通过Meta分析的方法系统比较锁定加压钢板治疗锁骨（中段）骨折时采用微创经皮内固定（minimally invasive per cutaneous plate osteosynthesis，MIPPO）与传统切开复位内固定（open reduction and internal fixation，ORIF）术式的临床疗效差异。

检索1997年1月至2018年7月已发表的临床对照研究。所检索的数据库包括Embase、Pubmed、中国知网、维普、万方等数据库。采用Revman 5.3软件进行Meta分析。

MIPPO术式与ORIF比较，切口长度（MD=-4.21，95%CI：-4.68、-3.75）、术中出血量（MD=-36.36，95%CI：-64.06、-8.66）、骨折愈合时间（MD=-1.53，95%CI：-2.42、-0.64）、术后局部皮肤不适（OR=0.20，95%CI：0.11、0.38）、VAS评分（MD=-0.59，95%CI：-1.10、-0.07）、患者对肩功能恢复满意度（OR=4.99，95%CI：1.85、3.43）、切口外观满意度（OR=6.19，95%CI：3.27、11.71）两组间差异有统计学意义（P值均<0.05）。手术时间（MD=-6.74，95%CI：-16.03、2.54）、住院天数（MD=-3.05，95%CI：-6.65、0.55）、术后内固定松动或失败（OR=0.45，95%CI：0.18、1.17）、骨折延迟或畸形愈合（OR=0.47，95%CI：0.12、1.78）、Constant评分（MD=0.69，95%CI:-0.20、1.58）、DASH评分（MD=-0.40，95%CI：-0.85、0.04，P=0.08）、肩功能恢复优良率（OR=2.07，95%CI：0.83、5.13，P=0.12），两组间差异无统计学意义。

应用锁定加压钢板治疗锁骨（中段）骨折时，MIPPO术式与ORIF相比具有切口长度小、术中出血少、骨折愈合快、术后皮肤相关并发症少、VAS评分低、患者对肩功能恢复及切口外观满意度高等优点。

Accounting for 4%-10% of total body fractures, clavicular fracture is a relatively common fracture in clinical practice, among which about 80% are midshaft clavicular fractures. The traditional surgical method for midshaft clavicle fracture is conducted by cutting open the skin outside clavicle, exposing its fracture ends and then placing the plate for fixation after reduction. With the continuous development of minimally invasive concept and internal fixation technology, especially the emergence of locking compression plate (LCP) , minimally invasive percutaneous plate osteosynthesis (MIPPO) is gradually applied in the treatment of clavicular fracture to maximize the protection of periosteum and blood supply and to reduce postoperative scar. Currently, traditional open reduction and internal fixation (ORIF) is still the standard method for the treatment of displaced midshaft clavicular fractures. Whether the clinical efficacy of MIPPO is better than that of ORIF still lacks strong evidence support. Therefore, from the perspective of evidence-based medicine, it is of great clinical significance to explore the difference in the efficacy of the two surgical methods for the treatment of midshaft clavicular fractures.

