Description of the SUPPORT Intervention

BACKGROUND: The purpose of Study to Understand Prognoses and Preferences for Outcomes and Risks of Treatments (SUPPORT) was to improve outcomes for seriously ill hospitalized adults by improving information and decisionmaking. The SUPPORT intervention has been characterized only briefly in previous publications. OBJECTIVE: To characterize the intervention in SUPPORT and its implementation. DESIGN: Reports derived from training and administrative materials, quantitative descriptions of implementation activities, and qualitative analysis of narrative reports and focus group participation by the intervention nurses. SETTING AND PATIENTS: SUPPORT enrolled 2652 patients in the intervention arm and 2152 in the control arm of a block‐randomized trial of enhanced information, counseling, and support. The patients were hospitalized with one of nine serious illnesses in one of five US teaching hospitals between 1992 and 1994. MEASUREMENTS: (1) Reports on training and supervisory materials; (2) Rates of intervention component completion from contact logs and reports completed by the intervention nurses; and (3) grounded theory analysis of patient narratives, overview questionnaires, and focus group transcripts from the intervention nurses. RESULTS: Prognostic reports were delivered on time to the caregiving team in 83% of cases on Day 3. Reports of surrogate interviews of patient preferences and understanding were delivered on time to the caregiving team in 72% of first week cases. Patients' own reports of preferences were unavailable for 56% of cases in the first week. Overall, 39% of the rest of the patients had their interview information delivered on time to the caregiving team. The SUPPORT intervention nurses averaged 8.5 contacts with patients, 7.6 with surrogates, 3.5 with physicians, and 11.7 with other staff. The intervention nurses felt that they were fully involved in 81% of cases and had a limited role in another 14%. The major issues for patients were: understanding their situation, communication and decision‐making, advance planning, do not resuscitate orders, and general support, including support for loss and grieving. The narrative sources showed that the nurses were enthusiastic, dedicated, and strong in their support of the study objectives. They identified various barriers to effectiveness and voiced doubt that the analytic targets would show an effect from the intervention. CONCLUSIONS: The SUPPORT intervention was implemented vigorously and completely.