Pectus excavatum: Treatment

INTRODUCTION — Pectus excavatum (PE) is a deformity of the chest wall characterized by sternal depression, which typically begins over the midportion of the manubrium and progressing inward through the xiphoid process.

The clinical significance of PE depends on three issues:

●Severity of the chest wall defect

●Cardiopulmonary morbidity

●Psychosocial impact of the defect and its appearance on the patient

Despite decades of experience with surgical and nonsurgical treatment and efforts to quantify outcomes in each of these areas, the decision of when and how to treat PE remains controversial. These issues are discussed in this topic review.

The etiology, natural history, and evaluation of a child with PE are discussed in a separate topic review (see "Pectus excavatum: Etiology and evaluation"). The diagnosis and treatment of pectus carinatum and other chest wall deformities are also reviewed separately. (See "Pectus carinatum and arcuatum" and "Chest wall diseases and restrictive physiology".)

INDICATIONS — Surgical correction of PE improves physical appearance in most patients and cardiorespiratory function in some, but the indications for intervention are not fully standardized.

Variables measured as surgical outcomes include objective improvements in PE severity and cardiopulmonary function and the patient's subjective report of symptoms and concerns about appearance.

The evaluation of a patient with PE is discussed in a separate topic review. For patients with moderate or severe PE on visual inspection, a computed tomography (CT) scan should be performed to determine the pectus severity index (PSI), which is the ratio of the lateral diameter of the chest to the distance between the sternum and spine, at the point of maximal depression. (See "Pectus excavatum: Etiology and evaluation", section on 'Evaluation'.)

We and others generally consider corrective surgery in patients with two or more of the following criteria [1-3]:

●PSI of >3.25 (measured on CT scan)

●Cardiac compression, displacement, mitral valve prolapse, murmurs, or conduction abnormalities

●Pulmonary function testing showing restrictive respiratory disease

●Failed previous repair of PE

Other factors supporting surgery in some cases are documented progression of the PE deformity with worsening of physiologic symptoms or significant patient concern about appearance [1,4]. A retrospective review of patients at a single center who underwent surgical correction of PE explored the correlation between physical deformity (measured as PSI) and both physical and psychological ailments [5]. While over 90 percent of the patients had physical complaints and over 75 percent had psychological stress related to their PE, there was no correlation between PSI and either the physical or psychological complaints.

TIMING — When surgical intervention to treat PE was first contemplated more than 60 years ago, experts advocated surgical intervention early in life, ideally before 20 months of age. Early intervention was advocated because PE was hypothesized to be caused by imbalance of muscular forces on the sternum and could be treated by modifying the cartilage and bone structure and relationship of the sternum, costal cartilages, and ribs [6,7]. Moreover, surgery performed during or after adolescence had high morbidity and failure rates in that era [8]. This led to the philosophy that "prophylaxis is better than cure" [7]. However, it became clear that children undergoing early surgical intervention had a high risk for recurrence of PE during adolescence, and extensive resection of costal cartilage in young children also restricted chest wall growth in some cases [9]. This experience has led to later timing of surgical intervention for PE.

The general consensus is that the optimal time to correct a pectus defect is, at the earliest, eight years of age and certainly before the end of adolescence [10]. Children in this age range are young enough to have compliant costal cartilage that is amenable to remodeling but old enough to reduce potential for recurrence during the pubertal growth spurt [11-14]. This consensus is based on collective experience and published case series but has not been established through organized clinical trials. It is possible that the ideal timing for surgical repair varies with the surgical approach, but this question has not yet been examined.

TYPES OF SURGERY

Historical approaches — Surgical options to correct PE have evolved from experience with historical approaches.

During the 1920s, Sauerbruch resected both deformed ribs and cartilage, but this approach left a flail chest and had a 24 percent mortality rate [2,15].

During the 1930s, Brown performed a more limited resection, removing abnormal cartilage, performing a sternal osteotomy, and releasing the sternum from the inward pull of the diaphragm. Postoperatively, he used external traction to keep the repair in place. However, patients were prone to infection, presumably due to the external traction [15]. The external traction was discarded in the 1940s; as a result, the PE often recurred postoperatively.

During the 1940s, theories about abnormal diaphragmatic connections led to an even more limited surgical intervention, in which the central tendon of the diaphragm and substernal ligament were released [16]. The surgery was performed during infancy or early childhood. This approach was discarded because it was not effective. Furthermore, modern imaging and thoracoscopy have refuted the mechanistic theory underlying the approach [2].

