Proximal adjacent segment pathology (ASP) can manifest itself in many different ways both clinically and radiographically. Proximal junctional kyphosis (PJK) is 1 specific form of radiographical ASP that can occur after reconstructive spine surgery for kyphosis and scoliosis.1–5 Originally described radiographically as kyphosis greater than 10° between the upper instrumented vertebra (UIV) and the vertebral body 2 levels above, PJK in severe cases can lead to compromised outcomes and the need for revision surgery.2 Since it was first described, multiple studies have attempted to elucidate the risk factors associated with its development; however, no study has been able to definitively identify a single variable associated with the development of PJK, thus its etiology is likely multifactorial.1,3–5 Identified risk factors include combined anteroposterior surgical approach, type of instrumentation, magnitude of correction and sagittal imbalance correction, inclusion of pelvic fixation, osteoporosis, and integrity of the posterior ligamentous complex as well as the proximally instrumented vertebral level.1,3–7
The purpose of this systematic review was to examine the following key questions regarding PJK:
- In persons who have had primary or revision fusion for scoliosis or kyphosis extending to the thoracolumbar or mid or proximal thoracic spine:
- What is the risk and timing of PJK development?
- What risk factors are associated with PJK development?
- Do persons with PJK experience decreased function or quality of life?
MATERIALS AND METHODS
Electronic Literature Search
We performed a systematic search of PubMed, CINAHL, EMBASE, the Cochrane Library, and Google Scholar for literature published through February 15, 2012. The search was limited to studies written in English. Identification of studies explicitly designed to evaluate PJK and its risk and risk factors was the primary focus. Terms specific to PJK (proximal junctional kyphosis) (Title/Abstract) OR PJK OR (proximal and junction* and kypho*) as well as those related to complications of scoliosis or kyphosis surgery, diagnosis, and treatment of PJK were used. Meeting abstracts/proceedings, white papers, and editorials were excluded. Studies of adults who had primary or revision surgery for spinal deformity (scoliosis or kyphosis) that involved 5 or more vertebral levels were considered for inclusion. Articles were included if they defined PJK on the basis of the criteria outlined by Glattes et al,2 namely, proximal junction sagittal Cobb angles between the lower endplate of the UIV and the upper endplate of the 2 supra-adjacent vertebrae 10° and more and at least 10° greater than the preoperative measurement. Studies not meeting this definition or not describing the number of surgically treated levels were excluded. In instances in which multiple reports published on the same underlying population existed, the report with the most complete data was selected. Studies that included more than 20% of patients with trauma, tumor, infection, neuromuscular scoliosis, or ankylosing spondylitis were excluded. Studies with fewer than 10 patients were excluded. Biomechanical, cadaver, and computer simulation studies were excluded.
From the included articles, the following data were extracted: patient demographics, inclusion and exclusion criteria, follow-up duration and the rate of follow-up, risk (%) of PJK, potential risk factors and related effect sizes, and data comparing quality of life and function in those who developed PJK and those who did not (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703).
Study Quality and Overall Strength of Body of Literature
Level-of-evidence (LoE) ratings were assigned to each article independently by 2 reviewers using criteria set by The Journal of Bone & Joint Surgery, American Volume,8 for prognostic studies and modified to delineate criteria associated with methodological quality and described elsewhere.9 The overall strength of the body of evidence (SoE) with respect to each clinical question was determined on the basis of precepts outlined by the Grading of Recommendations Assessment, Development and Evaluation working group10 and recommendations made by the Agency for Healthcare Research and Quality.11 Risk of bias was evaluated during the individual study evaluation described earlier in the section “Study Quality and Overall Strength of Body of Literature.” This system, which derives a strength-of-evidence grade of “high,” “moderate,” “low,” or “insufficient” for each outcome or clinical question, is described in further detail in the methods article for this focus issue.9 The supplemental digital material contains the details of how we arrived at the strength of evidence for each key question (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703).
When data were provided, the risk (cumulative incidence) of PJK was reported. Data were summarized but were not pooled between studies because of the limited number of studies for specific populations available and the heterogeneity of patient populations and definitions for various factors.
