The Association Between Mental Disorders and Postoperative Outcomes of Scoliosis Surgery: A Systematic Review and Meta-analysis (2024)

Literature Search Strategy

A comprehensive literature search was conducted in PubMed, EMBASE, PsycINFO, and Cochrane Library from inception to February 2023 using a combination of controlled vocabulary terms and free text keywords related to “scoliosis”, “surgery”, “mental disorders”, “psychological factors”, “postoperative complications”, and “treatment outcome”. The full search strategy included the following terms:

(scoliosis[Title/Abstract] OR scoliotic[Title/Abstract]) AND (surgery[Title/Abstract] OR surgical[Title/Abstract] OR operation[Title/Abstract] OR operative[Title/Abstract]) AND (mental[Title/Abstract] OR psychological[Title/Abstract] OR psychiatry[Title/Abstract] OR emotion*[Title/Abstract] OR mood[Title/Abstract] OR depress*[Title/Abstract] OR anxi*[Title/Abstract] OR stress[Title/Abstract] OR psychosis[Title/Abstract] OR schizophrenia[Title/Abstract] OR bipolar[Title/Abstract] OR somatoform[Title/Abstract] OR somatization[Title/Abstract] OR conversion[Title/Abstract] OR OCD[Title/Abstract] OR ADHD[Title/Abstract] OR inattention[Title/Abstract] OR hyperactiv*[Title/Abstract]) AND (outcome*[Title/Abstract] OR complication*[Title/Abstract] OR pain[Title/Abstract] OR disabilit*[Title/Abstract] OR satisf*[Title/Abstract] OR qualit*[Title/Abstract]) AND (female*[Title/Abstract] OR woman*[Title/Abstract] OR girl*[Title/Abstract]) AND (adolescen*[Title/Abstract] OR teen*[Title/Abstract] OR youth*[Title/Abstract]) AND (age[Title/Abstract] OR aged[Title/Abstract] OR aging[Title/Abstract])

Additional studies were identified through handsearching reference lists of relevant reviews and included studies.

This study was guided by the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) statement.

Study Selection and Eligibility Criteria

The inclusion criteria for this systematic review were refined to ensure a comprehensive and relevant analysis of the literature. Only randomized controlled trials or observational studies were considered. Reviews, case reports, animal studies, and non-English studies were excluded.

Participant Characteristics

Scoliosis surgery: Participants must have undergone scoliosis surgery.

Age range: Studies including participants of any age were considered, with specific attention to categorizing findings based on age groups (e.g., children, adolescents, adults).

Mental disorder exposure: Studies examining patients with a documented history of mental disorders, such as depression or anxiety, were included. The duration of exposure to these mental health conditions was noted where available.

Outcome Measures

Studies must have examined the associations between preoperative mental disorders and postoperative outcomes.

Outcome measures should include validated assessment scales for mental health, pain, and functional status post-surgery. Examples include, but are not limited to, the Modified Somatic Perceptions Questionnaire (MSPQ),⁶ which has a Cronbach alpha of 0.861, the Patient Health Questionnaire (PHQ),⁷ which has a sensitivity of 88% and a specificity of 88% for major depressive disorder, the State-Trait Anxiety Inventory (STAI),⁸ which has a test-retest reliability of 0.923, and the Scoliosis Research Society-22 (SRS-22) questionnaire,⁹ which has a construct validity of 0.834.

Data Requirements

Studies must have reported data that could be extracted for meta-analysis. This includes both quantitative and qualitative data types. Quantitative data refers to numerical values for pre- and postoperative assessments, such as scores on the assessment scales, pain intensity, and functional level. Qualitative data refers to non-numerical data that describe the subjective experiences, opinions, and attitudes of the patients. Statistical analysis results and correlations between mental health status and postoperative outcomes should also be reported for both data types.

The selection process involved 2 reviewers independently screening titles and abstracts, followed by full-text reviews using Covidence systematic review software Revman 5.2 (The Nordic Cochrane Centre, Copenhagen, Denmark). Any disagreements were resolved through discussion to ensure a consistent and thorough selection process.

Data Extraction and Risk of Bias Assessment

Data were extracted into a standardized form capturing study characteristics, mental disorder measures, and postoperative outcome results. In our data extraction process, 2 independent reviewers systematically collected study details using a standardized form. Discrepancies were resolved through discussion or consultation with a third reviewer. Risk of bias was assessed using the Newcastle–Ottawa Scale, with any disagreements resolved through consensus or involvement of a senior reviewer. Literature quality was evaluated based on reporting, methodology, and relevance to the research question, ensuring a comprehensive and unbiased assessment.

