Registry for Evaluation of Lumbar Arthrodesis Sagittal alignmEnt: Prospective Observational Study of the Relationship Between Pelvic Parameters, Lumbar Lordosis Mismatch-pelvic Incidence, and Lordosis Ratio vs. Post-surgical Clinical Outcome for Degenerative Disease
Registry for Evaluation of Lumbar Arthrodesis Sagittal alignmEnt: Prospective Observational Study of the Relationship Between Pelvic Parameters, Lumbar Lordosis Mismatch-pelvic Incidence, and Lordosis Ratio vs. Post-surgical Clinical Outcome for Degenerative Disease
The study goal of RELApSE is to evaluate the relationships between radiological data and patients reported outcome. Restoration of Sagittal alignment and Pelvic Index (PI)-Lumbar Lordosis (LL) mismatch is closely associated with a better outcome in spinal deformities, while there is still a lack of consistent evidence regarding short-segment arthrodesis for lumbar degenerative pathology.
Over the past 10 years, an increasing number of publications reported associations between the presence of PI-LL mismatch, reduced lumbar lordosis, increased pelvic tilt, and outcome of lumbar arthrodesis for degenerative lumbar disease. Other authors, on the other hand, reported an absence of correlation between the same parameters and clinical outcome. In addition, several authors have been reported evidence regarding association of adjacent level disc degeneration and elevated pelvic tilt, persistent PI-LL mismatch and altered LL4-S1/LL ratio. Also on this aspect, other studies identify different elements as predisposing factors for junctional pathology. The definitive value for lumbar degenerative pathology of these aspects in relation to the surgical outcome remains to be clarified without consolidated evidence. The RELApSE study is the first prospective and multicenter study on these topics.
Starting from a very heterogeneous population in terms of clinical conditions, pathology and surgical treatment options, the study methods is to make the population homogeneous on some data available in all patients end that can be analyzed independently. These data are: pelvic parameters (pelvic incidence, pelvic tilt, sacral slope), segmental lumbar lordosis (LS), global lumbar lordosis (LL), PI-LL mismatch and L4-S1/LL lordosis ratio; clinical results based on administered questionnaires (Oswestry disability index, Short Form-12) and overall outcome assessment at FU (6 point scale: excellent (completely resolved symptoms), good (good clinical improvement, minor symptoms), fair (improvement compared to preoperative but still with relevant symptoms), unchanged (symptoms similar to preoperative), negative (worsening of symptoms compared to preoperative); severely worsened (reduction of personal autonomy compared to preoperative due to neurological deficits); occurrence of symptomatic junctional pathology (yes / no), need for surgical revision of the operated level (yes / no) or of the adjacent level (yes / no).No interference is foreseen on the patient's diagnostic-therapeutic path or technical treatment options chosen by partecipating surgeons. Furthermore, no form of experimentation with techniques or materials is envisaged. Data collection is prospective in the context of normal clinical activity.
BACKGROUND AND STUDY RATIONAL
Degenerative pathology of the lumbar spine is very frequent affecting approximately 5.7% of the European population. Arthrodesis is a frequently adopted surgical treatment with different technical option, accounting for 30,000 procedures performed every year in Italy (trovare riferimento bibliografico) and over 450,000 procedures performed in the United States. However, lumbar vertebral arthrodesis for degenerative disease still remains burdened by clinical problems with significant occurrence of negative results consisting in lack of clinical improvement, late symptoms relapse or even clinical worsening. The evidence based medicine on this topics is still insufficient and generally does not exceed evidence class B.
The available guidelines are mainly based on level II studies and only offer recommendations that may support practitioners in clinical activity. Thus, surgical treatment of lumbar degenerative disease still remains extremely heterogeneous also considering the number of technical options available for the single pathology. Lack of standard of care leads to treatment stratetegies based on institutional, departmental, or personal experience, planned on inconsistent scientific evidences. The introduction in the clinical practice of the sagittal balance assessment, the evaluation of the meaning of pelvic parameters and of spino-pelvic mismatch has also offered a new evaluation criterion for lumbar degenerative pathology and for outcome of short lumbar arthrodesis surgery. Restoration of Sagittal alignment and Pelvic Index (PI)-Lumbar Lordosis (LL) mismatch is closely associated with a better outcome in spinal deformities, while there is still a lack of consistent evidence regarding short-segment arthrodesis for lumbar degenerative pathology.
