Tailored Surgical Stabilization of Rib Fractures Matters More Than the Number of Fractured Ribs

Background: Patients sustaining multiple rib fractures have a significant risk of developing morbidity and mortality. More evidence is emerging that the indication of surgical stabilization of rib fractures (SSRF) should expand beyond flail chest. Nevertheless, little is known about factors associated with poor outcomes after surgical fixation. We reviewed patients with rib fractures to further explore the role of SSRF; we matched two groups by propensity score (PS). Method: A comparison of patients with blunt thoracic trauma treated with SSRF between 2010 and 2020 was compared with those who received conservative treatment for rib fractures. Risk factors for poor outcomes were analyzed by multivariate regression analysis. Results: After tailored SSRF, the number of fractured ribs was not associated with longer ventilator days (p = 0.617), ICU stay (p = 0.478), hospital stay (p = 0.706), and increased nonprocedure-related pulmonary complications (NPRCs) (p = 0.226) despite having experienced much more severe trauma. In the multivariate regression models, lower GCS, delayed surgery, thoracotomy, and flail chest requiring mechanical ventilation were factors associated with prolonged ventilator days. Lower GCS, higher ISS, delayed surgery, and flail chest requiring mechanical ventilation were factors associated with longer ICU stays. Lower GCS and older age were factors associated with increased NPRCs. In the PS model, NPRCs risk was reduced by SSRF. Conclusions: The risk of NPRCs was reduced once ribs were surgically fixed through an algorithmic approach, and poor consciousness and aging were independent risk factors for NPRCs.


Introduction
Thoracic trauma is the second most common injury in major trauma [1,2]. Ribs are one of the most vulnerable structures and the number of fractured ribs has been shown to be proportional to the severity of thoracic trauma [2][3][4][5][6]. The chest wall is highly innervated by pairs of intercostal nerves. Patients with broken ribs secondary to traumatic insult suffer excruciating pain, which limits the ability to take deep breaths and expectorate and contributes to pulmonary complications such as atelectasis and pneumonia after thoracic trauma. Patients with multiple rib fractures, even without flail, have a significant risk of developing morbidity and mortality [2,7,8].
Introduction of multimodal analgesia and surgical stabilization has changed the landscape of rib fracture management. Despite limited evidence, multimodal analgesia has been shown to decrease opioid use, ventilator days, pulmonary complication incidence, and the length of both hospital and ICU stay [3,4,[9][10][11][12]. Advances in regional anesthesia such as myofascial plane blocks and paravertebral blocks further broaden the options in pain control [13][14][15]. Similarly, more evidence is emerging that the indication of surgical stabilization of rib fractures (SSRF) should expand beyond flail chest [12,[16][17][18]. Compared to conservative treatment, SSRF has been observed to be of benefit for reducing the patientreported pain score, decreasing ventilator dependence, and shortening the length of ICU stay [19][20][21]. Nevertheless, little is known about the factors associated with poor outcomes in surgically fixed patients [16]. In addition, there is not enough discussion on whether the number of fractured ribs still matters after SSRF. Using the trauma register database in a tertiary trauma center, we retrospectively reviewed outcomes of patients with rib fractures. Firstly, we analyzed the impact of rib fracture numbers in the conservative treatment group (SSRF (−) group) and the tailored SSRF group (SSRF (+) group) in order to identify potential factors contributing to poor surgical outcomes. To further explore the role of SSRF, we matched the two groups by propensity score.

