The importance of anatomical reduction in the functional outcome of open ankle fractures

Abstract

Introduction

Whilst closed ankle fracture injuries are common, well characterized and have treatment guidelines derived from the literature, open ankle fractures are rare and are not well studied. The few studies that exist are from the developed First World and these well resourced circumstances are not available in most parts of the world. The purpose of this prospective study was to assess the functional outcomes of open ankle fractures and the factors associated with their outcome in an adult population in rural South Africa.

Methods

This prospective study was done at a referral hospital in a rural and semi-rural part of South Africa during the study period from January 1, 2013 until April 30, 2019 and was a retrospective chart review of 59 adult patients who sustained open ankle fractures. Pre-, peri- and postoperative factors in the chart review were assessed for each patient. Twenty-one patients were available for follow-up and were reviewed at least 6 months post injury for a functional assessment using the American Orthopedic Foot and Ankle Society Score (AOFAS).

Results

The average AOFAS was 68,2 out of 100 (fair outcome). The lowest score recorded was 38 (one patient) and the highest score 95 (two patients). Nine patients presented with a poor outcome (<60), six with a fair outcome (60-79), two with a good outcome (80-89) and four with an excellent outcome (90-100). The only significant finding, using a Fisher's exact test, indicated that anatomical reduction of the fracture lead to a fair, good or excellent outcome compared to a non-anatomical reduction leading to a poor functional outcome (p = 0,046).

Conclusions

Our results showed that open ankle fractures are serious injuries with a very variable outcome when managed in a rural or semi-rural setting and that anatomical reduction is required to achieve an optimal functional outcome.

1. Introduction

Ankle fractures are common orthopaedic injuries,¹ with numerous studies available on this topic. Open ankle fractures on the other hand are uncommon injuries with a limited number of studies published. More recent studies looking at demographics and complications in open ankle fractures found these injuries to be more common in elderly females mostly presenting as low energy injuries with Gustilo-Anderson 2, Weber B fracture patterns being the most common.^2,3,4 This differed from earlier studies indicating high energy injuries in young male populations being the more common mechanism of injury.^5,6,7 Studies involving treatment guidelines and subsequent outcomes of these injuries are also limited.^4,8,9,10 Studies have been conducted in high level trauma units with dedicated foot and ankle surgeons and ortho-plastic services available for a holistic management of these injuries. It is advised that these injuries should be managed in units such as these.¹¹ Our study centre presents a typical South African public service centre with Ngwelezana Hospital the main orthopaedic trauma referral centre for 15 district hospitals in the Northern KwaZulu-Natal province. Almost three million people live in the area of northern KwaZulu-Natal and there is a local, catchment population of 334,459 people in the area served directly by Ngwelezana hospital with an average of 400 outpatients seen in various orthopaedic clinics per week. Between 250 and 300 orthopaedic operations are performed every month with approximately 90% for trauma orthopaedics and 10% for cold orthopaedics.¹²

Open ankle fractures form part of the high orthopaedic trauma burden that is managed at the centre. Referral to specialized foot and ankle units with ortho-plastic surgery services is often not available, with most of these injuries managed by junior trainees and general orthopaedic surgeons. It is important to identify the functional outcome of open ankle fractures in this setting, as well as factors which result in a poor outcome in order to identify possible modifiable factors to decrease morbidity and improve clinical outcome. The aim of the study was to assess the functional outcome of open ankle fractures and the factors affecting their outcome in this setting. The secondary aim of this study was to assess how the demographic data and details surrounding the injury in our setting compare with other studies. It was a prospective study design which included a retrospective chart review assessing a number of demographic and peri-operative factors that could possibly affect functional outcome. Patients included in the retrospective chart review were assessed in the orthopaedic clinic at least 6 months post injury. The data collected from the retrospective chart review as well as the functional assessments were analyzed to assess the functional outcome as well as factors affecting the outcome.

