Following posterior spinal fusion for adolescent idiopathic scoliosis, patients with asthma are at increased odds of pulmonary adverse events

Abstract

Background

Posterior spinal fusion (PSF) is a highly successful treatment for adolescent idiopathic scoliosis (AIS). Asthma is one of the most common illnesses affecting young patients and has been shown to affect outcomes in orthopedic procedures. The correlation of asthma and postoperative adverse outcomes following PSF had not been well characterized.

Methods

Adolescent patients undergoing PSF for AIS between 2015 and 2023 were identified using the national administrative PearlDiver M165 database. Those with versus without asthma were matched 1:4 based on demographic and comorbid factors and compared for incidence of 90-day postoperative adverse events using multivariable logistic regression. Revision surgery within 5 years was compared using Kaplan–Meier analysis and the log-rank test.

Secondary multivariable analysis was performed comparing odds of 90-day postoperative adverse events for different levels of asthma severity (mild intermittent, mild persistent, moderate/severe) relative to non-asthma patients.

Results

Of 10,196 PSF patients meeting study criteria, asthma was noted for 1,616 (15.8%). After matching, those with asthma were at higher odds of respiratory failure (odds ratio [OR] 1.98, p < .001), pneumonia (OR 1.72, p = .004), and atelectasis (OR 1.52, p = .002), as well as emergency department visit (OR 1.66, p < .001) and readmissions (OR 1.58, p < .001). Differences were not identified for non-pulmonary adverse events. As asthma severity increased from mild intermittent, to mild persistent, to moderate/severe, odds ratios incrementally increased pneumonia (OR not significant, not significant, 4.52, respectively), respiratory failure and atelectasis (OR not significant, not significant, 2.38, respectively) (p < .05 for each listed).

Conclusion

Overall, asthma is a common comorbidity for adolescent patients undergoing PSF for AIS and is associated with higher odds of postoperative pneumonia, respiratory failure, and atelectasis. These adverse outcomes increase with increased asthma severity. These findings may help in patient counselling and expectations for postoperative management.

Introduction

Posterior scoliosis fusion (PSF) may be considered for larger curves associated with adolescent idiopathic scoliosis (AIS) [[1], [2], [3]]. While PSF is typically well-tolerated with relatively low morbidity [4], efforts are made to identify risk factors for adverse events in order to optimize surgical planning and minimize perioperative morbidities.

Several comorbidities have been identified as risk factors for adverse events following PSF for AIS. For example, Rudic et al. and Basques et al. used large databases to demonstrate that obesity significantly increased odds of experiencing any adverse event following PSF, including strong associations with surgical site infection and post-operative readmission [5,6]. Carreon et al. performed a study using a multicenter database and reported patients with renal disease were 7.9-fold more likely to experience an adverse event after surgical correction of AIS. Finally, Ramos et al. [7] studied AIS complications using the Nationwide Inpatient Sample database and also identified anemia, hypertension, and hypothyroidism as risk factors for complications.

Specific to pulmonary complications, Elsamadicy et al. [8] used the American College of Surgeons National Surgical Quality Improvement Program pediatric database to demonstrate that pediatric patients with asthma, chronic lung disease, structural pulmonary or airway abnormalities, or requiring airway support had significantly longer hospital stays following PSF for AIS. Despite this, asthma has not been independently examined in this surgical context.

Asthma is a common and clinically relevant comorbidity in this population. Overall, an estimated 11.5% of adolescents in the United States have self-reported asthma [9]. Characterized by chronic airway inflammation and remodeling, asthma leads to symptoms ranging from shortness of breath, wheezing and cough to respiratory distress and, in severe cases, death [10]. Furthermore, asthma has been shown to influence surgical outcomes. Pediatric patients with asthma have been found to have a significantly greater risk of postoperative pneumonia following inpatient surgery in general [11], while children undergoing adenotonsillectomy were at greater risk of respiratory complications, including need for mechanical ventilation [12].

