European Journal of Cardiovascular Medicine

Total knee arthroplasty (TKA) reliably relieves pain and restores function in patients with end-stage knee osteoarthritis, but early wound problems and postoperative infections remain among its most serious complications. Contemporary studies report the incidence of surgical-site infection (SSI) following TKA to range between 0.3% and 2.6%, while periprosthetic joint infection (PJI) occurs in approximately 1–2% of cases and represents a leading cause of revision arthroplasty worldwide [1]. These complications are associated with substantial morbidity, prolonged hospital stay, reoperations, and increased healthcare costs, making prevention and early risk stratification an essential clinical priority [2].

Delayed wound healing (DWH) manifesting as prolonged wound drainage, edge separation, or delayed epithelialization represents a critical early indicator that may precede deep infection [3]. Although definitions vary among studies, DWH is typically characterized by healing beyond 10–14 days, persistent exudation beyond 72 hours, or clinically visible gapping of wound margins [4]. Persistent drainage and wound-edge separation are clinically significant because they have been consistently implicated as precursors to periprosthetic infection and poor functional outcomes [5].

A variety of interacting risk factors contribute to DWH and postoperative infection following TKA. Patient-related factors include diabetes mellitus and inadequate perioperative glycaemic control, obesity and increased subcutaneous fat thickness, and malnutrition, often indicated by hypoalbuminaemia, which has been repeatedly linked to both SSI and PJI after total joint arthroplasty (TJA) [6]. Recent literature highlights the predictive value of preoperative albumin levels in forecasting wound complications and infection risk after primary TKA and total hip arthroplasty (THA) [7].

Procedure-related factors also play a major role. Longer operative time has been identified as an independent risk factor for postoperative infection after TKA. This is biologically plausible due to extended tissue exposure, increased soft-tissue trauma, and longer anaesthetic duration [8]. Additionally, bilateral or revision surgeries, excessive intraoperative bleeding, and blood transfusion have been shown to increase the risk of postoperative complications [9].

Postoperative factors such as persistent wound drainage, improper dressing, or inadequate wound protection can also precipitate delayed healing and infection. The introduction of negative-pressure wound therapy (NPWT) dressings has demonstrated benefits in reducing early wound complications, particularly in high-risk populations such as obese or diabetic patients [10].

Despite an expanding body of literature addressing wound complications and infection after arthroplasty, several gaps remain. First, most studies assess DWH and infection separately, despite their overlapping and sequential pathophysiology. Integrated analyses evaluating predictors for both outcomes within the same cohort are limited [11]. Second, variations in the definition of DWH across studies compromise comparability and generalizability. Third, although advanced predictive tools such as nomograms and machine-learning algorithms have been proposed for PJI risk assessment, few incorporate basic, easily obtainable parameters such as serum albumin, HbA1c, BMI, and operative time into practical models particularly in South Asian populations, where nutritional status, diabetes prevalence, and microbial flora may differ [12].

Aim of the Study of the study was to evaluate preoperative, intraoperative, and early postoperative predictive factors for delayed wound healing and postoperative infections in patients undergoing primary total knee arthroplasty, and to develop a clinically applicable risk-stratification model integrating patient characteristics (including nutritional and glycaemic markers), operative parameters (including surgical duration), and immediate postoperative wound status.

Study Design and Setting

This prospective observational study was conducted in the Department of Orthopedics, Mamata Academy of Medical Sciences, Hyderabad, between January 2023 to June 2024. The objective was to evaluate predictive factors for delayed wound healing (DWH) and postoperative infections among patients undergoing total knee arthroplasty (TKA). A total of 30 patients fulfilling the eligibility criteria were enrolled consecutively after obtaining informed consent.

Study Population

Patients scheduled for primary unilateral or bilateral TKA under elective settings were screened for eligibility. All surgeries were performed by experienced orthopedic surgeons following a standardized operative protocol.

Inclusion Criteria

• Patients aged 18 years and above undergoing primary total knee arthroplasty for osteoarthritis or rheumatoid arthritis.
• Patients medically optimized and fit for elective surgery (ASA I–III).
• Patients who provided written informed consent to participate.

Exclusion Criteria

• Revision TKA or previous surgery on the same knee.
• Active infection at any site or chronic skin ulceration near the operative limb.
• Inflammatory arthropathies with active systemic disease (e.g., lupus, psoriatic arthritis).
• Immunocompromised patients (e.g., on chronic steroids or chemotherapy).
• Peripheral vascular disease or severe varicose veins affecting lower limbs.
• Poorly controlled diabetes mellitus (HbA1c > 8.0%) or uncontrolled hypertension.
• Patients refusing follow-up or with incomplete records.

