BLOOD DONORS AND BLOOD COLLECTION Determinants of return behavior: a comparison of current and lapsed donors Marc Germain, Simone A. Glynn, George B. Schreiber, Stéphanie Gélinas, Melissa King, Mike Jones, James Bethel, and Yongling Tu
BACKGROUND: There is a need to identify factors explaining why some people stop donating blood. STUDY DESIGN AND METHODS: A random mail survey of first-time (FT) and repeat (RPT) current (donating within 6 months before survey) and lapsed (donating >2 years prior) donors was conducted. The self-administered questionnaire included questions on personal, social, and behavioral characteristics. RESULTS: Among 1280 current and 1672 lapsed donors with valid addresses, the participation rate was 66.8 and 39.2 percent, respectively. In FT donors, the odds of lapsing increased with education (odds ratio [OR], 2.18; 95% confidence interval [CI], 1.34-3.55 for college or higher vs. Grade 12 or less education). Lapsed FT donors were more often asked to donate (OR, 1.89; 95% CI, 1.32-2.70) and had less interest in incentives (p < 0.001) than current FT donors. In RPT donors, lapsed status was associated with being younger (p < 0.001) and female (OR, 1.19; 95% CI, 1.00-1.42). Lapsed status was inversely associated with satisfaction with the last donation experience in both FT (p = 0.043) and RPT (p < 0.001) donors. Lapsed and current donors did not differ in perceived need for blood, personal transfusion experience, or mean reported altruistic behavior score. CONCLUSION: A positive donation experience appears to be a major determinant of donor return behavior. Lapsed donors do not appear, on average, to engage in fewer altruistic behaviors than currently active donors. Retention marketing strategies that appeal solely to altruistic values need to be further evaluated for their effectiveness.
1862 TRANSFUSION
Volume 47, October 2007
T
he task of recruiting blood donors remains a major challenge for blood collection agencies. In North America, only 3 to 5 percent of the total adult population give blood each year, and new donors contribute a relatively small portion of blood collections. For example, annual statistics at Héma-Québec for 2004 to 2005 indicate that only 9.6 percent of all donations came from first-time (FT) donors. In addition to their significant quantitative contribution, habitual donors also have lower rates of infectious disease markers. Thus, the availability and the safety of blood largely depend on the continuing contribution of repeat (RPT) donors. A significant number of donors stop donating, however, after one or more successful donations. To develop effective retention strategies, it is therefore crucial to understand factors that deter people from becoming habitual donors. Previous studies of blood donation behavior have focused mainly on the comparison of donors with nondonors. For example, London and Hemphill1 surveyed a sample of donors, compared them with census data and concluded that blood donors were more likely to be male, younger, and of higher socioeconomic status compared to the general population. In another study, Bettinghaus and Milkovich2 compared various personal and sociodemographic characteristics and also concluded that
ABBREVIATIONS: FT = first-time (donors); PAR = populationattributable risk; RPT = repeat (donors). From Héma-Québec, Montréal, Québec, Canada; and Westat, Rockville, Maryland. Address correspondence to: Marc Germain, MD, PhD, FRCPC, Scientific Advisor, Microbiology & Epidemiology, HémaQuébec, 1009 du Vallon Road, Ste-Foy, Quebec, Canada G1V 5C3; e-mail:
[email protected]. This work was supported by Héma-Québec and by NHLBI Contract N01-HB-97082. Received for publication January 17, 2007; revision received April 2, 2007, and accepted April 3, 2007. doi: 10.1111/j.1537-2995.2007.01409.x TRANSFUSION 2007;47:1862-1870.