In this study, Meta-analysis was used to compare the two surgical methods from 11 aspects including operation time, incision length, intraoperative blood loss, length of hospital stay, fracture healing time, postoperative complication, VAS score, Constant score, DASH score, excellent and good rate of shoulder function recovery and patient satisfaction, so as to determine the difference in efficacy.I. Inclusive and exclusive criteria. 1. Inclusive criteria: (1) Research types: domestic and foreign published clinical control studies on MIPPO and ORIF (when using LCP) in the treatment of midshaft clavicular fractures such as observational studies, case analysis and randomized controlled trials; (2) Research subjects: adult patients diagnosed as displaced midshaft clavicular fractures based on medical history, physical examination and imaging examination in need of surgical intervention, without major nerve and blood vessel injury and regardless of age, gender and race; (3) Evaluation indicators: including at least fracture healing time, related complications and postoperative functional status. 2. Exclusive criteria: (1) Pathological fractures; (2) Open fractures; (3) Literatures incapable of accessing to original data; (4) Repeated literatures; (5) Non-clinical results, incomplete preoperative and postoperative measurement results, animal experiment literatures and cadaver mechanics experiments; (6) Research types: review, comment, lecture and reply from readers.II. Treatment methods.The clinical efficacy of MIPPO and ORIF was compared in the treatment of midshaft clavicular fractures with LCP.III. Efficacy evaluation indicators. The 11 indicators were operation time, incision length, intraoperative blood loss, length of hospital stay, fracture healing time, postoperative complication, VAS score, Constant score, DASH score, excellent and good rate of shoulder function recovery and patient satisfaction. IV. Retrieval strategy.The retrieved databases included Embase, Pubmed, Central, Cinahl, PQDT, CNKI, CQVP, Wanfang Data, Cochrane Library, CBM, etc.The directories of periodical and reference were manually retrieved, and the grey literatures such as chapters in unpublished academic papers, monographs, etc. were retrieved as much as possible. Meanwhile, the language was not limited in all relevant literatures, and the translation was conducted if necessary. The English keywords were Clavicle Fracture, MIPPO, MIPO and ORIF, and the retrieval strategy was Clavicle Fracture AND (MIPPO OR MIPO) AND ORIF. The Chinese keywords were clavicular fracture, open reduction and internal fixation, minimally invasive percutaneous plate internal fixation and locked compression plate.V. Statistical analysis.The Meta-analysis of the extracted data was conducted using the Review manager5.3 software provided by the Cochrane collaboration. The odds ratio (OR) and 95% confidence interval (CI) were used to represent the dichotomy variables. The mean difference (MD) , standard mean difference (SMD) or 95% CI were used to represent the continuity variables. The I² value was calculated to test the heterogeneity among different studies. Fixed effect model was adopted if I²≥50% , indicating that the heterogeneity among studies was relatively large. Meanwhile, the reasons for heterogeneity were analyzed and the random effect model was adopted. Sensitivity analysis was conducted by removing some studies, and funnel plots were made to assess publication bias. A P value of <0.05 was considered statistically different.