Where authors provided effect sizes from multivariate analysis (i.e., adjusted effect size estimates) and/or of significance based on adjustment for confounders, these were reported. In studies that did not use multivariate analysis, crude risk ratios (RRs) and 95% confidence intervals (95% CIs) or risk differences were calculated to provide an estimate of effect size where data were available. Calculations were performed using Stata 9.0.12 Odds ratios (ORs), RRs, or hazard ratios whose CI included the value of 1 were not statistically significant. Values above 1 suggest increased risk and values below 1 suggest decreased risk for PJK. Wide CIs suggest greater variability in estimates and call into question the certainty of the effect estimate. For continuous data, the mean change and percent change relative to preoperative values were calculated if the data were not part of the multivariate analysis.
Clinical Recommendations and Consensus Statements
Clinical recommendations or consensus statements were made through a modified Delphi approach by applying the Grading of Recommendations Assessment, Development and Evaluation/Agency for Healthcare Research and Quality criteria that imparts a deliberate separation between the strength of evidence (i.e., high, moderate, low, or insufficient) and the strength of recommendation. When appropriate, recommendations or statements “for” or “against” were given “strong” or “weak” designations based on the quality of evidence, the balance of benefits/harms, and the values and patient preferences. In some instances, costs may have been considered. A more thorough description of this process can be found in the Focus Issue methods article.9
The systematic literature search yielded 85 potentially relevant citations that were evaluated against the inclusion/exclusion criteria set a priori (Figure 1 and Table 1). Two investigators independently considered the studies for inclusion. Discrepancies in selection were resolved by discussion and when necessary, evaluation of the manuscript. The majority of studies (n = 63) were excluded on the basis of title and abstract evaluation. The full text of 22 articles was reviewed, leading to exclusion of 15. Some studies provided data to answer more than 1 question. Seven studies were included, forming the basis of this report.1,3–7,13 The cohort studies with data available to answer the key questions were graded as LoE III. A list of excluded studies, including definitions of PJK used in those studies, can be found in the supplemental digital material along with the details of LoE determination (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703).
Table 2 summarizes characteristics of included studies and their populations.
Reported Risk and Timing of PJK Development
The risk of PJK ranged from 17% to 39% in patients with adult scoliosis,3,4,6 27% to 28% in patients with adolescent scoliosis,5,7 and 30% to 32% in patients with Scheuermann kyphosis1,13 (Figure 2).
The timing for the development of PJK was not uniformly reported across studies. In 1 study, 43% of adults who developed PJK did so within the first year of follow-up.3 Furthermore, the average increase in the PJK angle by 8 weeks postoperative accounted for 59% of the total increase at final follow-up.4 Among studies of adolescent scoliosis, the greatest increase in PJK angle occurred within 1.5 to 2.0 years. Neither of the studies in persons with Scheuermann kyphosis provided information on the timing of PJK development (Table 3).
What Factors Are Associated With PJK Development?
Potential risk factors were evaluated across multiple studies and are illustrated in Table 4. Tables 5 to 9 summarize these risk factors and provide additional information on effect sizes on the basis of available data. Some studies evaluated and adjusted for potential confounding factors with the use of multivariate analyses.3,6
Patient and Demographic Factors
As reported in 2 studies that employed multivariate analysis,3,6 age at surgery was not significantly associated with PJK although it was significant in the third study based on unadjusted estimates.4 Sex, body mass index, and comorbidities were not statistically significant risk factors when adjusting for other confounding variables3 (Table 5).
No statistical association between age and the occurrence of PJK was seen in studies of adolescent scoliosis.5,7 Risser sign was used to grade skeletal maturity (scale range, 0–5, where 5 indicated completion of fusion). Lower grades were associated with PJK in multivariate analysis from 1 study,7 but no association was found in the other.5 In 1 study, males were at higher risk for PJK based on unadjusted RRs5 but not significantly so in the other, which adjusted for other factors7 (Table 6).
None of the studies of Scheuermann kyphosis evaluated patient demographic factors or characteristics as risk factors.
Two studies in adults evaluated fusion to S1 as a risk factor for PJK, but this association failed to reach significance in a multivariate analysis. Combined anteroposterior approach was associated with an increased risk of PJK in 2 studies based on adjusted estimates, and the UIV level at T1–T3 was associated with PJK in 1 study, which provided adjusted estimates.3 Similarly, the UIV at T8 or below was not4 based on univariate analysis. After adjustment for age, use of pedicle screws did not significantly increase PJK risk versus the use of hybrid constructs or hooks. Performing an osteotomy did not increase the risk for PJK in that same study (Table 5).
