Prolonged grief in ten sessions: what a modular group protocol actually does

For prolonged grief disorder a clinician usually has two options: a full individual manual that requires training most services do not fund, or an open support group with warmth and no defined procedure. This pilot trial occupies the space between them, and it publishes the session grid rather than describing the intervention in the abstract.

How the ten sessions run

The architecture follows the dual process model, alternating loss-oriented and restoration-oriented work. Sessions 1 and 2 establish the alliance and affect regulation: a brief description of the relationship with the deceased, an emotional thermometer, diaphragmatic breathing, then psychoeducation separating acute from integrated grief, with mindfulness practice and a culturally embedded health-preservation element. Sessions 3 and 4 move to meaning reconstruction and to identity: naming sources of meaning, documenting resilience strategies, working the question of who the patient now is. Sessions 5 and 6 are the cognitive block, targeting grief-specific appraisals, self-blame and low self-worth, and ending with a written list of the patient's own counter-cognitions. Session 7 addresses continuing bonds, with homework that is a letter to the deceased expressing love, gratitude, apology, unresolved questions, a promise and a farewell, followed by imagining the reply. Session 8 introduces traumatic grief, builds a personal reminder list and prepares the narrative. Session 9 is the only individual session, an in-depth narrative of the death itself. Session 10 covers relapse planning, managing oscillations in cognition, emotion and behaviour, and short, medium and long-term goals.

Participants were 20 to 75 years old, at least four months post-loss, scoring above 24 on the PG-13-R, with mean age 40.24 years, 82.7% women, mean 27.89 months since the death and 36.8% violent deaths. Delivery was online, by eight clinicians (three psychotherapists, three clinical psychologists, two social workers), all with ten years of bereavement experience and supervised through the trial. The control group received access to an online peer group and self-help materials. Effects held at three months, and the diagnostic transition rates are the more clinically legible number: 61.8% of completers versus 20.0% of controls no longer scored in the probable-PGD range.

The design limits how far this can be read. Allocation was neither randomised nor concurrent: the control group was recruited only after the intervention had finished and matched to the treated group on relationship to the deceased, which makes this a historical comparison. The comparator also received minimal support rather than an active alternative, so part of the difference reflects attention and expectancy. The control group was younger and less heavily female than the treated group at baseline, and started about four points higher on the trauma questionnaire; age and sex were adjusted for in the analyses, so the between-group effect sizes are adjusted values, but the baseline gap on the trauma questionnaire was not, which makes the posttraumatic stress comparison the weaker of the two. This is a credible signal that the programme works as designed, not a definitive efficacy estimate.

What this changes at the table

The transferable content is the ordering, not the branding: stabilise and psychoeducate, then rebuild meaning and identity, then work the appraisals, then address the bond in writing, and only then approach the death itself. If a service already runs open grief groups, that sequence is a defensible spine to impose on them. Note that the death narrative sits in an individual session, not a group one, which is worth preserving when scheduling pressure invites collapsing it back into the group.

Two operational details matter. All 14 dropouts left within the first five sessions, before the exposure-adjacent material, so retention effort belongs in the first half, not the second. And the exclusions define the patient: no active suicidality, no severe comorbid illness, no concurrent psychological treatment. For patients with violent-death losses, who were over a third of this sample, the modest movement in posttraumatic stress suggests planning trauma-focused work alongside rather than assuming grief work will carry it.