Abstract
Introduction
This study investigates the incidence, risk factors for, and clinical outcomes of hematoma following total elbow arthroplasty.
Methods
We retrospectively reviewed patient and surgical characteristics as predictors of post-operative hematoma in 382 total elbow arthroplasty (196 primary, 157 revision, and 29 conversion) between May 2004 and February, 2017. For comparison of outcomes, cases were matched (1:2) with controls by age, gender, type of surgery, and surgical indication.
Results
Nineteen post-operative hematomas (5.0%; 19/382) were identified. Total elbow arthroplasty for post-traumatic arthritis (6.7%; 4/60), aseptic loosening (7.9%; 3/38), sequelae of periprosthetic joint infection (6.1%; 5/81), and non-union (28.6%; 2/7) had the highest incidence of hematoma. Clinic aspiration and compressive wrap was performed in 14 patients and avoided a return to the operating room in 78.6% (11/14). Seven patients (36.8%) required a return to the operating room, of which five (71.4%) had positive cultures and required treatment for prosthetic joint infection. Compared to the matched controls, hematoma formation predicted a higher rate of reoperation (36.8% versus 7.9%; p = 0.007) and a higher rate of subsequent prosthetic joint infection (35.7% versus 0%; p = 0.008).
Conclusion
Hematoma formation is associated with both prosthetic joint infection and return to the operating room after total elbow arthroplasty. Strategies to prevent hematoma formation after total elbow arthroplasty may reduce complication rates.
Keywords: hematoma, total elbow arthoplasty, wound complications, periprosthetic joint infection, osteoarthritis
Introduction
The rates and types of complications following total elbow arthroplasty (TEA) are well described in the literature. Revision surgery can be required due to aseptic loosening, polyethylene wear and osteolysis, periprosthetic fracture, and triceps insufficiency.1–4 Wound complications and ulnar nerve complications are also relatively common after TEA, but infection is the most significant complication affecting long-term implant survival. 1 The reported rate of infection following TEA ranges from 1 to 13%.1,3,5,6
Prior studies have shown that hematoma formation after major joint arthroplasty can be a devastating complication as it increases the rate of return to the operating room and is associated with periprosthetic joint infection (PJI).7,8 One study specifically evaluated complications after TEA and found a 5.5% incidence of wound complications, of which approximately one-third were from hematomas. Interestingly, 27% of the wound hematomas progressed to secondary deep infection. 9
Given the paucity of data regarding post-operative hematoma formation and outcome after TEA, the purpose of this study was to (a) identify the incidence of post-operative hematoma formation after TEA, (b) identify risk factors for hematoma formation, and (c) evaluate outcomes associated with post-operative hematoma.
Methods
With institutional review board approval, this single-institution retrospective review was undertaken. Using an institutional database, all patients undergoing TEA (primary or revision) between May 2004 and February 2017 were identified. Patients undergoing removal of fracture hardware or partial elbow arthroplasty (e.g. radial head arthroplasty) and conversion to TEA were categorized as conversion arthroplasty. This review identified 196 (189 patients) primary TEA, 157 (90 patients) revision TEA, and 29 (29 patients) conversion arthroplasties. The total cohort consisted of 284 patients with an average age of 66.8 years (range: 22–95 years). There were 218 (76.8%; 218/284) female and 66 male (23.2%) patients. All patient charts were reviewed for post-operative complications, with a specific focus on post-operative hematoma formation. Patients were considered to have a hematoma when noted by the treating surgeon in follow-up office notes—clinical evidence of a peri-incisional fluid collection often confirmed by aspiration. In addition, charts were reviewed for details from the index surgery including indication for surgery, components utilized, surgeon, length of clinical follow-up, and range-of-motion at latest follow-up. Any intervention for the hematoma was also retrospectively identified from chart review and collected.
During the study time period, 90.8.% (345/380) of all TEA and 98.1% (153/156) of revision TEA were performed by shoulder and elbow fellowship-trained surgeons with a single surgeon performing greater than half (55.8%; 212/380) of all TEA procedures and 67.9% (106/156) of revisions. Overall, TEA cases were performed by 13 surgeons (seven shoulder and elbow fellowship trained). A uniform surgical technique (triceps management) or implant design was not utilized. Tranexamic acid (TXA) was not used.
To identify risk factors for post-operative hematoma formation, the entire arthroplasty cohort was utilized. Studied variables were age, gender, primary versus revision arthroplasty, drain use, triceps management, post-operative splint use, tourniquet use, and surgical indication for TEA.