Current approaches — Two surgical procedures are in active use today, with minor variations in technique:

●Nuss procedure – The Nuss procedure (referred to as the "minimally invasive" reconstruction of PE) is the most common approach to PE repair. It corrects the pectus defect without cartilage resection by applying outward pressure to the sternum at the point of maximal inward deflection, using a custom-contoured steel bar ("Nuss bar"). The Nuss bar is placed in the pleural space, passed behind the sternum, rotated 180 degrees, and then attached laterally to the outer edge of the rib cage (figure 1) [15].

The Nuss procedure has also been modified to reduce the sternal stress, by performing "relaxing" cartilage incisions [17] to better stabilize the bar on the rib cage using lateral stabilizers [13,18] and to correct longer and more complicated defects by using multiple bars [18-22]. Older patients (eg, older adolescents or adults) have less flexible rib cages, but the Nuss bar can be used successfully in this age group with some modifications of technique. For example, the Nuss bar has been used successfully in older patients, with successful correction in over 96 percent of patients from 18 to 29 years and in over 88 percent of patients 30 years of age or older [23]. In one case series, patients in the late teenage years and older required more bars for correction than younger patients (eg, younger adolescents), but the duration of bar placement did not differ between these age groups [24]. A "morphology-tailored" modification of the procedure has been used effectively, in which differently shaped bars are selected according to the morphology of the patient's defect [25].

The Nuss bar is typically removed after approximately two years. There is no objective evidence to guide the timing of the bar removal. Some authors have suggested that bar removal should be delayed until the patient is ≥18 years and has had the bar in place for at least four years to minimize the risk of recurrence [26]. Others suggest that measures of bone density, as a marker for skeletal maturation, might be a clinically useful predictor of success after bar removal [27]. This approach has not been fully validated in an organized clinical trial.

●Modified Ravitch procedure – The Ravitch procedure (referred to as the "open procedure" in some reports) consists of resection of the subperichondrial cartilage, a posterior sternal osteotomy, and temporary internal fixation to support the sternum (figure 2) [2,15]. The Ravitch procedure was first performed during the 1940s and has been modified since then to improve internal stabilization of the sternum, maximize the quality of the correction, and minimize complications [12,28,29]. Various forms of sternal fixation have been used for this procedure, including removable sternal fixation along the rib cage with a connecting strut through the sternum [29], and a synthetic mesh placed under the sternum and attached to the rib cage laterally (figure 3) [30]. The sternal support is then kept in place for at least 12 months [31].

The relative benefits and risks of these approaches are discussed below. (See 'Outcomes' below and 'Complications' below.)

Other approaches

●Sternal suction – A preliminary case series suggests that suction applied externally to the sternum can reduce the sternal depression by approximately 1 cm per month [32,33]. The suction device is used for one or more hours daily for 12 to 15 months. Several case series suggest significant improvement in PE during the course of treatment [32-35]. In two studies, 20 percent of patients achieved an "excellent" result (defined as a normal chest wall depth); this outcome was more likely among patients who had less severe deformity, were younger, or had more flexible chest walls [36] and/or had longer duration of treatment [37]. The device has also been used as an adjunct to conventional surgical correction.

●Sternal magnet – The magnetic Mini-Mover procedure (3MP) has been designed to lift the sternum using magnetic attraction between a magnet attached to the sternum and another magnet on an external sternal brace [38]. This approach is a paradigm shift in that it applies a small force over a long period of time, as opposed to a large force over a short period of time as applied in the Nuss procedure. The persistent gentle force allows the sternum to slowly remold. This contrasts with the large forces employed in the modified Ravitch and Nuss procedures and potentially avoids the associated surgical trauma and pain. Two pilot studies examined the safety and efficacy of the 3MP over an 18- or 24-month treatment period, with an additional one-year follow-up period after the 3MP was removed [39,40]. There was no significant change in the pectus severity index (PSI) in either group of study subjects, but there was a trend toward improvement in younger subjects. Patients reported good comfort while wearing the brace and were generally satisfied with the outcome [40]. Adverse effects included one postoperative pneumothorax requiring chest tube placement, one surgical site infection, and one pericardial effusion. Unfortunately, the device broke in nearly one-half of the subjects, despite a redesign between the two studies.

