Original Clinical ScienceçGeneral

Community-Based Interventions and Individuals'

Willingness to be a Deceased Organ Donor:
Systematic Review and Meta-Analysis
Andrew T. Li,1,2,3 Germaine Wong,2,3,4 Michelle Irving,2,3 Stephen Jan,5 Allison Tong,2,3 Angelique F. Ralph,2,3
and Kirsten Howard2,6

Background. Widespread in-principle community support for organ donation does not necessarily translate to individuals be-
coming organ donors after death. Previous studies have identified factors that influence individuals' decisions to become organ
donors, which may be effectively targeted by interventions. We aimed to describe and evaluate the effectiveness of community-
based interventions to increase the willingness of individuals to be a deceased organ donor. Methods. We systematically reviewed
all randomized controlled trials (RCTs), non-RCTs (NRCTs), and before-after studies that assessed the impact of interventions on
increasing the willingness to be a deceased organ donor (measured as commitment to donate and/or intention to donate). We
searched MEDLINE, Embase, PsycINFO, and CINAHL, without language restriction, to December 2013 and the reference lists
of the included articles. We conducted a risk of bias assessment using the Cochrane risk of bias tools and assessed confidence
in the evidence using the Grading of Recommendations Assessment, Development, and Evaluation framework. Results. We
identified 63 studies (11 RCTs, 8 cluster-RCTs, 4 NRCTs, 8 cluster-NRCTs, 27 before-after studies) with over 170000 participants.
Overall, the quality of the evidence was low. Participants who received a broad range of community-based interventions were more
likely to commit as donors (7 cluster-RCTs; 6015 participants; relative risk, 1.70; 95% confidence interval [95% CI], 1.22-2.36;
I2 = 94%, P = 0.002), and had higher levels of willingness to donate (3 RCTs, 393 participants; standardized mean difference,
0.29; 95% CI, 0.01-0.56; I2 = 45%; P = 0.04) than those who did not receive the interventions, but not the intention to donate
(315 participants; relative risk, 1.19; 95% CI, 0.94-1.51; P = 0.14). Conclusions. Community partnerships and active learning
community-based interventions may be effective in increasing the commitment, but not intentions to donate. However, the overall
risk of bias for was high, and this may have led to overestimation of the relative treatment effects of these interventions.
(Transplantation 2015;99: 2634–2643)

F or many patients with end-stage organ failure, transplan-

tation is the only lifesaving treatment option. However,
a critical shortage of donor organs exists worldwide, and
in most countries, deceased organ donation rates are inade-
quate to address the growing demand for organs. On aver-
age, more than 1 person in the United Kingdom1 and
almost 15 people in the United States die each day on the
transplant waiting list.2 In Australia, the median waiting time
on dialysis for a donor kidney is 3.3 years.3 The rates of de-
ceased organ donation vary greatly internationally, from
36.5 donors per million populations in Croatia to 0.09 do-
nors per million population in Morocco. 4
Received 12 February 2015. Revision received 20 May 2015.
Accepted 3 June 2015.
1
Monash School of Medicine, Monash University, Victoria, Australia.
2
Sydney School of Public Health, University of Sydney, New South Wales, Australia.
3
Centre for Kidney Research, The Children's Hospital at Westmead, New South
Wales, Australia.
4
Centre for Transplant and Renal Research, Westmead Hospital, New South Wales,
Australia.
5
The George Institute for Global Health, Camperdown, New South Wales, Australia.
6
The Institute for Choice, University of South Australia, North Sydney, New South
Wales, Australia.
The PAraDOx study is funded by an Australian Research Council Discovery Project
Grant (DP0985187). The funders’ played no role in the research or the decision
to publish.
In many countries, strong in-principle support for organ
donation in the community does not always translate into a
personal decision to become an organ donor. 5,6 The factors
that influence such decisions may be multifactorial and in-
clude knowledge about organ donation, personal beliefs,
such as altruism or fear of medical neglect; external influ-
ences, such as family and culture; emotional influences, such
as grief or apathy; and the prevailing institutional and policy
context, such as the complexity of the consent system.6 These
factors may represent potential levers for intervention to in-
crease the community's willingness to become a deceased or-
gan donor. A considerable variety of approaches have been
taken across and within countries to increase deceased organ
The authors declare no conflict of interest.
K.H., S.J., M.I., and G.W. designed the study. A.T.L., A.F.R., G.W., and A.T. conducted
the data extraction and analyses. All authors contributed to the interpretation of the
analyses. A.T.L. drafted the article. All authors contributed to the writing and review of
the article.
Correspondence: Germaine Wong, Centre for Transplant and Renal Research,
Westmead Hospital, Cnr Hawkesbury Rd & Darcy Rd, Westmead New South Wales
2145, Australia. (germaine.wong@health.nsw.gov.au).
Supplemental digital content (SDC) is available for this article. Direct URL citations
appear in the printed text, and links to the digital files are provided in the HTML text
of this article on the journal’s Web site (www.transplantjournal.com).
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
ISSN: 0041-1337/15/9912-2634
DOI: 10.1097/TP.0000000000000897

