A large single-center propensity score-matched cohort study on outcomes and complications based on the number of corrective bars used in the Nuss procedure Authors: Piotr Jerzy Skrzypczak, Monika Rozmiarek, Tomasz Dobiecki, Magdalena Sielewicz, Michał Suchodolski, Magdalena Roszak, Cezary Piwkowski, Krystian Pawlak Published: 2024 ABSTRACT AbstractThe Nuss procedure is the most common corrective surgery for pectus excavatum. We analyzed treatment outcomes and complication rates in 1238 patients treated with the Nuss procedure from 2002 to 2021, focusing on the number of corrective bars used. Using Propensity Score Matching based on age, sex, BMI, pre-operative FEV1, and the Haller index, we created two groups: 546 patients with a single bar and 546 with two bars. Both groups achieved similar correction effects (Haller index: single bar = 2.58 vs. two bars = 2.56; p = 0.65). In the univariate analysis, in the two-bar group, the postoperative complications were observed more often (28.6% vs. 15.4%, p < 0.001), including pneumothorax (11.2% vs. 6.2%, p < 0.001), hemothorax (3.7% vs. 0.7%, p < 0.001), additional drainage (13.7% vs. 5.3%, p < 0.001), the need for thoracentesis (8.6% vs. 2.9%, p < 0.001), bar displacement (3.8% vs. 0.7%, p < 0.001), pleural effusion (10.6% vs. 3.1%, p < 0.001), and fever (6.6% vs. 3.8%, p < 0.041). In the logistic regression, two bars significantly increased the risk of postoperative complications (p = 0.019), including hematoma (p = 0.036), pleural effusion (p = 0.002), and the need for thoracentesis (p = 0.013). Using two corrective bars during the Nuss procedure is associated with a higher rate of postoperative complications but similar corrective results.Keywords: Nuss procedure, Pectus excavatum, Corrective bars, Haller index, Bar rotationSubject terms: Diseases, Health care, Medical research, Risk factors, Signs and symptoms Introduction Pectus excavatum (PE) is the most common congenital chest wall deformity in humans1,2. PE affects 0.1% of all live births and usually manifests during the neonatal period1–3. Despite being prevalent, the exact cause of PE remains unclear. Several hypotheses have been proposed, including developmental disorders, overgrowth of costal cartilage, and genetic predispositions1. PE decreases patients’ quality of life in both their mental and physical well-being4. While some PE patients may be asymptomatic, others could complain of physical disorders (dyspnea, chest pain, and palpitations) or psychosocial symptoms (body image concerns and depression)4–6. Both cosmetic and functional impairments are indications for surgical correction4–6. In the treatment of PE, depending on the patient’s age and the advancement of the deformation, conservative or surgical methods can be utilized7–10. Non-surgical interventions, such as the vacuum bell or silicone/polyethylene implants, are available options for less advanced deformities in younger patients7,8. However, surgery remains the primary intervention for older patients6,10. Since its introduction in 1998, the method proposed by Donald Nuss10 has become increasingly prevalent and is now one of the most widely practiced operative treatments for PE globally10. The Nuss procedure involves minimally invasive insertion of one or more appropriately adjusted metal bars behind the sternum to correct the depressed chest wall11,12. Advantages of the Nuss procedure include minimal invasiveness, reduced surgical trauma, and fewer PE recurrences after bar removal13, distinguishing it from the Ravitch method14. The most common postoperative complications after the Nuss procedure include bar displacement, pneumothorax, wound infection, pleural effusion, or chronic pain15,16. The number of bars used during the Nuss procedure has not been precisely determined. Many surgeons consider it dependent on the severity of the defect or the patient’s characteristics, such as age, weight, or height17–19. Furthermore, the issue of precise indications of when a patient should receive multiple bar treatment is also rarely discussed in the literature. The aim of the study was to analyze the treatment outcomes and the frequency of complications in patients with a PE treated with the Nuss procedure, depending on the number of corrective bars used. Methods The Bioethics Committee of Poznan University of Medical Sciences approved all methods and waived the requirement for obtaining informed consent for the collection and analysis of anonymized data, as well as for the publication of results from this single-center, retrospective cohort study. The study was conducted in accordance with the Declaration of Helsinki (as revised in 2013). The study included 1238 patients with PE who underwent surgery between 2002 and 2021 using the Nuss procedure. The patients were preoperatively examined and were qualified for the operation on an outpatient basis. The severity of the PE was evaluated using the Haller index, which was based on two chest X-ray projections: postero-anterior and right lateral. According to Khanna et al.20 and Rattan et al.21, the Haller index