DETERMINANTS OF DONATION RETURN
socioeconomically advancing young adults were more likely to donate. Through an extensive review of 60 published studies on donation behavior, Oswalt3 listed several factors that had been identified as motivations to donate, such as altruism, personal credit, social pressure, and material incentives. A similar review published some years later by Boe and Ponder4 listed essentially the same factors. Some recent studies have used more refined methods but were also based on comparisons of donors and nondonors.5,6 Altruism has been singled out as a major motivational factor for donating blood,4,7,8 although some have questioned the true altruistic nature of such behavior. Altruism is a poorly defined characteristic and these studies did not include objective measurements of altruistic behaviors. Few studies have attempted to identify differences between donors who are currently active and donors who stopped donating. By reviewing the donation histories of 179,409 community whole-blood donors, Schreiber and coworkers9 found that first-year donation frequency, older age, a higher education level, and white non-Hispanic race or ethnicity were strong predictors of long-term donor return. Their study, however, did not explore the role of motivational factors in the establishment of long-term donation behavior. In particular, the study did not compare altruistic behavior between currently active donors and donors who stopped donating. In collaboration with the National Heart, Lung, and Blood Institute’s Retrovirus Epidemiology Donor Study (REDS), we surveyed donors in the province of Quebec, Canada, on various personal characteristics and sociobehavioral traits that may be related to their donation behavior. We compared currently active donors with donors who stopped donating without any obvious cause for deferral. In particular, we attempted to measure and compare the degree of altruistic behavior between these two groups of donors. Based on the accepted theory of blood donation behavior, we hypothesized that donors who stop donating are less altruistic compared to people who remain active donors. We also explored other factors such as donor satisfaction with the donation process, their personal experience with transfusion and their perceived importance of donor incentives. Héma-Québec is a not-for-profit agency that has the exclusive responsibility for collecting and distributing blood and blood products in the province of Quebec. More than 250,000 units of blood are collected annually. With a total population of 7.5 million, the majority of Quebeckers are French-speaking (82%) and of Caucasian descent (>90%). Health care is almost entirely funded by the government but social and economic characteristics are otherwise comparable to the United States. Blood is collected mainly through mobile clinics that reach most populated areas. The procedures, standards, and regula-
tory requirements that apply to Héma-Québec are essentially the same as elsewhere in North America.
MATERIALS AND METHODS Study population Since 1987, all information concerning Héma-Québec donors and their donation history is contained in a computerized database (Progesa, Mak Systems). The database includes personal information on date of birth, sex, address of residence, dates of all previous donations, type of donations, previous and current deferrals, and screening test results. For the purpose of this study, all relevant data were extracted from the Progesa database and imported into computer software (SAS, Version 8.02, SAS Institute, Cary, NC) for the purpose of defining the study population, selecting study subjects, and performing the demographic analyses. The study population included all donors who gave one or more units of whole blood in the past, as recorded in the donor database. Autologous donations were not counted and donors were also excluded if they had a permanent or temporary deferral at the time of selection of the study sample. Four groups of donors were defined on the basis of their donation history: “Current” donors were those who had given one or more allogeneic donations during the 6 months preceding the survey. “Lapsed” donors were those who did not donate during the 2 years preceding the survey but who had donated once or more in the previous years. Each group of current and lapsed donors was further stratified into FT and RPT donors. For the purpose of defining FT donor status, lifetime donation history was considered, as recorded in Progesa. Lifetime donation history was not only based on Progesa-recorded donations but also on all other donations reported by the donor, including those made before the starting date of the computerized database or donations made at another transfusion agency. Within each of the four groups defined by their current or lapsed and FT or RPT status, a study sample was randomly selected. Based on previous experience with this type of survey, we expected differential participation rates between lapsed and current donors and study sample sizes were defined accordingly. We targeted an approximate number of 500 respondents within each group, which would provide adequate statistical power to address the main study question. Sample selection and survey mailings started in January 2003 for current donors and in April 2003 for lapsed donors. Within a few days after selecting the study sample, a notice was mailed to the donors simply mentioning that they were selected to participate in a survey. One week later, a self-administered questionnaire was sent to each study subject along with a letter of introduction, instructions for completing the survey, and a prepaid return Volume 47, October 2007
TRANSFUSION 1863
GERMAIN ET AL.
envelope. After 3 weeks, a reminder letter was sent to those who did not return the completed survey or who did not indicate their refusal to participate. A final reminder, including a second full study package, was sent 2 weeks later. This mailing schedule was done according to the method described by Dillman.10 The protocol and the questionnaire were reviewed and approved by the HémaQuébec’s Ethics Review Board.
the scale. This composite score was then categorized into two or more levels describing the strength of that item. For the “altruistic behavior” variable, we used 13 of the most relevant items from the scale proposed by Rushton and colleagues11 as listed in Table 1. One of the original questions on the Rushton scale on previous blood donation(s) was omitted from our survey since by definition, all survey participants had previously donated.