2. Methods

This was a prospective study design which included a retrospective chart review of 59 adult patients who sustained open ankle fractures. Several pre, peri and postoperative factors in the chart review were assessed for each patient. These patients were then contacted and 21 available for follow up were reviewed at least 6 months post injury in the orthopaedic clinic for a functional assessment by the principal investigator. Informed consent was obtained from all participants. The patients were from the Northern KwaZulu-Natal drainage area that were treated and followed up at Ngwelezana Hospital. The study period was from January 1, 2013 until April 30, 2019.

2.1. Inclusion criteria

• -All adult patients (18 years and older) who sustained open ankle fractures and treated at Ngwelezana Hospital from January 1, 2013 until April 30, 2019.

2.2. Exclusion criteria

• -Younger than 18 years of age.
• -Closed ankle fractures.
• -Open distal tibia fractures/pilon fractures/foot injuries.
• -Mental health care users.

The retrospective chart review included the following information that was correlated to the functional outcome: age, sex, occupation, body-mass index (BMI), pre-injury osteoarthritis, diabetes mellitus (DM), human immunodeficiency virus (HIV) status, smoking, alcohol use, ankle fracture classification (Danis-Weber & Gustilo-Anderson classifications), time to antibiotics administration, time to debridement (1. Initial washout 2. Definitive debridement), type of skeletal fixation, anatomical reduction of ankle fracture achieved or not during definitive treatment, soft tissue cover, and presence of sepsis.

The clinical assessment of the patients that were available for follow up was done in the orthopaedic outpatient clinic at least six months after the injury. The assessment was done by the principal investigator. The functional assessment tool used was the American Orthopedic Foot and Ankle Society Score (AOFAS).

Descriptive statistics were used to describe the sample groups. A factor analysis was done in order to find significance of factors with regards to functional outcome. A p-value of <0.05 was regarded as statistically significant. The sample size was 59 for the retrospective chart review with 21 patients available for the functional assessment. The study was underpowered but reflects the rarity of this type of injury and provides information about this particular pathology in our setting.

3. Results

Of the 59 patients in the retrospective chart review, 34 (57,6%) were females and 25 males (42,4%). Thirty two (54,2%) were between the ages of 18–44, 23 (39%) between 45 and 64 years old and three (5,1%) over 65 years old. The age was not clearly documented in one patient. The average age was 43,1 years old. Thirty (50,8%) were unemployed, four (6,8%) had office jobs and 11 (18,6%) were manual labourers. Employment statistics were unknown in 14 patients (23,7%). There were no patients with body mass index (BMI) < 18,5, 7 patients (11,9%) with BMI between 18,5 and 24,9, 4 patients (6,8%) with BMI between 25 and 29,9 and 12 patients (20,3%) with BMI of more than 30. The BMI was not documented in 36 patients (61%). Pre-injury osteoarthritis was present in one out of 59 patients (1,7%). Diabetes Mellitus was present in two out of 59 patients (3,4%). Seventeen patients (28,8%) were HIV positive, 39 (66,1%) HIV negative and the status was unknown in three (5,1%). Eight out of 59 patients (13,6%) were smokers. Nine out of 59 patients (15,3%) reported using alcohol. Table 1 details these factors further.

Table 1.

Patient factors.

Patient factors	Age (years)	Gender	Employment	Weight	Osteoarthritis	Diabetes mellitus	HIV	Smoker	Alcohol use
	18–44 n = 32	Male n = 25	Unemployed n = 30	BMI <18,5 n = 0	Present n = 1	Present n = 2	Positive n = 17	Yes n = 8	Yes n = 9
	45–64 n = 23	Female n = 34	Office job n = 4	BMI 18,5-24,9 n = 7	Absent n = 58	Absent n = 57	Negative n = 39	No n = 51	No n = 50
	>65 n = 3		Manual labour n = 11	BMI 25-29,9 n = 4			Unknown n = 3
	Unknown n = 1		Unknown n = 14	BMI > 30 n = 12
				BMI unknown n = 36

BMI = body mass index, HIV = human immunodeficiency virus, n = frequency.