Over the past 10 years, an increasing number of articles reported about associations between the presence of PI-LL mismatch, lumbar lordosis, pelvic tilt and outcome of lumbar arthrodesis for degenerative lumbar disease or occurrence of adjacent segment disease (ASD. A recent systematic meta-analysis highlighted a strong relationship between LBP and reduced lumbar lordotic curve especially when petients were analyzed with age-matched healthy controls. In 2000 Lazennec et al firstly analyzed the relationship between radiological parameters and postfusion pain focusing attention on the vertical position of the sacrum. Patients with pain persisting after arthrodesis showed both reduced SS and increased PT with PT reaching almost twice the normal value. Authors postulated that " achieving a strong fusion should not be the only goal. Appropriate position of the fused vertebrae is also of paramount importance to minimize muscle work during posture maintenance". An increased LL and SL was associated with better outcome (VAS, ODI) after unilateral instrumented TLIF single-level lumbar degenerative disease and after 1 level or 2-level PLIF for 2-level (L3-L4 and L4-L5) degenerative spondylolisthesis. Hioki et al described a positive linear correlation between increase in lordotic angle and improvement of JOA score at final outcome evaluation after two level PLIF.
Furthermore, reduction of pelvic tilt was associated with better outcome after PLIF surgery for spondylolisthesis as well as sacral slope increased to more than 30° after single level TLIF. The review by Le Huec et al. suggested that the increase in PT after surgery is associated with significant low back pain while the restoration of a normal PT results in a good clinical outcome. The study of Aoki et al was the first investigating influence of PI-LL mismatch on postoperative residual symtoms after 1 or 2 level TLIF. Larger PI-LL mismatch was significantly associated at 1 year follow-up with VAS for LBP, leg pain and leg numbness but not with postoperative disability (ODI). Detailed VAS analysis highlights association of mismatch with stending low back pain but not with LBP while sitting or in motion. Worse ODI scores were associate at 2 years follow-up with PI-LL mismatch after four-level lumbar fusion surgery (L2-S1); moreover, a significant correlations between PI-LL mismatch and improvement in both JOA score-VAS for LBP at 2 year follow up has been described for patients with lumbar degenerative scoliosis treated with short segment fusion (1, 2 or 3 level TLIF) at the affected levels. Radovanovic et al reported a better patient reported outcome in patients with SVA less than 50 mm. Patients with an SVA ≥ 50 mm presented reduced lumbar lordosis with increased mismatch and had a worse SF-36 PCS and Oswestry Disability Index (ODI; p = 0.043) as well as more back pain.
Better improvement of PI-LL mismatch with reduced PT and higher LL was found in patients without residual back pain after OLIF procedure and similarly a significant linear association with Oswestry Disability Index was reported for independent variables LL, delta LL and PI-LL status after OLIF for degenerative disc disease.
A cutoff value of 27,5 ° for preoperative PI-LL mismatch is reported as a negative factors for outcome of patients underwent second PLIF surgery for ASD and as predisposing factor for need of subsequent long corrective surgery.
Finally, sagittal malalignment with PI-LL mismatch greater than 10° was also associated with occurrence of pseudarthrosis.
On the contrary, HSU et al reported no correlation between LL or LL restoration ratio and the outcome of patients undergoing PLIF for degenerative spondylolisthesis while JIA et al concluded that PI-LL mismatch is not associated with clinical outcome (VAS, ODI) after MIS-TLIF for lumbar stenosis (25). A comparative study between ALIF and PLIF also showed no relationship between LL and patients outcome. A 2017 systematic review by Rhee et al. select only 4 article for final statistical analysis and point out the lack of well-powered studies on this topic. No statistically significant improvement in both ODI or VAS was related to restoration of segmental lordosis. So correction of malalignment does not seem to yield clinical improvements for lumbar short arthrodesis.
A recent large retrospective study show that patient outcomes in short-segment lumbar fusion for degenerative lumbar disease are equivalent in patients with and without a postoperative PI-LL mismatch at 1 year follow-up. No differences between groups in preoperative, postoperative or delta outcome scores (PCS-12, ODI, VAS back, VAS leg) were noted. PI-LL mismatch was not found to be an independent predictor for patient reported outcome on multivariate analysis (P> 0.05). This study suggest that limited surgery adressed to focal neurological disease has equivalent outcomes of corrective surgery.