Patient Enrollment and Data Collection
The data were queried from the trauma registry database at our hospital which was prospectively maintained by 2 full-term trauma registries. Patients with blunt thoracic trauma admitted to our institution between January of 2010 and December of 2020 were reviewed for eligibility. Those with rib fractures undergoing SSRF were enrolled in the SSRF (+) group. Since conservative treatment has advanced a lot in recent years, such as multimodal analgesia, only patients who received conservative treatment between January of 2019 and December of 2020 were enrolled in the SSRF (−) group. A number of exclusion criteria included trauma victims without fractured ribs and patients who died within 48 h of severe trauma. Medical charts and surgical details were reviewed. Preoperative computed tomography (CT) and serial chest radiograms were reviewed till latest follow-up (at least 1 month postoperatively). All rib fractures were documented as their location and type based on the standard Müller AO (Arbeitsgemeinschaft für Osteosynthesefragen) classification system and the shaft part was further subdivided to anterolateral and posterolateral sections as proposed by Love et al. [22,23].

Algorithmic Approach with Tailored Surgical Planning
Algorithmic approach: As depicted in Figure 1, early SSRF is encouraged unless there is the presence of absolute contraindication such as unstable hemodynamics or severe traumatic brain injury with high potential to progress. For major trauma victims suffering multiple injuries, a sequence of surgeries should be tailored individually and determined by collaborative specialties. For patients with unstable spinal fractures, spine surgery is usually performed first to facilitate the following lateral decubitus position in thoracic surgery. For those with nonspine fractures, the order of surgery for different injuries is planned to minimize intraoperative position change and movement of the injured area, while facilitating the next surgical procedure without mutual interference. Temporary splinting is possible and concomitant fixation in one surgery is favored in our trust.
Surgical indication: in our institute, SSRF indications consist of flail chest, nonflail fractures with respiratory compromise or difficulty in weaning from ventilator, chest wall deformity or severely displaced fractures, thoracotomy for associated thoracic injury (e.g., hemothorax, lung laceration), and intractable pain despite appropriate analgesia.
Consideration for SSRF: Preoperative CT is obtained and carefully checked. Fixation site and rib plating number are considered case by case. In general, flail segments, severely displaced ribs, and ribs 3 to 8 are fixated in priority [17]. Ideally, both ends of flail segments are addressed [24]. All patients undergoing SSRF were generally anesthetized, and most of them were intubated with a single-lumen endotracheal tube. Double-lumen endotracheal intubation is preferred in patients with concomitant lung resection or eradication of lobulated hemothorax unless severe contralateral lung contusion or hypoxia is present. We usually place the patient in a lateral decubitus position with mild anterior tilt on a beanbag and the operative arm positioned at 100 degrees of flexion to maximally expose the surgical field. The pre-existing thoracostomy tube is removed to avoid hardware infection. The skin disinfection area covers sternum anterior and spinal process posterior. Video-assisted thoracoscopic (VATS) examination is introduced to facilitate fracture site localization or intrathoracic procedures. The incision and surgical approach are tailored according to the distribution of the fracture sites. A linear incision along the fracture sites is the option for patients with linearly arrayed single fracture on each rib. For fractures beneath the scapula, the incision would be made along the scapular border. The fixation could be accomplished using a right-angled drill and screwdriver with the scapula elevated by a Kocher retractor. For rib fractures close to the spine, the incision is made along the border of the erector spinae muscle, followed by dissection between the muscle and bony structure to expose the transverse process of spine. Another incision along the corresponding spinal process is made followed by the dissection between the muscle and the spinal process to facilitate fixation if necessary. For ribs with multiple and fragmented fractures, a posterolateral incision along the scapula is made for maximal exposure of the fracture sites. Surgical indication: in our institute, SSRF indications consist of flail chest, nonflail fractures with respiratory compromise or difficulty in weaning from ventilator, chest wall deformity or severely displaced fractures, thoracotomy for associated thoracic injury (e.g., hemothorax, lung laceration), and intractable pain despite appropriate analgesia.
Consideration for SSRF: Preoperative CT is obtained and carefully checked. Fixation site and rib plating number are considered case by case. In general, flail segments, severely displaced ribs, and ribs 3 to 8 are fixated in priority [17]. Ideally, both ends of flail segments are addressed [24]. All patients undergoing SSRF were generally anesthetized, and most of them were intubated with a single-lumen endotracheal tube. Double-lumen endotracheal intubation is preferred in patients with concomitant lung resection or eradication of lobulated hemothorax unless severe contralateral lung contusion or hypoxia is present. We usually place the patient in a lateral decubitus position with mild anterior tilt on a beanbag and the operative arm positioned at 100 degrees of flexion to maximally expose the surgical field. The pre-existing thoracostomy tube is removed to avoid hardware infection. The skin disinfection area covers sternum anterior and spinal process posterior. Video-assisted thoracoscopic (VATS) examination is introduced to facilitate fracture site localization or intrathoracic procedures. The incision and surgical approach are tailored according to the distribution of the fracture sites. A linear incision along the fracture sites is the option for patients with linearly arrayed single fracture on A rehabilitation program is essential after a major trauma [25]. Physiotherapists are routinely consulted for all trauma patients in our Trauma team. However, it is up to the specialty of the physiotherapists to decide what kind of rehabilitation they need.

