Abstract
Background
Surgery can be psychologically challenging for pediatric patients and their families, yet the impact of pediatric orthopaedic surgery on the future development of mental health disorders (MHDs) is incompletely understood. The purpose of this study was to determine whether the rate of the 2-year cumulative risk of common MHDs differed by the type of pediatric orthopaedic procedure.
Methods
All patients aged ≤18 years undergoing pediatric orthopaedic surgery at a single academic medical center over a 10-year period were retrospectively identified by Current Procedural Terminology codes. Patients were categorized into 5 cohorts based on procedure type: lower extremity, upper extremity, spine, general, or multiple procedures. MHD-coded health care encounters (mood, anxiety, and behavioral/emotional disorders) were identified using International Classification of Diseases 10th Revision (ICD-10) codes and were assessed at baseline (12 months before surgery) and in 6-month intervals for up to 2 years after surgery. Individuals were considered to have a positive MHD outcome if they had at least 1 encounter containing an ICD-10 code for 1 of the included MHD categories. As the MHD outcome was evaluated as a cumulative risk, evidence of having MHD (ie, “yes”) was carried forward over the remaining intervals following the first interval of a positive MHD diagnosis. Risk ratios (RR) of MHDs were estimated for each interval using the presurgical period as the reference. The ratio of risk ratios (RRR), a difference-in-difference estimand, was estimated to assess differences in the 2-year rate of change in MHD risk across cohorts, adjusting for age, sex, surgery time, and the COVID-19 period.
Results
A total of 2,825 patients were included. Cohorts included lower extremity (n = 1,300; mean age: 11.4 years), upper extremity (n = 816; 8.7), spine (n = 386; 13.4), general (n = 170; 12.3), and multiple procedures (n = 153; 9.9). Presurgical baseline MHD prevalence ranged from 13% (upper extremity) to 18% (spine). All cohorts demonstrated higher cumulative MHD-coded encounter prevalence over the 2-year follow-up period. Pairwise comparisons found no meaningful difference in the rate of change in MHD risk between groups from 0 to 24 months, except for the upper extremity cohort which had a blunted increased rate of change in MHD risk compared to the lower extremity and spine cohorts.
Conclusions
At a single academic medical center, there was a similar rate of MHD risk accumulation following different pediatric orthopaedic procedure types, with the exception of a slightly blunted increased rate of MHD risk accumulation in upper extremity surgeries.
Key Concepts
-
(1)
The rate of change in the mental health disorder (MHD) risk was largely similar across procedure types, with the exception of a blunted increase in the upper extremity cohort.
-
(2)
Although younger patients had the lowest baseline prevalence of MHDs, they experienced the greatest relative increase in risk over time.
-
(3)
Although males had a higher presurgery MHD prevalence than females, females demonstrated a greater relative increase in risk over time, resulting in similar prevalence by 2 years postoperatively.
Level of Evidence
4
Keywords: Mental health diagnoses, Anxiety, Depression, Relative risk ratio, Psychosocial
Introduction
Surgery can be psychologically challenging for pediatric patients and their families, yet the impact of pediatric orthopaedic surgery on subsequent development of mental health disorders (MHDs) remains incompletely understood. Potential psychosocial stressors in this population include exposure to surgery and anesthesia, postoperative pain, hospitalization, and prolonged recovery with associated activity restrictions [1].
Population-based studies suggest an association between early surgical and anesthetic exposure and subsequent MHDs [2]. Evidence from pediatric surgical populations also supports an association between surgery and post-traumatic stress responses [3]. This distress can trigger downstream functional impairment [4]. As compared to other surgical subspecialties, orthopaedic surgery may pose unique psychosocial challenges due to pain, functional limitations, prolonged recovery, and repeated health care encounters. For example, Rabbitts et al. [5] found that nearly 20% of adolescents undergoing major musculoskeletal surgery developed chronic postsurgical pain, with presurgical depressive symptoms and poor sleep quality predicting worse outcomes. In addition, children with musculoskeletal conditions experience higher rates of anxiety, depression, behavioral disorders, and attention deficit hyperactivity disorder than their peers without such conditions, with more severe conditions further increasing these risks [6].
Despite this growing body of literature, longitudinal data on mental health trajectories following pediatric orthopaedic surgery remain limited. In a systematic review, Charla et al. note that most mental health research within pediatric orthopaedics focuses on specific diagnoses (eg, attention-deficit/hyperactivity disorder) or specific conditions (eg, adolescent idiopathic scoliosis or fractures). In contrast, few studies have investigated MHDs in more heterogenous (and younger) pediatric orthopaedic populations, and to our knowledge, none have investigated MHD trajectories across different procedure subtypes [7]. The purpose of this study was to determine whether the rate of the 2-year cumulative risk of common MHDs differed by the type of pediatric orthopaedic procedure.
Materials and methods
This retrospective cohort study was conducted at a single academic medical center in the Midwest region of the USA and was approved by the institutional review board. The electronic medical record (EMR) was queried for all encounters across the health system between January 01, 2016 and June 30, 2025. This study used an “identification” period between January 01, 2017 (to allow for a 1-year presurgical baseline period) and June 30, 2023 (to allow for up to 2 years of follow-up for the outcome) to identify the first eligible orthopaedic surgery among patients who were 18 years of age or younger by the date of the surgery. Orthopaedic patients were identified using Current Procedural Terminology (CPT) codes in the 20000-29999 range, corresponding to musculoskeletal system procedures. Each procedure was subsequently categorized into 1 of 5 cohorts by the senior author based on specific CPT codes: lower extremity, upper extremity, spine, general, or multiple (if multiple procedures across categories were performed in a single setting). General procedures broadly included foreign body removal, exploration of penetrating wounds, excision of ganglion cysts, bone grafting or biopsy, and manipulation of external fixators, among others. Fracture care was included within their respective extremity cohort. To reduce diagnostic detection bias from including patients who do not receive most of their care within our health system, this study required patients to have health care encounters at the medical center spanning from at least 1 year prior to surgery to at least 2 years following surgery [4,5].
