Evidence-based medicine (EBM), the term and concept, originated from clinical epidemiologists at McMaster University.1 A current definition of EBM is “the explicit, judicious, and conscientious use of current best evidence from health care research in decisions about the care of individuals and populations.”2 EBM has evolved over the past 20 years to provide a framework for the evaluation of therapy and the application of that assessment to a particular patient or a disease. There are 5 steps involved in the direct practice of EBM: define the question or problem, search for the evidence, critically appraise the literature, apply the results, and audit the outcome.
The evolution of surgical practice has been largely based on single-center innovations. Adaptation of new techniques is significantly affected by local opinion and preferences. The same disease is managed quite differently from one locale to another. In addition, surgical procedures are in constant evolution, with each surgeon adding small technical refinements or new steps that suit their approach to the problem at hand that can be innovative. Major technical innovations tend to be reported as a series and form the core of the clinical surgical literature. New operations may be adopted before comparison with previous techniques. Surgeons have been accused of being enthusiasts and criticized for not evaluating new procedures, techniques or technology, as one would a new drug. Spodick in 1975 argued for an “FDA for the surgeon.”3,4
Application of EBM analysis to spinal instrumentation, and specifically to pedicle screw constructs, is challenging. Spinal techniques have evolved extremely rapidly since Harrington’s first description of spinal instrumentation.5 Changes have not been limited to specific instrumentation techniques. Other advances have also improved our ability to manage patients with spinal deformity. More aggressive release of anatomic structures, pedicle subtraction, better anesthesia, improvements in spinal cord monitoring, better rod reduction instruments, surgical positioning, intraoperative imaging, and improved medical management all have enhanced our ability to treat deformity patients. These temporal changes make comparison of retrospective studies extremely flawed. Unfortunately, prospective, randomized studies comparing instrumentation techniques are very challenging to complete in spinal deformity. As a result, there are no randomized prospective trials specifically examining pedicle screw constructs. Another major limitation of studies of spinal deformity management is the lack of a discriminative outcome measure. The Scoliosis Research Society (SRS) outcome questionnaires, such as the SRS-24, have limited ability to detect advantages of one instrumentation system over another. These factors limit the body of literature available for an evidence based analysis. The objective of this paper is to evaluate the current EBM analysis of the literature in the use of pedicle screw constructs in patients with adolescent idiopathic scoliosis (AIS).
AIS or pedicle screws were keyword searched in the Cochrane database with no results. Ovid Medline and PubMed were searched using the combination of terms “pedicle screws” and “adolescent idiopathic scoliosis.” A total of 55 articles were listed. Of these 55 articles, 28 relevant articles were retrieved based on the information provided in the abstract. The following combination of terms were also searched (and found): pedicle screws + child + spine (34 total), pedicle screws + scoliosis + child (45 total), and pedicle screws + AIS + evidence (2 total). The syntax ‘adol$ adj5 scoli$’ was searched in Ovid Medline but did not retrieve any new articles. As well, to ensure that no articles were missed, the reference list of the major papers by authors Lenke, Suk, and Kim were hand searched. One author (AP) was responsible for the literature search, and all 3 authors determined whether an article was to be included or excluded in this review.
Relevant articles were included in this review if they described a pedicle screw construct to correct AIS and compared pedicle screw constructs with another technique.
Articles that did not have patients with AIS in their subject groups, used patients with a spinal deformity from a mixed diagnosis, did not use pedicle screws as a part of their deformity correction, or did not have a comparison group were excluded from the study.
The primary outcome used for comparison was mean deformity correction in the coronal plane. The secondary outcomes were safety (accuracy, complication rate), patient outcome scores (SRS-24), cost, apical vertebral translation (AVT), lowest instrumented vertebral tilt (LIVT), and fusion level.
Based on the search strategy described above, 40 articles met the inclusion criteria and were selected for review in this manuscript. Of these, 32 studies are retrospective reviews, including 2 studies that do not define their data collection technique. Six studies have a prospective study design, 1 is a case report, and 1 is a cadaveric study.
