Comparative effectiveness research has recently been the subject of intense discussion. With congressional support, there has been increasing funding and publication of studies using comparative effectiveness and related methodology. The neurosurgical field has been relatively slow to accept and embrace this approach. The author outlines the procedures and rationale of comparative effectiveness, illustrates how it applies to neurosurgical topics, and explains its importance.
Sherman C. Stein
The neurosurgical profession has taken a circuitous route to attain its current knowledge about timing for aneurysm surgery. While addressing the timing issue, neurosurgeons were beset by many pressures simultaneously. They were forced to justify not only optimal surgical techniques but the need for surgery at all in the treatment of ruptured aneurysms. The beliefs of surgeons with strong personalities, in addition to intuitive guesses, often served to guide surgery in the absence of scientific evidence. That any progress could be made against a background of desperately ill patients and frustrating early results is remarkable. The author briefly outlines the controversies and misdirection that accompanied this progress toward understanding surgical timing in the treatment of ruptured aneurysms.
Sherman C. Stein and Stewart Apfel
✓ A method of measuring flow rate through cerebrospinal fluid (CSF) shunts is reported. It consists of two thermistors in series applied to the skin over the shunt tubing. The thermistors respond by a drop in measured temperature following application of an ice cube placed on the skin overlying the proximal shunt tube. The time required for the thermal response to travel between the two thermistors is related to the velocity of flow through the shunt tubing. Flow rate can then be calculated using the internal diameter of the tubing. A series of animal experiments employing a constant infusion of mock CSF through subcutaneously implanted shunt tubing showed excellent correlation between calculated flow rates and actual infusion rates. The device is noninvasive and easily adapted to use in patients. The measurements are readily repeatable.
Sherman C. Stein and Mark G. Burnett
Medical decisions often depend, in part, on cost-effectiveness concerns. Decision analysis is frequently used to help resolve these questions. Unfortunately, this technique has received little attention in neurosurgery. Using an example of moderate head injury, the authors illustrate the utility of this powerful tool in estimating the cost effectiveness of neurosurgical management options.
Sherman C. Stein and Steven E. Ross
✓ The purpose of this study is to determine the initial treatment of patients who appear to have sustained moderate head injuries when first evaluated. The authors reviewed the records of 341 patients whose initial Glasgow Coma Scale (GCS) scores ranged from 9 to 12, as well as another 106 patients with GCS scores of 13. All patients underwent cranial computerized tomography (CT) at the time of admission.
In 40.3% of these patients the CT scans were abnormal (30.6% had intracranial lesions), and 8.1 % required neurosurgical intervention (craniotomies for hematoma in 12, elevation of depressed fractures in five, and insertion of intracranial pressure monitors in 19). Four patients died of their intracranial injuries. A similar incidence of lesions found on CT and at surgery suggests that an initial GCS score of 13 be classified with the moderate head injury group. Skull fractures were found to be poor indicators of intracranial abnormalities.
These results suggest that all patients with head injury thought to be moderate on initial examination be admitted to the hospital and undergo urgent CT scanning. Patients with intracranial lesions require immediate neurosurgical consultation, surgery as needed, and admission to a critical-care unit. Scans should be repeated in patients whose recovery is less rapid than expected and in all patients with evidence of clinical deterioration; this was necessary in almost half of the patients in this group, and 32% were found to have progression of radiological abnormalities on serial CT scans.
Predicting the results of cerebrospinal fluid shunting
Sherman C. Stein and Thomas W. Langfitt
✓ After shunting, 43 patients with normal-pressure hydrocephalus were followed for 6 to 30 months. All patients had complete preoperative clinical evaluations, pneumoencephalograms, and isotope cisternography; in addition, 21 had saline infusion studies, and 15 had biopsies. Of the 10 patients in whom the etiology of the hydrocephalus was known eight (80%) were significantly improved. Of the other 33 patients, 21 (64%) showed some improvement; this was substantial and sustained in only eight (24%). The whole series of patients with idiopathic normal pressure hydrocephalus was divided into two groups on the basis of shunt response and the relative predictive values of preoperative tests. No significant association was found between the results of shunting and preoperative clinical factors, pneumoencephalography, isotope cisternography, saline infusion tests, or various combinations of clinical and laboratory abnormalities. The significance of these findings is discussed.
Sherman C. Stein and Wensheng Guo
The object of this study was to mathematically model the prognosis of a newly inserted shunt in pediatric or adult patients with hydrocephalus.
A structured search was performed of the English-language literature for case series reporting shunt failure, patient mortality, and shunt removal rates after shunt insertion. A metaanalytic model was constructed to pool data from multiple studies and to predict the outcome of a shunt after insertion. Separate models were used to predict shunt survival rates for children (patients < 17 years old) and adults.
Shunt survival rates in children and adults were calculated for 1 year (64.2 and 80.1%, respectively), 5 years (49.4 and 60.2%, respectively), and the median (4.9 and 7.3 years, respectively). The longer-term rates predicted by the model agree closely with those reported in the literature.
This model gives a comprehensive view of the fate of a shunt for hydrocephalus after insertion. The advantages of this model compared with Kaplan–Meier survival curves are discussed. The model used in this study may provide useful prognostic information and aid in the early evaluation of new shunt designs and techniques.