Abdominoplasty is presently one of the most commonly performed plastic surgery interventions. Although men are increasingly undergoing this type of surgery, it is still performed mainly on women (i.e., one man for every 21 women in the United States in 20151). This intervention, which is often performed on formerly obese or overweight patients with scarred abdomens, requires substantial undermining and is associated with a risk of blood loss and injury to the lymphatic vessels. Its practice has been refined over the years, with a decrease in the associated morbidity and mortality rates.2–5 Indeed, in the United States in 2001, the mortality was estimated to be one of 3281 interventions (approximately 0.03 percent).2 The primary aim of this study was to analyze the incidence of the main postoperative complications, and to search for cause-and-effect links between the complications and the patient’s background and the abdominoplasty techniques used.
PATIENTS AND METHODS
This retrospective study was performed on a population of 1210 women who underwent abdominoplasty between January of 1990 and June of 2014 in the plastic surgery unit at Saint Louis Hospital, Paris, France. Male patients were excluded from the study, as were abdominoplasties performed in the first instance at a unit other than ours, repeated operations, body lifts, and abdominoplasties associated with other surgical procedures. All of the patients received general anesthesia. Indications for surgery were considered after assessment of the area, including looseness of the skin, excess fat, and anomalies of the musculoaponeurotic wall.
The incision was performed following an upward concave line 7 cm from the anterior commissure of the labia majora. This was carried out above the anterior rectus fascia centrally, and above the fascia superficialis laterally. The undermining continued above the umbilicus, which was detached from the upper abdominal skin flap but left attached to the rectus muscles, forming a central channel up to the xiphoid process. The umbilicus was brought out through a new incision in the flap, 10 to 12 cm above the suture line.
The detachment allowed for a resection and closure without tension. Undermining was carried out following the same technique as full abdominoplasty, with dissection below the umbilicus. The umbilicopubic distance remained approximately 10 to 12 cm so as not to compromise the aesthetic outcome. Otherwise, an umbilical transposition was performed.
This technique resulted in a hard-to-disguise xiphopubic scar. Its use was restricted to patients who already had a vertical scar and aponeurotic muscular defects. A pinch test was performed to simulate the resection, taking a conservative approach to avoid tension on the scar. The detachment was limited to the cutaneous area to be resected to preserve the lymphatics and the vasculature.
This technique starts as for an abdominoplasty with umbilical transposition. Once the layer of cutaneous fat was detached upward and the excess resected, a line of staples were put in. The lateral skin was lifted medially as much as possible to obtain a single medial flap, which was resected to form a vertical closure without tension. Undermining was limited as much as possible to avoid compromising the vasculature.
This technique was performed only when a rectus diastasis larger than two fingerbreadths was found during the preoperative examination. It was performed in conjunction with the aforementioned techniques, from the pubis to the xiphoid process. In the case of subumbilical abdominoplasty, only the subumbilical part of the diastasis was treated. Resorbable stitches were used.
Liposuction was performed on the deep and subumbilical layers of the flank and abdomen. Saline solution mixed with epinephrine (1 mg/liter) was injected before liposuction in those areas. Between 500 and 1000 ml of lipoaspirate was removed, but never more than 1000 ml, to reduce operative time and blood loss.
No progressive tension sutures were performed. All of the patients received two corrugated sheet drains (Delbet) positioned laterally near each endpoint of the horizontal suture line. The drains were removed at day 1, before the patient was discharged from the hospital. Two drainage apertures were left on the lateral side of the horizontal scar, which allowed for the passive drainage of seromas. No other drains were used. Healing of the aperture was achieved in 2 to 3 weeks. If needed, in follow-up consultation, we used a sterile cotton wick to painlessly reopen the scar, in the case of early healing. This technique eliminated the need for needle puncture.
All of the patients received the same thromboembolic prophylaxis based on preventative anticoagulation for 2 weeks using low-molecular-weight heparin at 0.4 ml/day, which started 6 hours after the end of surgery. Patients also used antithrombotic compression stockings. Perioperative antibiotic prophylaxis was administered at the induction as a single dose of cefazolin. Patients were encouraged to get up and about from postoperative day 1 to regain their autonomy as soon as possible.