Twenty-three out of 59 patients (39%) sustained high energy injuries, with 34 out of 59 (57,6%) sustaining low energy injuries. The mechanism of injury was unknown in two patients (3,4%). With regards to open ankle fracture injury patterns, one patient (1,7%) sustained a Weber A ankle fracture pattern, 37 (62,7%) a Weber B ankle fracture pattern and 14 (23,7%) a Weber C ankle fracture pattern. Two patients (3,4%) sustained isolated medial malleolar fractures and in five patients (8,5%) the Weber ankle fracture pattern was not documented. There was a single (1.7%) Gustilo-Anderson grade 1 open fracture, 43 (72,9%) Gustilo-Anderson grade 2, 7 (11,9%) Gustilo-Anderson grade 3A, 5 (8,5%) Gustilo-Anderson grade 3B and no Gustilo-Anderson grade 3C fractures. In three patients (5,1%) the grading was not documented. Table 2 summarizes the data on the pattern of injury.

Table 2.

Injury pattern.

Injury pattern	Mechanism of injury	Weber classification	Open fracture grading (Gustilo-Anderson)
	High n = 23	A n = 1	1 n = 1
	Low n = 34	B n = 37	2 n = 43
	Unknown n = 2	C n = 14	3A n = 7
		Other n = 2	3B n = 5
		Unknown n = 5	3C n = 0
			Unknown n = 3

n = frequency.

The time to antibiotic administration was less than 3 h in 20 patients (33,9%), between three and 6 h in 11 patients (18,6%), between six and 12 h in 12 patients (20,3%) and more than 12 h in four patients (6,8%). This data was unavailable in 12 patients (20,3%). Time from injury until initial informal washout, either at the base hospital or at Ngwelezana hospital, was less than 6 h in 28 patients (47,5%), 6–12 h in 1 (1,7%), 12–24 h in 2 (3,4%) and more than 24 h in four patients (6,8%). This data was unavailable in 14 patients (23,7%). No patient was in theatre for definitive debridement in less than 6 h, one patient (1,7%) between six to 12 h, 12 patients (20,3%) between 12 and 24 h and 37 patients (62,7%) after 24 h. The data was unavailable in nine patients (15,3%).

The type of definitive fixation was manipulation and plaster of paris in one patient (1,7%), external fixation in five (8,5%), open reduction and internal fixation with plating and screws in 21 (35,6%), Kirchner wiring in 15 (25,4%) and a combination of Kirchner wiring, plating and screws in 12 patients (20,3%). The choice of fixation was unknown in five patients (8,5%). Anatomical reduction was achieved in 53 out of 59 patients (89,8%). And non-anatomical reduction in six patients (10,2%). With regards to soft tissue cover, 11 (18,6%) wounds were primarily closed, four (6,8%) had a delayed primary closure, 10 (16,9%) wounds were left to heal by secondary intention and nine (15,3%) underwent closure with a skin graft. No flap coverage was performed. Data regarding soft tissue coverage was unavailable in 25 patients (42,4%). In-patient sepsis was noted in one out of 59 patients (1,7%), and chronic sepsis in two out of 59 patients (3,4%). Table 3 summarizes the data on the management of these injuries.

Table 3.

Management.

Management	Time to antibiotics	Time to initial washout	Time to definitive debridement	Type of fixation	Anatomic reduction	Soft tissue cover
	<3 h n = 20	<6 h n = 28	<6 h n = 0	MUA&POP n = 1	Yes n = 53	Primary closure n = 11
	3–6 h n = 11	6–12 h n = 11	6–12 h s n = 1	Exfix n = 5	No n = 6	Delayed primary closure n = 4
	6–12 h n = 12	12–24 h n = 2	12–24 h n = 12	Plating & screws n = 21		Healing by secondary intention n = 10
	>12 h n = 4	>24 h n = 4	>24 h n = 37	K-wiring n = 15		SSG n = 9
	Unknown n = 12	Unknown n = 14	Unknown n = 9	Combination (Plating and screws + K-wiring) n = 12		Flap n = 0
				Unknown n = 5		Unknown n = 25

MUA = manipulation under anaesthesia, POP = plaster of Paris, Exfix = external fixation, SSG = split skin graft, n = frequency.