2634 www.transplantjournal.com Transplantation ■ December 2015 ■ Volume 99 ■ Number 12

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

© 2015 Wolters Kluwer
donation. The types of interventions for increasing community
willingness to be a deceased organ donor that have been stud-
ied are diverse and include mass media campaigns, educa-
tional materials, and community partnerships, among others.
We aimed to identify and evaluate the totality of evidence re-
garding the effectiveness of community-based interventions
to increase deceased organ donation.
MATERIALS AND METHODS
We conducted a systematic review based on standard methods
and reporting in accordance with the Preferred Reporting Items
for Systematic Reviews and Meta-Analyses statement. 7
Inclusion and Exclusion Criteria
We included studies (randomized controlled trials [RCTs],
non-RCTs [NRCTs], and before-after studies) that quantita-
tively assessed the impact of interventions to increase the will-
ingness of members of the community to become solid organ
donors after death. We included studies that used individual
level assignment (RCTs and NRCTs) and those that used
group level assignment (cluster RCTs [CRCTs] and cluster
NRCTs). We imposed no restrictions on article language.
We excluded abstracts and non-research articles.
Search Strategies
We searched MEDLINE, Embase, PsycInfo and CINAHL
from inception to December 2013. MeSH terms and key-
words for organ donation, interventions, and comparative
studies (eg, “RCT”) were used. The search strategies are pro-
vided in Appendix 1, SDC, http://links.lww.com/TP/B187.
Data Extraction
Two authors (A.T.L. and M.J.I.) independently assessed all
titles and abstracts for eligible studies and discarded those
that did not meet the inclusion criteria. We extracted data
on study design, geographic location, setting, sample size,
participants, interventions and comparators, and the rates
of commitment to donate and intention to donate. Where
outcome data were reported in a form that could not be
meta-analyzed, we contacted corresponding authors to seek
the relevant data including the number of events and group
sample sizes for the dichotomous outcomes, and the mean es-
timates, standard deviations, and group sample sizes for the
continuous outcome.
Outcomes Measures
The 2 prespecified outcomes that we assessed were com-
mitment to donate and intention to donate. We defined com-
mitment to donate as documentation of being a donor on an
organ donor registry, driver's licence, donor card, or other
official record, communication to the next of kin that one
wishes to become an organ donor and/or otherwise self-
reporting that one is an organ donor.
Intention to donate could be reported as a dichotomous
outcome (“positive intent to donate”) or a continuous out-
come (“willingness to donate”). We defined positive intent
to donate as any action or statement that indicates one is will-
ing to be an organ donor, document being a donor on an
organ donor registry, driver's licence, donor card, or other of-
ficial record, and/or communicate to the next of kin that they
wish to become an organ donor, but without committing to
donation. We defined willingness to donate as self-assessed
willingness to become an organ donor, document being a
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Li et al 2635
donor on an organ donor registry, driver's licence, donor
card, or other official record, and/or communicate to the next
of kin that one wishes to become an organ donor.
Risk of Bias
Two authors (A.T.L. and A.F.R.) independently assessed
the risk of bias and resolved all discrepancies by consensus
for all included studies. For trials, we used the Cochrane Risk
of Bias Tool, 8 and for before-after studies, we used a modi-
fied version of the Cochrane Effective Practice and Organi-
sation of Care Group suggested risk of bias criteria for
interrupted time series studies. 9
Quality of Evidence
We assessed the quality of the evidence informing the sum-
mary estimates using the Grading of Recommendations As-
sessment Development and Evaluation guidelines. 10
Data Synthesis and Statistical Analyses
We expressed results from individual trials and before-
after studies as risk ratios (RR) with 95% confidence inter-
vals (95% CI) for dichotomous outcomes and as standard-
ized mean differences (SMD) with 95% CI for continuous
outcomes. We obtained the summary estimates of binary out-
comes using the DerSimonian and Laird random effects
model. We considered P values ≤0.05 to be statistically sig-
nificant. We quantified heterogeneity using the χ2 test and
I2 statistic and used preplanned subgroup analyses by inter-
vention types and target populations, and retrospective sub-
group analyses by geographic location of studies and study
quality, to explore the possible sources of heterogeneity.
Where possible, we undertook a sensitivity analysis based
on the quality of the studies and assessed publication bias
using funnel plots. We conducted all analyses using Review
Manager 5. Where appropriate, we also calculated the num-
ber needed to treat from the population attributable risk
(PAR) using the following formula, where Pe represents the
total proportion of participants who were exposed to the in-
terventions, and RRe represents the relative risks (RRs) given
by the meta-analyses:
1 1 1 þ Pe ðRRe ‐1Þ
NNT ¼ ¼ ¼
PAR Pe ðRRe ‐1Þ Pe ðRRe ‐1Þ
1þPe ðRRe ‐1Þ
RESULTS
Characteristics of Studies
We included 49 articles, incorporating a total of 63 stud-
ies as some articles reported findings on multiple studies
(Figure 1). All articles were from primary sources. The study
characteristics are summarized in Table 1. Five studies did
not report their sample size. The remaining 58 studies in-
volved 174,279 participants. Thirty-one studies were trials
(8 RCTs, 11 CRCTs, 4 NRCTs, 8 cluster NRCTs) and 32
were before-after studies. In total, 42 of 63 studies (67%)
were included in the meta-analyses. The majority of the stud-
ies originated from the United States, and with an average
follow-up time of 60.5 days (Table 1).
Risk of Bias
Appendix 2 (SDC, http://links.lww.com/TP/B187) shows
the risk of bias assessments for all included studies. Of the
31 trials, 11 (35%) were judged as high risk of bias, 18 trials
2636 Transplantation ■ December 2015 ■ Volume 99
FIGURE 1. Process of studies selection.
as unclear risk of bias (58%), and 2 trials (6%) as low risk of
bias. No trials blinded participants or personnel to the inter-
ventions. Sixteen trials (52%) had high risk of reporting bias.
Allocation concealment was adequate in 9 trials (29%). The
methods of randomization were adequate in 7 trials (23%),
unclear in 14 trials (45%), and inadequate in 10 trials (32%).
Overall, 8 of the 32 before-after studies (25%) were con-
sidered as high risk of bias, 23 studies as unclear risk of bias
(72%), and a single study as low risk of bias (3%).
Visual assessment of funnel plots (not shown) showed that
the studies were distributed asymmetrically around the com-
bined effect size, suggesting publication bias (Table 2).
Quality of Evidence
Table 3 summarizes the quality of evidence for the key find-
ings of our review. The findings were based on randomised
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
■ Number 12 www.transplantjournal.com
controlled trials or before-after studies, for which the risk of bias
varied between unclear and high. The consistency of the esti-
mates of the effects was also variable, with heterogeneity
ranging from low to high. Publication bias was present.
Hence, the overall quality of the evidence was low, indicating
the confidence we placed in the effect estimates was limited,
and the true estimates were likely to be different from the es-
timate of effects.10
Types of Interventions
The included studies examined a diverse range of interven-
tions (n = 12). Education was a key strategy in the majority
of interventions, conveying factual, and/or emotive messages
relating to the need for organs and/or the process of organ
donation. These interventions included advertising, active learn-
ing, community partnerships, a computer-tailored intervention,
© 2015 Wolters Kluwer Li et al 2637

