Development of a machine learning algorithm predicting discharge placement after surgery for spondylolisthesis.

PURPOSE: We aimed to develop a machine learning algorithm that can accurately predict discharge placement in patients undergoing elective surgery for degenerative spondylolisthesis. METHODS: The National Surgical Quality Improvement Program (NSQIP) database was used to select patients that underwent surgical treatment for degenerative spondylolisthesis between 2009 and 2016. Our primary outcome measure was non-home discharge which was defined as any discharge not to home for which we grouped together all non-home discharge destinations including rehabilitation facility, skilled nursing facility, and unskilled nursing facility. We used Akaike information criterion to select the most appropriate model based on the outcomes of the stepwise backward logistic regression. Four machine learning algorithms were developed to predict discharge placement and were assessed by discrimination, calibration, and overall performance. RESULTS: Nine thousand three hundred and thirty-eight patients were included. Median age was 63 (interquartile range [IQR] 54-71), and 63% (n = 5,887) were female. The non-home discharge rate was 18.6%. Our models included age, sex, diabetes, elective surgery, BMI, procedure, number of levels, ASA class, preoperative white blood cell count, and preoperative creatinine. The Bayes point machine was considered the best model based on discrimination (AUC = 0.753), calibration (slope = 1.111; intercept = - 0.002), and overall model performance (Brier score = 0.132). CONCLUSION: This study has shown that it is possible to create a predictive machine learning algorithm with both good accuracy and calibration to predict discharge placement. Using our methodology, this type of model can be developed for many other conditions and (elective) treatments. These slides can be retrieved under Electronic Supplementary Material.

Reoperation of decompression alone or decompression plus fusion surgeries for degenerative lumbar diseases: a systematic review.

PURPOSE: The objective of this paper was to compare the reoperation rates, timing and causes between decompression alone and decompression plus fusion surgeries for degenerative lumbar diseases through a systematic review of the published data. METHODS: A search of the literature was conducted on PubMed/MEDLINE, EMBASE and the Cochrane Collaboration Library. Reports that included reoperations after decompression alone and/or decompression plus fusion surgeries were selected using designed eligibility criteria. Comparative analysis of reoperation rates, timing and causes between the two surgeries was conducted. RESULTS: Thirty-two retrospective and three prospective studies were selected from 6401 papers of the literature search. The analysis of data reported in these studies revealed that both surgeries resulted in similar reoperation rates after the primary surgery. However, majority of reoperations following the fusion surgeries were due to adjacent-segment diseases, and following the decompression alone surgeries were due to the same-segment diseases. Reoperation rates were not found to decrease in patients operated more recently than those operated in early times. CONCLUSIONS: Reoperation rates were similar following decompression alone or plus fusion surgeries for degenerative lumbar diseases. However, different underlying major causes exist between the two surgeries. There is no evidence showing that the reoperation rate has a trend to decline with newer surgical techniques used. The exact mechanisms of reoperation after both surgeries are still unclear. Further researches are necessary to investigate the mechanisms of reoperation for improvement of surgical techniques that aim to delay or prevent reoperation after lumbar surgery. These slides can be retrieved under Electronic Supplementary Material.

Decompressive surgery for lumbar spinal stenosis across the Atlantic: a comparison of preoperative MRI between matched cohorts from the US and Norway.

BACKGROUND: There are no uniform guidelines regarding when to operate for Lumbar Spinal Stenosis (LSS). As we apply findings from clinical research from one population to the next, elucidating similarities or differences provides important context for the validity of extrapolating clinical outcomes. The aim of this study was to compare the morphological severity of lumbar spinal stenosis on preoperative MRI in patients undergoing decompressive surgery in Boston, USA, and Trondheim, Norway. METHODS: In this observational retrospective study, we compared morphological severity on MRI before surgical treatment between two propensity score-matched patient populations with single or two-level symptomatic LSS. We assessed the radiographic severity of LSS utilizing the Schizas classification (grade A to D). RESULTS: Following propensity score matching, demographics are balanced. In the Trondheim cohort, two levels decompression were present in 36.2% of the patients vs. 41.9% in Boston, (p = 0.396). There was no significant difference in grades A to D concerning central stenosis (p = 0.075). When dichotomized in mild/moderate (A/B) and severe /extreme (C/D), there were no significant differences in the rate of levels operated for high-grade stenosis (C/D), 67.6% in the Boston group compare to 78.1% in the Trondheim group (p = 0.088). CONCLUSIONS: Trondheim, Norway, and Boston, US, have similar radiographic thresholds of LSS for offering surgery.

Ranges of Cervical Intervertebral Disc Deformation During an In Vivo Dynamic Flexion-Extension of the Neck.