Of the 59 patients in the retrospective chart review, 21 patients (35,6%) were available for follow up and a functional assessment at least six months post injury. The functional outcome tool used was the AOFAS. The average AOFAS was 68,2 out of 100 (fair outcome). The lowest score recorded was 38 (one patient) and the highest score 95 (two patients). Nine patients presented with a poor outcome (<60), six with a fair outcome (60-79), two with a good outcome (80-89) and four with an excellent outcome (90-100). The average AOFAS was 69,2 if the Gustilo-Anderson IIIB injuries were excluded (fair outcome).

A statistical analysis was done on the 21 patients to correlate demographic factors, premorbid conditions, type of injury sustained and peri-operative management of the injury and complications to the functional outcome. The only significant finding, using a Fisher's exact test, indicated that anatomical reduction of the fracture lead to a fair, good or excellent outcome compared to a non-anatomical reduction leading to a poor functional outcome (p = 0,046).

4. Discussion

The aim of the study was to assess the functional outcomes of open ankle fractures and the factors relating to their outcome. In our study population the average AOFAS was 68,2 out of 100. This correlates with a fair outcome (60-79). Nine patients presented with a poor outcome (<60), six with a fair outcome (60-79), two with a good outcome (80-89) and four with an excellent outcome (90-100). Studies on functional outcomes of open ankle fractures for comparison are limited. Different scoring systems are also used which makes comparisons difficult. In a systematic review of the literature ranging from 1968 until 2010, which included two prospective and 13 retrospective studies, it was found that in patients treated by immediate internal fixation, 81% had a satisfactory result. A limitation in this review was that different scoring systems for outcomes were used by the different authors and so outcomes could only be reported on as “good”, “satisfactory” or “poor”.⁵ A recent study on low energy open ankle fractures in elderly patients reported a mean Enneking score of 36 out of 40, with better results in the group that underwent internal fixation (score of 37 out of 40) compared to those with external fixation (score of 31 out of 40), and those with definitive closure (score of 36 out of 40) compared to those with non-definitive closure (score of 31 out of 40).⁹

The statistically significant finding in our study regarding factors related to a poor functional outcome was failure of anatomical reduction of the ankle fracture. This is an important finding which correlates with the limited literature available, where Hulsker et al. found poor results to be commonly due to non-anatomic reductions, articular surface damage or deep infection in their systematic review of the literature.⁵ With a low chronic infection rate (3,4%) reported in our study despite only 20 out of 59 patients (33,9%) receiving early antibiotic administration within 3 h, this re-emphasizes the importance of early definitive fixation with appropriate techniques and implants to ensure an anatomical reduction of the ankle fracture which positively correlates with a better functional outcome.

Our study population was younger (mean age of 43,1 years) than currently reported in the literature (mean age of 60 years)⁴ and obesity was common, but this factor was poorly reported. The concerns with obese patients presenting with open ankle fractures are the higher risk of complications, longer length of hospital stay as well as higher total costs.¹³ In our study gender (57,6% females) was similar to another recent study (56% females) with females sustaining these injuries more commonly.⁴ There was a high unemployment rate in a younger cohort of patients which should be considered when looking at the overall functional outcomes. It could be speculated that the functional outcomes could be higher, because Ngwelezana Hospital drainage area involves large rural areas where people often have a limited income and seek a disablility grant. The AOFAS system uses both subjective and objective parameters as part of the score, and the subjective assessment could be inaccurate if the patient reports higher levels of pain and disability than which is truly present.

