Traumatic knee dislocations: a national database analysis

Abstract

Background

Knee dislocation is a rare but severe orthopedic emergency with a substantial risk of limb-threatening complications. This retrospective national database cohort study aimed to describe the epidemiology of traumatic knee dislocations and identify risk factors for vascular and neurological injuries.

Methods

A total of 164 patients diagnosed with knee dislocation between 2016 and 2023 were included in this retrospective study. Data were obtained from the Ministry of Health system, and diagnoses were confirmed. Demographics, injury mechanisms, associated fractures, and complications were evaluated using International Classification of Diseases (ICD) codes. Logistic regression was used to identify independent risk factors.

Results

The study population consisted primarily of males (81.7%) with an average age of 31.74 years. Vascular injuries occurred in 19.5% of patients, most commonly involving the popliteal artery. Neurological injury was present in 20.1%. In patients with open injuries, infection, vascular injury, amputation, debridement, and soft tissue defect repair were found to be significantly higher. Despite these findings, open injury was not an independent predictor of vascular injury in multivariate analysis. Charlson Comorbidity Index (CCI) > 0 independently predicted the need for vascular procedures. Younger age was associated with neurological injury. Intensive Care Unit (ICU) admission was associated with systemic injuries. The overall amputation rate was 1.2%.

Conclusion

Traumatic knee dislocations frequently involve serious neurovascular complications. While open injuries were associated with a more complicated clinical course, only CCI independently predicted the need for vascular intervention. Early recognition of patients with comorbidities and significant concomitant injuries may improve outcomes.

Level of evidence

Level IV.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12891-025-09409-x.

Introduction

Traumatic knee dislocation is a severe injury in orthopedic trauma, characterized by the disruption of the tibiofemoral articulation, typically accompanied by at least one ligament injury [1]. These are rare injuries, with an incidence reported as 0.2%−0.02% of all musculoskeletal injuries [2, 3]. However, the actual incidence may be higher, as spontaneous reduction can occur, leading to misdiagnosis in some cases [2]. Regardless, traumatic knee dislocation is an orthopedic emergency, with the potential to cause limb loss [4].

Knee dislocations most commonly result from high-energy trauma, with motor vehicle accidents being the leading cause. However, they can also occur less frequently due to sports injuries or low-energy trauma, particularly in elderly, obese, and female individuals [1, 3–6]. These injuries are often associated with multiple traumas [3]. Knee dislocations can be classified in various ways, including by the direction of dislocation (as suggested by Kennedy), by the presence of additional injuries, or by the involvement of ligamentous injuries [3, 7]. Fractures may also accompany knee dislocations, and fracture-dislocations are typically associated with high-energy trauma [8]. Additionally, vascular and nerve damage may also occur in these injuries [3, 4, 9].

Given the rarity of these injuries, their characteristics are generally derived from small series or database studies, which may include misdiagnoses. Therefore, the aim of this study is to examine and describe in detail the demographic characteristics, complications, and general features of knee dislocations, as well as to identify the risk factors for vascular and nerve injuries, using data from a national database. Our hypothesis is that in patients with traumatic knee dislocations, the presence of open injuries, additional comorbidities, and fractures leads to an increased incidence of vascular and neurological complications.

Materials and methods

In this study, patients diagnosed with knee dislocation according to the International Classification of Diseases (ICD) across Turkey between 2016 and 2023 were screened. Data were obtained from the Ministry of Health system under special permission. A total of 239 patients were initially identified. Inclusion criteria were confirmed acute knee dislocation through ICD codes. Exclusion criteria were misclassified cases, non-traumatic knee dislocations, iatrogenic knee dislocations, cases with insufficient documentation and imaging to verify diagnosis. After examination a final study population of 164 patients was achieved. All radiological images were re-evaluated by single orthopedic surgeon, experienced in knee trauma, to confirm the diagnosis.

Patient characteristics, including sociodemographic information and the etiology of the injury, were recorded. Additionally, details such as intensive care unit (ICU) admission, open injuries, additional systemic injuries, fractures, neurological and vascular injuries, mortality, initial interventions, infections, amputations, debridement, external fixation treatment, and subsequent surgeries were meticulously documented throughout the patient’s initial admission and treatment process.

