WSO - June 2023 - 535

Dalton et al.
535
Figure 1. A conceptualized model of the sampling process.
(Table 1). A criterion related to upper limb, cognition, language,
and contaminant condition were present in 70% or
more of the trials in the published data set and were also
included in the prospective data set (Table 1). An agerelated
criterion that restricted inclusion beyond adult (e.g.
age 20-80)16 was present in 50% of the published literature
but was not included in the prospective data set (Table 1).
Eligibility criteria filter (results)
An internal validity criterion was the primary exclusion
reason for 80.9% of patients from the published data set
(data available n = 6/30 trials) and 74.9% of patients from
the prospective data set. Internal validity criteria related to
upper-limb impairment (published: 20.4%, prospective:
45.9%) and no confirmed stroke (published: 15.4%, prospective:
22.9%) were the most common across both data
sets (Table 1). Primary and secondary exclusion reasons
were only available in the prospective data set; therefore,
the remaining eligibility criteria filter results are based on
the prospective data set only. Nearly 70% of patients were
excluded for more than one reason, with the combination of
internal validity and feasibility criteria most common
(n = 461, 45.8%), while a safety criterion alone was least
common (n = 7, 0.7%; Figure 2(a)).
The number of times individual criteria were implemented
as either a primary or secondary exclusion reason were captured
(Figure 2(b)). A SAFE score outside trial inclusion was
the most common exclusion criterion (n = 676), followed by
discharge outside of catchment (n = 293), discharge before
screening (n = 289, most before day 3), and no confirmed
stroke (n = 231). All but one exclusion criterion (palliation)
had a majority proportion of patients excluded for more than
just that criterion, as demonstrated by the stacked bar chart
(Figure 2(b)). The top-4 criteria accounted for nearly threequarters
(74.0%) of all exclusion reasons (Figure 2(b)).
To understand the impact of the top-4 criteria, the accessible
population with a confirmed stroke and reported
SAFE score (n = 669) were further examined. Scores were
grouped based on the SAFE eligibility criterion (too severe
SAFE 0-1; trial eligible SAFE 2-7; too good SAFE 8-10)
and patient characteristics described in Table 2. The majority
(84.2%) of patients fell outside the SAFE range (2-7),
with 38.1% of SAFE scores collected equalling 10. Just
under half (45.0%) received one or more acute stroke interventions
(Table 2). The median length of stay was 5 (3.0,
10.0) days. The length of stay for the too severe group was
over double (11 days) that of the too good group (4 days).
Patients
discharged
outside catchment
and
before
screening were the next two most prevalent exclusion reasons.
Figure 3 depicts the discharge location of the sample
grouped by SAFE score. Most of the too severe group
(42.6%) and trial-eligible group (39.6%) were discharged
to another acute hospital. These groups were often unable
to be recruited due to a length of stay shorter than the
required 8 days to complete the trial protocol.
International Journal of Stroke, 18(5)
WSO - June 2023

Table of Contents for the Digital Edition of WSO - June 2023
