Suicide Note Classification Using Natural Language Processing: A Content Analysis.

Suicide is the second leading cause of death among 25-34 year olds and the third leading cause of death among 15-25 year olds in the United States. In the Emergency Department, where suicidal patients often present, estimating the risk of repeated attempts is generally left to clinical judgment. This paper presents our second attempt to determine the role of computational algorithms in understanding a suicidal patient's thoughts, as represented by suicide notes. We focus on developing methods of natural language processing that distinguish between genuine and elicited suicide notes. We hypothesize that machine learning algorithms can categorize suicide notes as well as mental health professionals and psychiatric physician trainees do. The data used are comprised of suicide notes from 33 suicide completers and matched to 33 elicited notes from healthy control group members. Eleven mental health professionals and 31 psychiatric trainees were asked to decide if a note was genuine or elicited. Their decisions were compared to nine different machine-learning algorithms. The results indicate that trainees accurately classified notes 49% of the time, mental health professionals accurately classified notes 63% of the time, and the best machine learning algorithm accurately classified the notes 78% of the time. This is an important step in developing an evidence-based predictor of repeated suicide attempts because it shows that natural language processing can aid in distinguishing between classes of suicidal notes.

Developing the careers of clinician-educators in psychiatry.

OBJECTIVE: This article describes initial efforts by the newly developed Clinician-Educator Section (CES) of the Association for Academic Psychiatry (AAP) to support the career advancement of academic faculty within psychiatry. The CES provided its first workshop at the 2005 AAP annual meeting, focusing on early career development and academic promotion. METHOD: The CES obtained input from members regarding their vision for how this newly developed section could support their career development. Workshops were among the suggestions provided and therefore this article will describe the planning, implementation, and outcome measures of the initial workshop provided by the CES. RESULTS: The authors describe a summary of the workshop, descriptive statistics obtained from a needs assessment, and formal evaluation ratings of the workshop. Results indicate that this initial effort was a helpful professional development opportunity for academic psychiatrists. CONCLUSIONS: The AAP-led initiative described in this article is a positive first step in the field of academic psychiatry to leverage the valuable educational and facilitative role of a professional organization in promoting the career advancement of clinician-educators in psychiatry.

Utilization of a formative evaluation card in a psychiatry clerkship.

OBJECTIVE: This article discusses how formative feedback to medical students during their clinical rotations facilitates their successfully meeting the rotation's educational objectives. To help students initiate mid-rotation feedback and to help preceptors structure that feedback, the authors designed the Instant Feedback Card (IFc). The goal of this study was to examine the degree to which the IFc facilitated the formative feedback process from the perspectives of both students and faculty. METHOD: The IFc contains a checklist of the 13 clinical competencies that are also used to provide summative evaluations to students. One hundred and sixty-five medical students completing a 6-week psychiatry clerkship between January and December 2004 were given an IFc to present to their inpatient preceptor at one of their two inpatient sites. At the end of each 3-week inpatient rotation, all students were e-mailed a brief questionnaire about the formative feedback received during the preceding rotation. At the end of the 12-month study period, faculty were e-mailed a brief questionnaire about how useful the IFc was in helping to facilitate the feedback process. RESULTS: One hundred thirty-eight students (80%) participated in the study and returned 267 questionnaires (1.9/student, 97% response rate). Two hundred five students indicated on 77% of the questionnaires returned that they received mid-rotation feedback and of those, 84% revealed that feedback was helpful. Eighty-five percent of the students received mid-rotation feedback when they used the IFc and only 69% received mid-rotation feedback without the IFc. A majority of the 14 faculty serving as inpatient preceptors during the study period found the cards useful for stimulating feedback discussions, for reducing the stress of providing feedback, and for getting students to request feedback. CONCLUSIONS: The results of this pilot study are not definitive, but they do indicate that the Instant Feedback card (IFc) did facilitate the process and structure of providing and receiving formative feedback as indicated by a significantly greater number of students receiving formative feedback when they used the IFc.

Use of standardized patients during a psychiatry clerkship.

OBJECTIVE: Standardized patients are used in teaching medical students and evaluating their clinical skills during the psychiatric clerkship. The purpose of this study was to evaluate whether the addition of a Psychiatry Clinical Standardized Patient Examination (PCX) during the third-year clerkship improved students' performances on the psychiatry component of the Clinical Competency Examination (CCX) that is used at our institution to prepare fourth-year students for the National Board of Medical Examiners Step 2 Clinical Skills Exam (Step 2 CSA). METHOD: Scores on the CCX were analyzed by two-by-two tables (yes/no response to each clinical competency item versus took/did not take PCX) using Fisher's exact test. RESULTS: Of the 469 students who took the CCX between 2002 and 2004, 246 students also took the PCX. Compared to students who did not take the exam, students who took the PCX were significantly more likely (p<0.05) to receive positive ratings on the CCX items for professionalism, addressing patients' stressors at work and home and inquiring about thoughts of harm to self or others. CONCLUSIONS: The implementation of the PCX during the psychiatry clerkship significantly improved student performance on several important components of the CCX, which is used to prepare students for the Step 2 CSA. This suggests that the use of standardized patients during the clerkship may help students improve their interviewing and interpersonal skills.

Professionalism deficits among medical students: models of identification and intervention.

OBJECTIVE: This study compares the instruments and interventions utilized to identify and remediate unprofessional behaviors in medical students across U.S. psychiatry clerkships. METHODS: A 20-item questionnaire was distributed to 120 psychiatry clerkship directors and directors of medical student education, in the U.S., inquiring into the procedures utilized for identifying, monitoring, and remediating unprofessional behaviors among medical students during their psychiatry clerkship. RESULTS: Fifty seven (47.5%) clerkship directors responded to the questionnaire. Professionalism is evaluated by 96% of the clerkships with the most frequent goals being to provide feedback to students and to specifically identify problematic behaviors. Seventy percent of the clerkships identify one to three students per year with unprofessional behaviors that warrant intervention. The majority (86%) of the respondents note that unprofessional behaviors may impede the advancement of students during their medical education. A recommendation for a mental health evaluation occurs in at least 76% of cases. Dismissal from medical school for unprofessional behaviors is most likely recommended based on the severity and/or repetitiveness of the acts. CONCLUSIONS: There is concordance among clerkship directors regarding the importance of identifying unprofessional behaviors among medical students although there exists a range of modalities for monitoring, remediating, and disciplining such behaviors.

Use of a computerized evaluation system in a psychiatry clerkship.

OBJECTIVE: Paper-based evaluations of medical students on clerkships are inefficient and time consuming to compile and analyze. To improve the evaluation process, the authors developed a web-based program. METHODS: A secure web-based system was designed to collect evaluations of students and their feedback about the clerkship. Acceptance, compliance, and impact of the web-based system on the clerkship were assessed. RESULTS: The web-based program was universally accepted by medical students who have been 100% compliant with prompt completion of their evaluations of the rotation. Most of the preceptors have utilized the program, resulting in a more timely receipt of their evaluations of students. The web-based program computes numerical scores, compiles comments, and formats reports. The clerkship director has unlimited access to all of the data, which allows for efficient and timely communication with preceptors, medical students, and the dean's office. CONCLUSIONS: During the clerkship, the web-based program for evaluation and feedback has improved the efficiency of the evaluation process and the timely communication with students and preceptors.

A placebo-controlled, randomized trial of fluoxetine in the treatment of binge-eating disorder.

BACKGROUND: The purpose of this study was to assess the efficacy and safety of fluoxetine in the treatment of binge-eating disorder. METHOD: Sixty outpatients with a DSM-IV diagnosis of binge-eating disorder were randomly assigned to receive either fluoxetine, 20 to 80 mg/day (N = 30), or placebo (N = 30) in a 6-week, double-blind, flexible-dose study. The primary measure of efficacy was frequency of binge eating. Secondary measures included body mass index, weight, Clinical Global Impressions-Severity of Illness score, Hamilton Rating Scale for Depression (HAM-D) score, and response categories. The outcome measures were analyzed using 2 random regression methods, a time trend analysis (primary analysis) and an endpoint analysis. In addition, response categories were analyzed using an exact trend test. RESULTS: Compared with placebo-treated subjects, subjects receiving fluoxetine (mean +/- SD endpoint dose = 71.3 +/- 11.4 mg/day) had a significantly greater reduction in frequency of binge eating (p =.033), body mass index (p <.0001), weight (p =.001), and severity of illness (p =.032) and a marginally significant reduction in HAM-D scores (p =.061). Differences between groups on response categories were not statistically significant. CONCLUSION: In a 6-week, placebo-controlled, flexible-dose trial, fluoxetine was efficacious in reducing binge-eating frequency, weight, and severity of illness and was generally well tolerated in subjects with binge-eating disorder.