An analysis of attrition from Canadian neurosurgery residency programs.

PURPOSE: To determine the magnitude of and reasons for attrition from neurosurgical residency programs in Canada. METHOD: Directors of the 13 Canadian neurosurgery residency programs were asked to complete questionnaires on their programs, magnitude of attrition, reasons for attrition, and selection criteria. Open-ended questions were assessed with content analysis and quantified with dual-scaling techniques. Similar questionnaires were sent to 30 residents who had completed training; six residents who had voluntarily withdrawn were interviewed. RESULTS: Twelve of the 13 directors (92%) responded. Forty-two residents voluntarily withdrew from residency training between 1980 and 1992; withdrawal rates grew during that period. The number of dismissals--approximately 1.8 per year--remained constant. Reasons for voluntary withdrawal focused on excessive workloads and unexpected residency demands, whereas reasons for dismissal related primarily to deficits in professional attitudes and behaviors such as interpersonal skills and ethics. In selecting residents, programs with low attrition rates gave more importance to a candidate's work ethic than did programs with high attrition rates. The low-attrition programs also gave more importance to the relationship developed with residents during training. CONCLUSION: These results suggest that voluntary attrition from neurosurgical residency is significant and is related to issues of lifestyle control. Dismissal is rarely related to cognitive or psychomotor deficits, but usually occurs for concerns about professionalism such as ethics and interpersonal skills and behaviors. Further studies are necessary to confirm these findings across specialties and countries.

Guidelines for the use of carotid endarterectomy: current recommendations from the Canadian Neurosurgical Society.

OBJECTIVE: To develop guidelines on the suitability of patients for carotid endarterectomy (CEA). OPTIONS: For atherosclerotic carotid stenosis that has resulted in retinal or cerebral ischemia: antiplatelet drugs or CEA. For asymptomatic carotid stenosis: CEA or no surgery. OUTCOMES: Risk of stroke and death. EVIDENCE: Trials comparing CEA with nonsurgical management of carotid stenosis. VALUES: Greatest weight was given to findings that were highly significant both statistically and clinically. BENEFITS, HARMS AND COSTS: Benefit: reduction in the risk of stroke. Major harms: iatrogenic stroke, cardiac complications and death secondary to surgical manipulations of the artery or the systemic stress of surgery. Costs were not considered. RECOMMENDATIONS: CEA is clearly recommended for patients with surgically accessible internal carotid artery (ICA) stenoses equal to or greater than 70% of the more distal, normal ICA lumen diameter, providing: (1) the stenosis is symptomatic, causing transient ischemic attacks or nondisabling stroke (including retinal infarction); (2) there is no worse distal, ipsilateral, carotid distribution arterial disease; (3) the patient is in stable medical condition; and (4) the rates of major surgical complications (stroke and death) among patients of the treating surgeon are less than 6%. Surgery is not recommended for asymptomatic stenoses of less than 60%. Symptomatic stenoses of less than 70% and asymptomatic stenoses of greater than 60% are uncertain indications. For these indications, consideration should be given to (1) patient presentation, age and medical condition; (2) plaque characteristics such as degree of narrowing, the presence of ulceration and any documented worsening of the plaque over time; (3) other cerebral arterial stenoses or occlusions, or cerebral infarcts identified through neuroimaging; and (4) surgical complication rates at the institution. CEA should not be considered for asymptomatic stenoses unless the combined stroke and death rate among patients of the surgeon is less than 3%. VALIDATION: These guidelines generally agree with position statements prepared by other organizations in recent years, and with a January 1995 consensus statement by a group of experts assembled by the American Heart Association.

Granulocytic sarcoma of the spine in nonleukemic patients: report of three cases.

OBJECTIVE AND IMPORTANCE: Granulocytic sarcomas involving the spine in patients without myelogenous leukemia are rare. We report three cases and review the literature. CLINICAL PRESENTATION: Three patients presented with spinal epidural tumors, which caused spinal cord compression in one and cauda equina compression in two. INTERVENTION: All patients underwent surgery, and biopsies revealed histological features of granulocytic sarcomas. Bone marrow aspirates and biopsies showed no evidence of acute leukemia at initial presentation, for all three patients. CONCLUSION: Granulocytic sarcomas in nonleukemic patients are rare, and when they affect the spine they are frequently misdiagnosed. Appropriate therapy for these tumors requires early identification.

Benign skull lesions.

