Edward C. Benzel and Zoher Ghogawala
Anthony L. Asher, Matthew J. McGirt and Zoher Ghogawala
JNSPG 75th Anniversary Invited Review Article
Zoher Ghogawala, Melissa R. Dunbar and Irfan Essa
There are a wide variety of comparative treatment options in neurosurgery that do not lend themselves to traditional randomized controlled trials. The object of this article was to examine how clinical registries might be used to generate new evidence to support a particular treatment option when comparable options exist. Lumbar spondylolisthesis is used as an example.
The authors reviewed the literature examining the comparative effectiveness of decompression alone versus decompression with fusion for lumbar stenosis with degenerative spondylolisthesis. Modern data acquisition for the creation of registries was also reviewed with an eye toward how artificial intelligence for the treatment of lumbar spondylolisthesis might be explored.
Current randomized controlled trials differ on the importance of adding fusion when performing decompression for lumbar spondylolisthesis. Standardized approaches to extracting data from the electronic medical record as well as the ability to capture radiographic imaging and incorporate patient-reported outcomes (PROs) will ultimately lead to the development of modern, structured, data-filled registries that will lay the foundation for machine learning.
There is a growing realization that patient experience, satisfaction, and outcomes are essential to improving the overall quality of spine care. There is a need to use practical, validated PRO tools in the quest to optimize outcomes within spine care. Registries will be designed to contain robust clinical data in which predictive analytics can be generated to develop and guide data-driven personalized spine care.
Paul M. Arnold, Zoher Ghogawala and Candan Tamerler
Zoher Ghogawala, Daniel K. Resnick, Steven D. Glassman, James Dziura, Christopher I. Shaffrey and Praveen V. Mummaneni
Anthony L. Asher, Paul C. McCormick, Nathan R. Selden, Zoher Ghogawala and Matthew J. McGirt
Patient care data will soon inform all areas of health care decision making and will define clinical performance. Organized neurosurgery believes that prospective, systematic tracking of practice patterns and patient outcomes will allow neurosurgeons to improve the quality and efficiency and, ultimately, the value of care. In support of this mission, the American Association of Neurological Surgeons, in cooperation with a broad coalition of other neurosurgical societies including the Congress of Neurological Surgeons, Society of Neurological Surgeons, and American Board of Neurological Surgery, created the NeuroPoint Alliance (NPA), a not-for-profit corporation, in 2008. The NPA coordinates a variety of national projects involving the acquisition, analysis, and reporting of clinical data from neurosurgical practice using online technologies. It was designed to meet the health care quality and related research needs of individual neurosurgeons and neurosurgical practices, national organizations, health care plans, biomedical industry, and government agencies. To meet the growing need for tools to measure and promote high-quality care, NPA collaborated with several national stakeholders to create an unprecedented program: the National Neurosurgery Quality and Outcomes Database (N2QOD). This resource will allow any US neurosurgeon, practice group, or hospital system to contribute to and access aggregate quality and outcomes data through a centralized, nationally coordinated clinical registry. This paper describes the practical and scientific justifications for a national neurosurgical registry; the conceptualization, design, development, and implementation of the N2QOD; and the likely role of prospective, cooperative clinical data collection systems in evolving systems of neurosurgical training, continuing education, research, public reporting, and maintenance of certification.
Claire Blumenthal, Jill Curran, Edward C. Benzel, Rachel Potter, Subu N. Magge, J. Frederick Harrington Jr., Jean-Valery Coumans and Zoher Ghogawala
It is not known whether adding fusion to lumbar decompression is necessary for all patients undergoing surgery for degenerative lumbar spondylolisthesis with symptomatic stenosis. Determining specific radiographic traits that might predict delayed instability following decompression surgery might guide clinical decision making regarding the utility of up-front fusion in patients with degenerative Grade I spondylolisthesis.