Commonly studied pediatric mental health outcomes were identified based on the International Classification of Diseases 10th Revision (ICD-10) codes and included mood (affective) disorders (ICD-10 codes: F30.x-F39.x); anxiety, dissociative, stress-related, somatoform, and other nonpsychotic mental disorders (F40.x-F48.x); and behavioral and emotional disorders with the onset usually occurring in childhood and adolescence (F90.x-F94.x, F96.x-F98.x), excluding tic disorders (F95.x). The MHD outcome was dichotomous (ie, yes/no), and patients with multiple diagnoses were not further separated in this study. MHD diagnoses were assessed during the 1-year presurgical baseline period and across the postsurgical follow-up in 6-month intervals: 0-6, 7-12, 13-18, and 19-24 months after surgery. For the purposes of this study, the outcome was defined as a MHD-coded health care encounter rather than a clinically confirmed psychiatric diagnosis. Individuals were considered to have a positive MHD outcome if they had at least 1 encounter containing an ICD-10 code for 1 of the included MHD categories. As the MHD outcome was evaluated as a cumulative risk, evidence of having MHD (ie, “yes”) was carried forward over the remaining intervals following the first interval of a positive MHD diagnosis. For example, if a patient had evidence of an MHD diagnosis during the presurgical baseline period, they were considered to have the MHD outcome throughout the rest of the postsurgical follow-up.
Age at the time of surgery, sex, race, ethnicity, body mass index, and surgical duration (minutes) were retrieved from the EMR. A dichotomous variable was also created for statistical adjustment to indicate whether the surgery occurred before or after the onset of the COVID-19 pandemic in the USA, using January 1st, 2020, as the cutoff date.
Baseline characteristics and MHD prevalence, with 95% binomial confidence intervals (CIs), were summarized for each cohort during the presurgical period and each postsurgical 6-month interval. Risk ratios (RRs with 95% CIs) of MHDs were estimated separately for each cohort by comparing each postsurgical 6-month interval with the presurgical period to quantify changes in risk over time.
A pairwise difference-in-difference analysis was conducted to determine if the time-related changes in MHD risk from the presurgical baseline differed among surgical cohorts. Specifically, generalized linear models with repeated measures, a binomial distribution, and a log link function were constructed before and after adjustment for age (continuous), sex, surgery duration (<120 vs ≥120 minutes), and date of surgery (before vs on/after January 1st, 2020). A surgical duration of 120 minutes was used as a generalized proxy for procedural complexity. The effect of interest from this model was the interaction between the time variable (eg, postsurgical 6-month interval vs presurgical baseline) and cohort variable, which was interpreted as the ratio of risk ratios (RRR). The RRR quantifies the difference in the rate of change in the outcome between cohorts. The major strength of this quasi-experimental analytic approach is that it allows each cohort to serve as their own internal control, thus reducing bias from sample selection and confounding sources and reducing the need for matching or risk adjustment [[8], [9], [10], [11]]. The difference-in-difference design has been demonstrated to be a robust method for comparing outcomes using real-world clinical data, even among cohorts that have vastly different baseline demographic and clinical characteristics [9].
Additionally, an exploratory analysis was conducted using the same modeling framework as in the main analysis to examine the association between baseline demographics, procedural characteristics, and the cumulative risk of MHDs [12,14]. For these analyses, the prevalence of MHDs was visually presented across all postsurgical 6-month intervals; however, the RR and RRR were estimated for the final postsurgical interval (19-24 months) relative to the 12-month presurgical baseline period. To facilitate clinically meaningful comparisons, the adjusted model adjusted for the procedure type (ie, lower extremity, upper extremity, spine, general, or multiple), surgical duration (as <120 minutes vs ≥120 minutes), sex, COVID-19 (as before or on/after January 1st, 2020), and age which was grouped to reflect developmental stages: <6 years (early childhood), 6-11 years (middle childhood/school-aged), 12-14 years (early adolescence), and ≥15 years (later adolescence) [15].
Finally, because the primary analysis captured an MHD based on ≥1 claim containing a relevant ICD-10 code, which may reflect diagnostic evaluation rather than a confirmed diagnosis, we conducted a sensitivity analysis requiring ≥2 claims for the same MHD on different days within each period (ie, the 12-month presurgical baseline and each postsurgical 6-month interval). All analyses were performed using SAS version 9.4 (SAS Institute, Cary, NC, USA), and a two-tailed P value ≤ .05 was considered statistically significant.
Results
Of 5,453 patients with an eligible surgery, 2,628 patients were excluded due to incomplete preoperative or postoperative care. The final cohort included 2,825 patients, with a mean (standard deviation) age of 10.9 (4.7) years.
Most patients (about 95%) had a single surgery, while approximately 5% underwent 2 or more procedures on the same day (“multiple procedures”). The spine cohort was the oldest (13.4 [3.2] years), and the upper extremity cohort the youngest (9.9 [4.6] years) (Table 1).
Table 1.