●Prosthetic inserts – Some patients with PE but no significant symptoms have been treated with silicone prosthetic inserts, which improve the appearance of the chest without changing the sternal contour [41-43]. Initial problems with migration of the inserts were addressed by coarsening the surface and perforating the prosthesis to promote attachment to native tissue [41]. One group has reported perioperative problems with seroma in 31 percent of patients, all but one of which was treatable and temporary [42].

●Physical therapy – Physical therapy interventions have been offered to patients with mild PE who do not desire or are not candidates for surgical correction [44,45]. Physical therapy strategies attempt to improve the interplay between respiratory muscles and chest wall by improving posture and increasing chest wall expansion, harkening back to some of the early hypotheses that suggested that PE is caused by abnormal muscle forces acting on the rib cage. The intervention includes exercises designed to strengthen the thoracic musculature and promote use of specific muscle groups during breathing and breath holding [44,45]. Outcomes have been published only as scattered case reports, so the efficacy of this technique is unclear.

OUTCOMES — There are no large prospective studies comparing outcomes of the Nuss and Ravitch procedures. The surgical approach usually depends on the preferences of the surgeon and institution. Therefore, robust comparisons between the procedures will require a well-organized multicenter study, rather than the retrospective case series that are in the literature.

Based on these indirect comparisons, outcomes of appearance, self-reported exercise capacity, and objective measures of cardiopulmonary function appear to be similar after the two procedures, as described below.

Appearance — Most patients experience subjective improvement in self-image and appearance after either the modified Ravitch or Nuss procedure [12,26,46]. As examples:

●A large series of adult and pediatric patients undergoing the modified Ravitch procedure at a single center reported that 94 percent of patients felt that their results were very good or excellent [12].

●A large multicenter study that included 327 patients, most of whom underwent the Nuss procedure, reported improvement in the pectus severity index (PSI) from a mean of 4.4 preoperatively to 3.0 postoperatively [47]. A separate study that included many of the same patients demonstrated that patient self-image concerns decreased from 68 to 6 percent in female patients and from 40 to 4 percent in male patients when measured two years after Nuss bar placement [1]. In addition, 97 percent of patients liked the way their chest looked, 90 percent were happy they underwent surgery, and there was a significant decrease in overall parental concern.

●In his initial decade of experience using the procedure, Nuss reported that 87 percent of patients had a normal chest appearance or mild residual pectus after surgery [15]. However, the outcomes in adolescent patients tended to be worse than in younger children, perhaps because younger children have a more compliant chest wall. The outcomes were worse initially as the surgical technique was developed and refined, reflecting a learning curve associated with the new procedure.

●A retrospective series described self-reported outcomes among 43 patients who had either the Nuss or modified Ravitch procedure performed at a single institution [48]. At the time of follow-up (average 16.3 months postoperatively), patients who had the Nuss procedure were more satisfied with their appearance and reported less frequent chest pain or discomfort than those who had the modified Ravitch procedure. However, the differences were modest and the lack of randomization in this and other studies precludes firm conclusions about the relative success of the procedures.

●In a series of 1170 patients at a single center, a "morphology-tailored" Nuss procedure was used [25]. Patient-reported satisfaction was excellent in 93 percent and good in 6 percent. Complication rates were modest and decreased over time, as discussed below. (See 'Complications' below.)

Cardiopulmonary function — Subjective improvement in exercise tolerance and subtle improvements in some objective measures of cardiopulmonary function have been reported after either the Nuss or modified Ravitch procedure.

●Nuss procedure – Following the Nuss procedure, pulmonary function outcomes are variable. In some cases, there is a transient decrease in lung function after Nuss bar insertion, followed by a return to or improvement over preoperative values after Nuss bar removal [49-52]. One study in which most patients underwent a Nuss procedure reported a modest improvement in FVC (from 88 to 93 percent predicted) and in total lung capacity (from 94 to 100 percent predicted) after hardware removal [47]. Most patients also report subjective improvement in exercise capacity [51,52]. As with the modified Ravitch procedure, objective measures of exercise capacity (eg, VO₂max) either do not change or improve only slightly after surgery [47,49,51,53,54]. In one study of 20 subjects, VO₂max improved by 10.2 percent [47]. In a separate study of 19 patients, there was no significant improvement in either pulmonary function testing or exertional capacity [55]. In a cohort of 151 patients who underwent PE repair after 30 years of age, preoperative evaluation revealed abnormal cardiopulmonary exercise testing in 67 percent and cardiac compression in 77 percent [23]. After PE repair, there was a significant improvement in right ventricular output. The cardiac outcomes are more consistently positive, with an increase in stroke volume and right ventricular filling [50]. Electrocardiographic abnormalities also tend to improve, probably because the surgery relieves the preoperative cardiac compression [56].

