CAEP/ACMU 2001 Scientific Abstracts - Poster Presentations: 60-74

CAEP Abstracts

CJEM 2001;3(2):135-149

| 001-014 | 015-029 | 030-044 | 045-059 | 060-074 | 075-093 |

060 Comparison of emergency room asthma care to current Canadian asthma guidelines.
Thompson D, McCauley W. University of Western Ontario, London, Ontario.

OBJECTIVES: The emergency department (ED) management of asthma is challenging to the physician owing to the spectrum of illness severity, the variable response to treatment, and the need for the safe disposition of patients. In 1999, the Canadian Asthma Consensus Report issued guidelines for the diagnosis and management of asthma in the ED. The objective of this study is to compare the documented care given in our department with the Level I recommendations for asthma management proposed in these guidelines. METHODS: Two hundred and fifty ED records with the discharge diagnosis of asthma were randomly selected from those charts presenting between Jan. 1, 1999, and Dec. 31, 1999. One hundred and ninety eligible charts were reviewed for documentation of practice pertaining to the Level I recommendations published in the Canadian Asthma Consensus Report (1999). RESULTS: The use of B2-agonists was documented almost universally (98%), but compliance with metered dose inhaler (MDI) delivery was documented in only 6% of cases. Systemic steroids were given to 61% of patients, with 52% being administered within 1 hour of presentation. Steroids were continued after discharge in 64% of eligible patients, with only 20% receiving the recommended 7-14 day course. Nineteen percent of eligible patients were started de novo on inhaled steroid therapy from the ED. CONCLUSIONS: There are significant differences in the ED management and treatment of asthmatic patients when compared to Level I recommendations for asthmatic care in the Canadian Asthma Consensus Report. Notable lack of compliance exists in our use of MDI, rapid administration of systemic steroids, initiation of inhaled steroids from the ED and duration of oral steroid therapy at discharge. Future plans include the prospective collection of data, which will include patient outcomes.
Key words: asthma, therapy

061 Proton pump inhibitors in acute upper gastrointestinal bleeding peptic ulcers: a meta-analysis.
Zed PJ, Loewen PS, Slavik RS, Marra CA. CSU Pharmaceutical Sciences, Vancouver Hospital, University of British Columbia, Vancouver, British Columbia.

OBJECTIVES: To evaluate the efficacy of proton pump inhibitors (PPIs) compared to placebo and histamine receptor antagonists (H2RA) for reducing the incidence of rebleeding, surgery and death in acute gastrointestinal bleeding (GIB) associated with peptic ulcer disease. METHODS: A systematic search of the English-language literature using MEDLINE, EMBASE, and Pre-MEDLINE and a manual search of references. Selected prospective, randomized, controlled trials evaluating any PPI for acute GIB in adults with the endpoints of rebleeding, surgery or death. Data synthesis: 9 trials (1,829 patients) were included. RESULTS: The relative odds of rebleeding indicated a 50% reduction in the PPI-treated group (OR 0.50 [95% CI 0.33-0.77], p = 0.002, NNT = 9, [95% CI NNT 6-13]). The relative odds of surgery indicated a 53% reduction in the PPI-treated group (OR 0.47 [95% CI, 0.29-0.77], p = 0.003, NNT = 17 [95% CI 12-35]). The relative odds for mortality indicated a non-significant 8% decrease in the odds of death in the PPI-treated group (OR 0.92 [95% CI, 0.46-1.83], p = 0.26). CONCLUSIONS: PPIs are superior to H2-blockers and placebo in preventing rebleeding and the need for surgery in patients with GIB although do not appear to reduce mortality.
Key words: proton pump inhibitors, peptic ulcer, gastrointestinal bleeding

062 Efficacy of once daily cefazolin/probenecid for the outpatient management of skin and soft tissue infections.
Zed PJ, Harder C, Harrison DW, Purssell RA. CSU Pharmaceutical Sciences, Vancouver Hospital, University of British Columbia, Vancouver, British Columbia