Adjacent segment degeneration (ASD) after lumbar arthrodesis, both clinical and radiological, is a well known problem. Many studies in the literature discuss about this topics and several aspects were analyzed as causes or risk factors. Prospective study by Ekman shows that fusion accelerates occurrence of degenerative discopathy at the adjacent level compared with natural history. Reported incidence have a wide range between 2,62 and 84% with prevalence of proximal level and main associated factors were old age, body mass index (BMI), previous degenerative disc o facet disease, type of pathology, multiple-level fusio, male, intraoperative superior facet joint violation, laminectomy, sagittally oriented facet joint angle, PLIF and progressive fatty degeneration of the multifidus muscle. In 2001 Kumar et al firstly reported significant association of ASD and C7 sagittal plumb line. A vertical sacrum was highly associated also with ASD even with normal C7 plumb line. Sacral inclination was considered an important aspect of sagittal alignment as expression of compensation mechanism. Several authors reported a correlation between ASD and reduced lumbar lordosis as a major independent factor or in association with other factors. Also reduced postoperative segmental lordosis was associated with occurrence of ASD. Soh et al and Bae et al. suggested that the most important factor for prevention of ASD is restoration of segmental lordosis. Kim KH et al described association between reduced segmental lordotic angle and symptomatic ASD in isolated L4-L5 spondylolisthesis treated with interbody fusion and pedicle screw fixations. Again the pre-postoperative segmental lordosis delta was reported as a major risk factor for ASD especially with early onset (50,76).
Relationship between pelvic parameters and ASD was also studied. Nakashima et al. identified high degree of pelvic incidence as a risk factor for early onset radiological ASD probably in relation to the reduced probability of obtaining appropriate lordosis after surgery. This association was also confirmed by a recent meta-analysis.
Some authors found a significant reduction of sacral slope angle in patients with occurrence of ASD and similarly Di Martino et al reported in ASD patients significantly lower SS and consequent higher PT values related to pelvic retroversion and hyperlordosis compensation mechanism. In this study Authors defined SS value below 39° or PT above 21° as a strong risk factor for occurrence of symptomatic ASD (relative risk 1.73 for SS and 3.663 for PT). Another study confirmed that a PT greater than 24.1 ° can be considered predictive of ASD after lumbar TLIF (77). Also preoperative PT with 22.5° cutoff was strongly associated with ASD (risk 5.1 greater).
A meta-analysis by Phan et al conclude that development of ASD may be predicted from evaluation of Spinopelvic alignment paremeters (PT, SS, PI-LL mismatch, and LL) in patients with lumbar fusion for degenerative disease. An elevated PI-LL mismatch havee been reported closely associated with development of symptomatic ASD or radiological ASD. A PI-LL mismatch greater than 15° was identified as significant independent risk factors for radiographical ASD also in patients with L5-S1 spondylolytic spondylolisthesis treated with single level PLIF.
Morover Wang et al reported strong association between symptomatic ASD after lumbar fusion and greater PI-LL mismatch but identified different PI-LL mismatch cutoff in patiens below 60 years (PI-LL >10°) and older patients (PI-LL > 20°) to reach statistical significance. The authors hypothesize that the ideal correction of LL may vary with increasing age. Rothenfluh et al reported 10-times higher risk of ASD occurrence for patients with elevated PI-LL mismatch (> 10°). Patients with ASD have higher PI, higher PT and lower lumbar lordosis. Authors conclude that when fusion surgery is carried out without treatment of intrinsic deformity and PI-LL mismatch patients present elevated risk for ASD.
Finally lumbar distribution index (LDI=L4-S1 lordosis/lumbar lordosis x 100) was strongly associated with occurrence of ASD: patient with reduced distal L4-S1 lordosis and consequent low LDI present greater risk of developing ASD (36). Kim et al reported frequent occurrence of ASD in patients with LDI less than 50% also when PI-LL was satisfactorily corrected to less than 10°. Obtaining an appropriate postoperative LDI in L4-S1 may have a crucial role in the prevention of ASD.
In both clinical analysis and mechanistic simulation environment the increased loading and biomechanical shear forces at fusion adjiacent level have been postulated and discussed for patinets with PI-LL mismatch. These experimental data offer a further element in the evaluation of association between ASD occurrence and sagittal malalignment.
