CAEP/ACMU 2000 Scientific Abstracts: 48-50

CAEP Abstracts

CJEM 2001;2(3):169-187

048 Interobserver Agreement in the Assessment of Potential Cervical Spine Injuries.
Stiell I, Wells G, Brison R, Greenberg G, Vandemheen K, Clement C, Cass D, Dreyer J, Eisenhauer M, MacPhail I, McKnight D, Morrison L, Reardon M, Schull M, Worthington J. Clinical Epidemiology Unit, University of Ottawa, Ottawa.

OBJECTIVES: To determine the interobserver agreement or consistency in physician assessment of clinical signs and symptoms in potential neck injury patients. This methodological sub-study was an important component in the derivation of the Canadian C-Spine Rule for cervical spine radiography in the Canadian C-Spine/CT Head (CCC) Study.
METHODS: This prospective cohort study was conducted in 10 Canadian EDs and involved alert and stable adult trauma patients. Physicians prospectively evaluated patients for 22 standardized clinical findings prior to radiography and performed blinded and independent interobserver assessments when feasible. Analyses included calculation of the kappa coefficient with 95% CIs. RESULTS: 300 physician assessments were conducted on 150 patients whose characteristics were similar to those of the main study population for mean age (36.5; range 16­90), female (52.2%), MVC (63.1%), falls (15.9%), ambulance arrival (73.2%), admission (22.3%), but with a higher incidence of clinically important cervical spine injury (9.6%). Kappa values for the clinical findings were:

CONCLUSIONS: Findings with only fair to moderate agreement were immediate neck pain, intoxication, and distracting painful injury. All components of the Canadian C-Spine Rule showed excellent interobserver agreement and this suggests that physicians should be able to consistently interpret the overall rule. This reliability will be explicitly and prospectively evaluated in ongoing studies.

Key words: cervical spine, radiography, utilization, clinical prediction rule

049 Comparison of the Canadian CT Head Rule to Physician Judgement.
Stiell I, Worthington J, Dreyer J, Vandemheen K, Clement C, De Maio V, Schull M, Reardon M, Morrison L, McKnight D, MacPhail I, Greenberg G, Eisenhauer M, Cass D, Brison R, Wells G. Clinical Epidemiology Unit, University of Ottawa, Ottawa.

OBJECTIVES: To compare the predictive accuracy of emergency physicians' clinical judgement to the Canadian CT Head Rule, a recently developed and highly sensitive clinical decision rule for the use of CT in patients with minor head injury.
METHODS: This prospective cohort study was conducted as a component of the Canadian C-Spine/CT Head (CCC) Study in 10 Canadian EDs and involved adults with loss of consciousness, amnesia, or confusion and a GCS score of 13­15. Physicians completed a 22-item assessment form prior to CT scan. The outcome standards were "need for neurological intervention" and "clinically important brain injury" on CT. Physicians also estimated, based upon clinical judgement alone, the probability, from 0% to 100%, of brain injury and of neurological intervention. Analyses included comparison of areas under the receiver operating characteristic (ROC) curve with 95% CIs. RESULTS: Among 1,416 patients enrolled over 18 months, the mean age was 38.1 (range 16­96), 66.6% were male, 97 (6.9%) had a clinically important brain injury, and 11 (0.8%) underwent neurological intervention. Comparing physician judgement to the Canadian CT Head Rule for predicting brain injury on CT, the respective areas under the ROC curve were 0.77 (95% CI 0.72­0.83) vs. 0.87 (0.85­0.89) (p < 0.05); sensitivities were 91.8% vs. 97.2% (p < 0.05); and specificities were 36.5% vs. 49.0% (p < 0.001). Comparing physician judgement to the Canadian CT Head Rule for predicting neurological intervention, the respective areas under the ROC curve were 0.75 (0.58­0.92) vs. 0.96 (0.95­0.97) (p < 0.01); sensitivities were 72.7% vs. 100% (p < 0.01); and specificities were 72.4% vs. 80.9% (p < 0.001). CONCLUSIONS: Compared to the Canadian CT Head Rule, physicians fared poorly in predicting the presence of important brain injury on CT or the need for neurological intervention among minor head injury patients.
Key words: brain injury, computed tomography, utilization, clinical prediction rule

050 Validity Evaluation of the Brain Injury Proxy Outcome Assessment Tool in the CCC Study.
Stiell I, Morrison L, Schull M, Vandemheen K, Clement C, Brison R, Cass D, Dreyer J, Eisenhauer M, Greenberg G, MacPhail I, McKnight D, Reardon M, Worthington J. Clinical Epidemiology Unit, University of Ottawa, Ottawa.

OBJECTIVES: This methodological sub-study assessed the validity of a "Proxy Outcome Assessment Tool," administered to patients in the Canadian C-Spine/CT Head (CCC) Study, which is designed to develop a decision rule for imaging minor head injury patients. In Canada, approximately 35% of minor head injury patients do not undergo CT in the ED and require follow-up to determine outcome.
METHODS: The CCC Study is a prospective cohort study conducted in 10 Canadian EDs and includes patients with loss of consciousness, amnesia, or confusion and a GCS score of 13­15. The outcome measures were "need for neurological intervention" and "clinically important brain injury" demonstrated on CT. In this sub-study, all patients with abnormal CT scans and a sample of patients with normal scans were assessed by telephone at 14 days. The "Proxy Outcome Assessment Tool" consists of 5 general questions as well as the validated Katzman Short O-M-C Test (scored from 0 to 28 errors). Descriptive statistics with 95% CIs were calculated.
RESULTS: Of 2,536 CCC Study minor head injury patients, 172 (6.8%) were entered in this sub-study. These patients were similar to the overall study group except for a higher proportion with important brain injury (39.5%) and need for neurological intervention (3.5%).

CONCLUSIONS: This sub-study confirms the validity of the "Proxy Outcome Assessment Tool" to rule out important negative outcomes in neurologically intact CCC Study patients. This was an essential step in developing the Canadian CT Head Rule.

Key words: brain injury, computed tomography, utilization, clinical prediction rule