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. Prolonged grief disorder in adults, at least twelve months after the death. The programme also suits grief complicated by a violent or sudden death, or a death by suicide, with the adaptations in section 30.
What this programme is not. Support for every bereaved person. The great majority of adults get through a bereavement without treatment, and interventions offered to all bereaved people achieve negligible effects — the meta-analysis by Currier, Neimeyer and Berman (2008) is clear on this, and the prevention studies confirm it. This manual is addressed to those who stay stuck, whose life has stopped, and whose functioning is impaired a year on.
Reference model. The complicated grief treatment of Shear and colleagues, evaluated in three controlled trials (2005, 2014, 2016), which remains the best established treatment. It is complemented by the cognitive behavioural work of Boelen and colleagues (2007), which showed the contribution of exposure and of cognitive restructuring, and by the dual process model of Stroebe and Schut (1999), which supplies the alternation principle on which the whole programme is built.
Format. Sixteen individual sessions of 45 to 60 minutes, weekly, over about four months, followed by two booster sessions at one month and three months. Sixteen sessions is the evaluated dose: it is not an arbitrary division, and shortening it degrades the results.
Target mechanism. Neither forgetting, nor acceptance understood as resignation. Four changes: that the person can think about the death without being overwhelmed, that they stop avoiding what recalls the person who died and what recalls life, that the thoughts that weigh on them are examined, and that they take up plans that belong to them.
| Session |
Focus |
What the session produces |
| 1 |
Assessment, history of the loss, measures |
Daily record opened |
| 2 |
Formulation, the two processes |
Diagram written with the patient |
| 3 |
Personal goals |
Three goals, each with a first step |
| 4 |
The death narrative, first revisiting |
Recording of the revisiting |
| 5 |
Second revisiting |
Curve compared with the first |
| 6 |
Third revisiting, the worst moment |
Complete narrative, hardest moment approached |
| 7 |
The avoided situations: the hierarchy |
Hierarchy built, first exposure |
| 8 |
Situational exposure, mid-point |
Two situations taken back, measures |
| 9 |
Memories and photographs |
Positive memories accessible again |
| 10 |
"I should have": guilt |
Belief examined, share reattributed |
| 11 |
Injustice, anger, meaning |
What is said and what was not being said |
| 12 |
Rumination and proximity seeking |
Two habits reduced |
| 13 |
The imaginal conversation |
Dialogue conducted and recorded |
| 14 |
Rebuilding: relationships and activities |
Three things taken up |
| 15 |
Identity, the future, the continued bond |
What I keep, written |
| 16 |
Stocktake, plan, the dates that come round |
Summary sheet written by the patient |
What the patient takes away. Seven printable worksheets, listed in section 33 and downloadable from this page: where I am, my goals, my narrative and my listenings, what I avoid, my memories, what I tell myself, my plan.
What sets this programme apart from bereavement support. The target is not sadness, which is not a symptom and does not have to disappear. The target is what prevents grief from running its course: avoidance of the reality of the death, avoidance of life, rumination, and the proximity seeking that keeps the person who died present. Someone who cries for their wife three years on is doing well; someone who has never gone back into their bedroom is not. 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, recognise a major depression underneath a grief, and assess suicide risk.
It is not written for bereaved people. It contains descriptions of violent deaths, guidance on what must not be said, and criteria for non-response. Handed to someone affected, it would read as a judgement on their grief.
The prior question: should you treat?
This is the most important decision in the manual, and it is taken before session 1.
Do not treat ordinary grief. Bereaved adults are mostly doing well: they suffer, for a long time, and they take their lives up again. Interventions offered to everyone achieve effects close to zero, and there are data suggesting they can do harm to those who would have come through on their own. A bereaved person who cries, who talks about their dead, who has bad days and who works does not need you.
Treat grief that is stuck. Beyond twelve months: a life that has stopped, massive avoidance, a job lost, isolation, continuous rumination, thoughts of death, or a sense that nothing means anything any more.
Do not wait twelve months where there is risk. The diagnostic threshold is not a threshold for care. Someone three months bereaved with suicide risk, severe depression or post-traumatic stress disorder is treated immediately, for those reasons.
The first month is not the time for a protocol. What is useful then: information, a hearing, a little practical support, and an assessment of risk. Set an appointment for three months' time.
What this programme is not
It is not a treatment for depression. A full depressive episode is treated as such, before or alongside.
It is not a treatment for post-traumatic stress disorder. When the death was witnessed and re-experiencing dominates, the trauma is treated first or jointly, with a dedicated protocol.