OBJECTIVE: To document the epidemiologic and clinical features of benign skull lesions. DESIGN: A case series. SETTING: St. Michael's Hospital, a tertiary care facility affiliated with the University of Toronto. PATIENTS: Thirty-one patients who had a neurosurgical consultation and were discharged from hospital after excision of a benign skull lesion during a 10-year period. MAIN OUTCOME MEASURES: Patient demographics, clinical signs and symptoms, radiographic and pathological tumour characteristics, surgical procedure, length of hospital stay, outcome and follow-up. RESULTS: The 31 patients (6 men, 25 women) had 32 lesions excised. The mean age of the patients was 41.9 years. Osteomas accounted for 63% of the tumours. The most frequent location was the parietal bone. Neurologic symptoms were absent in the majority of calvarial tumours. Useful diagnostic studies included plain skull radiography and computed tomography. Nuclear bone scanning was done in 7 patients. All patients underwent craniectomy, with cranioplasty in most cases. Three patients had new neurologic symptoms postoperatively and 1 patient had incomplete resolution of symptoms. CONCLUSIONS: Benign skull lesions are infrequent, but they require neurosurgical intervention. When necessary, surgical excision can serve to confirm the diagnosis, improve cosmesis and retard the progression of neurologic dysfunction. Of primary importance is the recognition of such lesions by primary care physicians and referral to the surgeon so that an appropriate treatment plan can be made.

"Unexplained" delayed death from fungal meningitis after meningioma resection.

BACKGROUND: Progressive deterioration and ensuing death following a neurosurgical procedure often represents a diagnostic challenge to the team responsible for patient care. Many, but not all, causes are treatable if a diagnosis is made early. METHODS: A 69-year-old woman who died 6 weeks post-operatively following a meningioma resection is reported. An initial routine post-operative course became complicated by progressive neurological deterioration 3-4 weeks later. Despite extensive investigation she died 6 weeks post-operatively without a diagnosis. RESULTS: Autopsy demonstrated extensive Candida meningitis. A review of the literature demonstrates this to be a reported complication in high risk patients, difficult to diagnose, but treatable when identified. CONCLUSIONS: Fungal meningitis should be high in the differential diagnosis in the post-operative patient with delayed, unexplained neurological deterioration, especially when associated with negative CSF cultures.

Somatosensory evoked potential monitoring following severe closed head injury.

This paper describes the results of somatosensory evoked potential (SSEP) monitoring in 65 patients with severe head injury. Intracranial pressure (ICP) monitoring data were available for 63 patients, and arterial-jugular oxygen content (AVDO2) data for 52 patients. Eighty-nine percent of patients with no SSEP activity beyond 50 msec post-stimulus in either hemisphere died or were vegetative survivors (3 month Glasgow Outcome Score). All 17 patients with a good or moderate outcome had long latency cortical activity (i.e. > or = 70 msec post-stimulus) in both hemispheres. Among patients with absent activity in 1 hemisphere, 53% died and 47% were severely disabled (chi 2 = 40, p = 0.0000). In the latter group, age was a significant factor among patients who died or were severely disabled (p < 0.02). Forty-four of 65 patients had either clear-cut deterioration or improvement in SSEPs over the course of monitoring. There were no significant differences in peak ICP between patients with improving or deteriorating SSEPs. In contrast, those with deteriorating SSEPs had a significant drop in AVDO2, compared with patients with improving SSEPs (p < 0.01). Long-term continuous monitoring of SSEPs shows that following severe injury, neurologic function may undergo significant change in approximately two-thirds of patients. Furthermore, ICP does not appear to play a prominent role in neurologic deterioration. AVDO2 measurements indicate that deterioration is more likely associated with perturbation of cerebral oxidative metabolism. SSEP monitoring following severe head injury has proven prognostic value, and is recommended for patients who must be pharmacologically paralyzed for ICP or ventilator management.

Correlation of trauma scoring and outcome in a Canadian trauma centre.

OBJECTIVE: To study the effectiveness of treatment and the outcome in trauma patients, and to correlate these with trauma scoring systems. DESIGN: A prospective study over 1 year of the probability of survival after trauma, based on TRISS methodology. A computerized database provided outcome statistics. SETTING: A regional trauma centre in a Canadian university teaching hospital serving regional and referral patients. PATIENTS: Three hundred consecutive patients treated at a single trauma unit. Two patients were excluded because of lack of physiologic data. Blunt injuries (94%) were most frequently from motor vehicle accidents (46%). INTERVENTIONS: Those appropriate to multidisciplinary trauma management in a level 1 trauma centre. MAIN OUTCOME MEASURES: Survival (Z values) and injury severity (M values) for the total group and subsets were calculated for comparison of outcomes with the Multiple Trauma Outcome Study baseline and other Canadian centres for multisystem and single-system injuries. RESULTS: The mean Injury Severity Score was 21.16 and the mean Revised Trauma Score was 6.75. There were 51 (17%) deaths--a Z value for the study group of 2.26. The M value was 0.78. Forty of the 51 single-system injuries were head injuries and accounted for 15 deaths. The Z value for multisystem injuries only was 0.54, and the M value was 0.77. CONCLUSIONS: The TRISS method for analysing blunt trauma is comparable to other trauma scoring systems. The correlation of outcome analysis with other scoring systems is affected by exclusion rates, pre-referral resuscitation and single- versus multiple-injury mix of cases.