Patients with Grade I degenerative lumbar spondylolisthesis (3–14 mm) with symptomatic stenosis were prospectively enrolled from a single site between May 2002 and September 2009 and treated with decompressive laminectomy without fusion. Patients with mechanical back pain or with gross motion (> 3 mm) on flexion-extension lumbar radiographs were excluded. The baseline radiographic variables measured included amount of slippage, disc height, facet angle, motion at spondylolisthesis (flexion-extension), and sagittal rotation angle. Data were analyzed using multivariate forward selection stepwise logistic regression, chi-square tests, Student t-test, and ANOVA.
Forty patients were enrolled and treated with laminectomy without fusion, and all patients had complete radiographic data sets that were available for analysis. Reoperation was performed in 15 (37.5%) of 40 patients, with a mean follow-up duration of 3.6 years. Reoperation was performed for pain caused by instability at the index level in all 15 cases. Using multivariate stepwise logistic regression with a threshold p value of 0.35, motion at spondylolisthesis, disc height, and facet angle were predictors of reoperation following surgery. Facet angle > 50° was associated with a 39% rate of reoperation, disc height > 6.5 mm was associated with a 45% rate of reoperation, and motion at spondylolisthesis > 1.25 mm was associated with a 54% rate of reoperation. Patients with all 3 risk factors for instability had a 75% rate of reoperation, whereas patients with no risk factors for instability had a 0% rate of reoperation (p = 0.14).
Patients with motion at spondylolisthesis > 1.25 mm, disc height > 6.5 mm, and facet angle > 50° are more likely to experience instability following decompression surgery for Grade I lumbar spondylolisthesis. Identification of key risk factors for instability might improve patient selection for decompression without fusion surgery. Clinical trial registration no.: NCT00109213
Nathan R. Selden, Zoher Ghogawala, Robert E. Harbaugh, Zachary N. Litvack, Matthew J. McGirt and Anthony L. Asher
Outcomes-directed approaches to quality improvement have been adopted by diverse industries and are increasingly the focus of government-mandated reforms to health care education and delivery. The authors identify and review current reform initiatives originating from agencies regulating and funding graduate medical education and health care delivery. These reforms use outcomes-based methodologies and incorporate principles of lifelong learning and patient centeredness.
Important new initiatives include the Accreditation Council for Graduate Medical Education Milestones; the pending adoption by the American Board of Neurological Surgery of new requirements for Maintenance of Certification that are in part outcomes based; initiation by health care systems and consortia of public reporting of patient outcomes data; institution by the Centers for Medicare & Medicaid Services of requirements for comparative effectiveness research and the physician quality reporting system; and linking of health care reimbursement in part to patient outcomes data and quality measures. Opportunities exist to coordinate and unify patient outcomes measurement throughout neurosurgical training and practice, enabling effective patient-centered improvements in care delivery as well as efficient compliance with regulatory mandates. Coordination will likely require the development of a new science of practice based in the daily clinical environment and utilizing clinical data registries. A generation of outcomes science and quality experts within neurosurgery should be trained to facilitate attainment of these goals.
Khoi D. Than, Jill N. Curran, Daniel K. Resnick, Christopher I. Shaffrey, Zoher Ghogawala and Praveen V. Mummaneni
To date, the factors that predict whether a patient returns to work after lumbar discectomy are poorly understood. Information on postoperative work status is important in analyzing the cost-effectiveness of the procedure.
An observational prospective cohort study was completed at 13 academic and community sites (NeuroPoint–Spinal Disorders [NeuroPoint-SD] registry). Patients undergoing single-level lumbar discectomy were included. Variables assessed included age, sex, body mass index (BMI), SF-36 physical function score, Oswestry Disability Index (ODI) score, presence of diabetes, smoking status, systemic illness, workers' compensation status, and preoperative work status. The primary outcome was working status within 3 months after surgery. Stepwise logistic regression analysis was performed to determine which factors were predictive of return to work at 3 months following discectomy.