based on radiography correlates well with that calculated on computed tomography (CT). Routine preoperative thoracic CT scans were not performed due to their potential malignancy and cumulative lifetime radiation exposure21–23. Chest CT scans were limited to cases of advanced symmetrical or significantly asymmetrical deformities. The main indications for surgical treatment were cosmetic concerns, dyspnea on exertion, or a Haller index ≥ 3.5. The surgical procedure was performed under general anesthesia and the modified Nuss method initially described in 199810. All patients received additional analgesia using an epidural infusion of 0.25% bupivacaine. Each operation was performed by an experienced team of 2–3 thoracic surgeons, who performed 70–90 Nuss procedures annually, Correction of the deformity involved the insertion of one or two appropriately bent correctional steel bars (BBH Mikromed, Dąbrowa Górnicza, Poland). Videothoracoscopy was utilized during each operation, and the bars were inserted through a single incision, in a parallel position, retrosternally. An additional stabilizing bar was introduced to prevent implant displacement, typically following the insertion of the first bar, and secured to the adjacent ribs using absorbable sutures. After discharge, patients were provided with comprehensive instructions on gradually increasing physical activity, along with specific guidelines on activities to avoid during the early postoperative period (e.g., contact sports, martial arts, weightlifting). Subsequent follow-up visits were conducted at the thoracic surgery clinic, with the first of which took place approximately one month after discharge. The corrective bars were usually maintained for 2–3 years following the initial surgery. Then, the patient was readmitted to the thoracic surgery department, and under general anesthesia, the corrective bar was removed. In the first 24 h after the operation, the patient underwent a chest X-ray. The X-ray was used in our article to measure the final Haller index. In our paper, we evaluated the frequency of postoperative complications and the outcomes of the correction. Identification of the surgical complications consisted of identifying the most important complications in the early postoperative period (pneumothorax, hemothorax, the need for additional drainage or thoracentesis, fever), as well as assessing later complications that were found in a one- or two-month follow-up, after the patient was discharged home, during a follow-up visit in the thoracic surgery clinic (bar rotation, pleural effusion). The need for patient reoperation was also assessed, which most often resulted from the rotation of the corrective bat or insufficient correction performed during the first operation. The comparison of surgical outcomes involved evaluating the depth of the chest defect using the Haller index before the corrective bars were inserted and after 2–3 years, when the corrective bar was removed. The basic characteristics are presented in Table 1. Table 1Basic characteristics of the studied group.Single-bar group (n = 692)Two- bars group (n = 546)Age (years, median, Q3-Q1)16 (20 − 15)18 (23 − 15)Sex, n (%)- Male,- Female,− 562 (81.2%),− 130 (18.8%),422 (77.3%),124 (22.8%),BMI (kg/m2, median)19,8 (20.5–17.9)19,8 (24.3–20.1)FEV1% (median)89 (93.1 − 78.8)89 (90 − 80)Pre-operative Haller index (median)3.4 (4.3-3)3.7 (4.4-3) Basic characteristics of the studied group. Sex, n (%) - Male, - Female, − 562 (81.2%), − 130 (18.8%), 422 (77.3%), 124 (22.8%), Statistical analysis The analyzed data were expressed as mean ± standard deviation, median, minimum, maximum values, interquartile range (Q1 lower quartile, Q3 upper quartile), or percentage, as appropriate. The relationship between variables was analyzed using Spearman’s rank correlation coefficient. The normality of distribution was tested using the Shapiro–Wilk test, and equality of variances was checked using Levene’s test. A comparison of two unpaired groups was performed using the unpaired t-test for data that followed a normal distribution and had homogeneity of variances or the Mann–Whitney U-test. Categorical data were analyzed using the χ2 test when the sample size was larger than 40, and all expected values were greater than ten; for other situations, the exact test of Fisher or χ2 test with Yate’s correction was used. All results were considered significant at p < 0.05. To create two comparable groups, we used the Propensity Score Matching analysis. To calculate the predicted probability, the following variables were used: age, sex, BMI, pre-operative FEV1, and the Haller index based on the closest value of the predicted probability. The most appropriately matched pairs were selected using a “nearest neighbor” matching algorithm without replacement with a caliper of 0.01. Data manipulation and all calculations were performed in IBM® SPSS® Statistics version 27th (PS Imago Pro 8). Results Using the propensity score matching analysis, we obtained two groups that are similar in terms of basic characteristics (Table 2). The