TABLE 1. to lead discussion on organ donation within their church

Characteristics of included studies members. The computer-tailored intervention (n = 1, 2%)
used a computer program that asked participants questions
Characteristic No. Studies (%) relating to donation and provided feedback based on their in-
Geographic location put. Educational materials (n = 9, 14%) included written ma-
United States of America 46 (73.02) terials such as brochures, videos, and websites that promoted
Netherlands 4 (6.35) organ donation. Interventions at DMV offices (n = 4, 6%)
Germany 3 (4.76) encompassed a range of educational interventions conducted
Turkey 3 (4.76) at DMVs (or equivalent government agency) offices to en-
United Kingdom 3 (4.76) courage customers to commit as deceased organ donors at
Colombia 1 (1.59) these locations where their decisions can be recorded. Promo-
Croatia 1 (1.59) tional materials and staff training to improve communication
South Africa 1 (1.59) with the public about organ donation were a few examples of
Sweden 1 (1.59) such interventions. Workplace interventions (n = 3, 5%) in-
Year of publication cluded educational interventions conducted in the workplace
≤2000 2 (3.2) setting. Some methods had multiple main components (n = 6,
2001-2005 10 (15.9) 10%), for example, utilization of passive learning combined
2006-2010 32 (50.8) with opportunistic registration procedures. Other nonspecific
≥2011 19 (30.2) interventions did not disseminate facts or arguments in favor
Study design of donation. This category encompassed anticipated regret,
RCT 8 (12.7) (n = 3, 5%) in which participants were asked to consider their
CRCT 11 (17.5) regret if they were to not register as organ donors, and opportu-
NRCT 4 (6.3) nistic registration procedures, (n = 2, 3%), whereby the partici-
CNRCT 8 (12.7) pants were given an opportunity to register immediately.
Before-after study 32 (50.8) The specific interventions used in each included study are sum-
Sample size marized in Appendix 3, SDC, http://links.lww.com/TP/B187.
0-500 38 (60.3)
500-1000 4 (6.3) Efficacy of Interventions
1000-10,000 15 (23.8)
Randomized Controlled Trials
>10,000 1 (1.6)
Not stated 5 (7.9) The types of interventions assessed in the meta-analyses in-
Setting 26 (41.3) cluded anticipated regret (n = 2) and educational materials
General community (n = 1). Participants who received community-based inter-
Schools 11 (17.5) ventions experienced a statistically significant improvement
Universities 12 (19.0) in the willingness to donate compared to those that did not
Churches 3 (4.8) receive the interventions (3 studies, 393 participants: SMD,
Primary care 2 (3.2) 0.29; 95% CI, 0.01-0.56; I2 = 45%; P = 0.04) (Figure 2A).
Motor registry 4 (6.3) Participants who received a personalized letter from the Ohio
Military 1 (1.6) Secretary of State (89413 participants: RR, 1.90; 95% CI,
Workplaces 3 (4.8) 1.78-2.02; P < 0.001) and those who received a personalized
Online 1 (1.6) letter and a brochure (89625 participants: RR, 1.91; 95%
Outcomes CI, 1.80-2.03; P < 0.001) experienced a significant improve-
Commitment to donate 36 (57.1) ment in the commitment to donate compared to those who
Intention to donate 34 (54.0) only received a brochure alone. No significant differences
Follow-up time ( mean, SD), d 60.5 (113.3) were observed in the proportion of participants committing
c
to deceased organ donation between those who received a
Nonrandomized controlled trial.
CNRCT indicates cluster non-RCT. personalized letter and a brochure and those who received
a personalized letter only (85790 participants: RR, 1.01;
95% CI, 0.96-1.06; P = 0.73).11 There were also no signifi-
educational materials, education by health professionals, inter- cant differences in the proportion of participants' expressing
ventions at Department of Motor Vehicles (DMV) offices, passive a positive intention to donate between those who received
learning, and workplace interventions. Advertising interventions passive education compared to those that did not (315 partic-
were disseminated through media, such as television, radio, print, ipants: RR, 1.19; 95% CI, 0.94-1.51; P = 0.14).12 Three ad-
billboard, and/or online (n = 5, 8%). The majority of advertising ditional RCTs13–15 assessed the positive intent to donate,
interventions were targeted at ethnic minorities (n = 4, 80%). Ac- but only 2 of these studies reported a significant increase
tive learning (n = 5, 8%) involved participants to complete tasks in the proportion of participants' expressing a positive inten-
that promote and educate organ donation. This contrasted with tion to donate between the intervention and comparator
interventions that involved passive learning (n = 14, 22%), such groups. The types of interventions found to be effective in-
as presentations and school lessons. cluded educational materials on organ donation compared
Community partnerships (n = 9, 14%) involved lay mem- with education materials on avoiding the common cold15
bers from specific communities to promote organ donation, and passive learning combined with a computer-tailored
such as engagement with African-American church leaders intervention.14