OBJECTIVES: To evaluate the efficacy of once daily cefazolin/probenecid for an emergency department (ED) based outpatient treatment program for the management of skin and soft tissue infections (SSTIs). METHODS: A retrospective chart review of all patients treated with cefazolin 2 g IV daily in combination with probenecid 1 g PO daily for treatment of SSTI over a 1-year period in the ED--based outpatient program were evaluated. The primary outcome was clinical treatment success defined as discharge from the outpatient program on either no therapy or on oral antibiotics alone following improvements or resolution of the infection. Patients who required hospitalization, failed to return to the program for follow-up or were enrolled in the home IV antibiotic program for longer treatment duration were deemed to have failed on the cefazolin/probenecid regimen in the outpatient program. RESULTS: 346 patients received cefazolin/probenecid for SSTI over a 1-year period for a mean treatment duration of 3.5 days (standard deviation [SD] = 1.9 days). Mean age of patients was 44.8 years (SD = 17.5 years) and 67% were male. Overall, treatment success was achieved in 304 (87.9%) of patients who were discharged from the outpatient program. Treatment failures occurred in 42 (12%) patients. Twenty-eight (8.1%) patients failed to return for follow-up to the outpatient program and 14 (4.0%) required hospitalization for failure to improve in the outpatient program. Overall, 1225 hospital days of admission were avoided by treating these patients in the outpatient program. CONCLUSIONS: Once daily cefazolin/probenecid is an effective regimen for the outpatient treatment of SSTI in the ED. These data suggest that many patients with uncomplicated SSTI can be treated as outpatients, avoiding a more costly hospitalization.
Key words: cellulitis, cefazolin

063 Prevalence and prognosis of traumatic intraventricular hemorrhage in blunt head injury patients.
Atzema C, Mower WR. UCLA, Los Angeles, California.

OBJECTIVES: Traumatic intraventricular hemorrhage (tIVH) has received limited attention in the medical literature, with the largest study containing only 43 cases. Knowledge of the prevalence and prognosis of tIVH could improve our ability to treat affected patients, potentially decreasing the morbidity and mortality of tIVH. METHODS: As part of the NEXUS II study we prospectively enrolled all blunt trauma patients undergoing computed tomography (CT) at 8 trauma centres. Age, sex, presence of coagulopathy, and CT diagnosis were recorded for each patient. Long-term follow-up was obtained for patients from 3 sites. The data was analyzed for prevalence of tIVH, associated intracranial injuries, coagulopathies, hydrocephalus, and outcomes. RESULTS: Of 492 patients having CT findings, 70 (14%) exhibited tIVH, including 13 (19%) who exhibited isolated tIVH. Patients ranged in age from 1 to 95 years, and 70% were male. Of 21 cases with long-term follow-up, 9 (43%) were discharged to home with minimal or no disability, 7 (33%)went to long-term rehabilitation facilities, and 5 (24%) died. Six of the 9 patients discharged to home had only minor additional injuries on CT. Four of the 7 patients sent to rehabilitation facilities and 1 of the 5 patients who died exhibited only minor associated injuries on CT (relative risk [RR] 0.65; 95% confidence interval [CI] 0.30-1.39). Three of the 21 patients had isolated tIVH, and 1 required long-term rehabilitation (RR 0.55; CI 0.11-2.82). Three patients had coagulopathies, including 2 with tIVH as their only finding (RR 4.06; CI 1.55-10.66). Six patients developed hydrocephalus (9%), 3 of whom had only minor associated CT findings (RR 2.18; CI 0.48-9.96). CONCLUSIONS: tIVH is found in 14% of blunt head injury patients with CT abnormalities. Isolated tIVH is common and is associated with a better prognosis. Coagulopathy may be an antecedent factor in many cases of isolated tIVH.
Key words: brain injury, hemorrhage

064 What if the Canadian CT Head Rule derivation had been restricted to cases without the proxy outcome?
Stiell IG, Clement C, Cass D, Schull M, Morrison L, Wells GA, et al, for the CCC Study Group. University of Ottawa, Ottawa, Ontario.