Preoperative Assessment

All patients underwent a thorough clinical evaluation and routine investigations including complete blood count, random blood sugar, HbA1c (for diabetics), serum albumin, and renal function tests. Nutritional status, BMI, and comorbidities such as diabetes, hypertension, and smoking history were recorded.

Perioperative Protocol

All procedures were performed under regional anesthesia using a standard medial parapatellar approach and cemented prosthesis. Prophylactic antibiotics (first-generation cephalosporin) were administered 30–60 minutes prior to incision. Skin closure was done with subcuticular sutures or staples depending on surgeon preference. Drains and negative-pressure wound dressings were used as per intraoperative assessment.

Postoperative Care and Follow-up

Dressings were inspected daily for signs of discharge or hematoma. Physiotherapy was initiated on the first postoperative day. Patients were followed up on day 14 and at 6 weeks postoperatively.

Definition of Outcomes

• Delayed Wound Healing (DWH): wound drainage >72 hours, edge separation, or incomplete epithelialization by 14 days.
• Postoperative Infection: presence of purulent discharge, redness, warmth, fever, or need for antibiotics or surgical debridement within 30 days post-surgery.

Data Collection

Demographic details, comorbidities, laboratory parameters, intraoperative details (duration, blood loss, drain use), and postoperative outcomes were recorded in a structured proforma.

Statistical Analysis

Data were analyzed using SPSS version 25.0. Continuous variables were expressed as mean ± SD, and categorical data as percentages. Independent t-test and chi-square tests were used for comparison. Logistic regression was performed to identify predictors for DWH and infection, with p<0.05 considered significant.

Ethical Considerations

The study was approved by the Institutional Ethics Committee of Mamata Academy of Medical Sciences, Hyderabad. Written informed consent was obtained from all participants. Confidentiality of patient data was maintained throughout the study.

Table 1. Demographic and Baseline Characteristics of Study Participants (n = 30)

The demographic data of the 30 patients included in the study show that the mean age was 66.2 ± 7.8 years, indicating that most participants were elderly, consistent with the typical age profile of patients undergoing total knee arthroplasty (TKA). A majority were female (60%), reflecting the higher prevalence of advanced knee osteoarthritis among women. The mean body mass index (BMI) was 30.4 ± 4.6 kg/m², suggesting that many patients were overweight or obese, a known predisposing factor for wound-related complications and delayed recovery.

Among comorbidities, diabetes mellitus was present in 36.7% and hypertension in 50%, both common systemic illnesses influencing wound healing. A small proportion (10%) were active smokers. Most patients belonged to ASA class II (60%), indicating mild-to-moderate systemic disease, while 33.3% were ASA class III, suggesting higher operative risk. The mean serum albumin level was 3.9 ± 0.5 g/dL, and 20% of patients (not shown in table) had hypoalbuminaemia (<3.5 g/dL), reflecting mild nutritional compromise in part of the cohort. Among diabetic patients, the mean HbA1c was 7.3 ± 0.8%, indicating suboptimal glycaemic control.

In terms of surgical laterality, unilateral TKA was performed in the majority (86.7%), with only 13.3% undergoing bilateral procedures. Overall, the demographic profile represents a typical elderly, comorbidity-burdened arthroplasty population at moderate nutritional and metabolic risk for delayed wound healing and infection (Table 1).

Table 2. Comorbidities among Study Participants (n = 30)

Table 2 summarises the comorbid conditions observed among patients who underwent total knee arthroplasty (TKA) in this study. The most prevalent comorbidity was hypertension, affecting 50% of the cohort, followed by obesity (46.7%) and diabetes mellitus (36.7%). These conditions are frequently encountered in elderly TKA populations and contribute significantly to impaired wound healing and infection risk.

A notable proportion of patients (20%) had hypoalbuminaemia (< 3.5 g/dL), indicating mild malnutrition, which may adversely influence tissue regeneration and immune response. Anaemia was present in 16.7%, reflecting underlying nutritional or chronic disease status. Smoking was reported in 10%, known to impede microcirculation and delay wound healing.

Less common comorbidities included hypothyroidism (6.7%), coronary artery disease (6.7%), and chronic kidney disease (3.3%), all of which can prolong postoperative recovery. Two patients (6.7%) had other systemic conditions such as osteoporosis or chronic obstructive pulmonary disease (COPD).

