A tertiary analysis of the AVERT study reveals divergent findings suggesting that mortality increases with VEM compared to usual care after restricting follow-up to 14 days.
The high dose, intensive training of very early mobilization (VEM) increased mortality in patients within 14 days post-stroke when compared to usual care (UC), according to tertiary results of the prospective, parallel group, randomized clinical trial, AVERT (A Very Early Rehabilitation Trial).
After adjusting for age and stroke severity according to the National Institute of Health Stroke Scale score (NIHSS), patients treated with VEM had an adjusted odds ratio (aOR) of death of 1.76 (95% CI, 1.06–2.92; P = .029) within 14 days post-stroke compared to patients treated with UC. Within this time frame, 48 patients had died in the VEM group (4.56%) and 32 in the UC group (3.05%), for an overall fatality rate of 3.8% (n = 80). Most deaths were caused by stroke-related events (VEM, n = 29; UC, n = 16).
Study author Julie Bernhardt, PhD, laboratory head, Avert Early Rehabilitation Research Group, and director, NHMRC Centre of Research Excellence in Stroke Rehabilitation and Recovery, stated that following this research, “clinical guidelines have been revised to reflect this finding and, while not uniform, recommendations to delay or modify mobilization practices now exist.”
Bernhardt and colleagues performed a tertiary analysis of their earlier study, AVERT, published in Neurology, in which no significant differences between VEM and UC deaths were observed within 3 months. Bernhardt and colleagues then restricted follow up to 14 days post-stroke for this analysis, with significant results. A total of 2104 participants with a median age of 72 years (interquartile range [IQR], 63–80) and a median NIHSS of 7 (IQR, 4–12) were recruited to the study within 24 hours of stroke onset.
An outcomes committee judged whether any fatal or non-fatal serious adverse events (SAEs) were unrelated, probably not related, possibly related, or probably related to treatment. Deaths were assessed to be “probably” or “possibly” related to treatment in 26 (54%) patients in the VEM group and 14 (44%) in the UC group.
Selection and pattern mixture models used for sensitivity analyses supported the primary analysis (aOR, 1.76; 95% CI, 1.06–2.92; P = .029). Subgroup analyses revealed several trends that were not statistically significant (P >.05). Patients faced greater odds of death when treated with VEM instead of UC if they were over 80 years of age (aOR, 2.38; 95% CI, 1.12–5.08) or suffered an intracerebral hemorrhage stroke (aOR, 4.17; 95% CI, 1.06–16.43). Treatment by subgroup interactions were not significant due to the small number of deaths in patients over 80 years old (VEM, n = 27; UC, n = 13) and patients with intracerebral hemorrhage (VEM, n = 12; UC, n = 3).
Bernhardt and colleagues stress that the associations of VEM with age and intracerebral hemorrhages should be further explored, stating that “understanding who is most at risk with early activity is important for development of safe protocols for acute stroke patients.” They hypothesize that upright activity may relate to cerebral autoregulation soon after stroke. Various studies have investigated the effect of head position on cerebral blood flow without reaching a consensus. Prior to this analysis of AVERT, no other studies analyzing VEM have been able to show statistical significance.
Prior to any mobilization, 9 VEM and 10 UC participants suffered SAEs. In patients who died, the median time from stroke onset to death was 4 days in the VEM group, with 73% (n = 35) mobilized within 24 hours and 5 days in the UC group, with 34% (n = 11) mobilized within 24 hours. Nineteen participants died prior to any mobilization; 9 in VEM and 10 in UC. Bernhardt and colleagues did not collect any physiological measures of the patients or assess any ongoing occlusion and call for future studies to take these factors into account.
Bernhardt and colleagues also investigated the effect of VEM on non-fatal SAEs, of which 124 occurred in the VEM group and 130 occurred in the UC group. There was no significant difference between groups of non-fatal, stroke-related, and immobility-related SAEs. Non-fatal serious falls (VEM, n = 2; UC, n = 4) were too few to be analyzed.
The authors concluded that more extensive research must be conducted as “the clinical and research community continue to seek greater guidance for clinical protocols in this early period post stroke.”