OBJECTIVES: The Canadian CT Head Rule has been criticized because the original derivation determined outcomes with the proxy outcome telephone measure, rather than computed tomography (CT), for 1/3 of cases. This methodological sub-study repeated the derivation analyis using only the sub-set of minor head injury patients who underwent CT. METHODS: This secondary data analysis was based upon a prospective cohort study that enrolled adults with loss of consciousness, amnesia, or confusion and a Glasgow Coma Scale (GCS) score of 13-15 at 10 Canadian emergency departments (EDs). Data collection included a 22-item physician data form, CT reports, ambulance reports, and in-hospital records. The outcome measures were clinically important brain injury and need for neurological intervention. For this study, chi-square recursive partitioning analyses (KnowledgeSEEKER) assessed the sub-set of patients who underwent CT. Sensitivity, specificity, and CT rates were estimated. RESULTS: Comparing patients who underwent CT (n = 2,078) to the original CCC Study patient group (n = 3,121), respectively: initial GCS score 13%-5.2% vs. 3.5% (p < 0.01), initial GCS score 14%- 3.6% vs. 16.7% (p < 0.0001), arrived by ambulance 79.5% vs. 72.7% (p < 0.0001), transferred 17.2% vs. 12.8% (p < 0.0001), admitted 36.4% vs. 27.0% (p < 0.0001), any acute brain injury on CT 16.6% vs. 11.2% (p < 0.0001), important brain injury 12.1% vs. 8.1% (p < 0.0001), required neurological intervention 2.1% vs. 1.4% (p = 0.05). In the CT only sub-set, the 5 "high-risk" factors from the CT Head Rule predicted need for neurological intervention with sensitivity 100% (95% confidence interval [CI] 92%-100%), specificity 57.5% (55%-60%), and required CT rate 43.7%. The additional 2 "medium-risk" factors predicted important brain injury with sensitivity 98.4% (95% CI 96%-99%), specificity 36.3% (34%-38%), and required CT rate 67.9%. CONCLUSIONS: The patients in the CT only sub-set had more serious characteristics than those in the original study population. The Canadian CT Head Rule performed equally as well in this group as in the original derivation study.
Key words: brain injury, computed tomography, clinical prediction rule

065 How valid is the concept of "clinically unimportant" brain injury in patients with minor head injury?
Stiell IG, Lesiuk H, Brison RJ, Clement C, De Maio VJ, Wells
GA, et al, for the CCC Study Group. University of Ottawa, Ottawa, Ontario.

OBJECTIVES: The Canadian CT Head Rule was developed to help physicians predict which minor head injury patients have important brain injury (IBI) on computed tomography (CT). This sub-study evaluated the clinical validity of the concept "clinically unimportant brain injury" (CUBI), as previously endorsed by academic neurosurgeons. METHODS: The prospective cohort study enrolled adults with loss of consciousness, amnesia, or confusion and a Glasgow Coma Scale (GCS) score of 13-15 at 10 Canadian EDs. Data collection included MD dataforms, CT reviews by study neuroradiologists, in-hospital records, and 14-day telephone follow-up. Patients were considered to have CUBI, and therefore require neither admission nor specialized follow-up, if neurologically intact with 1 of these CT lesions: solitary contusion <5 mm in diameter, localized subarachnoid blood <1 mm thick, smear subdural hematoma <4 mm thick, or closed depressed skull fracture not through inner table. This study compared the characteristics of IBI and CUBI patients by chi-square and Student's t-test analyses. RESULTS: Among the 3,121 patients, there were 254 (8.1%) IBI cases and 94 (3.0%) CUBI cases:

CONCLUSIONS: Patients with CUBI had fewer CT lesions, admissions, and follow-up problems. No patient with CUBI required neurological intervention or died from head injury. This study confirms the validity of the concept of CUBI for minor head injury patients.
Key words: brain injury, computed tomography, clinical prediction rule

066 Comparison of recursive partitioning and logistic regression modelling in the derivation of the Canadian CT Head Rule.
Stiell IG, Wells GA, De Maio VJ, Clement C, Brison RJ, Cass D, et al, for the CCC Study Group. University of Ottawa, Ottawa, Ontario.

