Blood cultures have a low yield and do not appear to be helpful when collected in all patients admitted to hospital with community-acquired pneumonia.

A 48-year-old man with severe Guillain-Barré syndrome suffered complete paralysis, and for 31 days could not communicate with the outside world, while remaining fully conscious. After recovery, he provided feedback on aspects of his care, such as mechanical ventilation, physical therapy, and communication. Conventional low tidal volume normocapnic ventilation induced ongoing and profound dyspnoea, occasionally relieved by modest increases in minute ventilation. Routine and apparently benign physical therapy was extremely painful, which was not reflected in heart rate or blood pressure changes. When he eventually re-established communication after many weeks, via slight eye movements, his first message was to express a particular distressing symptom. His case is a valuable reminder of the sometimes large gap between clinical measurements and assumptions and the subjective patient experience. We propose several approaches to address such issues in other paralysed but conscious patients.