/liter) are given in parentheses. Once the univariate prognostic ability of the variables was determined, a stepwise selection procedure was applied to the same parameters. This started with the best univariate predictor, eCK-BB, and the other variables were added one by one to verify whether a combination of some or all risk factors could improve outcome prediction. The most important results of the stepwise selection are summarized in Table 2 . The percentages of correct predictions for each outcome group and for the total population are also given. Thoracic
Pol Hans, Adelin Albert, Colette Franssen and Jacques Born
Nikhil Paliwal, Prakhar Jaiswal, Vincent M. Tutino, Hussain Shallwani, Jason M. Davies, Adnan H. Siddiqui, Rahul Rai and Hui Meng
will identify IAs that may not occlude with a single, uniformly implanted FD, promoting the interventionalist to consider these alternative methods. To find the best predictive models, we have also compared the performance of 4 different ML algorithms for FD treatment outcome prediction, including LR, SVM, K-NN, and NN. Our results show that NN and G-SVM (90% accuracy) performed slightly better than LR, L-SVM, and K-NN (85% accuracy). The almost similar performances of all of the algorithms suggest that a larger dataset is required to identify the best algorithm
Vengalathur Ganesan Ramesh
more accurate in outcome prediction than logistic regression models. In connection with this, I would like to state that the Madras Head Injury Prognostic Scale (MHIPS) of Ramesh et al. 1 is also a simple and accurate prognostic model that is very useful for routine bedside outcome prediction, performed using readily available parameters. The MHIPS is regularly used by us, and the accuracy of prediction of outcome in traumatic brain injury is 87.5%. Disclosure The author reports no conflict of interest. References 1 Ramesh VG , Thirumaran KP , Raja
Paolo Ferroli and Morgan Broggi
this somewhat startling and revealing result is heightened because our own neurosurgical group has been committed to finding an evidence-based methodology that might lay the foundation for outcome prediction of any neurosurgical intervention on more solid, scientifically based ground. Like the Norwegian group, we have been acutely aware of the influence that many “human factor” variables have on a neurosurgeon’s prediction of surgical outcome. 4 In a study recently published in Neurosurgical Focus , 3 we compared the pre- and postoperative (at discharge) KPS scores
Sung C. Choi, John D. Ward and Donald P. Becker
objective basis for family counseling. Several techniques proposed for prognosis in head injury involve relatively complex statistical formulas. 3, 7, 12, 15 Although such formulas can be applied using calculators or computers, it would be convenient to have a simple chart from which outcome prediction is easily determined from a simple clinical examination performed on postinjury Days 1 and 4. Clinical Material and Methods Patients The analysis is based on 264 patients with head injury who were admitted to the Neurosurgical Service of the Medical College of
A comparative analysis of the clinical examination, multimodality evoked potentials, CT scanning, and intracranial pressure
Raj K. Narayan, Richard P. Greenberg, J. Douglas Miller, Gregory G. Enas, Sung C. Choi, Pulla R. S. Kishore, John B. Selhorst, Harry A. Lutz III and Donald P. Becker
of these influences renders even the best prognostic system imperfect. Outcome predictions altered by such complications always err toward over-optimism. Although undesirable, such over-optimistic predictions can be accounted for and do not have adverse implications with regard to patient management. Far more serious, however, are errors of undue pessimism. These are not related to unexpected complications, and represent an inherent weakness of the prognostic system. The optimal prognostic indicant, or combination of indicants would, therefore, predict outcome with
Anand I. Rughani, Travis M. Dumont, Zhenyu Lu, Josh Bongard, Michael A. Horgan, Paul L. Penar and Bruce I. Tranmer
, individual components of the GCS score at the scene of injury (eye, verbal, and motor score), total score and individual components of the GCS in the ED, and first systolic blood pressure measured in the ED. Variables were selected on the basis of known or expected influence on outcome, including age, 11 sex, 6 and hypotension. 3 It should be emphasized that several variables crucial to outcome prediction are absent from the NTDB, such as pupillary reactivity and specific radiographic findings. The solitary output variable we examined was in-hospital survival. All
Benjamin C. Warf and Abhaya V. Kulkarni
In the setting of a developing country where preoperative imaging may be limited, the authors wished to determine whether cisternal scarring or aqueduct patency at the time of surgery was sufficiently predictive of the failure of endoscopic third ventriculostomy (ETV) to justify shunt placement at the time of the initial operation.