Baseline descriptive characteristics of the study cohorts based on the type of pediatric surgery.
| Lower extremity surgery (n = 1,300) | Upper extremity surgery (n = 816) | Spine surgery (n = 386) | General pediatric orthopaedic surgery (n = 170) | Multiple procedures (n = 153) | |
|---|---|---|---|---|---|
| Age | |||||
| Mean (SD) | 11.4 (4.7) | 8.7 (4.4) | 13.4 (3.2) | 12.3 (4.0) | 9.9 (4.6) |
| Median (IQR) | 12.4 (7.7, 15.1) | 7.8 (5.3, 12.4) | 13.6 (11.4, 15.6) | 12.7 (9.6, 15.3) | 9.9 (6.2, 13.8) |
| <6 years, % (n) | 18.7 (243) | 31.6 (258) | 3.4 (13) | 7.7 (13) | 23.5 (36) |
| 6-11 years, % (n) | 27.1 (352) | 41.7 (340) | 25.1 (97) | 33.5 (57) | 41.8 (64) |
| 12-14 years, % (n) | 28.8 (374) | 15.4 (126) | 38.6 (149) | 30.0 (51) | 20.3 (31) |
| ≥15 years, % (n) | 25.5 (331) | 11.3 (92) | 32.9 (127) | 28.8 (49) | 14.4 (22) |
| Sex, % (n) | |||||
| Female | 44.4 (577) | 42.2 (344) | 63.7 (246) | 36.5 (62) | 48.4 (74) |
| Male | 55.6 (723) | 57.8 (472) | 36.3 (140) | 63.5 (108) | 51.6 (79) |
| Race, % (n) | |||||
| Black | 12.6 (164) | 6.4 (52) | 11.9 (46) | 8.8 (15) | 11.8 (18) |
| White | 78.5 (1,020) | 81.1 (662) | 78.8 (304) | 82.9 (141) | 79.7 (122) |
| Other | 8.9 (116) | 12.5 (102) | 9.3 (36) | 8.2 (14) | 8.5 (13) |
| Ethnicity, % (n) | |||||
| Non-Hispanic or Latino | 91.5 (1,189) | 91.2 (744) | 92.8 (358) | 90.0 (153) | 92.8 (142) |
| Hispanic or Latino | 8.5 (111) | 8.8 (72) | 7.3 (28) | 10.0 (17) | 7.2 (11) |
| Body mass index, kg/m2 | |||||
| Mean (SD) | 22.2 (6.5) | 19.1 (5.1) | 20.8 (5.3) | 20.8 (5.6) | 20.5 (5.5) |
| Median (IQR) | 20.7 (17.4, 25.7) | 17.4 (15.7, 21.4) | 19.6 (17.3, 22.9) | 20.2 (16.4, 23.5) | 19.2 (16.7, 22.9) |
| Missing, % (n) | 10.8 (141) | 11.8 (96) | 6.2 (24) | 5.9 (10) | 14.4 (22) |
| Surgery time (min) | |||||
| Mean (SD) | 98.9 (65.7) | 54.5 (45.5) | 258.4 (116.5) | 43.6 (35.9) | 153.7 (83.9) |
| Median (IQR) | 82 (51, 134) | 39 (23, 73) | 242 (190, 314) | 36 (24, 51) | 135 (98, 200) |
IQR, interquartile range; MHD, mental health disorder; RR, risk ratio; SD, standard deviation.
The prevalence of MHDs during the 1-year presurgical period was highest in the spine cohort (17.9%), followed by the lower extremity (16.7%), general (15.3%), multiple procedures (13.1%), and upper extremity cohorts (12.9%); differences were not statistically significant (all P > .05; Table 2; Fig. 1).
Table 2.
Crude prevalence and risk ratio (RR) of presurgical and postsurgical mental health disorders among the study cohorts based on the type of pediatric surgery.
| Lower extremity surgery (n = 1,300) | Upper extremity surgery (n = 816) | Spine surgery (n = 386) | General pediatric orthopaedic surgery (n = 170) | Multiple procedures (n = 153) | |
|---|---|---|---|---|---|
| Prevalence, % (95% CI) | |||||
| Presurgery (12 mo) | 16.7 (11.7, 21.7) | 12.9 (6.5, 19.3) | 17.9 (8.8, 26.9) | 15.3 (1.5, 29.1) | 13.1 (0.0, 27.8) |
| Postsurgery, 0-6 mo | 21.2 (16.3, 26.0) | 14.5 (8.1, 20.8) | 24.4 (15.7, 33.0) | 19.4 (5.9, 32.9) | 16.3 (1.8, 30.8) |
| Postsurgery, 7-12 mo | 23.3 (18.5, 28.1) | 16.1 (9.8, 22.3) | 26.4 (17.9, 35.0) | 22.4 (9.1, 35.6) | 17.0 (2.6, 31.4) |
| Postsurgery, 13-18 mo | 25.2 (20.5, 29.9) | 18.4 (12.2, 24.6) | 29.0 (20.6, 37.4) | 25.3 (12.3, 38.3) | 18.3 (4.0, 32.6) |
| Postsurgery, 19-24 mo | 27.3 (22.7, 31.9) | 18.9 (12.7, 25.1) | 30.8 (22.5, 39.1) | 26.5 (13.6, 39.4) | 20.3 (6.1, 34.4) |
| RR (95% CI) | |||||
| Presurgery (12 mo) | Reference | Reference | Reference | Reference | Reference |
| Postsurgery, 0-6 mo | 1.27 (1.06, 1.51) | 1.12 (0.86, 1.46) | 1.36 (1.00, 1.86) | 1.27 (0.76, 2.12) | 1.25 (0.69, 2.25) |
| Postsurgery, 7-12 mo | 1.40 (1.17, 1.66) | 1.25 (0.97, 1.61) | 1.48 (1.09, 2.01) | 1.46 (0.89, 2.41) | 1.30 (0.73, 2.33) |
| Postsurgery, 13-18 mo | 1.51 (1.27, 1.79) | 1.43 (1.11, 1.83) | 1.62 (1.20, 2.19) | 1.65 (1.02, 2.69) | 1.40 (0.79, 2.49) |
| Postsurgery, 19-24 mo | 1.64 (1.38, 1.94) | 1.47 (1.14, 1.88) | 1.72 (1.28, 2.32) | 1.73 (1.07, 2.80) | 1.55 (0.88, 2.72) |
CI, confidence interval.