●Ravitch procedure – The effect of the modified Ravitch procedure on forced vital capacity (FVC) is variable. Two small studies reported a decrease or no change in FVC [49,57]. There is often a subjective improvement in exercise symptoms [12,49,57], and one study showed improvement in measures of inspiratory muscle strength [58]. Cardiac outcomes include an improvement in right ventricular filling [50], stroke volume [49,57], and shifting of the displaced heart back to midline [59]. Improvement in oxygen pulse is detectable in some patients [57,58]. Meta-analyses suggest borderline improvement in maximum oxygen consumption (VO₂max) [49,60]. These findings suggest that the improvement in exercise tolerance may be related to improved cardiac function and not pulmonary function [50,61].

This variable improvement in cardiopulmonary function may be due in part to an absence of linkage with the severity of PE defect to more accurately determine postoperative change. Even in the most severe defects, the absolute volume gained in the correction of a defect will be small relative to the total lung volume and any improvement in cardiopulmonary function will likely be related to less restricted cardiopulmonary motion.

COMPLICATIONS — The Nuss and modified Ravitch procedures have different types of complications, but the overall frequency and severity of complications are similar. Each procedure has been successfully modified to address some of these complications.

Nuss procedure

Postoperative pain — Several series suggest that patients undergoing the Nuss procedure have somewhat more postoperative pain than those undergoing the Ravitch procedure and require more analgesia [15,31,59,62,63]. However, in a large series of patients undergoing the Nuss or Ravitch procedure, perioperative pain was managed equally effectively after either procedure and postoperative pain resolved within one month in all patients [64]. Because of their increased need for analgesia, patients undergoing the Nuss procedure generally have slightly longer hospital stays than those undergoing the Ravitch procedure [15,48,59,62,63].

One explanation for the additional postoperative pain after the Nuss procedure is that the act of forcing the sternum into a normal orientation puts a strain on the costal cartilage as well as stress on both the ribs and sternum. The rib stress and cartilage strain can cause small rib and sternum fractures seen on ultrasound [65] and nuclear medicine scans [66]. The risk of fracture is related to age (with an increased risk in patients older than 9 to 10 years [65,66]) and also to the amount of sternal elevation [65].

An important advance in pain management is cryoablation of the intercostal nerves in the region of the repair (termed cryoanalgesia), which prevents afferent conduction and therefore the perception of pain. The nerves eventually heal, and sensation and efferent function is restored. The most prominent advantage is a decrease in the postoperative length of stay and need for postoperative narcotics [67-70]. The combination of cryoablation with intercostal nerve blocks achieves excellent postoperative pain control while reducing or eliminating opioids [69].

Bar displacement — Bar displacement or migration is an important complication of the Nuss procedure. In some cases, the bar is displaced by rotation (90 or 180° posteriorly) and, in other cases, it is displaced laterally [71]. Historically, bar displacement occurred in 10 to 20 percent of patients, but advances in surgical experience and some technical improvements in stabilizer design have substantially reduced the likelihood of this complication [15,26,72-74]. Stabilizing techniques include metal stabilizers, pericostal suture or wire fixation, and use of two bars when indicated. When stabilization techniques are used, bar displacement occurs in 0.5 to 3 percent of patients [25,72,75]. As an example, in a case series of more than 1000 patients undergoing Nuss repair at a single center, use of stabilizing techniques reduced the incidence of bar displacement from 13 percent to 3.2 percent, one-half of which required reoperation [75]. Another series reported fewer episodes of bar displacement with pericostal suture fixation rather than bar stabilizers or wire fixation [26].

Cardiac complications — There have been reports of cardiac perforation during the retrosternal dissection that is performed before passing the Nuss bar into the retrosternal space [76,77]. To avoid this complication, some surgeons elevate the sternum with an external device to better expose the retrosternal space. Transesophageal echocardiography may also be used to ensure safe retrosternal dissection and placement of the Nuss bar.