The status of the prepontine cistern and aqueduct at the time of ventriculoscopy was prospectively recorded in 403 children in whom an ETV had been completed. Kaplan-Meier methods were used to construct survival curves. A Cox proportional hazards model was used to provide estimates of HRs for the time to ETV failure. Several independent variables were tested in a single multivariable model, including those previously shown to be associated with ETV survival, that is, age, hydrocephalus etiology, and extent of choroid plexus cauterization (CPC). In addition, intraoperative variables of particular interest were included in the analysis: status of the aqueduct at surgery (closed vs open) and status of the prepontine cistern at surgery (scarred vs clean/unscarred). Multicollinearity was not a concern since the variance inflation factors for all variables were < 2. The examination of stratified survival curves confirmed the appropriateness of the proportional hazards assumption for each variable.
Overall actuarial 3-year success was 57%. Consistent with previous results, age, hydrocephalus etiology, and extent of CPC were significantly associated with ETV success. A closed aqueduct and an unscarred cistern were each independently associated with significantly better ETV success (HRs of 0.66 and 0.44, respectively). The presence of cisternal scarring more than doubled the risk of ETV failure, and an open aqueduct increased the risk of failure by 50%.
Intraoperative observations of the aqueduct and prepontine cistern are independent predictors of the risk of ETV failure and can be used to further refine outcome predictions based on age, hydrocephalus etiology, and extent of CPC. Further studies will test validity in several African centers and determine what threshold of failure risk should prompt shunt placement at the initial operation.
Michael G. Fehlings, Lindsay Tetreault, Patrick C. Hsieh, Vincent Traynelis and Michael Y. Wang
Aditya Vedantam, Ashish Jonathan and Vedantam Rajshekhar
Few studies have evaluated the prognostic significance of different types of T2-weighted MR imaging changes in patients with cervical spondylotic myelopathy (CSM). The object of this study was to determine whether the type of increased signal intensity (ISI) was an independent predictor of outcome following central corpectomy in patients with CSM or ossification of the posterior longitudinal ligament (OPLL).
Magnetic resonance images obtained in 197 patients who had undergone central corpectomy for CSM or OPLL were assessed for ISI within the cord on sagittal T2-weighted images and hypointensity on T1-weighted images. The T2-weighted changes were categorized as no change (Type 0), fuzzy (Type 1), or sharp (Type 2) based on the ISI characteristics. Outcomes were assessed as a change in Nurick grade of 1 grade or more from preoperatively to postoperatively, and cure as a follow-up Nurick grade of 0 or 1. Multilevel regression analysis was performed to identify predictors of change in Nurick grade ≥ 1 and cure.
There were 30 patients (15.2%) with Type 0, 104 patients (52.8%) with Type 1, and 63 patients (32%) with Type 2 ISI on MR images. Age, duration of symptoms, and preoperative Nurick grade were similar among the groups. A preoperative Nurick grade of 4 or 5 (OR 0.23, p < 0.001) and presence of Type 2 ISI on T2-weighted images (OR 0.48, p = 0.04) negatively influenced the probability of cure after surgery. Hypointensity on T1-weighted images was only seen in patients who had Type 2 ISI changes. Among the 63 patients with Type 2 ISI, the presence of T1-weighted hypointensity (16 patients) was found to negatively impact cure (OR 0.1, p = 0.04).
Increased signal intensity on preoperative T2-weighted MR images was seen in more than 80% of the cases. However, only Type 2 ISI on T2-weighted images had a prognostic significance of being associated with a decreased likelihood of cure in patients with CSM or OPLL. Hypointensity on T1-weighted images predicted a lower probability of cure among patients with Type 2 ISI on T2-weighted images.