Outcomes
Outcomes included pulmonary complications and time-related parameters. Pulmonary complications consisted of nonprocedure-related pulmonary complications (NPRCs) and procedure-related pulmonary complications (PRCs). Adult respiratory distress syndrome (ARDS), pulmonary embolism, pneumonia, and pulmonary edema were considered as NPRCs. PRCs were defined as any event developing during the SSRF or resulting from implants, screws, and surgical wounds. Time-related outcomes comprised ventilator days, length of ICU and hospital stay, and time from trauma to ambulation.

Statistical Analyses
Numeric outcome variables were assessed by the Kolmogorov-Smirnov test for normality. Non-normally distributed numeric variables were presented as median (interquartile range, IQR) and were analyzed by the Mann-Whitney U test. Categorical variables were presented as counts (percentage). Pearson X2 and Fisher exact tests were applied to compare categorical variables. If the variable was skewed, natural log transformation was used before multiple linear regression analysis. In order to adjust for the complex interactions between variables, both multivariate regression and propensity score (PS) models were used.
Categorical outcomes were analyzed by multivariate logistic regression, followed by the Hosmer-Lemeshow test for goodness of fit. Those factors with p-value < 0.05 in univariate analysis were entered into multivariate analysis. Based on established factors associated with outcomes in the literature and key factors addressed in this study, fracture numbers and surgical approaches were also included in the analyses.
A PS matching was applied since the SSRF (+) group differed greatly from the SSRF (−) group. Among the variables considered in the PS model were sex, age, comorbidity, transfer from district hospitals, unilateral and bilateral rib fractures, number of rib fractures, concomitant intrathoracic, scapular, sternal, and clavicular injuries, flail chest requiring mechanical ventilation, trauma mechanism, injury severity score (ISS), and abbreviated injury scale (AIS). The Wilcoxon signed rank test was used. A p < 0.05 was considered to indicate significance. Statistical analyses were performed using SPSS Version 25.0 Windows (IBM SPSS Corp., Armonk, NY, USA).

Results
There were 1468 consecutive patients with rib fractures between 2010 and 2020. A total of 177 of them (12.1%) received rib plating with A Plus titanium implants (SSFR (+) group, 84 of them receiving SSRF between 2019 and 2020). A total of 1291 patients received conservative treatment, 256 of whom were admitted between 2019 and 2020 and were enrolled as SSFR (−) group ( Figure 2). All patients were followed for at least 1 month after hospital discharge. Median (IQR) follow-up time was 18.1 (5.2-29.4) months. Table S1 provides unmatched demographics and preoperative variables between the SSRF (−) and the SSRF (+) groups. Outcomes of unmatched groups are provided in Table S2.

