RESUMO
OBJECTIVE: We tested the feasibility of a randomized controlled trial for comparing primary nursing with standard care. RESEARCH METHODOLOGY: Elective cardiac surgical patients were eligible for inclusion. Patients with an intensive care unit stay of ≥ 3 days were followed up until intensive care unit discharge. Recruitment period was one year. SETTING: Two intensive care units at a university hospital specialized in cardiovascular and diabetic diseases. MAIN OUTCOME MEASURES: Primary outcomes were recruitment and delivery rate. Primary clinical outcome was duration of delirium, as assessed by the Confusion Assessment Method for Intensive Care Units. Secondary outcomes included the incidence of delirium, anxiety (10-point Numeric Rating Scale), and the satisfaction of patient relatives (validated questionnaire). RESULTS: Of 369 patients screened, 269 could be allocated to primary nursing (n = 134) or standard care (n = 135), of whom 46 patients and 48 patients, respectively, underwent an intensive care unit stay ≥ 3 days. Thus, recruitment and delivery rates were 73 and 26 %, respectively. During primary nursing and standard care, 18 and 24 patients developed a delirium, with a median duration of 32 (IQR: 14-96) and 24 (IQR: 8-44) hours (P = 0.10). The risk difference of delirium for primary nursing versus standard care was 11 % and the relative risk was 0.65 (95 % CI: 0.28-1.46; P = 0.29). The extent of anxiety was similar between groups (P = 0.13). Satisfaction could be assessed in 73.5 % of relatives, without substantial differences between groups. CONCLUSION: Data demonstrate that a trial for comparing primary nursing with standard care is generally feasible. However, the incidence of delirium may be a better primary outcome parameter than delirium duration, both in terms of long-term patient outcome and robustness of data quality. IMPLICATIONS FOR CLINICAL PRACTICE: A randomized clinical trial regarding nursing organization during intensive care unit stay requires detailed planning of patient recruitment, data evaluation, and power calculation.