The fracture status of the patients was assessed using the relevant ICD codes within one month of the initial diagnosis. Infection parameters were screened using the appropriate ICD codes from the time of the initial diagnosis of knee dislocation to the present. Neurological injuries were also reviewed using relevant ICD codes from the initial diagnosis to the present. Vascular injuries were assessed using ICD codes within the first month following the initial diagnosis of knee dislocation. Vascular systemic complications were screened from the initial diagnosis to the present. Vascular and neurological procedures were identified using the relevant codes. Additionally, the Charlson Comorbidity Index (CCI) was calculated by identifying comorbidity-related ICD codes [10]. The CCI is a validated comorbidity scoring system originally developed by Charlson et al. and later adapted for ICD-coded databases by Quan et al. [11, 12]. It consists of 19 items, each representing a separate comorbid condition, with the total score reflecting the overall comorbidity burden. A higher CCI score not only indicates increased mortality risk but also reflects the presence of clinically significant comorbidities.

The study was conducted in accordance with the Declaration of Helsinki and received approval from the Turkish Ministry of Health with a waiver of informed consent for retrospective data analysis and the health information privacy law (ID: 95741342-020/27112019).

Statistical methods

Data were analyzed using IBM SPSS V23. Normality was assessed using the Shapiro-Wilk test and the Kolmogorov-Smirnov test. The comparison of categorical variables between groups was performed using the Chi-square test, Yates correction, Fisher’s Exact test, and Fisher-Freeman-Halton test. Multiple comparisons of proportions were analyzed using the Bonferroni corrected Z test. The Mann-Whitney U test was used for comparing non-normally distributed data between two groups. For comparisons of non-normally distributed data among three or more groups, the Kruskal-Wallis test was employed, followed by Dunn’s test for multiple comparisons. The impact of independent risk factors on the dependent variable was examined using binary logistic regression analysis. Results were presented as frequency (percentage) for categorical data, and as mean ± standard deviation and median (minimum-maximum) for quantitative data. A significance level of p < 0.050 was considered statistically significant.

Results

A total of 164 patients were included in the study, of whom were male (81.7%), with a mean age of 31.74 years. The most affected age group was 20–30 years (Fig. 1). The distribution of trauma cases by years and months is illustrated in Fig. 2. CCI score of zero was observed in 60.4% of the patients (Fig. 3). Patients with CCI = 0 were significantly younger (mean: 26.9 ± 9.3; median: 25 [12–55]) compared with those with CCI > 0 (mean: 39 ± 16; median: 37 [15–74]) (p < 0.001). Additional systemic injuries were identified in 9.8% of the patients (9 head traumas, 11 thoracic traumas, and 1 abdominal trauma) out of the 163 patients for whom this information was available. Fractures were not observed in 51.8% of the patients; however, 17.7% had periarticular fractures of the knee, 14.6% had fractures both around the knee and in other regions, and 15.9% had fractures in areas other than around the knee. Of the 119 patients whose initial intervention could be determined, external fixator use was identified in 33 patients (27.7%) during the initial intervention, while the number of patients who had an external fixator used at any stage of the treatment was 49 out of 164 (29.9%). The Schenck and Kennedy classifications were used in this study, with the Schenck classification determined for 121 patients and the Kennedy classification determined for 83 patients; and vascular systemic complications such as deep vein thrombosis and pulmonary embolism were observed in 9.8% of the patients (Table 1).

Fig. 1.

Age distribution of patients with knee dislocation

Fig. 2.

Distribution of knee dislocation cases by month and year

Fig. 3.

Distribution of Charlson Comorbidity Index (CCI) scores among patients

Table 1.

