Artificial intelligence predicts disk re-herniation following lumbar microdiscectomy: development of the "RAD" risk profile.

PURPOSE: Surgical treatment of herniated lumbar intervertebral disks is a common procedure worldwide. However, recurrent herniated nucleus pulposus (re-HNP) may develop, complicating outcomes and patient management. The purpose of this study was to utilize machine-learning (ML) analytics to predict lumbar re-HNP, whereby a personalized risk prediction can be developed as a clinical tool. METHODS: A retrospective, single center study was conducted of 2630 consecutive patients that underwent lumbar microdiscectomy (mean follow-up: 22-months). Various preoperative patient pain/disability/functional profiles, imaging parameters, and anthropomorphic/demographic metrics were noted. An Extreme Gradient Boost (XGBoost) classifier was implemented to develop a predictive model identifying patients at risk for re-HNP. The model was exported to a web application software for clinical utility. RESULTS: There were 1608 males and 1022 females, 114 of whom experienced re-HNP. Primary herniations were central (65.8%), paracentral (17.6%), and far lateral (17.1%). The XGBoost algorithm identified multiple re-HNP predictors and was incorporated into an open-access web application software, identifying patients at low or high risk for re-HNP. Preoperative VAS leg, disability, alignment parameters, elevated body mass index, symptom duration, and age were the strongest predictors. CONCLUSIONS: Our predictive modeling via an ML approach of our large-scale cohort is the first study, to our knowledge, that has identified significant risk factors for the development of re-HNP after initial lumbar decompression. We developed the re-herniation after decompression (RAD) profile index that has been translated into an online screening tool to identify low-high risk patients for re-HNP. Additional validation is needed for potential global implementation.

Comparison of clinical outcomes in patients undergoing one- and two-level minimally invasive lumbar microdiscectomy.

OBJECTIVE: Herniated nucleus pulposus (HNP) is one of the most common lumbar spine conditions treated surgically, often through a minimally invasive surgery (MIS) microdiscectomy approach. This technique attempts to reduce damage to the paraspinal muscular-ligamentous envelope. However, there are currently limited data regarding comparative outcomes using patient-reported outcome measures (PROMs) for one- and two-level MIS discectomies. The aim of this study was to quantify comparative clinical outcomes in patients undergoing one-level and two-level MIS lumbar microdiscectomy for HNP using PROMs. METHODS: The authors performed a retrospective review of patients undergoing MIS lumbar microdiscectomy between 2004 and 2019 for the primary diagnosis of HNP at a single academic institution. All patients had a minimum 1-year follow-up. Patient demographics and comorbidities were collected to establish baselines between cohorts. PROMs and minimal clinically important differences (MCIDs) were used to examine the patient's perception of operative success. Bivariate and multivariate linear/logistic regression analyses were used to compare one- and two-level discectomies. The bivariate analysis included the t-test and chi-square test, which were used to assess continuous and categorical variables, respectively. Statistical significance was established at p < 0.05. RESULTS: A total of 293 patients underwent one-level (n = 250) or two-level (n = 43) MIS discectomies. The mean follow-ups for the one- and two-level cohorts were 50.4 (SD 35.5) months and 61.6 (SD 39.8) months, respectively. Fewer female patients underwent two-level discectomies, and BMI and operative duration were higher in the two-level group (p < 0.001). Recurrent herniation requiring reoperation was recorded at rates of 6.80% and 11.6% in the one- and two-level groups, respectively (p = 0.270). Pre- and postoperative PROMs were largely similar between the cohorts; however, patients undergoing one-level discectomy had greater improvement in leg pain, and a significantly greater proportion of these patients achieved MCID for the leg pain visual analog scale score (p < 0.001). CONCLUSIONS: At the 1-year clinical follow-up, patients who underwent two-level discectomy had significantly less improvement in leg pain scores with lower achievement of MCID for leg pain improvement than patients undergoing one-level procedures. At the 1-year follow-up, there were no other significant differences in PROMs between the two cohorts. Given these findings, patients should be counseled regarding the anticipated outcomes to better manage expectations. Further studies are warranted to examine the long-term clinical outcomes associated with single- and multilevel MIS discectomy.

