Site Characteristics
A total of 55 sites obtained ethics approval and signed clinical trial agreements. Forty-eight (87%) of these open sites recruited at least one patient during the study period and were deemed active, while seven sites did not enroll any patients in the study. There were 44 other sites that expressed interest and/or received study contracts but never completed the process of becoming an open site. Table 1 and Figure 1 demonstrate the number of active sites and patients enrolled per month over the study period. A total of 36 active sites were North American and 12 were international. When characterized by healthcare model, 19 sites were predominately publicly funded and the remaining 29 were predominately privately funded. Of the privately funded sites, 28 were based in the United States and one site was based in Argentina.
Table 1: Average recruitment over the study period.
Year
|
Average Monthly Enrollment
|
Open Sites
|
2013
|
2.8
|
13
|
2014
|
3.1
|
20
|
2015
|
4.2
|
33
|
2016
|
6.5
|
37
|
2017
|
10.2
|
45
|
2018
|
14.0
|
54
|
2019
|
11.0
|
55
|
Patient enrollment rate
A total of 602 patients from 48 clinical sites in 12 countries were recruited from January 2013 through to October 2019. While the average number of patients enrolled throughout the study was 7.3 patients/month (SD +/- 5.2), average monthly enrollment increased every year until the last year of the study and enrollment was greater than 10 patients/month in 2017, 2018 and 2019 (Table 1 and Figure 1). While 58 months were required to reach 50% of target enrollment (January 2013 to October 2017), only an additional 24 months were required to achieve 100% enrollment. The top 20% of sites in terms of patient enrollment accounted for almost 60% of all included patients (355/602) (Table 2). The highest enrolling single site for the trial accounted for 14% of all patients (84/602).
Table 2. Active sites in order of total recruitment.
Site Number
|
Location
|
Site total
|
1
|
International
|
84
|
2
|
North America
|
58
|
3
|
North America
|
46
|
4
|
North America
|
36
|
5
|
North America
|
36
|
6
|
International
|
33
|
7
|
North America
|
22
|
8
|
North America
|
20
|
9
|
North America
|
20
|
10
|
North America
|
19
|
11
|
North America
|
15
|
12
|
International
|
13
|
13
|
North America
|
13
|
14
|
North America
|
13
|
15
|
North America
|
12
|
16
|
International
|
12
|
17
|
North America
|
11
|
18
|
North America
|
11
|
19
|
North America
|
10
|
20
|
International
|
9
|
21
|
North America
|
9
|
22
|
International
|
8
|
23
|
North America
|
8
|
24
|
International
|
6
|
25
|
North America
|
6
|
26
|
North America
|
5
|
27
|
International
|
5
|
28
|
North America
|
5
|
29
|
International
|
5
|
30
|
North America
|
5
|
31
|
North America
|
5
|
32
|
North America
|
4
|
33
|
North America
|
4
|
34
|
North America
|
4
|
35
|
North America
|
4
|
36
|
North America
|
3
|
37
|
North America
|
3
|
38
|
North America
|
3
|
39
|
International
|
3
|
40
|
North America
|
2
|
41
|
International
|
2
|
42
|
North America
|
2
|
43
|
North America
|
2
|
44
|
International
|
2
|
45
|
North America
|
1
|
46
|
North America
|
1
|
47
|
North America
|
1
|
48
|
North America
|
1
|
Site opening and enrollment
There was no correlation between TTSU and the site’s average monthly enrollment (Pearson R = 0.27, p =0.069) [Figure 2]. However, once set up, the time needed for a site to recruit their first patient (TTFP) was predictive of the number of total patients recruited and the average recruitment rate of that site. Generally, the time needed to recruit the first patient was inversely correlated to the average monthly enrollment rate of that site (Pearson R = -0.44, p = 0.0033). In fact, sites that required longer than one year to recruit their first patient had an average recruitment rate three times lower than that of sites that were able to recruit their first patient within one year (0.09/month vs 0.32/month, p= 0.04).
Site location and health care system
There were 36 North American and 12 international active sites that enrolled patients. While North American sites were able to become fully enrollment-ready significantly faster than international sites (TTSU 19.5 vs 27.0 months, p=0.04), international sites had a significantly higher average monthly enrollment rate once enrolling (0.2/month vs. 0.59/month, p=0.023) (Tables 3 and 4). The difference in TTFP between North American and international sites was not statistically significant (8.7 vs. 4.5 months, p=0.25). When compared to North American sites, there were a significantly higher proportion of international sites that demonstrated interest but did not become active or recruit a single patient (25/61 international vs. 26/38 North American, p =0.007). When considering health care model, there was no significant difference in TTSU (20.1 vs. 23.4 months, p=0.40) or TTFP (8.1 vs. 6.9 months, p=0.70) between privately and publicly funded sites. However, the 19 publicly funded sites had significantly higher monthly enrollment compared to privately funded sites (0.4/month vs. 0.17/month, p=0.03).
Research Personnel
Of active enrolling sites, 40 (83%) had research support personnel and 8 sites did not. Sites with research personnel were able to reach enrollment ready status significantly faster than sites without research support (TTSU of 19.3 vs. 30.3 months, p=0.03). However, there was no significant difference between in recruitment rate per month once the sites were actively enrolling (0.28/month vs. 0.2/month, p=0.63). There was no significant difference in TTFP between centers with and without research support (8.0 vs. 5.9 months, p=0.62).
Table 3. Differences in monthly recruitment.
Group
|
Enrollment
(patients/month ± SD)
|
p-value
|
North American (36)
International (12)
|
0.20 ±0.18
0.59 ±0.73
|
0.023*
|
Public/Mixed (19)
Private (29)
|
0.40 ± 0.56
0.17 ± 0.14
|
0.03*
|
Research Personnel (40)
No Research Personnel (8)
|
0.28 ± 0.42
0.2 ± 0.16
|
0.63
|
Two-tailed non-paired t-tests were to test statistical significance. * indicates statistical significance with alpha=0.05. SD = standard deviation.
|
Table 4. Time to start-up (TTSU) and Time to first patient (TTFP) by site factors
Group
|
Time to start up (months ± SD)
|
p-value
|
North American (36)
International (12)
|
19.5 ± 9.0
27.0 ± 19.5
|
0.04*
|
Public/Mixed (19)
Private (29)
|
23.4 ± 16.5
20.1 ± 9.4
|
0.39
|
Research Personnel (40)
No Research Personnel (8)
|
19.3 ± 9.6
30.3 ± 22.3
|
0.03*
|
Group (n)
|
Time to first patient (months ± SD)
|
p-value
|
North American (36)
International (12)
|
8.7 ± 11.9
4.5 ± 5.1
|
0.25
|
Public/Mixed (19)
Private (29)
|
6.9 ± 15.0
8.1 ± 7.02
|
0.7
|
Research Personnel (40)
No Research Personnel (8)
|
8.0 ± 11.3
5.9 ± 7.3
|
0.62
|
Two-tailed non-paired t-tests were to test statistical significance. * indicates statistical significance with alpha=0.05. SD = standard deviation
|