Bikinis, Tissue Protectors, and the Standard Longitudinal Incision: Comparable Wound Complications and Safety in Direct Anterior Approach Total Hip Arthroplasty Authors: William K Crockatt, Riley Sevensky, Mark Ren, Nana O Sarpong, Alexander L Neuwirth, H John Cooper, Carl L Herndon Published: 2026 ABSTRACT AbstractBackgroundWound complications in direct anterior approach (DAA) total hip arthroplasty (THA) may differ by incision type. Bikini incisions may improve cosmesis and healing, minimizing panniculus effects while single-use soft tissue protectors may prevent damage from surgical instruments. This study compared wound complication rates in DAA THA with longitudinal incisions with and without a tissue protector, and bikini incisions.MethodsWe retrospectively reviewed 925 primary DAA THAs performed via standard longitudinal incision (n = 444, 48.0%), bikini incision (n = 375, 40.5%), and longitudinal incision with a tissue protector (n = 106, 11.5%). Wound data included complication type, depth, and treatment, including return to the operating room (RTOR). Analyses included pairwise Fisher exact testing and multivariable regression, controlling for age, sex, body mass index, smoking, surgical indication, American Society of Anesthesiologists, operative time, and performing surgeon.ResultsThere were 37 (4.0%) wound complications—34 (3.7%) superficial wound issues (eg, dehiscence without deep infection) and 3 (0.3%) deep infections. Seven (0.8%) patients required RTOR, with no significant difference between cohorts. Compared to the standard longitudinal incision (6.5%), both the bikini incision (2.4%) and the tissue protector (0.0%) cohorts had fewer overall wound complications. When controlling for the performing surgeon, the bikini incision performed similarly to the standard longitudinal approach (adjusted Odds Ratio: 0.49; P = .23).ConclusionsDAA THA demonstrates an excellent safety profile, with a low overall incidence of wound complications. While initial reductions in superficial complications with bikini incisions and tissue protectors were influenced by surgeon-specific factors, all techniques proved to be highly safe and effective, with low RTOR rates.Keywords: Direct anterior approach, Total hip arthroplasty, Wound complications, Incision type, Tissue protector/retractor, Bikini incision/approach Introduction Total hip arthroplasty (THA) via the direct anterior approach (DAA) has become increasingly popular due to its potential for reduced postoperative pain, faster recovery, and lower dislocation rates compared to posterior and lateral approaches [1,2]. However, despite its advantages, wound complications remain a common concern following DAA THA [3]. Several factors may influence the likelihood of complications, including patient demographics, comorbidities, and the type of incision used. Incision choice may affect wound healing, cosmesis, and the risk of infection, making it an important consideration in DAA THA procedures. The traditional approach to DAA involves a standard longitudinal incision, which is typically made over the muscle belly of the tensor fascia lata just distal and lateral to the anterior superior iliac spine (Fig. 1a). This incision remains the standard utility approach for the DAA due to its extensile nature, allowing for proximal or distal extension should intraoperative complications arise. However, because it runs perpendicular to the natural skin tension lines (Langer’s lines), the longitudinal incision may be associated with visible scarring and concerns regarding wound dehiscence, although recent reviews have found conflicting evidence regarding complication rates [[3], [4], [5]]. To address some of these concerns, the oblique, bikini incision offers an alternative [6]. The bikini incision is performed parallel to the inguinal crease, following skin tension lines (Fig. 1b), which may not only improve cosmesis but also reduce or avoid the detrimental effects of panniculus and inguinal crease on incisional healing, potentially leading to better wound healing in obese patients [[7], [8], [9]]. However, it is nonextensile and has been associated with a potentially higher risk of lateral femoral cutaneous nerve neurapraxia [5,7,10]. (a) Standard longitudinal incision—approximately 8 cm in