In the orthopedic context, while not specific to asthma, preoperative lung disease has been identified as a risk factor for postoperative pulmonary complications following spinal fusion for idiopathic and congenital scoliosis [13,14]. In adult populations, asthma has been associated with elevated odds of 90-day postoperative complications after total knee arthroplasty, including pulmonary and other medical adverse events like sepsis and venous thromboembolism [15]. Notably the same study demonstrated the severity of asthma correlated with increasing risk of adverse events, highlighting the importance of preoperative optimization to mitigate to surgical risk for patients with asthma [15].

Taken together, asthma is a common condition that may be present for patients undergoing PSF for AIS. The present study aimed to leverage a national, administrative database to evaluate the association between asthma and adverse events in this context.

Methods

Database and cohorts

The present study leveraged the large sample size of the January 1, 2010 to October 31, 2022 M165 PearlDiver Mariner Patient Claims Database (PearlDiver Technologies). This is a commercially available administrative US database frequently used in studies investigating complications following spine surgery [[16], [17], [18], [19]]. Since the data in the database is deidentified and aggregated, it is compliant with the Health Insurance Portability and Accountability Act and our Institutional Review Board deemed studies using this database as exempt from review.

Adolescent patients with AIS were identified using International Classification of Diseases of the Ninth and Tenth Revision (ICD-9, ICD-10). The following ICD codes were used: ICD-9-D-73730, ICD-10-D-M41122, ICD-10-D-M41123, ICD-10-D-M41124, ICD-10-D-M41125, ICD-10-D-M41126, ICD-10-D-M41127, ICD-10-D-M41129, ICD-10-D-M4120, ICD-10-D-M4122, ICD-10-D-M4123, ICD-10-D-M4124, ICD-10-D-M4125, ICD-10-D-M4126, ICD-10-D-M4127. AIS patients undergoing PSF were then identified using the Current Procedural Terminology (CPT) codes 22843 and 22844. Patients were excluded for the following criteria: age <10, age >21, <90 days of follow-up in the database, concurrent pelvic fixation surgery, or history of trauma, neoplasm, or infectious diagnosis within 90 days before PSF. Patient factors extracted included age, sex, obesity, tobacco use history, number of levels treated (7–12, or >13), length of stay, and Elixhauser Comorbidity Index (ECI, a measure of comorbidity burden) [20].

PSF patients with prior history of asthma at the time of surgery were similarly identified with ICD-9, ICD-10 codes. For sub-analyses, asthma of varying severity was also determined based on ICD coding for mild intermittent asthma, mild persistent asthma, moderate asthma, and severe asthma based on prior literature of asthma patients using this same administrative database [15].

Postoperative outcomes

The incidence of individual and aggregate adverse events within 90 days of PSF was determined using ICD coding, consistent with other database studies [21,22]. Severe adverse events were defined as the occurrence of at least one of the following: cardiac events (cardiac arrest/myocardial infarction), venous thromboembolism (deep vein thrombosis/pulmonary embolism), sepsis, or surgical site infection (SSI). Minor adverse events were defined as the occurrence of at least one of the following: urinary tract infection (UTI), wound dehiscence, transfusion, or hematoma, or acute kidney injury (AKI). Pulmonary adverse events were defined as the presence of any of the following complications: respiratory failure, pneumonia atelectasis, or pleural effusion. The incidence of any adverse event was recorded if there was the occurrence of any of the above adverse events. Emergency department (ED) visits and hospital readmissions within 90 days of surgery were also determined.

Reoperations within 5 years were also determined. These cases were identified based on CPT codes 22010, 22015, 22558, 22612, 22830, 22842, 22850, 22852, and 22862.

Data analyses

Differences in patient demographic characteristics between asthma and control PSF patients were determined using Student’s t-tests and chi-squared tests. Matching between asthma and non-asthma patients was done in a 1:4 ratio based on age, sex, comorbid obesity, tobacco use history, and ECI using the PearlDiver MATCH function, an exact match methodology. A 1:4 ratio was chosen to increase control sample size and has been frequently used for national database studies [16,19,21,[22], [23]].

Perioperative adverse events were compared between the matched asthma and non-asthma populations. Chi-squared tests were used in univariable analyses, while multivariable regressions controlling for age, sex, and ECI were used to calculate odds ratios (OR) and 95% confidence intervals (95% CI) of 90-day adverse events. Kaplan–Meier analysis with log-rank test was used to compare 5-year reoperation rates.