Descriptive statistics of variables

	Frequency (n)/Mean ± SD	Percentage (%)/Median (min - max)
Gender
Male	134	81,7
Female	30	18,3
Age	31,7 ± 13,7	27,5 (12,0–74,0)
Charlson Comorbidity Index (CCI)	0,8 ± 1,5	0,0 (0,0–8,0)
CCI Categories
0	99	60,4
> 0	65	39,6
Etiology
Motorcycle accident (Mean Age;26,6 ± 9,4)	51	33,1
Fall from height (Mean Age;31,4 ± 11,7)	22	14,3
Occupational accident (Mean Age;41,1 ± 15,2)	16	10,4
Traffic collision (Mean Age;34,3 ± 15,8)	51	33,1
Sports and falls (Mean Age;30,5 ± 13,3)	14	9,1
Side
Right	78	49,7
Left	79	50,3
Schenck classification
KDI	23	19
KDII	15	12,4
KDIII	53	43,8
KDIV	4	3,3
KDV	26	21,5
Kennedy classification
Anterior	22	26,5
Posterior	21	25,3
Medial	5	6
Lateral	16	19,3
Rotational	19	22,9
Intensive Care Unit (ICU) Admission
No	136	82,9
Yes	28	17,1
Open injury
No	132	80,5
Yes	32	19,5
Initial intervention
Closed Reduction (CR) + Splint	80	67,2
CR + External fixation	33	27,7
Spontaneously reduced	1	0,8
Open surgery	5	4,2
Mortality
No	161	98,2
Yes	3	1,8
Additional systemic injuries
No	147	90,2
Yes	16	9,8
Fractures
No fracture	85	51,8
Peri-knee fracture	29	17,7
Non-knee fracture	26	15,9
Peri-Knee + Non-knee fracture	24	14,6
Infection
No	155	94,5
Yes	9	5,5
Neurological Injury
No	131	79,9
Yes	33	20,1
Neurological Procedure
No	145	88,4
Yes	19	11,6
Vascular Imaging
No	59	36
Yes	105	64
Vascular Injury
No	132	80,5
Yes	32	19,5
Vascular Procedure
No	146	89
Yes	18	11
Amputation
No	162	98,8
Yes	2	1,2
Arthrodesis
No	163	99,4
Yes	1	0,6
Debridement
No	144	87,8
Yes	20	12,2
External fixation treatment
No	115	70,1
Yes	49	29,9
Soft tissue defect repair
No	158	96,3
Yes	6	3,7
Fasciotomy
No	159	97
Yes	5	3
Systemic vascular complications
No	148	90,2
Yes	16	9,8

Advanced neurological investigations such as electromyography (EMG) were conducted for 40 patients. Of these patients, 33 received a neurological injury, and 19 underwent a neurological procedure. A total of 105 patients underwent vascular imaging. Among these, 86 patients had Computed Tomography (CT) angiography, and 36 had Doppler ultrasound; 19 of the Doppler ultrasound examinations were performed as the only imaging modality, while 16 were performed in addition to CT angiography. One patient underwent all three modalities: CT angiography, Doppler ultrasound, and Magnetic Resonance Imaging (MRI) angiography. Vascular diagnoses were made in 32 patients, with 18 of these patients undergoing a vascular procedure. Among the 32 patients diagnosed with vascular injury, popliteal artery injury was identified in 18 patients. Vascular injuries other than the popliteal artery were observed in 3 patients. In 11 patients, the injured vessel was not specified (Table 1).

When parameters evaluated, open injuries demonstrated significant associations with multiple adverse outcomes, including higher rates of ICU admission, additional fractures, infection, vascular injury, amputation, debridement, and soft tissue-defect repair, with multiple comparisons showing a difference in open injury rates between posterior and lateral dislocations in the Kennedy classification. Patients with peri-knee fractures exhibited significantly higher rates of neurological procedures and debridement. When patients diagnosed with neurological injury were evaluated, significant differences were identified in the Kennedy classification, vascular injury, vascular procedures, soft tissue defect repair, and fasciotomy. The highest proportion of neurological injury was observed in Kennedy anterior dislocations. Patients with neurological injuries were significantly more likely to have vascular injuries and required significantly more vascular procedures (Table 2). When patients were compared according to vascular injury, a significant difference was found in ICU admission rates. It was observed that these patients were more frequently admitted to the ICU, had a higher incidence of open injuries, and experienced significantly higher infection rates. Debridement and soft tissue defect repair were also significantly more common, and fasciotomy was significantly more frequent in this patient group (Table 3).

