Cognitive rehabilitation

After brain injury or stroke, attention, memory, planning, processing speed and insight can all be affected. These difficulties determine whether someone returns to work and manages independently far more often than physical impairment does, and they are far less often treated.

Cognitive rehabilitation works in two ways: restorative training of the underlying function, and compensatory strategies that work around it. In practice, compensation delivers most of the real-world benefit.

Attention and concentration. Memory, principally through external strategies — structured diaries, phone systems, routines — because these work better than internal memory training. Executive function: planning, sequencing, problem-solving and self-monitoring. Processing speed. Visual neglect after right hemisphere stroke. Awareness of one's own difficulties, which is itself frequently impaired and is the hardest to treat.

Fatigue management sits alongside all of it, because cognitive performance after brain injury collapses with fatigue and is often mistaken for a fixed deficit.

Attention training after traumatic brain injury, compensatory memory strategies, and structured problem-solving training for executive difficulty have reasonable support 1. Neglect rehabilitation has moderate support.

Computerised brain training programmes largely improve performance on the trained games and transfer poorly to daily life 2. They can be one component, never the programme.

Improvement on clinic tasks transfers poorly to a workplace or a household. Training has to happen in, or closely simulate, the environment where the skill is needed.

That means practising with the patient's own diary, their own kitchen, their own work tasks — and involving family, because the strategies only survive if the people around the patient understand and support them.

Some functions recover with training; others are better managed with strategies and external aids. A programme built entirely on drills ignores the second half, and one built entirely on compensation gives up on the first too early.

Which balance is right depends on the injury, the time since it and what the person needs to do. It is set at assessment and revisited.

Improvement on a task tends to stay on that task. Gains from repeated computerised training frequently do not transfer to daily activities, which is the outcome that matters 2.

So training is embedded in real contexts — managing appointments, following a recipe, planning a journey — and the family is taught the strategies too.

Mostly by teaching a routine for stopping and thinking before acting. In goal management training, for example, the person learns to pause, state the goal, break it into steps, carry them out and check the result against the goal — first on simple tasks, then on real ones such as cooking a meal or organising a day.

A meta-analysis of 19 treatment groups found small to moderate improvements in executive function, including on everyday tasks, usually maintained at follow-up — except in patients' and relatives' own ratings, where the gains did not last 3. The routine helps only if it is used outside the session, which is why it is practised on the person's own tasks and taught to the family.