Rates of Postoperative Complications and Approach-related Neurological Symptoms After L4-L5 Lateral Transpsoas Lumbar Interbody Fusion Compared With Upper Lumbar Levels.

STUDY DESIGN: This was a retrospective comparative study. OBJECTIVE: To compare the likelihood of approach-related complications for patients undergoing single-level lateral lumbar interbody fusion (LLIF) at L4-L5 to those undergoing the procedure at upper lumbar levels. SUMMARY OF BACKGROUND DATA: LLIF has been associated with a number of advantages when compared with traditional interbody fusion techniques. However, potential risks with the approach include vascular or visceral injury, thigh dysesthesias, and lumbar plexus injury. There are concerns of a higher risk of these complications at the L4-L5 level compared with upper lumbar levels. MATERIALS AND METHODS: A retrospective cohort review was completed for consecutive patients undergoing single-level LLIF between 2004 and 2019 by a single surgeon. Indication for surgery was symptomatic degenerative lumbar stenosis and/or spondylolisthesis. Patients were divided into 2 cohorts: LLIF at L4-L5 versus a single level between L1 and L4. Baseline characteristics, intraoperative complications, postoperative approach-related neurological symptoms, and patient-reported outcomes were compared and analyzed between the cohorts. RESULTS: A total of 122 were included in analysis, of which 58 underwent LLIF at L4-L5 and 64 underwent LLIF between L1 and L4. There were no visceral or vascular injuries or lumbar plexus injuries in either cohort. There was no significant difference in the rate of postoperative hip pain, anterior thigh dysesthesias, and/or hip flexor weakness between the cohorts (53.5% L4-L5 vs. 37.5% L1-L4; P =0.102). All patients reported complete resolution of these symptoms by 6-month postoperative follow-up. DISCUSSION: LLIF surgery at the L4-L5 level is associated with a similar infrequent likelihood of approach-related complications and postoperative hip pain, thigh dysesthesias, and hip flexor weakness when compared with upper lumbar level LLIF. Careful patient selection, meticulous use of real-time neuromonitoring, and an understanding of the anatomic location of the lumbar plexus to the working corridor are critical to success.

YouTube as a source of information on pediatric scoliosis: a reliability and educational quality analysis.

PURPOSE: To assess the reliability and educational quality of YouTube videos related to pediatric scoliosis. METHODS: In December 2020, searches of "pediatric scoliosis", "idiopathic scoliosis", "scoliosis in children", and "curved spine in children" were conducted using YouTube. The first 50 results of each search were analyzed according to upload source and content. The Journal of the American Medical Association (JAMA) Benchmark Criteria were used to assess reliability (score 0-4), and educational quality was evaluated using the Global Quality Score (GQS; score 0-5) and Pediatric Scoliosis-Specific Score (PSS; score 0-15). Differences in scores based on upload source and content were determined by Analysis of Variance (ANOVA) or Kruskal-Wallis tests. Multivariate linear regressions identified any independent predictors of reliability and educational quality. RESULTS: After eliminating duplicates, 153 videos were analyzed. Videos were viewed 28.5 million times in total, averaging 186,160.3 ± 1,012,485.0 views per video. Physicians (54.2%) and medical sources (19.0%) were the most common upload sources, and content was primarily categorized as disease-specific (50.0%) and patient experience (25.5%). Videos uploaded by patients achieved significantly lower JAMA scores (p = 0.004). Conversely, academic or physician-uploaded videos scored higher on PSS (p = 0.003) and demonstrated a trend towards improved GQS (p = 0.051). Multivariate analysis determined longer video duration predicted higher scores on all measures. However, there were no independent associations between upload source or content and assessment scores. CONCLUSION: YouTube contains a large repository of videos concerning pediatric scoliosis; however, the reliability and educational quality of these videos were low. LEVEL OF EVIDENCE: V.

Low back pain: What is the role of YouTube content in patient education?