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

2638 Transplantation ■ December 2015 ■ Volume 99 ■ Number 12
TABLE 2.
Outcomes of meta-analyses
Outcome Studies Participants
RCTs
Intention to donate (willingness to donate) 3 393
CRCTs
Commitment to donate 7 6015
NRCTs
Commitment to donate 2 320
Before-after studies
Commitment to donate 15 3709
Intention to donate (positive intent to donate) 10 9572
Intention to donate (willingness to donate) 5 645
Cluster-Randomized Controlled Trials
The types of interventions assessed included in the meta-
analyses were community partnership (n = 2), opportunis-
tic registration procedures (n = 2), educational materials
(n = 1), intervention at DMVoffices (n = 1), and a workplace
intervention (n = 1). Participants who received community-
based interventions experienced a significant improvement
in the commitment to donate compared to those not exposed
(7 studies, 6015 participants: RR, 1.70; 95% CI, 1.22-2.36;
I2 = 94%; P = 0.002) (Figure 2A).
Additionally, a single study that evaluated the effectiveness of
community partnership reported a significant improvement in
the willingness to donate compared to those that did not receive
the interventions (1043 participants: RR, 0.21; 95% CI, 0.09-
0.33; P < 0.001).16 Interventions in the form of passive learning
did not increase the proportion of participants with positive in-
tent to donate (187 participants: RR, 1.16; 95% CI, 0.80-1.67;
P = 0.44). 17 Also, an intervention at DMVoffices was more ef-
fective in improving the proportion of participants committed
to being an organ donor compared with passive display of or-
gan donation materials. 14
Non–Randomized Controlled Trials
Anticipated regret (n = 1) and education by health profes-
sional (n = 1) were the 2 major forms of interventions included
in the meta-analysis. There were no significant differences in
the participants' commitment to donate between the inter-
vention and control arms (2 studies, 320 participants: RR,
1.05; 95% CI, 0.45-2.45; I2 = 73%; P = 0.91). Two other
studies that examined the benefits of passive learning (17, 18)
did not report a significant increase in the willingness to do-
nate and positive intent to donate.
Cluster NRCTs
Participants who were shown a videotaped presentation
containing demographic information about potential organ
recipients were more likely to take a donor sticker and card
than participants who were shown the same video without
the demographic information (328 participants: RR, 1.34;
95% CI, 1.12-1.61; P = 0.002).18 On the contrary, passive
learning with classroom education did not improve the posi-
tive intent to donate (69 participants: RR, 0.73; 95% CI,
0.40-1.33; P = 0.31).19 A combined intervention of mass me-
dia and interpersonal approach was associated with a signif-
icant improvement in the number of participants signing
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.transplantjournal.com
P I2 Outcome (summary estimate)
0.04 45% SMD, 0.29 (0.01-0.56)
0.002 94% RR, 1.70 (1.22-2.36)
0.91 73% RR, 1.05 (0.45-2.45)
<0.001 78% RR, 1.51 (1.22-1.86)
0.01 86% RR, 1.14 (1.03-1.26)
<0.001 76% SMD, 0.54 (0.23-0.86)
donor cards compared with mass media only.20 Workplace
intervention also resulted in a significant increase in the rate
of signing donor cards when compared with no intervention.21
Workplace intervention comprising of educational materials
and onsite visits by staff was associated with a significant
increase in self-reported changes in donor status among
nondonors compared with no intervention (odds ratio, 1.32;
P < 0.001) When the same intervention was compared with
another intervention comprising only educational materials,
a similar result was reported (odds ratio, 1.17; P < 0.03).22
Before-After Studies
Twenty-six before-after studies were included in the meta-
analyses. After the interventions were introduced, a signifi-
cant improvement in the proportion of participants commit-
ting as donors (15 studies, 3709 participants: RR, 1.51; 95%
CI, 1.22-1.86; I2 = 78%; P < 0.001), intending to become
donors (10 studies, 9572 participants: RR, 1.14; 95% CI,
1.03-1.26, I2 = 86%, P = 0.01,), and in the willingness to do-
nate (5 studies, 645 participants: SMD, 0.54; 95% CI, 0.23-
0.86, I2 = 76%, P < 0.001) was observed in the participants
(Figure 2B). The types of interventions assessed for commit-
ment to donate included community partnership (n = 6), ac-
tive learning (n = 4), passive learning (n = 3), advertising
(n = 1), and education by health professional (n = 1). The
types of interventions assessed for positive intent to donate
included advertising (n = 5), passive learning (n = 4), and ed-
ucational materials (n = 1). The types of interventions
assessed for willingness to donate included active learning
(n = 2), educational materials (n = 1), passive learning (n =1),
and intervention that comprised both passive learning and
opportunistic registration procedures (n = 1).
Five other before-after studies examined changes in the
commitment to donate.23–27 Two studies that focused on ed-
ucational materials (n = 1) and passive learning combined
with opportunistic registration procedures (n = 1) reported
a significant increase in the proportion of participants agree-
ing and committing to being donors compared with baseline
results.24,25 Two studies examined changes in the intention to
donate28,29 and one reported a significant improvement in
the willingness to donate using educational materials.28
Subgroup Analyses
Types of Intervention
Compared with those not exposed to any form of interven-
tion, active partnership with the communities was associated
 © 2015 Wolters Kluwer

TABLE 3.
Grading of recommendations assessment, development, and evaluation evidence profile of studies for commitment to donate and intention to donate after community-based
interventions

Quality of assessment (decrease in quality score) Summary of findings

Summary outcome Quality of
No. studies (type of study) Risk of bias/quality of eEvidence Consistency Directness Precision Publication Bias (95% CI)* evidence
Community-based interventions (all types)
7 CRCTs All CRCTs; mostly unclear risk of bias (+3) Large inconsistency; I2 = 94% (−2) Direct (0) No serious imprecision Suspected publication bias (−1) Commitment to donate
RR, 1.70 (1.22-2.36) Very low
Active learning
4 Before-after studies Lack RCTs; unclear risk of bias (+2) No inconsistency; I2 = 0% Direct (0) No serious imprecision Suspected publication bias (−1) Commitment to donate
96.4% (91.6-98.5%) Very low
Anticipated regret
2 RCTs All RCTs; mostly unclear risk of bias (+3) Some inconsistency; I2 = 32% (−1) Direct (0) No serious imprecision Suspected publication bias (−1) Willingness to donate
SMD, 0.39 (0.09-0.69) Low
Community partnership
2 CRCTs All CRCTs; high risk of bias (+2) Some inconsistency; I2 = 42% (−1) Direct (0) No serious imprecision Suspected publication bias (−1) Commitment to donate
RR, 3.00 (2.14-4.20) Very low
Advertising
5 Before-after studies Lack RCTs; unclear risk of bias (+2) No inconsistency; I2 = 0% Direct (0) No serious imprecision No publication bias (0) Positive intent to donate

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

RR, 1.00 (0.97-1.03) Low
Li et al
2639
2640 Transplantation ■ December 2015 ■ Volume 99 ■ Number 12 www.transplantjournal.com

FIGURE 2. A, Meta-analysis of randomized controlled trials for commitment to donate and intention to donate. B, Meta-analysis of before-after
studies for commitment to donate and intention to donate.

