Reply to the comment by Dr. Ntoumenopoulos

Excerpt

Sir: Thank you for the opportunity to reply to Dr. Ntoumenopoulos who raises concerns regarding our prospective randomised controlled (PRCT) study of routine chest physiotherapy in ventilated patients. Taking the concerns in order, randomisation was by our pharmacy department who routinely prepare sealed opaque envelopes. Blinding was as rigorous as it could be given that staff did not know whether a patient was in the trial. Physiotherapists were not blinded, and if anything were in a position to bias data in favour of those receiving chest physiotherapy. Nurse allocation to patients changed every 12 h and was limited to three shifts per week, it is difficult to see how a consistent bias might be introduced through nursing care. Data were collected over 37 months, which is not unusual for a single-centre critical care study of this size. The single-centre design considerably reduces bias due to variance in practices this more than compensates for any within-centre medical practice changes over time which in any case would affect both groups. The fact that the data are from 1997 onwards does not invalidate them. As far as we are aware, physiotherapy techniques have not changed much. …