Study characteristics

The 18 included studies were conducted between 1993 and 2012. These either evaluated the effects of an ethnically targeted mass media campaign or community-based education. With the exception of one UK study, all were undertaken with ethnic minority populations in the USA, mainly African Americans or Hispanics.12–16 Whereas most interventions adopted a pre-evaluation and postevaluation design (table 1), studies published since 2009 mainly employed a cluster randomised design.12–14 ,17 ,18 Outcomes assessed were verified registration, self-reported registration status or knowledge about organ donation (see table 1).

Educational interventions

Eleven articles reported educational interventions. Recruitment was mainly through or conducted at places of worship and/or educational establishments, and delivered by a range of individuals including recipient families, trained lay individuals or organ transplant and procurement staff. Six studies were designed for a specific ethnic group, with the remainder aimed at a multiethnic audience (see table 2)

Verified registration

Three cluster randomised trials reported an outcome of verified enrolment on a donor register.12 ,13 ,17 Two of these studies trained individuals embedded within community settings to deliver the intervention,12 ,13 with both these interventions having a strong interpersonal element and being delivered in familiar settings of a local church or hairdresser. Peer educators in Andrews’ study13 mostly had a medical background and were involved in a Church Health Committee. They undertook a 4 h training package that comprised general information about organ donation and tips on integrating the topic into church discussions. In Resnicow's study,12 hair stylists were trained over 2 days in practical techniques to facilitate communication (motivational interviewing) and general organ donation information. There were wide variations in the duration of these study interventions: participants in the Church study were exposed over a period of 12 months, while interventions with a hair stylist ran for 1 session/month for 4 months. In contrast, the third intervention in this category did not have a strong interpersonal element, and trialled the use of a 5 min educational film to address common barriers to organ donation that was targeted at those attending the state vehicle licensing office.17 Across all three interventions, the exposed group were significantly more likely to register than the controls, although the effects of the video intervention among vehicle licensing office attendees were significant for White and African Americans but not Hispanic participants,17 possibly reflecting their small number (n=−28).

Further, three community educational studies based on a follow-up or cross-sectional design reported strategies to increase organ donor registration. Feeley et al19 reported a 1.6% increase in new organ donor registrations following campus-led peer education among students from six university campuses in the USA that were selected as having a high percentage of minority student enrolment. Similarly, a study in Arizona focused on Hispanics attending a large community event found that advertising and offering an immediate opportunity to register produced significantly greater registration rates compared with the use of a generic slogan and offering general information about organ donation (86% vs 54%).20 A further UK community-based educational intervention study focused on attendees at a range of venues and community events selected to maximise engagement with Black and minority (BME) communities (A Warrens, personal communication, 2013). The authors report that following discussion with a trained BME Peer Educator and the availability of facilities for registering, 9% of those not already on the Organ Donor Register joined on the day. The intervention was also shown to be most effective among those who had previously considered signing up but who did not know how to go about it. These community educational studies thus identify important influences on registration as both the individual's prior readiness and access to registration (table 3).

Changing knowledge

Previous research has shown that knowledge about organ donation is associated with attitudes, intention to donation and willingness to discuss organ donation intentions with family.21–23 Among the included studies, knowledge was assessed by asking participants to indicate whether a range of statements regarding organ donation were true or false. These studies conceived knowledge as comprising five facets (1) general knowledge about donation; (2) knowledge of the relevance of donation to the ethnic minority population; (3) knowledge about the procurement and allocation of organs; (4) religious and cultural knowledge about organ donation and (5) knowledge about the financial costs of organ donation. Presentations to participants were delivered by ethnic minority transplant surgeons and donor/recipient families. Interventions tended to be one-off and commonly lasted 1 h, with immediate follow-up questionnaires administered to participants.

Two studies tested the relationship between knowledge about organ donation and willingness to become an organ donor. These studies found that significant improvements in knowledge increased the willingness among participants to state that they intended to register as a donor. Among Hispanic participants, knowledge relating to the fair allocation of organs was a significant independent predictor of willingness24 (see table 4), while among a multiethnic sample of high school students improvements in composite knowledge scores significantly predicted a willingness to donate organs in the intervention group.18

The remaining two studies assessed a range of knowledge items before and after the delivery of an educational intervention. In a multiethnic sample of adults, a significant increase in knowledge was achieved for half of all knowledge items (see table 4).25 Of particular note, African Americans who comprised 60% of the sample achieved the greatest change postintervention in relation to trust in doctors when compared to trust scores for other ethnic groups. Overall, postintervention, participants in this study reported being significantly more likely to state that they were willing to donate their own and their family's organs after death (p<0.000).25 Unlike the aforementioned studies which focused on the lay population, Allen and Stillwater's26 holistic interventions addressed knowledge gaps and misinformation among Alaskan Community Health aides and school children. The authors do not report details about individual aspects of knowledge addressed but report a significant increase in knowledge among both students from 58% to 95% correct (p<0.0001) and community health practitioners (p<0.0001).

Table 4 shows the number of items where a significant increase in knowledge was observed post-test. All four interventions reported that it is possible to address knowledge deficits among participants. However, no assessment was made about the extent to which gains in knowledge are maintained and whether increased willingness necessarily results in actual donor registration.

Measured shift towards greater readiness

Two educational interventions were informed by the transtheoretical model (TTM) of behaviour change27 and measured change in an individual's ‘readiness to become an organ donor’.14 ,28 Both studies aimed to measure change across the five TTM stages from precontemplation to maintenance (box 2). In Fahrenwald's study, 56.9% of participants progressed to a more advanced stage post-test and none regressed, while Arriola reported that intervention participants were 1.53 times more likely to be in a later stage of readiness than control participants.

Collaboration with members of the targeted ethnic population was undertaken to ensure the relevance and acceptability of the intervention. For Arriola et al,14 this involved a specific focus on religious objections to organ donation, involvement of the church pastors and the use of a well-known gospel choir to present the intervention video, whereas Fahrenwald et al28 included respected members of Native American tribes and their storytelling traditions to convey donation messages. There were significant differences in participant exposure; participants in Arriola's study were exposed to intervention materials in the home at several time points over a year (eg, through video, pamphlets and greetings cards), whereas Fahrenwald's participants received a short intervention lasting between 30 and 60 min that involved watching a video about organ donation followed by a group discussion.

Although these studies provide evidence of positive movement towards readiness to become an organ donor, sign a donor card or talk to family members about their preferences, whether those at a later stage of readiness actually signed a donor card is less clear. This issue is reflected in the UK community study where only a small proportion of participants who stated an intention to register in the future had done so at follow-up (A Warrens, personal communication, 2013).

Mass media interventions

All media interventions attempted to detect changes in the willingness or intention to join a donor register or sign a donor card.16 ,29–32 However, interventions that solely utilised the mass media reported no significant change in the intention or willingness to register as an organ donor, despite focusing on a single ethnic group and undertaking formative research in the target population (see table 4 below). In contrast, a high level of success was achieved by one study15 which innovatively combined mass media techniques with interpersonal communication. This was conducted at the vehicle licensing office, a setting where individuals are able to join a state donor registry. Harrison et al set out with the clear aim of focusing on members of the population categorised as passive positives (ie, those favourable to organ donation but not yet joined the register). To determine the effect of each component of the intervention on registrations, the campaign occurred in three stages, each lasting 3 months. The interpersonal component of the intervention accounted for the greatest increase in registrants from 444 in the previous year to 5588 post campaign. Overall, the combined effect of the intervention components increased registrations by 700% (table 5).