CAEP/ACMU 2000 Scientific Abstracts: 8-10

CAEP Abstracts

CJEM 2001;2(3):169-187

008 Pain as a Predictor of Complications of Renal Colic.
Papa L, Stiell I, Lee J, Wells G, Mahoney J. Clinical Epidemiology Unit, University of Ottawa, Ottawa.

OBJECTIVES: To prospectively assess pain as a predictor of 4-week complications from delayed passage of urinary calculi.
METHODS: Consecutive patients with suspected renal colic were assessed prospectively at 2 teaching hospital EDs based on acute flank pain and hematuria. A 10-cm visual analogue scale (VAS) was used to assess patients' pain on arrival and at discharge from the ED. Patients were then followed via telephone until stone passage. Complications were defined by i) persistent or recurrent pain at 4 weeks, or ii) development of temperature >38°C, or iii) elevation of creatinine >150 mmol/L by 4 weeks. Appropriate univariate analysis with 95% CIs were performed.
RESULTS: From July to September 1998, a cohort of 70 patients with suspected renal colic were assessed prospectively. Seventy-nine (79%) of the patients were found to have renal calculi and 28.6% of these developed complications. Of those patients with complications the average number of hours of pain prior to presenting to the ED was significantly higher 6.7 hours (±SD 2.2) than those without complications 1.9 hours (±SD 12.1) (p = 0.009). In addition, mean VAS pain scores at discharge from the ED in those patients with complications from their stone were significantly higher than those without complications 19.3 mm (±SD 11.3) vs. 7.3 mm (±SD 23.9) (p = 0.018). Emergency physicians correctly estimated the likelihood of complications in these patients in only 38% of cases.
CONCLUSION: Physicians are unable to predict 4-week complications from calculi at initial ED visit. Duration of pain prior to presentation to the ED and severity of pain at discharge from the ED are significant indicators of complications. If we could derive a clinical decision rule, using tools such as pain scales, to identify those patients who will develop complications from their renal calculi, we could refer patients to a urologist for earlier intervention.
Key words: crenal colic, complications

009 The Revised Canadian CT Head Rule for Patients with Minor Head Injury.
Stiell I, Lesiuk H, Vandemheen K, Wells G, Clement C, De Maio V, Brison R, Cass D, Dreyer J, Eisenhauer M, Greenberg G, MacPhail I, McKnight D, Morrison L, Reardon M, Schull M, Worthington J. Clinical Epidemiology Unit, University of Ottawa, Ottawa.

OBJECTIVES: To revise a recently developed draft decision rule for CT head in patients with minor head injury, as part of the Canadian C-Spine/CT Head (CCC) Study.
METHODS: This prospective cohort study enrolled adults with loss of consciousness, amnesia, or confusion and a GCS score of 13­15. Physicians in 10 Canadian EDs completed a 22-item form prior to CT scan and in some cases 2nd physicians performed interobserver assessments. The outcome standards were "need for neurological intervention" and "clinically important brain injury." Analyses included kappa coefficient, appropriate univariate tests, and chi-square recursive partitioning.
RESULTS: The final study population of 2,536 patients included an additional 817 (47.5%) cases and 80 (60.6%) new cases with brain injury. Characteristics were: mean age 38.6 (range 16­96); male 68.1%; GCS scores: 13- 3.5%, 14- 15.7%, 15- 80.8%; admitted 26.7%; important brain injury 8.4%; neurological intervention 1.5%. A rule with 3 "high-risk" factors has 100% sensitivity (95% CI 91%­100%) and 80.9% specificity (95% CI 47%­51%) for predicting neurological intervention: a) failure to reach GCS of 15 within 2 hours of injury, b) suspected depressed skull fracture, or c) repeated vomiting. Adding 4 "medium-risk" factors gives 97.7% sensitivity (94%­99%) and 49.0% specificity (47%­51%) for predicting important brain injury: d) signs of basilar skull fracture, e) age >55, f) high-risk mechanism, or g) amnesia before injury >30 min. Additional factors could raise sensitivity but with an unacceptable loss in specificity. Potential CT rates for the "high-risk" and "medium-risk" models would be 21.9% and 51.3%.
CONCLUSIONS: This revised Canadian CT Head Rule is a highly sensitive decision rule for use of CT head in minor head injury. Prospective validation of accuracy, reliability, and acceptability is required before clinical use.
Key words: brain injury, computed tomography, clinical prediction rule

010 Evaluating Clinical Clerks in Emergency Medicine: Does the Knowledge Subdomain Predict Examination Marks?
Bandiera GW, Regehr G, Morrison LJ. University of Toronto, Toronto, Ont.

OBJECTIVES: To determine if knowledge marks predict examination marks better than overall clinical marks.
METHODS: The marks from clinical Global Assessment Forms (GAFs) and written examinations for clinical clerks in emergency medicine were collected prospectively for the academic years 1997­98 and 1998­99. Marks for each of ten clinical subdomains as well as the separate overall clinical marks were retrieved from original GAFs. Written examination marks were obtained from the original datasheet for written examinations. Correlations between marks for each subdomain and the written examination were determined, as was the correlation between overall clinical mark and written examination mark. Olkin's Z score was used to check for significance between correlations.
RESULTS: Marks were reviewed for 347 clerks. Nine clerks were excluded (incomplete evaluations); 338 clerks were analyzed. Data entry error was 0.058%. Mean marks out of 5 on each of ten subdomains on the GAF ranged from 3.86 to 4.34 (SD 0.62­0.71). Mean overall clinical mark was 80.11(SD = 4.375) percent. The mean examination mark was 81 (SD = 7.66) percent. The knowledge subdomain had the highest correlation with the examination mark (0.19). The overall clinical marks had lower correlation with the examination marks (0.169). The difference was not significant (Olkin z = 0.40, p = 0.35). The correlation of the average marks of all subdomains excluding knowledge with written examination marks was even lower (0.12). The ten-item alpha for the GAF was 0.92.
CONCLUSIONS: Marks were generally high with low standard deviations. Knowledge marks correlated with examination marks better than overall clinical marks, but the difference was not significant. There is a large amount of redundancy in the data, suggesting a significant halo effect.
Key words:medical education