N = 177
Ventilator As demonstrated in Figure 3, the PL to anterolateral section (AL) of the third to eighth ribs were most often involved, whereas when the first and second ribs were injured, the PL to posterior section (p) was the most vulnerable. Additionally, the first and second rib injuries were frequently associated with clavicle (81.0% vs. 36.3%; p < 0.01), sternum (14.3% vs. 1.5%; p < 0.01), and scapula fractures (33.3% vs. 17.0%; p = 0.02). The fourth to eighth ribs were surgically fixated most often. As demonstrated in Figure 3, the PL to anterolateral section (AL) of the third to eighth ribs were most often involved, whereas when the first and second ribs were injured, the PL to posterior section (p) was the most vulnerable. Additionally, the first and second rib injuries were frequently associated with clavicle (81.0% vs. 36.3%; p < 0.01), sternum (14.3% vs. 1.5%; p < 0.01), and scapula fractures (33.3% vs. 17.0%; p = 0.02). The fourth to eighth ribs were surgically fixated most often.  Table S3. Age, gender, comorbid conditions, and trauma mechanism did not differ significantly between the more than six fractured ribs and no more than six fractured ribs groups. Shown in Table  4, patients with more than six fractured ribs had longer ventilator days, length of ICU stay, hospital stay, and time from trauma to ambulation (p < 0.01, respectively). However, the number of rib fractures was not an independent factor for ventilator days (p = 0.617),  Table S3. Age, gender, comorbid conditions, and trauma mechanism did not differ significantly between the more than six fractured ribs and no more than six fractured ribs groups. Shown in Table 4, patients with more than six fractured ribs had longer ventilator days, length of ICU stay, hospital stay, and time from trauma to ambulation (p < 0.01, respectively). However, the number of rib fractures was not an independent factor for ventilator days (p = 0.617), ICU stay (p = 0.478), hospital stay (p = 0.706), and NPRCs (OR, 0.734; 95% CI, 0.444-1.211; p = 0.226) after SSRF in multivariate models despite having experienced much more severe trauma (higher ISS (21.5 vs. 17.0; p < 0.01), more bilateral rib fractures (21.7% vs. 5.9%; p < 0.01), more flail segments (62.0% vs. 17.6%; p < 0.01), and more associated lung parenchyma injuries (47.8% vs. 28.2%; p < 0.01)). Details of univariate analyses of the SSRF (+) group are provided in Table S4. In the multivariate regression models, longer ventilator days was associated with lower initial GCS (p = 0.028), initial flail chest requiring mechanical ventilation (p = 0.020), longer time from trauma to SSRF (p = 0.025), and thoracotomy approach (p = 0.038) ( Table 5). Longer ICU stay was associated with poorer GCS (p = 0.008), higher ISS score (p = 0.003), flail chest requiring mechanical ventilation (p < 0.001), and longer time from trauma to SSRF (p = 0.007) ( Table 6). Poorer GCS (p = 0.037), higher ISS score (p < 0.001), and flail chest requiring mechanical ventilation (p = 0.049) were independent factors associated with longer hospital stay (Table 7). Presented in Table 8, risk factors for NPRCs were older age (OR, 1.097; 95% CI, 1.025-1.173; p = 0.007) and lower GCS (OR, 0.648; 95% CI, 0.499-0.843; p = 0.001). The Hosmer-Lemeshow test suggested the logistic model was accurate (χ2 = 1.757; p = 0.99). Table 4. A comparison of the outcomes of SSRF (+) group with and without more than six fractured ribs.     For those receiving conservative treatment, the number of fractured ribs was associated with longer ICU stay (p = 0.029) in the multivariate model, but not with ventilator days (p = 0.282), hospital stay (p = 0.861), and NPRCs (p = 0.418). Details of univariate analyses of the SSRF (−) group are provided in Table S5. Multivariate model results of ventilator days, ICU stay, hospital stay, and NPRCs are provided in Tables S6-S9, respectively.  Table 9 provides PS matched demographic and perioperative variables. Table 10 shows that matched SSRF (+) patients sat up earlier (p = 0.002) and had lower NPRCs (17.0% vs. 6.6%; p = 0.019) despite not having shorter ventilator days, ICU stays, and hospital stays. Although not statistically significant, SSRF reduced the need for tracheostomies as well (3.8% vs. 0.0%; p = 0.121). Table 9. Postpropensity scores matching demographic and perioperative variables of the patients.  Flail chest, the presence of three or more contiguous ribs fractured in two or more places; fracture fixation ratio, total fixated ribs divided by total fractured ribs; SSRF, surgical stabilization of rib fractures.