Table 2.

Comparison results according to neurological injury

Parameter	No	Yes	Total	Test Statistic	p-value
Schenck Classification
KDI	16 (16,5)	7 (29,2)	23 (19)	6,482	0,146t
KDII	12 (12,4)	3 (12,5)	15 (12,4)
KDIII	43 (44,3)	10 (41,7)	53 (43,8)
KDIV	2 (2,1)	2 (8,3)	4 (3,3)
KDV	24 (24,7)	2 (8,3)	26 (21,5)
Kennedy Classification
Anterior	12 (54,5)a	10 (45,5)I	22 (26,5)	12,227	0,010t
Posterior	18(85,7)ab	3 (14,3)	21 (25,3)
Medial	4 (80)ab	1 (20)	5 (6)
Lateral	16 (100)b	0 (0)	16 (19,3)
Rotational	14(73,7)ab	5 (26,3)	19 (22,9)
Vascular Injury
No	116 (88,6)	16 (48,5)	132 (80,5)	24,451	< 0,001y
Yes	15 (11,5)	17 (51,5)	32 (19,5)
Vascular Procedure
No	125 (95,4)	21 (63,6)	146 (89)	—	< 0,001z
Yes	6 (4,6)	12 (36,4)	18 (11)
Soft Tisue Defect Repairs
No	129 (98,5)	29 (87,9)	158 (96,3)	—	0,016z
Yes	2 (1,5)	4 (12,1)	6 (3,7)
Fasciotomy
No	130 (99,2)	29 (87,9)	159 (97)	—	0,006z
Yes	1 (0,8)	4 (12,1)	5 (3,1)

^yYates correction, ^zFisher’s Exact Test, ^tFisher-Freeman-Halton test, n(%), a-b: There is no difference among the ratios that share the same letter within each column. The underlined data are given as row percentages, while the others are given as column percentages

Table 3.

Comparison results according to vascular injury

Parameter	No	Yes		Total		Test Statistic	p-value
Schenck classification
KDI	18 (17,7)	5 (26,3)		23 (19)		4,480	0,321t
KDII	13 (12,8)	2 (10,5)		15 (12,4)
KDIII	46 (45,1)	7 (36,8)		53 (43,8)
KDIV	2 (2)	2 (10,5)		4 (3,3)
KDV	23 (22,6)	3 (15,8)		26 (21,5)
Kennedy classification
Anterior	18 (27,7)	4 (22,2)		22 (26,5)		9,110	0,045t
Posterior	16 (24,6)	5 (27,8)		21 (25,3)
Medial	3 (4,6)	2 (11,1)		5 (6)
Lateral	16 (24,6)	0 (0)		16 (19,3)
Rotational	12 (18,5)	7 (38,9)		19 (22,9)
ICU Admission
No	117 (88,6)	19 (59,4)		136 (82,9)		13,578	< 0,001y
Yes	15 (11,4)	13 (40,6)		28 (17,1)
Open injury
No	111 (84,1)	21 (65,6)		132 (80,5)		4,478	0,034y
Yes	21 (15,9)	11 (34,4)		32 (19,5)
Infection
No	128 (97)	27 (84,4)		155 (94,5)		—	0,015z
Yes	4 (3)	5 (15,6)		9 (5,5)
Neurological procedure
No	116 (87,9)	29 (90,6)		145 (88,4)		—	1,000z
Yes	16 (12,1)	3 (9,4)		19 (11,6)
Debridement
No	121 (91,7)	23 (71,9)		144 (87,8)		—	0,005z
Yes	11 (8,3)	9 (28,1)		20 (12,2)
Soft Tissue Defect Repair
No	131 (99,2)	27 (84,4)		158 (96,3)		—	0,001z
Yes	1 (0,8)	5 (15,6)		6 (3,7)
Fasciotomy
No	131 (99,2)	28 (87,5)		159 (97)		—	0,005z
Yes	1 (0,8)	4 (12,5)		5 (3,1)

^yYates correction, ^zFisher’s Exact Test, ^tFisher-Freeman-Halton test, n(%)

When the risk factors for ICU admission were examined using logistic regression, open injuries and fractures showed significant associations in the univariate analysis; however, only the presence of additional systemic injuries remained an independent risk factor in the multivariate model. For infection, fractures were a significant factor in univariate analysis, but open injury was the only independent predictor with higher infection risk in multivariate analyses. For neurological injuries, age and Schenck KDV were risk factors, with increasing age associated with a lower likelihood of neurological deficit. In the analysis of vascular injuries, open injuries were significant in the univariate model but did not retain significance in multivariate model. The only independent predictor in multivariate analysis for requiring a vascular procedure was a higher CCI (Table 4).