Statistical Analysis

Random-effects (Tibco Software Inc., California, USA) meta-analyses were performed to pool standardized mean differences (SMDs) between patients with and without mental disorders for postoperative outcomes using the meta package in R statistical software Revman 5.2 (The Nordic Cochrane Centre, Copenhagen, Denmark). We assessed the heterogeneity among the studies using the I² statistic, and Cochran’s Q statistic was also calculated to assess the presence of heterogeneity. Publication bias was evaluated through funnel plots, Egger’s test, and Begg’s test. We set the level of significance at α = 0.05.

Summary of Studies on Mental Disorders and Scoliosis Surgery Outcomes

The study identified 405 records through database searching and 32 additional records through other sources, resulting in 437 potentially relevant articles. After screening the title and abstract, 195 articles were excluded for various reasons, such as being unrelated to the study, being review articles or case reports, being unable to translate into English, or being unable to retrieve. This left 242 articles for full-text assessment. After screening the full text, 151 articles were excluded for various reasons, such as having an unsuitable study design, being unable to construct a table, having possible overlapping data, including less than 5 patients, or being unable to derive specificity. This left 12 articles for eligibility assessment (Figure 1).

The study selected 5 articles out of the 12 eligible articles for meta-analysis. These articles were Catanzano et al., 2023,¹⁰ Voepel-Lewis et al., 2018,¹¹ Yakut et al., 2022,¹² Hwang et al., 2019,¹³ and Lau et al., 2023.¹⁴ These articles met the criteria for meta-analysis, such as having comparable outcome measures, having sufficient data for pooling, and having low heterogeneity.

The results of the included studies are summarized in Table 1. The studies varied in terms of sample size, age range, region, country, and main findings. The studies used different instruments to measure mental disorders, such as the Patient Health Questionnaire (PHQ), the Distress and Risk Assessment Method (DRAM), the Children’s Depression Inventory (CDI), the State-Trait Anxiety Inventory (STAI), and the Scoliosis Research Society questionnaire (SRS-22r). The studies also used different outcome measures to assess the postoperative results of scoliosis surgery, such as the Oswestry Disability Index (ODI), the SF-36 Physical and Mental Component Summary (PCS and MCS), the Visual Analogue Scale (VAS), and the SRS-22r.

The meta-analysis of the pooled data showed that mental disorders had a significant influence on the postoperative results of scoliosis surgery in children. Specifically, mental disorders were associated with:

• Higher preoperative pain intensity and disability scores

• Lower preoperative quality of life and satisfaction scores

• Higher postoperative pain intensity and disability scores

• Lower postoperative quality of life and satisfaction scores

• Higher risk of chronic back pain and insomnia in adulthood

• The meta-analysis also showed that some factors moderated the influence of mental disorders on the postoperative results of scoliosis surgery in children. These factors included:

• The type and severity of mental disorder: patients with more severe depression or anxiety, or those with somatization symptoms, had worse outcomes than those with mild or no depression or anxiety, or those without somatization symptoms. The DRAM was found to be a useful screening tool to identify patients at risk of poor outcomes based on their mental health status. The MSPQ score was found to be a predictor of ODI improvement at 2 years.

• The type and phase of treatment: patients who underwent surgery had more improvement in their Health-Related Quality Of Life (HRQOL) scores than those who were treated conservatively or observed. However, patients who underwent surgery also had higher levels of self-criticism, neuroticism, depressiveness, and kinesiophobia than those who did not. The duration of brace application was positively correlated with depression. Preoperative and postoperative interventions to reduce stress and pain were suggested to improve the outcomes.

• The age and sex of the patient: older patients and female patients had higher rates of back pain, depression, and anxiety than younger patients and male patients. These factors were also associated with worse spinal deformity parameters and lower HRQOL scores. Age- and sex-specific strategies to address the psychological needs of scoliosis patients were recommended.

Meta-Analysis on the Influence of Mental Disorders on Scoliosis Surgery Outcomes

The results of the meta-analysis showed that mental disorders had a significant influence on the postoperative results of scoliosis surgery. The meta-analysis included 3 studies with a total of 145 participants. The studies measured the SMDs between the test group (patients with mental disorders) and the control group (patients without mental disorders) on various outcome measures, such as pain, disability, quality of life, and satisfaction.