In the 2 studies of adolescents,5,7 no significant association was found between the number of fused vertebrae and the risk of PJK. Increased risk of PJK was associated with performed thoracoplasty.5,7 Both studies suggested that the use of screws might be associated with higher PJK risk than the use of hooks. However, the wide CI for the adjusted OR reported by Wang et al7 suggested great variability in the estimate and calls its stability into question. The study by Wang et al7 suggested that the use of distraction (vs. a rotating bar) increases PJK risk and the use of allogenic bone or biomaterials for fusion decreases the risk compared with autogenous bone for grafting. The proportions of persons who developed PJK and those who did not were similar with respect to the UIV in 1 study5 (Table 6).
Limited information on the influence of surgical factors on the risk of PJK from unadjusted estimates was found in the 2 Scheuermann kyphosis studies. In both, shorter fusions were associated with increased PJK risk. Lonner et al13 evaluated the risk of PJK for levels fused in relation to the proximal Cobb end vertebra. An increase in PJK risk was associated with fusion levels caudal to and including the Cobb end vertebra compared with fusion cephalad to the Cobb end vertebra, but statistical significance was not achieved. Denis et al1 described the “level of proximal instrumented vertebra” and the “proximal end vertebra (PEV)” and the risk associated with the proximal instrumented vertebra being short of the PEV. The PEV was determined by using the “best fit line” technique on the radiographs. Fusions short of the PEV were associated with a significant increase in PJK risk; however, the wide CI should be considered. In this study, correction of more than 50% was associated with increased risk of PJK. The authors state that this finding lost significance when PJK was related to failure to fuse to the actual upper end of the vertebra of the curve or when there was damage to the ligamentum flavum; however, they do not provide supportive data. Lonner et al13 reported that there was no association with proximal anchor type (hook or screw) and PJK development (P = 0.36) but did not provide supportive data (Table 7).
Radiographical Factors: Angles
Authors reported various measured angles from radiographs. Tables 8 and 9 summarize findings related to these measurements and additional data are available in the supplemental digital content (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703). Kim et al5 found that preoperative thoracic kyphosis of more than 40° along with a postoperative decrease of thoracic kyphosis more than 5° demonstrated an increased risk of PJK development. It may be that if excessive kyphosis is removed during thoracic spinal deformity surgery, there is a risk of PJK developing to compensate for the kyphosis that has been removed.
Two studies of adults evaluated the PJK angle. It was not a significant risk factor in multivariate analysis in 1 study,3 but the other, using unadjusted analyses, found significant differences postoperatively between those who did and did not develop PJK4 (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703). In studies using multivariate analyses, the sagittal sacral vertical line,3 the thoracic kyphosis-lumbar lordosis Cobb angle, and the C7 plumb line6 remained significantly different between groups (Table 8). Kim et al4 also reported on the following measures: thoracic (T5–T12) kyphosis (significant differences at 8 wk and ultimate follow-up but not preoperatively), lumbar lordosis (no statistical differences at any time point), and sagittal vertical axis (no statistical differences at any time point); additional details can be found in the supplemental digital content (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703).
In the 2 studies of adolescents, increased preoperative thoracic Cobb angle and postural thoracic curvature angle were associated with increased risk of PJK5,7 (Table 9). The multivariate analysis of Wang et al7 included preoperative postural curvature of thoracic vertebrae: compared with those with less than 20°, those with more than 40° were at greater risk for PJK (the odds of >40° angle were 4 times greater for those who had PJK). On the contrary, Kim et al5 reported no significant differences between those who did and did not develop PJK in the proximal thoracic Cobb angle, main thoracic Cobb angle, or C7 to the plumb line to the sacrum. In addition, Kim et al5 examined radiographical factors (additional details available in the Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703), preoperatively, immediate postoperatively, and at 2 years postoperatively, comparing values for those who did and did not develop PJK.5 In this study, no significant differences were found in the parameters addressed.
Scheuermann kyphosis studies provided limited radiographical findings. Denis et al1 did not compare thoracic kyphosis, lumbar lordosis, or C7 plumb line between those with and without PJK but did report that 10 patients overall had a residual kyphosis of more than 55° following surgery. Lonner et al13 reported a preoperative kyphosis difference between groups (P = 0.07), a statistically significant difference in kyphosis at follow-up (P < 0.001) with the percent change also significant (P = 0.02) but no difference in pelvic incidence.
Do Patients With PJK Have Worse HRQOL Outcomes?