OBJECTIVES: The Canadian CT Head Rule for use of CT in minor head injury was derived by recursive partitioning (RP) analysis. This study compared the accuracy of the CT Head Rule to a model developed by an alternate statistical technique, logistic regression (LR). METHODS: This secondary data analysis was based on a prospective cohort study conducted in 10 Canadian emergency departments (EDs) and involved adults with loss of consciousness, amnesia, or confusion and a Glasgow Coma Scale (GCS) score of 13-15. Physicians completed a 22-item data form for all patients who then underwent CT scan. The outcome measures were need for neurological intervention and important brain injury. Variables correlated with these outcomes on univariate analysis and having kappa values >0.6 were then assessed by 2 multivariate statistical techniques. Chi-square RP analysis (KnowledgeSEEKER) was used for the original CT Head Rule derivation and forward stepwise LR (SAS) was used for this analytic study. RESULTS: The CT Head (CCC) Study dataset contained 3,121 minor head injury cases, including 44 (1.4%) requiring neurological intervention and 254 (8.1%) with important brain injury. 27 variables demonstrated univariate p-values <0.05 and 10 had kappa values >0.60. The RP model lost no cases to missing values and contained 7 variables. The LR model lost 80 cases to missing values and contained 8 variables, including the new "drop in GCS score." Comparing the RP to the LR models for identifying important brain injury: area under receiver operating characteristic (ROC) curve 0.893 vs. 0.894 (p = NS), sensitivity 98.4% vs. 98.4% (p = NS), specificity 48.4% vs. 48.2% (p = NS), and required CT rate 55.4% vs. 55.4%. Both models performed with 100% sensitivity for identifying need for neurological intervention. CONCLUSIONS: Both statistical approaches lead to models with very similar measures of performance. As in previous studies, RP provides a more parsimonious model that is likely to be more acceptable to clinicians.
Key words: brain injury, computed tomography, clinical prediction rule

067 Interobserver agreement in the assessment of patients with minor head injury.
Stiell IG, Wells GA, Clement C, Brison R, McKnight RD, Worthington JR, et al, for the CCC Study Group. University of Ottawa, Ottawa, Ontario.

OBJECTIVES: To determine interobserver agreement in the MD assessment of clinical findings in minor head injury patients. This methodological sub-study was an important component in the derivation study for the Canadian CT Head Rule. METHODS: This prospective cohort study was conducted in 10 Canadian EDs and involved adults with loss of consciousness, amnesia, or confusion and a Glasgow Coma Scale (GCS) score of 13-15. MDs evaluated patients for 22 standardized clinical findings before imaging and performed blinded interobserver assessments when feasible. Analyses included the simple or weighted kappa coefficient with 95% confidence intervals (CIs). RESULTS: 202 assessments were conducted on 101 patients who were similar to those of the main study population for: mean age (37.7; range 16-90), male (71.3%), falls (36.6%), motor vehicle collision (MVC) (24.8%), ambulance arrival (72.3%); they had a higher rate of admission (42.6%) and clinically important brain injury (14.9%). Kappa values for the clinical findings were:

CONCLUSIONS: Findings with only moderate agreement were "any amnesia" and "unreliable exam due to ETOH." All components of the Canadian CT Head Rule showed excellent interobserver agreement suggesting that physicians should be able to consistently interpret the overall rule. This reliability will be explicitly and prospectively evaluated in ongoing studies.
Key words: brain injury, computed tomography, clinical prediction rule

068 The Nova Scotia Cellulitis Guidelines: a pilot study.
Campbell SG, Burton-MacLeod R, Pierce S, Ackroyd S, Gerami D. Dalhousie University, Halifax, Nova Scotia.