Of the 32 retrospective studies that are included for review in this manuscript, only 10 studies met our inclusion criteria, which included only AIS patients in their study group (Table 1). These studies compared pedicle screw constructs to sublaminar wires, hooks, and/or hybrid constructs, some matching for age, gender, and curve size. The remaining 22 studies include a mixture of adolescent and adult subjects or include mixed types of spinal deformities.
The prospective studies reviewed were difficult to compare because they have different subjects and outcome measures. Liljenqist et al,6 Halm et al,7 and Bullman et al8 used a mixed group of adults and adolescents with idiopathic scoliosis. Shufflebarger et al9 prospectively collected data on 61 patients with AIS that underwent segmental posterior screw instrumentation but did not have a comparison group. Lee et al10 compared a homogeneous group of subjects with AIS and similar curve types and magnitude. Their primary outcome was to compare the surgical results between direct vertebral rotation with pedicle screw fixation and simple rod derotation. Delorme et al11 were the only group to attempt a prospective randomization protocol on subjects with AIS. The patients were randomly referred to 1 of the 3 surgeons, two of whom were routinely using Cotrel-Dubousset (CD) instrumentation and one who was using Colorado instrumentation. Although this is the first group to attempt randomization, the pedicle screw constructs were not well described in AIS.
Mean Deformity Correction
All pedicle screw constructs provide better, or comparable, correction in terms of mean absolute degree curve correction (range, 1°–20°) and percent of curve correction (range, 2%–26%) in comparison to other constructs (Table 1).
Suk et al categorized the complications into 3 phases.12 Phase I categories are complications attributable to the events before the preparation of the pedicle holes and comprise soft tissue injury, adjacent facet joint injuries, and fractures of the transverse process. Phase II categories refer to the complication occurring during hole preparation and screw placement. Phase III refers to the complications occurring after placement of the screws (e.g., those occurring during the rod derotation maneuver or osteotomies).
The overall complication rate associated with pedicle screw instrumentation ranges from no reported complications to 25%.6,7–10,12–37 The most common complication was the malposition of screws with pedicle screws breaching the medial or lateral walls. Suk et al reported that 10.4% of all patients and 1.5% of all screws were malpositioned.12 The incidence of neurologic injury, either root or cord, appears to be small. However, it is likely subject to a significant reporting bias. A detailed understanding of the incidence of neurologic injury would be an important determinant of strong EBM recommendations and would dominate any financial analysis because of the long-term cost of neurologic injury.
Patient Outcome Scores
The 2-year postoperative SRS-24 scores did not show any differences between the following compared groups: sublaminar wires or pedicle screws,14 pedicle screws or hooks,38 pedicle screws or hybrid instrumentations,39 CD instrumentation or Universal Spine System (USS),19 or between those patients treated with anterior-posterior spinal fusion or posterior spinal fusion alone.34 Luhmann et al34 noted that those patients with pedicle screws only in the posterior spinal fusion group had significantly better scores on postoperative self-image and function. Storer et al23 found no difference in quality of life scores as measured by the Child Health Questionnaire between patients treated with either pedicle screw or hook constructs.
There are no cost-effectiveness or cost utility analyses comparing all pedicle screw constructs with other constructs in AIS. Storer et al23 examined the costs associated with screw constructs in addition to the radiographic and overall quality of life scores. Although the total operative time was similar, there was a significant difference in the cost of implants with the screw constructs being much more expensive than the hook constructs.23 Surgical intervention using screw constructs had an additional cost of U.S. $1711 on average, although this represented less than 5% of the total cost of treating a patient with AIS.23
AVT, LIVT, and Fusion Level
Liljenqvist et al found the correction of both AVT and LIVT was significantly greater in the all pedicle screw group compared to the hook group (AVT 64% vs. 54%, LIVT 70% vs. 60%).21 Barr et al found an even greater difference in LIVT between the screw and hook group (66% vs. 11%).26 Shufflebarger et al found 81% of LIVT correction with all pedicle screws; they did not have a hook comparison group in their study.9 The authors theorized that better horizontalization of LIV can be achieved with screw constructs because the more lateral position of the pedicle screw provides a considerably better leverage7 and because the tangential fixation strength of pedicle screws is significantly greater than that of the laminar hooks. Improved apical vertebral rotation was found in the all pedicle screw group compared with the hook group.40 Suk et al found that all pedicle screw constructs save an average of 1.1 motion segments compared with standard CD hook fixation fused down to the stable vertebra.20 Kim et al also noted that all pedicle screw constructs saved more distal motion segments compared with hooks.38 Kim et al found no difference in distal fusion levels between all pedicle screws and hybrid fixation in AIS.39 The LIV averaged 0.6 vertebrae below the lowest end vertebra in the screw group and 0.4 in the hybrid group. Degenerative change distal to the lowest instrumented vertebra will likely be the most important determinant of long-term outcome in AIS patients.41–44 Unfortunately, the value of improved LIVT and the importance of saving a motion segment at the thoracolumbar junction have not yet been demonstrated in the literature.