Overall, the data reflect a multimorbid elderly population, where overlapping metabolic, nutritional, and vascular factors contribute to variable wound-healing capacity. The coexistence of diabetes, obesity, and hypoalbuminaemia underscores the importance of preoperative optimisation of metabolic and nutritional parameters to minimise delayed wound healing and postoperative infections in TKA patients.

Table 3. Laboratory Parameters of Study Participants (n = 30)

Table 3 presents the laboratory parameters of the patients included in the study. The mean hemoglobin level was 12.6 ± 1.1 g/dL, with 16.7% of patients showing mild anaemia. The mean total leukocyte count and platelet count were within normal reference ranges, indicating that none of the patients had active infection or hematological abnormalities preoperatively.

The mean random blood sugar was 132 ± 26 mg/dL, and 26.7% of the participants had elevated values. Among diabetic patients, the mean HbA1c was 7.3 ± 0.8%, and 63.6% had HbA1c levels > 6.5%, reflecting suboptimal glycaemic control. This suggests a high baseline metabolic risk profile in this cohort, which may predispose to delayed wound healing.

The mean serum albumin level was 3.9 ± 0.5 g/dL, with 20% of patients having hypoalbuminaemia (<3.5 g/dL), indicating mild malnutrition. Similarly, 13.3% had reduced total protein levels. These findings point to a moderate nutritional deficit among TKA candidates, consistent with other Indian and international reports linking hypoalbuminaemia to poor wound outcomes.

Renal parameters (serum creatinine and blood urea) were largely normal, indicating preserved renal function in most patients. The inflammatory markers CRP (4.2 ± 2.1 mg/L) and ESR (26 ± 10 mm/hr) were mildly raised in a few cases (10–16.7%), likely reflecting chronic degenerative joint disease rather than acute infection.

Overall, these results show that a significant subset of patients had nutritional and metabolic derangements (low albumin, high HbA1c), both of which are known to adversely affect wound healing and infection risk following total knee arthroplasty.

Table 4. Intraoperative Details of Study Participants (n = 30)

Table 4 summarises the intraoperative characteristics of the 30 patients who underwent total knee arthroplasty (TKA). The mean operative duration was 92.3 ± 14.6 minutes, ranging from 70 to 130 minutes, which aligns with typical durations reported for standard primary TKA procedures. The mean intraoperative blood loss was 280 ± 65 mL, reflecting effective hemostasis and standard surgical technique.

A tourniquet was used in 90% of cases, with a mean inflation time of 68.4 ± 12.2 minutes, consistent with accepted orthopedic practice to maintain a bloodless field and reduce intraoperative bleeding. Surgical drains were placed in 70% of patients, with a mean drain output of 125 ± 40 mL, and were typically removed within 24–48 hours postoperatively.

Spinal anesthesia was the preferred modality (86.7%), while a combined spinal–epidural technique was used in the remaining 13.3% of patients. Skin closure was achieved using staples in 60% and subcuticular sutures in 40% of cases, depending on the surgeon’s preference and wound characteristics. Negative-pressure wound dressings were selectively applied in 4 patients (13.3%) identified as high-risk due to obesity or diabetes, demonstrating an evidence-based approach to wound protection.

Most procedures were unilateral (86.7%), with bilateral surgeries performed in 13.3%. Intraoperative complications were minimal only two cases (6.7%) experienced minor bleeding, and no major vascular, nerve, or anesthetic complications were encountered.

Overall, these findings indicate that the intraoperative course was uneventful in the majority of cases, and surgical and anesthetic practices adhered to standard protocols, supporting the safety and reproducibility of TKA at the study centre.

Table 5. Postoperative Outcomes of Study Participants (n = 30)

Table 5 presents the postoperative outcomes of the patients who underwent total knee arthroplasty (TKA). The mean hospital stay was 6.8 ± 1.4 days, ranging from 5 to 10 days, which is comparable to reported durations in similar arthroplasty series. The average drain duration was 32.5 ± 8.3 hours, with a mean drain output of 125 ± 40 mL, indicating effective postoperative hemostasis. Early mobilization was achieved in most patients, with mean time to ambulation of 22.6 ± 6.2 hours, reflecting adherence to enhanced recovery protocols.