Survey instrument
Statistical analyses
For some basic demographic characteristics contained in our donor database, we were able to compare study participants to nonparticipants and to the entire Quebec donor population. To compensate for variable probabilities of selection and differential response rates, weights were calculated for each observation, according to previously described methods.12 Briefly, a base weight was calculated as the reciprocal of the probability of including the donor in the study. These base weights were first adjusted for nonresponse with nonresponse adjustment classes defined by age and region. The final nonresponse-adjusted weights were adjusted for differences due to variations in the original sample selection. Unless specified otherwise, the weighted observations were used in all TABLE 1. Potential motivation factors for blood donation and specific measures of association between variquestions that were included in the survey instrument ables of interests and current or lapsed Personal transfusion experience status. Has anyone close to you ever received a blood transfusion? The main comparisons of interest Have you ever received a blood transfusion? Perceived need for blood were between current and lapsed Do you feel that in Québec, there is always enough blood available for those who need donors. To describe the association it? between certain variables and current Having been asked to donate Before your last donation, did any person or organization ask you to donate blood? or lapsed status, we calculated weighted Satisfaction with last donation odds ratios (ORs) and their confidence Were you glad that you donated blood? intervals (CIs). A Satterthwaite-adjusted Were you nervous or scared about your last blood donation? How would you rate the way you were treated by staff? statistic, based on the Rao-Scott How would you rate the skill of the staff? method, was used for chi-square tests,13 How would you rate how you physically felt during or after your donation? which were implemented with a statisHow would you rate the level of privacy during the screening process? How would you rate your overall donation experience? tical package (WesVarPC, Westat, RockPrevious deferral ville, MD). For continuous variables, Have you ever been told that you could not give blood for a certain period of time, but means were also calculated and comcould return in the future? Importance of incentives pared with t tests (nonweighted obserThinking about your last donation, how important was each factor in your decision to vations). Because sampling was done donate: Receiving a gift? Getting the result of a health screen? Receiving a reward, according to FT or RPT donor status, all such as a pin? Wanting tickets to an event? Wanting time off work? Wanting to reach a target number of donations? analyses were stratified according to Ease of access to a donation site this variable. Logistic regression was Is there a convenient place for you to donate blood now? used to derive weighted ORs adjusted Altruistic behavior How often have you: Given directions to a stranger? Given money to charity? Given for potential confounders. Statistical money to a stranger who needed it or asked for it? Donated goods or clothes to a significance was defined at the charity? Done volunteer work for a charity? Helped carry a stranger’s belongings? 95 percent confidence level. We also Delayed an elevator and held the door for a stranger? Pointed out a clerk’s error in undercharging me for an item? Bought “charity” holiday cards? Allowed someone to go evaluated the relative contribution of ahead in a line? Let a neighbor borrow an item of some value to you? Volunteered for specific factors found to be indepenlooking after a neighbor’s pet or children without being paid? Offered to help a dently associated with lapsed behavior handicap or elderly across a street? by estimating the percent population-
The questionnaire queried the donor about personal and sociodemographic characteristics that were not included in the donor database, such as level of education and ethnicity. The rest of the survey included various questions about items that could be related to motivations or deterrents for blood donation. Table 1 shows selected items of interest that were the main focus of this analysis, along with the specific questions that were included under each item. Depending on the question, donors were asked to give a “yes” or “no” answer or to rate their level of agreement on a five-point scale. For items with multiple questions, a composite score was obtained by adding the number of “yes” answers and/or the number of points on
1864 TRANSFUSION