Since the advent of Harrington instrumentation,5 several new instrumentation systems and corrective methods have been developed. Pedicle screw fixation appears to provide the surgeon with improved ability to correct the 3-dimensional deformity present in idiopathic scoliosis. Advocates of pedicle screw constructs report that the advantages for all pedicle screw constructs in AIS include 3-column fixation, improved coronal, sagittal, and rotational correction, lower pseudarthrosis rates, lower implant failures, and fewer postoperative bracing requirements when compared with conventional hook and wire constructs.25,37,45,46 The use of pedicle screws also has advantages over hook and wire implants with respect to vertebral anatomy. When placed optimally, screws are completely external to the spinal canal.47–49 The low profile head design of most screw implants facilitates segmental fixation, which is more difficult to accomplish with hooks. Also, facet joints, lamina, and transverse processes are free of implants such as wires or hook blades, leaving more surface area available for aggressive decortication and grafting without destabilizing the instrumentation.50
There were no randomized controlled trials found in this review that addressed pedicle screw use in AIS. Few of the retrospective comparative studies that qualified for inclusion in this report had a defined primary outcome. Most studies compared different spine constructs with respect to degree correction achieved in the coronal and sagittal planes, blood loss, time for surgery, lowest instrumented vertebra, complications, and the SRS-24 outcome measure. The most common outcome measure used in all studies is mean deformity correction. We have used this as a primary outcome measure for this study; it is the only common outcome that provided any opportunity for an EBM analysis of this body of literature. Only 5 studies in this review used SRS-24 outcome scores (Table 2).14,19,34,38,39
There are no cost-effectiveness or cost utility analyses comparing all pedicle screw constructs with other constructs in AIS.51,52 While the high costs of healthcare interventions need to be considered, it is important to first consider the impact of an intervention on clinical outcome measures for the affected patient. Because of the extremely low mortality and long expected life span of these patients, factors affecting long-term morbidity may have the greatest implications on cost. However, accurate cost-effectiveness analysis has not been completed in this area.
The rules of EBM are strict and difficult to apply to a review of an operative technique already used in clinical practice. However, the exercise is an important one. Does the wide variation in operative procedures for this problem make a difference? The large numbers of operations testify to the reality that, in many situations, the technique does not matter despite strenuous and occasionally acrimonious discussions in support of one operation or another. There must be a “best solution” in the hands of the “best surgeon.” We should then know if one procedure should be used by every surgeon. Sometimes, however, a less complicated operation may serve patients better when done by most surgeons.53 A robust patient-centered outcome measure would be a powerful tool in an EBM approach. Unfortunately, only 5 studies in this review used the SRS outcome score,14,19,34,38,39 and only 1 study found any significant difference.34
A frequent argument against the use of all pedicle screw constructs is that the cost and risks involved in pedicle screws may not justify the potentially marginal improvement in correction achieved. A prospective randomized trial would provide the best evidence to answer this question. However, to ethically proceed with a surgical randomized controlled trial, there must be equipoise within the surgical community. Given the complexity of spinal deformity surgery and the extensive training required to safely manage this problem, few surgeons may be able to proceed with a randomized trial. In the absence of randomized trials, we must strive to do well-designed and methodologically strong observational studies. It is important for surgeons to consider the best evidence in managing their patients, realizing however, that given variations in experience, skill, practice environment, and patient population, a specific procedure may not be best practice for a specific surgeon.