The fixed effect model estimated that the mean deviation (MD) was −1.5147, meaning that patients with mental disorders had worse outcomes than patients without mental disorders by 1.5147 points on average. The 95% confidence interval (CI) was −1.9597 to −1.0697, meaning that there was a 95% chance that the true MD was between these values. The z-test showed that the MD was statistically significant (z = −6.67, P < .001), meaning that there was a very low probability of observing such a large MD by chance. The random effects model estimated that the MD was −1.9617, meaning that patients with mental disorders had worse outcomes than patients without mental disorders by 1.9617 points on average. The 95% CI was −3.0784 to −0.8449, meaning that there was a 95% chance that the true MD was between these values. The z-test showed that the MD was also statistically significant (z = −3.44, P < .001), meaning that there was a very low probability of observing such a large MD by chance.

The meta-analysis also quantified the heterogeneity among the studies, which is the variation in the effect sizes due to differences in study characteristics or methods. The meta-analysis used 3statistics to measure heterogeneity: Tau², I², and H. Tau², which is the estimated variance of the true effect sizes across studies, reflects the amount of heterogeneity and was estimated to be 0.731, with a 95% CI of 0.000 to 35.249. Tau is the square root of tau², which reflects the standard deviation of the true effect sizes across studies. Tau was estimated to be 0.8550, with a 95% CI of 0.0000 to 5.9371. I² is the percentage of total variation in the observed effect sizes that is due to heterogeneity rather than sampling error. I² can range from 0% to 100%, with higher values indicating more heterogeneity. I² was estimated to be 76%, with a 95% CI of 22%-93%. H is the ratio of the standard error of the observed effect sizes to the standard error assuming no heterogeneity. H can range from 1 to infinity, with higher values indicating more heterogeneity. H was estimated to be 2.05, with a 95% CI of 1.13-3.71.

The meta-analysis also tested the hypothesis that there was no heterogeneity among the studies using the Q statistic, which is the weighted sum of squares of the differences between the observed effect sizes and the pooled effect size. The Q statistic follows a chi-square distribution with k-1 degrees of freedom, where k is the number of studies. The Q statistic was calculated to be 8.41, with 2 degrees of freedom. The P-value for testing heterogeneity was .01, meaning that there was strong evidence to reject the null hypothesis of no heterogeneity.

As shown in Figure 2, the forest plot demonstrates the pooled standardized mean differences (SMDs) between patients with and without mental disorders, indicating the substantial influence mental disorders have on postoperative outcomes in scoliosis surgery.

The funnel plot (Figure 3) shows that 2 studies (Voepel-Lewis et al., 2018 and Yakut et al., 2022) are located on the left side of the funnel plot, while 1 study (Catanzano et al., 2023) is located on the right side. This means that these studies have more negative MDs than the pooled effect size, indicating that patients with mental disorders had worse outcomes than patients without mental disorders by a larger margin than the average. These studies also have larger standard errors (SEs) than the other study, indicating that they have less precision and more variability in their effect sizes. The asymmetry of the funnel plot suggests that there may be some factors that influence the effect of mental disorders on the postoperative results of scoliosis surgery, such as the type and severity of mental disorder, the type and phase of treatment, the age and sex of the patient, or other methodological or clinical characteristics of the studies.

Meta-Analysis on the Influence of Postoperative Pain on Mental Disorders in Scoliosis Patients

The results of the meta-analysis showed that pain had a significant influence on the mental disorders of scoliosis patients. The meta-analysis included 6 datasets. The studies measured the SMDs between the test group (patients with pain) and the control group (patients without pain) on various outcome measures, such as depression, anxiety, fatigue, pain interference, catastrophizing, and pain DETECT scores. The meta-analysis used both fixed effect and random effects models to estimate the pooled MD and its 95% CI. The fixed effect model assumes that the true effect size is the same across all studies, while the random effects model allows for heterogeneity among the studies. The fixed effect model estimated that the MD was 0.6290, meaning that patients with pain had worse mental disorders than patients without pain by 0.6290 points on average. The 95% CI was 0.5653-0.6928, meaning that there was a 95% chance that the true MD was between these values. The z-test showed that the MD was statistically significant (z = 19.33, P < .001), meaning that there was a very low probability of observing such a large MD by chance. The random effects model estimated that the MD was 0.3914, meaning that patients with pain had worse mental disorders than patients without pain by 0.3914 points on average. The 95% CI was 0.1369-0.6459, meaning that there was a 95% chance that the true MD was between these values. The z-test showed that the MD was also statistically significant (z = 3.01, P = .003), meaning that there was a very low probability of observing such a large MD by chance.

