. 2018 Feb 28;8(2):e020368. doi: 10.1136/bmjopen-2017-020368

Table 1.

Overview of study characteristics

Publication location date of study number of centres	Randomised mean age % female	Intervention treatment	Control treatment	Pain assessment adherence to treatment Losses to follow-up
Bruun-Olsen et al¹⁴ Norway 2008–2010 Two centres	n=57 (29:28) 68:69 years 62:50%	Group-based physiotherapist-led walking skills programme (2–6 patients per group). Commenced 6 weeks after surgery. Twelve sessions over 6–8 weeks.	1:1 usual physiotherapy care consisting of 12 individual physiotherapy sessions. Commenced 6 weeks after surgery. Twice-weekly sessions until 12–14 weeks after surgery.	KOOS pain scale 28/29 completed intervention. 28/28 received control treatment. 6 (2:4) lost to follow-up.
Buhagiar et al¹⁵ Australia 2012–2015 Two centres	n=165 (81:84) 67:67 years 69:68%	Inpatient rehabilitation at rehabilitation facility with twice daily supervised sessions of 1:1 and group-based exercises. Commenced after hospital discharge for 10 days. Home exercise programme after discharge from rehabilitation facility.	Home exercise programme comprising 2–3 group-based outpatient sessions to practice and progress exercises. Commenced 2 weeks after surgery.	KOOS pain scale 72/81 adhered to intervention. 74/84 adhered to control treatment. 6 (2:4) lost to follow-up.
Buker et al*³⁰ Turkey 2009–2011 One centre	n=34 (18:16) 64:68 years 89:94%	20 sessions of supervised physiotherapy and rehabilitation including range of motion and strengthening exercises, application of heat and TENS application. Five days a week for 4 weeks.	Home exercises including range of motion and strengthening exercise for 1 hour per day. Five days a week for 4 weeks.	Pain VAS Adherence not reported. Losses to follow-up not reported.
Chen et al¹⁶ China 2013–2014 One centre	n=202 (101:101) 66:67 years 63:67%	Structured telephone follow-up by nurse at 1, 3 and 6 weeks after hospital discharge to improve adherence to home exercise routine.	No telephone follow-up	Pain VAS Adherence not reported. 15 (7:8) lost to follow-up.
Fransen et al¹⁷ Australia 2009–2012 12 centres	n=422 (212:210) 64:65 years 54:52%	Group-based circuit exercise classes supervised by physiotherapist. Up to six patients per class. Commenced 6 weeks after surgery. Twice-weekly sessions for at least 8 weeks.	Usual physiotherapy care. Twenty-two per cent of participants reported six or more occasions of physiotherapy during the 6–12-week period after TKR.	WOMAC pain scale 140/212 participants attended≥12 classes 210/210 received control treatment 74 (43:31) lost to follow-up
Frost et al¹⁸ UK 1995–1996 Not reported	n=47 (23:24) 72:71 years 48:50%	Home-based functional exercise. Commenced following discharge from hospital. Duration not reported.	Home-based traditional exercise.	OKS item (pain on walking) Adherence not reported. 20 (7:13) lost to follow-up.
Kauppila et al¹⁹† Finland 2002–2005 One centre	n=86 (44:42) 71:71 years 76:79%	Group-based multidisciplinary rehabilitation programme. Up to eight patients per group. Commenced 2–4 months after surgery for 10 days.	Usual physiotherapy care. Supervised exercise programme at 2-month outpatient visit, with provision of further rehabilitation based on needs assessment.	WOMAC pain scale 44/44 attended intervention. 42:42 received control treatment. 11 (8:3) lost to follow-up.
Ko et al ²⁰ Australia 2008–2010 Three-arm trial Four centres	n=249 (85:84:80) 67:68:67 years 68:60:61%	1:1 physiotherapy with home-based sessions. Commenced 2 weeks after surgery. Twice-weekly 1:1 and home-based sessions over 6 weeks.	(1) Group-based circuit classes supervised by physiotherapist with home-based sessions. Up to eight patients per class. Commenced 2 weeks after surgery. Twice-weekly group and home-based sessions over 6 weeks. (2) Monitored home programme, two 1:1 physiotherapy sessions and one telephone follow-up call. Commenced 2 weeks after surgery. Four sessions per week for 6 weeks.	WOMAC pain scale 80% participants attended 9 or more 1:1 sessions, 77% attended nine or more group sessions, 83% attended both sessions in monitored home programme group. 16 (7:3:6) lost to follow-up.
