Short Term Complications in Geriatric Ankle Fractures Using a Protocolized Approach to Surgical Treatment: Is Early Weight Bearing Safe?

Abstract

Background:

Surgical management of geriatric ankle fractures requires unique considerations in addressing operative risks. Prior studies have reached varying conclusions regarding optimal treatment strategies. The primary aim of this study was to determine if surgical fixation following a predetermined treatment protocol was safe and effective. The secondary aim was to determine if immediate weight bearing as tolerated (IWBAT) in a subset of patients was safe or conferred any short-term benefits.

Methods:

This retrospective study included all patients over the age 65 treated surgically for an ankle fracture by a single surgeon over a five-year period. A protocol was used including: augmented fixation techniques, IWBAT for select patients, and specific strategies to minimize soft tissue damage. Complications associated with operative treatment were analyzed. A subgroup analysis of patients with isolated ankle injuries was carried out to compare patients made IWBAT to patients made non-weight bearing (NWB) postoperatively.

Results:

Thirty-four patients were included in the study. Fracture types were predominantly OTA 44B2 (18/34, 53%) and 44B3 (8/34, 24%). Union rate was 100%. Augmented fixation techniques were used in 14/34, 41% of patients. Twenty-one of 34, 62% of patients were allowed IWBAT. There were 4 complications, 12%: 1 malunion, 1 superficial infection, and 2 wound dehiscence. Two patients returned to the operating room for removal of hardware and irrigation and debridement. In the subgroup analysis, the IWBAT group was discharged to a rehabilitation facility at a significantly lower rate than the NWB group, 25% (4/16) vs 90% (9/10; p=0.0036). There were no differences in the complication rates between the two groups

Conclusion:

Acceptable outcomes can be reliably obtained when following a standardized approach to geriatric ankle fracture management. In addition, immediate weight bearing in select patients does not seem to increase complications and may benefit patients by increasing rate of discharge to home.

Level of Evidence: IV

Introduction

The incidence of low-trauma geriatric ankle fractures is rising at a rapid rate.¹ The optimal treatment protocol for unstable ankle fractures in the elderly remains debatable. Closed management and casting was once the mainstay of treatment, but has decreased over time due to higher rates of malunion and nonunion, and lower long-term functional scores compared to surgical fixation.^2–6 While operative management can better restore and maintain anatomy; it carries the inherent risks of surgery. This is particularly challenging in geriatric patients, who tend to have higher perioperative risk due to comorbidities, diabetes, peripheral vascular disease, poor bone stock, diminished soft tissue healing, and lower cardiac reserve.^2,7–14 Given this reality, the postoperative complication rate in surgically treated geriatric ankle fractures has been reported as high as 13-59%.^{3–5,15–18}

Treatment protocols in the literature vary and lack consensus regarding many aspects of care such as fixation strategies, weight bearing recommendations, and soft tissue considerations.^{3–5,15–18} The anecdotal experience of the senior author was that following a predetermined treatment algorithm for geriatric ankle fractures—surgical fixation followed by immediate weight bearing for a subset of patients—seemed to have decreased short-term complications than what has been reported while also providing the early ambulation crucial to this specific population.

To validate the anecdotal experience, this study was formulated. The primary aim of the study was to determine if surgical fixation following a predetermined treatment protocol was safe and effective. The secondary aim was to determine if immediate weight bearing as tolerated (IWBAT) in a subset of patients was safe or conferred any short-term benefits.

Methods

Upon receiving Institutional Review Board approval, a retrospective evaluation of the institutional orthopedic patient database was conducted to identify all patients 65 years or older who were treated with open reduction internal fixation of an ankle fracture by the senior author at a level one trauma center from 2012 through 2017. Patients were identified using applicable Current Procedural Terminology (CPT) codes (Table 1). For all included patients, a standardized treatment protocol was used. Data for this analysis was generated by thorough review of patients’ medical charts including their initial evaluation, details of surgical intervention, pre- and postoperative radiographs, and post-operative follow-up. From this demographic data, fracture type, treatment strategy, hospital course, and complications were recorded. The primary outcome of this analysis was short-term complications. Patients with less than six weeks follow-up were excluded. Fractures were classified according to the OTA/AO fracture and dislocation compendium.

Table 1.