The aim of this study was to characterize the educational quality and reliability of YouTube videos related to low back pain (LBP) as well as to identify factors associated with the overall video quality. A review of YouTube was performed using two separate search strings. Video-specific characteristics were analyzed for the first 50 videos of each string. Seventy-seven eligible videos were identified as a result. The mean Journal of the American Medical Association score was 2.25 ± 1.09 (range: 0-4) out of 4. The mean Global Quality Score (GQS) score was 2.29 ± 1.37 (range: 1-4) out of 5. The mean LBP score (LPS) score was 3.83 ± 2.23 (range: 0-11) out of 15. Video power index was a predictor of GQS score (ß = 55.78, p = 0.048), whereas the number of likes (ß = -2.49, p = 0.047) and view ratio (ß = -55.62, p = 0.049) were associated with lower quality scores. Days since initial upload (ß = 0.32, p = 0.042) as well as like ratio (ß = 0.37, p = 0.019) were independent predictors of higher LPS scores. The results of this study suggest that the overall reliability and educational quality of videos uploaded to YouTube concerning LBP are unsatisfactory. More popular videos demonstrated poorer educational quality than their less popular counterparts. As the prevalence of LBP rises, more accurate and thorough educational videos are necessary to ensure accurate information is available to patients.

Assessing the Quality and Credibility of Publicly Available Videos on Cervical Fusion: Is YouTube a Reliable Educational Tool?

BACKGROUND: YouTube has become a popular source for patient education, though there are concerns regarding the quality and reliability of videos related to orthopaedic and neurosurgical procedures. This study aims to evaluate the credibility and educational content of videos on YouTube related to cervical fusion. Secondarily, the study aims to identify factors predictive of higher or lower quality videos. METHODS: A YouTube query using the search terms "cervical fusion" was performed, and the first 50 videos were included for analysis. Reliability was assessed using the Journal of the American Medical Association (JAMA) criteria. Educational quality was assessed using the Global Quality Score (GQS) and the Cervical Fusion Content Score (CFCS). Videos were stratified by content and source, and differences in JAMA, GQS, and CFCS scores were assessed. Multivariable linear regression was used to identify predictors of higher or lower JAMA, GQS, and CFCS scores. Statistical significance was established at P < 0.05. RESULTS: Total number of views was 6 221 816 with a mean of 124 436.32 ± 412 883.32 views per video. Physicians, academic, and medical sources had significantly higher mean JAMA scores (P = 0.042). Exercise training and nonsurgical management videos had significantly higher mean CFCS scores (P = 0.018). Videos by physicians (ß = 0.616; P = 0.025) were independently associated with higher JAMA scores. Advertisements were significant predictors of worse CFCS (ß = -3.978; P = 0.030), and videos by commercial sources predicted significantly lower JAMA scores (ß = -1.326; P = 0.006). CONCLUSIONS: While videos related to cervical fusion amassed a large viewership, they were poor in both quality and reliability. Videos by physicians were associated with higher reliability scores relative to other sources, whereas commercial sources and advertisements had significantly lower reliability and educational content scores. Currently, YouTube seems to be an unreliable source of information on cervical fusion for patients. LEVEL OF EVIDENCE: 4. CLINICAL RELEVANCE: The results of this study aid surgeons in counseling patients interested in cervical fusion, and suggest that publicly available videos regarding cervical fusion may not be an adequate tool for patient education at this time.

Imaging in Spine Surgery: Current Concepts and Future Directions.

OBJECTIVE: To review and highlight the historical and recent advances of imaging in spine surgery and to discuss current applications and future directions. METHODS: A PubMed review of the current literature was performed on all relevant articles that examined historical and recent imaging techniques used in spine surgery. Studies were examined for their thoroughness in description of various modalities and applications in current and future management. RESULTS: We reviewed 97 articles that discussed past, present, and future applications for imaging in spine surgery. Although most historical approaches relied heavily upon basic radiography, more recent advances have begun to expand upon advanced modalities, including the integration of more sophisticated equipment and artificial intelligence. CONCLUSIONS: Since the days of conventional radiography, various modalities have emerged and become integral components of the spinal surgeon's diagnostic armamentarium. As such, it behooves the practitioner to remain informed on the current trends and potential developments in spinal imaging, as rapid adoption and interpretation of new techniques may make significant differences in patient management and outcomes. Future directions will likely become increasingly sophisticated as the implementation of machine learning, and artificial intelligence has become more commonplace in clinical practice.