Influence of intraoperative reduction quality on functional outcome and quality of life in treatment of tibial plafond fractures: a retrospective case-control study

The aim of the study was to evaluate the impact of reduction quality, using intraoperative 3D imaging, on quality of life and functional outcome in the operative treatment of tibial plafond fractures. A group of patients with tibial plafond fractures was postoperatively examined. The operative treatment was performed between September 2001 and October 2011 under reduction control using an intraoperative 3D C-arm. A categorization with regard to the type and the size of joint surface irregularities was carried out after intraoperative reduction. Postoperative results were evaluated using: Olerud and Molander (O & M) score, Short-Form-36 (SF-36) score, leg circumference, movement deficit, Kellgren and Lawrence grade of osteoarthritis, and pain intensity.

with intraoperative 3D imaging -plays the most important role in postoperative quality of life and functional outcome.

Background
Tibial plafond fractures occur in approximately 5 per 100,000 people and account for about 5 to 7% of all tibial fractures. [1] More than 30% of all tibial plafond fractures are associated with high-velocity trauma, which makes operative treatment challenging due to the complex fragment dislocation and severe soft tissue damage. [2] An anatomically incorrect reduction, in the sense of axial deviation or graduation of the joint surface, leads to a relevant functional limitation of joint movement and premature arthrosis. [3][4][5][6] Therefore, operative intervention with anatomical reconstruction of the joint structures is often indispensable to achieve a satisfactory clinical outcome. [5,[7][8][9] The intraoperative assessment of the articular surface and implant placement with conventional fluoroscopy is demanding. Studies using the cadaver model have shown that, even under optimal conditions, the analysis of the joint surface and implant placement using conventional fluoroscopy may not be sufficient.[10-12] The current gold standard for preoperative planning and postoperative assessment of reduction quality and implant placement is computed tomography (CT), which is not regularly available for intraoperative evaluations. [13] Postoperative complications in tibial plafond fractures are well known and described in detail in the literature. [14,15] Postoperative detection of a relevant fragment dislocation or implant misplacement therefore usually leads to revision surgery. A reliable intraoperative examination regarding the quality of reduction should make it possible to recognize and correct a malalignment already during the present operation.
Intraoperative 3D imaging using a mobile C-arm can be used to assess the reduction result and implant placement to identify intraoperative conditions requiring correction. [16][17][18][19][20][21] Several studies have already demonstrated that the use of intraoperative 3D imaging may lead to a relevant intraoperative revision rate between 14.6 and 36% of the cases, despite the lack of evidence of malreduction or implant misplacement in conventional fluoroscopy.
[18 -20, 22, 23] Furthermore, only a few studies exist that have investigated the functional outcome and health-related quality of life after operations due to tibial plafond fractures, without referring to the quality of reduction. [24,25] A study investigating the postoperative outcome of tibial plafond fractures, taking into account the reduction quality using intraoperative cone-beam CT, has not yet been conducted.
The aim of the study was to investigate the influence of reduction quality in the operative treatment of tibial plate fractures, depending on type and size of the joint surface irregularity, using intraoperative 3D imaging, on quality of life and functional outcome.