OBJECTIVES: To assess the safety of a 4-level grading system and algorithm to treat adults with cellulitis and to monitor physician compliance and approval. METHODS: A prospective study observing physician compliance with the algorithm and patient outcome for an 8-week period after introduction of the algorithm was performed. Consenting patients were contacted in 7-10 days to assess outcome. Physician compliance was graded according to the number of elements of the algorithm followed. RESULTS: A total of 146 patients visited the ED during the study period (182 visits). The mean age was 47 years and 63.7% were female. From the visits made, 57.1% were Grade 1, 16.5% Grade 2, 12.9% Grade 3, and 0.6% Grade 4 infections. 12.9% had insufficient information to establish a grade. 51.3% of patients received oral antibiotics, 20.9% received intravenous antibiotics, and 27.8% received both. Of 57 consenting patients, 54 were followed-up. From these patients contacted 90.7% reported improvement in their condition, 5.6% no change, and 3.7% worse. Following the physician questionnaire, 84.6% had relied on the algorithm during the study period. 85.7% felt the algorithm should be permanently implemented in the ED. The chart review showed compliance with the algorithm in 25.4% of evaluable visits. Of compliant visits, 81.8% of the patients followed-up indicated their condition was improved. CONCLUSIONS: This pilot study suggests that the use of a 4-level grading system and algorithm for the ED management of cellulitis is safe for patients, and its use is acceptable to physicians. Further studies are planned.
Key words: cellulitis

069 Vehicle effects on efficacy of salicylate adsorption by activated charcoal in a gastrointestinal tract model.
Dagnone D, Matsui D, Rieder MJ, Freeman DJ. Children's Hospital of Western Ontario, London, Ontario.

OBJECTIVES: Activated charcoal (AC) is the antidote of choice for therapy of ingestions in children. Charcoal is often given with vehicles to improve palatability. Our hypothesis was that AC can be mixed with flavouring vehicles without altering its ability to adsorb toxic levels of acetylsalicylic acid (ASA) in vitro using a gastrointestinal model. METHODS: The 4 vehicles chosen were water, chocolate milk, orange juice and a cola beverage. ASA was added to 5 ml of human gastric fluid from each of 10 children aged 7 to 15 years. The dose used was calculated from a 10-kg child at a toxic dose of 10 mg/kg. Activated charcoal was mixed with an equal volume of vehicle to achieve a concentration of 1 gm/kg. Incubation was continued for a further 5 hours and aliquots were taken. The pH of the solution was initially adjusted to 3.0 and then increased at 180 minutes to 7.0 to mimic the transit of gastric fluid to the intestine. Concentrations of ASA and total salicylate were determined for the 4 vehicles studied using an HPLC assay system. RESULTS: All 4 vehicles lowered the concentration of salicylates from toxic concentrations (3000 ug/ml) to significantly lower concentrations (18 ± 4% of original concentration, p > 0.05) within 30 minutes. There was no significant difference in effect between the 4 vehicles. After the increase in pH at 180 minutes, there was a sharp and significant increase in salicylate concentrations for all 4 vehicles (64 ± 5% of original concentration, p > 0.01, at time 300 min). This increase was most marked for chocolate milk (71 ± 3%, p > 0.05). CONCLUSIONS: The results of this study suggest that activated charcoal may be less useful in the reduction of toxic concentrations than is currently believed, and that vehicles may have more effects than previously appreciated.
Key words: intoxication, salicylate, activated charcoal

070 Accuracy of prehospital assessment of acute pulmonary edema.
Lett DA, Petrie D, Ackroyd S. Queen's University, Kingston, Ontario.

OBJECTIVES: Acute pulmonary edema (APE) due to congestive heart failure (CHF) is a common and serious reason for emergency medical services (EMS) activation. However, only appropriate interventions will reduce morbidity, mortality and patient discomfort. The purpose of this study was to determine the accuracy of paramedics in determining whether APE was present in the prehospital setting. In addition, we sought to describe and compare patients that used EMS versus those that did not, when presenting with chest pain or respiratory distress. METHODS: The study involved a 5-week prospective review of ED patient records and paramedic records for all 806 patients presenting to the Queen Elizabeth II Health Sciences Centre with chest pain or respiratory distress. A predetermined data abstracting form was used. Data collected for all patients included age, gender, chief complaint, emergency department (ED) diagnosis, and disposition. Several additional variables were collected for patients who came by ambulance with an ED diagnosis or paramedic impression of CHF. Paramedics were not aware that this study was taking place. Data was analyzed using t-tests, X², and Fisher's exact tests. Significance was achieved if p < 0.05. RESULTS: There were 48 cases of ED confirmed CHF, of which 25 came by ambulance. Of the ambulance transports, the paramedics recorded an impression of CHF in 14 cases; 6 of these were later confirmed by ED physicians, 6 were clearly incorrect, and 2 were indeterminant due to a vague recorded impression. The impression of CHF by a paramedic had a specificity of 0.95, a sensitivity of 0.24 and an accuracy of 86%. All of the patients with correct paramedic impressions of CHF were admitted to hospital, versus 66% of those with the incorrect impression. CONCLUSIONS: These findings show that the accuracy of paramedics in recognizing APE could, and should, be improved prior to a randomized controlled evaluation of the effectiveness of prehospital interventions for APE.
Key words: congestive heart failure, pulmonary edema