The mean duration of analgesia was 10.2 ± 2.8 hours, and rescue analgesia was administered when the visual analogue score (VAS) exceeded 4. In terms of complications, delayed wound healing (DWH) occurred in 4 patients (13.3%), and superficial surgical site infection (SSI) was observed in 2 patients (6.7%). Importantly, no cases of deep infection or periprosthetic joint infection (PJI) were reported during the 30-day follow-up period, indicating effective infection-control practices. Wound edge necrosis or gapping was observed in 2 patients (6.7%), and postoperative fever occurred in 3 patients (10%), all of which were mild and self-limiting. There were no thromboembolic events (DVT) or major systemic complications reported.

Functional recovery was satisfactory, with mean knee flexion at discharge of 95.6 ± 8.5°, enabling early rehabilitation. Overall, these findings suggest favorable short-term surgical outcomes, with a low incidence of infection and wound-related complications. The timely ambulation, low morbidity, and absence of deep infection reflect effective perioperative management and adherence to aseptic and rehabilitation protocols at the study center.

Table 6. Logistic Regression Analysis for Predictors of Delayed Wound Healing (DWH)

Table 6 summarises the results of logistic regression analysis for predictors of delayed wound healing (DWH) among patients undergoing total knee arthroplasty (TKA). On multivariate analysis, serum albumin <3.5 g/dL and HbA1c >7% emerged as significant independent predictors of DWH, with adjusted odds ratios (AOR) of 4.86 (95% CI: 1.06–22.2; p=0.041) and 3.40 (95% CI: 1.01–11.8; p=0.048), respectively. These findings indicate that patients with hypoalbuminaemia or poor glycaemic control were nearly four times more likely to experience delayed wound healing.

Operative time >100 minutes also showed a positive association (AOR 3.05; p=0.058), suggesting that longer surgical duration might increase the likelihood of wound complications, though it did not reach statistical significance in this small sample. Other factors, including age, gender, BMI, diabetes, and drain use, did not significantly influence wound healing outcomes in this cohort.

The results emphasise that nutritional status and metabolic control are critical modifiable determinants of postoperative wound recovery. Ensuring adequate serum albumin levels and maintaining HbA1c below 7% before surgery may substantially reduce the incidence of delayed wound healing after TKA.

Table 7. Logistic Regression Analysis for Predictors of Postoperative Infection

Table 7 shows the logistic regression analysis performed to identify predictors of postoperative infection following total knee arthroplasty (TKA). Among the variables assessed, serum albumin <3.5 g/dL and operative time >100 minutes emerged as statistically significant independent predictors. Patients with low serum albumin had a 4.2-fold higher risk of infection (AOR 4.21, 95% CI: 1.02–18.9; p = 0.047), underscoring the strong influence of nutritional status on postoperative wound outcomes. Likewise, those with prolonged operative duration had a 3.6-fold increased risk of developing infection (AOR 3.66, 95% CI: 1.00–14.0; p = 0.049).

Other factors such as diabetes mellitus, BMI ≥30 kg/m², and age >65 years showed positive but non-significant associations, likely due to the limited sample size. Use of drain was not significantly related to infection risk. These results reinforce the importance of preoperative nutritional optimization and intraoperative efficiency to minimize infection risk after TKA.

In this prospective observational study of 30 patients undergoing primary total knee arthroplasty (TKA) at the Department of Orthopedics, Mamata Academy of Medical Sciences, Hyderabad, the overall incidence of delayed wound healing (DWH) and postoperative infection was comparable to rates reported worldwide. DWH occurred in 13.3 % of cases, superficial surgical-site infection (SSI) in 6.7 %, and no deep periprosthetic joint infection (PJI) was observed within 30 days of surgery. Similar ranges have been documented by Luthra et al. (2023) and Bagheri et al. (2025), who reported DWH between 8–14 % and SSI between 3–6 % (13). These findings affirm that wound-related morbidity remains an important yet largely preventable challenge after TKA.

The present study identified low pre-operative serum albumin, poor glycaemic control, and prolonged operative duration as the most significant predictors of postoperative wound complications. Serum albumin < 3.5 g/dL was independently associated with both DWH and infection. This observation corroborates several large-scale analyses linking hypoalbuminaemia with higher postoperative infection and reoperation rates after arthroplasty (14, 15). Telang et al. (2025) found that patients with albumin < 3.1 g/dL had a three-fold increase in PJI risk, while Almeida et al. (2024) emphasised that even “mild” malnutrition significantly delays wound epithelialisation. The underlying mechanism involves impaired collagen synthesis, reduced fibroblast activity, and compromised immune response (15).