For the study cohort, the management of surgical site hematoma was determined and correlated with clinical outcomes. Clinical outcomes included a need for surgical intervention, development of surgical site or PJI, and clinical range-of-motion at last clinical follow-up as judged by the treating surgeon. Management of surgical site hematomas over this study period included compressive wrap, in-office aspiration and compressive wrap, or surgical irrigation and debridement (I&D). The treatment course was dictated by the treating surgeon and was based upon wound appearance and signs of infection or imminent wound-breakdown. In-office aspiration was performed after betadine prep with an 18-gauge needle targeting subcutaneous hematomas with near complete evacuation and decompression of the soft-tissues. Following aspiration, the patient was placed into a compressive sleeve and re-evaluated in two to three weeks. Aspirated fluid was discarded. In this study, infection was defined as one positive culture or evidence of gross purulence. For comparison of outcomes, patients who developed a hematoma post-operatively were matched (1:2) with patients who did not develop a hematoma by age, gender, type of surgery, and indication for surgery. These groups were then compared for clinical outcomes: return to the operating room, infection, and final range-of-motion. Follow-up was a minimum of six months for both the study and control cohorts. There were two patients in the study cohort who were included in the analysis, but lost to follow-up at 59 and 96 days following non-operative management. These patients both underwent revision arthroplasty for infection (resection arthroplasty and reimplantation arthroplasty, respectively).
For this analysis, descriptive statistics (proportions, mean, and ranges) were first used to describe the cohort and the incidence of hematoma formation. To compare proportions between groups, a Chi-squared analysis was utilized. However, to avoid inaccurate results, two-by-two analyses in which a specific outcome had less than 10 occurrences utilized a Fisher’s exact test instead. To compare continuous variables (e.g. range-of-motion) between groups, the data were first checked for parametric distribution (skewness and kurtosis less than 2 and 12, respectively). For non-parametric distributions, a Wilcoxon signed-rank test was utilized and medians were reported for each group. For normal data, a paired Student’s t-test was utilized and mean and standard deviation are reported.
Results
Nineteen post-operative hematomas (5.0%; 19/382) were identified following TEA procedures. These presented on average 11.3 days (range: 6–19 days) following surgery. Seven patients (36.8%; 7/19) with post-operative hematomas presented with concurrent serosanguinous wound drainage. Hematomas complicated 7 (3.6%; 7/196) primary TEA, 10 (6.4%; 10/157) revision TEA, and 2 (6.9%; 2/29) conversion TEA.
No statistically significant differences were noted with regards to patient age or gender and the development of post-operative hematoma following TEA (Table 1). Patients undergoing revision or conversion TEA did have a higher, but not statistically significant, rate of post-operative hematoma (6.5%; 12/186) compared to primary TEA (3.6%; 7/196; p = 0.24). With regards to surgical indication, hematoma was most common following TEA for post-traumatic arthritis, aseptic loosening, infection, and non-union (7.5%; 14/186) compared to osteoarthritis, fracture, rheumatoid arthritis, and remaining surgical indications (2.6%; 5/196; Table 1; p = 0.03). Drain placement was not statistically associated with an increased rate of post-operative hematoma (6.3%; 16/256 versus 2.4%; 3/126; p = 0.13). Tourniquet use provided a decreased rate of hematoma (4.0%; 13/321 versus 9.3%; 6/61; p = 0.10).
Table 1.
Comparison of patients with and without post-operative hematomas following total elbow arthroplasty.