Postoperative pericarditis or pericardial effusions have been reported in 0.6 to 4 percent of cases, presumably due to the inadvertent entry of the pericardial space [21,74,78].

Pleural effusion — Pleural effusions occur in approximately 3 percent of patients after the Nuss procedure, typically occurring approximately two weeks after the operation [71,79]. In one case series, 80 percent of patients with an effusion were symptomatic, presenting with progressive dyspnea [79]. All affected patients were adults, and most had double bars. The effusion was exudative and responded well to treatment with thoracentesis (for symptomatic patients) and antiinflammatory agents (nonsteroidal antiinflammatory agents or glucocorticoids).

Modified Ravitch procedure

Secondary thoracic dystrophy — Perhaps the most concerning complication of the modified Ravitch procedure is the risk for developing a constricted thorax or secondary thoracic dystrophy (acquired Jeune syndrome). This is probably caused by excessive resection of the costal cartilage, including the growth centers, excessive traction on the perichondrium, or damage of the bone and cartilage growth centers [9,59,80,81]. In this case, normal chest wall growth does not occur and the thoracic cage restricts lung growth. In the initial series of 12 patients with thoracic dystrophy after PE repair, the mean forced vital capacity (FVC) was 33 to 55 percent of predicted and the total lung capacity 32 to 64 percent of predicted [9].

Risk factors for secondary thoracic dystrophy include surgery at an early age (<4 years), resection of five or more costal cartilages, interruption of the growth center between the sternum and the second rib, and attachment of the perichondrial sheaths together behind the sternum [9,80].

To reduce this risk, it has been recommended to delay PE repair until at least until eight years of age and to resect four or fewer cartilages [9]. With these modifications, secondary thoracic dystrophy was not reported in more recent series [12].

There has been some success in repairing secondary thoracic dystrophy in adulthood by reconstructing the anterior chest wall and placing a metal brace inside the sternum [82].

Either procedure — Complications associated with either the Nuss or modified Ravitch procedures include:

Pneumothorax — Clinically significant pneumothorax has been reported in 1 to 7 percent of patients undergoing either the Nuss procedure or Ravitch procedure [62,64,74,77,79,83,84]. In one large series of patients undergoing the Nuss procedure at a single center, pneumothorax rates improved over time, dropping from 7.5 percent at the beginning of the series to 0.8 percent 10 years later [25]. The authors attributed the improvement to improved postoperative drainage (hemo-vac).

Recurrence — When PE recurs after the Nuss procedure, it is usually attributable to bar displacement, premature bar removal, insufficient number of bars, having the bars enter the thorax too lateral, or use of an excessively compliant bar [15,85,86]. In a large series from a single center, only 0.6 percent of patients had recurrences after bar removal, and these were minor [25]. In another series, recurrence was associated with younger age at operation and earlier removal of the bar (mean 2.4 versus 3.8 years with the bar in place); the authors suggested that the bar should not be removed until the patient is at least 18 years old and has had the bar in place for at least four years [26].

Reported rates of recurrence of PE after the modified Ravitch procedure range from 2 to 20 percent. However, these historical statistics may not reflect current risks, because of improvements in surgical technique and changes in patient selection and timing [2,85]. Patient factors that appear to increase the risk for PE recurrence after the modified Ravitch procedure include Marfan syndrome and the patient's age at operation; patients who are either younger or older than the optimal age range have somewhat higher risk of recurrence [12,87,88]. There also can be significant problems due to chondrodystrophy, which is felt to worsen chest wall compliance [86] (see 'Timing' above). Surgical factors that increase the risk for recurrence include inadequate support for the sternum after cartilage resection, failure to reattach the pectoralis and abdominal muscles over the cartilage, disrupting the vascular supply of the sternum, and excision of the xiphoid process [59].

For patients with recurrent PE after the modified Ravitch procedure, reoperation with the Nuss procedure is usually successful [62,85-87,89]. Conversely, for patients with unsatisfactory results after a Nuss procedure, reoperation with a modified Ravitch procedure is often successful [85,89]. A repeat Nuss procedure also has been used effectively in such patients [89]. In one series of 93 subjects in which a Nuss procedure was performed to correct a failed repair of either the modified Ravitch, Leonard, or the Nuss procedure, outcomes were considered "excellent" (based on surgeon assessment) in 75 percent of patients with a prior Nuss procedure and in 63 percent of patients with a previous modified Ravitch procedure [86]. While these outcomes are promising, the outcomes are subjective and without any physiologic or functional correlate.

