Stephens, Byron F.

OBJECTIVE Mechanical complications are a vexing occurrence after adult spinal deformity (ASD) surgery. While achieving ideal spinal alignment in ASD surgery is critical, alignment alone may not fully explain all mechanical complications. The authors sought to determine which combination of inputs produced the most sensitive and specific machine learning model to predict mechanical complications using postoperative alignment, bone quality, and soft tissue data. METHODS A retrospective cohort study was performed in patients undergoing ASD surgery from 2009 to 2021. Inclusion criteria were a fusion ≥ 5 levels, sagittal/coronal deformity, and at least 2 years of follow-up. The primary exposure variables were 1) alignment, evaluated in both the sagittal and coronal planes using the L1–pelvic angle ± 3°, L4–S1 lordosis, sagittal vertical axis, pelvic tilt, and coronal vertical axis; 2) bone quality, evaluated by the T-score from a dual-energy x-ray absorptiometry scan; and 3) soft tissue, evaluated by the paraspinal muscle–to–vertebral body ratio and fatty infiltration. The primary outcome was mechanical complications. Alongside demographic data in each model, 7 machine learning models with all combinations of domains (alignment, bone quality, and soft tissue) were trained. The positive predictive value (PPV) was calculated for each model. RESULTS Of 231 patients (24% male) undergoing ASD surgery with a mean age of 64 ± 17 years, 147 (64%) developed at least one mechanical complication. The model with alignment alone performed poorly, with a PPV of 0.85. However, the model with alignment, bone quality, and soft tissue achieved a high PPV of 0.90, sensitivity of 0.67, and specificity of 0.84. Moreover, the model with alignment alone failed to predict 15 complications of 100, whereas the model with all three domains only failed to predict 10 of 100. CONCLUSIONS These results support the notion that not every mechanical failure is explained by alignment alone. The authors found that a combination of alignment, bone quality, and soft tissue provided the most accurate prediction of mechanical complications after ASD surgery. While achieving optimal alignment is essential, additional data including bone and soft tissue are necessary to minimize mechanical complications.

BACKGROUND AND OBJECTIVES: Although risk factors for unplanned readmission after cervical spine surgery have been widely reported, less is known about how readmission itself affects patient-reported outcome measures (PROMs). Using the Quality Outcomes Database registry of patients undergoing elective cervical spine surgery, we sought to (1) determine the impact of unplanned readmission on PROMs and (2) compare the effect of specific readmission reasons on PROMs. METHODS: An observational study was performed using a multi-institution, retrospective registry for patients undergoing cervical spine surgery. The occurrence of 90-day unplanned readmission classified into medical, surgical, pain only, and no readmissions was the exposure variable. Outcome variables included 12-month PROMs of Neck Disability Index (NDI), Numeric Rating Scale (NRS)-neck/arm pain, EuroQol-5D (EQ-5D), and patient dissatisfaction. Multivariable models predicting each PROM were built using readmission reasons controlling for demographics, clinical characteristics, and preoperative PROMs. RESULTS: Data from 13 355 patients undergoing elective cervical spine surgery (82% anterior approach and 18% posterior approach) were analyzed. Unplanned readmission within 90 days of surgery occurred in 3.8% patients, including medical (1.6%), surgical (1.8%), and pain (0.3%). Besides medical reasons, wound infection/dehiscence was the most common reason for unplanned readmission for the total cohort (0.5%), dysphagia in the anterior approach (0.6%), and wound infection/dehiscence in the posterior approach (1.5%). Based on multivariable regression, surgical readmission was significantly associated with worse 12-month NDI, NRS-neck pain, NRS-arm pain, EQ-5D, and higher odds of dissatisfaction. Pain readmissions were associated with worse 12-month NDI and NRS-neck pain scores, and worse dissatisfaction. For specific readmission reasons, pain, surgical site infection/wound dehiscence, hematoma/seroma, revision surgery, deep vein thrombosis, and pulmonary embolism were significantly associated with worsened 12-month PROMs. CONCLUSION: In patients undergoing elective cervical spine surgery, 90-day unplanned surgical and pain readmissions were associated with worse 12-month PROMs compared with patients with medical readmissions and no readmissions.