| Post-operative hematoma | No hematoma | p-value | |
|---|---|---|---|
| Number (%) | 19 (5.0%) | 363 (95.0%) | N/A |
| Age (years; range) | 62.4 (23.4–80.1) | 65.3 (22.0–95.4) | 0.32 |
| Gender (%) | |||
| Male | 6 (5.8%) | 97 (94.2%) | 0.22 |
| Female | 13 (4.7%) | 266 (95.3%) | |
| Surgical type (%) | |||
| Primary | 7 (3.6%) | 189 (96.4%) | 0.44 |
| Revision | 10 (6.4%) | 147 (93.6%) | |
| Conversion | 2 (6.9%) | 27 (93.1%) | |
| Surgical indication (%) | |||
| Post-traumatic arthritis | 4 (6.7%) | 56 (93.3%) | 0.08 |
| Rheumatoid arthritis | 1 (1.6%) | 62 (98.4%) | |
| Loosening | 3 (7.9%) | 35 (92.1%) | |
| Infection | 5 (6.1%) | 76 (93.9%) | |
| Periprosthetic fracture | 0 | 11 (100%) | |
| Osteoarthritis | 0 | 16 (100%) | |
| Fracture | 1 (1.4%) | 71 (98.6%) | |
| Non-union | 2 (28.6%) | 5 (71.4%) | |
| Miscellaneous | 3 (8.8%) | 31 (91.2%) | |
| Drain use | 16 (6.3%) | 240 (93.7%) | 0.13 |
| Tourniquet use | 13 (4.0%) | 308 (96.0%) | 0.10 |
| Post-operative splint use | 17 (5.3%) | 304 (94.7%) | 0.75 |
| No DVT chemoprophylaxis | 15 (5.7%) | 248 (94.3%) | 0.45 |
| Triceps management | |||
| Reflecting | 1 (1.8%) | 54 (98.2%) | 0.52 |
| Splitting | 5 (6.4%) | 73 (93.6%) | |
| Sparing | 12 (5.7%) | 197 (94.3%) | |
| Other | 1 (2.5%) | 39 (97.5%) | |
A single hematoma, without wound drainage, required no intervention beyond a compressive wrap and resolved spontaneously. Fourteen (73.7%; 14/19) of the patients with post-operative hematoma underwent aspiration in the clinic at an average of 11.5 days (range: 6–19) following surgery. Aspirated volume ranged from negligible to 300 ml. Four patients (28.6%) underwent repeat aspiration for recurrent hematoma an average of five days following their initial aspiration. Two of these patients (50%; 2/4) required a return to the operating room for formal I&D. In total, aspiration was a successful definitive treatment in 11 (78.6%; 11/14) cases with complete resolution of the hematoma and no reoperation. In total, seven patients (36.8%; 7/19) required a return to the operating room for hematoma evacuation at an average of seven (range: 8–23) days following TEA. Five of these patients (71.4%; 5/7) had positive cultures (all including Staphylococcus aureus (a single polymicrobial infection with Coagulase-negative Staphylococcus)) from intra-operative samples. Of patients undergoing TEA for non-infectious etiology, post-operative hematoma was associated with a higher rate of subsequent PJI (35.7%; 5/14) compared to matched controls (0%; 0/28; p = 0.002). Patients with post-operative hematoma also had a greater risk of unexpected return to the operating room (36.8%; 7/19) compared to matched controls (7.9%; 3/38; p = 0.01).
Patients with post-operative hematomas had no long-term deficit in flexion (122.7° versus 118.0°; p = 0.65) or terminal extension (13.8° versus 9.7°; p = 0.28) compared to their matched controls at an average 538 versus 937 days of follow-up, respectively. All of the patients with an infected post-operative hematoma (5/5) had wound drainage compared to none of those patients with an aseptic hematoma (0/9; p = 0.0005; Table 2).
Table 2.
Comparison of patients with post-operative hematomas following aseptic elbow arthroplasty surgery that developed periprosthetic joint infection to those that did not.
| Infected hematoma | Aseptic hematoma | p-value | |
|---|---|---|---|
| Number (%) | 5 (35.7%) | 9 (64.3%) | N/A |
| Age (years; range) | 68.0 (54.9–80.1) | 58.1 (23.4–73.5) | 0.26 |
| Gender (%) | |||
| Male | 2 (40%) | 3 (60%) | 1.0 |
| Female | 3 (33.3%) | 6 (66.7%) | |
| Surgical type (%) | |||
| Primary | 3 (42.9%) | 4 (57.1%) | 0.64 |
| Revision | 1 (20%) | 4 (80%) | |
| Conversion | 1 (50%) | 1 (50%) | |
| Surgical indication (%) | |||
| Loosening | 1 (33.3%) | 2 (66.7%) | 0.34 |
| Post-traumatic arthritis | 1 (25%) | 3 (75%) | |
| Rheumatoid arthritis | 1 (100%) | 0 (0%) | |
| Miscellaneous | 0 | 3 (100%) | |
| Non-union | 1 (50%) | 1 (50%) | |
| Fracture | 1 (100%) | 0 | |
| Time to hematoma | 10.4 days (6–18) | 11.1 days (6–13) | 0.73 |
| Wound drainage | |||
| Yes | 5 (100%) | 0 | 0.0005 |
| No | 0 | 9 (100%) | |
Discussion
Complications, including delayed or failed wound-healing, are not uncommon following TEA.1–4 PJI, however, is a well-known, devastating complication following TEA that may significantly impact long-term outcome.1,3,5,6 Prior research has demonstrated the nature of acute post-operative hematoma following major joint arthroplasty and its link with increased return to the operating room and PJI.7–9 This study investigates the incidence and risk factors, treatment strategies, and outcomes of hematoma formation following TEA.