process of performing the Propensity Score Matching analysis is illustrated in Fig. 1. A similar correction effect assessed with the Haller index was achieved (single-bar group = 2.58 vs. two-bars group = 2.56; p = 0.65). The basic characteristics of the groups after Propensity Score Matching are presented in the Table 2. Table 2The basics characteristics of the studied groups after propensity score matching and the results of the comparison in terms of the complications rates and the effects of the correction.Single-bar group (n = 546)Two- bars group (n = 546)p-valueAge (years, median, Q3-Q1)16 (20 − 14)18 (23 − 15)0.08Sex, n (%)- Male,- Female,− 446 (81.6%),− 100 (18.4%),422 (77.3%),124 (22.8%),P = 0.1BMI (kg/m2, median)19,6 (21.5–18.0)19.8 (21.9–18.1)0.440FEV1% (median)91.1 (101 − 82.1)91 (100.5–82.5)0.562Pre-operative Haller index (median)3.6 (4.0-2.9)3.8 (4.4-3)0.07Post-operative Haller Index2.6 (2.8–2.2)2.6 (2.9–2.2)0.650Complications Total numer of the complications84 (15.4)156 (28.6) < 0.001 Pneumothorax34 (6.2)61 (11.2) 0.004 Hematoma4 (0.7)20 (3.7) < 0.001 Pleural effusion17 (3.1)58 (10.6) < 0.001 Additional drainage29 (5.3)75 (13.7) < 0.001 Thoracocenthesis16 (2.9)47 (8.6) < 0.001 Bar displacement4 (0.7)21 (3.8) < 0.001 Wound infection3 (0.5)2 (0.4)0.5 Pain in the chest5 (0.9)11 (2.0)0.131 Fever21 (3.8)36 (6.6) 0.041 Deformation reccurence9 (1.6)9 (1.6)0.99 Lack of corrections3 (0.5)4 (0.7)0.5 Re-hospitalization23 (4.2)45 (8.2) 0.06 Reoperation19 (3.5)39 (7.2) 0.07 Other complications14 (2.6)19 (3.5)0.374Significant values are in bold. The basics characteristics of the studied groups after propensity score matching and the results of the comparison in terms of the complications rates and the effects of the correction. Sex, n (%) - Male, - Female, − 446 (81.6%), − 100 (18.4%), 422 (77.3%), 124 (22.8%), Significant values are in bold. The flow diagram summarizing the subsequent stages of propensity score matching analysis. In the two-bars group, postoperative complications were observed significantly more often (28.6% vs. 15.4%, p < 0.001), including pneumothorax (11.2% vs. 6.2%, p < 0.001), hemothorax (3.7% vs. 0.7%, p < 0.001), additional drainage (13.7% vs. 5.3%, p < 0.001), thoracentesis (8.6% vs. 2.9%, p < 0.001), bar rotation (3.8% vs. 0.7%, p < 0.001), pleural effusion (10.6% vs. 3.1%, p < 0.001), and fever (6.6% vs. 3.8%, p < 0.041). In the logistic regression analysis, the following factors were analyzed: age, sex, preoperative Haller index, BMI, and the number of bars used. Implantation of two corrective bars significantly increased the risk of postoperative complications (p = 0.019), including hematoma (p = 0.036), effusion (p = 0.002), and the need for thoracentesis (p = 0.013). The results of the logistic regression are presented in the Table 3. Table 3Results of the logistic regression.Type of the complicationp-valueOR (95% CI)Total number of the post-operative complications Age0.7930.99 (0.916–1.069) Sex0.9111.057(0.4-2.792) BMI0.6080.958 (0.814–1.128) Pre-operative Haller index0.6061.094 (0.777–1.542) Number of bars used 0.019 2.636 (1.172–5.929) Pneumothorax Age 0.049 0.785 (0.617–0.999) Sex0.2232,399 (0.588–9.794) BMI0.2740.862 (0.66–1.125) Pre-operative Haller index0.7351.092 (0.655–1.821) Number of bars used0.3052.014 (0.528–7.680)Hematoma Age0.4051.051 (0.934–1.183) Sex0.9490.942 (0.155–5.738) BMI0.1581.270 (0.911–1.771) Pre-operative Haller index 0.026 1.954 (1.082–3.528) Number of bars used 0.036 6.874 (1.129–41.844) Pleural effusion Age0.9771.002 (0.9-1.115) Sex0.1770.226 (0.026–1.960) BMI0.2351.172 (0.902–1.522) Pre-operative Haller index0.7271.120 (0.592–2.119) Number of bars used 0.002 8.914 (2.243–35.431) Additional drainage Age 0.026 0.783 (0.631–0.971) Sex0.3431.944 (0.492–7.681) BMI0.920.988 (0.776–1.257) Pre-operative Haller index0.6271.128 (0.695–1.831) Number of bars used0.0852.887 (0.864–9.653)Thoracocentesis Age 0.024 1.121 (1.015–1.238) Sex0.160.287 (0.05–1.635) BMI0.8220.971 (0.755–1.25) Pre-operative Haller index0.1901.379 (0.853–2.228) Number of bars used 0.013 4.42 (1.369–14.272) Bar displacement Age0.5090.04 (-0.054-0.11) Sex0.7230.04 (-0.067-0.096) BMI0.6610.043 (-0.103-0.07) Pre-operative Haller index0.3060.026 (-0.08-0.025) Number of bars used 0.048 0.043 (0.001–0.17) Fever Age 0.049 0.842 (0.708–1.001) Sex0.6990.728 (0.143–3.699) BMI0.5451.072 (0.854–1.346) Pre-operative Haller index0.1270.54 (0.243–1.201) Number of bars used0.082.954 (0.872–10.009)The influence of age, sex, BMI, pre-operative Haller index, and number of bars used on the occurrence of post-operative complications.Significant values are in bold. Results of the logistic regression. The influence of age, sex, BMI, pre-operative Haller index, and number of bars used on the occurrence of post-operative complications. Significant values are in bold. Discussion We found that patients who were operated on with one correction bar and those who received two correction bars achieved similar correction effects as measured by the Haller index. However, the use of two bars during the Nuss procedure was associated