Data from 3 studies3–5 suggest that there is no difference in function between those who do and do not develop PJK, based on the Scoliosis Research Society (SRS) questionnaire scores at follow-up. Two studies, 1 in adults4 and 1 in adolescents,5 reported total SRS scores as well as those for the pain, self-image, function, and satisfaction. Kim et al5 used the SRS-24 questionnaire, and the study by Kim et al4 used 3 questionnaires, the SRS-24, 29, and 30, normalized to 100 for purposes of analysis.4 The overall mean scores at follow-up were similar for those who did and did not develop PJK, and no differences between groups were seen when separate domain scores were compared (Figure 3 and Table 10).
The third study reported total SRS outcomes score differences from preoperative to postoperative for each group, with the note that for the score differences, comparison between the patients with PJK and non-PJK did not show a statistically significant difference.3 However, there are certainly patients with PJK who are symptomatic and require either nonoperative or even operative management, but these facts are not quantified well in any current literature.
Studies on Scheuermann kyphosis did not assess HRQOL outcomes.
The overall summary of the available evidence (Table 11) focuses on findings across multiple studies. The strength of evidence was rated as moderate for overall risk (cumulative incidence) of PJK, indicating that future research is likely to have an important impact on the confidence of the estimate and may change the estimate. The strength was upgraded from low to moderate on the basis of estimate size. A rating of low for all of the risk factors across studies of adolescents and adults suggests that future research is very likely to have an important impact on our confidence of the effect estimate and likely to change the estimate. The overall evidence for the majority of factors evaluated in persons with Scheuermann kyphosis was considered insufficient, indicating that any estimate is uncertain.
PJK is a form of ASP that occurs in 17% to 39% of cases after spinal deformity surgery. Although multifactorial in etiology, combined anteroposterior surgery, thoracoplasty, UIV at T1–T3, and those without anatomical restoration of thoracic kyphosis are at higher risk and should be followed closely. Interestingly, the occurrence of PJK had no effect on HRQOL outcomes, and there is a paucity of literature on methods for assessing PJK when it occurs.
It seems that combined anteroposterior surgery, thoracoplasty, UIV at T1–T3, and those without anatomical restoration of thoracic kyphosis postoperatively demonstrated a higher risk for the development of PJK. We decided to concentrate on PJK, because this is the most distinct form of radiographical ASP that allowed a formal systematic review to be successfully carried out. These results should be considered within the context of the overall low (LoE III) quality of studies. Because of the multifactorial etiology of PJK and because of the potential for confounding variables, studies with multivariate analysis were initially considered more strongly in our risk factor assessment. Control for potential confounding factors was seen in several studies,3,6,7 and 1 reported age-adjusted P values for some risk factors.4 Studies that controlled for confounding factors are generally considered to provide the best estimates for determining the strength of association. Some risk factors (i.e., age, fusions to the sacrum), which were identified on univariate analysis, were no longer significantly associated with PJK in multivariate analysis. However, due to the paucity of studies using multivariate analysis, some of these risk factors were still considered important on the basis of the unadjusted estimates and clinical differences in risk between groups.
Our primary focus was on factors that were reported across multiple studies. Risk factors identified in single studies included osteoporosis and obesity, but those that were not significant in our systematic review. Surgery-related risk factors described previously in the literature included those who have had significant corrections in the sagittal plane and pedicle screw instrumentation at the proximal end vertebra.
There are other risk factors that are well accepted by the spinal reconstructive surgical community that are not necessarily quantified in the literature, because these factors are either difficult to quantify or so complex that research to evaluate their effect is challenging. In addition, some of these risk factors may have implications for the development of radiographical or clinical ASP that is different from PJK, such as acute fractures at or directly above constructs with or without neurological sequelae. One such surgical issue is disruption of muscular/ligamentous/bony tissue at or cephalad to the upper instrumented/fused level during spinal reconstructive surgeries. Although all surgeons would agree that minor degrees of this occur in nearly every standard surgery performed, quantification of this disruption is difficult, making a scientific basis for our theory on this difficult to formulate. In addition, the type and degree of corrective forces used during actual rod placement that are a part of spinal deformity corrective surgery probably play an important role as well but are impossible to quantify currently. Thus, the load that the proximal portion of the implants and adjacent tissues is experiencing after a deformity correction can become nonphysiologic and may play a role in the alterations between a rigid/instrumented proximal region of the spine and the abrupt transition to the mobile/uninstrumented and fused immediate cephalad. These transitional forces are real, long lasting, and certainly a component of PJK development.