Poor glycaemic control (HbA1c > 7 %) also emerged as a significant determinant of DWH in our cohort. Hwang et al. (2015) demonstrated that HbA1c ≥ 8 % markedly increased SSI rates in diabetic TKA patients, and later threshold analyses suggested risk escalation beginning around 7.5 %. Chronic hyperglycaemia alters leukocyte function and microcirculatory perfusion, thereby slowing granulation and epithelialisation (16). The similarity of our results supports a perioperative HbA1c target below 7 % for elective arthroplasty, consistent with current recommendations.

Another independent factor was prolonged operative time (> 100 min), which significantly increased infection risk. This finding is consistent with the meta-analysis by Shin et al. (2024), who concluded that every 20-minute increment beyond 90 minutes raised the odds of PJI by nearly 30 % (17). Longer surgeries reflect greater soft-tissue exposure, increased blood loss, and higher bacterial burden, highlighting the importance of operative efficiency and team coordination in infection prevention.

Our DWH findings align with the mechanism proposed by Wagenaar et al. (2019), who noted that persistent wound drainage > 72 hours is a strong precursor to infection (18), and with Luthra et al. (2023), who stressed early recognition and evidence-based use of negative-pressure wound therapy (NPWT) for high-risk wounds (11). In the present study, NPWT was used in selected obese and diabetic patients and appeared to reduce drainage duration similar to results reported by Helito et al. (19).

When compared with global data, the mean operative time (≈ 92 min) and blood loss (≈ 280 mL) in this cohort were within the expected range reported by Helito et al. (2020) and Li et al. (2022). The 30-day infection rate of 6.7 % was slightly higher than the pooled mean of 4 % (Bagheri et al., 2025) but can be attributed to the small sample size and inclusion of multiple comorbidities such as diabetes and obesity factors repeatedly implicated in SSI risk (20).

Overall, the present findings reinforce the clinical relevance of nutritional and metabolic optimisation before elective arthroplasty. Correction of hypoalbuminaemia through protein supplementation or dietetic referral, strict perioperative glucose control, and avoidance of unnecessary prolongation of operative time are pragmatic measures that may substantially reduce early wound complications. For patients who remain high-risk, selective NPWT and closer follow-up during the first two postoperative weeks are advisable (11, 19).

This study’s strengths include its prospective design, uniform surgical protocol, and systematic documentation of risk variables. However, the small sample size limits statistical power, and short-term (30-day) follow-up may underestimate late-onset PJI. In addition, multivariate modelling in small samples risks over-fitting. Nevertheless, the direction of associations mirrors larger series, suggesting external validity.

Delayed wound healing and postoperative infection after total knee arthroplasty were predominantly associated with low serum albumin, poor glycaemic control, and prolonged operative time. These are modifiable predictors that can be addressed through preoperative optimisation and efficient intraoperative practice. Maintaining albumin ≥ 3.5 g/dL and HbA1c < 7 %, combined with minimising surgical duration and employing NPWT in selected cases, can substantially improve postoperative wound outcomes and reduce the risk of periprosthetic infection.