Volume 47, October 2007
DETERMINANTS OF DONATION RETURN
attributable risk (PAR). The percent PAR of each factor was calculated as
(risk factor prevalence) × (OR −1)/OR, where the OR is used to estimate the relative risk and where the risk factor prevalence is calculated from the current donor population.14
RESULTS A total of 1300 and 2000 study packages were mailed to current and lapsed donors, respectively. Some mailings were returned to the study center because of an invalid address: 20 (1.5%) among current donors and 328 (16.4%) among lapsed donors. Among current donors who were assumed to have received the study package, 855 returned the completed questionnaire for a participation rate of 66.8 percent. Of those, 334 were FT donors and 521 were RPT donors. The participation rate was 39.2 percent (656 respondents) among lapsed donors, with 256 FT and 400 RPT donors. Based on the demographic data for the whole study sample, current donors who participated in the survey were older (p < 0.0001), had a larger number of previous donations (p < 0.0001), and lived more often outside the Montreal and Quebec metropolitan areas (p = 0.05) but were similar in terms of their sex distribution (p = 0.85), compared to current donors who did not return the survey. Among lapsed donors, participants were older (p < 0.0001), were more likely to be female (p < 0.001), and
had donated more (p < 0.0001) but were similar for region of residence, compared to nonparticipants. Table 2 shows the comparison of selected characteristics between current and lapsed donors, within each stratum of FT and RPT donors. Among FT donors, the weighted analysis showed that lapsed donors were more likely to be Caucasian (chi-square p < 0.001) and more highly educated (p = 0.001) than current donors. The mean age (years) of current (32.7; standard deviation [SD] = 11.9) and lapsed (32.5; SD = 12.0) FT donors was similar (t test p = 0.79). FT donors also showed no association between lapsed status and sex (p = 0.25), language (p = 0.57), and region of residence (p = 0.13). Among RPT donors, lapsed status was strongly associated with female sex (p < 0.001) and English language (p = 0.001) but not with education (p = 0.15). The mean age was also significantly lower in lapsed donors (39.1; SD = 12.3) than in current donors (46.0; SD = 10.8), with a p value of less than 0.0001. When the data were analyzed without weighing the observations, the observed associations did not change meaningfully (data not shown), except for an even stronger association between age and lapsed status among RPT donors. Table 3 compares motivational factors between current and lapsed FT and RPT donors. Among FT donors, factors that were not significantly different between lapsed and current donors included: a personal experience with transfusion (p = 0.61), the perceived need for blood (p = 0.93), a previous deferral (p = 0.74), the ease of access to the clinic (p = 0.84), and the mean altruistic
TABLE 2. Comparison of sociodemographic characteristics between current and lapsed donors within each stratum of FT and RPT donors and OR of being a lapsed donor* Characteristic Age (years) 18-29 30-39 40-49 ⱖ50
Lapsed 117 58 53 28
Sex Female Male
(45.7) (22.7) (20.7) (10.9)
FT donors (n = 590) Current Weighted OR (95% CI) 162 69 70 33
Lapsed
(48.5) (20.7) (21.0) (9.9)
1.10 (0.93-1.31) 1.01 (0.85-1.19) 0.93 (0.72-1.21) Reference
108 83 122 87
162 (63.3) 94 (36.7)
196 (58.7) 138 (41.3)
1.14 (0.91-1.42) Reference
Language English French
21 (8.2) 235 (91.8)
31 (9.3) 303 (90.7)
Ethnicity Caucasian Other
252 (99.2) 2 (0.8)
Region of residence Montreal Quebec Other Education College or higher High school Grade 12 or less
(27.0) (20.8) (30.5) (21.8)
RPT donors (n = 921) Current Weighted OR (95% CI) 42 86 187 206
(8.1) (16.5) (35.9) (39.5)
2.74 (2.35-3.19) 1.54 (1.33-1.79) 1.10 (0.95-1.27) Reference
223 (55.8) 177 (44.3)
191 (36.7) 330 (63.3)
1.51 (1.39-1.64) Reference
0.85 (0.49-1.50) Reference
27 (6.8) 373 (93.3)
11 (2.1) 510 (97.9)
3.11 (1.54-6.27) Reference
312 (93.7) 21 (6.3)
8.30 (1.08-63.8) Reference
388 (98.2) 7 (1.8)
510 (97.9) 11 (2.1)
1.38 (0.49-3.88) Reference
104 (40.6) 36 (14.1) 116 (45.3)
138 (41.3) 58 (17.4) 138 (41.3)
1.06 (0.84-1.35) 0.73 (0.51-1.06) Reference
170 (42.5) 72 (18.0) 158 (39.5)
194 (37.2) 91 (17.5) 236 (45.3)
1.30 (1.07-1.58) 1.18 (0.83-1.69) Reference
87 (34.3) 89 (35.0) 78 (30.7)
64 (19.3) 130 (39.2) 138 (41.6)
2.27 (1.48-3.49) 1.21 (0.78-1.90) Reference
100 (25.2) 131 (32.9) 167 (42.0)
113 (21.7) 146 (28.0) 262 (50.3)
1.33 (0.97-1.82) 1.25 (0.92-1.70) Reference
* Data are reported as number (%).
Volume 47, October 2007
TRANSFUSION 1865
GERMAIN ET AL.