Statistically significant comparisons, at a P value <.05, are presented in bold for ease of interpretation.
Figure 1.
Crude prevalence of mental health disorders based on the type of surgery at the presurgical 12-month period and postsurgical 2-year period by 6-month blocks.
By the final postsurgical interval, all surgical cohorts except multiple procedures exhibited a statistically significant increase in MHD risk over the 2-year postsurgical period, though the magnitude of increase differed. The largest significant increase was observed in the general procedures cohort, with a 73% increase in the MHD risk in the final postsurgical interval (RR = 1.73; 95% CI = 1.07-2.80; P = .026). The smallest significant increase was upper extremity, with a 47% increase in the MHD risk by the final postsurgical interval (RR = 1.47; 95% CI = 1.14-1.88; P = .003; Table 2).
The unadjusted and adjusted RRRs are presented in Table 3, which compares the 2-year rate of change (eg, slope) in MHD risk between cohorts. There were no meaningful differences in the rate of change in MHD risk between most groups from 0 to 24 months (adjusted RRRs ranged from 0.83 to 1.19, all P > .05), except for the upper extremity cohort for postsurgery intervals 0-6 and 7-12 months which had a blunted increased rate of change in MHD risk compared to the lower extremity (pairwise comparison set as upper extremity vs lower extremity, both RRR = 0.89, both P < .05) and spine (pairwise comparison set as spine vs upper extremity, 0- to 6-month interval RRR = 1.22, 7- to 12-month interval RRR = 1.18, both P < .05) cohorts.
Table 3.
The ratio of the risk ratios (RRR) of presurgical (12-months) and postsurgical (in 6-month blocks) mental health disorders among the study cohorts based on the type of pediatric surgery.
| Unadjusted | Fully adjusted | |
|---|---|---|
| RRR (95% CI) | RRR (95% CI) | |
| Postsurgery 0-6 mo vs presurgery | ||
| Upper extremity vs lower extremity | 0.89 (0.81, 0.97) | 0.89 (0.81, 0.97) |
| Spine vs lower extremity | 1.08 (0.94, 1.23) | 1.08 (0.94, 1.23) |
| General vs lower extremity | 1.00 (0.83, 1.21) | 1.01 (0.83, 1.22) |
| Multiple procedures vs lower extremity | 0.99 (0.80, 1.21) | 0.98 (0.80, 1.20) |
| Spine vs upper extremity | 1.21 (1.06, 1.39) | 1.22 (1.06, 1.39) |
| General vs upper extremity | 1.13 (0.94, 1.36) | 1.13 (0.93, 1.36) |
| Multiple procedures vs upper extremity | 1.11 (0.91, 1.37) | 1.10 (0.90, 1.33) |
| General vs spine | 0.93 (0.75, 1.15) | 0.93 (0.75, 1.14) |
| Multiple procedures vs spine | 0.92 (0.73, 1.16) | 0.91 (0.73, 1.15) |
| Multiple procedures vs general | 0.98 (0.76, 1.28) | 0.97 (0.75, 1.27) |
| Postsurgery 7-12 mo vs presurgery | ||
| Upper extremity vs lower extremity | 0.89 (0.80, 0.99) | 0.89 (0.80, 0.99) |
| Spine vs lower extremity | 1.06 (0.91, 1.23) | 1.05 (0.91, 1.23) |
| General vs lower extremity | 1.05 (0.83, 1.31) | 1.05 (0.84, 1.32) |
| Multiple procedures vs lower extremity | 0.93 (0.75, 1.16) | 0.93 (0.75, 1.15) |
| Spine vs upper extremity | 1.18 (1.01, 1.39) | 1.18 (1.01, 1.39) |
| General vs upper extremity | 1.17 (0.93, 1.48) | 1.17 (0.93, 1.47) |
| Multiple procedures vs upper extremity | 1.04 (0.83, 1.31) | 1.03 (0.83, 1.27) |
| General vs spine | 0.99 (0.77, 1.28) | 1.00 (0.77, 1.28) |
| Multiple procedures vs spine | 0.88 (0.69, 1.13) | 0.88 (0.69, 1.13) |
| Multiple procedures vs general | 0.89 (0.66, 1.20) | 0.87 (0.65, 1.18) |
| Postsurgery 13-18 mo vs presurgery | ||
| Upper extremity vs lower extremity | 0.95 (0.83, 1.08) | 0.94 (0.83, 1.07) |
| Spine vs lower extremity | 1.08 (0.91, 1.27) | 1.08 (0.91, 1.27) |
| General vs lower extremity | 1.10 (0.85, 1.41) | 1.11 (0.86, 1.43) |
| Multiple procedures vs lower extremity | 0.93 (0.73, 1.19) | 0.93 (0.73, 1.18) |
| Spine vs upper extremity | 1.14 (0.95, 1.36) | 1.14 (0.96, 1.37) |
| General vs upper extremity | 1.16 (0.89, 1.51) | 1.17 (0.90, 1.52) |
| Multiple procedures vs upper extremity | 0.98 (0.76, 1.27) | 0.98 (0.76, 1.25) |
| General vs spine | 1.02 (0.77, 1.35) | 1.03 (0.78, 1.37) |
| Multiple procedures vs spine | 0.86 (0.65, 1.14) | 0.86 (0.66, 1.13) |
| Multiple procedures vs general | 0.85 (0.60, 1.19) | 0.83 (0.60, 1.17) |
| Postsurgery 19-24 mo vs presurgery | ||
| Upper extremity vs lower extremity | 0.90 (0.78, 1.03) | 0.89 (0.78, 1.02) |
| Spine vs lower extremity | 1.05 (0.89, 1.25) | 1.05 (0.89, 1.26) |
| General vs lower extremity | 1.06 (0.81, 1.38) | 1.06 (0.82, 1.39) |
| Multiple procedures vs lower extremity | 0.95 (0.72, 1.25) | 0.95 (0.72, 1.24) |
| Spine vs upper extremity | 1.18 (0.98, 1.42) | 1.18 (0.98, 1.42) |
| General vs upper extremity | 1.18 (0.90, 1.55) | 1.19 (0.91, 1.56) |
| Multiple procedures vs upper extremity | 1.06 (0.80, 1.40) | 1.06 (0.80, 1.39) |
| General vs spine | 1.00 (0.75, 1.35) | 1.01 (0.75, 1.35) |
| Multiple procedures vs spine | 0.90 (0.66, 1.22) | 0.90 (0.67, 1.21) |
| Multiple procedures vs general | 0.90 (0.62, 1.29) | 0.89 (0.62, 1.27) |
CI, confidence interval.