Discussion
Conservative treatment of multiple rib fractures has been associated with high incidence of pulmonary complications, prolonged ICU and hospital length of stay, and persistence of chronic pain [2,[26][27][28][29][30]. Without surgical intervention, the overall pulmonary complications and mortality ranged between 35% and 48%, and 3% and 18%, respectively [2][3][4]8]. Compared to conservative treatment, SSRF has been shown to improve short-term outcomes and quality of life with flail chest being most extensively studied [19][20][21]26,29]. Both our multivariate regression and PS models did not predict shorter ventilator days, shorter ICU stays, or shorter hospital stays. Despite this, patients who received SSRF sat out of bed earlier (p = 0.002) and had lower NPRCs (p = 0.019). A decrease in tracheostomy requirements was observed without significance, which might have been due to a small number of cases.
An increased number of fractured ribs has been directly associated with the severity of thoracic injury and higher pulmonary morbidity and mortality [3][4][5]. Flagel et al. demonstrated that mortality rate and incidence of pulmonary complications increased significantly when six or more ribs were fractured, so did length of hospital stay for up to seven rib fractures [4]. We found that the number of fractures was correlated with longer ICU stays (p = 0.029) in the SSRF (−) group. However, in the SSRF (+) group, the number of fractures was not associated with ventilator days (p = 0.617), length of ICU stay (p = 0.478), hospital stay (p = 0.706), and NPRCs (OR, 0.734; 95% CI, 0.444-1.211; p = 0.226) after adjusting for age, comorbid conditions, and injury severity. Furthermore, overall NPRCs (7.3%) after SSRF was low in our cohort as compared to both the SSRF (−) group (18.8%; p < 0.01) and previous studies without SSRF. It could also provide indirect evidence of the benefit of SSRF for decreasing NPRCs due to early ICU liberation regardless of the number of broken ribs. Nevertheless, it is noteworthy that, without being adjusted, length of hospital stay was still longer in the group with more than six rib fractures despite comparable ventilator days and ICU stay. We later adjusted the numbers of fractured ribs with both a multivariable regression model (Table 7) and a propensity score model ( Table 10) to show that after surgery fixation, the number of fractured ribs is no longer important, as our title states. This suggested that the number of fractured ribs correlated with the frequency of concomitant lung parenchyma injury (47.8% vs. 28.2%; p < 0.01) and extrathoracic injuries (extremity AIS ≥ 3: 21.7% vs. 4.7%; p < 0.01) and was an indicator of overall trauma severity (ISS: 21.5 vs. 17.0; p < 0.01) resulting in prolonged hospital stay at first glance despite leaving the ICU early. However, after adjustment, the number of rib fractures was not an independent variable of these outcomes.
While the results of SSRF seem positive, the optimal approach is still under debate. Single lung ventilation is advocated by most surgeons because of better exposure and less risk of lung parenchyma injury [17]. Our cohort was in line with this idea with around 70% of patients receiving single lung ventilation. This choice was independent of the number of fractured ribs (p = 0.922) and did not affect operating time (p = 0.242) but was correlated with ISS (single vs. double lung ventilation, 17.0 vs. 19.5; p = 0.047). In the situation of hemoptysis, double lumen intubation is required to maintain a patent airway in advance and to prevent soiling the contralateral lung. Nevertheless, due to increased intrapulmonary right-to-left shunt, pre-existing lung diseases should be evaluated carefully. With the aid of CT images and proficiency in SSRF, preoperative rib thickness measurements and temporary lowering tidal volume or apnea while placing screws is another feasible choice whenever one encounters intolerance of one lung ventilation. Likewise, double lumen intubation is not without risk. Postoperative hoarseness and bronchial injuries had been observed [31]. Furthermore, double lumen intubation is much more technically demanding, time consuming, and requires replacement with a single lumen tube postoperatively once longer ventilation support is necessary. In case of emergency, single lumen intubation followed by thoracotomy is necessary. In our cohort, patients undergoing thoracotomy were almost twice as likely as those receiving video-assisted surgery to have single lumen intubation (40.5% vs. 22.8%; p = 0.026).