Dural cavernous hemangioma: an under-recognized lesion mimicking meningioma.

We report a 77-year-old woman who presented with partial seizures and was found to have an enhancing dural-based parietal convexity mass. The lesion enlarged on serial examination by computed tomography (CT) over a one year period. The clinical features and radiologic appearance were compatible with a pre-operative diagnosis of meningioma; however, pathologic findings were typical of a dural cavernous hemangioma. Accumulating evidence suggests that these lesions are an uncommon but distinct type of vascular malformation most often arising from the cavernous sinus, tentorium, or cerebello-pontine angle. With CT, magnetic resonance imaging and angiography, these lesions can closely resemble meningioma in terms of signal characteristics, enhancement pattern, and location. This is of importance both in the practical management of meningiomas where the diagnosis is often based on radiologic studies alone, and in clinical trials where incorrect entry diagnosis should be avoided.

The value of reoperation for recurrent glioblastoma.

To determine the value of reoperation alone (no further surgical procedures or radiotherapy), 43 patients (27 men, 16 women) with recurrent supratentorial glioblastomas who underwent a second craniotomy for recurrent tumour were reviewed retrospectively. The patients ranged in age from 27 to 66 years (median 53 years). All patients were treated initially by surgical resection and external radiation (50 Gy in 25 fractions through parallel opposed regional fields). In addition, 10 patients (23%) received chemotherapy, 3 patients (7%) received photodynamic therapy and 9 patients (21%) received interstitial brachytherapy postoperatively. Although none of the patients had further surgical procedures or radiotherapy after reoperation for tumour recurrence, 5 of the 43 did receive single-agent chemotherapy. The median survival after the first operation was 57 weeks. The median interval between first and second operations was 32 weeks. Median survival after reoperation was 19 weeks. An interval of more than 50 weeks between the two operations correlated with a significant (p < 0.05) increase in survival. The death rate for reoperation was 4.6%. The infection rate was 9.3%. The authors conclude that reoperation alone confers a modest but valuable increase in survival, especially if the interval between operations is greater than 50 weeks.

Risk factors predicting operable intracranial hematomas in head injury.

A study was performed to examine the incidence of operable traumatic intracranial hematomas accompanying head injuries of differing degrees of severity, and to see if factors predicting operable mass lesions could be identified. Logistic analysis was used to identify independent predictors of operable traumatic intracranial hematomas. Data were gathered prospectively on 1039 patients admitted with head injury between January, 1986, and December, 1990. Patient age, Glasgow Coma Scale (GCS) score, pupillary inequality, and injury by falling were all independent predictors of the presence of operable intracranial hematomas (p = 0.0000, 0.0000, 0.0182, and 0.0001, respectively). Injury to vehicle occupants was less likely to result in operable mass lesions (p = 0.0001) than injury by other means. The incidence of traumatic intracranial hematomas in patients over 50 years old was three to four times higher than in those under 30 years of age. Not surprisingly, the incidence of operable hematomas increased with decreasing GCS scores. However, even at a GCS score of 13 to 15, patients with other risk factors had a substantial incidence of operable mass lesions. There was a 29% incidence of operable intracranial hematomas for patients with a GCS score of 13 to 15, aged over 40 years and injured in a fall. It is suggested that patients who are middle-aged or older, or those injured in falls, are at particular risk for traumatic intracranial hematomas even if their GCS score is high. These patients should have early definitive investigation with computerized tomography in order to identify operable hematomas and to initiate surgical treatment prior to neurological deterioration from mass effect.

Relative incidence of intracranial mass lesions and severe torso injury after accidental injury: implications for triage and management.

In a previous study of head injury patients we found that old age, low Glasgow Coma Scale (GCS) score, pupillary inequality, and falls were significant predictors of intracranial mass lesions (IMLs). Injury to motor vehicle occupants was less likely to result in IML. The present study defines predictors of severe torso injury (STI) in 646 patients admitted to a trauma unit and compares these with predictors of IML obtained in the previous study. Tachycardia and low blood pressure were associated with an increased incidence of STI (p = 0.003, p = 0.0000). The incidence of STI in falls differed from that of IML (13.2% vs. 47.7%, p less than 0.001). There was a greater incidence of STI than IML in MVAs (33.6% vs. 14.8%, p less than 0.001). Patients 70 years of age or older had a higher incidence of IMLs than STIs (p less than 0.001). Patients less than 30 years old had a significantly greater incidence of STIs than IMLs (p less than 0.001). These data suggest that in MVA victims who are less than 30 years old, are hypotensive, and tachycardic, the diagnosis and emergent treatment of severe torso injury should take precedence over measures designed to detect and treat intracranial mass lesions. The converse is true for older patients injured in falls.