At the follow-up examination, we evaluated six types of complications: hemorrhage, infection of the operative site, skin necrosis, seroma, secondary dehiscence of the scar, and thromboembolic accidents.
The quantitative data were compared between groups with Wilcoxon or Kruskal-Wallis rank test, and the qualitative data were compared with the Fisher’s exact test. Association between complications and patients’ characteristics was studied using a logistic regression model. Significant variables at a 10 percent significance threshold in univariate analysis were candidates for multivariable analysis, performed using a backward stepwise selection procedure.
Of the 1210 eligible patients who were operated on during the inclusion period, 82 were lost to follow-up. The final analysis involved a total of 1128 patients.
The median age of the patients in the cohort was 41 years (range, 16 to 72 years). Fifty-eight patients who were active smokers were included. They claimed to have ceased smoking at least 1 month before the intervention. The median body mass index was 26.80 kg/m2 (range, 23.69 to 30.46 kg/m2). A weight loss of more than 10 kg before the operation occurred in 17.8 percent of the patients (n = 201). Of these, 4.6 percent lost weight after undergoing bariatric surgery (n = 52). Two hundred sixty-eight of the patients had undergone prior abdominal surgery (Table 1).
The most commonly performed intervention was an abdominoplasty with umbilical transposition (71.2 percent), followed by a subumbilical abdominoplasty (23.5 percent). Inverted-T plastic surgery procedures (43 patients) and purely vertical abdominoplasties (17 patients) were rare. These procedures were performed in association with liposuction for 353 patients, a diastasis treatment for 345 of them, and an umbilical hernia treatment for 124 (Tables 2 and 3). The most overweight patients belonged to the group undergoing abdominoplasty with umbilical transposition (average body mass index, 28.2 kg/m2). The patients in the vertical and subumbilical abdominoplasty groups on average had normal body mass indexes (between 22.8 and 24.8 kg/m2) (p < 0.0001). The patient population receiving liposuction had a significantly lower body mass index (p < 0.0001).
One hundred ninety-three patients exhibited a postoperative complication. The most frequent complications were hematomas [n = 64 (5.7 percent)], infections [n = 51 (4.5 percent)], cutaneous necrosis [n = 31 (2.7 percent)], and seromas [n = 30 (2.7 percent)]. We did not encounter a single pulmonary embolism or death (Table 4).
The rate of complications was higher in the case of inverted-T abdominoplasty [n = 9 (20.9 percent)]. Rates were lower for vertical abdominoplasties (17.6 percent) and those involving umbilical transposition (20.2 percent). Subumbilical abdominoplasty had the lowest rate of complications, at 7.2 percent (Table 5).
Obesity was significantly associated with more infections (OR, 3.43; p = 0.0007) and a higher overall rate of complications (OR, 2.35; p = 0.0004). An age of 40 years and older was associated with a higher rate of immediate complications (p = 0.019), and an elevated risk of seroma in particular (OR, 8.19; p = 0.0006), and wound dehiscence (OR, 3.69; p = 0.044). Active smoking was also associated with a higher overall rate of complications (p = 0.0006), although it was mainly with a more substantial rate of cutaneous necrosis (OR, 4.81; p = 0.001) and an increased incidence of infections (OR, 2.62; p = 0.035). All of the other complications had a tendency to occur more frequently with smokers than with nonsmokers, although this difference did not reach statistical significance.
A weight loss greater than or equal to 10 kg was the factor most predictive for the following complications: hematoma (OR, 2.4; p = 0.002), infections (OR, 2; p = 0.030), and seromas (OR, 2.77; p = 0.009). Prior abdominal surgery was not a risk factor for any of the complications that were investigated. In multivariable analysis, smoking (OR, 2.20; p = 0.040), obesity (OR, 1.75; p = 0.026), and age older than 40 years (OR, 2.04; p =0.004) remained the foremost risk factors for all complications. Smoking increased cutaneous necrosis (OR, 4.83; p = 0.001); age older than 40 years (OR, 8.82; p = 0.0004) and major weight loss (OR, 2.48; p = 0.024) represented the two main risk factors for seroma; and obesity was the main risk factor for infection (OR, 3.43; p = 0.0007) (Table 6).