with a significant increase in the proportion of participants
documenting as committed donors (2 CRCTs, 4043 partici-
pants: RR, 3.00; 95% CI, 2.14-4.20; I2 = 42%; P < 0.001;
PAR, 49.6%). Interventions that involved anticipated regret
were associated with an improvement in the willingness to
donate compared to questionnaires that did not use antici-
pated regret (2 RCTs, 284 participants: SMD, 0.39; 95%
CI, 0.09-0.69; I2 = 32%; P = 0.010). Participants who re-
ceived interventions that required active learning, such as im-
plementing their own donation campaigns, experienced
improved commitment to donate after being exposed to the
interventions (4 before-after studies, 322 participants: RR,
2.82; 95% CI, 2.13-3.72; I2 = 0%; P < 0.001). There was
no significant improvement in the proportion of participants

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

© 2015 Wolters Kluwer Li et al 2641

FIGURE 2, continued. B. Meta-analysis of before-after studies for commitment to donate and intention to donate.

with positive intentions to donate after the introduction of
the advertising interventions (5 before-after studies, 6380 par-
ticipants: RR, 1.00; 95% CI, 0.97-1.03; I2 = 0%; P = 0.93).
Targeted Populations
Interventions targeted at ethnic minorities did not show a
significant improvement in commitment to donate (4 CRCTs,
4536 participants: RR, 1.83; 95% CI, 0.99-3.37; I2 = 95%;
P = 0.05; PAR, 29.2%). A significant improvement in the
commitment to donate (3 CRCTs, 4199 participants: RR,
2.28; 95% CI, 1.22-4.27; I2 = 91%; P = 0.01) was observed
after exclusion of a single study in which ethnically targeted
educational materials with a religious focus were compared
with ethnically targeted educational materials without a reli-
gious focus.

Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.