Overall, post-operative hematoma following TEA was not uncommon, occurring in 5.0% of all patients and leading to a 1.8% rate of return to the operating room for hematoma I&D. This rate of hematoma formation is slightly higher than previously published rates after joint arthroplasty. One group found a post-operative incidence of wound problems of 5.5% after TEA, only one-third of which were hematoma. 9 Another study evaluating incidence of post-operative hematoma following total knee arthroplasty identified a rate of 0.24% within the first 30 days which required a return to the operating room. 10
Additionally, in this study post-operative hematomas were more common following revision and conversion TEA as compared to primary TEA. Similarly, hematomas were more common following TEA surgery for post-traumatic arthritis, aseptic loosing, infection, and non-union. Contrary to this, one prior study found an increased incidence of post-operative wound complications following TEA in patients with rheumatoid arthritis as compared to post-traumatic arthritis. 9 A higher rate of complications including wound problems and infection following revision total hip arthroplasty and knee arthroplasty has been previously established;11–13 however, higher rates of hematoma following revision have not previously been reported. Based upon this large cohort analysis, hematoma formation is not an uncommon complication, especially following revision and conversion TEA.
Of the 19 patients with post-operative hematoma, 11 underwent aspiration in the clinic and did not require further intervention. This indicates that nonsurgical intervention for a non-draining hematoma can be successful in the majority of cases. In contrast to matched controls, 35.7% of patients who developed hematoma after TEA were subsequently treated for culture-positive PJI. This is similar to findings of PJI risk in the setting of post-arthroplasty hematoma of other major joints.13,14 One study found that development of a hematoma following primary total hip arthroplasty was an independent risk factor for adverse outcomes including PJI as well as increasing morbidity and mortality. 7 Another study, Cheung et al., 8 found a strong association between hematoma formation after total shoulder arthroplasty and development of PJI. However, our analysis specifically distinguishes that post-operative hematoma with wound drainage following TEA poses a high risk of PJI.
Given the association with PJI, aggressive attempts to prevent this seemingly minor complication must be undertaken. As hematomas are associated with revision and conversion TEA, it is likely that greater operative time, surgical dissection, significant scar tissue, and bony debridement play a role in the development of post-operative hematomas. Attempts to minimize these factors must be undertaken. Unfortunately, there is little literature to guide surgeons in these efforts. Hematoma prophylaxis may come from use of peri-operative TXA,15,16 post-operative immobilization, minimizing aggressive pharmacologic anticoagulation, 17 meticulous hemostasis with or without a tourniquet, and minimizing unnecessary soft-tissue dissection and flap formation. Despite abandoning routine drain use in lower extremity arthroplasty after evidence showed its lack of benefit, 18 routine drain use in upper extremity arthroplasty continues—with a paucity of evidence to support it. It is logical to consider that the use of drains in TEA helps to prevent hematoma formation in this superficial joint.
Despite the fact that hematomas have a higher risk of developing PJI and returning to the operating room, patients with post-operative hematomas had no long-term deficit in flexion or terminal extension compared to their matched controls at last follow-up. It is possible that range-of-motion in this at-risk cohort was protected by aggressive non-operative management (aspiration and decompression). Future studies, with observed controls, may be better able to determine the true impact of post-operative hematoma on functional outcomes.
In reviewing this evidence, readers must be cognizant of this study’s limitations. Foremost, this is a retrospective review of a heterogeneous cohort of TEA cases. While this cohort provides the opportunity to capture post-operative complications, there is a lack of uniform intra-operative and post-operative protocols with variation in post-arthroplasty hematoma management. The retrospective nature also complicates a detailed comparison of patients with and without post-operative hematomas, namely from the standpoint of patient-satisfaction and functional scores. Furthermore, we were unable to assess the impact of systemic factors such as specific patient comorbidities, overall health measures, or preoperative anticoagulation. Due to the relatively low utilization of TEA, the sample size of the cohort is limited, making statistical conclusions difficult, and potentially resulting in type-II errors—specifically regarding primary versus revision TEA, drain use, and tourniquet use.
Conclusion
Hematoma formation following TEA is not an infrequent complication in the early post-operative period. As with other major joint arthroplasties, a hematoma is associated with an increased risk of PJI. Strategies to prevent hematoma formation are important and require further investigation.
Footnotes
Declaration of Conflicting Interests: The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Ethical Review and Patient Consent: This study received approval from the Thomas Jefferson University IRB Control Number: 18D.164.