Adult peripheral neuroepithelioma in Meckel's cave.

A case of peripheral neuroepithelioma arising from the trigeminal nerve in Meckel's cave is presented. The discussion emphasizes the pathological criteria for the diagnosis of a peripheral neuroepithelioma and the current controversy about the classification of this and related tumors.

Petrous meningiomas: a review of seventeen cases.

Seventeen patients with petrous meningiomas managed at St. Michael's Hospital, during the years 1973-1987, were retrospectively reviewed. There were 15 females and 2 males; their ages ranged from 42 to 68 years (mean age: 53 years). The clinical presentation most commonly included headache and eighth cranial nerve dysfunction; the average duration of symptoms was 6 years (3 month-27 years). Computed tomography was performed in 15 cases. The mean tumour size was 2.5 centimeters (0.5-4 cm). The most common site of tumour origin was at or medial to the porus acousticus. Meningioma was suspected preoperatively in 10 of the 15 patients who had preoperative CT scans. Complete excision was obtained in 12 cases. There were no operative deaths after initial resections. Postoperative morbidity included worsening of pre-existing hearing loss in six patients, transient facial nerve palsies in six, permanent facial nerve palsies in four and new facial or corneal hypesthesia in three. Two patients developed cerebrospinal fluid fistulae. Tumour recurrence occurred into two patients in whom a complete resection was anticipated. Also, in two patients with incompletely resected tumours second operations were required. Fourteen patients are alive, 13 of whom care for themselves independently. The average follow-up was 5 years (6 months-9 years). It seems appropriate to recommend initial radical surgical excision of these benign tumours, where possible, in order to prevent tumour recurrence.

Effect of population characteristics on head injury mortality.

We have analyzed predictors of mortality following closed head injury in a series of 1,031 consecutive patients with closed head injury admitted to hospital from January 1986 through December 1990. All patients were treated in a uniform manner and surgical intervention was performed as soon as possible in patients with intracranial mass lesions. Logistic analysis was used to identify patient and injury characteristics that were independent predictors of mortality within this patient group. Significant predictors were Glasgow Coma Score at admission (p = 0.0000), age (p = 0.0000), bilaterally unreactive pupils (p = 0.0000), presence of multiple systemic injuries (p = 0.0004), presence of an intracranial mass lesion (p = 0.0006), and presence of unilateral pupillary abnormalities (p = 0.0279). In an attempt to clarify the relationship between the incidence of these characteristics in series of severely head-injured patients reported during the last 2 decades and the mortality reported in those series, regression analysis was carried out comparing the mean age reported in the series, incidence of mass lesions, and reported mortality. Sixty-four percent of the variability in reported mortality rates could be accounted for by differences in mean age of the patients and mass lesion incidence (p = 0.0035). We conclude that apparent improvements in head injury mortality in the last 2 decades may be partly or wholly due to different population characteristics in the reported series. Multiple injuries appear to be important contributors to patient mortality, and in the interest of improved description of head injury populations, the Injury Severity Score should be reported with age, mass lesion incidence, and Glasgow Coma Score.

Multiple-level noncontiguous spinal fractures.

Double-level noncontiguous spinal injury is a more common occurrence than generally appreciated. A large number of these patients have neurological injury. Because these injuries occur in patients who meet the criteria for categorization as multiple trauma patients, we feel that a complete radiographic survey of the spine must be accomplished in the emergency room in any situation where clinical assessment is impaired. Treatment must be individualized with the same guidelines for treatment as for the isolated injury in the majority of circumstances.

Euthyroid hyperthyroxinemia due to familial excess of thyroxine-binding globulin.

The correct diagnosis of benign hyperthyroxinemia in this patient and his family members will spare them the unnecessary testing and treatment for thyrotoxicosis that has befallen some such patients. Results of the usual blood tests for assessment of thyroid function, such as T4, T3, and thyrotropin determinations, were not uniformly diagnostic, and were potentially misleading. An increased T4 level, a nonsuppressed TSH level, normal levels of FT4 and FT4D, and a low level of T3RU were clues that led to a request for specific measurement of serum TBG levels in multiple family members; family testing was essential for the diagnosis of euthyroid hyperthyroxinemia due to familial hepatic overproduction of TBG.