ORIGINAL ARTICLE Understanding noncompliance with selective donor deferral criteria for high-risk behaviors in Australian blood donors Tarana T.A. Lucky,1 Clive R. Seed,2 Daniel Waller,3 June F. Lee,2 Ann McDonald,1 Handan Wand,1 Stephen Wroth,2 Glen Shuttleworth,3 Anthony J. Keller,2 Joanne Pink,4 and David P. Wilson1
BACKGROUND: Using a predonation screening questionnaire, potential blood donors are screened for medical or behavioral factors associated with an increased risk for transfusion-transmissible infection. After disclosure of these risks, potential donors are deferred from donating. Understanding the degree of failure to disclose full and truthful information (termed noncompliance) is important to determine and minimize residual risk. This study estimates the prevalence of, and likely reasons for, noncompliance among Australian donors with the deferrals for injecting drug use, sex with an injecting drug user, male-to-male sex, sex worker activity or contact, and sex with a partner from a highHIV-prevalence country. STUDY DESIGN AND METHODS: An anonymous, online survey of a nationally representative sample of Australian blood donors was conducted. Prevalence of noncompliance with deferrable risk categories was estimated. Factors associated with noncompliance were determined using unadjusted and adjusted odds ratios. RESULTS: Of 98,044 invited donors, 30,790 donors completed the survey. The estimated prevalence of overall noncompliance (i.e., to at least one screening question) was 1.65% (95% confidence interval CI, 1.51%-1.8%). Noncompliance with individual deferrals ranged from 0.05% (sex work) to 0.54% (sex with an injecting drug user). The prevalences of the disclosed exclusionary risk behaviors were three to 14 times lower than their estimated prevalence in the general population. CONCLUSION: The prevalence of noncompliance is relatively low but our estimate is likely to be a lower bound. The selected high-risk behaviors were substantially less common in blood donors compared to the general population suggesting that self-deferral is effective. Nevertheless, a focus on further minimization should improve the blood safety.
A
fter the emergence of human immunodeficiency virus (HIV)/AIDS in the 1980s, blood services around the world introduced questions about high-risk lifestyle behaviors associated with transfusion-transmissible infections (TTIs) including HIV and hepatitis.1 The importance of deferral measures focusing on preventing donation by “high-risk” individuals was first illustrated in the early 1980s before universal HIV antibody screening when Busch and colleagues2 demonstrated a marked decline in the “per-unit” risk of HIV infection in the United States coinciding with the progressive implementation of donor selection measures. ABBREVIATIONS: ACASI = audio computer-assisted structured interview; AOR = adjusted odds ratio; ASHR = Australian Study of Health and Relationships; DHQ = donor health questionnaire; HHPC = high-HIV-prevalence country; IDU = injecting drug use; MSM = men who have sex with men; STI = sexually transmitted infection; TTI(s) = transfusiontransmissible infection(s). From 1The Kirby Institute, The University of New South Wales, Sydney, New South Wales; 2Australian Red Cross Blood Service, Perth, Western Australia; 3Australian Red Cross Blood Service, Sydney, New South Wales; and 4Australian Red Cross Blood Service, Brisbane, Queensland, Australia. Address reprint requests to: David P. Wilson, The Kirby Institute, The University of New South Wales, Sydney, NSW 2052, Australia; e-mail: [email protected]. The authors acknowledge funding jointly from the Australian Red Cross Blood Service and the Kirby Institute. The Australian Red Cross Blood Service is fully funded by the Australian Government for the provision of blood products and services to the Australian community. The Kirby Institute is funded by the Australian Government Department of Health and Ageing and is affiliated with the Faculty of Medicine at the University of New South Wales. Received for publication September 27, 2013; revision received November 27, 2013, and accepted November 27, 2013. doi: 10.1111/trf.12554 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
LUCKY ET AL.