Methods
In the scope of a retrospective, monocentric study, a group of patients with tibial plafond fractures, classified as AO/OTA type B and C according to the preoperative CT data, was postoperatively examined. The operative treatment was performed between September 2001 and October 2011 under reduction control using an intraoperative 3D C-arm (cone beam CT) (Siremobil-Iso-C-3D, Arcadis-Orbic-3D; Siemens Healthcare GmbH, Erlangen, Germany). The surgical interventions were performed by experienced surgeons from a level I trauma center. The therapeutic procedure included immobilization with a cast or the application of an external fixator, especially in cases of poor soft tissue conditions. The operative treatment included open reduction and internal fixation followed by an intraoperative cone beam CT to analyze reduction quality. If the reduction or implant placement was unsatisfactory, an intraoperative revision was performed and the cone beam CT scan repeated ( Figure 1). Only the final 3D image data set after the final surgical procedure was included in the study. The overall rate of intraoperative revision of the reduction or implant placement -based on intraoperative cone beam CT imaging -was 29.2% in total. The follow-up period was at least 2 years. The following exclusion criteria were also applied: concomitant injuries of the same extremity, spinal injuries with neurological symptoms, polytrauma with craniocerebral trauma higher than grade I, preexisting primary and secondary osteoarthritis of the ankle joint and previously suffered injuries of the same anatomical region (e.g. ankle fractures), postoperative complications (infection, thrombosis, compartment syndrome, flap plastic, revision in external clinics, wound healing disorder, bleeding, necrosis or amputation), and patients who had already died.
The collective was differentiated according to the following parameters: Age, gender, BMI, concomitant diseases, profession, type of accident (private/work-related), fractured side, type of fracture (type B/C), and concomitant injuries.
Furthermore, patients were retrospectively categorized into two groups regarding the reduction quality. The evaluation of the joint surface was carried out conducting a dynamic inspection of the complete 3D data set in all 3 planes. The first group was defined as reduction results with articular surface incongruencies (steps, gaps or defects) of less than or equal to 2 mm in the scan images. The second group included all patients whose incongruencies exceeded 2 mm in size of step, gap or defect. According to this classification, 15 patients were placed in Group I and 19 patients in Group II. From the SF-36 score, several scores can be derived, which are assigned to specific categories. In this case, all four domains of the score (Physical Functioning, Role Physical, Bodily Pain and General Health) were used, which mainly cover physical aspects of health, as well as a final component (Physical Component Summary) summarizing them once again. In addition, the Mental Component Summary (MCS) was analyzed, a comprehensive domain that summarizes the mental areas of health. The two overarching scores serve as an overview.
The severity of osteoarthritis of the ankle joint was determined using the radiographic classification by Kellgren and Lawrence.
The range of motion of both ankle joints was measured with a goniometer applying the neutral-zero method and subsequently compared to the healthy contralateral side by forming differences to determine any deficit. This was calculated separately for flexion and extension, and these values were added together to give an overall value.
Furthermore, the circumferences of both legs of each patient were measured. This was also performed in a side-by-side comparison and differences were again formed. The circumference of the healthy leg was subtracted from the circumference of the injured leg in order to identify a possible atrophy of the lower leg muscles or a swelling of the ankle joint. Measurements were taken at three different locations: 15 cm below the medial knee joint gap, at the narrowest point of the lower leg, and at ankle level.  Among the follow-up subjects with at least one risk factor, 5 were obese, 9 consumed nicotine, 12 consumed alcohol, 3 were diagnosed with diabetes mellitus, 1 had gout and 1 had hyperthyroidism.
The professions of patients were divided into four categories, depending on the level of physical activity: 4 pensioners/students/unemployed, 11 with sedentary work, 11 with basic physical work and 8 with heavy physical work.
In the follow-up collective, referring to the AO/OTA Classification, 20 of the subjects had type B fractures and 14 suffered from type C fractures. In Group I there were 11 type B fractures and 4 type C fractures. In Group II there were 9 type B fractures and 10 type C fractures.
During the follow-up, patients completed two questionnaires (the Olerud and Molander score and the SF-36 score): On average, the patients surveyed scored 69.12 points (SD: 24.79, range: 10-100) in the Olerud and Molander score. The comparison of both groups according to the reduction quality is shown in Table 1    Descriptive statistics on the circumferential differences 15 cm below the medial knee joint gap, at the narrowest point of the lower leg and at ankle level within the two groups are summarized in Table 3.
The illustration of the group-specific results for the visual analogue scale is shown in Table 4.