Statistically significant comparisons, at a P value <.05, are presented in bold for ease of interpretation. The fully adjusted model adjusted for age (continuous), sex, surgery time (as ≥120 minutes vs <120 minutes), and date of surgery (before January 01, 2020 vs on or after January 01, 2020).
For the exploratory analysis assessing the association between baseline demographics, procedural characteristics, and the cumulative risk of MHD, the MHD prevalence in the 1-year presurgical period was lowest in the youngest group (<6 years, 4.3%) and increased progressively with age (≥15 years, 25.8%) (Table 4). There was an inverse stepwise pattern of the relative increase in MHD risk over the 2-year follow-up period by age group, where the youngest age group (<6 years) had the greatest increase in the relative risk (RR = 3.00; 95% CI = 2.67-3.37; P < .001) and the oldest age group (≥15 years) had the lowest increase in the relative risk (RR = 1.42; 95% CI = 1.27-1.59; P < .001). When comparing the 2-year rate of cumulative MHD risk across age groups after adjusting for covariates and using the youngest age group as the reference, each older age group had an RRR of 0.47-0.55 (all, P < .05), indicating that their rate of increased MHD risk was outpaced by that of the youngest age group (Supplementary Table 5). There was also a statistically significant difference in the 2-year rate of cumulative MHD risk when comparing the oldest age group (≥15 years) to the second youngest age group (6-11 years) (RRR = 0.86; 95% CI = 0.74-0.99; P = .034). When comparing sex, males had a higher presurgery MHD prevalence than females (16.5% vs 14.3%). However, females demonstrated a greater relative increase in risk over time than males (RR = 1.75 vs 1.51, both P < .001) such that males had a 14% lower rate of MHD risk accumulation (RRR = 0.86; 95% CI = 0.77-0.97; P = .014), resulting in similar prevalence by 2 year postoperatively. Finally, longer surgery duration (≥120 minutes) had higher baseline prevalence and a greater relative increase in risk over time (RR = 1.75 vs 1.53, both P < .001) such that a surgery duration ≥120 minutes had a 14% higher rate of MHD risk accumulation (RRR = 1.15; 95% CI = 1.01-1.30; P = .029) (Supplementary Table 5).
Table 4.
Exploratory analysis—Crude prevalence and risk ratio (RR) of the 12-month presurgical baseline period and the last postsurgical 6-month interval (postsurgery months 19-24) based on age group, sex, and surgery time.
| Prevalence | RR | ||
|---|---|---|---|
| Presurgery (12 mo) | Postsurgery, 19-24 mo | Postsurgery, 19-24 mo (ref: presurgery) | |
| % (95% CI) | % (95% CI) | RR (95% CI) | |
| Age group | |||
| <6 years old | 4.3 (0.0, 12.3) | 12.8 (5.1, 20.5) | 3.00 (2.67, 3.37) |
| 6-11 years old | 13.2 (7.1, 19.2) | 21.9 (16.1, 27.6) | 1.66 (1.51, 1.82) |
| 12-14 years old | 18.2 (11.6, 24.8) | 28.2 (22.0, 34.3) | 1.55 (1.40, 1.72) |
| ≥15 years old | 25.8 (19.0, 32.5) | 36.6 (30.3, 42.8) | 1.42 (1.27, 1.59) |
| Sex | |||
| Females | 14.3 (9.2, 19.3) | 25.0 (20.3, 29.7) | 1.75 (1.62, 1.89) |
| Males | 16.5 (11.9, 21.1) | 24.8 (20.5, 29.2) | 1.51 (1.40, 1.62) |
| Surgery time | |||
| <120 min | 14.6 (10.4, 18.7) | 22.3 (18.4, 26.3) | 1.53 (1.44, 1.63) |
| ≥120 min | 17.4 (11.5, 23.2) | 30.3 (24.9, 35.8) | 1.75 (1.60, 1.92) |
CI, confidence interval.
Statistically significant comparisons, at a P value <.05, are presented in bold for ease of interpretation.
To limit misclassification error, we performed a sensitivity analysis requiring ≥2 claims for MHD. While MHD prevalence estimates were slightly lower than in the primary analysis, overall patterns in the 2-year cumulative risk, RRs, and RRRs remained consistent. The only exception was that the early postsurgical difference between upper extremity and spine cohorts was no longer statistically significant (RRR = 1.05; 95% CI, 0.94-1.19).
Discussion
Orthopaedic surgery in pediatric patients can introduce major stressors, including pain, functional limitation, and the emotional burden of recovery. Although prior research has suggested that surgery and anesthesia exposure may be associated with later MHD diagnoses, particularly for large, standardized procedures like spinal fusion, little is known about mental health risk across the full range of pediatric orthopaedic procedures. The present study adds to the literature by comparing how MHD risk changes over time after different types of pediatric orthopaedic surgery. Analyzing 2,825 patients over a 10-year span, we found that the rate of change in the MHD risk in the first 2 years after pediatric orthopaedic surgery was largely similar across procedure types, with the exception of a blunted increase in the upper extremity cohort. To our knowledge, this is among the first pediatric studies to describe postoperative mental health risk trajectories across heterogenous procedure types defined by anatomic region.
