How intensive should aphasia therapy be?

Ten hours a week for three weeks

The trial published by Breitenstein and colleagues in The Lancet involved 156 people living with chronic aphasia after a stroke — that is, well beyond the acute episode, at the point where recovery is commonly assumed to have plateaued. The protocol compared three weeks of intensive, individualised therapy, at a rate of at least ten hours a week, with a waiting period of the same length.

The primary outcome was not a naming score obtained under test conditions, but verbal communication in everyday life. That shift in the outcome measure is what gives the trial its weight: it does not merely show that people improve on what they practise, it shows that something carries over into use.

The treated group improved significantly where the waiting group remained stable, and a six-month follow-up found the benefit still present. The variable being manipulated was not the total duration of care but its concentration in time.

Breitenstein et al., The Lancet, 2017↗

Twenty to fifty hours — not a prescription

The RELEASE consortium reanalysed the individual data of 959 patients drawn from several trials, rather than the published aggregate results alone. This method makes it possible to examine what varies from one patient to the next, and not only a group average.

The greatest language gains were associated with a cumulative total above twenty hours, in a range extending to around fifty, delivered at high frequency, adjusted to the person’s needs and complemented by work between sessions. The range is wide, and that is a finding in itself: the literature describes a favourable region, not a dosage to be applied.

RELEASE Collaborators, Stroke, 2022↗

What actually happens in clinical services

That leaves the question the trials do not ask: how many hours does a patient actually receive? Monnelly and colleagues surveyed speech and language therapists in the United Kingdom about what they understand by intensive therapy, what they consider desirable, and what they manage to deliver.

The reported average comes to 128 minutes of therapy per week. A little over two hours, where the trials speak of ten. The gap is not a failure of clinicians to follow the evidence: the same professionals cite intensity as a decisive factor. It comes down to the time they have, the size of their caseloads and the format of funded care.

This is the most useful point on this page, and the least discussed. The open question is not ‘what intensity should be recommended’ — that has been settled since 2017. It is: by what means does a service reach a dose it cannot produce in session.

Monnelly et al., International Journal of Language & Communication Disorders, 2023↗

‘Intensive’ does not mean the same thing everywhere

One difficulty runs through this entire literature, and Monnelly’s survey brings it into the open: the word intensive has no single definition. Depending on the authors, it refers to a number of hours per week, a number of sessions, a density of practice within the session, or the total length of the programme. The clinicians surveyed use the term with widely varying thresholds, often below those of the trials.

The consequence is direct for anyone reading a paper. Two publications can both announce intensive therapy and describe protocols separated by a factor of three. Comparing their results without looking at the actual number of hours amounts to comparing something other than what one believes one is comparing.

This is also why the figures quoted above are worth remembering as they stand, with their units, rather than as a label. Ten hours a week over three weeks in Breitenstein, a range of twenty to fifty cumulative hours in RELEASE: these are quantities, not qualifiers.

The best-practice framework

“People with aphasia should be offered intensive and individualised therapy.” Simmons-Mackie et al., Aphasiology, 2017↗

These recommendations, the product of an international consensus process, also state that care should aim at participation in social life and involve those close to the patient. The wording is long-standing and commands broad agreement. Its practical reach depends entirely on the previous question: a recommendation on intensity commits to nothing as long as the number of hours is out of reach.

Work between sessions is not an extra

One point deserves to be singled out, because it is often treated as secondary in how these trials are read. In the RELEASE synthesis, the protocols associated with the best results include work carried out by the patient outside the sessions. It is not a comfort variable; it is a component of the treatment.

The arithmetic is easy to set out. One weekly forty-five-minute session amounts to three hours a month. The volumes described above require, over the same period, four to ten times more. No amount of timetable reshuffling closes such a gap: it can only be closed by working time that is not clinician time.

This conclusion is not a comfortable one, because it shifts part of the effort onto the person and those around them. It does not say the session becomes optional; it says the session alone does not produce the volumes the literature associates with the clearest gains.

What this means in everyday practice

Four consequences emerge, of unequal weight.

The first concerns the calendar: for the same number of hours, the literature favours a dense period over an even spread across several months. A course of one session a week for a year does not produce the same documented effect as the same quantity concentrated.

The second concerns work between sessions. In the RELEASE synthesis, home practice is not a supplement: it is part of what distinguishes the protocols associated with the best results. It is also, in practice, the only resource that makes it possible to approach the volumes described without multiplying appointments.

The third concerns how progress is read. Since the documented effect bears on communication in everyday life, that is where it should be looked for, and not only in the scores obtained on the trained items. A patient can make clear gains in naming on their training material without those around them perceiving any change, which is another question, addressed in material renewal.

One limitation remains, and it should be stated plainly. None of this work says that more is always better: it describes thresholds beyond which an effect becomes measurable, not a relationship that would continue indefinitely. The trials also recruit people able to follow a dense programme, which in practice excludes the most fatigable, the oldest and the least supported patients. Fatigue, mood and the person’s own goals remain clinical parameters that the literature on dose does not replace, and that no figure on this page settles.

How doMind puts this into practice

doMind is a cognitive and speech & language rehabilitation platform designed so that the work continues outside the practice: the clinician puts together a series of activities adjusted to the patient, the patient completes them at home, and the results come back to the clinician. The aim is not to replace the session but to extend its effect on the days it does not take place — which is where the gap between two hours and ten is played out. Discover the tool, or read the companion page on remote rehabilitation.

Book a demonstration →