underlying question, which is almost always "am I normal, and can I get out of this". Say what is true: only one of these presentations is a recognised diagnosis, theirs either meets it or does not, and that changes nothing about what we are going to do. Then shift to the concrete: "what matters is not the label, it is that you are not sleeping any more and that you have stopped doing sport. That is what we are going to work on." Avoid confirming the word: it brings an identity and no lever.
Should a maximum number of hours be set?
No. It is the first reflex and the first error. A quota says nothing about the function, ignores what the use replaces, and sets the patient up to fail every week against a figure they will end up circumventing — by watching on another device, for example. Keep the measured time as an indicator, never as an objective. The objectives are: sleeping, having taken back three activities, being able to get through an evening without.
The patient hugely underestimates their use. Should I confront them?
No: show them the counter and say nothing. The underestimation is massive and unintentional — the work comparing reported time with logged time finds considerable discrepancies. Ask for their estimate before opening the record, look at it together, then one question only: "what do you make of that?" It is one of the most effective moments in the programme, and it is spoilt by any comment.
Is a complete break needed to start?
Almost never. Total-break cures have virtually no data in their favour, and they fail for a simple reason: they replace nothing. Abstinence has precise indications and it applies to an object — a game, an app, a site — never to screens in general. The criterion that settles it: has the patient ever managed moderate use of this specific object? If every attempt has failed in the same way, propose abstinence from that one.
The patient came pushed by his wife and does not really want to change.
That is the most frequent case, and it is not an obstacle: it is the starting point. Begin no restriction. Spend the whole of session 3 on the ambivalence, take seriously what they would lose, and look for a goal that belongs to them. If you push, they will defend their use — resistance in therapy is almost always manufactured by the therapist. And ask the question frankly in the first session: many are relieved to be able to say they are here for someone else.
Blocking apps — do they work?
As friction, yes; as a barrier, no — the patient will get round them, and that does not matter. Two conditions for them to be of use: that the code is not in their hands alone, or that circumventing it takes long enough to let the impulse pass. They are not treatments, they are pieces of environment, in the same way as getting the phone out of the bedroom — which remains, by far, the most profitable intervention.
What do I do when gaming is the patient's only social life?
Do not devalue it: they are real relationships, and a patient who hears you say otherwise will not come back. Two directions. Look for what those friendships cannot give them, and add it — without removing anything first. And propose an often excellent and rarely considered avenue: actually meeting the people they have been playing with for years. Check for social anxiety too: if all real contact is avoided, it is that which has to be treated, and reducing the gaming first would simply have isolated them.
A thirty-year-old plays twenty hours a week and his partner wants him to get help. Should he be treated?
Look at functioning, not duration. If he sleeps, works, sees people and has no consequences, he does not have a disorder, and saying so is the right intervention. The real subject is then often elsewhere: the time the couple no longer spends together, a disappointed expectation, a division of tasks. That can be worked on, and it is not a treatment of the behaviour.
Pornography: where do I start?
By establishing what this is. Two different presentations arrive identically: a behaviour that genuinely has got out of hand, and a suffering born of a conflict of values in someone whose use is ordinary. The second is not relieved by reducing the durations — it is the conflict that has to be worked on, without your arbitrating their values. In both cases, do not take sides on pornography itself: the patient will detect your position and will fall silent.
Should the attention disorder be treated first?
Yes when it is present and untreated: it is the first diagnosis to rule out in every adult with this presentation, and the use often partly eases as soon as it is managed. Refer for assessment, and say so clearly to the patient rather than doing both by halves. If you carry on in parallel, two adaptations: much shorter tasks, and a more radically modified environment — in this patient, willpower is not a reliable lever.
Is there a medication?
No, not for the behaviour. A few trials of bupropion or methylphenidate in players, often with an attention disorder, are too limited to found a prescription. What is treated by medication is the comorbidity: depression, anxiety, attention disorder. Treating the behaviour with a drug amounts to confirming to the patient that the problem is beyond them.
Do screens really damage the brain and attention?
The question is open, the data are far thinner than the surrounding discourse, and confirming the claim makes the patient fatalistic. What is certain and sufficient: they have lost the habit of sustained attention and of boredom, and a habit can be taken up again — that is the object of session 9. Promise an experience, not a cognitive restoration: being able to read a chapter, hold a conversation, wait with nothing.
How long before a change is visible?
The environment and sleep often produce an effect from the second or third week, and that is what supports engagement. The rest is slower, because it requires rebuilding activities and links: allow two to three months. Warn the patient of that timetable in session 1, or they will lose heart at the moment when the easy settings have given all they had to give.
The patient responded well, then everything came back during their holidays.
That is the most frequent resumption scenario, along with sick leave and working from home: the structure disappears and the use takes its place again. It is not a failure of the treatment. Three to four targeted sessions are generally enough, and the essential thing is to check the five settings — device in hand, yourself, since the patient will not flag what has been undone. Warn them about it in session 14: a patient who has been warned comes back early.
Are fourteen sessions enough?
It is the dose of the only properly evaluated protocol, and it suits a majority. If improvement is partial at the end, extend by four to six targeted sessions, naming what remains, rather than running the whole thing again. If you have only eight sessions, keep sessions 1, 3, 5, 7, 8 and 14 intact: the sorting, the ambivalence, the environment, the replacement, the sleep and the plan. That is the skeleton of the programme.