071 Towards a "No Pain Zone": Does a pain awareness program in triage targeting pediatric musculoskeletal injuries increase pain assessment and intervention?
Duggan L, Vukov I, Littlejohn A, Chu L, Wood C, Bruce E, et al. Hospital for Sick Children, University of Toronto, Toronto, Ontario.

OBJECTIVES: Patients with musculoskeletal injuries may wait for hours to be assessed by a physician, therefore early pain assessment and intervention are essential. The "No Pain Zone" program was created to encourage pain assessment during triage. This study was conducted to compare pain assessment and intervention rates before and after program implementation. METHODS: A prospective, interventional, non-randomized cohort chart review design was used. All patients presenting with musculoskeletal injuries to our tertiary-care pediatric emergency department from July 1 to August 31, 2000, were included. Data was collected for 2 weeks before and 4 weeks after initiating the program. There was a 2-week suspension of data collection midway, during the introduction of the program. It consisted of a general pain awareness campaign and intensive one-on-one teaching of age-appropriate pain scales. Incorporation of pain scales into triage assessment was encouraged. Appropriate interventions for pain were left to the discretion of the triage nurse. Primary outcomes were frequency of pain assessment, either using pain scales (formal) or subjective clinical assessment (informal), and frequency of interventions. These variables were compared before versus after the program using the 2-tailed chi-square test. RESULTS: There were 567 patients in total, with 233 presenting before, and 334 presenting after the program. Pain assessment, including formal and informal methods, increased from 43.8% to 56.0% (p = 0.005). Total interventions increased from 17.6% to 26.9% (p = 0.011). Patients assessed by formal pain scale were more than twice as likely to receive an intervention as compared to those assessed informally (48.4% vs. 21.5%, p < 0.0005). CONCLUSIONS: The "No Pain Zone" program was associated with an increase in pain assessment and intervention. The use of formal pain assessment was associated with a considerable increase in interventions. This provides support for the addition of formal pain assessment in triage to reduce pain in the emergency department.
Key words: analgesia, pain, pediatric

072 Canadian Emergency Department Triage and Acuity Scale triage: reliability for high acuity patients.
Grafstein E, Innes G, Christenson J, Clarke L. St. Paul's Hospital, Vancouver, British Columbia.

OBJECTIVES: The Canadian Emergency Department Triage and Acuity Scale (CTAS) is a national triage standard. CTAS acuity levels correlate with patient morbidity and mortality, nursing workload, admission rate, hospital length-of-stay and diagnostic test use. In future, the CTAS will be used to compare acuity and utilization across emergency departments (EDs), but coding variability may limit the validity of such comparisons. Our objective was to determine nurse-assigned acuity levels in a cohort of ED patients with complaints defined in the CTAS as Level 1. METHODS: A 1-year retrospective cohort study was performed at St. Paul's Hospital, an urban Vancouver teaching centre. Our triage nurses code patients by selecting the most appropriate presenting complaint from an electronic menu, which then suggests a CTAS-based acuity level (1-5). Nurses may override the suggested level and enter their own, based on subjective assessment. We recorded nurse-assigned acuity levels in consecutive patients with the triage complaints below, specifically defined as Level 1 within the CTAS. RESULTS: Of 50,406 patients seen, 350 had CTAS-defined Level 1 complaints. Of these, 101 (29%) were coded as Level 1, 164 as Level 2 (47%), 75 as Level 3 (21%), and 10 (3%) as Level 4 or 5. Of 50,056 patients with CTAS Levels 2-5 complaints, 30 patients were coded as Level 1.