Strength of Recommendations
The strength of any recommendation depends on 2 factors: the trade-off between risks and benefits and the quality of the methodology that leads to estimates of the treatment effect. The McMaster approach to grading recommendations captures the magnitude of random error in the decision about the confidence in the trade-off between benefits, harms, and cost54 (Table 3). The uncertainty associated with this trade-off will determine the strength of recommendations. The grades that experts generate using the McMaster approach are 1A, 1B, 1C, 2A, 2B, and 2C.54 If experts are very certain that benefits do, or do not, outweigh harms and cost, they will make a strong recommendation, Grade 1, in the McMaster formulation. If they are less certain of the magnitude of the benefits and harms, and their relative impact, they must make a weaker Grade 2 recommendation. Grade 2 recommendations are those in which variation in patient values or individual physician values often will mandate different treatment choices, even among average or typical patients. The McMaster approach expresses the primacy of risk of harm by placing it first in the grade of recommendation.
Quality of Evidence
The assignment of Grades A to C, the second part of the grade, focuses on the likelihood of bias based on the methodologic quality of the underlying evidence.55 The approach classifies randomized controlled trials with consistent results as Grade A. If results across randomized trials are inconsistent, or if trials have major methodologic weaknesses, one assigns Grade B. Evidence leading to Grade C recommendations comes from observational studies and from generalization of results from randomized trials in one group of patients to a different group. When experts find the generalization from randomized trials secure or the data from observational studies overwhelmingly compelling, they choose Grade C. For example, there are no randomized controlled trials for use of insulin in diabetic ketoacidosis, but no responsible clinician would ask for a randomized controlled trial of insulin treatment because of the large treatment effect observed in observational studies. Indeed, most people would consider such a trial unethical. Although observational studies are likely to overestimate the true effect, the study design is very unlikely to explain the entire benefit. Therefore, in instances such as the one described, the evidence from observational studies is overwhelmingly compelling and would generate Grade C in terms of methodologic quality.
In our review of current literature, there is weak evidence (Level III) to show that all pedicle screw constructs improve and maintain scoliosis curve correction in patients with AIS. From this, we can only give 2C recommendation for the use of all pedicle screw constructs in the treatment of AIS.
In the absence of evidence from randomized trials, surgeons must rely on the best available information to guide patient management decisions. Although there have been many publications on the topic of all pedicle screw constructs in AIS, evidence regarding the advantage of all pedicle screw constructs remain limited to case series, biomechanical studies, and expert opinions. There appear to be a number of advantages of all pedicle screw constructs in AIS, such as curve correction, lowest instrumented vertebra, vertebral tilt, and the avoidance of anterior surgery. However, those outcomes may not be associated with an improved patient-related quality of life outcome or long-term outcome.
- There is weak evidence (Level III) to show that all pedicle screw constructs improve and maintain scoliosis curve correction in patients with AIS.
- The evidence regarding the advantage of all pedicle screw constructs remains limited to case series, biomechanical studies, and expert opinions.
- Even though there appear to be advantages in a number of parameters, such as curve correction, those outcomes may not be associated with an improved patient-related quality of life outcome or long-term outcome.
1. Guyatt G. Evidence-based medicine
. ACP J Club
2. Sackett DL, Straus S, Richardson SR, et al. Evidence-Based Medicine
: How to Practice and Teach EBM
. London: Churchill Livingstone; 2000.
