Comparison of 30-day outcome following anterior cervical discectomy and fusion with or without instrumentation for cervical spondylosis: A review of 2352 elective cases.

BACKGROUND: Anterior cervical discectomy and fusion (ACDF) is a commonly performed procedure to address cervical myeloradiculopathy. However, 30-day outcomes after additional plating/instrumentation are not very clear.
METHODS: The authors reviewed The National Surgical Quality Improvement Program database to identify all elective ACDF cases with or without instrumentation for patients having cervical spondylosis with or without myelopathy from 2011 to 2013 using current procedural terminology and International Classification of Disease-9 codes. We identified 2352 cases and subdivided these into two cohorts based on instrumentation procedures (588 cases without instrumentation and 1764 cases with instrumentation). Baseline differences in two cohorts were adjusted by propensity score matching analysis, yielding well-matched 583 pairs.
RESULTS: Following propensity matching, the authors observed no significant difference in 30-day complication rates (prematch, 2.4% vs. 2.4%; and postmatch, 2.4% vs. 1.7%), readmission (prematch, 4.1% vs. 3.2%; and postmatch, 3.9% vs. 3.3%), and reoperation (prematch 0.9% vs. 1.8%; and postmatch 0.9% vs. 1.5%).
CONCLUSION: Our results demonstrate similar 30-day outcomes in both cohorts and suggest that instrumentation can be safely implemented in the setting of ACDF. Copyright:

INTRODUCTION

Cervical spondylosis is a common degenerative condition of the spine. It often correlates clinically with radiculopathy and less commonly, myelopathy. Anterior cervical discectomy and fusion (ACDF) is commonly utilized to address cervical myeloradiculopathy, although the role of additional plating/instrumentation is less clear.[2] For single-level procedures, several studies demonstrated similar clinical outcomes and fusion rates, but better sagittal alignment with instrumentation.[6,9,17] However, there is limited evidence regarding short-term outcomes following instrumented versus noninstrumented ACDF procedures. Therefore, this study was designed to assess short-term outcomes following instrumented versus noninstrumented ACDF utilizing The American College of Surgeons National Surgical Quality Improvement Program (NSQIP) database and to investigate whether instrumentation affects short-term postoperative outcomes.

MATERIALS AND METHODS

Data acquisition

The authors reviewed the NSQIP database to identify all elective ACDF cases with or without instrumentation for patients with cervical spondylosis with or without myelopathy from 2011 to 2013. We utilized Current Procedural Terminology (CPT) and International Classification of Disease (ICD-9) codes to capture 2352 elective ACDF procedures [Table 1]. These cases were divided into two cohorts based on instrumentation (588 cases without instrumentation and 1764 cases with instrumentation).

Table 1:

CPT codes used for case selection/exclusion.

We tracked multiple demographic and operative variables for adequate propensity score matching analysis to compare 30-day postoperative outcomes following instrumented versus noninstrumented ACDF procedures [Tables 2-3].

Table 2:

Unmatched patient demographics, comorbidities, and operative characteristics.

Table 3:

Propensity-matched patient demographics, comorbidities, and operative characteristics.

Statistical analysis

Continuous variables were compared using student t-test or Manny–Whitney U-test based on normality test. For categorical variables, we used Pearson’s Chi-square test or Fischer’s exact test. We also utilized propensity score matching analysis to adjust for baseline difference between two cohorts. This process yielded well-matched 583 pairs and they were analyzed using the McNemar exact test for categorical variables, and Wilcoxon Signed-rank test or paired t-test for continuous variables [Table 4].

Table 4:

Thirty-day outcomes following anterior cervical discectomy and fusion.

For all analyses performed in this study, we used IBM SPSS Statistics version 22.0 (IBM Corp., Armonk, NY).

RESULTS

Unadjusted dataset

Of the 2352 patients included in this study, 588 were in the noninstrumentation cohort and 1764 in the instrumentation cohort. Patients who had instrumentation were more likely to be obese, Caucasian, were neurosurgical cases, had lower ASA classifications, involved two or more levels, and had a higher total RVU value [Table 2]. No significant differences were appreciated between the two cohorts in postoperative outcome including 30-day complication rates, unplanned reoperation, and readmission [Table 3].

Propensity score-matched dataset

Propensity score matching yielded 583 well-matched cases. After matching, there was no significant baseline difference other than total operative duration and total RVU. The mean difference propensity scores between the two cohorts before and after matching were 0.05 and <0.001, respectively [Table 4]. There was also no significant difference in 30-day complication rates, unplanned reoperation, and readmission rate [Table 4].

DISCUSSION

The authors relied on a prospectively maintained, national surgical database to assess how instrumentation may influence short-term postoperative outcome following single- and multi-level, elective ACDF for cervical spondylosis. This study sought to eliminate the baseline differences by utilizing a propensity score matching algorithm, which adjusted patient demographic profile, comorbidities, and major operative variables such as corpectomy or multi- level procedures. We further minimized procedural bias by selecting elective cases and tracking both individual CPT and ICD-9 codes.

Here, we demonstrated that patients in both noninstrumentation and instrumentation cohorts had comparable adverse outcome rates 30-day complication rates (prematch, 2.4% vs. 2.4%; and postmatch, 2.4% vs. 1.7%), readmission (prematch, 4.1% vs. 3.2%; and postmatch, 3.9% vs. 3.3%), and reoperation (prematch 0.9% vs. 1.8%; and postmatch 0.9% vs. 1.5%).

Complication and reoperation rates in this study were lower than previously reported rates, which were as high as 10%–13%.[1,3,4,8,10] The discrepancy is most likely attributed to a shorter follow-up period. Notably, our results showed that additional instrumentation is not significantly associated with adverse events including infection, soft tissue injury, or neurological deficit.[7,9,14,16,17] Previous studies demonstrated similar outcomes that instrumentation-related complication rates were well below 5% including unsecure screws, plate bending, and dysphagia.[5,7,11-15]

Our matched analysis also demonstrated that average operative duration was significantly longer than the noninstrumentation cohort (131.6 ± 69.6 vs. 122.1 ± 61.4; P = 0.014). However, 9-min difference in operative duration likely does not have any meaningful clinical influence despite the statistical difference arising from narrow standard deviation. This difference likely occurred during the matching process where operative duration was purposely not accounted for.

Here, we present a short-term multicenter analysis of outcomes following elective instrumented versus noninstrumented ACDF. We found similar 30-day outcomes in both cohorts which suggest that instrumentation can be safely implemented for ACDF resulting in comparable outcomes without incurring significant morbidity or adverse events.

CONCLUSION

Our analyses demonstrate similar 30-day outcomes in both cohorts, and suggest an additional instrumentation step can be safely implemented in the setting of cervical spondylosis with little concern for postoperative complication.