While abnormal loading is widely believed to cause cervical spine disc diseases, in vivo cervical disc deformation during dynamic neck motion has not been well delineated. This study investigated the range of cervical disc deformation during an in vivo functional flexion-extension of the neck. Ten asymptomatic human subjects were tested using a combined dual fluoroscopic imaging system (DFIS) and magnetic resonance imaging (MRI)-based three-dimensional (3D) modeling technique. Overall disc deformation was determined using the changes of the space geometry between upper and lower endplates of each intervertebral segment (C3/4, C4/5, C5/6, and C6/7). Five points (anterior, center, posterior, left, and right) of each disc were analyzed to examine the disc deformation distributions. The data indicated that between the functional maximum flexion and extension of the neck, the anterior points of the discs experienced large changes of distraction/compression deformation and shear deformation. The higher level discs experienced higher ranges of disc deformation. No significant difference was found in deformation ranges at posterior points of all the discs. The data indicated that the range of disc deformation is disc level dependent and the anterior region experienced larger changes of deformation than the center and posterior regions, except for the C6/7 disc. The data obtained from this study could serve as baseline knowledge for the understanding of the cervical spine disc biomechanics and for investigation of the biomechanical etiology of disc diseases. These data could also provide insights for development of motion preservation surgeries for cervical spine.

The effect of short (2-weeks) versus long (6-weeks) post-operative restrictions following lumbar discectomy: a prospective randomized control trial.

PURPOSE: We sought to evaluate how short (2-week) versus long (6-week) post-operative restrictions following lumbar discectomy impacted outcomes and reherniation rates for a period up to 1 year following surgery. METHODS: This study included 108 patients undergoing index lumbar discectomy. Patients were randomized immediately following surgery. Outcomes included back and leg visual analog pain scales (VAS), Oswestry Disability Index (ODI), and reherniation rates at 2-weeks, 6-weeks, 3-months, and 1-year following surgery. Differences in reherniation rates were analyzed using Fisher's exact test. VAS and ODI scores were evaluated using Student's t test. RESULTS: Six patients (11%) in the 2-week restriction group had a reherniation event and four patients (7%) in the 6-week restriction group experienced a reherniation (p = 0.52). VAS back pain (p < 0.001), leg pain (p < 0.001), and ODI scores (p < 0.001) were significantly improved for both cohorts as compared to baseline at the 2-week time point and remained significantly improved through 1-year [VAS back (p < 0.001); VAS leg (p < 0.001); ODI (p < 0.001)]. No significant differences in ODI, VAS back, or VAS leg scores were detected at any of the time points between the 2- and 6-week restriction groups. CONCLUSIONS: The results of this randomized trial suggest equivalent clinical outcomes irrespective of the length of post-operative restriction. From a clinical perspective, if patients are deemed at low risk for a reherniation event they may be confidence that early return to activity at 2 weeks will not compromise outcomes and may not adversely impact the risk of reherniation. Level of Evidence II.

The influence of adjacent level disc disease on discectomy outcomes.

PURPOSE: The state of adjacent level discs and its impact on surgical outcomes following single-level lumbar discectomy have not been previously investigated. The purpose of the present study was to determine if a significant relationship exists between the degree of preoperative adjacent level disc degeneration and post-operative clinical outcomes following lumbar discectomy. METHODS: This study retrospectively used preoperative magnetic resonance imaging (MRI) and prospectively collected data from a randomized clinical trial at two tertiary-care academic hospitals. Patients who underwent a primary, single-level lumbar discectomy were included. Exclusion criteria included prior lumbar surgery. Outcome measures were the Modified Oswestry Disability Index (ODI) score and Visual Analog Scale (VAS) scores for back and leg pain. These were recorded at baseline and at 3 months, 1, and 2 years postoperatively. An independent reviewer graded adjacent level disc degeneration on all preoperative MRIs using the Pfirrmann grading scale. These data were then analyzed for correlation with each outcome measure. RESULTS: Forty-seven patients were included in the study. No statistically significant correlations were found when comparing preoperative 3-month or 1-year postoperative scores or change from baseline of any outcome measure between Pfirrmann grades. Only about half the patients had 2-year follow-up, but at that time point a statistically significant difference in back VAS scores was observed between Pfirrmann groups. No other significant differences were observed at that point. CONCLUSIONS: The degree of preoperative adjacent level degeneration does not significantly affect functional or pain relief outcomes following lumbar discectomy up to 1 year after surgery.

In-vivo T2-relaxation times of asymptomatic cervical intervertebral discs.