Table 4.

Investigation of the effect of independent risk factors on various parameters using binary logistic regression analysis

Parameter	Univariate		Multiple
ICU Admission	OR (95% CI)	p-value	OR (95% CI)	p-value
Age	1,009 (0,98 − 1,038)	0,563	0,983 (0,933–1,036)	0,528
CCI	1,156 (0,91 − 1,468)	0,236	1,33 (0,864–2,049)	0,195
Etiology (Reference: Motorcycle accident)
Fall from height	0,629 (0,12 − 3,299)	0,583	0,83 (0,128–5,38)	0,845
Occupational accident	1,451 (0,328–6,419)	0,624	1,483 (0,243–9,047)	0,669
Traffic collision	2,619 (0,964–7,115)	0,059	1,698 (0,497–5,804)	0,399
Sports and falls	0,484 (0,054 − 4,298)	0,514	0,968 (0,096 − 9,804)	0,978
Open injury (Reference: No)	2,879 (1,173–7,064)	0,021	3,129 (1–9,789)	0,050
Additional systemic injury (Reference: No)	12,745 (4,112 − 39,507)	< 0,001	13,111 (3,405 − 50,494)	< 0,001
Fractures (Reference: No)	2,674 (1,129–6,334)	0,025	1,648 (0,541–5,014)	0,379

Infection
Age	0,954 (0,892–1,021)	0,174	0,962 (0,879–1,053)	0,400
CCI	0,474 (0,14 − 1,607)	0,231	0,572 (0,14 − 2,336)	0,436
Open injury (Reference: No)	9,923 (2,331 − 42,241)	0,002	7,894 (1,583 − 39,374)	0,012
Additional systemic injury (Reference: No)	1,158 (0,135–9,905)	0,893	0,662 (0,05–8,768)	0,754
Fractures (Reference: No)	9,465 (1,156 − 77,502)	0,036	5,826 (0,631 − 53,813)	0,120

Neurological Injury
Age	0,99 (0,961–1,019)	0,499	0,906 (0,846–0,969)	0,004
CCI	1,122 (0,889–1,416)	0,334	1,464 (0,823–2,605)	0,194
Etiology (Reference: Motorcycle accident)
Fall from height	1,537 (0,479–4,931)	0,469	---	---
Occupational accident	0,273 (0,032 − 2,321)	0,235	---	---
Traffic collision	0,879 (0,324–2,384)	0,799	---	---
Sports and falls	3,075 (0,869 − 10,887)	0,082	---	---
Schenck (Reference: KDI)
KDII	0,571 (0,122–2,681)	0,478	0,375 (0,064 − 2,204)	0,278
KDIII	0,532 (0,173–1,635)	0,270	0,336 (0,092 − 1,229)	0,099
KDIV	2,286 (0,266 − 19,658)	0,451	2,478 (0,183 − 33,462)	0,494
KDV	0,19 (0,035 − 1,036)	0,055	0,077 (0,011 − 0,525)	0,009
Open injury (Reference: No)	0,897 (0,336–2,399)	0,829	0,518 (0,115–2,329)	0,391
Additional systemic injury (Reference: No)	0,535 (0,115–2,478)	0,423
Fractures (Reference: No)	1,182 (0,551–2,539)	0,667	2,335 (0,783–6,967)	0,128

