Association between clinical observations and a mobile crisis team's level of care recommendations

Mobile mental health crisis programs are a widely used and valuable community resource. Literature analyzing the service, however, is sparse and descriptive in nature. This study uses multinomial logistic regression to analyze clinical data from a mobile crisis program in Pennsylvania over 12 months. 793 individuals recommended to various levels of care were analyzed. Clinical and demographic presentations were used as predictor variables and level of care recommendation as outcome variable. Several clinical presentations were found to increase the likelihood of various levels of care recommendations. These findings are discussed in light of current suicide intervention and data-driven practice.