View Full Version : Summary: Attrition Rates Survey

Amy Courtney
04-13-1995, 04:37 AM
Dear BIOMCH-L readers:

On 3/21/95 I posted an informal survey related to attrition rates in
studies with human volunteers as subjects. Thanks to those willing to share
their experience. The general points will probably not come as a surprise
to any of you, but I didn't receive enough input to make quantitative
recommendations. However, those who did reply had diverse experiences, and
I hope some of you will be interested in the vignettes provided below.

Some main conclusions from responses and my own experience:

=-= Subjects with a perceived or actual connection with or health benefit from
the institution or experiment are more likely to complete the study.
If this is not present, even money is not enough to keep people coming back
(although there is arguably some amount of money which would keep em'
coming back!) For example, only about 25% of the subjects recruited from the
emergency room in one study returned for any given follow-up visit. On the
other hand, 75% of controls recruited from med students and staff completed
the study. In another study, only 2 of 12 women in the study of abdominal
muscles in pregnancy described below dropped out. The investigator noted
the usefulness of "enthusiasm and really talking about how great it was that
they were volunteering." I have seen this tactic work well in experiments
here at the Cleveland Clinic. In this sense, retaining subjects is a skill,
but not a scientific one; rather, it has a lot to do with the "bedside manner"
(benchside manner?) of the person collecting the data!

=-= Repeated visits for physical training etc. can result in large attrition
rates over time; 50% dropout is not unusual among college students in
exercise studies.

=-= Monetary incentives ranged from $0 to $10 to $100 per trial among the
responses I received. Apparently, whatever the market requires and the
budget allows is what goes. It seems that the greater the personal benefit
and the less invasive the study to the subject, the lower the monetary
incentive needs to be. College students often come rather cheaply
(I once participated in a study for a $5 gift certificate to the favorite
local ice cream store), although one respondent noted that paying subjects
"well" is "getting harder at an expensive university where the students
are already well off and they think they are doing YOU a favor by
participating." (to which I have to respond, "well, they ARE").

=-= Of the five studies described below, three investigators recruited
additional subjects "after the fact", one did not, and the question did
not apply to one.

Brief summaries of the studies (in my own words, so investigators please
pardon any minor mis-interpretations):

=-= Comparison of the effects of a 10-week land-based or deep water based run
training program on VO2peak, strength, and running performance
Randy Jensen jensen@coe.unt.edu

11/24 subjects who were recruited from classes and posted flyers
completed the study, which offered no incentive. Subjects exercised
35 min/day 3 days/wk, plus maximal treadmill and deep water running
max tests pre and post; two 1.5 mile runs pre and one post; two
200 m runs pre and one post; isokinetic strength measures of knee,
hip and shoulder flex/ext. 8/13 subjects who dropped out did so because
they didn't want to do the post-tests.

=-= Long-term changes in masticatory performance, jaw muscle strength, and
range of mandibular movement
Gaylord Throckmorton throckmo@utsw.swmed.edu

Part I: 60/80 medical students, oral surgery residents and staff
completed the study. Subjects were recruited via class announcements
and word-of-mouth and were paid $25 for each of three trials over a
one-year period.

Part II: Approximately 130/200 patients who elected to have
orthognathic surgery to correct skeletal malocclusions completed
the study. Subjects were asked to participate by their surgeon
and were paid $50 for each of 6 trials over up to 4 years.
Approximately 120/200 patients showed up at any given time point.

Part III: Approximately 32/90 patients who came to the emergency
room for treatment of jaw fractures completed the study. Subjects
were asked to participate by the resident surgeon running the study
and were paid $100 per trial for each of 6 trials over up to
4 years. Approximately 23/90 patients showed up at any given time

=-= Longitudinal investigation into the structure and function of the abdominal
muscles during pregnancy and immediate post-birth period.
Wendy Gilleard W.Gilleard@cchs.su.edu.au

Controls: 10/12 female college students completed the study. Subjects
were recruited via posters and were given no monetary incentive for
being tested once a month for 4 months. Control subjects were not
available as long as the experimental subjects due to end of semester.

Subjects: 3/6 pregnant women completed the study as planned, but
some data were obtained for all subjects. Subjects were recruited
by word of mouth and newspaper ads. No monetary incentive was offered,
but subjects were highly interested in the study, and even suggested
post-birth testing, which was incorporated into the protocol.

=-= Eye movements recorded using EOG's while subjects were seated in a
rotating chair and viewing a rotating optokinetic field of light spots.
Grace Peng g-peng@nwu.edu

10/11 male and female subjects in their 20's completed the study.
Subjects were recruited by personal solicitation, and 70% of those
asked to participate agreed. Subjects were paid $30 for a single,
6-hour long test session. The investigator noted the value of making
sure that subjects are clearly aware of all the experimental conditions.

=-= Study of maximal isometric strength in the workspace reach envelopes of
male and female paraplegics
Nancy Black blackn@umoncton.ca

13/17 paraplegic adults 18-65 years of age (8M, 9F) were recruited
via an association newsletter or were solicited based on previous
study participation. Subjects were paid $10/hour for one session
of 2-3 hours participation. 3 subjects did not complete the study
due to illness; 1 subject did not show up.

=-= One year prospective clinical trial on back pain
Kim Burton kburton@cix.compulink.co.uk

186/252 subjects completed the study. Patients attending a
primary care facility were verbally asked to participate by their
physician; 99% of those so asked agreed to participate. No
monetary incentive was offered for filling out questionnaires
at 0, 3, 6, and 12 months or a one-year follow-up visit.
Care, consideration, and stressing the importance of the study
was viewed as helpful in minimizing attrition rates.


Amy Courtney

Amy C. Courtney, Ph.D.
Department of Biomedical Engineering
Research Institute
The Cleveland Clinic Foundation
Cleveland, OH