A total of 243 patients (130 female, 113 male) were included in this study; all patients completed the functional outcome and pain scoring measures at a mean follow-up of 63.4 ± 21.3 months (range 26–110), postoperatively. The mean age of the patients at the time of surgery was 54.8 ± 15.4 years (range: 17–88 years), and the mean BMI was 25.2 ± 4.8 kg/m2. As a mechanism of injury, n = 104 patients (42.8%) sustained injuries during activities of daily living, n = 86 (35.4%) at work, n = 32 (13.1%) due to traffic accidents and n = 6 (2.5%) during sports, and n = 15 (6.2%) for other reasons.

As a primary radiological diagnostic, nearly all patients (93.0%) received a preoperative x-ray of the knee, while 29.6% also received a CT scan instead of or in addition to the x-ray. In 99.6% of patients, a postoperative native radiograph was obtained, with 10.7% receiving postoperative CT scans. According to the AO/OTA classification, we analyzed three cases of 34-type A1 fracture (1.2%), 22 cases of 34-type B fracture (B1: n = 12 (4.9%), B2: n = 10 (4.1%)), and 218 cases of 34-type C fracture (C1: n = 67 (27.6%), C2: n = 45 (18.5%), C3: n = 106 (43.6%)). Patients with AO 34-type C3-fractures were distributed equally among those under 65 years (43.3%) and those over 65 years (43.7%). The correlation between the AO/OTA classification and preoperative CT scans indicated that CT scans were exclusively performed for C fractures, depending on the severity of the fracture. Specifically, 4.5% of 34-type C1 fractures, 28.9% of 34-type C2 fractures, and 40.6% of 34-type C3 fractures underwent preoperative CT scans. Additionally, 19.8% of all fractures were classified as open fractures according to Gustilo and Anderson’s classification. A total of 17.7% of patients had associated injuries, including additional fractures or soft tissue injuries of the upper or lower extremities.

Surgery was performed on a mean of 3.0 ± 5.2 days (range: 0–28 days) after trauma. The procedures included TBW in 66.9% of patients, LPO in 19.0%, and SO in 14.1%. FiberWire sutures alone were used exclusively in one patient. Supplemental stabilization material was used in 20.3% of patients: 8.0% received a McLaughlin-Cerclage and 12.3% received additional single screws. A total of 76.1% of 34-type C1 fractures, 64.4% of 34-type C2 fractures, and 68.9% of 34-type C3 fractures underwent TBW. LPO was performed in 7.5% of 34-type C1 fractures, 24.5% of 34-type C2 fractures, and 26.4% of 34-type C3 fractures, while SO was used in 14.9% of 34-type C1 fractures, 11.1% of 34-type C2 fractures, and 4.7% of 34-type C3 fractures.

All patients followed a standardized rehabilitation protocol, which involved either partial or full weight bearing (in extension) while wearing a brace that limited flexion to 30° for the initial 2 weeks, 60° for another 2 weeks, and 90° over the following 2 weeks. Implant removal was performed in 122 patients (50.8%) at 11.3 ± 10.7 months postoperatively for various reasons, such as discomfort caused by the implant or patient’s request (TBW: 56,7%, LPO: 45,6%, SO: 23,5%). Removal was not considered a complication if no material insufficiency or dislocation was found.

Overall, complications occurred in 38 patients (15.6%). Among these, TBW had the highest rate of complications (16.7%) compared to LPO (15.2%) and SO (11.8%). However, these differences were not statistically significant (p = 0.733). The two main causes of complications in this cohort, accounting for 50% of all complications, were atraumatic fragment dislocation and material insufficiency/dislocation. These complications occurred significantly more common after TBW than LPO (p = 0.015) (Fig. 2). In contrast, complications such as limited range of motion, traumatic refracture, and wound healing problems were slightly more common in patients with LPO or SO. Figure 2 provides a detailed distribution of the types of complications and surgical procedures.