length, 2 fingerbreadths distal, and lateral to the anterior superior iliac spine (ASIS) aiming toward the fibular head; (b) Transverse “bikini” incision—parallel to skin tension lines in the flexion crease, with two-thirds of the incision length lateral to the ASIS and one-third medial. Soft tissue protectors, such as single-use retractors (Fig. 2), have been employed in various surgical specialties to minimize soft tissue trauma and the risk of wound complications [11,12]. These devices are designed to shield the skin from mechanical injury caused by surgical instruments, thereby preventing damage that can lead to wound breakdown. In orthopaedic surgery, the use of soft tissue protectors has shown promise, particularly in minimizing infection rates and improving wound healing in hip and knee arthroplasty [[13], [14]]. However, evidence is still limited regarding their efficacy in these cases, and further evaluation is needed. Single-use soft tissue protector. Typically placed deep to the tensor fascia lata muscle following superficial dissection with a longitudinal incision to provide additional exposure and protect soft tissues from abrasive instruments and debris. This study aimed to compare the incidence of postoperative wound complications following THA with DAA using 3 different variations in the technique: the standard longitudinal incision, the bikini incision, and the longitudinal incision with the use of a soft tissue protector. We hypothesized that both the bikini incision and the soft tissue protector would be associated with a lower incidence of superficial wound complications compared to the standard longitudinal incision without increasing the rate of major complications requiring reoperation. Methods Patient selection This was a retrospective cohort analysis of all primary THAs performed through a DAA by 4 fellowship-trained arthroplasty surgeons. A total of 1020 surgeries were consecutively performed between September 2022 and September 2024. Exclusion criteria included any patients with operative notes that did not specify the incision (n = 66) or dressing (n = 7) utilized, conversion THA (n = 13), simultaneous bilateral THA (n = 4), or patients with prior infection of the hip joint (n = 1). Two patients who had a “hockey-stick” style incision and another 2 individuals who died prior to discharge from their admission were also excluded; thus 925 hips were included in this analysis. Patients were classified into 3 cohorts based on surgical technique—444 (48.0%) were performed through a standard longitudinal incision, 375 (40.5%) through a bikini incision, and 106 (11.5%) through a longitudinal incision with a soft tissue protector (Alexis Orthopedic Protector, Applied Medical, Rancho Santa Margarita, CA). For the soft tissue protector cohort, the device was placed deep to the tensor fascia lata muscle following dissection and coagulation of the lateral femoral circumflex vessels; it may be utilized with either a bikini or standard longitudinal incision, however was only used with standard longitudinal in this study. Incision choice and whether to use a soft tissue protector was left to surgeon discretion. All wounds were closed in layered fashion using monofilament barbed suture for the capsule, deep layers, and monofilament subcuticular for the skin layer. Dressing choice was also left to surgeon discretion in this retrospective cohort. Demographic data included age at the time of surgery, body mass index, sex, medical comorbidities (including presence of diabetes, chronic kidney disease, liver disease, preoperative anemia, smoking status, use of immunosuppressive agents [eg, steroids, disease-modifying antirheumatic drugs [DMARDs], antirejection medications for organ transplant, etc.]). Surgical information included indication for surgery, laterality, type of incision/use of tissue protector, and operative time. Specific operative techniques, including incision type, use of a tissue protector, implant selection, type of dressing, and postoperative anticoagulation choice were nonrandomized and left to the discretion of the surgeon. The demographic and baseline clinical characteristics of the 3 cohorts are detailed in Table 1, and a surgeon-specific breakdown is detailed in Table 2. Table 1Patient demographics.VariableStandard