It is not a stage model. The "five stages of grief" do not describe what bereaved people go through, have never been validated as a sequence, and do harm when presented: a patient who is not "at the right stage" becomes anxious, and the people around them become impatient. Do not use them.
It is not a support group. Groups have their use against isolation, and they do not treat prolonged grief.
How to use it
Read the whole thing before the first session. Sections 3 to 12 lay down the frame; sections 13 to 28 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 disorder. The session can pass very agreeably listening, every week, to an account of the week just gone and of the missing. The patient leaves relieved, you leave with the sense of having accompanied them, and nothing changes. This is the central drift of grief work, and the alternation principle in section 4 is what prevents it.
3. The clinical picture
What the patient describes
They do not say they are unwell. They say their wife died, and that it is normal to be like this. They often come because someone pushed them — a child, a doctor, an employer.
Asked, they describe a missing that has not diminished, time stopped, the impression that life ended at the same moment, and an inability to imagine a future. The sentences recur from one patient to the next: "I have not moved a millimetre", "I am pretending", "I still do not believe it".
The two forms of avoidance
This is the key to the picture, and the part most often missed.
Avoidance of the reality of the death. The patient does not go to the cemetery, has not emptied the wardrobe, keeps the bedroom untouched, does not say the word "dead", changes the subject, does not look at the photographs, does not listen to the voicemails they kept — or listens to them compulsively, which serves the same function. They have often never told anyone the whole story of the death.
Avoidance of life. Less often spotted and just as important. The patient no longer sees their friends, has stopped their activities, makes no plans, turns down invitations, does not laugh, takes no holidays, and denies themselves every pleasure. Often explicitly: it would be a betrayal.
Both avoidances are treated, and the programme devotes sessions 4 to 8 to the first, and 14 and 15 to the second.
Rumination, and proximity seeking
Rumination is the central mental activity of these patients: going over the last days, looking for what could have been done, imagining what would be if, replaying the scenario. It passes for grief work, and it is not: Eisma and colleagues showed that it functions as an avoidance — it fills the mind with the past so as to avoid facing both the reality of the death and the demands of the present.
Proximity seeking is its behavioural counterpart: sleeping with a piece of clothing, phoning the voicemail to hear the voice, keeping the place at the table, talking to the dead for hours a day, visiting the grave every day, moving nothing. These behaviours are not pathological in themselves — many bereaved people who are doing well have them — and they become so when they serve to keep the person present and to avoid admitting the absence.
What keeps the disorder going
Five mechanisms, all of them accessible.
A reality insufficiently integrated. The death is known and not assimilated. The patient knows, and does not believe it.
The two avoidances described above, which prevent integration on the one hand and reconstruction on the other.
Negative beliefs about oneself, the world and the future: it is my fault, I will not survive it, life has no meaning, I have no right to be happy, if I stop suffering it means I am forgetting them.
Rumination and proximity seeking, which hold the system in place.
The people around them. They tire, they avoid the subject, they advise moving on, or on the contrary they maintain the shrine. Either way the patient falls silent.
Epidemiology, in two useful figures
Among bereaved adults, roughly 7 to 10 % develop full prolonged grief (Lundorff et al., 2017). The proportion is markedly higher after a violent death, after a suicide, after the death of a child, and in people who were already vulnerable.
The impact is major: prolonged grief is associated with impaired functioning, with increased suicide risk, and with damage to physical health — sleep, blood pressure, substance use.
4. The model that guides this programme
Why this one
The complicated grief treatment of Shear and colleagues is the best established treatment. Its three controlled trials show superiority over interpersonal psychotherapy (2005), over supportive grief therapy (2014), and no added benefit from citalopram alongside the therapy (2016). Its components are identified: the repeated revisiting of the death narrative, taking back the avoided situations, work on memories, the imaginal conversation with the person who died, and — the most often neglected point — explicit work on the patient's personal goals from the earliest sessions.
The dual process model of Stroebe and Schut (1999) supplies the principle that organises the whole: a grief that runs its course alternates between loss orientation — crying, remembering, facing the reality — and restoration orientation — the new tasks, the new roles, life resuming. Stuck grief is characterised by the absence of alternation: the patient is fixed on one of the two, almost always on the loss, sometimes on restoration with complete avoidance of sorrow.
This principle has an immediate practical consequence: every session in this programme contains both. You work on the loss and you work on life, in the same week, from beginning to end. That is what stops the therapy becoming one long loss orientation, and it is what distinguishes this protocol from support.