Vascular Injury
Age	1,009 (0,982–1,037)	0,506	0,987 (0,93 − 1,048)	0,678
CCI	0,969 (0,739–1,269)	0,818	1,014 (0,58 − 1,771)	0,962
Etiology (Reference: Motorcycle accident)
Fall from height	1,194 (0,319–4,473)	0,792	---	---
Occupational accident	1,792 (0,46 − 6,982)	0,401	---	---
Traffic collision	2,034 (0,768–5,384)	0,153	---	---
Sports and falls	0,413 (0,047 − 3,619)	0,425	---	---
Schenck (Reference: KDI)
KDII	0,554 (0,093 − 3,312)	0,517	0,403 (0,06 − 2,724)	0,351
KDIII	0,548 (0,154–1,952)	0,353	0,321 (0,078 − 1,321)	0,115
KDIV	3,6 (0,4–32,366)	0,253	2,675 (0,225 − 31,756)	0,436
KDV	0,47 (0,099 − 2,231)	0,342	0,223 (0,037 − 1,329)	0,099
Open injury (Reference: No)	2,769 (1,165–6,581)	0,021	2,31 (0,672–7,936)	0,184
Additional systemic injury (Reference: No)	0,939 (0,251–3,513)	0,926	0,834 (0,082 − 8,459)	0,878
Fractures (Reference: No)	1,754 (0,801–3,842)	0,160	2,422 (0,735–7,98)	0,146

Vascular Procedure
Gender (Reference: Male)	1,319 (0,401–4,332)	0,648	1,589 (0,309–8,175)	0,579
Age	0,993 (0,957–1,031)	0,709	0,938 (0,867–1,014)	0,108
CCI	1,196 (0,915–1,563)	0,191	1,848 (1,027 − 3,325)	0,041
Etiology (Reference: Motorcycle accident)
Fall from height	0,75 (0,139–4,043)	0,738	---	---
Occupational accident	0,5 (0,056 − 4,495)	0,536	---	---
Traffic collision	0,815 (0,232–2,862)	0,750	---	---
Sports and falls	2,045 (0,441–9,491)	0,361	---	---
Schenck (Reference: KDI)
KDII	1,615 (0,202 − 12,91)	0,651	1,305 (0,131 − 12,946)	0,820
KDIII	1,094 (0,196–6,097)	0,919	0,883 (0,138–5,652)	0,895
KDIV	3,5 (0,238 − 51,46)	0,361	2,307 (0,129 − 41,21)	0,570
KDV	0,875 (0,113–6,767)	0,898	0,422 (0,042 − 4,243)	0,464
Open injury (Reference: No)	2,308 (0,793–6,713)	0,125	2,548 (0,6–10,82)	0,205
Additional systemic injury (Reference: No)	0,51 (0,063 − 4,107)	0,527	---	---
Fractures (Reference: No)	1,803 (0,662–4,909)	0,249	1,894 (0,433–8,295)	0,397

KD Knee Dislocation, ICU Intensive Care Unit

ICU Admission (Omnibus Test (=27,751; p=0,004); Hosmer and Lemeshow Test (=6,478; p=0,594)); Infection (Omnibus Test (=18,503; p=0,010); Hosmer and Lemeshow Test (=3,231; p=0,919)); Neurological Injury (Omnibus Test (=21,905; p=0,016); Hosmer and Lemeshow Test (=11,041; p=0,199)); Vascular Injury (Omnibus Test (=11,243; p=0,423); Hosmer and Lemeshow Test (=8,405; p=0,395)); Vascular Procedure (Omnibus Test (=8,692; p=0,562); Hosmer and Lemeshow Test (=11,023; p=0,200))

“For categorical variables, odds ratios are calculated relative to the reference category”

Discussion

Several factors were associated with a worse clinical course in traumatic knee dislocations. Open injuries, higher CCI scores, younger age, dislocation classification, additional injuries and associated fractures were all linked to different complications.

Arom et al. [4] reported an overall incidence of 0.072 dislocation events per 100 patient-years and Chowdhry et al. [13] found an incidence rate of 18 cases per 10.000 admissions. In our study, 164 cases were confirmed, which is relatively low compared to the literature. This may be due to reliance on ICD codes, as they can be entered incorrectly, or can be missing. For example, a knee dislocation might be incorrectly coded as a patella dislocation. Also, 20–80% of patients present with spontaneous reduction and may not receive a correct diagnosis [8].