longitudinal (n = 444)Longitudinal + tissue protector (n = 106)Bikini (n = 375)P valueSex<.001 Men2454788 Women19959287Age (mean, y)64.967.969.4<.001Body mass index (mean, kg/m2)27.628.37.7.51ASA score (mean)2.42.42.5.14Comorbidities Diabetes652052.003 Chronic kidney disease481238.92 Preop anemia592028.019Smoking.31 Former14243135 Current3057Liver disease1044.17DMARDs/corticosteroid use36927.85Preop antiplatelet/coagulant use1453097.080Indication for surgery<.001 Osteoarthritis36476334 Inflammatory arthritis002 Fracture252611 AVN45322 Dysplasia806 Other210Operative time (mean, min)76.692.767.4<.001Dressing type Self-adhering mesh + skin adhesive26075217 NPWT alone271920 Ag-antibacterial dressing alone7812137 NPWT + skin adhesive3001 Ag-antibacterial dressing + skin adhesive4900Average follow-up (mean, d)174.2146.1167NAThe bolded values are those results that are statistically significant.ASA, American Society of Anesthesiologists; AVN, avascular necrosis; NPWT, negative pressure wound therapy. Patient demographics. The bolded values are those results that are statistically significant. ASA, American Society of Anesthesiologists; AVN, avascular necrosis; NPWT, negative pressure wound therapy. Table 2Surgeon-specific breakdown.VariableSurgeon ASurgeon BSurgeon CSurgeon DTotal cases (n)537106168114Incision preference (n)Bikini (375)standard (162)Tissue protector (106)Standard (168)Standard (114)Age (mean)68.268.162.361Sex Male220476944 Female317609970BMI (mean, kg/m2)27.928.228.425.5Mean ASA score2.42.52.42.4Smoking statusFormer - 182Current - 27Former - 43Current - 5Former - 58Current - 11Former - 37Current - 9Indication for surgery Primary osteoarthritis473 (88.1%)77 (72.0%)139 (82.7%)86 (75.4%) Inflammatory arthritis2 (0.4%)0 (0.0%)0 (0.0%)0 (0.0%) Fracture19 (3.5%)26 (24.3%)6 (3.6%)11 (9.6%) Dysplasia9 (1.7%)0 (0.0%)5 (3.0%)0 (0.0%) AVN33 (6.1%)3 (2.8%)18 (10.7%)16 (14.0%)Comorbidities Diabetes mellitus61 (11.4%)20 (18.7%)18 (10.7%)11 (9.6%) Chronic kidney disease (any)53 (9.9%)12 (11.2%)21 (12.5%)12 (10.5%) Liver disease6 (1.1%)4 (3.7%)7 (4.2%)1 (0.9%) Preoperative anemia43 (8.0%)20 (18.7%)23 (13.7%)21 (18.4%) DMARDs/corticosteroid use37 (6.9%)9 (8.4%)13 (7.7%)13 (11.4%) Preop antiplatelet/coagulant use153 (28.5%)30 (28.0%)57 (33.9%)32 (28.1%)Operative time (mean, min)69.392.772.985.9AVN, avascular necrosis, BMI, body mass index; ASA, American Society of Anesthesiologists. Surgeon-specific breakdown. AVN, avascular necrosis, BMI, body mass index; ASA, American Society of Anesthesiologists. Primary and secondary outcomes The primary outcome was presence or absence of a postoperative wound complication within 90 days. Secondary endpoints included type of complication (eg, cellulitis, wound dehiscence, drainage, infection, etc.), depth of soft tissue involvement (superficial vs deep), and treatment required (eg, none, dressing changes alone, dressing changes plus oral antibiotics, negative pressure wound therapy plus oral antibiotics, image-guided aspiration, return to the operating room [RTOR] for debridement or component exchange, etc.). Data analyses Statistical analyses included pairwise Fisher exact testing and multivariable regression, controlling for the performing surgeon and known factors associated with increased postoperative complications including age, sex, body mass index, smoking history, surgical indication, American Society of Anesthesiologists classification, and operative time. Variables were also selected for inclusion in the multivariate model if they demonstrated a P value of <0.20 in the initial univariate analysis. A time-to-event survival analysis was beyond the scope of this study. Statistical analyses were conducted using R Statistical Software (R Core Team, Vienna, Austria) [15]. A post hoc power analysis was calculated based on the incidence rate of wound complications in each of the 3 cohorts and concluded a 93.5% power comparing tissue protector vs standard longitudinal incision for an alpha of 0.05, and 80.5% power comparing bikini to standard longitudinal incisions. Our study was underpowered to compare tissue protector to bikini incision at 25.5% power. Results A total of 37 wound complications (4.0%) occurred within the 90-day postoperative period. Of these, 34 (3.7%) were superficial, limited