Results of search (study selection):
We searched 4604 records up to December 2020 on CRUS from electronic databases Cochrane Central Register of Controlled Trials (CENTRAL) (0), CINAHL (114), DOAJ (162), Embase (724), Medline (703), Proquest (863), Pubmed (1625), and Scopus (413). We isolated 19 studies meeting the inclusion criteria from electronic databases (2, 4,6–11,13,14,16, 19–26). Another four studies were isolated from the references of the isolated studies (27–30). We assessed 23 studies in qualitative and quantitative analysis. There were four prospective and 16 retrospective studies/case series included for assessment in this review. Three studies did not specify the nature of the study. Figure 1 is the PRISMA diagram for the studies in this review.
Study characteristics. The characteristics of included studies are presented in Table 1.
Risk of Bias in the included studies: Table 2 presents the Risk of bias assessment for this review. We found that all reports were either prospective or retrospective case series with high-Risk of bias. We had four studies, each rated good and poor while 15 reports fair on the Risk of bias assessment with the NIH tool for this review. Figure 2 presents the Risk of bias assessment on the NIH assessment tool for this review. The mean score from the modified Coleman methodology for Risk of assessment of bias was 40.30±8.8. The description of Surgical technique and follow-up domains scored high across all studies; however, the reports were rated poor on sample size (number of forearms operated), description of outcome criteria, and procedure of assessing outcomes. The studies uniformly described the mean deformity before and after CDO; however, functional outcomes on a measurable scale were reported by five studies (2, 10, 11, 16, 23).
Demographic outcomes: There were 4 studies each from Europe (14,20,25,26), America (4,6,8,13), and Africa (19,21,29,30). There were 11 studies from Asia (2, 7,9–11,16, 22–24,27,28). We identified 383 forearms/318 participants from 23 studies on CDO for CRUS. There were 114 right, 98 left, and 112 bilateral forearms with CRUS. Nine studies described hand dominance (2, 11, 13, 19, 20–24). There were 189 males and 129 females. The mean age of the participants at CDO was 6.28±1.75 years, ranging from 2 to 22 years (10, 13,30). There were five classification schemes from 15 studies for CRUS in this review (1, 22, 26, 37, 38). The classification schemes for CRUS are based on the shape, position of the radial head in the elbow joint, and the radiographic pattern of synostosis (Table 3). Twelve studies classified 4, 25, 152, and 5 forearms under Cleary and Omer types I II III and IV, respectively (Table 1). Two studies classified patients as per the scheme reported by Tachdjian and Wilkie (Table 1), while another study devised a new classification for identifying patients (26). The mean follow-up after CDO was 66.81±48.73 (6 to 312) months. The mean time to union was 7.27±1.67weeks (4 to 16 weeks).
Primary outcomes from the included studies: The mean preoperative pronation deformity was 72.83±15.640 (62.920 to 820) from 22 studies (2, 4, 6–11, 13, 14, 18–21, 23–30). The postoperative deformity after CDO was reported by 22 Studies. Twelve studies (4,8,13,14,18–21,23,26,29,30) corrected the forearm to a mean pronation deformity of 10.4±5.90 (30 to 220) degrees, while nine studies (2,6,9–11,24,25,27,28) corrected the forearm to a mean supination deformity of 13.47±9.510 (4.20 to 27.20). One study corrected the forearms to a neutral position (7). The mean correction achieved was 73.13±16.540 (350 supination to 1300 pronation). The correction achieved and the pooled mean difference between the deformity before and after CDO is presented in Figure 3. The overall mean difference was -68.26 [95% CI-86.87, -49.66]. The mean difference shows correction favouring supination. We got an I2 index of 100%, depicting significant heterogeneity among the studies. The pooled effect estimate was calculated with the random-effects model.
Secondary outcomes from the included studies: Six studies used plaster of Paris splint after CDO for immobilization (2, 7, 13, 16, 21,22). Fifteen studies used various metallic fixation devices for osteotomy site fixation (4, 6,9–11,14,19,20,23–25,26,28–30). Two studies were heterogeneous on the mode of immobilization, which included plaster of Paris splint and implant both and hence were not included in comparison (8, 27). Five studies used long intramedullary wire/ nail for CDO fixation (11, 19,23–25). Six studies used crossed K wire for osteotomy fixation (4, 6, 14, 21, 28, 29). One study used tension band wire (9). Two studies each used Plate fixation (10, 30) and heterogeneous methods for CDO fixation (8, 26). One study compared plaster of Paris and intramedullary fixations in both bone double level CDO (11). Overall, there were 22 incidents of significant loss of correction (>5 degrees) after CDO (Table 4). There were 11 incidents each of loss of correction in plaster of Paris splint and implant fixation groups. Six forearms had a loss of correction where intramedullary wires were used for CDO fixation (23,24). There were two incidents of loss of correction among plate fixations (26) and one each among crossed K wires (29), staple (8), and external fixation (26). We classified the studies into three osteotomy types based on the osteotomy location and number of bones osteotomised in the forearm: 1. There were seven studies (112 forearms) that osteotomised synostosis site to achieve desired correction (4, 6, 9, 10, 20, 29, 30) 2. Nine studies (163 forearms) osteotomised the forearm bones at different levels to achieve the desired derotation of the forearm (2, 11, 13, 19, 21, 23–25, 26) 3. However, four studies (54 forearms) osteotomised one of the forearm bones for correction of deformity (radius three studies and ulna 1 study) (7, 16, 22, 28). Three heterogeneous studies (54 forearms) used more than one osteotomy technique described above to derotate the forearm in the desired correction (8, 14, 27). We did pooled estimate for each of the subgroups of osteotomies (Supplementary material 1).
Neurovascular complications and compartment syndrome dominated complications in CDO for CRUS. In this review, four studies of synostosis site osteotomy recorded eight transient/reversible nerve palsies and six incidents of compartment syndrome/muscle herniation/circulatory compromise (4, 8–10). Another four incidents of compartment syndrome were reported by other studies (6, 25, 28, 29).
Functional outcomes in CDO: Four reports described functional outcomes in CDO by assessment scales/questionnaire (Table 1) (2, 10, 11, 16). The Liverpool Elbow Score, the classification system used by Failla et al., Jebsen-Taylor hand-function test, and quick disabilities of the arm, shoulder, and hand score (QuickDASH), described outcomes assessment tools. Significant improvements were recorded in the described assessment scales/questionnaire after CDO. Three studies devised novel scoring systems to assess functional improvements after CDO (6, 8, 23). Other studies evaluated the outcomes based on the parent and patient satisfaction in executing ADL after CDO.