Meta-Analysis on Comparison of Therapeutic Effects of Different Treatment Methods on Multiple Rib Fractures

Objective The treatment of multiple rib fractures is worthy of consideration. A Meta-analysis on the surgical and conservative treatment of multiple rib fractures was carried out in this paper to explore the curative effect of these methods and provide some medical evidence. Methods Articles about randomized controlled trials (RCT) on the surgical and conservative treatment of multiple rib fractures were retrieved from the PubMed, Embase and Cochrane library databases, and were screened then to extract the data. The results were displayed by forest plots. Results This study included 5 articles and 337 cases which were divided into the surgical group (161 cases) and conservative group (176 cases). The Meta-analysis showed that differences between the two groups in respirator support time, intensive care unit (ICU) stay time, length of stay, and the incidence of pulmonary infection, chest wall deformity, chest tightness and breathlessness on exertion were statistically significant, and the differences between the two groups in the rate of tracheotomy and in-hospital mortality were not statistically significant. Conclusion The surgical treatment of multiple rib fractures can shorten the respirator support time, ICU stay time and length of stay, and reduce the incidence of pulmonary infection and chest disability, but this should be verified further by more randomized controlled trials.

these studies, include more randomized controlled trials, and arrive at more convincing results. Therefore, the significance of this paper is to include more high-quality studies that have been published continuously and provide more medical evidence for this question.

Literature Retrieval
Articles published from the establishment date of the PubMed, Embase and Cochrane library databases to December 2019 about the surgery and traditional treatment of multiple rib fractures were retrieved from these databases. The keywords included "surgery", "operative", "internal fixation", "conservative treatment", "non-surgical treatment", "rib fracture", "flail chest", and "randomized controlled trial", which were connected with "AND" or "OR".

Inclusion and Exclusion Criteria
Inclusion criteria: a. Object of this study were patients diagnosed as having multiple rib fractures after a variety of inspection; b. This study was a randomized controlled trial (RCT); c. The interventions for the surgical treatment group included thoracoscopy or thoracotomy, and those for the conservative treatment group included methods like chest bandage for fixation, rest in bed, sedation, and respiratory support. Exclusion criteria: a. Non-RCT studies; b. Personal operative experience, animal model tests, conference papers, summaries, and case reports; c. Studies without the control group (i.e. lacking the conservative treatment group or the surgical treatment group).