Limited research exists on T2-mapping techniques for cervical intervertebral discs and its potential clinical utility. The objective of this research was to investigate the in-vivo T2-relaxation times of cervical discs, including C2-C3 through C7-T1. Ten asymptomatic subjects were imaged using a 3.0 T MR scanner and a sagittal multi-slice multi-echo sequence. Using the mid-sagittal image, intervertebral discs were divided into five regions-of-interest (ROIs), centered along the mid-line of the disc. Average T2 relaxation time values were calculated for each ROI using a mono-exponential fit. Differences in T2 values between disc levels and across ROIs of the same disc were examined. For a given ROI, the results showed a trend of increasing relaxation times moving down the spinal column, particularly in the middle regions (ROIs 2, 3 and 4). The C6-C7 and C7-T1 discs had significantly greater T2 values compared to superior discs (discs between C2 and C6). The results also showed spatial homogeneity of T2 values in the C3-C4, C4-C5, and C5-C6 discs, while C2-C3, C6-C7, and C7-T1 showed significant differences between ROIs. The findings indicate there may be inherent differences in T2-relaxation time properties between different cervical discs. Clinical evaluations utilizing T2-mapping techniques in the cervical spine may need to be level-dependent.

Coronal Spinal Balance in Adult Spine Deformity Patients With Long Spinal Fusions: A Minimum 2- to 5-Year Follow-up Study.

STUDY DESIGN: This study was a retrospective one. OBJECTIVE: The objective of the study was to analyze the causes, prevalence of, and risk factors for coronal decompensation in long adult lumbar spinal instrumentation and fusion (from thoracic or upper lumbar spine) to L5 or S1. SUMMARY OF BACKGROUND DATA: Coronal and sagittal decompensation after long fusions for spinal deformities can affect outcomes negatively. There is no study reporting the natural history of coronal spinal balance after long spinal fusions. METHODS: A single-center retrospective review of data from 54 patients with spinal deformity was performed. Inclusion criteria were patients over 18 years with long fusions (>4 segments) to L5 or the pelvis who had full spine standing radiographs before surgery and up to 2-5 years postoperatively. Radiographic data included C7PL, magnitude of scoliotic curve, shoulder or pelvic asymmetry in the coronal plane, thoracic kyphosis, lumbar lordosis, and pelvic parameters (pelvic incidence, pelvic tilt, sacral slope). Coronal imbalance (CI) was considered if the C7PL was >4 cm lateral to the central sacral line, and sagittal imbalance (SI) was considered when the C7 plumbline was >4 cm anterior to the middle of the upper sacral plate. Paired t test, χ test, and repeated measures regression analysis using demographic data (age, sex, body mass index), operative (previous fusion, posterior only or anteroposterior fusion, iliac fixation or not, decompression or not, osteotomy or not) and postoperative (complications, use of bracing) data, and radiographic parameters (including SI) were performed. RESULTS: Patients showing CI equaled 11 (19.3%) preoperatively, remained 11 (19.3%) (4 of whom were new patients with CI) at 6 weeks postoperatively, and increased (P<0.001) to 18 (31.6%) (8 of them without initial CI) at 2-5 years follow-up. However, in terms of numeric distance of C7PL from the midsacrum, there was no statistically significant change (P>0.05) from preoperative to last follow-up. SI showed significant improvement (P<0.05) from preoperative to 6 weeks postoperative and no statistical significant change (P>0.05) from 6 weeks to 2-5 years postoperatively. Repeated measures regression analysis showed that the presence of osteoporosis and the combination of anterior approach surgery with a history of previous surgery were significant (P<0.05) factors predictive of changes in coronal balance. CONCLUSIONS: After surgical correction of spinal deformities, coronal spinal decompensation appears in an increased number of patients at last follow-up postoperatively but without significant differences in coronal plane C7PL during the postoperative period. Attention should be paid to patients with osteoporosis and those with a combination of previous same site posterior spine surgery and new anterior approach surgery for changes of coronal balance postoperatively.

A proposed set of metrics for standardized outcome reporting in the management of low back pain.

BACKGROUND AND PURPOSE: Outcome measurement has been shown to improve performance in several fields of healthcare. This understanding has driven a growing interest in value-based healthcare, where value is defined as outcomes achieved per money spent. While low back pain (LBP) constitutes an enormous burden of disease, no universal set of metrics has yet been accepted to measure and compare outcomes. Here, we aim to define such a set. PATIENTS AND METHODS: An international group of 22 specialists in several disciplines of spine care was assembled to review literature and select LBP outcome metrics through a 6-round modified Delphi process. The scope of the outcome set was degenerative lumbar conditions. RESULTS: Patient-reported metrics include numerical pain scales, lumbar-related function using the Oswestry disability index, health-related quality of life using the EQ-5D-3L questionnaire, and questions assessing work status and analgesic use. Specific common and serious complications are included. Recommended follow-up intervals include 6, 12, and 24 months after initiating treatment, with optional follow-up at 3 months and 5 years. Metrics for risk stratification are selected based on pre-existing tools. INTERPRETATION: The outcome measures recommended here are structured around specific etiologies of LBP, span a patient's entire cycle of care, and allow for risk adjustment. Thus, when implemented, this set can be expected to facilitate meaningful comparisons and ultimately provide a continuous feedback loop, enabling ongoing improvements in quality of care. Much work lies ahead in implementation, revision, and validation of this set, but it is an essential first step toward establishing a community of LBP providers focused on maximizing the value of the care we deliver.

