Return to 2001 Abstacts List

DW Coombs, School of Public Health, University of Alabama at Birmingham; L Davis, VA Medical Centers, Tuscaloosa and Birmingham, AL; L Fish, Private Consultant, Birmingham, AL; SN Barton, VA Medical Center, Minneapolis, MN; E Rand, Dept. of Psychiatry, University of Alabama, Tuscaloosa; M Stephane, VA Medical Center, Minneapolis, MN

Objectives: To report a study of criterion validity for a new instrument, the Suicide Assessment and Management Prtocol for Primary Care (SAM-PC). Our basic objective is to determine if the SAM-PC provided an accurate and valid measure of suicide risk with Beck Scale for Suicide Ideation (BSSI) scores as criterion. The SAM-PC is designed to help primary care providers accurately identify suicidal patients, assess their risk of suicide, and develop a management plan appropriate to the patient's risk level.

Methods: The study was conducted with 21 patients in a psychiatric partial hospital entering the program in May, 2000. Patients in the PPH experience varying degrees of suicide risk and thus are appropriate for validation of this instrument.The SAM-PC consists of:patient questions to assess current and past suicidal ideation/behavior; provider questions to assess comorbidities; an algorithm using pt./provider answers to estimate suicide risk; management guidelines keyed to the patient's risk level.SAM-PC and BSSI questions were adminstered to PPH patients; PPH providers answered SAM-PC provider questions.Investigators used the algorithm to compute patient's suicide risk.Total and subscale BSSI scores were compared with SAM-PC risk scores. Patient answers to conceptually similar items from the two instruments were also

compared.

Results: A significant positive association between BSSI and SAM-PC risk scores was observed(Spearman's p=0.60, one-tailed p<0.01.) Pts designated Immediate Danger by the SAM-PC had high BSSI risk scores and pts. designated Low Risk on the SAM-PC had BSSI scores of zero. Associations between SAM-PC and BSSI scores at intermediate risk levels were not as strong as at the extremes.Answers to conceptually similar items on both instruments were the same except in the case of one pt.

Conclusions: The SAM-PC is a potentially valid instument for assessing suicide risk given the widely-used, validated BSSI as criterion. Due to the small sample ,however, we conclude that further research on larger, more diverse samples is required and justified.

Impact: This study is a small but important step in validating the SAM-PC. Larger studies, for which this serves as a prototype, will help us identify and correct any waeknesses that may be found in the SAM-PC and justify the next step of testing the instrument's feasibility

for widespread use in primary care and other settings.