There were 127 patients (of 148 total) with data collected 3 months postoperatively. The patients' average age at the time of surgery was 46 ± 1 years, and 66.9% of patients were working 3 months postoperatively. Statistical analyses demonstrated that the patients more likely to return to work were those of younger age (44.5 years vs 50.5 years, p = 0.008), males (55.3% vs 28.6%, p = 0.005), those with higher preoperative SF-36 physical function scores (44.0 vs 30.3, p = 0.002), those with lower preoperative ODI scores (43.8 vs 52.6, p = 0.01), nonsmokers (83.5% vs 66.7%, p = 0.03), and those who were working preoperatively (91.8% vs 26.2%, p < 0.0001). When controlling for patients who were working preoperatively (105 patients), only age was a statistically significant predictor of postoperative return to work (44.1 years vs 51.1 years, p = 0.049).
In this cohort of lumbar discectomy patients, preoperative working status was the strongest predictor of postoperative working status 3 months after surgery. Younger age was also a predictor. Factors not influencing return to work in the logistic regression analysis included sex, BMI, SF-36 physical function score, ODI score, presence of diabetes, smoking status, and systemic illness.
Clinical trial registration no.: 01220921 (clinicaltrials.gov)
Leah Y. Carreon, Steven D. Glassman, Zoher Ghogawala, Praveen V. Mummaneni, Matthew J. McGirt and Anthony L. Asher
Transforaminal lumbar interbody fusion (TLIF) has become the most commonly used fusion technique for lumbar degenerative disorders. This suggests an expectation of better clinical outcomes with this technique, but this has not been validated consistently. How surgical variables and choice of health utility measures drive the cost-effectiveness of TLIF relative to posterolateral fusion (PSF) has not been established. The authors used health utility values derived from Short Form-6D (SF-6D) and EQ-5D and different cost-effectiveness thresholds to evaluate the relative cost-effectiveness of TLIF compared with PSF.
From the National Neurosurgery Quality and Outcomes Database (N2QOD), 101 patients with spondylolisthesis who underwent PSF were propensity matched to patients who underwent TLIF. Health-related quality of life measures and perioperative parameters were compared. Because health utility values derived from the SF-6D and EQ-5D questionnaires have been shown to vary in patients with low-back pain, quality-adjusted life years (QALYs) were derived from both measures. On the basis of these matched cases, a sensitivity analysis for the relative cost per QALY of TLIF versus PSF was performed in a series of cost-assumption models.
Operative time, blood loss, hospital stay, and 30-day and 90-day readmission rates were similar for the TLIF and PSF groups. Both TLIF and PSF significantly improved back and leg pain, Oswestry Disability Index (ODI) scores, and EQ-5D and SF-6D scores at 3 and 12 months postoperatively. At 12 months postoperatively, patients who had undergone TLIF had greater improvements in mean ODI scores (30.4 vs 21.1, p = 0.001) and mean SF-6D scores (0.16 vs 0.11, p = 0.001) but similar improvements in mean EQ-5D scores (0.25 vs 0.22, p = 0.415) as patients treated with PSF. At a cost per QALY threshold of $100,000 and using SF-6D–based QALYs, the authors found that TLIF would be cost-prohibitive compared with PSF at a surgical cost of $4830 above that of PSF. However, with EQ-5D–based QALYs, TLIF would become cost-prohibitive at an increased surgical cost of $2960 relative to that of PSF. With the 2014 US per capita gross domestic product of $53,042 as a more stringent cost-effectiveness threshold, TLIF would become cost-prohibitive at surgical costs $2562 above that of PSF with SF-6D–based QALYs or at a surgical cost exceeding that of PSF by $1570 with EQ-5D–derived QALYs.
As with all cost-effectiveness studies, cost per QALY depended on the measure of health utility selected, durability of the intervention, readmission rates, and the accuracy of the cost assumptions.