Table 1.

 Description of main characteristics of 17 included randomised controlled trials (listed alphabetically by type of rehabilitation) of inpatient rehabilitation specifically designed for geriatric patients. In all studies, patients in control group received standard inpatient care, except in study of Karppi et al,²² in which they received usual supervised home care

	Type of hospital for initial acute care	Reason for initial hospital admission	Mean length of hospital stay before randomisation (days)	Main criteria for selection of study sample	Mean age (years)	No of patients (intervention/control)
General geriatric rehabilitation
Applegate 1990, USA28 43	Community and city hospitals	Functional impairment from acute illness	21	Age ≥65, at risk for admission to nursing home, reversible functional problem, medically stable, not terminally ill or severely demented	78.8	78/77
Cohen 2002, USA21 49	Medical centre (medical or surgical ward)	Acute medical or surgical illnesses	NA	Age ≥65, admitted from home, not terminally ill or severely demented	74.2	694/694
Fleming 2004, UK44	Community hospital	Acute medical or orthopaedic disorders	35	Age ≥65, admitted from home, at risk for admission to nursing home, not severely disabled or demented, medically stable	81.5*	81/84
Karppi 1995, Finland22	Direct admissions from home	Medical problems with anticipated benefit from geriatric intervention	0†	Age >65, admitted from home, at risk for admission to nursing home, reversible functional problem	78.5	104/208
Rubenstein 1984, USA5 29 30	Medical centre	Acute medical or surgical illnesses	≥7	Age ≥65 (all men), at risk for admission to nursing home, reversible functional problem, not terminally ill or severely demented	77.9	63/60
Saltvedt 2002, Norway14-16	University hospital	Acute medical illnesses	0†	Age ≥75, admitted from home, frail, not terminally ill or severely demented	82.1	127/127
White 1994, USA27	University hospital	Acute medical or surgical illnesses	17	Age >65, at risk for admission to nursing home, reversible functional problem, not severely demented, medically stable	76.5	20/20
Young 2007, UK19 12 48	Community hospital	Acute medical or surgical illnesses	6*	Age >75, reversible functional problem, medically stable	86.0*	280/210
Orthopaedic geriatric rehabilitation
Cameron 1993, Australia11	General hospital	Acute hip fracture‡	2	Age ≥65	84.9	127/125
Gilchrist 1988, UK45	University hospital	Acute hip fracture‡	10	Age ≥65 (all women), expected prolonged hospital stay	81.3	97/125
Huusko 2002, Finland46 47	Local health centre hospital	Acute hip fracture‡	≥1	Age ≥65, admitted from home, good function before fracture, not terminally ill or severely demented	80.0	120/123
Kennie 1988, UK23 24	District hospital	Acute hip fracture‡	5	Age ≥65 (all women), expected length of hospital stay >7days	82*	54/54
Naglie 2002, Canada13	University affiliated hospital	Acute hip fracture‡	0†	Age ≥70, admitted from home, good function before fracture, not terminally ill	84.2	141/139
Shyu 2005, Thailand17	Community based medical centre	Acute hip fracture‡	0†	Age ≥60, good function before fracture, not terminally ill or severely demented	77.6	72/87
Stenvall 2007, Sweden25	University hospital	Acute hip fracture‡	0†	Age >70, good function before fracture	82.2	102/97
Swanson 1998, Australia18	University hospital	Acute hip fracture‡	0†	Age ≥55, admitted from home, good function before fracture	78.1	38/33
Vidan 2005, Spain26	University hospital	Acute hip fracture‡	0†	Age >65, admitted from home, good function before fracture, not terminally ill	81.9	155/164

NA=not available.

*Median.

†Zero (0) indicates patients were enrolled and randomised directly after hospital admission (in these studies, because of initial screening and informed consent procedures, though not explicitly reported, patients might have had few inpatient hospital days before randomisation).

‡Most studies explicitly defined “acute hip fracture” as uncomplicated single sided proximal femur fracture (that is, no additional fractures, non-pathological fracture).