Importantly, as tests for TTIs are unable to detect donors with very recently acquired infection (known as the “window period”) the donor health questionnaire (DHQ) remains important for maximizing the safety of blood transfusions even where universal testing for a TTI is undertaken.3,4 In Australia, each donor is required to self-complete a comprehensive DHQ every time he or she donates. The DHQ is reviewed at a private and confidential interview with the donor and a legally binding declaration form is signed before donation. The Australian Red Cross Blood Service (Blood Service) is highly reliant on the donor’s complete and truthful answers to all interview questions (i.e., “compliance”). This is particularly important for questions relating to TTI risk behavior given the existence of the testing window period. Should donors in the window period fail to answer truthfully a question that would normally result in their deferral from donation, they may place recipients at risk because a potentially infectious unit of blood that testing will not identify could be transfused. The risk has been substantially reduced as a result of state-of-the-art TTI testing including nucleic acid testing (NAT) for HIV, hepatitis C virus (HCV), and hepatitis B virus in Australia.5 Despite this, the risk is real as evidenced by occasional cases of TTI transmission associated with noncompliance, which continue to occur even where NAT is implemented.6 Comprehensive data on noncompliance with highrisk donor deferral criteria are limited. In Australia, noncompliance among test-positive donors ranges between 12.9 and 24.4%.5 However, before this study no data were available for noncompliance among TTI test–negative donors. Internationally, studies conducted in Canada,7,8 Germany,9 the United Kingdom,10 the United States,11-14 and Hong Kong15 have examined various aspects of risk nondisclosure during predonation assessment. The majority focused on permanent deferral for male-to-male sex with a range of noncompliance of 0.8% to 2.3% among male donors.7,12,15 A few studies assessed other high-risk nonsexual behaviors such as injecting drug use (IDU).8,9,13 However, there is a paucity of data concerning noncompliance with high-risk heterosexual deferrals (e.g., sex workers and their sexual contacts). In Australia, a history of IDU (even one episode) results in permanent deferral while sex with an injecting drug user defers a person temporarily (12 months since last sexual contact).16 Similarly, a 12-month exclusion also applies for men who have sex with men (MSM), females who have sex with MSM, sex worker activity or contact, and sexual contact with a person from a high-HIVprevalence country (HHPC). These policies are reviewed annually and revised where appropriate. The current 12-month exclusion for the previously defined sexual activity–based deferrals became nationally applicable during the late-1990s when the Blood Service was estab2
TRANSFUSION Volume **, ** **
lished as a national service. The latest change to the duration of deferral occurred in 2000 when the current 12-month deferral for MSM was implemented nationwide.17 Notably, the deferral policy for IDU has not been revised since the 1980s, when the original 5-year deferral was changed to permanent deferral. A previous Blood Service study of HIV-positive donors highlighted the importance of compliance with the current 12-month deferral for male-to-male sex on HIV residual risk.17 Subsequent modeling studies have supported the critical role of compliance as a risk determinant.18,19 Understanding the level of, and possible motivators for, noncompliance are important both in the context of minimizing TTI risk and underpinning evidence-based review of deferral policies. The lack of reliable estimates of noncompliance in TTI test–negative Australian donors has been a barrier to comprehensive review of existing deferral policies. An independent expert committee has reviewed sexual activity–based deferrals and recommended that before shortening the deferral duration, the Blood Service consider data on the level of compliance with the existing deferral and possible impacts of any reduction in deferral duration.20 This study aimed to address the knowledge gap and underpin evidence-based policy review. We conducted a large, online, anonymous national survey to estimate the prevalence of, and factors associated with, noncompliance with the existing deferrals for IDU; sex with an injecting drug user; MSM; and females with MSM partners, sex worker activity or contact, and new sexual partner from a HHPC, among TTI test–negative Australian donors. Using a subset of data from this survey we have previously reported results on noncompliance with our MSM deferral.21 Here we report data associated with the other highrisk sexual and nonsexual deferral criteria for which information was collected in the survey.
