Letter to the Editor. Minimally invasive spine surgery improves outcomes in the obese population

Lei Zhao MD, PhD1, Chao Cheng MD, MS1, Wei Zuo MD, MS1, Peng Wang MD, PhD1, and Weixin Li MD, PhD1
View More View Less
  • 1 Tangdu Hospital, The Fourth Military Medical University, Xi’an, China
Online Publication Date:
24 Dec 2021
Page Range:
1038
Volume/Issue:
Volume 36: Issue 6
DOI link:
https://doi.org/10.3171/2021.10.SPINE211281
Free access

TO THE EDITOR: We read with great interest the article in your journal by Carroll et al.1 regarding the outcomes of minimally invasive spine surgery (MIS) in obese patients (Carroll AH, Dowlati E, Molina E, et al. Does minimally invasive spine surgery improve outcomes in the obese population? A retrospective review of 1442 degenerative lumbar spine surgeries. J Neurosurg Spine. 2021;35[4]:460-470). In this article, the authors describe the retrospective study they performed to investigate whether MIS can improve patient outcomes in the obese population. The authors found that obese patients have poorer outcomes than nonobese patients after posterior lumbar MIS, although the use of an MIS technique for posterior decompression with or without instrumented fusion can be beneficial in obese patients, as evidenced by decreased estimated blood loss (EBL), operative time, length of stay (LOS), and surgical site infections (SSIs). Carroll et al.1 report an outstanding retrospective study, focusing on MIS in obese patients. However, we still have concerns regarding their investigation.

As SSI is a binary outcome, the use of a multivariable logistic regression model is suggested. LOS, EBL, and operative time are continuous outcomes, so the use of a multivariable linear regression model is indicated. According to Table 4 in the article, it seems that multivariable linear regression was performed to analyze SSI and multivariable logistic regression was performed to analyze LOS, EBL, and operative time, which are inappropriate.

Additionally, Table 2 indicates the use of multivariable zero-inflated Poisson regression to analyze LOS. However, the indication for the use of this model2 (which might be used to model count data for which the proportion of zero counts is greater than expected on the basis of the mean of the nonzero counts) was not included in the description of the statistical analysis in this article, which might be somewhat confusing for the readers, especially for those who are not so familiar with the statistical methods.

We greatly appreciate this study, which indicates the significance of BMI in the patient outcomes for MIS. In addition, we suggest that the details mentioned above be considered.

Disclosures

The authors report no conflict of interest.

References

  • 1↑

    Carroll AH, Dowlati E, Molina E, et al. Does minimally invasive spine surgery improve outcomes in the obese population? A retrospective review of 1442 degenerative lumbar spine surgeries. J Neurosurg Spine. 2021;35(4):460–470.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2↑

    Li CS, Lu JC, Park J, Kim K, Brinkley PA, Peterson JP. Multivariate zero-inflated Poisson models and their applications. Technometrics. 1999;41(1):29–38.

    • Crossref
    • Search Google Scholar
    • Export Citation

Response

Ehsan Dowlati1 and Jean-Marc Voyadzis1
View More View Less
  • 1 MedStar Georgetown University Hospital, Washington, DC

Response

We thank Dr. Zhao and colleagues for their interest and letter regarding our study. We aimed to evaluate MIS in posterior lumbar procedures in obese patient populations and compared this cohort to nonobese patients undergoing MIS and to obese patients undergoing open approach posterior lumbar surgeries. We specifically evaluated intraoperative parameters, perioperative outcomes, and complication rates. We would like to take this opportunity to address the two concerns raised by the authors in their letter.

The authors note that Table 4 in our article indicates that a multivariate logistic regression was performed to analyze LOS, EBL, and operative time, and a multivariate linear regression was used to assess SSI. This information is provided in the Table 4 legend, which shows that an asterisk (*) indicates linear regression and a dagger (†) indicates logistic regression. However, these symbols as shown in the article should be flipped because of an error that occurred during production and proofing of the manuscript. The accompanying erratum includes the corrected Table 4.

The authors also note that we used a zero-inflated Poisson regression model to assess the effect on LOS. The use of this model in our statistical analysis was not apparent to readers in our Methods section. This statistical model was used due to the fact that a significant proportion of data from the MIS cohort consisted of values of zero. For these patients, the LOS was “0 days” as they were typically discharged and sent home the day of surgery. As noted by Dr. Zhao and colleagues in their letter, the multivariate zero-inflated Poisson regression model would provide a good fit for statistical analysis.

We appreciate the authors’ thoughts and input and hope that our explanation has brought clarity to readers. We also note that the results and conclusions of our study remain unchanged.

Acknowledgments

We acknowledge Sameer Desale and the Biostatistics and Biomedical Informatics Department at the Medstar Health Research Institute for their work with statistical calculations and guidance.

Volume 36: Issue 6 (Jun 2022)

in Journal of Neurosurgery: Spine
Cover Journal of Neurosurgery: Spine

Contributor Notes

Correspondence Weixin Li: neuropub@163.com.

INCLUDE WHEN CITING Published online December 24, 2021; DOI: 10.3171/2021.10.SPINE211281.

Disclosures The authors report no conflict of interest.

Page Count:
2
  • 1

    Carroll AH, Dowlati E, Molina E, et al. Does minimally invasive spine surgery improve outcomes in the obese population? A retrospective review of 1442 degenerative lumbar spine surgeries. J Neurosurg Spine. 2021;35(4):460–470.

    • Crossref
    • PubMed
    • Search Google Scholar
    • Export Citation
  • 2

    Li CS, Lu JC, Park J, Kim K, Brinkley PA, Peterson JP. Multivariate zero-inflated Poisson models and their applications. Technometrics. 1999;41(1):29–38.

    • Crossref
    • Search Google Scholar
    • Export Citation

Metrics

All Time Past Year Past 30 Days
Abstract Views 0 0 0
Full Text Views 401 401 58
PDF Downloads 343 343 70
EPUB Downloads 0 0 0