Dynamic motion characteristics of the lower lumbar spine: implication to lumbar pathology and surgical treatment.

PURPOSE: Many studies have reported on the segmental motion range of the lumbar spine using various in vitro and in vivo experimental designs. However, the in vivo weightbearing dynamic motion characteristics of the L4-5 and L5-S1 motion segments are still not clearly described in literature. This study investigated in vivo motion of the lumbar spine during a weight-lifting activity. METHODS: Ten asymptomatic subjects (M/F: 5/5; age: 40-60 years) were recruited. The lumbar segment of each subject was MRI-scanned to construct 3D models of the L2-S1 vertebrae. The lumbar spine was then imaged using a dual fluoroscopic imaging system as the subject performed a weight-lifting activity from a lumbar flexion position (45°) to maximal extension position. The 3D vertebral models and the fluoroscopic images were used to reproduce the in vivo vertebral positions along the motion path. The relative translations and rotations of each motion segment were analyzed. RESULTS: All vertebral motion segments, L2-3, L3-4, L4-5 and L5-S1, rotated similarly during the lifting motion. L4-5 showed the largest anterior-posterior (AP) translation with 2.9 ± 1.5 mm and was significantly larger than L5-S1 (p < 0.05). L5-S1 showed the largest proximal-distal (PD) translation with 2.8 ± 0.9 mm and was significantly larger than all other motion segments (p < 0.05). CONCLUSIONS: The lower lumbar motion segments L4-5 and L5-S1 showed larger AP and PD translations, respectively, than the higher vertebral motion segments during the weight-lifting motion. The data provide insight into the physiological motion characteristics of the lumbar spine and potential mechanical mechanisms of lumbar disease development.

Assessing the variation and drivers of cost in 1-level lumbar fusion: a time-driven activity-based costing analysis.

BACKGROUND CONTEXT: As value-based health care arrangements gain traction in spine care, understanding the true cost of care becomes critical. Historically, inaccurate cost proxies have been used, including negotiated reimbursement rates or list prices. However, time-driven activity-based costing (TDABC) allows for a more accurate cost assessment, including a better understanding of the primary drivers of cost in 1-level lumbar fusion. PURPOSE: To determine the variation of total hospital cost, differences in characteristics between high-cost and non-high-cost patients, and to identify the primary drivers of total hospital cost in a sample of patients undergoing 1-level lumbar fusion. STUDY DESIGN/SETTING: Retrospective, multicenter (one academic medical center, one community-based hospital), observational study. PATIENT SAMPLE: A total of 383 patients undergoing elective 1-level lumbar fusion for degenerative spine conditions between November 2, 2021 and December 2, 2022. OUTCOME MEASURES: Total hospital cost of care (normalized); preoperative, intraoperative, and postoperative cost of care (normalized); ratio of most to least expensive 1-level lumbar fusion. METHODS: Patients undergoing a 1-level lumbar fusion between November 2, 2021 and December 2, 2022 were identified at two hospitals (one quaternary referral academic medical center and one community-based hospital) within our health system. TDABC was used to calculate total hospital cost, which was also broken up into: pre-, intra-, and postoperative timeframes. Operating surgeon and patient characteristics were also collected and compared between high- and non-high-cost patients. The correlation of surgical time and cost was determined. Multivariable linear regression was used to determine factors associated with total hospital cost. RESULTS: The most expensive 1-level lumbar fusion was 6.8x more expensive than the least expensive 1-level lumbar fusion, with the intraoperative period accounting for 88% of total cost. On average. the implant cost accounted for 30% of the total, but across the patient sample, the implant cost accounted for a range of 6% to 44% of the total cost. High-cost patients were younger (55 years [SD: 13 years] vs.63 years [SD: 13 years], p=.0002), more likely to have commercial health insurance (24 out of 38 (63%) vs. 181 out of 345 (52%), p=.003). There was a poor correlation between time of surgery (i.e., incision to close) and total overall cost (ρ: .26, p<.0001). Increase age (RC: -0.003 [95% CI: -0.006 to -0.000007], p=.049) was associated with decreased cost. Surgery by certain surgeons was associated with decreased total cost when accounting for other factors (p<.05). CONCLUSIONS: A large variation exists in the total hospital cost for patients undergoing 1-level lumbar fusion, which is primarily driven by surgeon-level decisions and preferences (e.g., implant and technology use). Also, being a "fast" surgeon intraoperatively does not mean your total cost is meaningfully lower. As efforts continue to optimize patient value through ensuring appropriate clinical outcomes while also reducing cost, spine surgeons must use this knowledge to lead, or at least be active participants in, any discussions that could impact patient care.