On the contrary, some authors have highlighted the absence of a relationship between ASD and sagittal alignment. Anandjiwala et al reported that LL is not a risk factor for ASD occurrence while Chen et al reported no differences in lumbar lordosis between patients presenting ASD and those who did not. Masevin et al analyzing the risk factors for ASD reported the absence of a role of the sagittal balance in short fixations for which the only risk factor is preoperative degenerative changes. A meta-analysis by Wang et al, based on 19 papers, showed that postoperative PT and SS are not associated with ASD occurrence. Multivariate analyses showed that segment distraction was the most significant risk factor after L4-L5 PLIF or only multilevel surgery associated with high rate reoperation. Finally a recent study on the prognostic factors of ASD after L4-L5 fusion does not consider sagittal alignment at all, suggesting that these aspects are still underestimated by many surgeons.
The analysis of these literature data shows sparse and variable evidence and highlights how clarification and greater understanding on this argument is still needed, evoking the need for methodologically correct and high-level studies. In fact, most of the works in the literature are retrospective, generally monocentric, based on small populations and often address partially the topic evaluation. Certainly, the generation of standards based on scientific evidence remains very difficult for degenerative pathology of the lumbar spine. However, the evaluation of data collected on a sufficiently large population in a prospective, uniform and methodologically correct manner could allow to highlight and underline some associations. This prospective and multicentric study is based on the possibility of comparing an heterogeneous population by pathology and different surgical technical options on some homogeneous clinical and anatomo-radiological measures. The data analysis will contribute to the understanding of the value that global lumbar and segmental lordosis, distribution of lordosis, pelvic tilt and PI-LL mismatch may have, as independent factors, on clinical outcome in lumbar degenerative pathology and on occurrence of adjacent segment disease. Consequently, it will focus and enhance at least some rational aspects of lumbar arthrodesis such as, in particular, the need to adopt surgical strategies aimed to restoring segmental lordosis and correcting the sagittal profile. Carrying on the follow-up for several years, the study will finally also provide information on long-term evolution and in particular on the occurrence of symptomatic adjacent level degeneration.
STUDY DESIGN
3.1) Type of study
RELApSE is a purely observational study. No interference is foreseen on the patient's diagnostic-therapeutic path or on the technical treatment options of the chosen by the single surgeon participating in the data collection. Furthermore, no form of experimentation with techniques or materials is envisaged. Data collection is prospective in the context of normal clinical activity. The study is not sponsored.
The study goal is to evaluate the relationships between radiological data and patients reported outcome. The study method is to make the population comparable and homogeneous on some data (and related pre-postoperative delta) Starting from a very heterogeneous population in terms of clinical conditions, pathology and surgical treatment options, the study methods is to make the population homogeneous on some data available in all patients end that can be analyzed independently. These data are: pelvic parameters (pelvic incidence, pelvic tilt, sacral slope), segmental lumbar lordosis (LS), global lumbar lordosis (LL), PI-LL mismatch and L4-S1/LL lordosis ratio; clinical results based on administered questionnaires (Oswestry disability index, Short Form-12) and overall outcome assessment at FU (6 point scale: excellent (completely resolved symptoms), good (good clinical improvement, minor symptoms), fair (improvement compared to preoperative but still with relevant symptoms), unchanged (symptoms similar to preoperative), negative (worsening of symptoms compared to preoperative); much worsened (reduction of personal autonomy compared to preoperative due to neurological deficits); occurrence of symptomatic junctional pathology (yes / no), need for surgical revision of the operated level (yes / no) or of the adjacent level (yes / no).
3.2) Study endpoints
Primary end-point: analysis of the relationship between clinical results (ODI, SF-12, global outcome) and pelvic parameters, overall lumbar / segmental postoperative lordosis / ratio L4-S1-overall lumbar lordosis (delta pre-postoperative-follow-up values) in the hypothesis that the persistence of elevated PI-LL mismatch, reduced LL or altered pelvic parameters (pelvic tilt) may represent independent negative prognostic factors for patients reported outcome, occurrence of symptomatic junctional pathology and need for surgical revision.
Secondary end-points: 1) Comparison between different interbody fusion techniques (Anterior Lumbar Interbody Fusion (ALIF) vs Posterior Lumbar Interbody Fusion (PLIF) vs Transforaminal Lumbar Interbody Fusion (TLIF) vs Extreme lateral Lumbar Interbody Fusion (XLIF)) in terms of increasing segmental lordosis; 2) Comparison between different interbody fusion techniques (ALIF vs PLIF vs TLIF vs XLIF) in terms of changing the global lumbar lordosis, spino-pelvic parameters and PI-LL mismatch; 3) Long-term evaluation of lumbar lordosis stability in relation with occurrence of subsidence or pedicular screws failure.
