Sample characteristics.
The survey collected 1518 complete responses: 1141 (75.2%) from NDG and 377 (24.8%) from DG. The sociodemographic characteristics of the two groups are presented in Table 1. Most of the respondents in both groups were female (75.0% and 76.1%, in NDG and DG, respectively), although the average age was higher in NDG (38.15 vs 36.02, p=0.01), and over half of the sample in both groups was not married (51.3% in NDG; 56.8% in DG). In the DG, respondents lived predominantly in northern Italy (56.8%), while in the NDG, most lived in the South and Islands (51.7%), with a significant difference in distribution between the groups (p<0.001). Regarding education level, most respondents had a university degree (54.2% in NDG; 57.3% in DG) and were employed mainly in the public sector (32.3% in NDG; 25.2% in DG) or in the private sector (23.5% in NDG; 34.0% in DG). Of note, healthcare professionals accounted for 49.2% and for 49.3% of NDG and DG, respectively. In response to the question about religion, the majority in both groups answered that they were Christian (72.3% in NDG; 64.2% in DG), followed by atheists (24.1% in NDG; 28.9% in DG), with a significant difference in distribution (p=0.02), as reported in Table 1.
Table 1. Respondents’ sociodemographic characteristics
Please put Table 1 here.
Due to the size of the table, questionnaire items, response frequencies and percentages, and comparisons among groups are represented in Tables 2 and 3.
General knowledge of donation process
Compared with DG, NDG respondents knew less about the health problems that generate the need for compatible donors in Italy (75.6% in DG vs 48.3% in NDG, p<0.0001); NDG also knew significantly less frequently about the existence of the IBMDR (85.2% vs 96.0%, p<0.0001) and about the activities of volunteer associations promoting donation (90.2% vs 51.7%, p<0.0001). NDG respondents also knew less about CB donation (87.8% vs 83.4%, p=0.03) and about the activities of CB banks (67.1% vs 57.4%, p<0.0001). Interestingly, 36.5% of NDG respondents were found to be more willing to donate if they could choose the recipient. Table 1 shows other significant differences between the groups.
As reported by the DG, the main reasons leading to IBMDR enrollment were the desire to do something for others (82.2%), the awareness raised by volunteer associations (38.5%), the awareness raised by hospital facilities such as transfusion services (26.8%), having a relative/friend with cancer (19.9%), and other reasons detailed in figure 1. Figure 2 illustrates the main reasons for not enrolling in IBMDR as reported by NDG; major findings include the lack of knowledge (36.0%) and low level of concern about the problem (36.0%).
Fig. 1 Reasons for enrolling in the IBMDR
Please put figure 1 here.
Fig. 2 Reasons for not enrolling in the IBMDR
Please put figure 2 here.
Knowledge of stem cell role and function
DG demonstrated greater knowledge than did NDG of SC characteristics such as duplication and self-renewal (79.3% vs. 66.3%, p<0.0001) and differentiation and maturation activities (72.7% vs. 59.6%, p<0.0001). This trend was confirmed responding to more specific questions on SC functions, including red cells (75.3% vs 62.5% of correct answers, p<0.001), white cells (79.4% vs 59.6%, p<0.001) and platelet production (71.1% vs 60.0%; p<0.001). DG had greater knowledge of where the SCs can be harvested for or not, thus SC can be collected from the BM for 93.1% of respondents in DG versus 84.5% in NDG (p<0.001). Excluding that on UCB donation (p=0.06), significant differences with the same direction were found for all questions of this section. Look at table 2 for more details.
Knowledge of stem cell collection methods
NDG had limited familiarity with collection sources compared to DG. For instance, NDG demonstrated less knowledge of SC harvesting methods such as BM explant through repeated punctures of the iliac crests (67.7% vs. 86.5%, p<0.0001), stem cell harvesting through apheresis (62.1% vs. 83.8%, p<0.0001), and collection from the umbilical cord vessels (61.3% vs. 70.5%, p=0.004). Other differences between the groups are presented in Table 2, items 4.0–4.5.
Table 2 Respondents’ knowledge of stem cell donation.
Please put Table 2 here
Respondents’ beliefs about stem cell donation
Awareness of the safety of SC collection methods (PBSC collection, BM harvesting, and CB collection) was greater among DG than among NDG; 92.3% of DG and 85.5% of NDG (p<0.0001) believed that CB collection is a safe procedure for both the mother and the baby, and 89.1% of DG and 77.2% of NDG (p<0.0001) believed that the PBSC collection procedure was safe. Both groups had a little more fear of BM harvesting as a procedure (Table 3. Items 4.0–4.5).
Respondents’ feelings about stem cell donation
The emotional levels experienced when thinking about donation were explored by asking all respondents to rate their feelings using a 4-point Likert scale (not at all, a little, moderately, a lot) when considering a set of 22 human emotions applicable to the donation setting.
The results revealed differences between the 2 groups: DG experienced less fear (p=0.02), hesitation (p<0.001), feeling of fragility (p=0.01), uncertainty (p<0.001), vulnerability (p=0.04), refusal (p<0.001), and perplexity (p<0.001) and greater exaltation (p<0.001), curiosity (p=0.007), enthusiasm (p<0.001), and sense of solidarity (p<0.001). Table 3 reports a complete list of the results.
Opinions and values
This section of the questionnaire used a 5-point Likert scale for each question, with the extremes being “Extremely Disagree” and “Strongly Agree”. No difference between the groups was found regarding their agreement that SC should be used in clinical trials and research (p=0.14). The groups had similar opinions on the need to have more information on donation facilities (p=0.08) and on the usefulness of receiving information on transplant outcome after donation (p=0.32). Interestingly, the whole sample agreed that CB should be collected by default (81.4% of DG and 79.0% of NDG, p=0.37). A high variability in opinions was found regarding whether the identities of the donor and recipient should be shared, with no differences between groups (p=0.79). Significant differences in the distribution of answers between the groups were seen. A high percent of respondents in both groups agreed that HSCT is a lifesaving treatment (92.6% of DG and 91.9% of NDG, p<0.001), and that SC donation should be considered by everyone who is eligible (92.3% in DG and 88.3% in NDG, p<0.001). According to 55.7% of DG and 45.2% of NDG (p=0.04), IBMDR enrollment should be mandatory for everyone who is eligible.
Information and education needs
This section of the questionnaire investigated the willingness of participants to receive more information. In NDG, 66% of the sample did not consider themselves informed enough about donation, compared 33.1% of DG (p<0.001). Interestingly, both groups reported they were interested in receiving further information and education (82.2% and 86.4% of DG and NDG, respectively), with significantly more in NDG (p=0.04). However, even with high percentages in both groups (82.2% of DG and 75.5% of NDG), NDG appeared to be less interested in participating in awareness initiatives on SC donation.
Table 3. Beliefs, feelings, values, and education needs
Please put Table 3 here
Probability of becoming an SC Donor
A complex network of reasons and attitudes influencing the decision to enroll in the IBMDR emerged; key findings were the lack of information and feelings such as fear and lack of trust in the donation process. In contrast, respondents enrolled in IBMDR expressed their deep willingness to contribute to the wellbeing of others, with various factors supporting their choice. However, more and better knowledge appeared to be positively associated with the decision of whether to donate. The questionnaire findings were then used for ML of 6 classifiers to predict the probability of enrolling in the IBMDR; the result obtained was an average accuracy (AUC) of 76.7% (±5.4). Figure 3 illustrates the ROC curves generated by each classifier.
Fig. 3 ROC curves describing performance of each classifier.
Please put figure 3 here.