Magnetic Resonance Imaging Is Inadequate to Assess Cervical Sagittal Alignment Parameters.

STUDY DESIGN: Retrospective radiographic study. OBJECTIVE: To evaluate cervical sagittal alignment measurement reliability and correlation between upright radiographs and magnetic resonance imaging (MRI). SUMMARY OF BACKGROUND DATA: Cervical sagittal alignment (CSA) helps determine the surgical technique employed to treat cervical spondylotic myelopathy. Traditionally, upright lateral radiographs are used to measure CSA, but obtaining adequate imaging can be challenging. Utilizing MRI to evaluate sagittal parameters has been explored; however, the impact of positional change on these parameters has not been determined. METHODS: One hundred seventeen adult patients were identified who underwent laminoplasty or laminectomy and fusion for cervical spondylotic myelopathy from 2017 to 2019. Two clinicians independently measured the C2-C7 sagittal angle, C2-C7 sagittal vertical axis (SVA), and the T1 tilt. Interobserver and intraobserver reliability were assessed by intraclass correlation coefficient. RESULTS: Intraobserver and interobserver reliabilities were highly correlated, with correlations greater than 0.85 across all permutations; intraclass correlation coefficients were highest with MRI measurements. The C2-C7 sagittal angle was highly correlated between x-ray and MRI at 0.76 with no significant difference ( P =0.46). There was a weaker correlation with regard to C2-C7 SVA (0.48) and T1 tilt (0.62) with significant differences observed in the mean values between the 2 modalities ( P <0.01). CONCLUSIONS: The C2-C7 sagittal angle is highly correlated and not significantly different between upright x-ray and supine MRIs. However, cervical SVA and T1 tilt change with patient position. Since MRI does not accurately reflect the CSA in the upright position, upright lateral radiographs should be obtained to assess global sagittal alignment when planning a posterior-based cervical procedure.

Surgical Intervention is Associated With Improvements in the ASIA Impairment Scale in Gunshot-induced Spinal Injuries of the Thoracic and Lumbar Spine.

STUDY DESIGN: Retrospective cohort study of patients from the National Spinal Cord Injury Statistical Center (NSCISC). OBJECTIVE: The aim was to compare the outcomes of patients with gunshot-induced spinal injuries (GSIs) treated operatively and nonoperatively. SUMMARY OF BACKGROUND DATA: The treatment of neurological deficits associated with gunshot wounds to the spine has been controversial. Treatment has varied widely, ranging from nonoperative to aggressive surgery. METHODS: Patient demographics, clinical information, and outcomes were extracted. Surgical intervention was defined as a "laminectomy, neural canal restoration, open reduction, spinal fusion, or internal fixation of the spine." The primary outcome was the American Spinal Injury Association (ASIA) Impairment Scale. Statistical comparisons of baseline demographics and neurological outcomes between operative and nonoperative cohorts were performed. RESULTS: In total, 961 patients with GSI and at least 1-year follow-up were identified from 1975 to 2015. The majority of patients were Black/African American (55.6%), male (89.7%), and 15-29 years old (73.8%). Of those treated surgically (19.7% of all patients), 34.2% had improvement in their ASIA Impairment Scale score at 1 year, compared with 20.6% treated nonoperatively. Overall, surgery was associated with a 2.0 [95% confidence interval (CI): 1.4-2.8] times greater likelihood of ASIA Impairment Scale improvement at 1 year. Specifically, benefit was seen in thoracic (odds ratio: 2.5; 95% CI: 1.4-4.6) and lumbar injuries (odds ratio: 1.7; 95% CI: 1.1-3.1), but not cervical injuries. CONCLUSIONS: While surgical indications are always determined on an individualized basis, in our review of GSIs, surgical intervention was associated with a greater likelihood of neurological recovery. Specifically, patients with thoracic and lumbar GSIs had a 2.5 and 1.7-times greater likelihood of improvement in their ASIA Impairment Scale score 1 year after injury, respectively, if they underwent surgical intervention.

Practice Variation Within a Single Institution in Management of Degenerative Spondylolisthesis.