3.3) Sample size e statistical considerations
The sample size is assessed on in relation to the primary objective. We suppose we want to compare the variation in clinical parameters before and after the intervention (ODI, SF-12: MCS, PCS, global outcome) between the groups finally identified by the presence / absence of LL-PI mismatch. In the hypothesis of no difference, the data necessary for the calculation can be referred to as reported by Divi et al (ref). As regards to MCS-12, the study reports a delta equal to 4.1 in subjects without mismatch and equal to 4.1 in subjects with mismatch, consequently no difference is expected and this parameter is not included in the calculation of sample size. In relation to ODI, the study reports a delta equal to -22.8 in subjects without mismatch and equal to -20.9 in subjects with mismatch (hypothesized sd equal to 5). For PCS-12, the study reports a delta equal to 9 in subjects without mismatch and equal to 10 in subjects with mismatch (hypothesized sd equal to 2). With a significance of 0.05, a test power of 0.95, assuming to compare the two samples using a two-tailed t test, a minimum sample size of 362 subjects is required. Since the study is multicentric, the enrollment of a total of 500 patients is considered to be an exhaustive objective. The evaluation of the secondary end-points, which is not essential, will be carried out on an the same population sample.
4) PROCEDURE
4.1) Inclusion criteria
All patients with the following requirements can be included in the observational study
4.2) Exclusion or Withdrowal criteria
Patients who, although already enrolled, do not meet the inclusion criteria will be excluded. Patients who, regardless of age, have a life expectancy of less than 5 years due to associated diseases, must be excluded from the study. Patients who, for various reasons, have to withdraw their consent to be included in the study in writing will also be excluded. There are no other exclusion criteria.
4.3) Informed Consent
The patient will be asked to express informed consent, in coded and anonymous form, for acceptance of study inclusion. At the same time, consent will be acquired for the processing of personal data in relation to the current italian privacy law. Consents must be acquired and preseved from Surgeon performing artrhodesis. Finally, the patient's consent will be acquired also to be contacted, again anonymously, by the Referring Doctor together with independent examiners for a remote telephone follow-up. The original informed consent forms containing the patient's personal details and the identification code assigned must be kept, for privacy reasons, only by the Referring Doctor. The patient will be given a copy of the consents containing his / her study identification code with which he will eventually be contacted by phone anonymously. the patient will also receive an information form with the aims and characteristics of the study.
4.4) Data collection
For the recruited patients, all the data required (preoperative, postoperative, follow-up) in the data collection form (CRF) must be entered together with the patient code:
All images must be centralized for measurements anonymously and coded as reported below. (4.5). All radiological data, measured as angles or derived numbers, will be calculated by independent data managers recruited in the form of volunteers among medical specialists or fellows (see Study Board). The calculation of the angle values will be done by performing three successive measurements on the same radiological image by 3 independent examiners not involved in the surgical management. Subsequent calculation of the average of the 9 values obtained. The calculation of pelvic angles and parameters can be done with freewere SurgimapÒ software developed by a group of vertebral surgeons and engineers to support anatomical evaluation and surgical planning (Nemaris Inc.TM innovation, New York, NY, USA). The software can be downloaded freely from the website www.surgimap.com where all the policies and conditions of use are visible.
4.5) Privacy protection
In order to ensure patient privacy, the data will be transmitted completely anonymously. Patients will thus be identified by a code consisting of the investigator's reference and a progressive number (eg. FIRENZE001, SIENA002, BERGAMO003 etc). The identification code will be attributed to the individual investigator at the time of joining the study and reported in the letter of intent. The list of investigator codes is shown in chapter 1.2 of this document. Likewise, the transmission of radiological images will be performed through private images of the patient's data in jpg or dcm format. The files will be given a name consisting of the patient's code followed by the type of exam (eg FIRENZE001RXpre).
4.6) Follow-up
All patients recruited will have to be followed up, unless the patient withdraws consent explicitly, for the entire duration of the study initially planned for 5 years. The follow-up evaluations must include the acquisition of all clinical data indicated in the data collection form (CRF-FU) including ODI and SF-12 evaluation questionnaires (results to be reported in the CRF-FU). These data together with the patient code are: months of follow-up, clinical outcomes, presence of radicular pain, presence of claudication, presence of neurological deficits, ODI, SF-12 MCS, SF-12 PCS, technical errors, complications, hardware faillure, need for reoperation, occurrence of junctional pathology, intervention-junctional pathology time.