to the skin and supra-fascial layers, while 3 (0.3%) involved deep infection. Seven patients (0.8%) required RTOR: 4 underwent debridement and revision closure, while the 3 with deep infections required revision/exchange of prosthetic components. A detailed breakdown of the complication profile of each cohort is detailed in Table 3. Table 3Complication profile.Complication/treatmentStandard longitudinal (n = 444)Longitudinal + tissue protector (n = 106)Bikini (n = 375)Depth2 Superficial2806 Deep102Wound complication Dehiscence/slow healing2104 Hematoma/seroma201 Cellulitis/superficial SSI501 Deep infection102Treatment required None500 Dressing changes + abx1502 NPWT + abx401 Aspiration102 RTOR403Other complications Neuropraxia325 Hematoma305Abx, antibiotics; NPWT, negative pressure wound therapy; SSI, surgical site infection. Complication profile. Abx, antibiotics; NPWT, negative pressure wound therapy; SSI, surgical site infection. The results of the univariate and multivariate logistic regression analyses are presented in Tables 4 and 5, respectively. The univariate analysis identified several patient and surgical factors with a preliminary, unadjusted association with an increased risk of wound complications. These included liver disease (P = .013), use of DMARDs or corticosteroids (P = .0025), use of “full” anticoagulant medications (eg, low-molecular-weight heparin, novel/direct oral equivalents, etc.) (P = .0005), inflammatory arthritis as a surgical indication (P = .022), and longer operative time (P = .043). On multivariate analysis, only the use of DMARDs or corticosteroids (adjusted Odds Ratio [aOR] 3.48, P = .01) and “full” anticoagulation remained significant (aOR: 4.75, P = .002). Table 4Univariate analysis.VariableOR95% CIP valueMale sex0.43−0.23 to 1.09.20Age0.0004−0.027 to 0.028.98Body mass index−0.001−0.062 to 0.060.97ASA score0.51−0.086 to 1.10.094Comorbidities Diabetes0.29−0.34 to 0.92.37 Chronic kidney disease0.51−0.38 to 1.42.26 Preop anemiaNA−0.87 to 1.09.82 Smoking0.22−0.29 to 0.74.40 Liver disease1.640.35-2.92.013 DMARDs/corticosteroid use1.270.44-2.09.0025Preop antiplatelet/coagulant use (vs none) Aspirin0.004−0.92 to 0.93.99 ClopidogrelNANAn.s.a DAPT1.32−0.80 to 3.4.22 Full AC1.470.65-2.30.0005Indication for surgery (vs primary osteoarthritis) Inflammatory arthritis3.280.49-6.08.022 Fracture0.61−1.47 to 1.69.27 AVN0.48−0.60 to 1.56.38 DysplasiaNANAn.s.a OtherNANAn.s.aOperative time0.0120.0003-0.023.043Incision (vs. standard longitudinal) Longitudinal + protectorNANA.0023a Bikini−1.16−1.96 to −0.037.0041Dressing type (vs mesh + adhesive) NPWT alone0.98−0.17 to 2.13.094 Ag-antibacterial dressing alone0.65−0.19 to 1.49.13 NPWT + skin adhesive1.810.63-3.0.0027 Ag-antibacterial dressing + skin adhesive1.760.74-2.77.0007Average follow-up (mean, d)N/AThe bolded values are those results that are statistically significant.AC, anticoagulation; ASA, American Society of Anesthesiologists; AVN, avascular necrosis; DAPT, dual antiplatelet therapy; NPWT, negative pressure wound therapy.aNo wound complications in group. Pairwise Fisher’s exact test with Bonferroni correction reported. Univariate analysis. The bolded values are those results that are statistically significant. AC, anticoagulation; ASA, American Society of Anesthesiologists; AVN, avascular necrosis; DAPT, dual antiplatelet therapy; NPWT, negative pressure wound therapy. No wound complications in group. Pairwise Fisher’s exact test with Bonferroni correction reported. Table 5Multivariate analysis.VariableaOR95% CIP valueMale sex1.480.69-3.17.32Body mass index1.030.96-1.11.42ASA score1.000.48-2.06.99Comorbidities Diabetes1.010.50-2.05.97 Smoking1.120.63-2.00.70 Liver disease4.080.96-17.4.06 DMARDs/corticosteroid use3.481.3-9.3.01Preop antiplatelet/coagulant use (vs none) Aspirin0.770.28-2.07.60 ClopidogrelNANANA DAPT1.80.16-20.27.63 Full AC4.751.76-12.81.002Indication for surgery (vs primary osteoarthritis) Inflammatory arthritis23.670.58-972.63.10 Fracture2.010.55-7.35.29 AVN1.050.32-3.52.93 DysplasiaNANANA OtherNANANA Operative time1.0040.990-1.02.60Incision (vs. standard longitudinal) Longitudinal + protectorNANANA Bikini0.490.15-1.57.23AC, anticoagulation; ASA, American Society of Anesthesiologists; AVN, avascular necrosis; DAPT, dual antiplatelet