The cognitive behavioural work of Boelen and colleagues (2007) showed the efficacy of exposure and cognitive restructuring, with one practical finding: exposure first, restructuring afterwards, produces better results than the reverse order. The programme follows that order.
The six levers
Personal goals. From session 3, before exposure. The patient must have something of their own, for themselves, independent of the person who died. Without it, you are asking them to give up their pain with nothing put in its place.
Revisiting the narrative. Telling the death, aloud, recorded, several times, rating as you go. This is the central exposure.
Taking back the avoided situations. Graded, negotiated, with written predictions.
The memories. Making accessible again the ordinary and pleasant memories, which have disappeared behind the memory of the end.
The beliefs. Guilt first, then injustice and meaning.
The continued bond. Turning a bond made of absence into a bond made of memory. The imaginal conversation in session 13 is the most direct tool.
What the model implies you should not do
Do not aim for the end of grief. Say so in session 1: nobody here will ask you to forget, or to move on, or to stop being sad.
Do not make the session a permanent loss orientation. If all your sessions are spent listening to the missing, you are accompanying a blockage.
Do not let rumination pass for work. A patient who spends forty minutes going over what they should have said at the hospital is not working, they are performing a ritual.
Do not ask them to give up the bond. The aim is not detachment. That idea, inherited from old theories, has done a great deal of damage.
Do not use the stages of grief.
5. The DSM-5-TR criteria, reformulated
The criteria below are a reformulation in our own words, as an aide-memoire. They do not replace the manual: consult the DSM-5-TR (American Psychiatric Association, 2022) for the letter of the text and the application notes.
Prolonged grief disorder was introduced in the text revision of DSM-5. It requires the following.
A death that occurred at least twelve months ago in adults — at least six months in children and adolescents.
A persistent grief reaction, present most days to a clinically significant degree, in one or other of two forms: intense yearning for the deceased, or preoccupation with thoughts and memories concerning them.
At least three symptoms out of eight, present most days for at least a month: a disturbance of the sense of identity, as though part of oneself had died; a marked sense of disbelief about the death; avoidance of reminders of the loss; intense emotional pain — anger, bitterness, sorrow; difficulty resuming relationships and activities; emotional numbness; a sense that life has no meaning; intense loneliness.
Clinically significant distress or impairment of functioning.
A reaction that exceeds what would be expected given the person's cultural and religious context.
The exclusion of another disorder that would better account for the picture: major depression, post-traumatic stress disorder, the effect of a substance.
What these criteria do not say, and what you must assess
They do not mention rumination, which is the central mental activity and a direct target.
They do not mention proximity seeking, which keeps the disorder going.
They do not mention guilt, which is the most frequent and most treatable cognition.
They do not distinguish the two forms of avoidance — of the death and of life — although the treatment attacks them differently.
One practical point: the diagnosis serves to avoid treating ordinary grief, and to avoid leaving without care someone whose life has stopped. It is not communicated to the patient as a label: many hear it as a judgement that their love was excessive.
Frequently asked questions
From when should you treat, and when should you hold off?
The diagnostic threshold is twelve months in DSM-5-TR and six in ICD-11, but the criterion for care is not duration: it is functioning. Do not treat someone who suffers and lives — who cries, talks about their dead, works, sees the people close to them. Treat someone whose life has stopped. And wait for no threshold where there is suicide risk, severe depression or post-traumatic stress disorder: those reasons are treated immediately, in their own right.
A patient says sorrow is normal and that they do not need therapy. Are they right?
Often, they are. Listen to what brings them, and look at functioning rather than at the intensity of the pain. If they work, go out and make plans, tell them so: "what you are going through is not an illness, and I see no reason to treat you." That is an intervention in itself, and it relieves. If they have not gone out for two years, show them so with facts, without calling their love excessive: "your sorrow is not the problem. The problem is that your life stopped at the same time as hers."
How do I stop the sessions becoming plain support?
By the agenda, and by the ten-minute rule on goals at the start of every session. A reliable indicator, to be checked at session 8: has at least one of the three personal goals visibly moved? If none has after eight sessions, the therapy has drifted, whatever the quality of the exchanges. And one question to ask yourself while rereading your notes: was there, every week, a task given and checked?
Do you really have to make someone who is already suffering tell the death?
Yes, and it is the best established component. Two precautions make the exercise bearable: session 3 on goals comes before, and every revisiting ends with a coming back and a return to the present. Patients dread this session and almost always come out of it saying they had never told it in full to anyone. What makes things worse is not telling it: it is telling it once and never going back.
The patient refuses to listen to the recordings.