1. Poultsides LA, Ma Y, Della Valle AG, Chiu YL, Sculco TP, Memtsoudis SG. In-hospital surgical site infections after primary hip and knee arthroplasty—incidence and risk factors. The Journal of arthroplasty. 2013 Mar 1;28(3):385-9.
2. Ma T, Jiao J, Guo DW, Lv SZ, Zhang D, Hou DC. Incidence of periprosthetic joint infection after primary total knee arthroplasty shows significant variation: a synthesis of meta-analysis and bibliometric analysis. Journal of Orthopaedic Surgery and Research. 2024 Dec;19(1):1-3.
3. Wagenaar FC, Löwik CA, Zahar A, Jutte PC, Gehrke T, Parvizi J. Persistent wound drainage after total joint arthroplasty: a narrative review. The Journal of arthroplasty. 2019 Jan 1;34(1):175-82.
4. Luthra JS, Habsi SA. Management of the Leaky Wound After Arthroplasty: A Review. Journal of Wound Management and Research. 2023 Jul 17;19(3):173-8.
5. Helito CP, Sobrado MF, Giglio PN, Bonadio MB, Pécora JR, Demange MK, Gobbi RG. The use of negative-pressure wound therapy after total knee arthroplasty is effective for reducing complications and the need for reintervention. BMC Musculoskeletal Disorders. 2020 Jul 25;21(1):490.
6. Song BM, Adil SA, Schneider AM, Bendich I. Surgical Site Fat Thickness is More Predictive of Postoperative Wound Complications Than Body Mass Index Following Posterior Approach Total Hip Arthroplasty. The Journal of Arthroplasty. 2025 May 23.
7. Telang SS, Palmer RC, Cooperman WS, Dobitsch A, Lieberman JR, Heckmann ND. Albumin as a Predictor of Periprosthetic Joint Infection Following Total Joint Arthroplasty: Identifying a New Data-Driven Threshold Utilizing a Continuous Variable Analysis. The Journal of Arthroplasty. 2025 Feb 11.
8. Shin KH, Kim JH, Han SB. Greater risk of periprosthetic joint infection associated with prolonged operative time in primary total knee arthroplasty: meta-analysis of 427,361 patients. Journal of Clinical Medicine. 2024 May 22;13(11):3046.
9. Ko MS, Choi CH, Yoon HK, Yoo JH, Oh HC, Lee JH, Park SH. Risk factors of postoperative complications following total knee arthroplasty in Korea: a nationwide retrospective cohort study. Medicine. 2021 Dec 3;100(48):e28052.
10. Luthra JS, Habsi SA. Management of the Leaky Wound After Arthroplasty: A Review. Journal of Wound Management and Research. 2023 Jul 17;19(3):173-8.
11. Eriksson HK, Lazarinis S. Patient-related factors associated with superficial surgical site infection and progression to a periprosthetic joint infection after elective primary total joint arthroplasty: a single-centre, retrospective study in Sweden. BMJ open. 2022 Sep 1;12(9):e060754.
12. Chong YY, Lau CM, Jiang T, Wen C, Zhang J, Cheung A, Luk MH, Leung KC, Cheung MH, Fu H, Chiu KY. Predicting periprosthetic joint infection in primary total knee arthroplasty: a machine learning model integrating preoperative and perioperative risk factors. BMC Musculoskeletal Disorders. 2025 Mar 11;26(1):241.
13. Bagheri A, Niknafs AS, Farhadi B, Mazhari SA, Karimian P, Pour NH, Hojjati H, Nosratabadi I, Emami A, Salehi R. Incidence and Risk Factors of Surgical Site Infection After Knee Arthroplasty; a Systematic Review and Meta-Analysis. Archives of Academic Emergency Medicine. 2025 Jan 18;13(1):e28.
14. Telang SS, Palmer RC, Cooperman WS, Dobitsch A, Lieberman JR, Heckmann ND. Albumin as a Predictor of Periprosthetic Joint Infection Following Total Joint Arthroplasty: Identifying a New Data-Driven Threshold Utilizing a Continuous Variable Analysis. The Journal of Arthroplasty. 2025 Feb 11.
15. Almeida PR, Mokete L, Sikhauli N, Mota A, Ndindwa B, Pietrzak JR. Malnutrition in total joint arthroplasty: what should the orthopaedic surgeon consider?. EFORT Open Reviews. 2024 Jul 1;9(7):615-24.
16. Hwang JS, Kim SJ, Bamne AB, Na YG, Kim TK. Do glycemic markers predict occurrence of complications after total knee arthroplasty in patients with diabetes?. Clinical Orthopaedics and Related Research®. 2015 May;473(5):1726-31.
17. Shin KH, Kim JH, Han SB. Greater risk of periprosthetic joint infection associated with prolonged operative time in primary total knee arthroplasty: meta-analysis of 427,361 patients. Journal of Clinical Medicine. 2024 May 22;13(11):3046.
18. Wagenaar FC, Löwik CA, Stevens M, Bulstra SK, Pronk Y, Wouthuyzen-Bakker M, Nelissen RG, Poolman RW, der WV, Jutte PC. Managing persistent wound leakage after total knee and hip arthroplasty. Results of a nationwide survey among Dutch orthopaedic surgeons. Journal of bone and joint infection. 2017 Nov 3;2(4):202-7.
19. Helito CP, Bueno DK, Giglio PN, Bonadio MB, Pécora JR, Demange MK. Negative-pressure wound therapy in the treatment of complex injuries after total knee arthroplasty. Acta Ortopédica Brasileira. 2017;25(2):85-8.
20. Carroll K, Dowsey M, Choong P, Peel T. Risk factors for superficial wound complications in hip and knee arthroplasty. Clinical microbiology and infection. 2014 Feb 1;20(2):130-5.