therapy. Multivariate analysis. AC, anticoagulation; ASA, American Society of Anesthesiologists; AVN, avascular necrosis; DAPT, dual antiplatelet therapy. When comparing the groups, the standard longitudinal incision had a wound complication rate of 6.5%. In contrast, the bikini incision group had a lower complication rate of 2.4%, and the group using a tissue protector had a complication rate of 0.0%. However, when incorporating the performing surgeon into the multivariate model, the bikini incision demonstrated comparable protections against wound complications relative to the standard longitudinal incision (aOR: 0.49, P = .23). Due to the zero-event rate in the protector group, an odds ratio could not be calculated, but its significance vs the standard group was established by the corrected Fisher's exact test (P = .0023). While the overall complication rates differed, the rate of RTOR was low across all cohorts, with no difference observed between the longitudinal, bikini, and tissue protector groups. Discussion This study demonstrates the overarching safety of multiple incision techniques for DAA THA. While the overall wound complication rate in this patient population was exceptionally low, univariate analysis initially suggested the bikini incision and tissue protectors were associated with lower complication rates. However, when controlling for the performing surgeon in multivariate analysis, the incision types performed comparably. This highlights that excellent wound healing outcomes are likely driven by individual surgeon technique, experience, and case volume rather than the incision type itself. Previous literature has shown that the bikini incision not only improves cosmesis but also reduces wound complications in THA procedures, particularly in patients with increased abdominal panniculus [4,5,[7], [8], [9]]. While the longitudinal incision provides excellent exposure, it may be more prone to wound healing complications, particularly in patients with obesity or increased panniculus size. The oblique bikini incision, in contrast, is placed parallel to the natural skin tension lines (Langer's lines), which minimizes tension on the wound and improves the potential for more reliable wound healing [5,9]. The use of soft tissue protectors has been demonstrated to reduce the incidence of surgical site infections and wound breakdown in other surgical specialties, such as gastrointestinal and urologic surgery [11,12]. Our results are in line with other studies where such devices have been associated with reduced tissue damage, lower rates of infection, and less hypertrophic scarring following hip and knee arthroplasties [13,14]. However, while our overall study cohort is large, the group with the soft tissue protector was considerably smaller, which, as noted in our post hoc power analysis, limited the power to detect a significant difference when comparing this group directly to the bikini incision cohort. Additionally, we observed a longer mean operative time in the tissue protector cohort which may reflect a learning curve associated with the initial implementation of the device, the additional time required for its setup and insertion, or potentially the selection of more complex cases for this specific technique during the study period. It is critical to note that while superficial wound healing issues were less frequent in the study cohorts, the rate of RTOR—a proxy for deep infection and major complications—did not statistically differ between groups. This suggests that while these techniques may optimize superficial soft tissue healing, the risk of deep periprosthetic infection remains low and comparable across all incision types in experienced hands. The present study highlights the need for additional high-quality studies to determine whether single-use soft tissue protectors can reduce the risk of wound complications in THA in conjunction with the DAA approach. Despite the promising results, this study has several limitations. First, the retrospective nature of the analysis introduces the potential for significant selection bias. As noted in our methods, the choice of surgical technique was nonrandomized and based on surgeon preference, which for some surgeons evolved during the data collection period. While