Does Hospital Design Heal? Architecture And Recovery

AFor most of the last century the design of hospitals answered to two masters, the containment of infection and the movement of trolleys, and the buildings that resulted were admired by managers and disliked by almost everybody else. The reaction, when it came, was gentler in tone and far larger in ambition: gardens, daylight, artwork and views of trees were said to shorten convalescence, and a handful of celebrated comparisons, most of them small, were repeated until they hardened into common sense. The phrase healing environment entered official guidance in 1998 and has not left it since. What the phrase quietly assumes is a direct line from the fabric of a room to the body of the person lying in it. A second possibility has been comparatively neglected, namely that a floor plan acts upon patients indirectly, by settling where the staff who look after them can stand, what they can see from there, and how far they must walk to fetch anything.

BAviva Renshaw, of the Talmadge School of Health Systems, spent 2017 observing twelve wards built to three different plans, recording where nurses went rather than what patients felt. The differences in effort were substantial: on long corridor wards a nurse covered 6.2 kilometres in a shift, against 4.1 on wards where charting stations were distributed among the beds. Walking less, however, did not by itself buy anything for the patient. Time spent at the bedside rose only where the medication store lay within eight metres of the beds, and where drugs were held centrally the minutes saved on the corridor were simply spent at the store instead. Renshaw argues that the influence of a building upon recovery is therefore borrowed rather than owned, since it operates through the working day of the staff, and she insists that a plan which shortens journeys without shortening errands will show no benefit whatever.

CColm Sarraf, a physician at the Wexbury trust, contends that comparisons of this kind confound the shape of a building with the people inside it. When his own hospital moved into new premises in 2018, falls among inpatients declined by 21 per cent within a year, a result the architects were quick to claim. Sarraf points out that the trust had also raised night staffing by a fifth in the very month of the move, and that no method available to him can divide the credit between the two. He is careful not to overstate his position. His analysis covers one site across sixteen months, which is too little to establish anything general, and he concedes that his own figures would look every bit as impressive if the building really were responsible. What he dismisses is not the claim itself but the confidence with which it is ordinarily made.

DThe patient's own experience has been examined most closely in relation to sleep. In shared bays the median number of night-time entries into a patient's space was eleven, against four in single rooms, and the occupants of single rooms slept for appreciably longer. That advantage, however, is not evenly distributed. Among patients who could not use a call button, whether through confusion or through weakness, the interval between a fall and its discovery lengthened from nine minutes in an open bay to twenty-one minutes behind a closed door. Single rooms, in other words, move a small quantity of risk from the many to the few, and the movement is invisible in any figure reported as an average. Guidance recommending private accommodation throughout new construction seldom says which patients ought to be exempted from it, and that omission has proved more consequential than the choice of finishes or the width of a corridor.

EAssessing any of this rigorously is close to impossible, because a building cannot be assigned at random and nobody can be blinded to the room they are in. A survey published in 2021 gathered thirty-four studies from twelve countries and found the pattern that usually appears in such fields: the more carefully a study had been conducted, the less it found. Only five of the thirty-four allowed for the changes in staffing that so often accompany a move, and those five reported an average saving of one day in hospital rather than the four days claimed elsewhere. The reviewers were sceptical of before-and-after designs in particular, since a new building arrives with new equipment, a rearranged rota and a temporary lift in morale, none of which the walls can be credited with. Nothing in the review suggests that design is irrelevant. It suggests that its contribution has been measured badly.

FThe practical consequence has been a shift of attention away from flagship projects and towards arrangements costing almost nothing: lowering the volume of overhead paging, replacing corridor alarms with pagers carried by staff, agreeing a protected period of quiet at night, and moving a bed so that its occupant can see daylight. Whether such measures count as architecture is a question best left open. The evidence assembled so far leans towards Renshaw's account, provided that the fabric itself already meets a decent standard, since a plan can waste good staff but cannot substitute for absent ones. Where a hospital is short of nurses no amount of glazing will repair the deficit, and where it has enough, the arrangement of rooms decides how much of their time actually reaches the bedside. Both mechanisms are presumably at work, in proportions that no study has yet managed to fix.