CPT Codes Utilized to Identify Patients Undergoing Surgical Treatment of Ankle Fractures

CPT code	Procedure Description
27766	Open treatment of medial malleolus fracture
27769	Open treatment of posterior malleolus fracture
27792	Open treatment of distal fibular fracture
27814	Open treatment of bimalleolar ankle fracture
27822	Open treatment of trimalleolar ankle fracture medial and/or lateral malleolus; without fixation of posterior lip
27823	Open treatment of trimalleolar ankle fracture medial and/or lateral malleolus with fixation of posterior lip

CPT=current procedural terminology

2.1 Treatment Protocol

All ankle fractures other than stable isolated lateral malleolus fractures were indicated for surgery. All injured ankles were immobilized in a well-padded splint with strict elevation until soft tissue swelling subsided and they were deemed appropriate for surgical fixation. Specific strategies to minimize soft tissue damage were employed. These included scalpel dissection, as opposed to blunt scissor dissection, no tourniquet use, and single layer closure with nylon sutures. A posterolateral incision was used for all distal fibula fractures. Supination external rotation Weber B distal fibula fractures were treated with a posterolateral antiglide plate. In patients with osteoporosis or poor bone stock, 3.5mm quadricortical screws were placed at the apex of the fracture across the tibia for augmented fixation. Syndesmotic injuries were treated in the same fashion. Fractures of the medial malleolus were fixed based on fracture pattern. Vertical fracture patterns were treated with a buttress plate, while horizontal fracture patterns were treated with lag screws. In patients with osteoporosis or poor bone stock, bicortical fixation into the lateral cortex of the distal tibia was achieved with fully threaded cannulated screws for augmented fixation (Figure 1). Large, nondisplaced posterior malleolus fractures were fixed percutaneously with a single anterior to posterior 2.7mm cortical screw (Figure 2). Displaced fractures were treated with open reduction internal fixation through a posterolateral approach and standard buttress plating. All patients were made weight bearing as tolerated in a Controlled Ankle Motion (CAM) Walker Boot unless they met one of the following criteria: large displaced posterior malleolus fragment, complete syndesmotic disruption, profound osteoporosis, or ipsilateral lower extremity injury. Patients who met these criteria were immobilized in a splint and made non-weight bearing (NWB) for 8 weeks.

Figure 1.

AP radiograph demonstrating ankle fracture fixation construct with 3.5mm quadricortical screw placed at the apex of the lateral malleolus fracture and bicortical cannulated screws in the medial malleolus for augmented fixation.

Figure 2.

Lateral radiograph demonstrating ankle fracture fixation construct with a single anterior to posterior 2.7mm cortical screw into a nondisplaced posterior malleolus fragment.

2.2 Data Analysis

Data were stored in and analyzed with Excel 2011 (Microsoft Corp, Redmond, WA, USA). Clinical characteristics were analyzed using descriptive statistics. A subgroup analysis of patients with only isolated ankle injuries was carried out to compare IWBAT patients to NWB patients. The student t- test was used to evaluate the means for each of the two groups and Fisher exact tests were used to analyze the differences in rates. Significance was set at p<0.05

Results

Thirty-six patients were identified who were 65 years or older and underwent open reduction internal fixation of an ankle fracture by the senior author. The cohort was 78% female (28 women, 8 men), had a mean age of 73.4 years (range, 65 to 97 years), and mean body mass index of 28.1 (range, 17.3 to 37.2 kg/m²) at time of injury. Eight (22.2%; 8/36) had diabetes mellitus and two (5.6%; 2/36) were active smokers. Fracture types were predominantly OTA 44B2 (19/36, 53%) and 44B3 (8/36, 22%; Table 2). Time from injury to surgical fixation was on average 4.1 days (range, 1 to 16 days).

Table 2.

Demographics and Complications

	Value
n=	34
Demographics, n(%)
Female	26(76)
OTA 44B2 or 44B3	26(76)
Augmented Fixation	14(41)
IWBAT	21(62)
Complications, n(%)
Nonunion	0(0)
Malunion	1(3)
Superficial infection	1(3)
Superifical dehiscence	2(6)
Return to OR	2(6)

IWBAT= immediate weight bearing as tolerated

Augmented fixation techniques were used in 15/36 (42%) of patients: 13 patients with lateral malleolus quadricortical screws and seven patients with medial malleolus bicortical screws. Twenty-two (61%; 22/36) patients were cleared for immediate full weight bearing. Patients were discharged on average postoperative day 3.8 and one-third (36%; 13/36) were discharged home while the rest were discharged to a rehabilitation facility. Thirty-four of 36 patients were available for clinical follow-up at least six weeks postoperatively, average 24 weeks (range, 6 to 108 weeks). Fracture union rate was 100%. There were four (12%; 4/34) complications: one malunion, two superficial wound dehiscence, and one superficial infection six weeks from surgery managed successfully with superficial irrigation and debridement and hardware retention. Patients with diabetes showed a higher risk for complications, 25% (2/8) vs 8% (2/26; p=0.21). Complications or return to the operating room were not associated with postoperative weight bearing status.