For the sub analysis based on asthma severity, adverse events of significance from previous analyses (aggregated pulmonary adverse events, pneumonia, respiratory failure, atelectasis, ED visits, and hospital readmissions) were assessed for each of these groups. Each asthma severity was compared to the non-asthma group using multivariable analysis controlling for age, sex, and ECI.

Statistical analyses were performed within the PearlDiver software. Figs. and tables were created using Microsoft PowerPoint (Microsoft Corporation) or GraphPad Prism version 10 (GraphPad Software). Significance was defined using a significance level of p < .05.

Results

Study cohort

There were 10,196 PSF for AIS patients meeting the study inclusion/exclusion criteria, of which asthma was diagnosed for 1,616 (15.8%). The asthma cohort was younger, more likely to be male, more likely to be obese, more likely to have more levels fused, more likely to have tobacco use history, and had higher ECI than non-asthma patients (left columns of Table 1). Before matching, patients with asthma had significantly longer average length of stay compared to non-asthma patients (11.5 ± 19.8 days vs 9.7 ± 9.81 days, p = .015).

Table 1.

Descriptive characteristics of matched and unmatched AIS patients undergoing PSF with and without asthma

	Unmatched population			1:4 matching
	No asthma	Asthma	p-value	No asthma	Asthma	p-value
N (%)	8,580	1,616		3,693	953
Age (mean ± SD)	14.55 ± 2.68	13.08 ± 3.36	p <.001	14.18 ± 2.51	14.20 ± 2.54	p = .821
Sex			p <.001			p = .630
Male (%)	2,158 (25.2%)	529 (32.7%)		965 (26.1%)	257 (27.0%)
Female (%)	6,422 (74.8%)	1,087 (67.3%)		2,728 (73.9%)	696 (73.0%)
ECI (Mean ± SD)	1.59 ± 1.71	3.27 ± 2.15	p <.001	2.41 ± 1.39	2.46 ± 1.44	p = .273
Levels Fused			p <.001			p = .127
7-12 (%)	5,369 (62.8%)	909 (56.2%)		2,151 (58.2%)	582 (61.1%)
13+ (%)	3,211 (37.4%)	707 (43.8%)		1,542 (41.8%)	371 (38.9%)
Obesity (%)	917 (10.7%)	313 (19.4%)	p <.001	446 (12.1%)	111 (11.6%)	p = .758
Tobacco Use (%)	180 (2.1%)	51 (3.2%)	p =.011	54 (1.5%)	13 (1.4%)	p = .941

After 1:4 matching of asthma to non-asthma patients, there were 953 asthma patients and 3,693 control patients. The two matched cohorts no longer had significant differences in the above characteristics for which they were matched (right columns of Table 1). The average length of stay of the two matched groups vs equivalent (10.2 ± 13.5 days vs 10.0 ± 11.2 days, p = .840). The median time duration between asthma diagnosis and PSF in the unmatched asthma cohort was 737.5 days, while in the matched asthma cohort the median was 475.0 days.

Postoperative outcomes

The results of univariable and multivariable analysis of asthma versus non-asthma patients are shown in Table 2 and Fig. 1. For 90-day aggregated any adverse event, aggregated severe adverse events, aggregated minor adverse events, and each of the individual adverse events in those categories, there were no differences by univariate or multivariate analysis.

Table 2.

Univariable and multivariable comparison of 90-day risk of adverse events following posterior spinal fusion for AIS patients with relative to without asthma