Kramer et al²¹ Canada Not reported Not reported	n=160 (80:80) 68:69 years 59:55%	1:1 clinic-based rehabilitation programme with home exercise programme. Commenced 2 weeks after surgery. Up to two sessions a week for 10 weeks.	Home-based rehabilitation, monitored by telephone calls from physiotherapist. Commenced 2 weeks after surgery. At least one telephone call in weeks 2–6 and 1 call in weeks 7–12.	WOMAC pain scale 76/80 received intervention. 78/80 received control treatment. 26 (15:22) lost to follow-up.
Liebs et al*²² Germany 2005–2006 Five centres	n=159 (85:74) 70:70 years 73:70%	Ergometer cycling supervised by physiotherapist. Commenced 2 weeks after surgery. Three sessions a week for at least 3 weeks.	No ergometer cycling.	WOMAC pain scale Adherence not reported. 33 (15:18) lost to follow-up.
Liebs et al*²³ Germany 2003–2004 Two centres	n=185 (87:98) 69:71 years 70:73%	Early aquatic therapy. Commenced 6 days after surgery. Three times a week up to fifth week postoperative.	Late aquatic therapy. Commenced 14 days after surgery. Three times a week up to fifth week postoperative.	WOMAC pain scale Adherence not reported. 41 (18:23) lost to follow-up.
Minns Lowe et al²⁴ UK 2006–2008 One centre	n=107 (56:51) 68:71 years 57:59%	Home-based functional rehabilitation with two visits from physiotherapist at 2 weeks and 6–8 weeks after hospital discharge. Twice daily exercises for at least 3 months.	Usual physiotherapy care involving provision of an exercise booklet, with outpatient physiotherapy on a needs-only basis. No additional home visits.	KOOS pain scale 46/56 patients received two visits. 47/51 received control treatment. 9 (7:2) lost to follow-up.
Moffet et al²⁵ Canada 1997–1999 Five centres	n=77 (38:39) 67:69 years 63:56%	Functional rehabilitation programme with individualised home exercises. Commenced at 2 months after surgery. Twelve supervised sessions over 6–8 weeks.	Usual physiotherapy care, which included supervised home rehabilitation visits for 26% of patients.	WOMAC pain scale 38/38 participated in 12 sessions. 39/39 received control treatment. 8 (0:8) lost to follow-up.
Monticone et al²⁶ Italy 2010–2013 One centre	n=110 (55:55) 67:68 years 65:62%	Home-based functional exercises aimed at managing kinesiophobia, with monthly phone calls to encourage adherence. Commenced after hospital discharge. Twice-weekly sessions for 6 months.	No physiotherapy, advice to stay active.	KOOS pain scale Adherence not reported. 0 lost to follow-up.
Petterson et al²⁷ USA 200–2005 One centre	n=200 (100:100) 65:65 years 47:45%	Combined neuromuscular electric stimulation (NMES) and volitional strength training programme. Commenced 3–4 weeks after surgery. Two or three sessions a week for 6 weeks.	Volitional strength training programme without NMES.	KOS ADL item (effect of pain on function) 84/100 completed intervention. 97/100 completed control treatment. 51 (32:19) lost to follow-up.
Szots et al²⁸ Denmark 2013 One centre	n=117 (59:58) 67:68 years 61:67%	Two nurse-led structured telephone follow-up calls. Telephone calls at 4 days and 14 days after hospital discharge.	No telephone follow-up.	WOMAC pain scale 54/59 patients had both telephone follow-up calls. 54/58 received control treatment. 9 (5:4) lost to follow-up.
Vuorenmaa et al²⁹‡ Finland 2008–2010 One centre	n=108 (53:55) 69:69 years 57:65%	Delayed monitored home exercises, with guidance from physiotherapist at 2, 3 and 6 months postoperative. Commenced at 2 months after surgery for 12 months.	Usual care, which involved no additional guidance from 2 months postoperative.	WOMAC pain scale Seventy-two per cent of patients performed the training sessions at least twice per week in the first 6 months. 53/53 received control treatment. 4 (2:2) lost to follow-up.

*24-month follow-up also conducted but data from 12-month follow-up included in table to be consistent with follow-up period of other studies.

†Pain-specific outcome data were provided by the authors for a previous review⁶ and was used again in this review.

‡Follow-up at 14 months postoperatively.

KOOS, Knee Injury and Osteoarthritis Outcome Score; KOS ADL, Knee Outcome Survey - Activities of Daily Living; OKS, Oxford Knee Score; TENS, Transcutaneous Electrical Nerve Stimulation; VAS, Visual Analogue Scale; WOMAC, Western Ontario and McMaster Universities Osteoarthritis Index.