The follow-up includes postoperative clinical evaluation 3 months after surgery, 1 year and up to 5 years annually. Any additional evaluation will be dictated only by clinical needs and must be added to the follow-up with related time of occurrence. Follow-up is the responsibility of single investigator. A telephone assessment may be performed on patients who are not available for outpatient assessments or who cannot be reached by completing ODI and SF-12 by interview.
The follow-up must also include the acquisition of a lumbar spine standing X-ray (methods indicated in point 4.1, 4.4) at 3 months and 12 months as normally suggested by good clinical practice. According to the literature, the implant settlement and cages subsidence occur mainly in the first 12 months (39,40). Further or subsequent radiological examinations will be eventually and exclusively performed for clinical needs (persistent symptoms, clinical worsening)
4.7) Statistical analysis
Quantitative variables will be expressed using means and standard deviations (medians and quartiles when appropriate). Qualitative data will be expressed as a raw number and a percentage of the population. Comparison of proportions were performed with Chi-squared test for categorical variables. Continuous variables will be compared using the Student t-Test and ANOVA (or non-parametric analogue). The correlation between numerical variables will be evaluated through Pearson's correlation coefficient and relative test (or similar Spearman coefficient and test, if appropriate). Any linear or non-linear logistic regressions will be evaluated for the relationship between lumbar or segmental lordosis, pelvic parameters and clinical data. Further statistical analyzes may be evaluated during the data evaluation. Depending on the results provided by the analyzes described, it may be appropriate to set up mixed effects models to evaluate the longitudinal trend of the study parameters. The possible implementation of non-parametric models of supervised machine learning (e.g. random forest, SVM) will be evaluated to verify the actual predictive power of the variables under consideration with respect to the clinical evaluation. Statistical significance was defined with a p-value < 0.05.
Due to the nature of the multicentre study and the high sample size, it is planned to create an interactive report (dashboard) for viewing the statistical results obtained.
5) METHOD OF IDENTIFICATION OF PARTICIPATING SURGEONS
The participants were identified in the Italian context among orthopedists or neurosurgeons with proven experience in the field of vertebral surgery by direct invitation from one of the members of the steering committee.
Surgeons who have chosen to participate in the study are required to sign a letter of intent to underline the commitment required and the roles of the investigator. Each participating investigator is required to obtain approval from their relevant ethics committee.
6) METHOD OF PATIENTS RECRUITING
Investigators will recruit patients in the context of their regular clinical and surgical activity. Recruitment must be prospective and the patient must be enrolled prior to surgery in order to avoid selection bias related to clinical result. Each participating surgeon will be required to recruit A minimum of 20 complete and evaluable patients is requested to be included in the investigators while the number of patients enrolled should not exceed 60 cases for single surgeon. This limits are aimed at homogenizing the contribution of individual investigators and avoiding, within the enrolled population, excessive imbalances and discrepancies related to the contribution of individual surgeons. Once the programmed sample size has been reached, recruitment will be closed while careful follow-up must continue until the study is closed (duration 5 years).
7) ETHICAL CONSIDERATIONS
The study is observational and not sponsored and does not involve any change in the diagnostic, therapeutic and follow-up path of the patients involved. The study only represent an analysis and comparison of data obtained from normal clinical and surgical activity. No additional radiological exposures are foreseen with respect to normally planned follow-up paths. No other additional diagnostic investigations are requested. Furthermore, no experimentation with surgical materials or techniques is envisaged. The study guarantees the transmission and collection of data in a completely anonymous form. There are no conflicts of interest for any of the members of the board. Consequently, there are no problems of an ethical nature.
8) STUDY MONITORING AND SUPERVISION
The supervision of the study provides for the verification of the patients record completeness, the adequacy of the radiographs for the purpose of the study as well as the correct anonymization of the images, the correct presence of follow-ups. Random evaluation of the accuracy of the measurements. Evaluation of correct data analysis procedures. Participation and supervision to the results analysis. Supervision is carried out by independent figures who are not part of the study board and are not investigators of the study.
A telephone evaluation will be performed anonymously on a sample of enrolled patients representative of all investigators to verify the accuracy and truthfulness of the entered data.
Inclusion Criteria:
Exclusion Criteria: