Testing the use of symptom-based terrorism triage algorithms with hospital-based providers.

INTRODUCTION: A set of symptom-based, all-hazards, decision-making algorithms was designed to aid the first-contact provider during early patient presentations after a terrorist incident. OBJECTIVE: The primary objective was to assess the usability of these algorithms. A secondary objective was to assess the psychometric properties of the testing scenarios. METHODS: This was a written, usability assessment of the algorithms employing a convenience sample of hospital-based, healthcare providers who had not taken any specific training in the use of the algorithms. A series of 26 paragraph-length, moderately difficult scenarios was created to reflect possible agents, means of attack, and types of patients. Each of the 26 scenarios requires that one make a triage choice on the "attack" algorithm (the trunk algorithm), then proceed to one of four other branch algorithms (dirty resuscitation, chemical agents, biological agents, bomb/blast/radiation dispersal device) to make a final triage choice. Conditional scores based on getting both the attack and final card correct were calculated for each algorithm. RESULTS: Nineteen attending physicians, 50 emergency medicine residents, and 41 nurses took the assessment. The total score was 45% correct for all participants. The score on the attack algorithm was 66% correct. Dirty resuscitation, biological, chemical, and bomb/blast scores were 46%, 54%, 46%, and 51% respectively. The probability of guessing the correct answer on the attack algorithm was 1/7 or 14%. The conditional probability of guessing both the attack algorithm and the final card correct ranged from 4.7% for the biological, chemical, and bomb/blast algorithms to 2.4% for the dirty resuscitation algorithm. Item discrimination, item difficulty, and Cronbach's alpha were acceptable for the overall test. Certain individual items had item difficulty levels suggesting they were too difficult and should be replaced in future versions of the test. CONCLUSIONS: Performance on the test suggests that participants did substantially better than would have been expected by chance alone. Future efforts will revise the algorithms with the goal of simplification. Revision of the testing instrument and testing algorithm use after instruction also are needed.

Nonpunitive medication error reporting: 3-year findings from one hospital's Primum Non Nocere initiative.

OBJECTIVE: To identify underlying practices and attitudes on medication error occurrences and reporting practices. BACKGROUND: In response to a hospital-wide quality improvement initiative, a task force was formed to facilitate a nonpunitive culture toward reporting medication errors. To identify underlying practices and attitudes on medication errors and medication error reporting, a baseline survey was conducted. Based on findings, an initiative that included modifications to clinical and administrative processes was developed and implemented. METHODS: A pre/post initiative questionnaire to measure staff practices and attitudes on medication error reporting was developed and administered. Findings from the presurvey were used to craft the Nonpunitive Patient Safety Policy and its implementation plan. Pre-post comparative analysis was performed following a baseline-postimplementation design. RESULTS: Conceptually, a medication error is qualified by its outcome severity. Medication errors with more serious outcomes are more likely to be reported than those with less serious ones. Staff perception that medication error reporting carries the risks of disciplinary action was identified as a primary barrier to the likelihood of reporting. CONCLUSION: Evaluation of the initiative suggests that a multicomponent approach facilitates positive movement in the direction of a nonpunitive culture toward reporting medication errors.

Impact of immediate reconstruction on the local recurrence of breast cancer after mastectomy.

The incidence of local recurrence of breast cancer in women who underwent mastectomy with or without reconstruction was examined. All female mastectomy patients were followed-up in a 10-year retrospective review. Groups consisted of patients who had mastectomy, mastectomy with immediate reconstruction, or delayed reconstruction. Reconstruction was performed using prostheses, latissimus dorsi musculocutaneous flaps with or without implants, or transverse rectus abdominis musculocutaneous flaps. Charts were reviewed for local breast cancer recurrence. Statistical analysis was performed using Pearson's chi-square and analysis of variance. Of the 1,444 mastectomies performed from 1988 to 1997, 1,262 breasts (87%) were not reconstructed, 182 (13%) were reconstructed, 158 (87%) were immediately reconstructed, and 24 (13%) were reconstructed later. There were no recurrences in the delayed reconstruction group, two recurrences (1.3%) in the immediate reconstruction group, and nine recurrences (0.7%) in the mastectomy without reconstruction group (p=0.746). Analyses of an additional time period from 1992 to 2000 yielded similar results. There is little relationship between local recurrence of breast cancer after mastectomy and reconstruction.