	Univariable			Multivariable, controlling for age, sex, ECI
	No asthma (n = 3,693)	Asthma (n = 953)	p-value	OR (95% CI) (n = 4,646)	p-value
Any adverse events	724 (19.6%)	212 (22.2%)	.077	1.16 (0.97, 1.39)	.098
Severe events	111 (3.0%)	37 (3.9%)	.204	1.26 (0.85, 1.85)	.249
Cardiac events	<11 (<0.3%)	<11 (<1.2%)	.452	2.22 (0.53, 9.37)	.276
Venous thromboembolism	26 (0.7%)	11 (1.2%)	.234	1.54 (0.75, 3.17)	.242
Sepsis	38 (1.0%)	<11 (1.2%)	1	0.95 (0.47, 1.94)	.888
Surgical site infection	60 (1.6%)	16 (1.7%)	1	0.95 (0.53, 1.68)	.848
Minor events	335 (9.1%)	85 (8.9%)	.934	0.97 (0.76, 1.25)	.828
Urinary tract infection	144 (3.9%)	43 (4.5%)	.444	1.15 (0.81, 1.64)	.426
Wound dehiscence	75 (2.0%)	19 (2.0%)	1	0.96 (0.58, 1.60)	.883
Transfusion	178 (4.8%)	43 (4.5%)	.755	0.93 (0.66, 1.31)	.680
Hematoma	23 (0.6%)	<11 (<1.2%)	.909	0.82 (0.31, 2.17)	.694
Acute kidney injury	17 (0.5%)	<11 (<1.2%)	.426	0.43 (0.10, 1.85)	.255
Pulmonary adverse events	406 (11.0%)	133 (14.0%)	.013	1.30 (1.05, 1.62)	.017
Respiratory failure	124 (3.4%)	61 (6.4%)	<.001	1.98 (1.42, 2.75)	<.001
Pneumonia	99 (2.7%)	44 (4.6%)	.003	1.72 (1.19, 2.49)	.004
Atelectasis	211 (5.7%)	81 (8.5%)	.002	1.52 (1.16, 2.00)	.002
Pleural effusion	133 (3.6%)	31 (3.3%)	.674	0.88 (0.59, 1.32)	.535
ED visits	786 (21.3%)	294 (30.8%)	<.001	1.66 (1.41, 1.95)	<.001
Readmissions	1,479 (40.0%)	486 (51.0%)	<.001	1.58 (1.36, 1.82)	<.001

Bold values indicate significant p-values less than 0.05.

OR, odds ratio (95% confidence interval); ED, emergency department.

Fig. 1.

Odds ratio and 95% confidence interval for postoperative complications for matched asthma patients relative to control.

Conversely, there were significant differences for most of the pulmonary related adverse events. On multivariable analysis, asthma patients had significantly higher odds of aggregate pulmonary adverse events (OR 1.30, p = .017) compared to control patients (Table 2 and Fig. 1). Asthma patients also had significantly higher odds of the following individual adverse events within 90 days of PSF: respiratory failure (OR 1.98, p < .001), pneumonia (OR 1.72, p = .004), and atelectasis (OR 1.52, p = .002).

On multivariable analysis, asthma patients also had significantly higher odds of ED visits (OR 1.66) and hospital readmissions (OR 1.58) within 90 days of PSF compared to non-asthma patients (p < .001 for both). Kaplan–Meier survival analysis and log-rank test showed that the 5-year survival rate to subsequent spinal surgery was equivalent in both groups (Fig. 2).

Fig. 2.

Kaplan–Meier survival curve of reoperations following posterior spinal fusion for matched asthma and non-asthma patients.

Asthma severity sub-analysis

Asthma patients were divided for sub-group analysis based on severity. Within the matched asthma cohort, the number of patients with mild intermittent asthma was 517, the number of mild persistent asthma patients was 387, the number of patients with moderate asthma was 139, and the number of patients with severe asthma was 16. Due to the low number of patients with severe asthma, in all subsequent analysis, they were combined with moderate asthma patients for a sub-cohort of moderate/severe asthma that was 155 patients.

Patients with mild, intermittent asthma had significantly higher odds of ED visits (OR 1.63, p < .001), and readmissions (OR 1.98, p < .001) compared to non-asthma patients, but did not have higher odds of aggregate pulmonary adverse events (p = .380), respiratory failure (p = .301), pneumonia (p = .120) or atelectasis (p = .249) (left column of Table 3, Fig. 3). Patients with mild, persistent asthma had significantly higher odds of respiratory failure (OR 2.90, p = .006), ED visits (OR 1.77, p = .015), compared to non-asthma patients, but did not have higher odds of aggregate pulmonary adverse events (p = .680), pneumonia (p = .257), atelectasis (p = .237), or readmissions (p = .189) (middle column of Table 3). Patients with moderate/severe asthma had significantly higher odds of aggregate pulmonary adverse events (OR 2.32, p = .009), pneumonia (OR 4.52, p < .001), atelectasis (OR 2.38, p = .027), and ED visits (OR 4.24, p < .001) compared to non-asthma patients, but did not have higher odds of respiratory failure (p = .089) or readmissions within 90 days (p = .824) (right column of Table 3). The incremental increase in adverse outcomes in patients of increasing asthma severity is depicted in Fig. 3.