The intervention associated with the lowest rate of complications was subumbilical abdominoplasty (OR, 0.29; p < 0.0001), with fewer hematomas in particular (OR, 0.26; p = 0.005). Inverted-T plastic surgery led to 3.5 times more cutaneous necrosis (OR, 3.48; p = 0.028). Abdominoplasty with umbilical transposition was the greatest risk factor for seromas, which remained rare for the other intervention types. Hematomas were also significantly more frequent for abdominoplasties with umbilical transposition compared with other interventions (p = 0.009). In multivariable analysis, liposuction led to fewer complications (OR, 0.24; p = 0.0007), fewer hematomas (OR, 0.19; p = 0.0002), and fewer seromas (OR, 0.09; p = 0.019). As no pulmonary embolisms or deaths occurred in our cohort, and there were only two cases of deep vein thrombosis, a statistical study of such events could not be carried out.
The main strength of our study lies in its large sample size. The largest monocentric studies found to date in the literature have involved 500 to 600 patients, less than half the size of our study population.3 This large patient cohort allowed statistical results to be obtained despite the rarity of certain events, and increased the power of the study and improved the representativeness of our results. There is no other study in the literature with a comparable duration of 24 years, which allows for the changes in abdominoplasty practice to be studied over time. The repartition of the abdominal plastic surgery operations over time was very variable, with a peak in 1997, and lower levels in 2005 and 2014. We attribute this variation to the increase in the practice of double procedures in recent years, which were excluded here to limit bias. The last inverted-T abdominoplasty and vertical abdominoplasty that was included took place in 2007 and 2008. We sought to avoid vertical scarring as much as possible. The study by Chaouat et al., published in 2000, provides evidence for a substantial level of cutaneous necrosis with T-shaped abdominal plastic surgery operations (35.5 percent) compared with the other techniques (p < 0.001). The results of this study were the basis for procedural changes4 (Fig. 1). We show a distinct decrease in complications in general relative to our preliminary study [e.g., seroma (2.7 percent versus 10.90 percent), infections (4.5 percent versus 7 percent), cutaneous necrosis (2.7 versus 6.6 percent), and dehiscence (1.3 percent versus 2.3 percent)]. Only hematomas occurred more frequently (i.e., 5.7 percent versus 1.20 percent).
The main shortcoming of our study is linked to its retrospective design. Some data (such as operative time and body mass index) were missing. However, relative bias was limited by the important size of the study cohort. The study took place in a university hospital, so we chose to focus on the procedures carried out by four senior surgeons. Although they used well-standardized techniques, this may have caused some variation.
Our female population was comparable to those found in the literature, albeit it had fewer smokers and a longer follow-up.5–7 We excluded men, to limit bias, as this type of surgery is performed on female patients in the majority of instances, to enhance the homogeneity of the population, and to allow for easier extrapolation of the results. We conducted a multivariable analysis to limit the bias induced by the different operative techniques.
The rate of abdominoplasty complications varies a lot in the literature, because this is based on how the authors define complications. Depending on the studies, complications have been reported to occur in 22 to 66 percent of cases.8–13 We explored several ways of minimizing postoperative complications. As is recommended in many studies, there was a very strict preselection of patients (e.g., individuals who were nonsmokers8 , 14–17 and not obese7 , 9 , 12 , 16 , 18 , 19). We were able to show that being aged 40 years and older led to twice as many complications and up to eight times more seromas. Because abdominoplasty is often performed with women who have been affected by postpartum weight gain, it is therefore not surprising that our population had an average age of 41 years.