STUDY DESIGN: This was a retrospective cohort study. OBJECTIVE: The objective of this study was to assess variation in care for degenerative spondylolisthesis (DS) among surgeons at the same institution, to establish diagnostic and therapeutic variables contributing to this variation, and to determine whether variation in care changed over time. SUMMARY OF BACKGROUND DATA: Like other degenerative spinal disorders, DS is prone to practice variation due to the wide array of treatment options. Focusing on a single institution can identify more individualized drivers of practice variation by omitting geographic variability of demographics and socioeconomic factors. MATERIALS AND METHODS: We collected number of office visits, imaging procedures, injections, electromyography (EMG), and surgical procedures within 1 year after diagnosis. Multivariable logistic regression was used to determine predictors of surgery. The coefficient of variation (CV) was calculated to compare the variation in practice over time. RESULTS: Patients had a mean 2.5 (±0.6) visits, 1.8 (±0.7) imaging procedures, and 0.16 (±0.09) injections in the first year after diagnosis. Thirty-six percent (1937/5091) of patients had physical therapy in the 3 months after diagnosis. CV was highest for EMG (95%) and lowest for office visits (22%). An additional spinal diagnosis [odds ratio (OR)=3.99, P <0.001], visiting a neurosurgery clinic (OR=1.81, P =0.016), and diagnosis post-2007 (OR=1.21, P =0.010) were independently associated with increased surgery rates. The CVs for all variables decreased after 2007, with the largest decrease seen for EMG (132% vs. 56%). CONCLUSIONS: While there is variation in the management of patients diagnosed with DS between surgeons of a single institution, this variation seems to have gone down in recent years. All practice variables showed diminished variation. The largest variation and subsequent decrease of variation was seen in the use of EMG. Despite the smaller amount of variation, the rate of surgery has gone up since 2007.

Laminoplasty-an underutilized procedure for cervical spondylotic myelopathy.

BACKGROUND CONTENT: Cervical laminoplasty (LP) and laminectomy and fusion (LF) are commonly used surgical techniques for cervical spondylotic myelopathy (CSM). Several recent studies have demonstrated superior perioperative metrics and decreased overall costs with LP, yet LF is performed far more often in the United States. PURPOSE: To determine the percentage of patients with CSM who are radiographically candidates for LP. STUDY DESIGN: Retrospective comparative cohort study. PATIENT SAMPLE: Patients >18 years old who underwent LF or LP for CSM at 2 large academic institutions from 2017 to 2019. OUTCOME MEASURES: Candidacy for LP based on radiographic criteria. METHODS: Radiographs were assessed by 2 spine surgeons not involved in the care of the patients to determine the C2-C7 Cobb angle and the presence and extent of cervical instability. Patients with kyphosis >13°, > 3.5 mm of listhesis on static imaging, or > 2.5 mm of motion on flexion-extension or standing-supine films were not considered candidates for LP. Intraclass coefficient (ICC) was calculated to assess the interobserver reliability of angular measurements and the presence of instability. The percentage of patients for whom LP was contraindicated was calculated. RESULTS: One hundred eight patients underwent LF while 142 underwent LP. Of the 108 patients who underwent LF, 79.6% were radiographically deemed candidates for LP, as were all 142 patients who underwent LP. The ICC for C2-C7 alignment was 0.90; there was 97% agreement with respect to the presence of instability. CONCLUSIONS: In 250 patients with CSM, 228 (91.2%) were radiographically candidates for LP. These data suggest that LP may be an underutilized procedure for the treatment for CSM.

Surgeon-level variance in achieving clinical improvement after lumbar decompression: the importance of adequate risk adjustment.

BACKGROUND CONTEXT: Patient-Reported Outcome Measurement Information System (PROMIS) scores are increasingly utilized in clinical care. However, it is unclear if PROMIS can discriminate surgeon performance on an individual level. PURPOSE: The purpose of this study was to examine surgeon-level variance in rates of achieving minimal clinically important difference (MCID) after lumbar decompression. PATIENT SAMPLE: This is a prospective, observational cohort study performed across a healthcare enterprise (two academic medical centers and three community centers). Patients 18 years or older undergoing one- to two-level primary decompression for lumbar disc herniation (LDH) or lumbar spinal stenosis (LSS) were included. OUTCOME MEASURES: The primary outcome was achievement of MCID, using a distribution-based method, on paired PROMIS physical function scores. METHODS: Descriptive statistics were generated to examine the baseline characteristics of the study cohort. Bivariate analyses were used to examine the impact of surgeon-level variance on rates of MCID. Multivariable analyses were used to examine the risk-adjusted impact of surgeon-level variance on rates of MCID. RESULTS: Overall, 636 patients treated by nine surgeons were included. The median patient age was 58 [interquartile range (IQR): 46-70] and 62.3% (n=396) were female. Among all patients, 56.9% (n=362) underwent surgery for LDH. The overall rate of achieving MCID was 75.8% (n=482). Of the surgeons, the median years in practice were 12 (range 4-31) and 55.6% (n=5) were in academic practice settings. On bivariate analysis, patients treated by one of the surgeons had lower rates of achieving MICD (odds ratio=0.37, 95% confidence interval: 0.15-0.91, p=.03). However, on multivariable analysis adjusting for operative indication (LDH vs. LSS), body mass index, number of comorbidities, percent unemployment in patient zip code, and preoperative PROMIS physical function scores, all surgeons were equally likely to obtain MCID. CONCLUSIONS: In this cohort, variance in PROMIS scores after primary lumbar decompression is influenced by patient-related factors and not by individual surgeon. Adequate risk adjustment is needed if ascertaining clinical improvement on an individual surgeon basis. LEVEL OF EVIDENCE: 2.