with a significantly higher incidence of complications, including hematoma, pleural effusion, and the necessity for thoracentesis. Preparing the intercostal space for the insertion of an additional corrective bar may lead to postoperative complications such as hematoma or pleural effusion. These complications often necessitate decompression and thoracocentesis, as evidenced in our results. Therefore, in our center, the standard practice is to introduce one correction bar if the advancement of the deformation, anatomical considerations, and the patient’s age permit. An additional bar is frequently added in older, taller, and more severely deformed patients. Since the introduction of the Nuss procedure in the recent century10, the method has gained significant popularity and has undergone several modifications6,24. Initially, the method involved using a single bar for the majority of patients with PE. However, as the technique became more widespread, the surgical indications were broadened, particularly to encompass adult patients. In their case, multiple correction bars were often utilized6,25. Moreover, the advancement of thoracoscopy and its potential for routine usage during correction bar insertion ensures the safety of the procedure and reduces the occurrence of severe complications26,27. The procedure yields favorable corrective outcomes for PE. While it is minimally invasive, it remains a surgical intervention associated with the potential for serious complications. The risk of complications increases with patient age or when utilizing multiple correction bars6,27–29. Bar displacement remains one of the most severe complications following the Nuss procedure30. The incidence of this complication is estimated to range from 1.8 to 16.6%10,31–33, with reoperation required in 3.4–27% of cases30. Several surgical methods are employed to prevent common mechanisms of bar displacement, such as bar flipping, lateral sliding, or hinge point disruption27. Most techniques facilitate the solid attachment of the bar to the ribs: stabilizers, bar fixation with stainless-steel wire or a shorter bar34–36. Other common complications include postoperative pneumothorax, pleural effusion, pericardial effusion, hemothorax, and infectious complications6,27,29. The number of bars used remains a topic of discussion, with some centers considering the insertion of two correction bars as a standard practice. Stanfill et al. emphasized in their study the efficacy of the Nuss procedure employing two corrective bars to mitigate bar rotation and enhance bar stability30. However, they involved relatively small sample sizes of 58 and 27 patients, which were also unequal in size. In our propensity score-matched analysis, we observed a statistically significant higher incidence of bar rotation in the two-bars group. A similar finding was reported by Uemura et al., who used two bars with lateral stabilizing in high-risk patients37. Conversely, achieving satisfactory correction of deep sternal depression in PE often requires the use of two corrective bars. Chen et al. noted more infections in patients who underwent insertion of two corrective bars17. Furthermore, they emphasized the importance of lower weight and the value of the Haller index as essential risk factors that elevate the likelihood of severe infections. They suggested a cut-off point of 50 kg as the target group of patients who can be safely and effectively undergo surgery with two correction bars17. An important challenge in corrective PE surgery is the issue of operating on adult patients. It is recommended to perform surgery on younger individuals, particularly adolescents aged 12–1618,19, as they tend to experience better cosmetic outcomes and fewer serious complications. This is attributed to the constitution of the skeletal system and anatomical factors, which contribute to a higher frequency of post-operative complications in adults6,27. However, experienced centers also operate on adult patients, using two corrective bars as a standard and only sporadically using two-bars in adolescents38. Ben et al.39 reviewed the effect of treatment on 153 patients with significantly asymmetric PE who underwent multiple-bar Nuss operations. However, their study exhibits a notable disparity in the size of the compared groups: 151 patients received treatment with two bars, while only 2 cases were treated with three bars. The authors obtained favorable results: displacement of a bar only in one case, bar exposure in seven cases, and high scores in post-operative satisfaction questionnaires. The authors of the study recommend the multiple-bar Nuss operation for patients with significantly deformed PE as a safe and effective method that yields optimal cosmetic results39. Although the authors acknowledge that one bar may suffice in many cases, they emphasize the superiority of employing two bars to achieve better cosmetic outcomes in both the lying and standing positions39. In their compelling study, Nagaso et al.40 conducted a detailed analysis of postoperative pain in patients undergoing