TABLE 3. OR of being a lapsed donor by motivational factors within each stratum of FT and RPT donors* FT donors (n = 590) Current Weighted OR (95% CI) 206 (61.7) 128 (38.3)
0.92 (0.65-1.30) Reference
239 (59.8) 161 (40.2)
276 (53.0) 245 (47.0)
1.31 (0.99-1.74) Reference
Perceived need for blood Low 28 (10.9) High 228 (89.1)
38 (11.4) 296 (88.6)
1.02 (0.61-1.73) Reference
81 (20.3) 319 (79.7)
105 (20.2) 416 (79.8)
1.04 (0.74-1.46) Reference
Donor was asked to donate Yes 147 (57.4) No 109 (42.6)
151 (45.2) 183 (54.8)
1.71 (1.24-2.36) Reference
293 (73.3) 107 (26.8)
367 (70.4) 154 (29.6)
1.13 (0.83-1.54) Reference
Satisfaction with last donation Low 91 (35.6) Medium 64 (25.0) High 101 (39.5)
87 (26.1) 102 (30.5) 145 (43.4)
1.55 (1.06-2.27) 0.95 (0.63-1.42) Reference
70 (17.5) 82 (20.5) 248 (62.0)
39 (7.5) 70 (13.4) 412 (79.1)
2.75 (1.85-4.09) 1.67 (1.17-2.36) Reference
Previous deferral Yes 34 (13.3) No 222 (86.7)
49 (14.7) 285 (85.3)
0.91 (0.51-1.62) Reference
103 (25.8) 297 (74.3)
88 (16.9) 433 (83.1)
1.52 (1.10-2.11) Reference
Importance of incentives Low 145 (56.6) Medium 56 (21.9) High 55 (21.5)
131 (39.2) 87 (26.1) 116 (34.7)
2.31 (1.51-3.55) 1.37 (0.86-2.19) Reference
191 (47.8) 105 (26.3) 104 (26.0)
230 (44.2) 131 (25.1) 160 (30.7)
1.25 (0.90-1.73) 1.12 (0.80-1.56) Reference
Easy access to donation clinic No 111 (43.4) Yes 145 (56.6)
143 (42.8) 191 (57.2)
1.03 (0.76-1.40) Reference
146 (36.5) 254 (63.5)
198 (38.0) 323 (62.0)
0.90 (0.65-1.24) Reference
Altruism score Low Medium High
54 (16.2) 215 (64.4) 65 (19.5)
1.41 (0.85-2.32) 1.32 (0.79-2.19) Reference
61 (15.4) 266 (67.2) 69 (17.4)
85 (16.3) 343 (65.8) 93 (17.9)
0.92 (0.60-1.42) 1.03 (0.75-1.42) Reference
46 (18.0) 170 (66.7) 39 (15.3)
Lapsed
RPT donors (n = 921) Current Weighted OR (95% CI)
Factor Lapsed Personal transfusion experience No 150 (58.6) Yes 106 (41.4)
* Data are reported as number (%).
behavior score (p = 0.38). FT donors who lapsed were more likely to report having been asked to donate before their donation (p = 0.001). They also reported a lower level of satisfaction with their donation experience (p = 0.043). Finally, FT lapsed donors attached a lower level of importance to receiving donation incentives than FT current donors (p < 0.001). Among RPT donors, there was a borderline nonsignificant association between the lack of personal transfusion experience and lapsed status (p = 0.06). There was, however, a strong inverse association between the level of satisfaction with the last donation experience and lapsed status (p < 0.001). A previous temporary deferral was also strongly associated with lapsed status (p = 0.009). There was no difference between lapsed and current donors with regard to the perceived need for blood (p = 0.84), the fact of having been asked to donate (p = 0.44), the importance of incentives (p = 0.36), the ease of access to the clinic (p = 0.50), or the altruistic behavior score (p = 0.80). In both FT and RPT donors, a lower level of satisfaction was associated with lapsed status. This association, however, was significantly stronger among RPT donors (p = 0.02 for the Breslow-Day test for homogeneity of the ORs). We conducted weighted logistic regression to evaluate if the observed associations were confounded by other variables. When adjusting for age, the association between 1866 TRANSFUSION
Volume 47, October 2007