CHOICE OF SURGERY — The Nuss procedure is performed approximately twice as frequently as the modified Ravitch procedure, and its use is increasing [31,90]. Some surgeons and institutions use the Nuss procedure exclusively [91,92]. Ravitch or open repair may be appropriate and necessary for more complex or asymmetric deformities [93]. The main appeal of the Nuss procedure is its minimally invasive nature, but this is balanced by substantially more pain and need for postoperative narcotic analgesia. Advances in anesthetic techniques, particularly use of cryoanalgesia, has allowed for better control of postoperative pain, has dramatically shortened length of stay, and may promote further adoption of the Nuss procedure [67,68]. One meta-analysis showed that there was a shorter operating room time and less blood loss using the Nuss procedure compared with the modified Ravitch procedure; however, the length of stay was no different [94]. (See 'Postoperative pain' above.)

However, it should be recognized that comparison of the modified Ravitch and Nuss procedures is based on case series and conclusions are limited by lack of randomization, evolution of surgical technique over time, and lack of standardized selection criteria. Based on this limited information, the outcomes of appearance, self-reported exercise capacity, and objective measures of cardiopulmonary function appear to be similar after the two procedures. The procedures have different types of complications, but the overall frequency and severity of complications appear to be similar. In the absence of organized trials comparing outcomes of the modified Ravitch and Nuss procedures, we suggest that the choice of procedure be based on the preference and experience of the surgeon and institution as well as anatomic considerations and patient preferences.

SUMMARY AND RECOMMENDATIONS — Increasing evidence suggests that surgical intervention for pectus excavatum (PE) has benefits to psychological function for many patients but variable physiologic benefit. The indications, timing, and optimal approach for intervention are not fully standardized.

●Clinical assessment – Most patients with mild PE do not require surgical treatment. Such patients should be reevaluated regularly to determine if the deformity is progressing, particularly during periods of rapid growth. They may be offered physical therapy, but the benefit of this intervention is not established. For patients with moderate or severe PE on visual inspection, a CT scan should be performed to determine the pectus severity index (PSI). (See "Pectus excavatum: Etiology and evaluation".)

●Indications for surgery – We suggest corrective surgery in patients with moderate or severe PE (PSI >3.25), significant cardiopulmonary symptoms, cardiac compression, patient concern about appearance, and/or progression of the deformity over time (Grade 2C). (See 'Indications' above.)

●Timing of surgery – For most patients with the above indications, we suggest performing surgical intervention between late childhood and mid-adolescence (Grade 2C). The consensus is that this age range allows for cartilage remodeling while minimizing the risk of recurrence during the puberty. (See 'Timing' above.)

●Techniques – Surgical correction for PE is done with either the Nuss procedure or modified Ravitch procedure. The Nuss procedure is a minimally invasive technique in which a curved bar is inserted to lift the sternum; the bar is removed at least two years later, but some authors suggest that longer time periods are preferable to reduce the risk of recurrence. The modified Ravitch procedure is an open resection of the subperichondrial cartilage and sternal osteotomy, with placement of an internal stabilizing device. (See 'Current approaches' above and 'Recurrence' above.)

●Outcomes – Outcomes for the Nuss and modified Ravitch procedures have not been directly compared in organized trials. Based on case series, overall success for both procedures is high and outcomes of appearance, self-reported exercise capacity, and objective measures of cardiopulmonary function appear to be similar. (See 'Outcomes' above.)

●Complications – The Nuss and modified Ravitch procedures have different types of complications, but the overall frequency and severity of complications appear to be similar. Modifications of the devices and surgical techniques and surgeon experience have reduced the risk of complications and recurrence. Advances in analgesia (particularly cryoablation of the intercostal nerves) have improved pain management after the Nuss procedure. (See 'Complications' above.)

●Choice of surgery – In the absence of organized trials comparing outcomes and complications of the Nuss and modified Ravitch procedures, the choice of procedure should be based on the preference and experience of the surgeon and institution. The Nuss procedure is preferred in many institutions; the modified Ravitch procedure may be appropriate and necessary for more complex or asymmetric deformities. (See 'Choice of surgery' above.)