Quantifying the increase in the relative risk for PJK for various factors across included studies was problematic because a mixture of ORs and RRs is used in this review on the basis of data available in individual studies. ORs were reported in some studies.3,6,7 Although these studies adjusted for potentially confounding factors, they may have overestimated the relative risk of PJK. (From a statistical standpoint, in which there is a high incidence of an outcome [>10%], the OR overestimates the relative risk, particularly with higher frequencies. The risk of PJK reported in the included studies ranged from 17% to 39%, which is not consistent with a “rare” outcome.) Unadjusted RRs were calculated for studies that did not perform multivariate analyses.1,4,5,13 Unfortunately, such estimates do not take into account the potential influence of confounding factors and therefore may be biased. Thus, neither measure may provide the true picture of the association between some factors and increased risk of PJK.
We excluded studies that did not define the PJK angle as the angle between the caudal endplate of the UIV to the cranial endplate of the 2 levels above with an angle of 10° or more and at least 10° greater than the preoperative measurement. Since its first description in the literature, the majority of studies have implemented this measurement. A handful of other studies that used different methods for measuring the PJK angle as well as those that defined PJK as an angle other than 10° were excluded from this study (see Supplemental Digital Content 1, available at http://links.lww.com/BRS/A703). Using a standardized definition for PJK allows for a uniform method of evaluating PJK and risk factors associated with its development across study populations.
From a radiographical perspective, few preoperative or postoperative risk factors were identified in this review. It is likely that there are obvious and subtle alterations in sagittal plane alignment and balance that will undoubtedly contribute to PJK development. The intricate interaction between spinal regional alignment (cervical lordosis, thoracic kyphosis, and lumbar lordosis) and the contributions from skull position and sacropelvic parameters most likely play a role in optimal sagittal balance and for the prevention of things such as PJK. For example, all spinal reconstructive surgeons have experienced patients undergoing a major sagittal realignment procedure in the lumbar spine, such as a pedicle subtraction osteotomy, in order to increase lumbar lordosis and optimize global sagittal balance, where the patient developed PJK cephalad to the construct postoperatively. This can be seen very early postoperatively in those patients in whom the degree of lordosis produced was somewhat excessive for what the patient's regional and global balance required and is reflected by a reciprocal PJK mentioned above. Unfortunately, these types of common clinical scenarios are not well referenced in our current literature. However, the notion that there is a delicate balance among all spinal regions is becoming increasingly accepted, with important contributions from the sacropelvic unit below as well as less understood contributions from the brain/brainstem balance centers above that all together help establish the physiologic set point of global sagittal alignment and balance for each person. When spinal reconstructive surgery alters that dynamic balance, there may be many subtle as well as obvious manifestations, and the development of PJK is a true reflection of that.
PJK is a complication with a multifactorial etiology that, in our systematic review, occurred with increased frequency in those of advanced age, fusions to the sacrum, combined anteroposterior surgery, thoracoplasty, UIV at T1–T3, and those without anatomical restoration of thoracic kyphosis postoperatively. Other radiographical and surgical risk factors probably also play a role in PJK development but have not been adequately quantified in current literature. Patients with these risk factors should be monitored closely for the potential development of PJK. Although the majority of PJK cases do not necessitate revision surgery, those cases that are severe and progressive can lead to a devastating neurological compromise, so close monitoring of patients at high risk is warranted (Figures 4 and 5).
Future directions for research should aim toward stratifying different severities of PJK to delineate those of various radiographical severities, symptomatic versus asymptomatic, as well as those requiring revision surgery versus continued observation. Risk factor analysis should stress multivariate analysis for an etiology, which is multifactorial and further outcomes assessment needs to be performed after stratification into these different subgroups and etiologies.
The overall strength of evidence regarding the cumulative risk of PJK was considered moderate on the basis of the large proportions of persons reported to having developed it across studies, indicating that we have moderate confidence that the evidence reflects the true effect; however, future research may change this confidence as well as the estimate. With regard to the timing of PJK development, various risk factors described, and impact of PJK on HRQOL, the overall strength of evidence across studies of adults and adolescents was low. Additional research is likely to change the estimate as well as our confidence that this represents the true effect. With the exception of surgical risk factors, the evidence across studies of patients with Scheuermann kyphosis was deemed insufficient, indicating that we have low confidence that the evidence reflects the true effect. Further research is likely to change our confidence in the effect estimate and it is likely to change the estimate of effect given that the evidence either is unavailable or does not permit a conclusion (Table 11).