Development of prediction models for clinically meaningful improvement in PROMIS scores after lumbar decompression.

BACKGROUND: The ability to preoperatively predict which patients will achieve a minimal clinically important difference (MCID) after lumbar spine decompression surgery can help determine the appropriateness and timing of surgery. Patient-Reported Outcome Measurement Information System (PROMIS) scores are an increasingly popular outcome instrument. PURPOSE: The purpose of this study was to develop algorithms predictive of achieving MCID after primary lumbar decompression surgery. PATIENT SAMPLE: This was a retrospective study at two academic medical centers and three community medical centers including adult patients 18 years or older undergoing one or two level posterior decompression for lumbar disc herniation or lumbar spinal stenosis between January 1, 2016 and April 1, 2019. OUTCOME MEASURES: The primary outcome, MCID, was defined using distribution-based methods as one half the standard deviation of postoperative patient-reported outcomes (PROMIS physical function, pain interference, pain intensity). METHODS: Five machine learning algorithms were developed to predict MCID on these surveys and assessed by discrimination, calibration, Brier score, and decision curve analysis. The final model was incorporated into an open access digital application. RESULTS: Overall, 906 patients completed at least one PROMs survey in the 90 days before surgery and at least one PROMs survey in the year after surgery. Attainment of MCID during the study period by PROMIS instrument was 74.3% for physical function, 75.8% for pain interference, and 79.2% for pain intensity. Factors identified for preoperative prediction of MCID attainment on these outcomes included preoperative PROs, percent unemployment in neighborhood of residence, comorbidities, body mass index, private insurance, preoperative opioid use, surgery for disc herniation, and federal poverty level in neighborhood of residence. The discrimination (c-statistic) of the final algorithms for these outcomes was 0.79 for physical function, 0.74 for pain interference, and 0.69 for pain intensity with good calibration. The open access digital application for these algorithms can be found here: https://sorg-apps.shinyapps.io/promis_pld_mcid/ CONCLUSION: Lower preoperative PROMIS scores, fewer comorbidities, and certain sociodemographic factors increase the likelihood of achieving MCID for PROMIS after lumbar spine decompression.

Development of machine learning and natural language processing algorithms for preoperative prediction and automated identification of intraoperative vascular injury in anterior lumbar spine surgery.

BACKGROUND: Intraoperative vascular injury (VI) may be an unavoidable complication of anterior lumbar spine surgery; however, vascular injury has implications for quality and safety reporting as this intraoperative complication may result in serious bleeding, thrombosis, and postoperative stricture. PURPOSE: The purpose of this study was to (1) develop machine learning algorithms for preoperative prediction of VI and (2) develop natural language processing (NLP) algorithms for automated surveillance of intraoperative VI from free-text operative notes. PATIENT SAMPLE: Adult patients, 18 years or age or older, undergoing anterior lumbar spine surgery at two academic and three community medical centers were included in this analysis. OUTCOME MEASURES: The primary outcome was unintended VI during anterior lumbar spine surgery. METHODS: Manual review of free-text operative notes was used to identify patients who had unintended VI. The available population was split into training and testing cohorts. Five machine learning algorithms were developed for preoperative prediction of VI. An NLP algorithm was trained for automated detection of intraoperative VI from free-text operative notes. Performance of the NLP algorithm was compared to current procedural terminology and international classification of diseases codes. RESULTS: In all, 1035 patients underwent anterior lumbar spine surgery and the rate of intraoperative VI was 7.2% (n=75). Variables used for preoperative prediction of VI were age, male sex, body mass index, diabetes, L4-L5 exposure, and surgery for infection (discitis, osteomyelitis). The best performing machine learning algorithm achieved c-statistic of 0.73 for preoperative prediction of VI (https://sorg-apps.shinyapps.io/lumbar_vascular_injury/). For automated detection of intraoperative VI from free-text notes, the NLP algorithm achieved c-statistic of 0.92. The NLP algorithm identified 18 of the 21 patients (sensitivity 0.86) who had a VI whereas current procedural terminologyand international classification of diseases codes identified 6 of the 21 (sensitivity 0.29) patients. At this threshold, the NLP algorithm had a specificity of 0.93, negative predictive value of 0.99, positive predictive value of 0.51, and F1-score of 0.64. CONCLUSION: Relying on administrative procedural and diagnosis codes may underestimate the rate of unintended intraoperative VI in anterior lumbar spine surgery. External and prospective validation of the algorithms presented here may improve quality and safety reporting.