In the subgroup analysis of patients with only isolated ankle injuries, 16 patients were made IWBAT compared to 10 who were NWB in the acute post-operative period. The two groups were similar in terms of age, sex, BMI, comorbidities, diabetes, smoking status, preadmission ambulatory aids, and preadmission living arrangements (Table 3). The IWBAT group was discharged earlier on average than the NWB group, postoperative day 2.2 vs 2.9 (p=0.26). When looking at discharge location, the IWBAT group was discharged to a rehabilitation facility at a significantly lower rate than the NWB group, 25% (4/16) vs 90% (9/10; p=0.0036). There were no differences in the complication rates between the two groups (Table 4).

Table 3.

Demographic Data Comparing Weight Bearing as Tolerated and Non-Weight Bearing Patients

	WBAT	NWB	p value
n=	16	10
Mean age	72.9	74.9	0.55
Female gender	81%	80%	1.00
Body mass index	28.4	29.2	0.71
Comorbidities	2.3	3.6	0.15
Diabetes	7%	40%	0.06
Smoker	13%	0%	0.51
Uses ambulatory aid	25%	30%	1.00
Lives in facility	0%	10%	0.39

Table 4.

Outcome Data Comparing Weight Bearing as Tolerated and Non-Weight Bearing Patients

	WBAT	NWB	p value
Postop time to discharge	2.2	2.9	0.26
Discharge to rehab facility	25%	90%	0.0036
All complications	13%	0%	0.51

Discussion

The rate of fragility ankle fractures in the geriatric population is increasing and has become a burden to the health care system.^1,19 The optimal treatment algorithm for these injuries remains unclear. Without established dogma, the decision for surgical intervention is largely surgeon dependent based upon their interpretation of this risk-benefit profile in this unique cohort. In the geriatric population, the risks of surgery is higher due to comorbidities, osteoporosis, and diminished healing potential.^2,7–14 Yet, as has been demonstrated in many other orthopedic injuries, the importance of early independent mobilization in the elderly cannot be overstated.^20–27

Historically, open reduction internal fixation of ankle fractures in the geriatric population was met with skepticism due to high complication rates. Beauchamp et al, in 1983, were the first to look at patients aged greater than 50, who underwent surgical treatment.¹⁵ Their patients were allowed immediate full weight bearing; they found an alarmingly high complication rate (59%) and recommended against surgery. Recently, with the advancement of fixation methods and implants, the complication rates in the literature have improved but are still relatively high. Srinivasan et al looked at patients over the age of 70 who underwent operative fixation of an ankle fracture and found a complication rate of 18%.¹⁸ Davidovitch et al reviewed their series of patients over the age of 60 and found complications in 13% of patients.¹⁶ In both of these studies, patients were made non-weight bearing for six to eight weeks postoperatively. In our series, even though two-thirds of patients were allowed immediate post-operative weight bearing, the complication rate was lower (11.8%) than what has been reported. One patient developed a malunion, which was asymptomatic without hardware failure and was observed. Two patients had superficial wound dehiscence, which were treated with local wound care and both healed uneventfully by secondary intention. And lastly, one patient developed a wound infection, which was successfully treated with a return trip to the operating room for a superficial irrigation and debridement with hardware retention. There was no identified association between post-operative weight bearing status and complications; however, this study is likely underpowered to detect such a difference.

An important aspect of the proposed surgical protocol was augmented fixation techniques in patients with poor bone stock. This included frequent use of antiglide or buttress plating, quadricortical screws through the fibula and tibia at the apex of the fracture, and bicortical medial malleolus screws. These techniques offer enhanced fixation and their use is supported in the literature. Schaffer and Manoli conducted biomechanical studies looking at posterolateral antiglide plates vs direct lateral neutralizing plates for short oblique fractures. They found an antiglide plate to be stiffer, stronger, and require more energy to failure compared to a lag screw and neutralization plate.²⁸ However, the difference was not clinically significant. More recently, in another biomechanical cadaveric study, Minihane et al studied this comparison specifically in the setting of osteoporotic distal fibular fractures and found similar results. When used in an anti-glide manner, one-third tubular plates show increased stability compared to lateral periarticular locking plates.²⁹ To further neutralize external rotation forces, Panchbhavi et al explored increasing stability with quadricortical screws through the fibula and tibia in an osteoporotic cadaveric biomechanical model.³⁰ Although their findings did not reach statistical significance, compared to just fibular screws, they found quadricortical screws increased failure torque by 9%; external rotation angle at failure by 24%, and energy observed before failure by 34%. With regards to medial malleolus fixation, Ricci et al investigated lag by design with partially threaded cancellous screws vs. lag by technique with fully threaded bicortical screws and found the latter to have superior biomechanical, radiographic, and clinical outcomes.³¹ In this cohort, there were no instances of hardware failure. One patient developed a malunion of the medial malleolus but was asymptomatic. Interestingly, this patient’s medial malleolus fracture was treated with a fully threaded unicortical 2.7mm screw. The results from this analysis are an early clinical validation that the application of these principals, in a standardized protocol, allow for acceptable outcomes and may even allow a subset of them to safely weight bear immediately.