Thoracotomy causes destruction of the already injured chest wall. Even modified by muscle-sparing technique, the larger incision of thoracotomy has been shown to increase morbidity and mortality. Since the introduction of video-assisted thoracoscopic surgery (VATS), VATS has been gaining popularity in different aspects of thoracic surgery because of less invasiveness and better postoperative recovery [32]. In our cohort, thoracotomy was associated with longer ventilator days (p = 0.038) but not ICU stay (p = 0.265) after adjustment for comorbidity and ISS in multiple linear regression models. This difference emphasized the great influence of thoracotomy on the integrity of chest wall muscles for weaning from the ventilator. Therefore, video-assisted minithoracotomy was often used in our institute unless contraindicated or the presence of clear indication for thoracotomy. Vertical incision along the medial border of the latissimus dorsi over the epicenter of the fracture site and one camera port at 7th intercostal space were made. As depicted in Figure 3, the third to eighth ribs were involved most often [17,33]. We plated ribs fourth to eighth in priority because they contribute the majority of rib cage stability and generate the most pain [12]. Furthermore, rib fractures most often involved PL (65.9%), an anatomical weak point and the point of impact loading, followed by AL (23.0%), which was consistent with the findings of Liebsch et al. [33]. Therefore, posterior vertical incision sufficed most of the time. Dissecting beneath muscle fascial planes anteriorly and posteriorly, AL and p injuries could be fixated by means of thoracoscopic localization and right-angled devices.
Our present study faced several limitations such as relatively small sample size, and a single-center experience. Because SSRF is not covered by national health insurance in our country, each rib plate costs around 2000 USD, resulting in selection bias. Consequently, many patients might not receive SSRF due to price rather than severity, which also impacts the choice of appropriate control group. In addition, some victims were not covered by private health insurance, which greatly limited the appropriate number of stabilizations. Patient-reported outcomes were not measured and long-term benefits or complications of SSRF could not be evaluated on account of a lack of chronic pain assessment. Further studies with a larger sample size are needed to find prognostic factors as to better define those who benefit most from SSRF.

Conclusions
The risk of NPRCs was reduced through an algorithmic approach. Once surgically fixated through a personalized approach, the number of fractured ribs, even more than six ribs, was not associated with longer ventilator days, length of ICU stay, length of hospital stay, and NPRCs. Poor consciousness and older age were risk factors for NPRCs independent of the number of rib fractures.

Supplementary Materials:
The following supporting information can be downloaded at: https:// www.mdpi.com/article/10.3390/jpm12111844/s1, Table S1: Preoperative variables between unmatched groups; Table S2: Outcomes between unmatched groups; Table S3: A comparison of the perioperative variables of SSRF group with and without more than six fractured ribs; Table S4: p value of univariate analyses of SSRF (+) group; Table S5  Institutional Review Board Statement: Ethical review and approval were waived for this study, due to retrospective observational studies based on databases using unlinkable anonymized data.
Informed Consent Statement: Patient consent was waived because of the retrospective nature of the study and the analysis used anonymous clinical data.

Data Availability Statement:
The data that support the findings of this study are available from National Cheng Kung University Hospital, but restrictions apply to the availability of these data, which were used under license for the current study, and so are not publicly available. Data are, however, available from the authors upon reasonable request and with permission of correspondence.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.