Group-specific analyses:
With the numbers available, no significant differences could be found with regard to age (P = 0.836), sex (P = 0.231), BMI (P = 0.151) or type of fracture (P = 0.127) in the groups differentiated according to reduction quality. Significant distribution differences were observed with regard to nicotine abuse (P = 0.002), profession with heavy physical work (P = 0.014) and concomitant injuries (P = 0.004), whereby these were predominantly found in the suboptimal reduction group (Group II).
Considering the number of patients available, the Olerud and Molander score could be significantly influenced by the reduction quality (P = 0.000): The mean difference was 33.79 points with a standard error difference of 6.32. The 95% confidence interval was 20.92 to 46.66 points.
Significant association of SF-36 score with reduction quality could also be observed (P = 0.001 to P = 0.02; without MCS domain): In the comparison of the PCS domain, the mean difference amounted to 10.24 points (P = 0.003). The standard error difference amounted to 3.25 points. The range of the 95% confidence interval was between 3.63 and 16.86 points. There were no significant differences with regard to the MCS domain of the SF-36 score (P = 0.142), when considering reduction quality.
Only the circumferential difference 15 cm below the knee joint gap showed a significant deviation within the two groups (P = 0.012): The mean difference was 1.29 cm with a standard error difference of 0.48 cm. The range of the 95% confidence interval was between 0.31 and 2.28 cm. The remaining circumferential differences were not significant (P = 0.729; P = 0.278).
Significant differences of movement deficit in comparison of reduction quality (P = 0.001): The mean ranks of the good reduction group were lower (11.50°) than the mean ranks of Analysis of variance (ANOVA): Table 5 lists the descriptive statistics regarding the Olerud and Molander scores depending on the type of articular surface irregularity. It was found that only the group of patients with steps differed significantly from those with the combination of gaps and defects (P = 0.034). All other groups did not provide significant differences regarding the comparison of their mean values in the Olerud and Molander scores.
Furthermore, it was examined whether there was a significant difference between the groups regarding their mean value in the SF-36 score domains. However, none of the components of the SF-36 score showed a significant result with respect to the comparison of the mean values of the different articular surface irregularity groups.
Thereafter, it was analyzed whether there were significant differences between the specific groups of joint surface irregularities in terms of circumference differences. The ANOVA analysis showed no significant differences. The width of the gaps ranged from 0 to 8.3 mm (SD: 1.74), the range of defects was 0 to 9 mm (SD: 2.7) and the steps varied from 0 to 4.7 mm (1.27).
The correlations between the Olerud and Molander Score and the step, gap and defect sizes revealed the results listed in Table 6 Furthermore, it was examined whether there was a significant correlation between the circumferential differences and the step, gap and defect sizes. However, this could not be confirmed by the Pearson correlation analysis. No significant correlations were found here.
The visual analogue scale did not correlate significantly with either the step size or the gap size. The defect size, however, showed a significant result (P = 0.012). The Pearson coefficient was positive (r = 0.425). Thus, it could be concluded that larger defects were associated with higher values on the VAS.

Multivariate linear regression analyses:
In this study, the reduction quality had the greatest influence on the functional result after operatively treated tibial plafond fracture determined by the Olerud and Molander score (P = 0.001) and the PCS domain of the SF-36 score (P = 0.018).

Discussion
The operative treatment of intra-articular tibial plafond fractures remains difficult even for the experienced trauma surgeon, since the intraoperative assessment of the tibial joint surface and the implant placement using conventional fluoroscopy is limited. Previous studies focused on whether an anatomically correct reduction of the distal tibial joint surface in tibial plafond fractures results in a prognostic difference in patient outcomes and concluded that remaining joint gaps or steps of more than 2 mm after the reduction and axial deviations in the frontal or sagittal plane of more than 5 degrees can lead to poorer clinical results and higher osteoarthritis rates. [5,[31][32][33][34][35]   Smoking as a risk factor correlated significantly with the Olerud and Molander score.
Nicotine consumption led to a lower score. The negative effect of nicotine consumption on osteogenesis and fracture healing has already been well demonstrated in vitro and in vivo test series.[36, 37] Furthermore, a recent study also shows that nicotine consumption has an influence on pain perception, so that smokers are dependent on significantly more analgesics postoperatively.
[38] Other studies reported smoking as a predictive factor for musculoskeletal complaints, defined as having pain and/or stiffness in muscles and joints. [39] The study had several limitations. The absolute number of 34 participants was quite low and therefore allows only a limited statement about the overall population. Furthermore, it was also disadvantageous in terms of the statistical evaluation of some results due to the resulting high range. Nevertheless, in view of the low incidence of the type of injury, the fact that type B and C fractures are very rare, and the long follow-up period, the number of patients examined compared to other studies is actually very high.
Two different devices have been used over the years to perform the 3D scans. However, the image quality of the two devices is comparable in terms of diagnostic evaluation and surgical analysis.

Conclusions
In conclusion, the established reduction criteria in intraoperative 3D imaging appear to have the highest impact on postoperative quality of life and functional outcome. This is despite other relevant factors such as nicotine consumption, concomitant injuries or gender, as well as the type of accident.
Furthermore, it is not always the type of joint surface irregularity that is decisive, but rather the size. This should be taken into account in the reduction analysis and corrected if necessary, especially if the surface irregularity is above 2 mm. 37.Fuegener S, Hoff P, Lang A, Gaber T, Rakow A, Simon P, Burmester G-R, Perka C,       Figure 1 Standardized workflow for the application of intraoperative 3D imaging for the assessment of reduction quality in trauma surgery visualized as flow chart.