That is frequent, and it can be dealt with. Look for the reason: apprehension most often, sometimes a belief — "I will not bear hearing my own voice", "it will open everything up again". Reduce the dose: one listening in the week, in your presence at first if necessary. And show them the figures: the fall between the first and third listening is the best argument available. Do not let it pass: without the listenings, the programme loses its active component.
What do I answer to "if I get better, it means I am forgetting her"?
This is the belief that blocks most, and it deserves time. Three interventions. Reframe: getting better does not mean loving her less, it means stopping dying with her. Turn the question round: "if the roles had been reversed, what would you have wanted for her?" — patients answer immediately and hear their own answer. And offer an experiment: a week in which they allow themselves one enjoyable thing, and a record of what happens to their memory of her. It does not fade.
Should a patient be encouraged to talk to their dead, keep their clothes, go to the cemetery every day?
Neither encourage nor forbid: assess the function. Many bereaved people who are doing very well talk to their dead and keep their things. The criterion is twofold: does it console, or does it serve to avoid admitting the absence? And: can the patient go a day without doing it? A flexible rite that soothes is not a target. An obligatory daily rite that stops them going out is one, and it is reduced a notch at a time, with their agreement.
Rumination — should it be left alone? Surely it is grief work?
No, and that is a widespread error. Going over what one should have done, replaying the scenario, imagining what would be if: the work of Eisma and colleagues shows that it functions as an avoidance. The test to give the patient: "after twenty minutes, have you learnt anything?" Distinguish it from remembering, which bears on what was and which has an end. Then postpone it to a period, without ever asking for it to be suppressed.
Should an antidepressant be prescribed?
There is no drug treatment for prolonged grief: the trial by Shear and colleagues (2016) showed that citalopram added nothing to the therapy on grief symptoms, while improving the associated depression. The practical conclusion: treat comorbid depression if it exists, and do not prescribe for sorrow. And watch benzodiazepines taken as needed, which function as an avoidance behaviour and compromise the exposure.
What do I do when the death was violent and the patient has re-experiencing?
Treat the trauma first or jointly, with a dedicated protocol. The mistake would be to conduct sessions 4 to 6 as though the intrusive image were a grief memory: it is not the same mechanism, and the exposure has to be conducted within the trauma frame. Once the re-experiencing is reduced, the programme resumes its course, and the narrative often takes a different form.
The imaginal conversation in session 13 makes me uncomfortable. Can it be skipped?
You can, and it is almost always a mistake. The discomfort is the therapist's, not the patient's: it is one of the few interventions patients talk about years later. Offer it simply, without staging, explaining that it is an exercise of imagination. If the patient refuses, offer it again the following week — those who hesitate often get the greatest benefit. And never suggest what the person who died would reply.
How many sessions are really needed?
Sixteen is the evaluated dose, and shorter formats have not been compared conclusively. If you have only ten sessions, keep sessions 1 to 7, session 10 and session 16 intact, and sacrifice the imaginal conversation and part of the reconstruction work — knowing what you are leaving out, and telling the patient. If improvement is partial at session 16, extend by four to six targeted sessions rather than running a whole programme again.
A patient talks about joining her husband. Is that suicide risk?
Always explore, never dramatise straight away. The wish to join the person who died is very common, often without intent, and many patients do not voice it for fear of being admitted to hospital. Distinguish the passive wish to die, intent, a plan and means; assess what holds them back — almost always children or grandchildren; and make a written safety plan where warranted. The risk is genuinely raised in people bereaved by suicide, in parents who have lost a child, and in isolated older widowers.
Bereavement groups, yes or no?
Useful against isolation, and little evaluated as a treatment. A patient with full prolonged grief needs treatment first, and the group is a good complement — not least because it provides a place to talk about their dead, which those around them have often stopped offering. A patient whose only problem is isolation may need nothing but the group.
Should a patient be pushed to empty the wardrobe, sell the house, take down the photographs?
No. These are not clinical goals, and there is no normal timetable. What is a clinical goal is that the patient can open the wardrobe, go into the bedroom, look at the photographs. Say so explicitly, or they will refuse the exposures for fear of what they commit them to: an exposure is not a decision. And discourage irreversible decisions taken under pressure from family or in a rush.
The people around the patient tell them they are overdoing it and should move on.
That is frequent and it does harm. Two useful things: acknowledge with the patient that those sentences are hurtful and that they do not have to conform to them; and invite a support person for one session, with a single precise request — talk about her, ask after her, invite him even if he refuses. A support person who leaves with a concrete instruction keeps it; one who has been given a lecture on grief changes nothing.