2642 Transplantation ■ December 2015 ■ Volume 99 ■ Number 12
Interventions that did not target specific ethnic groups also
resulted in a significant improvement in commitment to
donate (3 CRCTs, 1479 participants: RR, 1.43; 95% CI,
1.03-1.98; I2 = 88%; P = 0.03).
Geographic Location
The majority of included studies (n = 46, 73%) were orig-
inated from the United States. There were no significant dif-
ferences in outcomes, such as willingness to donate, the
commitment to donate or positive intent to donate between
studies conducted in the United States and those conducted
elsewhere.
Study Quality
Only 1 study30 was judged to have low-risk bias when con-
sidering the commitment to donate as an outcome. This study
reported a smaller but statistically significant improvement in
commitment to donate between the intervention and control
arms (1 study, 952 participants: RR, 1.16; 95% CI, 1.08-
1.24; P < 0.001). Similarly, the single RCT31 that reported
willingness to donate as an outcome and was deemed to have
a low risk of bias reported no significant differences in willing-
ness to donate between intervention and controlled arm.
Adverse Events
Adverse events were not reported in any of the included
studies.
DISCUSSION
There was considerable variation in the types of community-
based interventions used to promote and improve the rates of
deceased donation in the general population. Broadly, interven-
tions included advertising, active learning, anticipated regret,
community partnerships, a computer-tailored intervention, edu-
cational materials, education by health professionals, interven-
tions at DMV offices, opportunistic registration procedures,
passive learning, workplace interventions, and interventions
with multiple main components. Evidence from RCTs sug-
gested that community-based interventions, such as community
partnerships and anticipated regret were effective in improving
the commitment to donate. Participants who received an inter-
vention were 1.7 times (95% CI, 1.22-2.36) more likely to com-
mit to being an organ donor compared to those that did not
receive any specific intervention. However, the overall risk of
bias for individual studies was high, and this may have led to
overestimation of the true treatment effects of the interventions.
In general, strategies that involved community partnership
and active participation involving participants achieved the
greatest impact in promoting deceased organ donation. Inter-
ventions that included active partnership with the community
increased the number of committed donors by approximately
50%, compared with no intervention. Specific interventions
that targeted ethnic communities with culturally sensitive
approaches were also effective in increasing the number of
committed donors by 29.2% compared to those who did
not receive the interventions.
The studies examined in this review highlight how the deci-
sion to become an organ donor consists of 2 distinct steps:
first, the adoption of in-principle support for donation and sec-
ond, the translation of this support to a formal commitment
through registration. Our findings suggest that community-
based interventions are generally effective in relation to the
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
www.transplantjournal.com
latter, of translating a broad willingness to donate into a for-
mal commitment, but are less consistently effective in shifting
people's attitudes toward donation. The implication of this is
that intervention through community programs is best suited
to encouraging individuals already disposed to organ dona-
tion to formally register and in doing so, more generally, cap-
italizing on the high level of support for organ donation that is
known to exist in many countries. However, another possible
explanation for these findings is that actions such as registra-
tion and notifying family members are more objective and
hence may be more conclusively measured than studies that in-
cluded behavioural intentions.
Overall, our findings are broadly consistent with findings
of previous reviews that suggest specific interventions may
be effective in increasing the willingness to be a deceased or-
gan donor,32,33 in the general community and ethnic minor-
ities.32,34 Additionally, this review quantifies the potential
effectiveness of several specific intervention types, namely
community partnerships, active learning, and anticipated re-
gret, in improving the community willingness to be a deceased
organ donor.
Our study has several strengths. The inclusiveness of
our review enables it to assess the overall effectiveness of
community-based interventions in general and to compare
across different types of interventions and target populations.
We have also used the standardized approach of conducting