3. Spodick DH. Numerators without denominators. There is no FDA for the surgeon. JAMA
4. Love JW. Drugs and operations: some important differences. JAMA
5. Harrington PR. Treatment of scoliosis: correction and internal fixation by spine instrumentation. J Bone Joint Surg Am
6. Liljenqvist U, Halm H, Link T. Pedicle screw instrumentation of the thoracic spine in idiopathic scoliosis. Spine
7. Halm H, Niemeyer T, Link T, et al. Segmental pedicle screw instrumentation in idiopathic thoracolumbar and lumbar scoliosis. Eur Spine J
8. Bullman V, Halm H, Schulte T, et al. Combined anterior and posterior instrumentation
in severe and rigid idiopathic scoliosis. Eur Spine J
9. Shufflebarger HL, Geck MJ, Clark CE. The posterior approach for lumbar and thoracolumbar adolescent idiopathic scoliosis
: posterior shortening and pedicle screws
10. Lee SM, Suk S, Chung ER. Direct vertebral rotation: a new technique of three-dimensional deformity correction with segmental pedicle screw fixation in adolescent idiopathic scoliosis
11. Delorme S, Labelle H, Aubin CE, et al. A three-dimensional radiographic comparison of Cotrel-Dubousset and Colorado instrumentations for the correction of idiopathic scoliosis. Spine
12. Suk S, Kim WJ, Lee SM, et al. Thoracic pedicle screw fixation in spinal deformities: are they really safe? Spine
13. Faraj AA, Webb JK. Early complication of spinal pedicle screw. Eur Spine J
14. Cheng I, Kim YJ, Gupta MC, et al. Apical sublaminar wires versus pedicle screws
: which provides better results for surgical correction of adolescent idiopathic scoliosis
15. Kim YJ, Lenke LG, Cheh G, et al. Evaluation of pedicle screw placement in the deformed spine using intraoperative plane radiographs: a comparison with computerized tomography. Spine
16. Kuklo TR, Lenke LG, O’Brien MF, et al. Accuracy and efficacy of thoracic pedicle screws
in curves more than 90°. Spine
17. Kuklo TR, Potter BK, Polly D Jr, et al. Monaxial versus multiaxial thoracic pedicle screws
in the correction of adolescent idiopathic scoliosis
18. Dobbs MB, Lenke LG, Kim YJ, et al. Anterior/posterior spinal instrumentation versus posterior instrumentation
alone for the treatment of adolescent idiopathic scoliotic curves more than 90°. Spine
19. Remes V, Helenius I, Schlenzka D, et al. Cotrel-Dubousset (CD) or Universal Spine System (USS) instrumentation in adolescent idiopathic scoliosis
(AIS): comparison of midterm clinical, functional, and radiologic outcomes. Spine
20. Suk S, Lee SM, Chung ER, et al. Selective thoracic fusion with segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis: more than 5-year follow-up. Spine
21. Liljenqvist U, Lepsien U, Hackenberg L, et al. Comparative analysis of pedicle screw and hook instrumentation in posterior correction and fusion of idiopathic thoracic scoliosis. Eur Spine J
22. Rhee JM, Bridwell KH, Won DS, et al. Sagittal plane analysis of adolescent idiopathic scoliosis
23. Storer SK, Vitale MG, Hyman JE, et al. Correction of adolescent idiopathic scoliosis
using thoracic pedicle screw fixation versus hook constructs. J Pediatr Orthop
24. Laohacharoensombat W, Jaovisdha S, Wajanavisit W, et al. Apical derotation in the treatment of idiopathic scoliosis. J Med Assoc Thai
25. Hamill CL, Lenke LG, Bridwell KH, et al. The use of pedicle screw fixation to improve correction in the lumbar spine of patients with idiopathic scoliosis: is it warranted? Spine
26. Barr SJ, Schuette AM, Emans JB. Lumbar pedicle screws
versus hooks: results in double major curves in adolescent idiopathic scoliosis
27. Kim YJ, Lenke LG, Bridwell KH, et al. Free hand pedicle screw placement in the thoracic spine: is it safe? Spine
28. Girardi FP, Boachie-Adjei O, Burke SW, et al. Surgical treatment of adolescent idiopathic scoliosis