It is important to note that this study was designed to address the between-cohort RRR analyses, and as such, the within-cohort analyses, including the crude prevalence estimates and RR analyses, should be interpreted purely as descriptive and noncausal. For example, this study was unable to identify if MHDs were in cessation, remission, or being effectively treated. Furthermore, the carry-forward approach can only result in no change or increase in the outcome risk, whereas MHD can fluctuate in real life. With that said, and for context, the baseline MHD prevalence in this study varied modestly across surgical cohorts, ranging from 12.9% in the upper extremity cohort to 17.9% in the spine cohort. These findings are slightly higher than the estimated prevalence of mental disorders reported in the 2024 Global Burden of Disease analysis, which reported the prevalence of any mental disorder to be 6.8% in children aged 5-9 years, 12.4% in those aged 10-14 years, and 14.0% in those aged 15-19 years worldwide [13]. Specific to pediatric orthopaedics, there is substantial adolescent idiopathic scoliosis literature on perioperative mental health. For example, Sanders et al. [16] reported a 32% rate of clinically significant distress, while Lee et al. [17] noted a 7% prevalence of psychiatric disorders. Most relevantly, a recent administrative Adolescent Idiopathic Scoliosis (AIS) study reported a preoperative MHD prevalence of 15.92% that remained stable postoperatively—a different trajectory from the 1 reported here [18]. However, these comparisons should be interpreted cautiously as the differences in MHD risk profiles are likely due to differences in sample selection, outcome definition, and how the MHD outcomes were ascertained (eg, survey, diagnostic codes) as compared to the point-prevalence values reported in many of these studies.
Although all groups in this study showed an increased cumulative risk of MHD following surgery, the magnitude of RR varied between cohorts. By 2 years postoperatively, the general procedure cohort showed the largest relative increase in risk by 73% (RR = 1.73), while the upper extremity cohort showed the smallest increase in risk by 47% (RR = 1.47). Differences in MHD risk accumulation across surgical cohorts were most apparent during the early postoperative period. Patients undergoing upper extremity procedures demonstrated a significantly lower rate of risk accumulation than those undergoing lower extremity procedures during both the 0- to 6-month and 7- to 12-month intervals. In contrast, spine patients showed a greater increase in risk than upper extremity patients during the same early postoperative intervals. No statistically significant differences were observed between spine and lower extremity cohorts, or among general and multiple procedure cohorts, at any time point. These between-group differences narrowed over time and were less apparent by 13-24 months. Finally, although the younger age group demonstrated the largest relative increase in cumulative MHD risk over time, this finding should be interpreted in context of low baseline prevalence. As a result, this may contribute to the large relative effect. While there are differences in the methodology used, prior work supports the general notion of increasing MHD risk following traumatic events, such as surgery and injuries [2,3,18].
One possible explanation for the observed differences in the relative risk across procedure types is that spine procedures may carry greater postoperative pain, functional limitation, or recovery burden than upper extremity procedures. Similarly, lower extremity procedures may require periods of nonweightbearing that affect ambulation and sports and can require an obvious gait aid (eg, crutches or wheelchair) that may make some children feel different. Indeed, in this study, the spine and lower extremity cohorts not only had the highest presurgical MHD risk but also exhibited some of the highest relative changes over the 2-year follow-up, especially in the first 6 months following surgery. Prior orthopaedic literature demonstrated that spine surgery populations experience worse baseline and postoperative pain and functional impairment than other pediatric orthopaedic surgical groups [19]. Another possible explanation is the older mean age of patients in the spine cohort. However, this study used the quasi-experimental difference-in-difference approach that inherently mitigates confounding bias of this nature and we further adjusted models for age in the analyses [[8], [9], [10], [11]]. Nevertheless, the older mean age suggests that a higher proportion of participants in the spine cohort were in a different developmental stage, ie, adolescence, which is disproportionally more susceptible to MHD risk and may contribute to the higher relative rate of MHD risk accumulation observed. Currently, there is limited evidence suggesting that psychiatric risk may vary by anatomic location of orthopaedic surgery. In 1 adult orthopaedic trauma study comparing upper and lower extremity fractures, upper extremity fractures were associated with higher risk of certain psychiatric outcomes in some analyses, although this pattern has not been established in pediatric orthopaedic populations [20].
Overall, this study represents an early phase of a larger research program focused on developing the methodology needed to compare MHD trajectories between orthopaedic and nonorthopaedic cohorts. At this stage, our primary objective was to determine whether pediatric orthopaedic procedures can be meaningfully grouped for future analyses. Based on our findings, MHC trajectories across procedure types can be meaningfully grouped. Future work incorporating control and/or nonorthopaedic surgical groups is planned to expand clinical relevance and help establish measures of absolute risk. This will also help determine whether the observed increase in MHDs over time is due to natural causes or is a result of orthopaedic intervention.
On a broader scale, our findings should increase awareness that postoperative mental health risk is not limited to major procedures such as spine surgery. Even procedures often viewed as more routine may be followed by meaningful mental health burden over time. These patterns also highlight an opportunity to support pediatric mental health through screening as pediatric orthopaedists may be the only physicians routinely following up these otherwise healthy kids as they recover from injury and/or surgery. Even if MHDs cannot be directly attributed to surgery, rates are still high, and early recognition may support timely referral to primary care, psychology, psychiatry, or other supportive services, particularly during the first 2 years after surgery.