Predicting prolonged opioid prescriptions in opioid-naïve lumbar spine surgery patients.

IMPORTANCE: Preoperative determination of the potential for postoperative opioid dependence in previously naïve patients undergoing elective spine surgery may facilitate targeted interventions. OBJECTIVE: The purpose of this study was to develop supervised machine learning algorithms for preoperative prediction of prolonged opioid prescription use in opioid-naïve patients following lumbar spine surgery. DESIGN: Retrospective review of clinical registry data. Variables considered for prediction included demographics, insurance status, preoperative medications, surgical factors, laboratory values, comorbidities, and neighborhood characteristics. Five supervised machine learning algorithms were developed and assessed by discrimination, calibration, Brier score, and decision curve analysis. SETTING: One healthcare entity (two academic medical centers, three community hospitals), 2000 to 2018. PARTICIPANTS: Opioid-naïve patients undergoing decompression and/or fusion for lumbar disk herniation, stenosis, and spondylolisthesis. MAIN OUTCOME: Sustained prescription opioid use exceeding 90 days after surgery. RESULTS: Overall, of 8,435 patients included, 359 (4.3%) were found to have prolonged postoperative opioid prescriptions. The elastic-net penalized logistic regression achieved the best performance in the independent testing set not used for algorithm development with c-statistic=0.70, calibration intercept=0.06, calibration slope=1.02, and Brier score=0.039. The five most important factors for prolonged opioid prescriptions were use of instrumented spinal fusion, preoperative benzodiazepine use, preoperative antidepressant use, preoperative gabapentin use, and uninsured status. Individual patient-level explanations were provided for the algorithm predictions and the algorithms were incorporated into an open access digital application available here: https://sorg-apps.shinyapps.io/lumbaropioidnaive/. CONCLUSION AND RELEVANCE: The clinician decision aid developed in this study may be helpful to preoperatively risk-stratify opioid-naïve patients undergoing lumbar spine surgery. The tool demonstrates moderate discriminative capacity for identifying those at greatest risk of prolonged prescription opioid use. External validation is required to further support the potential utility of this tool in practice.

Laminoplasty versus laminectomy and fusion for cervical spondylotic myelopathy: a cost analysis.

BACKGROUND CONTEXT: Laminectomy with fusion (LF) and laminoplasty (LP) are commonly used to treat cervical spondylotic myelopathy (CSM). The decision regarding which procedure to perform is largely a matter of surgeon's preference, while financial implications are rarely considered. PURPOSE: We aimed to better understand the financial considerations of LF compared to LP in the treatment of CSM. STUDY DESIGN: Retrospective comparative study. PATIENT SAMPLE: Adult patients, 18 years of age or older, who had undergone LF or LP for CSM from 2017 to 2019 at 2 large academic centers were included. Patients who had undergone previous cervical spine surgery or procedures that extended above C2 or below T2 were excluded. OUTCOME MEASURES: The primary outcome was defined as the total cost of the procedure, which was calculated as the sum of the implant and non-implant supply costs. METHODS: Patient demographics, surgical parameters, including estimated blood loss and operative time, and length of stay were collected. Operating room material - both implant and non-implant - cost data was also obtained. Variables were analyzed individually as well as after adjustment based on the number of operative levels involved. Statistical analysis was performed using either Student t test with unequal variance or Wilcoxon rank sum test for continuous variables and chi-squared analysis for categorical variables. RESULTS: Two hundred fifty patients were identified who met inclusion criteria. There was no statistical difference in the mean age at time of surgery (p=.25), gender distribution (p=.33), or re-operation rate between the LF and LP groups (p=.39). Overall, operative time was similar between the LF (165.7 ± 61.9 min) and LP (173.8 ± 58.2 min) groups (p=.29), but the LP cohort had a shorter length of stay at 3.8 ± 2.7 days compared to the LF cohort at 4.8 ± 3.7 days. Implant costs in the LF group were significantly more at $6,204.94 ± $1426.41 compared to LP implant costs at $1994.39 ± $643.09. Mean total costs of LP were significantly less at $2,859.08 ± $784.19 compared to LF total costs of $6,983.16 ± $1,589.17. Furthermore, when adjusted for the number of operative levels, LP remained significantly less costly at $766.12 ± $213.64 per level while LF cost $1,789.05 ± $486.66 per operative level. Additional subgroup analysis limiting the cohorts to patients with either three or four involved vertebral levels demonstrated nearly identical cost savings with LP as compared to LF. CONCLUSIONS: This study demonstrates that LF is on average at least 2.4 times the total operative supply cost of LP and at least 2.3 times the operative supply cost of LP when adjusted for the number of operative levels. In patients deemed appropriate for either LP or LF, these data may be incorporated into decision-making for the treatment of CSM.