ICU memories and patient outcomes in a low middle-income country : a longitudinal cohort study

Abstract

OBJECTIVES: To study memories of ICU following discharge, their associations, and impact on mental health and quality of life in a low- and middle-income country.

DESIGN: Prospective observational cohort; data on memories (pain, fear, nightmare, factual), clinical and demographic variables, anxiety-depression, posttraumatic stress symptoms, and quality of life were collected 0, 7, 14, 30, 90, and 180 days post discharge. Home visits for assessment minimized loss to follow-up. Linear mixed-models and regression analyses were used to estimate adjusted effects of memories controlling for age, sex, time, and severity of illness.

SETTING: Twenty-five bedded ICU of a tertiary care center in East India.

PATIENTS: Adult ICU survivors between January 2017 and July 2018 able to communicate their memories.

INTERVENTIONS: Not applicable.

MEASUREMENTS AND MAIN RESULTS: Final sample consisted of 322 patients who completed 180 days follow-up. Pain, fear, factual, and nightmare memories dropped from 85%, 56%, 55%, and 45% at discharge to less than or equal to 5% at 180 days. Patients with gaps in ICU memory had worse anxiety-depression, posttraumatic stress symptoms, and quality of life at all follow-up points. Sedation (odds ratio, 0.54; CI, 0.4–0.7), steroids (odds ratio, 0.47; CI, 0.3–0.8), benzodiazepines (odds ratio, 1.74; CI, 1–3.04), and mechanical ventilation (odds ratio, 0.43; CI, 0.2–0.8) were independently associated with gaps in memory. Non-ICU factor such as substance addiction (odds ratio, 5.38; CI, 2–14) was associated with memories affecting mental health and quality of life.

CONCLUSIONS: Gaps in memory and various memory types were common after ICU admission, whose prevalence waned over time. Compared with nightmares and fearful memories, gaps in memories were most strongly associated with poor mental health and quality of life. Identifying patients with gaps in memories might be an objective way of planning interventions to improve their long-term outcomes.