A large percentage (57,6%) were caused by a low energy mechanism of injury. Gustilo-Anderson type 2 open injuries, with a Weber B ankle fracture pattern was the most common fracture pattern. These results correlated with current literature.2, 3, 4 Infection rates in this study was low. (In patient sepsis 1,7% and chronic sepsis 3,4% and an overall sepsis rate of 5,1%). The incidence of diabetes mellitus in our study was two out of 59 patients (3,4%). The prevalence of diabetes mellitus in adults is predicted to rise globally.¹⁴ Diabetic patients with ankle fractures require twice the time compared to non-diabetic patients to achieve fracture union.¹⁵ They have an increased risk for infection at the surgical site as well as amputation, so the skin should be carefully monitored for evidence of infection.¹⁵ Diabetic patients with ankle fractures achieve better outcomes with additional fixation and a minimally invasive approach.¹⁵ These factors should be kept in mind when dealing with diabetic patients with open ankle fractures.¹⁵ The HIV positive incidence is high in this area, but HIV positive patients are not at higher risk of implant sepsis compared to HIV negative patients, although CD4 counts and viral loads should be considered.¹⁶ Therefore, the presence of HIV shouldn't deter the operating surgeon from internal fixation techniques with the aim of an anatomical reduction.

Strong evidence exists regarding early antibiotic administration (less than 3 h from time of injury) in infection prevention.¹⁷ In our study 33,9% of patients (20 out of 59) received antibiotics early (within 3 h from time of injury), however a significant portion had a delay (more than 3 h) in antibiotic administration. Most patients received an informal washout early at the base hospital or in our casualty department, but definitive debridement in theatre was delayed beyond 24 h in most cases. This can be explained by the long travelling distances and delays in transport systems as well as theatre time constraints commonly experienced in rural hospitals, where after-hours theatres need to be shared between different specialities which is the case at our institution.

Immediate open reduction and internal fixation of open fractures is considered safe even in Gustilo-Anderson III injuries.⁵ Only when there is inadequate soft tissue to cover osteosynthesis materials, external fixation should be considered.¹⁸ Our choice of surgical fixation was open reduction and internal fixation with plating and screws in most cases. A combination of plating and screws with K-wiring were also used commonly. One of our common indications for this was when severe contamination was a concern. The fibula was plated in these cases with the medial malleolar fragment K wired. The idea was to be able to easily remove the K-wires on the re-assesment of the wound in 48 h so that the ankle joint could be better exposed and the wound more thoroughly debrided again from the medial side, which was the common side for the open ankle wound. An interesting finding in a study by Ngcelwane M.V. was debris found in the lateral region of the ankle, far away from the commonly encountered medial wound. This was thought to be due to a suction effect in the event of severe Weber B or C injuries. In this study it was advised that a separate lateral incision be made to aid in a more thorough washout.¹⁹ With the unavailability of a dedicated ortho-plastic service, no patients in our study received a flap. The five patients who sustained Gustilo-Anderson IIIB injuries all underwent split thickness skin grafts (SSG) after prolonged vacuum-assisted closure (VAC) therapy.²⁰ The functional outcomes in these patients was a mean of 63,2 with a deep infection rate of 20% (1 out of 5). With an overall follow up rate of 35,6% (21 out of 59 patients) yet a 100% follow up rate of these patients with Grade IIIB injuries (five out of five) we think that patients with less severe injuries and possibly a better functional outcome, were less inclined to return for later assessment and more regular follow up. Patients with associated infection had worse functional outcomes and soft tissue injury is an important risk factor for deep infection.^21,22

In summary, our study showed that open ankle fractures are severe injuries which often have a sub-optimal outcome in a rural and semi-rural setting. However, if a strong emphasis is placed on the anatomical reduction and stable fixation of these injuries, functional outcomes can be improved. In the presence of severe injuries with significant soft tissue involvement the ideal would be early referral to a dedicated ortho-plastic unit.

Funding

This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.

Declaration of competing interest

None.