- The risk of developing PJK above a spinal deformity fusion is 17% to 39%, with most noted by 2 years postoperative.Level of Evidence: Moderate
Strength of Statement: Strong
- The risk factors of PJK development include increased age, fusion to sacrum, combined ASF/PSF, thoracoplasty, UIV at T1–T3, and nonanatomic restoration of thoracic kyphosis.Level of Evidence: Low
Strength of Statement: Weak
- The development of PJK does not seem to have a detrimental effect on HRQOL outcomes, at least in milder/nonrevision forms.Level of Evidence: Moderate
Strength of Statement: Weak
- The risk of developing PJK above a spinal deformity fusion is 17% to 39%, with most noted by 2 years postoperative.
- Surgery-related risk factors are combined anterior/posterior surgery, thoracoplasty, UIV at T1–T3, and those without anatomical restoration of thoracic kyphosis postoperatively.
- The use of hooks, wires, or pedicle screws at the proximal level did not have a consistent statistically significant association with PJK across studies.
- The presence of PJK did not result in differences in HRQOL outcomes.
The author contribution is as follows: L.L. study concept, interpretation, manuscript preparation, manuscript revision; C.I.S. interpretation, manuscript preparation, manuscript revision; H.J.K. interpretation, manuscript preparation, manuscript revision; A.C.S. data analysis and interpretation, manuscript preparation, manuscript revision; and E.M.V. data analysis, manuscript preparation, manuscript revision.
Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal's Web site (www.spinejournal.com).
1. Denis F, Sun EC, Winter RB. Incidence and risk factors for proximal and distal junctional kyphosis following surgical treatment for Scheuermann kyphosis: minimum five-year follow-up. Spine (Phila Pa 1976) 2009;34:E729–34.
2. Glattes RC, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adult spinal deformity following long instrumented posterior spinal fusion: incidence, outcomes, and risk factor analysis. Spine (Phila Pa 1976) 2005;30:1643–9.
3. Kim HJ, Yagi M, Nyugen J, et al. Combined anterior-posterior surgery is the most important risk factor for developing proximal junctional kyphosis in idiopathic scoliosis. Clin Orthop Relat Res 2012;470:1633–9.
4. Kim YJ, Bridwell KH, Lenke LG, et al. Proximal junctional kyphosis in adult spinal deformity after segmental posterior spinal instrumentation and fusion: minimum five-year follow-up. Spine (Phila Pa 1976) 2008;33:2179–84.
5. Kim YJ, Lenke LG, Bridwell KH, et al. Proximal junctional kyphosis in adolescent idiopathic scoliosis after 3 different types of posterior segmental spinal instrumentation and fusions: incidence and risk factor analysis of 410 cases. Spine (Phila Pa 1976) 2007;32:2731–8.
6. Mendoza-Lattes S, Ries Z, Gao Y, et al. Proximal junctional kyphosis in adult reconstructive spine surgery results from incomplete restoration of the lumbar lordosis relative to the magnitude of the thoracic kyphosis. Iowa Orthop J 2011;31:199–206.
7. Wang J, Zhao Y, Shen B, et al. Risk factor analysis of proximal junctional kyphosis after posterior fusion in patients with idiopathic scoliosis. Injury 2010;41:415–20.
8. Wright JG, Swiontkowski MF, Heckman JD. Introducing levels of evidence to the journal. J Bone Joint Surg Am 2003;8:1–3.
9. Norvell DC, Dettori JR, Skelly AC, et al. Methodology for the systematic reviews on adjacent segment pathology. Spine (Phila Pa 1976) 2012;37:S10–7.
10. Atkins D, Best D, Briss PA, et al. Grading quality of evidence and strength of recommendations. BMJ 2004;328:1490.
11. West S, King V, Carey TS, et al. Systems to Rate the Strength of Scientific Evidence. Evidence Report/Technology Assessment No. 47 (Prepared by the Research Triangle Institute-University of North Carolina Evidence-based Practice Center, Contract No. 290-97-0011). Rockville, MD: Agency for Healthcare Research and Quality; 2002.
12. StataCorp. Stata Statistical Software: Release 9th ed. College Station, TX: StataCorp LP; 2005.
13. Lonner BS, Newton P, Betz R, et al. Operative management of Scheuermann's kyphosis in 78 patients: radiographic outcomes, complications, and technique. Spine (Phila Pa 1976) 2007;32:2644–52.