Funding: The author(s) received no financial support for the research, authorship, and/or publication of this article.
ORCID iDs
Benjamin Zmistowski https://orcid.org/0000-0001-7360-3775 Surena Namdari https://orcid.org/0000-0002-8222-554X
References
- 1.Voloshin I, Schippert DW, Kakar S, et al. Complications of total elbow replacement: a systematic review. J Shoulder Elbow Surg 2011; 20: 158–168. [DOI] [PubMed] [Google Scholar]
- 2.Sanchez-Sotelo J, O’Driscoll S, Morrey BF. Periprosthetic humeral fractures after total elbow arthroplasty: treatment with implant revision and strut allograft augmentation. J Bone Joint Surg Am 2002; 84: 1642–1650. [PubMed] [Google Scholar]
- 3.Yamaguchi K, Adams RA, Morrey BF. Infection after total elbow arthroplasty. J Bone Joint Surg Am 1998; 80: 481–491. [DOI] [PubMed] [Google Scholar]
- 4.Goldberg SH, Urban RM, Jacobs JJ, et al. Modes of wear after semiconstrained total elbow arthroplasty. J Bone Joint Surg Am 2008; 90: 609–619. [DOI] [PubMed] [Google Scholar]
- 5.Ewald FC, Simmons EDJ, Sullivan JA, et al. Capitellocondylar total elbow replacement in rheumatoid arthritis. Long-term results. J Bone Joint Surg Am 1993; 75: 498–507. [DOI] [PubMed] [Google Scholar]
- 6.Kasten MD, Skinner HB. Total elbow arthroplasty. An 18-year experience. Clin Orthop Relat Res 1993; 290: 177–188. [PubMed] [Google Scholar]
- 7.Mortazavi SMJ, Hansen P, Zmistowski B, et al. Hematoma following primary total hip arthroplasty: a grave complication. J Arthroplasty 2013; 28: 498–503. [DOI] [PubMed] [Google Scholar]
- 8.Cheung EV, Sperling JW, Cofield RH. Infection associated with hematoma formation after shoulder arthroplasty. Clin Orthop Relat Res 2008; 466: 1363–1367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Jeon I-H, Morrey BF, Anakwenze OA, et al. Incidence and implications of early postoperative wound complications after total elbow arthroplasty. J Shoulder Elbow Surg 2011; 20: 857–865. [DOI] [PubMed] [Google Scholar]
- 10.Galat DD, McGovern SC, Hanssen AD, et al. Early return to surgery for evacuation of a postoperative hematoma after primary total knee arthroplasty. J Bone Joint Surg Am 2008; 90: 2331–2336. [DOI] [PubMed] [Google Scholar]
- 11.Mahomed NN, Barrett JA, Katz JN, et al. Rates and outcomes of primary and revision total hip replacement in the United States medicare population. J Bone Joint Surg Am 2003; 85-A: 27–32. [DOI] [PubMed] [Google Scholar]
- 12.Mortazavi SMJ, Schwartzenberger J, Austin MS, et al. Revision total knee arthroplasty infection: incidence and predictors. Clin Orthop Relat Res 2010; 468: 2052–2059. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Poss R, Thornhill TS, Ewald FC, et al. Factors influencing the incidence and outcome of infection following total joint arthroplasty. Clin Orthop Relat Res 1984; 182: 117–126. [PubMed] [Google Scholar]
- 14.Pulido L, Ghanem E, Joshi A, et al. Periprosthetic joint infection: the incidence, timing, and predisposing factors. Clin Orthop Relat Res 2008; 466: 1710–1715. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Chevet I, Remérand F, Couvret C, et al. [Tranexamic acid reduces haematomas but not pain after total knee arthroplasty]. Ann Fr Anesth Reanim 2011; 30: 17–24. [DOI] [PubMed] [Google Scholar]
- 16.Sun C-X, Zhang L, Mi L-D, et al. Efficiency and safety of tranexamic acid in reducing blood loss in total shoulder arthroplasty. Medicine 2017; 96: e7015–e7015. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Parvizi J, Ghanem E, Joshi A, et al. Does ‘excessive’ anticoagulation predispose to periprosthetic infection? J Arthroplasty 2007; 22: 24–28. [DOI] [PubMed] [Google Scholar]
- 18.Parker MJ, Roberts CP, Hay D. Closed suction drainage for hip and knee arthroplasty: a meta-analysis. J Bone Joint Surg 2004; 86: 1146–1146. [DOI] [PubMed] [Google Scholar]