Data extraction and statistical methods
The basic feature collection table for studies included was made and filled out, and the data were extracted by two researchers separately. Then, they did a cross-check, and the inconsistencies or disagreements would be solved by them through discussion or by all researchers through consultation. The data extracted included the basic features of the studies (including the name of the first author, the year of publication, the number of cases, the number of cases with flail chest, the number of broken ribs and the follow-up time). Stata version 15.1 software (StatCorp, College Station, TX, USA) was used for the Meta-analysis. The test for heterogeneity of the studies included was carried out at first. If no significant heterogeneity among these studies was observed (P 0.1,I 2 50%), the fixed effect model was used to analyze the results; if the heterogeneity among these studies was observed (P 0.1 I 2 50%), the random effect model was adopted to analyze the results. Standardized mean difference (SMD) was taken as the statistic for the analysis of continuous variables. The relative risk (RR) was used as the statistic for the analysis of binary variables. 95% confidence interval (CI) was used as the interval estimation of each statistic. 0.05 (α = 0.05) was taken as the significant level. That is to say, the difference between the two groups was statistically significant when P < 0.05. The analytical results were shown by the forest plots.
Results 704 papers were retrieved in total, in which 83 papers were selected after excluding 621 papers that were identical and not in line with this study through reading the title and the abstract. After reading the whole article, 78 papers that did not meet the inclusion criteria were excluded and 5 papers were included finally [16][17][18][19][20]. The retrieval process is shown in Fig. 1. The basic features of the articles included are shown in Table 1.  [16][17][18][19][20] were included to analyze the respiratory support time (day). The heterogeneity among these 5 studies was observed (I 2 = 94.6%, P = 0.000 < 0.1), so the random effect model was adopted for Meta-analysis. The results indicated that the surgical group had shorter respiratory support time than the conservative group 5 articles [16][17][18][19][20] were included to analyze the intensive care unit (ICU) stay time (day).
The heterogeneity among these 5 studies was observed (I 2 = 86.6%, P = 0.000 < 0.1), so the random effect model was adopted for Meta-analysis. The results showed that the surgical group had shorter ICU stay time than the conservative group (SMD=-1.35; 95%CI: -2.07 to -0.64; P = 0.000 < 0.05), as shown in Fig. 3.
3 articles [17][18][19] were included to analyze the length of stay (day). The heterogeneity among these 3 studies was observed (I 2 = 96.3%, P = 0.000 < 0.1), so the random effect model was adopted for Meta-analysis. The results suggested that the surgical group had 3 articles [17,[19][20] were included to analyze the incidence of chest wall deformity. No significant heterogeneity among these 3 studies was observed (I 2 = 31.6%, P = 0.232), so the fixed effect model was used for Meta-analysis. The results showed that the surgical group had lower incidence of chest wall deformity than the conservative group (RR = 0.07; 95%CI: 0.03 to 0.17; P = 0.000 < 0.05), as shown in Fig. 8.
2 articles [16,19] were included to analyze the incidence of chest tightness. No significant heterogeneity between these 2 studies was observed (I 2 = 28.8%, P = 0.236), so the fixed effect model was used for Meta-analysis. The results showed that the surgical group had lower incidence of chest tightness than the conservative group (RR = 0.27; 95%CI: 0.17 to 0.44; P = 0.000 < 0.05), as shown in Fig. 9.
2 articles [16,19] were included to analyze the incidence of breathlessness on exertion.
No significant heterogeneity between these 2 studies was observed (I 2 = 0.0%, P = 0.337), so the fixed effect model was used for Meta-analysis. The results showed that the surgical group had lower incidence of breathlessness on exertion than the conservative group (RR = 0.32; 95%CI: 0.16 to 0.61; P = 0.001), as shown in Fig. 10. With so many advantages, the surgical treatment of multiple rib fractures also has negative effects. Firstly, the surgery should be undertaken after general anesthesia, which, however, requires good cardiopulmonary function. The surgery itself causes trauma to the body, and increases the risk in the patients with severe chest injury and with cardiopulmonary injury, but patients with severe chest injury tend to be treated with surgery. Secondly, as internal fixation implants of fractures increase the risk of tissue infection [26], the requirements for the implants are extremely strict. Granetzny [17] have reported 2 cases of wound infection and 2 cases of mediastinal infection after surgery.

Discussion
Thirdly, for some patients who may need secondary surgery, the implant had to be taken out, which increases the pain and the risk. Fourthly, the internal fixation may press the intercostal nerves of some patients, who will feel pain and numbness in the corresponding parts of the chest wall, probably reducing their quality of life.
In addition, indications and the timing of surgery for multiple rib fractures also need to be researched. There is no consensus on surgical indications, which depend on the general situation of patients, mainly including unbearable pain at fractured bones, respiratory failure caused by flail chest, the occurrence of chest wall deformity and lung function impairment [27]. In fact, some patients who have multiple and severely dislocated rib fractures do not necessarily have flail chest. They may not feel unbearable pain and have respiratory failure when they breathe slowly and stay in bed, but they also need surgery considering the high incidence of atelectasis, pulmonary infection and chest wall deformity in them. In addition, hemodynamic instability like the progressive hemothorax and delayed hemothorax, is also one of the indications of surgical treatment. Therefore, the change degree of chest volume within 24-48 hours, the dislocation degree of fracture end and the influence on the respiratory function should be considered when we assess whether a patient with multiple rib fractures should be treated surgically or non-surgically [5]. The timing of surgery is not unified at present. Some studies claim that the surgery should be performed within 24 hours or 8 days [28]. Though there is no clear consensus on the timing of surgery, it is commonly believed that the sooner to perform a surgery, the better for the patients. In theory, early surgical intervention can avoid prolonging the use of respirator and decrease the related complications for patients with multiple rib fractures and flail chest, while putting off surgery may cause these benefits to disappear and even greatly increase the incidence of complications [29], as proved by the Metaanalysis results of this paper. When to perform a surgery on patients with chest trauma,   Forest plot of length of stay (day) Figure 5 Forest plot of incidence of pulmonary infection Figure 6 Forest plot of rate of tracheotomy Forest plot of in-hospital mortality Figure 8 Forest plot of incidence of chest wall deformity Figure 9 Forest plot of incidence of chest tightness Figure 10 Forest plot of incidence of breathlessness on exertion