the Nuss procedure with the implantation of one or two corrective bars. They used a patient-controlled anesthetic system to assess the requirement for painkillers in both groups and observed more severe pain in the single-bar group. Furthermore, they performed a biomechanical analysis, revealing that utilizing double bars reduces stress on the thoraces. The authors recommend double-bar application for patients with advanced PE and difficult correction conditions. Such treatment ensures more effective cosmetic results, reduces postoperative pain, and improves the biomechanical correction conditions. However, their study involved very small groups of 14 vs. 10 patients. According to Nagaso’s study40, more favorable biomechanics of correction using double-bar correction causes pressure dispersion. According to the authors, it minimizes the possibility of displacement of the bar. In our study on a very large group of patients, it is difficult for us to agree - we found that in the case of double-bar correction, rotation of one of the bars occurred more often. Lo et al. analyzed a cohort comprising nearly 300 patients, with a majority undergoing the use of two or even three correction bars. By employing bilateral thoracoscopy for precise assessment of intraoperative deformation, they achieved an overall complication rate of 6.8%. Notably, they found that the postoperative complications rate was significantly related only to the Haller index24. Bar displacement necessitating reoperation emerged as the most common complication reported in their study. Early complications included pneumonia, delayed pleural effusion, and postoperative wound infections, while late complications involved chronic postoperative pain. Although patients with three bars experienced longer hospital stays, the difference was not statistically significant24. The authors acknowledge several limitations of the study. Firstly, our study has the character of a retrospective cohort study. Our study has no precise indications for using one or two corrective bars in patients. For the most part, the decision was based on the surgeon’s experience as well as the patient’s pre-operative imaging and physical examination. Consequently, there exists a possibility of selective bias. The surgeries were performed in a high-volume center by an experienced team. A satisfaction survey was not used in the studied cohort, which could constitute a valuable, additional assessment of surgical appearance results. However, we preferred to focus on more objective methods, such as radiological examination, Haller index, or postoperative course. but in the case of the assessment of resultsTo enhance the reliability of the data analysis, Propensity Score Matching was employed, incorporating key confounders mentioned in the literature to calculate predictive probabilities. Additionally, the study included one of the largest cohorts available, comprising over 1,000 patients operated on in a single center. While this is a single-center study, it draws upon multi-years of experience, thereby enriching the depth of the findings. Conclusions The utilization of two correction bars during the Nuss procedure is associated with a higher rate of severe postoperative complications. The utilization of two correction bars during the Nuss procedure is associated with a higher rate of severe postoperative complications. The correction effects demonstrated similarity across both patient groups, irrespective of the number of correction bars employed. The correction effects demonstrated similarity across both patient groups, irrespective of the number of correction bars employed. The use of a single corrective bar should be considered the standard management in the Nuss procedure performed for pectus excavatum. The use of a single corrective bar should be considered the standard management in the Nuss procedure performed for pectus excavatum. Acknowledgement The work presented in this publication was conducted with the participation of the Greater Poland Center of Digital Medicine. The Center is funded by the Medical Research Agency (S. Moniuszki 1A, 00-014 Warsaw, Poland) grant number 2023/ABM/02/00007-00, financed from the state budget (29 999 256,00 PLN). Author contributions Conceptualization, P.S. and K.P.; Data curation, M.R, T.D., M.Su. and M.S.; Formal analysis, P.S. and M.R.; Investigation, P.S., M.R. and M.Su.; Methodology, P.S and K.P.; Project administration, C.P.; Resources, M.R.; Software, P.S.; Supervision, K.P. and C.P.; Visualization, P.S.; Writing—original draft, P.S. and K.P.; Writing—review and editing, C.P., M.S., M.Su., K.P., M.R. and T.D. All authors have read and agreed to the published version of the manuscript. Data availability The datasets used and/or analysed during the current study available from the corresponding author on reasonable request. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. References Associated Data Data Availability Statement The datasets used and/or analysed during the current study available from the corresponding author on reasonable request.