The risks and benefits of early weight bearing in geriatric patients after ankle fracture surgery are not well understood. In the younger population, patients with stable osteosynthesis after ankle fracture experienced benefits from IWBAT including quicker return to full unassisted weight bearing, better mobility, decreased calf atrophy, and reduced osteoporotic changes without an increased rate of hardware failure or reduced functional outcomes.³² In the hip fracture literature, the benefits of early ambulation have clearly been shown in the recovery of elderly patients after surgery.^20–26 Although the morbidity of ankle fractures may be less than that of hip fractures in the elderly population, protracted immobilization in itself can be detrimental to the overall health of geriatric patients.³³ In geriatric ankle fractures, to our knowledge, there is only one other study comparing early weight bearing to NWB. Lynde et al retrospectively reviewed 211 geriatric ankle fracture patients who underwent open reduction internal fixation. One hundred and sixty-six patients were made NWB and 45 patients were allowed early (<15 days) weight bearing based on surgeon preference. They found no difference in complication rate or hardware failure.²⁷ This study echoes those findings. However, this study makes an important distinction by removing surgeon bias and replacing it with a standardized treatment protocol. Without differences in complication rates, it was found that the IWBAT group was discharged to home at a significantly higher rate than the NWB group despite being similar in terms of demographics, comorbidities, pre-injury use of ambulatory aids, and pre-injury living situation. The benefit of being sufficiently independent and functional to be discharged home postoperatively cannot be overstated in the elderly. Besides the obvious savings in health care expenditure, discharge to home vs a skilled nursing or rehabilitation facility has been associated with better outcomes in controlled studies in other orthopedic fields.^34–38 In a large Medicare database study, Kadakia et al looked specifically at outcomes in 19,648 geriatric ankle fracture patients, finding a significant increase in surgical and postoperative medical complications as well as one-year mortality in patients discharged to a facility compared to home. However, this significance was not seen with multivariate analysis suggesting other patient factors may play a more important role in determining one-year mortality following geriatric ankle fractures.³⁹

This study has several limitations inherent to any retrospective case series. This was a single surgeon analysis resulting in a relatively small sample size, which may have led to type II errors in detecting differences in outcomes based upon post-operative weight bearing. Furthermore, the evaluation of this treatment protocol was against historical controls rather than a direct comparison. Although we did a subgroup analysis comparing the IWBAT group to the NWB group was conducted, this was based on specific criteria and not randomized. In the subgroup analysis the assumption was made that our criteria for NWB (a disrupted syndesmosis or large posterior malleolus fragment) was not associated with greater injury severity that would impact the patient’s ability to be discharged home, otherwise this could be a confounding variable. Moreover, we did not investigate individual elements of our treatment protocol to determine their specific impact on outcomes. Two of 36 patients did not have appropriate follow up. If both patients had hardware failure, the complication rate would be similar to what has been reported (17%). Lastly, the decision for IWBAT or NWB was based, in part, on subjective criteria as no specific parameters (outside of intraoperative assessment) were employed to indicate the degree of osteoporosis or size of posterior malleolus fragment, which could lead to surgeon’s selection bias. Despite this study supporting immediate weight bearing following surgery for a subset of geriatric ankle fracture patients, a randomized controlled prospective trial looking further at delayed versus immediate postoperative weight bearing is warranted.

Conclusion

Acceptable outcomes can be reliably obtained when following a standardized approach to geriatric ankle fracture management. Key elements of the studied approach include augmented fixation techniques, immediate weight bearing for select patients, and specific strategies to minimize soft tissue insult. In addition, full weight bearing in select patients does not seem to increase complications and may benefit patients by increasing rate of discharge to home.