: a comparative study of two segmental instrumentation systems. J Spinal Disord
29. Wimmer C, Gluch H. Aseptic loosening after CD instrumentation in the treatment of scoliosis: a report about eight cases. J Spinal Disord
30. King AG, Mills TE, Chutkan NB, et al. Lumbar pedicle morphology in adolescent idiopathic scoliosis
31. Suk S, Lee CK, Min HJ, et al. Comparison of Cotrel-Dubousset pedicle screws
and hooks in the treatment of idiopathic scoliosis. Int Orthop
32. De Giorgi G, Stella G, Becchetti S, et al. Cotrel-Dubousset instrumentation for the treatment of severe scoliosis. Eur Spine J
33. Arlet V, Papin P, Marchesi D, et al. Adolescent idiopathic thoracic scoliosis: apical correction with specialized pedicle hooks. Eur Spine J
34. Luhmann SJ, Lenke LG, Kim YJ, et al. Thoracic adolescent idiopathic scoliosis
curves between 70° and 100°: is anterior release necessary? Spine
35. Suk S, Kim WJ, Lee CS, et al. Indications of proximal thoracic curve fusion in thoracic adolescent idiopathic scoliosis
: recognition and treatment of double thoracic curve pattern in adolescent idiopathic scoliosis
treated with segmental instrumentation. Spine
36. Smorgick Y, Millgram MA, Anekstein Y, et al. Accuracy and safety of thoracic pedicle screw placement in spinal deformities. J Spinal Disord Tech
37. Lonstein JE, Denis F, Perra JH, et al. Complications associated with pedicle screws
. J Bone Joint Surg Am
38. Kim YJ, Lenke LG, Cho SK, et al. Comparative analysis of pedicle screw versus hook instrumentation in posterior spinal fusion of adolescent idiopathic scoliosis
39. Kim YJ, Lenke LG, Kim JH, et al. Comparative analysis of pedicle screw versus hybrid instrumentation in posterior spinal fusion of adolescent idiopathic scoliosis
40. Liljenqvist U, Hackenberg L, Halm H. Pullout strength of pedicle screws
versus pedicle and laminar hooks in the thoracic spine. Acta Orthop Belg
41. Aaro S, Ohlen G. Harrington instrumentation on the sagittal configuration and mobility of the spine in scoliosis. Spine
42. Burton DC, Asher MA, Lai SM. The selection of fusion levels using torsional correction techniques in the surgical treatment of idiopathic scoliosis. Spine
43. Cochran T, Irstam L, Nachemson A. Long-term anatomic and functional changes in patients with adolescent idiopathic scoliosis
treated by Harrington rod fusion. Spine
44. Ginsburg HH, Goldstein L, Robinson SC. Back pain in postoperative idiopathic scoliosis: a long-term follow-up study. Spine
45. Gaines RW Jr. The use of pedicle-screw internal fixation for the operative treatment of spinal disorders. J Bone Joint Surg Am
46. Suk S, Kim WJ, Lee SM, et al. Segmental pedicle screw fixation in the treatment of thoracic idiopathic scoliosis. Spine
47. Vaccaro AR, Rizzolo SJ, Allardyce TJ, et al. Placement of pedicle screws
in the thoracic spine: I. Morphometric analysis of the thoracic vertebrae. J Bone Joint Surg Am
48. Gertzbein SD, Bobbins SE. Accuracy of pedicular screw placement in vivo. Spine
49. Brown CA, Lenke LG, Bridwell KH, et al. Complications of pediatric thoracolumbar and lumbar pedicle screws
50. Lenke LG, Newton PO, Marks MC, et al. Prospective pulmonary function comparison of open versus endoscopic anterior fusion combined with posterior fusion
in adolescent idiopathic scoliosis
51. Brauer CA, Rosen AB, Olchanski NV, et al. Cost-utility analyses in orthopaedic surgery. J Bone Joint Surg Am
52. Brauer CA, Neumann PJ, Rosen AB. Trends in cost effectiveness analyses in orthopaedic surgery. Clin Orthop
53. Meakins JL. Innovation in surgery: the rules of evidence. Am J Surg
54. Guyatt G, Schunemann HJ, Cook DJ, et al. Grades of recommendation for antithrombotic agents. Chest
55. Schunemann HJ, Bone L. Evidence-based orthopaedics: a primer. Clin Orthop