The meta-analysis also quantified the heterogeneity among the studies, which is the variation in the effect sizes due to differences in study characteristics or methods. Tau² was estimated to be 0.087, with a 95% CI of 0.015-0.448. Tau was estimated to be 0.2950, with a 95% CI of 0.1244-0.6690. I² was estimated to be 89%, with a 95% CI of 79%-94%. H was estimated to be 3.06, with a 95% CI of 2.21 to 4.25 (Figure 4).

The funnel plot (Figure 5) shows the effect sizes and CIs of different studies on the relationship between mental disorders and postoperative pain. One study (Hwang et al., 2019) is located on the right side of the funnel plot. This study has a large SMD (0.89) and a small SE (0.15), meaning that it has high precision and a large sample size. Another study (Lau et al., 2023) has 5 groups of data, based on different time periods: past 24 hours, past 7 days, past 1 month, past 6 months, and past 12 months. All these groups are located on the right side of the funnel plot, indicating a positive effect of postoperative pain on mental disorders. However, the effect sizes vary across the groups, with the past 24 hours group having the largest SMD (0.93) and the past 12 months group having the smallest SMD (0.31). The SEs also vary across the groups, with the past 24 hours group having the smallest SE (0.14) and the past 12 months group having the largest SE (0.29). The results of Egger’s test and Begg’s test confirmed that there was no significant publication bias among the studies for each outcome measure. The p-values for Egger’s test ranged from 0.27 to 0.86, and the P-values for Begg’s test ranged from 0.32 to 0.89, all above the significance level of 0.10.

Peer-review: Externally peer-reviewed.

Author Contributions: Concept – L.Y., L.S., L.Y., C.X.; Design –L.Y., L.S., G.C., C.X.; Supervision – L.Y.; Materials – L.Y.; Analysis and/or Interpretation – L.Y., C.X.; Literature Search – G.C.; Writing – S.L., Y.L., C.G., X.C.; Critical Review – L.Y., C.X.

Acknowledgement:

The authors would like to thank the patients and families who participated in the studies included in this meta-analysis. We are also grateful to the researchers and clinicians whose work contributes to our understanding of the impact of mental health on surgical outcomes. Their dedication to patient care and advancing medical knowledge is invaluable.

Declaration of Interests: The authors have no conflicts of interest to declare.

1. Negrini S, Donzelli S, Aulisa AG, et al. SOSORT guidelines: orthopaedic and rehabilitation treatment of idiopathic scoliosis during growth. Scoliosis Spinal Disord. 2016;13(3):2018. [PMC free article] [PubMed] [Google Scholar]

2. Smith JS, Fu KM, Polly DW, et al. Complication rates of three common spine procedures and rates of thromboembolism following spine surgery based on 108,419 procedures: a report from the Scoliosis Research Society Morbidity and Mortality Committee. Spine (Phila Pa 1976). 2010;35(24):21402149. ( 10.1097/BRS.0b013e3181cbc8e7) [PubMed] [CrossRef] [Google Scholar]

3. Smith JS, Shaffrey CI, Glassman SD, et al. Risk-benefit assessment of surgery for adult scoliosis: an analysis based on patient age. Spine (Phila Pa 1976). 2011;36(10):817824. ( 10.1097/BRS.0b013e3181e21783) [PubMed] [CrossRef] [Google Scholar]

4. Bateman RM, Sharpe MD, Jagger JE, et al. 36th International Symposium on Intensive Care and Emergency Medicine: Brussels, Belgium. 15-18 March 2016. Crit Care. 2016;20:94. ( 10.1186/s13054-016-1208-6) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

5. Aghdasi B, Bachmann KR, Clark D, et al. Patient-reported outcomes following surgical intervention for adolescent idiopathic scoliosis: a systematic review and meta-analysis. Clin Spine Surg. 2020;33(1):2434. ( 10.1097/BSD.0000000000000822) [PubMed] [CrossRef] [Google Scholar]

6. Main CJ. The Modified Somatic Perception Questionnaire (MSPQ). J Psychosom Res. 1983;27(6):503514. ( 10.1016/0022-3999(83)90040-5) [PubMed] [CrossRef] [Google Scholar]

7. Levis B, Benedetti A, Thombs BD, Depression Screening Data (DEPRESSD) Collaboration. Accuracy of Patient Health Questionnaire-9 (PHQ-9) for screening to detect major depression: individual participant data meta-analysis. BMJ. 2019;365:l1476. ( 10.1136/bmj.l1476) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

8. Julian LJ. Measures of anxiety: State-Trait Anxiety Inventory (STAI), Beck Anxiety Inventory (BAI), and Hospital Anxiety and Depression Scale-Anxiety (HADS-A). Arthritis Care Res (Hoboken). 2011;63:S467S472. ( 10.1002/acr.20561) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