There are several limitations to this study. First, the findings are derived from a single academic medical center, which may limit generalizability to broader pediatric orthopaedic populations in different geographic regions or countries. Second, this study relied on administrative coding rather than direct chart review to identify mental health outcomes. Accordingly, the outcome should be interpreted as an MHD-coded health care encounter rather than a clinically adjudicated psychiatric diagnosis. Prior validation studies have demonstrated that administrative claim algorithms for depression have imperfect sensitivity and positive predictive value, particularly when based on a single claim, although requiring multiple claims generally improves diagnostic validity [[21], [22], [23]]. This limitation is inherent to many large retrospective database studies. To improve the robustness of our findings, we performed a sensitivity analysis requiring 2 MHD-coded encounters on separate days, which demonstrated similar overall trends and conclusions. Regardless, some degree of outcome misclassification remains possible. Despite these measures, bias from residual confounding and incomplete data capture remain possible. Third, although the 2-year cumulative risk of MHDs increased across all study groups, this interpretation should be made with caution as this study opted for a methodology that used a “carry-forward” approach of MHD diagnoses over subsequent time intervals. This decision was made to begin understanding the potential impact of orthopaedic surgery on longitudinal MHD risk at the cohort level and drive new hypotheses to test in future studies, but this approach does not account for cessation of MHD or severity of MHD. Thus, the MHD risk estimates in this study could only increase over time rather than waver based on individual-level changes in mental health. Another limitation is the potential for differential detection bias. Children often have more health care encounters following surgery than during the preoperative period, creating additional opportunities for an MHD-coded encounter to be recorded. Although our cumulative-risk methodology intentionally used a “carry-forward” approach to characterize longitudinal MHD burden at the cohort level after the first recorded MHD-coded encounter, we were unable to quantify or adjust for differences in health care utilization across follow-up intervals. Consequently, increased health care contact may have contributed, in part, to the observed increase in cumulative MHD risk over time. However, we did perform sensitivity analyses requiring 2+ claims per time interval, with no significant change in our finding. Furthermore, it was not possible to identify MHD severity, remission, or cessation. Although prospective longitudinal designs could address some of these issues, they are often limited by feasibility and selection bias. In this setting, real-world clinical data offer a practical way to begin understanding MHD risk patterns over time in clinically relevant populations [24]. Moreover, this study did not include a nonsurgical cohort, which may have provided deeper insights into the MHD risk trajectories. However, as there has been little research on longitudinal MHD risk following pediatric orthopaedic surgery, we first needed to understand if there were variations in MHD risk trajectories based on the type of orthopaedic procedure. Such knowledge informs future study designs on how to group, or not group, orthopaedic surgeries in clinically meaningful ways to compare to other nonorthopaedic surgical cohorts. Finally, the general orthopaedic and multiple-procedure cohorts were substantially smaller than the other surgical cohorts, which may have limited statistical power to detect between-group differences. Therefore, nonsignificant findings in these cohorts should be interpreted cautiously.
In conclusion, the rate of the 2-year cumulative risk of MHDs in our study population was largely consistent across procedure types, except for the slightly blunted increase among the upper extremity surgical cohort. Furthermore, the risk trajectories may vary by age and sex, with younger patients and females showing greater relative increases over time. Future research incorporating nonorthopaedic comparison groups are needed to determine if orthopaedic surgery plays a unique role on pediatric MHD trajectories.
Additional links
Author contributions
Stanley Zhu: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Erin Nicholas: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Charlie X. Walsh: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Anya Singh: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Sanidhya Singh: Writing – review & editing, Writing – original draft, Methodology, Investigation, Data curation, Conceptualization. Daniel G. Whitney: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Matthew Stepanovich: Writing – review & editing, Supervision, Methodology, Investigation, Formal analysis, Conceptualization. Kristina Walick: Writing – review & editing, Supervision, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Alex Gornitzky: Writing – review & editing, Writing – original draft, Validation, Supervision, Methodology, Investigation, Formal analysis, Conceptualization.
Ethics approval and consent
The authors declare that no patient consent was necessary as no images or identifying information is included in the article.
Disclosure
There are no additional activities to disclose.
Declaration of the use of generative AI and AI-assisted technologies in the writing process
No generative AI or AI-assisted technologies were utilized in the preparation or writing of this manuscript.
Funding
This research received no external financial or nonfinancial support.
Declaration of competing interests
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Footnotes
Supplementary material related to this article can be found at https://doi.org/10.1016/j.jposna.2026.100431.