Table 3.

Multivariable comparison of 90-day risk of significant adverse events following PSF for AIS patients stratified by asthma severity relative to patients without asthma

	Mild intermittent asthma (n = 517)		Mild persistent asthma (N = 187)		Moderate/severe asthma (N = 155)
	OR (95% CI)	p-value	OR (95% CI)	p-value	OR (95% CI)	p-value
Pulmonary adverse events	1.20 (0.80, 1.79)	p = .380	1.14 (0.60, 2.16)	p = .680	2.32 (1.23, 4.36)	p =.009
Respiratory failure	1.43 (0.73, 2.80)	p = .301	2.90 (1.36, 6.18)	P =.006	2.34 (0.88, 6.21)	p =.089
Pneumonia	1.71 (0.87, 3.35)	p = .120	1.74 (0.67, 4.51)	p = .257	4.52 (1.94, 10.58)	p <.001
Atelectasis	1.34 (0.81, 2.22)	p = .249	1.57 (0.74, 3.33)	p = .237	2.38 (1.10, 5.12)	p =.027
ED visits	1.63 (1.22, 2.17)	p <.001	1.77 (1.11, 2.81)	p =.015	4.24 (2.51, 7.17)	p <.001
Readmissions	1.98 (1.53, 2.57)	P <.001	1.33 (0.87, 2.04)	p = .189	0.94 (0.56, 1.59)	p = .824

Bold values indicate significant p-values less than .05.

Fig. 3.

Odds ratio for pulmonary complications progressively increases as the severity of asthma increases.

Discussion

While PSF is a common and successful procedure to treat AIS, it can be associated with postoperative complications. Although asthma is a common comorbidity in adolescent patients, there has been a paucity of literature describing postoperative complications following PSF for patients with asthma. The present study utilized a national administrative database to compare the incidence of adverse events following PSF in patients with versus without asthma. The present study demonstrates that asthma patients face significantly higher rates of respiratory failure, pneumonia, atelectasis, ED visits, and hospital readmissions within 90 days of surgery compared to non-asthma patients.

The findings of higher pulmonary-related complications following PSF in asthma patients are consistent with much of the existing literature and mechanisms of asthma. The hyperresponsiveness of the airways in asthma can be exacerbated by some of the anesthetic stimuli during a surgical operation, such as the drugs and the endotracheal intubation process [24]. In particularly marked cases, scoliosis itself can have detrimental effects on respiratory function due to anatomical distortion that can obstruct the airway [25,26]. As patients with asthma had higher odds of these complications compared to control AIS patients, it appears that asthma is an additional risk factor for postoperative respiratory complications. Given the high morbidity of postoperative pulmonary complications [27], it is important to have close collaboration with pulmonologists and pediatricians to ensure that the disease state is well understood and appropriate tests, such as pulmonary function tests/spirometry, have been performed if indicated.

The present study also found that patients with asthma had higher odds of pneumonia, while not having higher odds of any other infectious complications, like SSI or UTI. This suggests that the slightly increased risk of pneumonia is not due to a generalized increased susceptibility to infections or immunodeficiency in asthma patients, but something specific about asthma and the lungs. The association between asthma and pneumonia has been well documented in previous literature, with a variety of proposed reasons [11,28]. While inhaled corticosteroid use is associated with increased risk of pneumonia in patients with chronic obstructive pulmonary disease [29], studies in patients with asthma have shown mixed results [30,31]. The increased risk of pneumonia in asthma patients is also thought to be related to impaired clearance of pathogenic bacteria within the diseased airways [32]. In other lumbar spine procedures such as posterior lumbar fusion, pneumonia has been shown to have significant morbidity and mortality risk highlighting the importance of awareness of this risk in asthma patients and early interventions for signs of pneumonia [33].