9. Asher MA, Lai SM, Glattes RC, Burton DC, Alanay A, Bago J. Refinement of the SRS-22 Health-Related Quality of Life questionnaire Function domain. Spine (Phila Pa 1976). 2006;31(5):593597. ( 10.1097/01.brs.0000201331.50597.ea) [PubMed] [CrossRef] [Google Scholar]

10. Catanzano AA, Newton PO, Shah VJ, Yaszay B, Bartley CE, Bastrom TP. From PHQ-2 to SRS-22: how a depression screening tool relates to SRS scores in patients with adolescent idiopathic scoliosis. Spine Deform. 2024;12(1):125131. ( 10.1007/s43390-023-00759-x) [PubMed] [CrossRef] [Google Scholar]

11. Voepel-Lewis T, Caird MS, Tait AR, et al. A cluster of high psychological and somatic symptoms in children with idiopathic scoliosis predicts persistent pain and analgesic use 1 year after spine fusion. Paediatr Anaesth. 2018;28(10):873880. ( 10.1111/pan.13467) [PubMed] [CrossRef] [Google Scholar]

12. Yakut Y, Pelin Z, Yagci G. An investigation of sleep profiles in individuals with idiopathic scoliosis. Sleep Sci. 2022;15(2):172178. ( 10.5935/1984-0063.20220038) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

13. Hwang SW, Pahys JM, Bastrom TP, et al. Lower SRS mental health scores are associated with greater preoperative pain in patients with adolescent idiopathic scoliosis. Spine (Phila Pa 1976). 2019;44(23):16471652. ( 10.1097/BRS.0000000000003162) [PubMed] [CrossRef] [Google Scholar]

14. Lau KKL, Kwan KYH, Cheung JPY, Law KKP, Cheung KMC. Impact of mental health components on the development of back pain in young adults with adolescent idiopathic scoliosis. Eur Spine J. 2023;32(11):39703978. ( 10.1007/s00586-023-07908-w) [PubMed] [CrossRef] [Google Scholar]

15. Rullander AC, Lundström M, Lindkvist M, Hägglöf B, Lindh V. Stress symptoms among adolescents before and after scoliosis surgery: correlations with postoperative pain. J Clin Nurs. 2016;25(7-8):10861094. ( 10.1111/jocn.13137) [PubMed] [CrossRef] [Google Scholar]

16. Anastasio AT, Farley KX, Rhee JM. Depression and anxiety as emerging contributors to increased hospital length of stay after posterior spinal fusion in patients with adolescent idiopathic scoliosis. N Am Spine Soc J. 2020;2:100012. ( 10.1016/j.xnsj.2020.100012) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

17. Theologis AA, Ailon T, Scheer JK, et al. Impact of preoperative depression on 2-year clinical outcomes following adult spinal deformity surgery: the importance of risk stratification based on type of psychological distress. J Neurosurg Spine. 2016;25(4):477485. ( 10.3171/2016.2.SPINE15980) [PubMed] [CrossRef] [Google Scholar]

18. Matamalas A, Figueras C, Pizones J, et al. How back pain intensity relates to clinical and psychosocial factors in patients with idiopathic scoliosis. Eur Spine J. 2022;31(4):10061012. ( 10.1007/s00586-022-07117-x) [PubMed] [CrossRef] [Google Scholar]

19. Duramaz A, Yılmaz S, Ziroğlu N, Bursal Duramaz B, Kara T. The effect of deformity correction on psychiatric condition of the adolescent with adolescent idiopathic scoliosis. Eur Spine J. 2018;27(9):22332240. ( 10.1007/s00586-018-5639-4) [PubMed] [CrossRef] [Google Scholar]

20. Misterska E, Glowacki M, Harasymczuk J. Personality characteristics of females with adolescent idiopathic scoliosis after brace or surgical treatment compared to healthy controls. Med Sci Monit. 2010;16(12):CR606CR615. [PubMed] [Google Scholar]

21. Wong AYL, Samartzis D, Cheung PWH, Cheung JPY. How common is back pain and what biopsychosocial factors are associated with back pain in patients with adolescent idiopathic scoliosis?Clin Orthop Relat Res. 2019;477(4):676686. ( 10.1097/CORR.0000000000000569) [PMC free article] [PubMed] [CrossRef] [Google Scholar]

22. Khan RS, Ahmed K, Blakeway E, et al. Catastrophizing: a predictive factor for postoperative pain. Am J Surg. 2011;201(1):122131. ( 10.1016/j.amjsurg.2010.02.007) [PubMed] [CrossRef] [Google Scholar]