Dodwell E, Snyder B, Wright J. Off-Label Use of Bone Morphogenetic Proteins in Pediatric Spinal Arthrodesis. JAMA. 2012;308(14):1429-1432. doi:10.1001/jama.2012.12929
Letters Section Editor: Jody W. Zylke, MD, Senior Editor.
Author Affiliations: Hospital for Special Surgery, New York, New York (Dr Dodwell; email@example.com); Hospital for Sick Children, Toronto, Ontario, Canada (Dr Snyder); and Children's Hospital Boston, Boston, Massachusetts (Dr Wright).
To the Editor: Arthrodesis of the spine is frequently performed in children. Although nonunion occurs frequently in adults, children rarely experience nonunion.1 We are aware of no evidence to support the need for augmentation beyond instrumentation and autograft bone grafting in pediatric spinal arthrodesis.
Bone morphogenetic proteins (BMPs) are approved for limited use in adults when healing may be suboptimal.2,3 In addition to concerns about possible carcinogenesis, complications include wound dehiscence, spinal stenosis, and respiratory complications.4 The US Food and Drug Administration has not approved the use of BMPs in children.
We determined the prevalence of BMP use, associated complications, costs, and potential predictors of use in pediatric spinal arthrodesis in the United States.
The Kidś Inpatient Database, Healthcare Cost and Utilization Project, a sample of 4121 US hospitals containing 10% of uncomplicated births and 80% of complicated pediatric admissions representing approximately 30% of total pediatric admissions, is weighted to allow calculation of national estimates.5 We included children aged 18 years or younger who had undergone primary or revision spine arthrodesis in 2009 using International Classification of Diseases, Ninth Revision, Clinical Modification diagnoses and procedure codes as previously described by Cahill et al.6
SAS version 9.2 (SAS Institute Inc) was used, accounting for the complex survey design to calculate appropriate standard errors and P values and to provide nationally representative weighted estimates of prevalence of BMP use, in-hospital complications, length of stay, and in-hospital costs. Costs were estimated from supplied charges data using hospital-specific cost-to-charge ratios from the Healthcare Cost and Utilization Project.
Multivariable logistic regression was used to determine the independent association of diagnoses, comorbidities, demographic factors, and insurance status with BMP use (Table). Complications assessed included medical, neurological, wound healing, infectious, and those related to breathing and dysphagia.6 The institutional review board at the Hospital for Sick Children approved the study and waived participant consent.
In 2009, 8289 pediatric spinal arthrodeses were in the Kidś Inpatient Database. Nationally, BMP was estimated to be used in 9.2% (95% CI, 7.3%-11.0%; unweighted n = 771) of cases. The estimated prevalence of in-hospital complications in those who received BMP was 3.0% (95% CI, 2.0%-4.1%; unweighted n = 24) and in those who did not receive BMP was 3.6% (95% CI, 3.0%-4.2%; unweighted n = 271) (Rao-Scott χ2 = 0.74; P = .39 for comparison across BMP groups).
The median total in-hospital adjusted costs for patients receiving BMP was $47 136 (interquartile range [IQR], $30 692-$73 848) and in those not receiving BMP was $43 126 (IQR, $31 246-$59 849) (P < .001). Adjusted analysis of log-transformed costs showed that surgeries using BMP were 19% (95% CI, 10%-28%) more costly than those not using BMP. Median length of stay for patients receiving BMP was 4.6 days (IQR, 3.2-6.9 days) and for those not receiving BMP was 4.6 days (IQR, 3.5-6.1 days) (P = .70).
Use of BMP was associated with older age, lumbosacral arthrodeses, fewer vertebrae fused, spondylolisthesis, neurofibromatosis, revision fusions, and surgeries performed in the Midwest. Use of BMP was less frequent in idiopathic scoliosis, specialized pediatric hospitals, and in patients with Medicare or Medicaid health coverage (Table).
Given the lack of indication for augmentation of pediatric spinal arthrodesis, the use of BMP in 9.2% of patients is surprising. Although no difference in the rate of in-hospital complications was demonstrated, most complications previously reported with BMP are late complications and would not be expected to occur during hospitalization. While it is possible that some unmeasured factors account for the increase in costs associated with BMP use, our adjusted analysis showed that surgeries using BMP were 19% more costly than those that did not involve BMP use. Although BMP was used for diagnoses considered higher risk for nonunion, such as neurofibromatosis, BMP was still used frequently in idiopathic scoliosis and for other low-risk diagnoses.
This study was limited by lack of longitudinal data in the Kidś Inpatient Database, as well as lack of details on dosage and type of BMP. Use of BMP should not be routine in pediatric spine arthrodesis until it has been shown to be safe and beneficial.
Author Contributions: Dr Dodwell had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study concept and design: Dodwell, Snyder, Wright.
Acquisition of data: Dodwell.
Analysis and interpretation of data: Dodwell, Snyder, Wright.
Drafting of the manuscript: Dodwell, Wright.
Critical revision of the manuscript for important intellectual content: Dodwell, Snyder, Wright.
Statistical analysis: Dodwell.
Study supervision: Snyder, Wright.
Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Dodwell reported receiving a grant from the Ruth Jackson Orthopedic Society/Orthopedic Education and Research Fund. Dr Wright reported receiving reimbursement for travel expenses incurred as a visiting professor. Dr Snyder did not report any disclosures.
Funding/Support: The source of the data, the Kids' Inpatient Database, is part of the Healthcare Cost and Utilization Project sponsored by the Agency for Healthcare Research and Quality.
Role of the Sponsor: The Agency for Healthcare Research and Quality had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.
Additional Contributions: We acknowledge Charles Victor, MSc, PStat (Institute for Clinical Evaluative Sciences and Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Ontario, Canada), for his assistance with the statistical analysis. Mr Victor received compensation for his work on this project.