CONCLUSIONS: Subjective nursing assessments differ substantially from CTAS-based acuity recommendations. Nurses often down-triage high acuity patients. A standardized computer algorithm may improve triage reliability and facilitate cross-site case-mix comparisons.
Key words: triage, acuity, case mix

073 Effect of a pneumococcal vaccine program on emergency department presentations.
Grafstein E, Daly P, Buxton J, Thorne A. St. Paul's Hospital and Vancouver Regional Health Board, Vancouver, British Columbia.

OBJECTIVES: Vancouver's Downtown East Side (DTES) is an area with many injection drug users, alcoholics, HIV and poor overall population health. St. Paul's Hospital, an urban, teaching centre, is the primary facility that serves the DTES. In November 1999, the Vancouver regional health board instituted a pneumococcal vaccine program and vaccinated 10,000 DTES residents, only 70% of whom had a permanent address. Our objective was to assess the impact of the pneumococcal vaccine program on the volume of pneumonia seen in our emergency department (ED). METHODS: A retrospective case control study was undertaken comparing the incidence of pneumonia 1 year before (Dec. 1998-Nov. 1999) and 1 year after (Dec. 1999-Nov. 2000) the vaccine campaign. For any patient, only the initial visit was counted as a case. Subsequent visits by the same patient >1 month after initial diagnosis of pneumonia were counted as a new case. Patient records were reviewed using an electronic ED database and linked to the regional immunization database. RESULTS: In the prevaccine period, 863 of 51,825 ED visits (1.67%) were pneumonia cases. In the postvaccine period, 646 of 49,981 ED visits (1.29%) were pneumonia cases (p < 0.001). There were fewer cases of pneumonia noted in almost every month and hospital admissions through the ED decreased by 25%. During this time there were no additional community clinics established and ED volumes at other nearby hospital EDs did not increase. In comparison, asthma incidence was 537 in 1999 and 524 in 2000 with a p = 0.87 and 95% confidence interval (-0.0014, 0.0010). The percentage of all patients with pneumonia living in the DTES before and after the vaccine program was 25%. There were 41 patients who received pneumococcal vaccine and developed pneumonia. CONCLUSIONS: A pneumococcal vaccination program may have contributed to a decrease in pneumonia seen in our high risk population.
Key words: pneumonia, pneumococcal vaccine

074 Clinical practice guideline (CPG) accessibility and interactivity using an electronic platform.
Meurer DP, Bullard M, Holroyd BR, Rowe BH. University of Alberta, Edmonton, Alberta.

OBJECTIVES: Clinical practice guidelines (CPGs), which are designed to apply results from related research to clinical care, are becoming important tools in reducing emergency medicine diagnostic and treatment practice variation. However, this environment demands that information be rapidly accessible, valid and evidence-based. Information systems that easily access user friendly CPGs, can facilitate this process. METHODS: R/O deep vein thrombosis (DVT) is a common emergency department (ED) presentation where wide diagnostic variability and changing treatments exist. Prospective assessment of R/O DVT patients was completed in the ED over an 18-month period (PRE). ED physicians were surveyed about the utility of a Swollen Limb (SL) CPG. Adobe PDF forms for assessing an SL were developed in 2000 and implemented in 2001 (POST). This form complimented an existing protocol for the treatment of DVT. The assessment is accessed on the desktop PC via an intranet Web site and completed on-line in PDF. RESULTS: Prior to the implementation, paper versions of the SL using the Wells' criteria were completed at rates of 60% (220/334) but significant practice variation in diagnostic testing was observed. Consensus from 20 staff was obtained for the SL form, and they requested online interactive fields and radio buttons. Summary scores and an MD estimate of pretest probability for DVT were requested. Some selections allowed for multiple functions to occur with a single click of the mouse (i.e., providing access to the appropriate order set, and providing radiographic recommendations). PRE diagnostic test utilization will be compared to POST. CONCLUSIONS: Recognizing the need for accessible and useful evidence-based tools for the practitioner is the force behind the development of effective electronic work platforms. While these applications have the potential to create efficiencies, formal iterative evaluation is required to document compliance and assess outcomes.
Key words: guideline, deep vein thrombosis

| 001-014 | 015-029 | 030-044 | 045-059 | 060-074 | 075-093 |