Supplementary material
The following are the supplementary data to this article:
References
- 1.Martins Esteves I., Pires R., Pestana-Santos M., Sampaio F., Reis Santos M. Preoperative interventions for reducing anxiety in school-age children undergoing surgery: a network meta-analysis. Cochrane Database Syst Rev. 2026;(2):CD016330. doi: 10.1002/14651858.CD016330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Ing C., DiMaggio C., Whitehouse A., Hegarty M.K., Sun M., von Ungern-Sternberg B.S., et al. Neurodevelopmental outcomes after initial childhood anesthetic exposure. Anesthesiology. 2014;120(4):840–852. doi: 10.1097/ANA.0000000000000121. [DOI] [PubMed] [Google Scholar]
- 3.Bhagat D., Pathak M., Gupta T., Rathod K., Jadhav A., Saxena R., et al. Surgical and psychosocial factors related to Pediatric Medical Traumatic Stress (PMTS) in children following surgery: a follow-up study. Pediatr Surg Int. 2025;41(1):77. doi: 10.1007/s00383-025-05975-2. [DOI] [PubMed] [Google Scholar]
- 4.Ari A.B., Peri T., Margalit D., Galili-Weisstub E., Udassin R., Benarroch F. Surgical procedures and pediatric medical traumatic stress (PMTS) syndrome: assessment and future directions. J Pediatr Surg. 2018;53(8):1526–1531. doi: 10.1016/j.jpedsurg.2017.10.043. [DOI] [PubMed] [Google Scholar]
- 5.Rabbitts J.A., Palermo T.M., Zhou C., Meyyappan A., Chen L. Psychosocial predictors of acute and chronic pain in adolescents undergoing major musculoskeletal surgery. J Pain. 2020;21(11-12):1236–1246. doi: 10.1016/j.jpain.2020.02.004. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Travis L.M., Prasad S., Khan F.A., Samineni A.V., Mei C., Grottkau B. Impact of pediatric neuropsychiatric disorders on postoperative outcomes following common nonspine orthopedic surgeries. J Pediatr Orthop. 2026 Mar 1;46(3):e288–e296. doi: 10.1097/BPO.0000000000003131. [DOI] [PubMed] [Google Scholar]
- 7.Nicholas Charla J., Horton B., Patton C.D., Bhatt A., Alvandi L.M., Gjonbalaj E., et al. Mental health and mental illness in pediatric orthopaedic surgery: a systematic review. J Pediatr Soc North Am. 2025 Sep 13;13 doi: 10.1016/j.jposna.2025.100268. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Dimick J.B., Ryan A.M. Methods for evaluating changes in health care policy: the difference-in-differences approach. JAMA. 2014;312:2401–2402. doi: 10.1001/jama.2014.16153. [DOI] [PubMed] [Google Scholar]
- 9.Zhou H., Taber C., Arcona S., Li Y. Difference-in-differences method in comparative effectiveness research: utility with unbalanced groups. Appl Health Econ Health Policy. 2016;14:419–429. doi: 10.1007/s40258-016-0249-y. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Daw J.R., Hatfield L.A. Matching and regression to the mean in difference-in-differences analysis. Health Serv Res. 2018;53:4138–4156. doi: 10.1111/1475-6773.12993. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Ryan A.M., Burgess J.F., Jr., Dimick J.B. Why we should not be indifferent to specification choices for difference-in-differences. Health Serv Res. 2015;50:1211–1235. doi: 10.1111/1475-6773.12270. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Bender R., Lange S. Adjusting for multiple testing - when and how? J Clin Epidemiol. 2001;54(4):343–349. doi: 10.1016/s0895-4356(00)00314-0. [DOI] [PubMed] [Google Scholar]
- 13.Kieling C., Buchweitz C., Caye A., Silvani J., Ameis S.H., Brunoni A.R., et al. Worldwide prevalence and disability from mental disorders across childhood and adolescence: evidence from the global burden of disease study. JAMA Psychiatry. 2024 Apr 1;81(4):347–356. doi: 10.1001/jamapsychiatry.2023.5051. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Rubin M. Do p values lose their meaning in exploratory analyses? It depends how you define the familywise error rate. Rev Gen Psychol. 2017;21(3):269–275. [Google Scholar]
- 15.American Academy of Pediatrics. Bright futures: guidelines for health supervision of infants, children, and adolescents. https://www.aap.org/en/practice-management/bright-futures/bright-futures-materials-and-tools/bright-futures-guideliens-and-pocketguide.com
- 16.Sanders A.E., Andras L.M., Iantorno S.E., Hamilton A., Choi P.D., Skaggs D.L. Clinically significant psychological and emotional distress in 32% of adolescent idiopathic scoliosis patients. Spine Deform. 2018 Jul-Aug;6(4):435–440. doi: 10.1016/j.jspd.2017.12.014. [DOI] [PubMed] [Google Scholar]
- 17.Lee S.B., Chae H.W., Kwon J.W., Sung S., Lee H.M., Moon S.H., et al. Is there an association between psychiatric disorders and adolescent idiopathic scoliosis? A large-database study. Clin Orthop Relat Res. 2021 Aug 1;479(8):1805–1812. doi: 10.1097/CORR.0000000000001716. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Abu-Sahra M.S., Karakash W.J., Kistler N., David McCavitt B.A., Henry Avetisian M.S., Brandon Yoshida M.D., et al. Mental health disorders in adolescent idiopathic scoliosis surgery: prevalence, postetoperative outcomes, and opioid use. Spine J. 2026;26(3):594–601. doi: 10.1016/j.spinee.2025.10.006. [DOI] [PubMed] [Google Scholar]
- 19.Spiwak R., Gawaziuk J.P., Chung D., Comaskey B., Cristall N., Chateau D., et al. Mental and physical health disorders following paediatric traumatic injury: a population-based longitudinal study in Manitoba, Canada. BMJ Open. 2025 Mar 5;15(3) doi: 10.1136/bmjopen-2024-097564. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Varlotta C., Fernandez L., Manning J., Wang E., Bendo J., Fischer C., et al. Evaluation of health-related quality of life improvement in patients undergoing spine versus adult reconstructive surgery. Spine (Phila Pa 1976) 2020 Sep 15;45(18):E1179–E1184. doi: 10.1097/BRS.0000000000003588. [DOI] [PubMed] [Google Scholar]
- 21.Chen K., Hynes K.K., Dirschl D., Wolf J.M., Strelzow J.A. Depression, anxiety, and post-traumatic stress disorder following upper versus lower extremity fractures. Injury. 2024 Mar;55(3) doi: 10.1016/j.injury.2023.111242. [DOI] [PubMed] [Google Scholar]
- 22.Young J.C., Conover M.M., Funk M.J. Measurement error and misclassification in electronic medical records: methods to mitigate bias. Curr Epidemiol Rep. 2018 Dec;5(4):343–356. doi: 10.1007/s40471-018-0164-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Fiest K.M., Jette N., Quan H., St Germaine-Smith C., Metcalfe A., Patten S.B., et al. Systematic review and assessment of validated case definitions for depression in administrative data. BMC Psychiatry. 2014;14:289. doi: 10.1186/s12888-014-0289-5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Doktorchik C., Patten S., Eastwood C., Peng M., Chen G., Beck C.A., et al. Validation of a case definition for depression in administrative data against primary chart data as a reference standard. BMC Psychiatry. 2019 Jan 7;19(1):9. doi: 10.1186/s12888-018-1990-6. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.