Another risk factor was a history of bariatric surgery. Our univariate analysis provided evidence for a statistically significant risk of more substantial complications, with double the number of complications for this subgroup. Lievain et al. carried out a review of 238 patients18 and showed that the surgery time was 11 minutes longer for post–bariatric surgery patients (p = 0.037). Although this difference seems minor in absolute terms, it was significantly associated with a higher rate of infections, dehiscence, and lymphorrhea. Manassa et al. also found a higher level of complications among patients who had lost weight because of bariatric surgery as compared to patients who had lost weight by dietary means only, and observed a much lower rate of complications when a stable weight was maintained at least 3 months before surgery.15 In contrast, and contrary to what has been previously reported in the literature,20 scars did not entail more problems with regard to wound healing. The delay between the first abdominal intervention and the abdominoplasty could come into play, with the onset of neoangiogenesis.19 , 21
This study shows a successful management of seroma rate, which is among the lowest reported to date: 2.37 percent in our study, compared with previously reported ranges from 1022 to 15 percent.13 , 14 Our use of a Delbet sheet for drainage is not common and is the key to this success (Figs. 2 and 3). This system allows for controlled wound healing of the opening and ensures prolonged passive drainage up to 2 weeks. In case of early healing, the scar can be reopened using a cotton wick. Therefore, there is no need for needle puncture. This process is painless for the patient, and leaves no anesthetic scar.
We also managed to nearly eliminate the risk of thromboembolic complications. Initially, the first injection of anticoagulant was performed on postoperative day 1. However, in light of the greater danger represented by venous thrombosis compared to the risk of bleeding in the surgical area, the protocol was later modified to an injection on the operative day. This could be the basis for the difference in the rate of hematomas seen in comparison with our preliminary study. Regardless, the 0.2 percent incidence of deep vein thrombosis and the 0 percent incidence of pulmonary embolism support this decision. The reported rates in the literature are 0.8 percent for deep vein thrombosis and 0.02 percent for pulmonary embolism.1 , 7 Furthermore, no retention girdle was prescribed for the patients if no diastasis cure was performed. In contrast, perioperative and postoperative wearing of support stockings was systematic. We believe that the purported benefit of a girdle does not appear to be warranted based on the available data. Rather, one can imagine that a girdle may impair venous return as a result of abdominal compression of the inferior vena cava. Some authors recommend performing diastasis treatment in a quasi-systematic manner to improve the aesthetic outcome. No surgical procedure is anodyne, and diastasis treatment can represent a reduction of venous return by abdominal pressure, and it also increases the length of the operation. Thromboembolic complications were successfully avoided in this cohort by strict preventative measures for all patients, namely the continuous wearing of support stockings, preventative anticoagulation from the operative day for every patient disregarding any risk factor, reduction of the surgical duration by a tried and tested surgical technique, absence of a girdle, and early ambulation.
The rate of necrosis and disunion for our cohort was lower than what has been reported in the literature. It is often 5 percent12 , 22 and can reach 10 percent,12 versus 2.7 percent for our cohort. Our rate of infections was low as well: 7 to 8 percent of cases in the literature.4 , 8 , 23 As the guidelines for aseptic surgery are relatively universal, the two main factors that can be controlled by the surgeon are systematic administration of an antibiotic prophylactic on induction of the anesthesia and a reduction in the length of the surgery. Some publications have shown a link with an increase in complications once the length of the surgery exceeds 104 minutes.18
Because of the substantial size of our sample, our data allowed for the analysis of a multitude of epidemiologic and statistical data. We were able to highlight three main risk factors for abdominoplasty complications: age older than 40 years, obesity, and smoking. A degree of prudence is needed when performing abdominoplasty on these patients. Prior abdominal surgery, which is often considered to be a major risk for cutaneous necrosis, did not appear to have a particular impact. The use of corrugated Delbet sheet drains led to the low seroma rates reported in this study. We recommend their use for controlled wound healing and the successful management of seromas following abdominoplasty. Our strict adherence to thromboembolic prevention measures led to very low rates of thromboembolic complications (two cases in 24 years); thus, we recommend a prevention strategy of early prophylaxis starting on the operative day for every patient. Complications can also be reduced by optimized surgical techniques to minimize operative time. Liposuction can lead to a distinct improvement in the aesthetic outcomes without it being associated with more complications. A second surgical intervention is often required for the final outcome to be fully aesthetically satisfactory. A detailed study of surgical reinterventions could lead to an improvement of the initial technique for a better outcome in a single procedure.
The authors thank Lucie Biard, M.D., and Matthieu Resche-Rigon, M.D., Ph.D., who carried out the statistical analysis of this study.
2. Matarasso A, Swift RW, Rankin MAbdominoplasty and abdominal contour surgery: A national plastic surgery survey. Plast Reconstr Surg. 2006;117:1797–1808.