a systematic review whereby a systematic search of medical
databases, data extraction and analysis, and trial quality as-
sessment by 2 independent reviewers was conducted based
on a prespecified protocol. Our review has several limita-
tions. First, the mean follow-up time of all included studies
is short (60.5 days) so longer term outcomes are not avail-
able. Additionally, the included studies were of generally
low quality. Publication bias may also exist leading to an
overestimation of treatment effect because trials or observa-
tional data with more favourable outcomes are more likely
to be published than studies with negative results. Although
enquiry of the authors did not reveal additional information,
there were insufficient studies to formally evaluate such bias.
Systematic bias, such as performance bias and sampling bias,
also exists when observational studies, particularly before
and after studies, were used to evaluate the effectiveness of
an intervention. We acknowledge that the limited number
of participants in the included studies may preclude accurate
assessment of heterogeneity beyond chance. Our subgroup
analyses, although limited by the small number of included
studies (n = 2-5), showed that the heterogeneity may partly
be explained by the effects of intervention types.
Data were sparse in a number of areas. Reporting of
potential harms was only reported in a single study.30 In
addition, given the relatively small number of studies avail-
able in the meta-analyses, we were unable to explore all
outcomes across all study designs, or to perform detailed
subgroup analyses to explore interactions between the
types of interventions on the treatment efficacy between
the various types of intervention. Additionally, our review
included only studies that directly reported on changes in
the willingness to be a deceased organ donor and did not
consider indirect outcomes such as knowledge or opinions
about donation. The extent to which changes in these indi-
rect outcomes translate into actual increases in deceased
organ donations remains unclear.
© 2015 Wolters Kluwer
Overall, the evidence suggests that interventions may be ef-
fective in improving the willingness to be a deceased organ
donor, however the quality of the current evidence is poor.
Well-powered and well-designed RCTs are needed to evalu-
ate the impact of opportunistic registration procedures, antic-
ipated regret, and active learning on improving the number
of committed donors in the community. Process evaluations
alongside these intervention studies are also needed to iden-
tify the barriers and facilitators to commit as organ donors
and the impact that specific types of interventions may ulti-
mately have on deceased organ donation rates.
CONCLUSIONS
Overall, we found a lack of consistency in methods and out-
come measures in studies. The overall quality of the evidence is
also low. Despite this, existing evidence suggests that commu-
nity partnerships, anticipated regret, and active learning may
be promising approaches to increase individuals' willingness
to be a deceased organ donor in both the general community
and ethnic minorities. Future studies in the forms of well-
powered randomised controlled trials comparing the various
forms of interventions, using consistent outcome measures,
are likely to have a major impact on improving our confi-
dence in the treatment effectiveness of these techniques.
REFERENCES
1. NHS Blood and Transplant. Organ Donation and Transplantation
Activity Report 2012/13. London: NHS; 2013. Available at: http://www.
organdonation.nhs.uk/statistics/transplant_activity_report/current_activity_
reports/ukt/activity_report_2012_13.pdf.
2. Based on OPTN data as of 17 January 2014.
3. Wright J, Narayan S. Analysis of kidney allocation during 2013. 2014.
4. International Registry in Organ Donation and Transplantation. IRODaT
Newsletter 2012. Available at: http://www.irodat.org/img/database/grafics/
newsletter/IRODaT Newsletter 2012.pdf.
5. The Gallup Organisation. National Survey of Organ and Tissue Donation
Attitudes and Behaviors. Rockville, MD: Health Resources and Services
Administration; 2005.
6. Irving MJ, Tong A, Jan S, et al. Factors that influence the decision to be an
organ donor: a systematic review of the qualitative literature. Nephrol Dial
Transplant. 2012;27:2526–2533.
7. Moher D, Liberati A, Tetzlaff J; PRISMA Group, et al. Preferred reporting
items for systematic reviews and meta-analyses: the PRISMA statement.
BMJ. 2009;339:b2535.
8. Higgins J, Green S. Cochrane Handbook for Systematic Reviews of Inter-
ventions. T.C. Collaboration; 2012.
9. Cochrane Effective Practice and Organisation of Care Group. Suggested
risk of bias criteria for EPOC reviews. 2012.
10. Guyatt GH, Oxman AD, Vist GE; GRADE Working Group, et al. GRADE:
an emerging consensus on rating quality of evidence and strength of rec-
ommendations. BMJ. 2008;336:924–926.
11. Quick BL, Bosch D, Morgan SE. Message framing and medium consider-
ations for recruiting newly eligible teen organ donor registrants. Am J
Transplant. 2012;12:1593–1597.
12. Smits M, van den Borne B, Dijker AJ, et al. Increasing Dutch adolescents'
willingness to register their organ donation preference: the effectiveness of
an education programme delivered by kidney transplantation patients. Eur
J Public Health. 2006;16:106–110.
Copyright © 2015 Wolters Kluwer Health, Inc. All rights reserved.
Li et al 2643
13. Reubsaet A, Brug J, De Vet E, et al. The effects of practicing registration of
organ donation preference on self-efficacy and registration intention: an
enactive mastery experience. Psychol Health. 2003;18:585–594.
14. Reubsaet A, Brug J, Nijkamp MD, et al. The impact of an organ dona-
tion registration information program for high school students in the
Netherlands. Soc Sci Med. 2005;60:1479–1486.
15. Vinokur AD, Merion RM, Couper MP, et al. Educational web-based inter-
vention for high school students to increase knowledge and promote
positive attitudes toward organ donation. Health Educ Behav. 2006;33:
773–786.
16. Andrews AM, Zhang N, Magee JC, et al. Increasing donor designation
through black churches: results of a randomized trial. Prog Transplant.
2012;22:161–167.
17. Rodrigue JR, Krouse J, Carroll C, et al. A department of motor vehicles
intervention yields moderate increases in donor designation rates. Prog
Transplant. 2012;22:18–24.
18. Singh M, Katz RC, Beauchamp K, et al. Effects of anonymous information
about potential organ transplant recipients on attitudes toward organ
transplantation and the willingness to donate organs. J Behav Med.
2002;25:469–476.
19. Weaver M, Spigner C, Pineda M, et al. Knowledge and opinions about or-
gan donation among urban high school students: pilot test of a health ed-
ucation program. Clin Transplant. 2000;14:292–303.
20. Morgan SE, Stephenson MT, Afifi W, et al. The University Worksite Organ
Donation Project: a comparison of two types of worksite campaigns on
the willingness to donate. Clin Transplant. 2011;25:600–605.
21. Morgan SE, Miller J, Arasaratman LA. Signing cards, saving lives: an eval-
uation of the worksite organ donation promotion project. Commun
Monogr. 2002;69:253.
22. Morgan SE, Harrison TR, Chewning LV, et al. The effectiveness of high-
and low-intensity worksite campaigns to promote organ donation: the
workplace partnership for life. (Author abstract)(Report). Commun Monogr.
2010;77:341–356.
23. D'Alessandro AM, Peltier JW, Dahl AJ. Use of social media and college
student organizations to increase support for organ donation and advo-
cacy: a case report. Prog Transplant. 2012;22:436–441.
24. Fahrenwald NL, Belitz C, Keckler A. Outcome evaluation of 'sharing the gift
of life': an organ and tissue donation educational program for American
Indians. Am J Transplant. 2010;10:1453–1459.
25. Siegel JT, Alvaro EM, Crano WD, et al. Passive-positive organ donor reg-
istration behavior: a mixed method assessment of the IIFF Model. Psychol
Health Med. 2010;15:198–209.
26. Harrison TR, Morgan SE, King AJ, et al. Saving lives branch by branch: the
effectiveness of driver licensing bureau campaigns to promote organ do-
nor registry sign-ups to African Americans in Michigan. J Health Commun.
2011;16:805–819.
27. Zaramo CE, Morton T, Yoo JW, et al. Culturally competent methods
to promote organ donation rates among African-Americans using venues
of the Bureau of Motor Vehicles. Transplant Proc. 2008;40:1001–1004.
28. Merion RM, Vinokur AD, Couper MP, et al. Internet-based intervention to
promote organ donor registry participation and family notification. Trans-
plantation. 2003;75:1175–1179.
29. Meier D, Schulz KH, Kuhlencordt R, et al. Effects of an educational seg-
ment concerning organ donation and transplantation. Transplant Proc.
2000;32:62–63.
30. Thornton JD, Alejandro-Rodriguez M, León JB, et al. Effect of an iPod
video intervention on consent to donate organs: a randomized trial. Ann
Intern Med. 2012;156:483–490.
31. McDonald DD, Ferreri R, Jin C, et al. Willingness to communicate organ
donation intention. Public Health Nurs. 2007;24:151–159.
32. Feeley TH, Moon S-i. A meta-analytic review of communication cam-
paigns to promote organ donation. Commun Rep. 2009;22:63–73.
33. Li AH, Rosenblum AM, Nevis IF, et al. Adolescent classroom education on
knowledge and attitudes about deceased organ donation: a systematic
review. Pediatr Transplant. 2013;17:119–128.
34. Deedat S, Kenten C, Morgan M. What are effective approaches to in-
creasing rates of organ donor registration among ethnic minority popula-
tions: a systematic review. BMJ Open. 2013;3:e003453.