An updated meta-analysis evaluating limb management after total knee arthroplasty—what is the optimal method?

Table 3 Results of subgroup analyses based on diverse postoperative knee flexion protocols

Clinical results	No. of trials	No. of participants			MD/RD	95% CI	P value	Heterogeneity I²	Model
Clinical results	No. of trials	Flexion	Extension	Total	MD/RD	95% CI	P value	Heterogeneity I²	Model
Total blood loss
LH group	4	153	153	306	− 185.44^a	− 211.59 to − 159.29	< 0.00001	43	Fixed
LM group	3	228	211	439	− 112.76^a	− 208.26 to − 17.15	0.02	77*	Random
SH group	3	204	210	414	− 245.45^a	− 419.93 to − 70.96	0.006	88*	Random
Hidden blood loss
LH group	3	123	123	246	− 95.77^a	− 138.21 to − 53.32	< 0.00001	85*	Random
LM group	3	228	211	439	− 83.70^a	− 186.28 to 18.88	0.11	98*	Random
SH group	2	80	80	160	− 196.35^a	− 259.06 to − 133.64	< 0.00001	87*	Random
Blood transfusion requirement
LH group	6	232	229	461	− 0.08^b	− 0.14 to − 0.03	0.003	40	Fixed
LM group	2	70	70	140	7.00^b	4.77 to 9.23	0.02	0	Fixed
SH group	2	84	84	168	0.02^b	− 0.07 to 0.12	0.62	0	Fixed
ROM (1 week after operation)
LH group	1	30	30	60	13.00	9.42 to16.58	< 0.00001	NA	–
LM group	4	258	241	499	4.75^a	3.10 to 6.40	< 0.00001	43	Fixed
SH group	3	205	208	413	2.63^a	− 2.80 to 8.06	0.34	93*	Random
DVT
LH group	6	232	229	461	0.00^b	− 0.02 to 0.02	1	0	Fixed
LM group	4	258	241	499	0.00^b	− 0.02 to 0.02	1	0	Fixed
SH group	4	175	179	354	0.00^b	− 0.03 to 0.03	0.98	0	Fixed
Wound-related infection
LH group	6	232	229	461	0.01^b	− 0.03 to 0.05	0.67	0	Fixed
LM group	3	204	187	391	0.00^b	− 0.02 to 0.02	1	0	Fixed
SH group	4	224	230	454	− 0.00^b	− 0.03 to 0.02	0.74	0	Fixed

Data in italics indicate a statistically significant P value
Abbreviations: LH long-term (≥ 24 h) high flexion (> 30°), LM long-term (≥ 24 h) mild flexion (≤ 30°), SH short-term (< 24 h) high flexion (> 30°), CI confidence interval, NA not applicable
^*Heterogeneity was statistically significant
^aMean difference (MD)
^bRisk difference (RD)

ISSN: 1749-799X