higher education and lapsed status remained significant among FT donors (and remained nonsignificant in RPT donors). Adjusting for age further decreased the magnitude of the association between previous personal experience with transfusion and lapsed status among RPT donors (age-adjusted OR, 1.22; 95% CI, 0.89-1.67). Table 4 shows the distribution of individual altruistic behavior items reported by lapsed and current donors, respectively, stratified according to FT or RPT status (unweighted data). As can be seen, the majority of these behaviors were reported with equal frequency by lapsed and current donors. Some specific behaviors were reported significantly less often by lapsed donors: giving directions or money to a stranger, buying charity cards, or lending an item to a neighbor. These findings, however, were not consistent between FT and RPT donors. Table 5 presents the individual components of the composite donor satisfaction variable, according to lapsed or current donor status (unweighted data). For most components, the reported level of satisfaction was significantly lower among lapsed donors, although this trend was less pronounced among FT donors. For FT and RPT donors separately, Table 6 shows weighted logistic regression models that include all sociodemographic variables and motivational factors that were identified as being significantly associated with
DETERMINANTS OF DONATION RETURN
TABLE 4. Comparison of reported altruistic behaviors between lapsed and current donors* Self-reported score† Reported behavior Giving directions to a stranger Giving money to charity Giving money to a stranger Donating goods or clothes Volunteering for a charity Carrying a stranger’s belongings Delaying an elevator Pointing out a clerk’s error Buying “charity” holiday cards Allowing someone to go ahead in a line Lending an item to a neighbor Looking after neighbor’s pet or children Helping a handicap or elderly across a street
Lapsed donors 3.57 (0.89) 3.34 (0.94) 2.19 (1.07) 3.89 (1.00) 2.07 (1.21) 2.41 (1.16) 3.62 (1.04) 3.06 (1.12) 2.14 (1.14) 3.21 (0.92) 2.77 (1.16) 2.49 (1.21) 2.61 (1.19)
FT donors Current donors 3.75 (0.85) 3.26 (1.06) 2.43 (1.09) 3.78 (1.12) 2.23 (1.28) 2.45 (1.04) 3.76 (0.99) 3.02 (1.13) 2.43 (1.20) 3.38 (0.89) 2.96 (1.23) 2.58 (1.28) 2.63 (1.26)
p Value‡ 0.02 0.33 0.009 0.21 0.14 0.66 0.09 0.66 0.004 0.02 0.06 0.41 0.85
Lapsed donors 3.62 (0.93) 3.53 (0.96) 2.22 (1.09) 3.90 (0.90) 2.14 (1.34) 2.37 (1.11) 3.65 (1.04) 2.99 (1.04) 2.43 (1.19) 3.30 (0.91) 2.81 (1.15) 2.48 (1.30) 2.70 (1.18)
RPT donors Current donors 3.62 (0.87) 3.59 (0.94) 2.25 (1.07) 3.85 (0.96) 2.08 (1.27) 2.43 (1.15) 3.54 (1.01) 2.98 (0.97) 2.52 (1.13) 3.26 (0.93) 3.09 (1.11) 2.60 (1.28) 2.76 (1.11)
p Value‡ 0.94 0.34 0.70 0.42 0.50 0.40 0.11 0.90 0.23 0.60 0.0001 0.16 0.44
* Data are reported as mean (SD). † 1 = never, 2 = once, 3 = more than once, 4 = often, 5 = very often. ‡ t test.
TABLE 5. Comparison of donor satisfaction variables between lapsed and current donors* Reported rate† FT donors Lapsed Current Satisfaction criterion donors donors Reported rate† How well the donor was treated by staff 4.56 (0.66) 4.67 (0.60) Skill of the staff 4.62 (0.60) 4.67 (0.67) How the donor felt physically during or after donation 3.83 (1.20) 4.08 (1.14) Level of privacy during the screening process 4.65 (0.56) 4.75 (0.48) Overall donation experience 4.17 (0.86) 4.42 (0.85) Percentage answering “yes” Were you glad you gave blood? 93.9 93.6 Were you nervous or scared about your last blood donation? 46.9 47.4 * † ‡ ||
p Value
Lapsed donors
0.03‡ 0.34‡ 0.01‡ 0.03‡ 0.0005‡
4.50 4.60 4.30 4.70 4.34
0.96|| 0.84||
92.5 21.8
RPT donors Current donors
(0.70) (0.69) (0.92) (0.51) (0.88)
4.65 4.74 4.64 4.78 4.61 98.6 9.8
(0.62) (0.51) (0.62) (0.48) (0.61)
p Value 0.0007‡ 0.0007‡