Asthma patients were found to have significantly higher odds of ED visits and readmissions within 90 days of PSF. These findings align with a previous finding of asthma as a significant predictor for unplanned hospital readmission after non-cardiac surgery [34]. Encouragingly, the present study also found that the survival to additional spinal surgery for asthma patients was equivalent to control patients, indicating that for patients with AIS in which PSF is indicated surgery should be pursued once the above risks are optimized for. While asthma was unlikely to affect implant longevity or curve correction, given the high burden and morbidity of revision surgery, it is reassuring to discover that there was no unexplained relationship between asthma and revision surgery.

A secondary analysis was performed with asthma patients stratified by severity. Three mutually exclusive groups of mild intermittent, mild persistent, and moderate/severe asthma were compared to non-asthma patients. As the severity of asthma increased, the association for many adverse events including aggregate pulmonary adverse events, respiratory failure, atelectasis, pneumonia, and ED visits increased. Patients with milder forms of asthma did not have higher odds of aggregate pulmonary adverse events or atelectasis compared to moderate/severe asthma patients. This relationship between asthma severity and risk of complications provides further evidence of the contribution of asthma to postoperative complications following PSF. These findings echo the results found in similar severity analysis in total knee arthroplasty analysis [15]. Interestingly, odds of readmission were highest in patients with mild intermittent asthma, despite this group having the lowest odds of other complications. This may suggest that many of the complications affecting patients with more severe forms of asthma are related to the index hospitalization. Alternatively, this may represent overly cautious postoperative management, or inaccuracies with coding for hospital readmissions in patients with more severe asthma.

There are several limitations to the present study. Like all retrospective studies based on administrative data, the results are dependent on the accuracy and specificity of the data encoded. As an example, not all asthma patients had coding to designate the severity of their asthma, excluding 166 patients (16.2%) from secondary analysis, which may result in potential selection bias. The low sample sizes for certain severities of asthma may have resulted in type II error. The wide confidence intervals also may create difficulties in interpretation. In addition, it is not possible to determine the type of anesthesia used during PSF. Other important surgical details, such as intraoperative time, intraoperative bleeding, and surgical approach could also not be determined, each of which could affect the incidence of adverse events [35,36]. Finally, management of asthma and adherence to treatment could not be determined from the database.

Overall, asthma is a relatively common comorbidity in adolescent patients undergoing PSF for AIS and it increases the risk of several pulmonary-related complications. These risks increase with the severity of asthma, but do not affect the risk for needing further spinal procedures. These findings are important in providing preoperative counseling for patients with asthma and should be combined with adequate asthma surveillance, including a recent exam by a pulmonologist/pediatrician and appropriate pulmonary function testing. Outcomes will be optimized with proper collaboration and vigilance between anesthesiologists, surgeons, and patients’ primary asthma physicians.

CRediT authorship contribution statement

Lucas Y Kim: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Project administration, Validation, Visualization, Writing – original draft, Writing – review & editing. Julian Smith-Voudouris: Conceptualization, Formal analysis, Methodology, Visualization, Writing – original draft. Scott J Halperin: Conceptualization, Validation, Writing – review & editing. Dominick Tuason: Conceptualization, Validation, Writing – review & editing. Jonathan N Grauer: Conceptualization, Methodology, Project administration, Validation, Writing – original draft, Writing – review & editing.

Declarations of competing interests

Lucas Kim: Research reported in this publication was supported by the National Institute of Aging of the National Institutes of Health under Award Number T35AG049685 and Richard K. Gershon Endowed Medical Student Research Fellowship. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institute of Aging of the National Institutes of Health and Richard K. Gershon Endowed Medical Student Research Fellowship. Julian Smith-Voudouris: Personal funding supported by Yale School of Medicine Medical Student Research Fellowship.Scott J Halperin: None. Dominick Tuason: Member of Quality, Safety, Value Committee for the Scoliosis Research Society, Member of Quality, Safety, Value Committee and JEDI Committee for Pediatric Orthopedic Society of North America, Consultant for OrthoPediatrics. Jonathan N Grauer: Editor-in-Chief of North American Spine Society Journal, Deputy Editor of Journal of American Academy of Orthopedic Surgeons, past member of the Board of Directors of North American Spine Society.