Distribution of types of complications and surgical procedures. TBW, tension band wiring; LPO, locking plate osteosynthesis; SO, screw osteosynthesis

In general, patient-specific factors such as age, sex, BMI, regular use of tobacco or alcohol, or osteoporosis were not significant predictors of complications (Table 1). Among fracture-specific factors, distal pole fracture (p = 0.801), open fracture (p = 0.102), and surgical procedure (p = 0.624) did not result in a significantly increased (general) complication rate. However, AO/OTA classification (p = 0.020) and concomitant injuries (p = 0.015) significantly influenced the occurrence of complications. Complications occurred exclusively in AO 34-type C fractures, resulting in a significantly increased complication rate for AO 34-type C fractures compared to B-type fractures (p = 0.033).

However, the analysis of the various types of complications revealed that the occurrence of some of these complications is significantly influenced by patient- and fracture-specific factors, as depicted in Table 2.

Notably, patients with an atraumatic fragment dislocation were significantly older (68 ± 13 years, p = 0.012) than patients with other types of complications (50 ± 19 years). Only 20% of patients underwent a preoperative CT scan, but a total of 70% of the reviewed patients had comminuted 34-type C3 fractures. A total of 90% of patients with an atraumatic fragment dislocation were initially treated surgically with TBW, and one patient (10%) received a LPO. An atraumatic fragment dislocation occurred at an early postoperative stage, around 4–5 weeks after surgery. Almost all complications were revised by implant removal and revision osteosynthesis.

Material insufficiency/dislocation was observed exclusively after TBW, which occurred either in the early (≤ 3 weeks) postoperative period (33.3%) or months/years later (66.7%) (mean time of 403 ± 625 days; overall range of 0–1685 days).

In contrast, limited range of motion was primarily observed in relatively younger (43 ± 21 years) men, which occurred after 140 ± 47 days and appeared to be independent of the surgical procedure. All patients with limited range of motion underwent revision surgery involving implant removal and/or arthroscopic arthrolysis after sufficient bony consolidation of the fracture was achieved.

Traumatic refracture mainly occurred in middle-aged patients (56 ± 19 years), independent of the surgical procedure. This complication primarily manifested within the first year (112 ± 148 days, range: 1–393), either early during (< 2.5 months; 66.7%) or late (> 5 months, 33.3%) after osseous consolidation.

Wound healing problems occurred relatively early after surgery (44 ± 42 days, range 14–92 days) and could mainly be treated conservatively. There were insufficient cases of wound healing problems (n = 4) and material malposition (n = 2) for detailed statistical analysis.

At the final follow-up (63.4 ± 21.3 months after surgery), analysis of pain levels measured by NRS showed significantly higher pain levels at rest after complications (1.3 ± 1.9) compared to patients without complications (0.5 ± 1.4, p = 0.0025) (Fig. 3). NRS values at motion were nearly three times higher. In contrast to patients without complications (1.5 ± 2.2, p = 0.004), patients with complications had significantly higher pain scores at motion (3.4 ± 3.2). Functional knee scores, including Lysholm score (88.4 ± 15.9 vs. 73.4 ± 20.4, p < 0.0001) and IKDC score (78.7 ± 16.8 vs. 62.5 ± 22.1, p = 0.0201), as well as the activity level measured by Tegner Activity Scale (3.9 ± 1.3 vs. 3.1 ± 1.3, 0.0021) were also significantly lower after complications.

Pain levels (NRS at rest and at motion), functional knee scores (Lysholm and IKDC score), and activity level (Tegner Activity Scale) of patients with and without complications (* p < 0.05)

The analysis of the outcome parameters for the different types of complications revealed that patients with complications of limited range of motion or traumatic refracture experienced significantly higher pain levels (at rest: p = 0.006; at motion: p = 0.001) and lower functional scores compared to implant-related complications such as atraumatic fragment dislocation and material insufficiency/dislocation. Although the final follow-up was 45 ± 13 months after the occurrence of complications and 62 ± 33 months after the initial surgery, patients reported higher pain scores. In addition, implant-related complications had significantly higher functional scores (Lysholm: p = 0.001 and IKDC: p < 0.001) and activity scores (Tegner: p = 0.049) compared to other types of complications (Table 3). Furthermore, no significant differences in functional knee score values were found between patients with atraumatic fragment dislocation and material insufficiency/dislocation and those without complications (Lysholm: p = 0.206; IKDC: p = 0.555; Tegner: p = 0.224).