3. Swanson EProspective clinical study of 551 cases of liposuction and abdominoplasty performed individually and in combination. Plast Reconstr Surg Glob Open 2013;1:e32.
4. Chaouat M, Levan P, Lalanne B, Buisson T, Nicolau P, Mimoun MAbdominal dermolipectomies: Early postoperative complications and long-term unfavorable results. Plast Reconstr Surg. 2000;106:1614–1618; discussion 16191623.
5. Swanson EProspective outcome study of 360 patients treated with liposuction, lipoabdominoplasty, and abdominoplasty. Plast Reconstr Surg. 2012;129:965–978.
6. Greco JA III, Castaldo ET, Nanney LB, et alThe effect of weight loss surgery and body mass index on wound complications after abdominal contouring operations. Ann Plast Surg. 2008;61:235–242.
7. Winocour J, Gupta V, Ramirez JAbdominoplasty: Risk factors, complication rates, and safety of combined procedures. Plast Reconstr Surg. 2015;136:597e–606e.
8. Hensel JM, Lehman JA Jr, Tantri MP, Parker MG, Wagner DS, Topham NSAn outcomes analysis and satisfaction survey of 199 consecutive abdominoplasties. Ann Plast Surg. 2001;46:357–363.
9. van der Beek ES, van der Molen AM, van Ramshorst BComplications after body contouring surgery in post-bariatric patients: The importance of a stable weight close to normal. Obes Facts 2011;4:61–66.
10. Neaman KC, Hansen JEAnalysis of complications from abdominoplasty: A review of 206 cases at a university hospital. Ann Plast Surg. 2007;58:292–298.
11. Espinosa-de-los-Monteros A, Torre J, Rosenberg LAbdominoplasty with total abdominal liposuction for patients with massive weight loss. Aesthetic Plast Surg. 2006; 30:42–46.
12. Stewart KJ, Stewart DA, Coghlan B, Harrison DH, Jones BM, Waterhouse NComplications of 278 consecutive abdominoplasties. J Plast Reconstr Aesthet Surg. 2006;59:1152–1155.
13. Neaman KC, Armstrong SD, Baca MEOutcomes of traditional cosmetic abdominoplasty in a community setting: A retrospective analysis of 1008 patients. Plast Reconstr Surg. 2013;131:403e–410e.
14. Mosely LH, Finseth FCigarette smoking: Impairment of digital blood flow and wound healing in the hand. Hand 1977;9:97–101.
15. Manassa EH, Hertl CH, Olbrisch RRWound healing problems in smokers and nonsmokers after 132 abdominoplasties. Plast Reconstr Surg. 2003;111:2082–2087; discussion 20882089.
16. van Uchelen JH, Werker PM, Kon MComplications of abdominoplasty in 86 patients. Plast Reconstr Surg. 2001;107:1869–1873.
17. Araco A, Gravante G, Sorge RWound infections in aesthetic abdominoplasties: The role of smoking. Plast Reconstr Surg. 2008;121:305e–310e.
18. Lievain L, Aktouf A, Auquit-Auckbur IAbdominoplasty complications: Particularities of the post-bariatric patients within a 238 patients series (in French). Ann Chir Plast Esthet. 2015;60:26–34.
19. Momeni A, Heier M, Bannasch H, Stark GBComplications in abdominoplasty: A risk factor analysis. J Plast Reconstr Aesthet Surg. 2009;62:1250–1254.
20. Karthikesalingam A, Kitcat M, Malata CMAbdominoplasty in patients with and without pre-existing scars: A retrospective comparison. J Plast Reconstr Aesthet Surg. 2011;64:369–374.
21. El-Khatib HA, Bener AAbdominal dermolipectomy in an abdomen with pre-existing scars: A different concept. Plast Reconstr Surg. 2004;114:992–997.
22. Matarasso AAwareness and avoidance of abdominoplasty complications. Aesthet Surg J. 1997;17:256, 258–261.
©2018American Society of Plastic Surgeons
23. Casaer B, Tan EK, Depoorter MThe role of antibiotic prophylaxis in abdominoplasty: A review of the infection rate in 300 cases treated without prophylaxis. Plast Reconstr Surg. 2009;123:42e.