Open injuries are reported in the literature to occur at rates between 4 and 17% [14]. Arom et al. [4] found a 17% incidence of open injuries, while Chowdhry et al. [13] reported a 13.6% incidence. In the latter study, 29% of patients underwent vascular investigation, 15% were found to have an injury, and 10.8% underwent a vascular procedure; associated fractures occurred in 41.4%, neurological injury in 6.2%, compartment syndrome in 2.7%, and amputations in 3.8%, with male sex, vascular injury, and open injury identified as risk factors for severe complications [13]. In our study, the open-injury rate was 19.5%, slightly higher than what has been reported in the literature. CCI greater than zero was the only independent variable associated with the need for vascular procedures. In a series of seven open knee dislocations by King et al., vascular injury occurred in 29% of cases, tibial fractures in five patients, femoral fractures in two, and infections in 43% [14]. In our study, there were 32 patients with open injuries. Among them, vascular imaging was performed in 71.8% and vascular injury was identified in 34.4%, which is comparable to or slightly higher than the literature. Vascular intervention, infection, neurological injury, and neurological procedures each occurred in 18.75% of cases. Debridement was performed in 28.1% of patients, external fixators were used in 40.6%, and soft tissue-defect repair was required in 15.6%. Notably, all amputations in our series occurred in patients with open injuries. Overall, the complication rates within our open-injury subgroup align with the high morbidity reported in the literature and underscore the severe clinical course typically associated with open traumatic knee dislocations.

The most common vascular injury associated with traumatic knee dislocation is popliteal artery injury. The popliteal artery is particularly vulnerable due to its relatively constrained anatomical positioning above and below the knee [9]. It is anchored at the adductor hiatus and the fascial arch of the soleus, making it susceptible to damage during significant displacement, such as that seen in knee dislocations [15]. Reported rates range from 20% to 40% in high-energy trauma, and 5–11% in low-energy mechanisms [1, 9]. Early diagnosis and repair are critical, as delays can result in limb loss and amputation. In our study, the most common injury was also the popliteal artery (18/32). Among the 21 patients with documented vascular injuries and identified vessels, 18 were found to have popliteal artery injuries. The optimal strategy for vascular imaging in knee dislocations remains unclear. Some authors advocate for routine angiography, while others recommend selective imaging based on specific criteria [3]. Chowdhry et al. [13] found that vascular investigation was performed in 29% of cases, with injuries documented in 15%, and 10.8% of patients received vascular procedures. In our study, the vascular imaging rate was 64.02% (105/164), with a vascular injury rate of 19.51% (32/164), and a vascular intervention rate of 10.97% (18/164). Hollis et al. reported an amputation rate of 8% in cases of traumatic knee dislocation, and Medina et al. reported an amputation rate of 12% in patients with vascular injuries [9, 16]. Our amputation rate was 1.21% (2/164), 6.25% (2/32) among those with vascular injuries, which is lower than the reported rates. Wright reported an amputation rate of 16%, with two-thirds of these amputations resulting from subsequent infection despite debridement [17]. Debridement procedures were performed in both amputations in our study, and a total of 20 patients underwent debridement at some point during their treatment.

The most injured nerve in traumatic knee dislocation is the common peroneal nerve. Similar to the popliteal artery, the common peroneal nerve is surrounded by tight fibrous structures above and below the knee, which limits its ability to adapt to changes in knee position during trauma [2]. Additionally, the peroneal nerve has inconsistent blood flow compared to other peripheral nerves and is prone to neuropraxia due to poor intraneural vascularity around the fibular neck [18]. Peroneal nerve damage can occur in approximately 10–40% of traumatic knee dislocations, and symptoms include foot drop and loss of sensation in the foot, which can also lead to neuropathic pain and require additional interventions as it significantly reduces the quality of life [2, 19, 20]. Harner et al. reported one case of permanent nerve palsy in 32 cases of closed knee dislocation [21]. Bozkurt et al. detected neurological injury in 8 patients in a series of 34 cases with rare etiology. In the same series, vascular injury was found in 5 of 34 patients, with no cases resulting in amputation [22]. In our series, the rate of neurological injury was 20.12% (33/164), which is consistent with the literature. However, the specific nerve involved in the injuries could not be identified and analyzed. Our multivariate analysis showed that younger age was independently associated with a higher likelihood of neurological injury. Although this may appear counterintuitive, younger patients are more frequently exposed to high energy trauma mechanisms. In contrast, older individuals more commonly sustain lower energy traumas. In addition to the complications already discussed, other complications associated with traumatic knee dislocation include compartment syndrome and deep vein thrombosis [19]. In our series, vascular complications such as deep vein thrombosis and pulmonary embolism were observed in 9.75% (16/164) of patients. Fasciotomy was required in 5 (3.04%) of these patients.