I. Basic information of the included literatures. According to the above search strategies, a total of 599 related literatures were retrieved. By reading the title and abstract, repeated literatures, non-control studies, repeated publications and literatures irrelevant to the research purpose were excluded, and 81 related literatures were preliminarily screened out. With the further reading of full text and screening in strict accordance with inclusive and exclusive criteria, finally 13 literatures were accorded with the inclusive criteria. Baseline conditions such as age and course of disease were compared in all the included literature studies, which were comparable (P>0.05) . II. Literature selection and quality evaluation.According to the Cochrane risk assessment criteria for bias, the included literatures were independently analyzed by two physicians to correspond the 6 items of standard respectively. If the literature was low-risk bias in each item, it should be regarded as low-degree bias. If the literature was high-risk bias or uncertain in two items, it should be considered as moderate bias. If the literature was high-risk or uncertain in two or more items, it should be regarded as high-degree bias. Differences were discussed or referred to a third senior physician to determine the quality of literature.III. Observation indicators and results. 1. Operation time: The operation times of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 10 literatures. Due to the heterogeneity among the results of each study (I²>50% ) , random effect model was used for Meta-analysis. The results showed that the operation time of MIPPO and ORIF in the treatment of clavicular fractures was similar (MD=-6.74, 95%CI: -16.03, 2.54, P=0.15) , and the difference was not statistically significant.2. Incision length:The incision lengths of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 6 literatures. Due to the great heterogeneity among the results of each study (I²>50% ) , fixation effect model was used for Meta-analysis. The results showed that the incision length of MIPPO in the treatment of clavicular fractures was less than that of ORIF (MD=-4.21, 95%CI: -4.68, -3.75, P<0.001) , and the difference was statistically significant.3. Intraoperative blood loss: The intraoperative blood losses of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 5 literatures. Due to the large heterogeneity among the results of each study (I²>50% ) , random-effect model was used for Meta-analysis. The results showed that the intraoperative blood loss of MIPPO in the treatment of clavicular fractures was less than that of ORIF (MD=-36.36, 95%CI: -64.06, -8.66, P=0.01) , and the difference was statistically significant.4. Length of hospital stays:The length of hospital stays of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 3 literatures. Due to the large heterogeneity among the results of each study (I²>50% ) , random effect model was used for Meta-analysis. The results showed that the length of hospital stays of MIPPO and ORIF in the treatment of clavicular fractures were similar (MD=-3.05, 95%CI: -6.65, 0.55, P=0.01) , and the difference was not statistically significant.5. Fracture healing time: The fracture healing times of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 12 literatures. Due to the heterogeneity among the results of various studies (I²>50% ) , random effect model was used for Meta-analysis. The results showed that the fracture healing time of MIPPO in the treatment of clavicular fractures was less than that of ORIF (MD=-1.53, 95%CI: -2.42, -0.64, P=0.0008) , and the difference was statistically significant.6. Postoperative complications:The postoperative complications of MIPPO and ORIF in the treatment of clavicular fractures were compared in the 10 included literatures. As the heterogeneity among the results of each study was not significant (I²<50% ) , fixation effect model was used for Meta-analysis. The results showed that the incidence of localized skin discomfort of MIPPO in the treatment of clavicular fractures was lower than that of ORIF (OR=0.20, 95%CI: 0.11, 0.38, P<0.001) , and the difference was statistically significant. The incidences of postoperative internal fixation loosening or failure (OR=0.45, 95%CI: 0.18, 1.17, P=0.10) and delayed union or malunion (OR=0.47, 95%CI: 0.12, 1.78, P=0.26) were similar between the two groups, and the differences were statistically significant.7. VAS score: The VAS scores of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 5 literatures. Due to the heterogeneity among the results of each study (I²>50% ) , random effect model was used for Meta-analysis. The results showed that the VAS score of MIPPO in the treatment of clavicular fractures was lower than that of ORIF (MD=-0.59, 95%CI: -1.10, -0.07, P=0.03) , and the difference was statistically significant.8. Constant score:The Constant scores of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 4 literatures. As the heterogeneity among the results of various studies was not significant (I²<50% ) , fixation effect model was used for Meta-analysis. The results showed that the Constant scores of MIPPO and ORIF in the treatment of clavicular fractures were similar (MD=0.69, 95%CI :-0.20, 1.58, P=0.13) , and the difference was not statistically significant.9. DASH score: The DASH scores of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 3 articles. As the heterogeneity among the results of each study was not significant (I²<50% ) , fixation effect model was used for Meta-analysis. The results showed that the DASH scores of MIPPO and ORIF in the treatment of clavicular fractures were similar (MD=-0.40, 95%CI: -0.85, 0.04, P=0.08) , and the difference was not statistically significant.10. Excellent and good rate of shoulder function recovery: The patient satisfactions of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 3 literatures. As the heterogeneity among the results of each study was not large (I²<50% ) , fixation effect model was used for Meta-analysis. The results showed that the excellent and good rates of shoulder function recovery of MIPPO and ORIF in the treatment of clavicular fractures were similar (OR=2.07, 95%CI: 0.83, 5.13, P=0.12) , and the difference was not statistically significant.11. Patient satisfaction:The patient satisfactions of MIPPO and ORIF in the treatment of clavicular fractures were compared in the included 5 literatures. As the heterogeneity among the results of each study was not large (I²<50% ) , fixation effect model was used for Meta-analysis. The results showed that the patient satisfaction of shoulder function recovery (OR=4.99, 95%CI: 1.85, 13.43, P=0.001) and incision appearance (OR=6.19, 95%CI: 3.27, 11.71, P<0.001) of MIPPO in the treatment of clavicular fractures were higher than those of ORIF, and the differences were statistically significant. III. Publication bias analysis.Review manager 5.3 statistical software provided by Cochrane collaboration was used to analyze the publication bias of operation time, fracture healing time, postoperative complications and other observation indicators. The results showed that all funnel plots were basically symmetrical, suggesting no significant publication bias.

The purpose of this meta-analysis was to compare the efficacies of MIPPO and ORIF in the treatment of displaced midshaft clavicular fractures. Compared with traditional ORIF, the treatment of MIPPO for clavicular fractures has the advantages of small incision, less intraoperative bleeding, faster fracture healing, less postoperative skin-related complication, low VAS score, and high level of satisfaction in shoulder function recovery and incision appearance. In conclusion, the clinical efficacy of MIPPO on clavicular fractures is better than that of ORIF.