we successfully controlled for the performing surgeon in our updated multivariate model, the strong multicollinearity between specific surgeons and their preferred techniques makes it difficult to completely isolate the incision’s independent effect from the surgeon’s overall skill, experience, and perioperative protocols. Similarly, the choice of postoperative dressing was also determined by surgeon preference and evolved over time. Due to this high variability and strong multicollinearity, dressing type was also excluded from the final regression model. Second, this study was conducted at a single, high-volume academic institution with experienced surgeons. Therefore, the excellent results observed may not be immediately generalizable to surgeons with less experience or those still on the learning curve for these specific DAA techniques. Future multicenter studies would be needed to corroborate these findings across different practice settings. Third, the 90-day follow-up period, while appropriate for assessing acute wound complications, is not long enough to evaluate longer-term outcomes such as the incidence of late infection or final scar maturation, which can take 6 months or longer. Finally, while we did attempt to collect and control for preoperative nutritional markers including albumin and prealbumin, the large amount of missing data prevented the inclusion of these variables in our final analysis, which could act as potential confounders. Future prospective studies with larger sample sizes and longer-term follow-up are necessary to confirm these results and further evaluate the outcomes associated with these techniques. Conclusions DAA THA demonstrates an excellent safety profile with a low overall incidence of wound complications. While initial reductions in superficial complications with bikini incisions and tissue protectors were influenced by surgeon-specific factors, all techniques proved to be highly safe and effective, with low RTOR rates. CRediT authorship contribution statement William K. Crockatt: Writing – review & editing, Writing – original draft, Visualization, Supervision, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Riley Sevensky: Writing – original draft, Methodology, Investigation, Data curation. Mark Ren: Writing – original draft, Formal analysis, Data curation. Nana O. Sarpong: Writing – review & editing, Supervision, Project administration, Conceptualization. Alexander L. Neuwirth: Writing – review & editing, Supervision, Project administration, Conceptualization. H. John Cooper: Writing – review & editing, Supervision, Project administration, Methodology. Carl L. Herndon: Writing – review & editing, Writing – original draft, Supervision, Project administration, Methodology, Conceptualization. Conflict of interest A.L. Neuwirth received research support from Smith & Nephew. C.L. Herndon is a paid consultant for DePuy and KCI; has stock or stock options in Polaris; and is a member of the medical/orthopaedic publications editorial/governing board at Journal of Arthroplasty. H.J. Cooper is on the speakers bureau/paid presentations for Solventum; is a paid consultant for Solventum and Zimmer Biomet; has stock or stock options in PolarisAR; received research support from Smith & Nephew; and is a board member/committee appointments for AAHKS. N.O. Sarpong is a paid consultant for Zimmer Biomet and has stock or stock options in Polaris. W.K. Crockatt is a board member/committee appointments for AAHKS; all other authors declare no potential conflicts of interest. For full disclosure statements refer to https://doi.org/10.1016/j.artd.2026.102037. Supplementary data Conflict of Interest Statement for Neuwirth Conflict of Interest Statement for Herndon Conflict of Interest Statement for Cooper Conflict of Interest Statement for Ren Conflict of Interest Statement for Sarpong Conflict of Interest Statement for Sevensky Conflict of Interest Statement for Crockatt References Associated Data Supplementary Materials Conflict of Interest Statement for Neuwirth Conflict of Interest Statement for Herndon Conflict of Interest Statement for Cooper Conflict of Interest Statement for Ren Conflict of Interest Statement for Sarpong Conflict of Interest Statement for Sevensky Conflict of Interest Statement for Crockatt