Fracture-dislocation cases are complex, high-energy injuries with a higher likelihood of complications [8]. Owens et al. reported that 11 out of 28 knees had associated fractures [23]. Another study found that 31% of patients had associated ipsilateral tibia or femur fractures [24]. In our study, no fractures were detected in 85 patients (51.82%). Fractures outside the knee area were detected in 26 patients (15.85%), while fractures around the knee were found in 53 patients (32.32%), which is consistent with the numbers in the literature.

Although this study reflects a nationwide Turkish cohort, the external validity of our findings should also be interpreted in the context of differing trauma mechanisms across countries. High-energy mechanisms such as motor-vehicle collisions remain the predominant cause of knee dislocations in many settings, particularly among young males [25, 26]. However, there are also some studies who have reported an increasing proportion of low-energy knee dislocations in obese or elderly patients, often following minor falls, which may influence neurovascular risk and outcomes [27–29]. Turkey’s younger working population and higher rates of traffic and occupational injuries likely contribute to the mechanism profile observed in our study, whereas countries with differing demographics or obesity prevalence may see different injury patterns. Regardless of country or regional differences in trauma patterns, the risk of neurovascular injury remains a critical and universally relevant concern in the management of knee dislocations.

This study has several limitations. First, its retrospective design inherently introduces the possibility of bias. Second, data extraction was based on ICD diagnostic and procedure codes, which may be affected by miscoding, variability in coding practices, and incomplete documentation across different centers and over the long study period. While all included cases were confirmed, some dislocations may have been missed due to coding inaccuracies. Third, clinical details, particularly the specific nerves involved in neurological injuries, were frequently incomplete, limiting our analysis. Forth, given the relatively small number of neurovascular injury events, our multivariate models may be subject to a reduced events-per-variable ratio, which carries an inherent risk of overfitting despite acceptable model fit indices. Finally, long-term functional outcomes, patient-reported measures, and recovery could not be assessed within the scope of this study.

Conclusion

This study provides valuable insights into the complex nature of traumatic knee dislocations, particularly highlighting the significant risk factors for severe complications. Additional systemic injuries, younger age, dislocation classification, CCI and open injuries were all linked to severe comorbidities. These results emphasize the importance of adopting a systemic approach in the management of knee dislocations, ensuring that all potential complications are carefully considered and addressed to improve patient outcomes.

Abbreviations

ICD

International classification of disease

CCI

Charlson Comorbidity Index

ICU

Intensive Care Unit

EMG

Electromyography

CT

Computed Tomography

MRI

Magnetic Resonance Imaging

Funding

No funding was received for conducting this study

Authors’ contributions

All authors contributed to the study conception and design. Material preparation, data collection, and analysis were performed by Gokhan Ayik, Ulas Can Kolac, and Saygin Kamaci. The first draft of the manuscript was written by Gokhan Ayik and all authors commented on previous versions of the manuscript. All authors read and approved the final manuscript.

Funding

No funding was received for conducting this study.

Data availability

The data that support the findings of this study are available upon reasonable request from the corresponding author.

Declarations

Ethics approval

The study was conducted in accordance with the Declaration of Helsinki and received approval from the Turkish Ministry of Health with a waiver of informed consent for retrospective data analysis and the health information privacy law (ID: 95741342-020/27112019).

Consent for publication

Not Applicable.

Competing interests

The authors have no competing interests to declare that are relevant to the content of this article.