Study Design. Retrospective review. Objective. The aim was to compare outcomes at 3 and 12 months for patients with lumbar spondylolisthesis treated with direct decompression (DD) versus indirect decompression (ID) techniques. Summary of Background Data. Debate persists regarding the optimal surgical strategy to treat lumbar spondylolisthesis. Novel techniques relying on ID have shown superior radiographic outcomes compared to DD, however, doubt remains regarding their effectiveness in achieving adequate decompression. Currently, there is a paucity of data comparing the clinical efficacy of DD to ID. Methods. The Quality Outcomes Database (QOD), a national, multicenter prospective spine registry, was queried for patients who underwent DD and ID between April 2013 and January 2019. Propensity scores for each treatment were estimated using logistic regression dependent on baseline covariates potentially associated with outcomes. The propensity scores were used to exclude nonsimilar patients. Multivariable regression analysis was performed with the treatment and covariate as independent variables and outcomes as dependent variables. Results. A total of 4163 patients were included in the DD group and 86 in the ID group. The ID group had significantly lower odds of having a longer hospital stay and for achieving 30% improvement in back and leg pain at 3 months. These trends were not statistically significant at 12 months. There were no differences in ED5D scores or Oswestry disability index 30% improvement scores at 3 or 12 months. ID patient had a significantly higher rate of undergoing a repeat operation at 3 months (4.9% vs. 1.5%, P =0.015). Conclusion. Our study suggests that both DD and ID for the treatment of lumbar spondylolisthesis result in similar clinical outcomes, with the exception that those treated with ID experienced a lower reduction in back and leg pain at 3 months and a higher 3-month reoperation rate. This data can provide surgeons with additional information when counseling patients on the pros and cons of ID versus DD surgery.

Study Design. Retrospective analysis on prospectively collected data. Objective. To determine the effectiveness of elective spine surgery in patients with Parkinson disease (PD). Background Context. PD has become increasingly prevalent in an aging population. While surgical treatment for degenerative spine pathology is often required in this population, previous literature has provided mixed results regarding its effectiveness. Methods. Data from the Quality Outcomes Database (QOD) was queried between April 2013 and January 2019. Three surgical groups were identified: (1) elective lumbar surgery, (2) elective cervical surgery for myelopathy, (3) elective cervical surgery for radiculopathy. Patients without PD were propensity matched against patients with PD in a 5:1 ratio without replacement based on American Society of Anesthesiology grade, arthrodesis, surgical approach, number of operated levels, age, and baseline Oswestry Disability Index, Numerical Rating Scale (NRS) extremity pain, NRS back pain, and EuroQol 5-Dimensions (EQ-5D). The mean difference was calculated for continuous outcomes (Oswestry Disability Index, NRS leg pain, NRS back pain, and EQ-5D at 3 and 12 mo after surgery) and the risk difference was calculated for binary outcomes (patient satisfaction, complications, readmission, reoperation, and mortality). Results. For the lumbar analysis, PD patients had a higher rate of reoperation at 12 months (risk difference=0.057, P=0.015) and lower mean EQ-5D score at 12 months (mean difference=−0.053, P=0.005) when compared with patients without PD. For the cervical myelopathy cohort, PD patients had lower NRS neck pain scores at 3 months (mean difference=−0.829, P=0.005) and lower patient satisfaction at 3 months (risk difference=−0.262, P=0.041) compared with patients without PD. For the cervical radiculopathy cohort, PD patients demonstrated a lower readmission rate at 3 months (risk difference=−0.045, P=0.014) compared with patients without PD. Conclusion. For the matched analysis, in general, patients with and without PD had similar patient-reported outcomes and complication, readmission, and reoperation rates. These results demonstrate that a diagnosis of PD alone should not represent a major contraindication to elective spine surgery.

Study Design: This was an observational study. Objectives: This study aims to determine the correlation between patient-reported outcomes (PROs) pulled from a national spine registry and physician ratings from physician review websites (PRWs). Summary of Background Data: PRWs are frequently utilized by patients to make health care decisions; however, many PRWs appear to incorporate subjective experiences unrelated to a surgeon’s clinical performance into ratings. As such, their utility as a health care decision-making tool remains unclear. Materials and Methods: This study evaluated 8834 patients from the Quality Outcomes Database (QOD) who underwent 1-level elective lumbar spine surgery. The lumbar module of QOD was queried to rank 124 surgeons using PROs (Oswestry Disability Index, EuroQOL, Numerical Rating Scale—back/leg pain, and patient satisfaction). The QOD PRO-ranking system was compared against PRWs including Healthgrades, Vitals, WebMD, and Google. The Spearman correlation coefficients, Kruskal-Wallis tests, and multiple linear regression models were used for statistical comparison. The primary outcome was the correlation between PRW scores and PROs. Results: Surgeon PRO-derived ranking showed high intercorrelational congruence with coefficients between the 3 PROs (Oswestry Disability Index, EuroQOL, Numerical Rating Scale back/leg) ranging between 0.70 and 0.88. Low correlations were observed between PRO-derived rankings and PRWs, ranging from 0.23 to 0.37. Healthgrades performed most similarly to PRO-derived rankings, correlating best with patient satisfaction, though the correlation was low (ρ=0.37). Conclusions: While PRWs are often used to evaluate surgeon competency, these results demonstrate they poorly correlate with a surgeon’s clinical ability measured by PROs. PRWs should be used with caution when making health care decisions by patients, payers, and administrators. Level of Evidence: Level III.