MATERIALS AND METHODS Data source This study assesses data from the DHQ accuracy survey, an online survey of Australian blood donors conducted between November 2012 and March 2013. Detailed methodology of the survey is described elsewhere.21 Briefly, donors who had donated within the previous 6 weeks, had successfully completed predonation “assessment,” and had tested negative for all mandatory TTI tests were eligible for invitation. A total of 98,044 whole blood donors (randomly selected from a sample representative of national donors) were invited nationwide. E-mails with secure links to the survey questionnaire were sent on a monthly basis. Participation was anonymous and voluntary. Participants were required to read an information statement and had to signal agreement to participation before gaining access to the survey. Ethics approval was
NONCOMPLIANCE WITH HIGH-RISK BEHAVIOR DEFERRALS
obtained from the human research ethics committees of both the Blood Service and the University of New South Wales. Information on key sociodemographics including age, state or territory of donation, country of birth, main language spoken at home (termed “language”), highest education level completed, and number of blood donations in Australia was collected. Questions on perception of donors about their latest predonation assessment experience and test-seeking behavior were also included. A donor was defined to have ever donated blood in Australia for the purpose of test-seeking (i.e., lifetime prevalence of test-seeking) if he or she replied “yes” to the question “Have you ever donated blood in Australia just to be tested for a possible infection?” Information on the reasons for not getting testing done elsewhere, and the infections they were most interested in, were also asked using two separate questions; these questions allowed survey respondents to select multiple answers as relevant (Appendix S1, available as supporting information in the online version of this paper). Finally, donors were asked to disclose if they had ever injected illicit drugs or if they had male-to-male sex (male donors only), sex with MSM (female donors only), sex with an injecting drug user, sex worker activity or contact, and a new sexual partner from a HHPC. Information on the timing of the noncompliant behavior in relation to their most recent donation was also collected. Noncompliance with any (termed “overall noncompliance”) and to each of the selective donor deferral criteria was defined according to the donor’s responses to the relevant questions in the survey (Table 1).
Statistical analysis A detailed analysis on noncompliance with the deferral for MSM among male donors has been described by Seed and colleagues.21 Here we present a similar analysis for other TTI-associated risk behaviors including MSM among male donors, other sexual activity–based deferrals, and IDU. We estimated prevalence and 95% confidence interval (CI) of overall noncompliance and noncompliance with each of the TTI-risk categories. Estimates were presented by sex and donor status (first-time vs. repeat donors) for the permanent deferral for IDU and 12-month deferral for sex with an injecting drug user, MSM, females with MSM partner, sex worker activity or contact, and new sexual partner from a HHPC. The prevalences of the selected high-risk behaviors were compared between blood donors and the age-eligible general population (persons aged 16-80 years) to understand the degree of self-deferral. Population data on the estimated prevalence of IDU,22 sex worker activity or contact,23 and MSM24 were obtained from the Australian Study of Health and Relationships (ASHR). Population data on sex with an injecting drug user
were estimated (Appendix S2A, available as supporting information in the online version of this paper) based on behavioral data from the Australian national survey of people who inject drugs.25 Population data on new sexual partners from a HHPC were estimated (Appendix S2C, available as supporting information in the online version of this paper) based on a published report.26 Population data on females with MSM partners were estimated (Appendix S2B, available as supporting information in the online version of this paper) based on findings from the ASHR.23 Descriptive statistics compared sociodemographic characteristics of overall self-reported compliers and noncompliers in the survey. Analyses were stratified by sex and included data on age, state or territory of donation, country of birth, language, and education. Chi-square and Fisher’s exact tests were used to compare the categorical variables. Factors associated with overall noncompliance were determined using unadjusted and adjusted odds ratios (AORs). A multivariate logistic regression model was built using a stepwise backward elimination approach where the base model contained all variables from the univariate analyses. p values less than 0.05 were considered significant. Analyses were conducted using statistical analysis software (SAS, Version 9.3, SAS, Inc., Cary, NC).
RESULTS Of 98,044 donors invited, 30,790 donors (31.4%) completed the survey. The response rate was similar among male (30.7%) and female (31.1%) donors but slightly differed by donor status (20.9% in first-time compared to 33% in repeat donors). A total of 516 responses were excluded from analyses (including 455 where responses came from links that were not matched with any of the unique links created and distributed). The remaining 30,274 responses were included in final analyses. Study participants comprised 11.7 and 88.3% first-time and repeat donors, and 47.82 and 52.18% male and female donors, respectively. Table 2 presents summarized data on prevalence of noncompliance, overall and to individual deferral category included in the study. Overall noncompliance among Australian TTI test–negative donors was low with a prevalence of 1.65% (95% CI, 1.51%-1.8%). Although not significant, male respondents (1.83%) showed a greater rate of noncompliance compared to females (1.48%). Among males, overall noncompliance varied significantly by donor status; first time 2.87% (2.03%-3.93%) versus repeat 1.73% (1.51%-1.97%). The same trend was also evident among female respondents. Overlap of noncompliance between deferral categories was low. Just over 10% (n = 55) of all (n = 499) noncompliers (to any deferral category) disclosed having more than one deferrable risk behavior. Volume **, ** **
TRANSFUSION
3
4
TRANSFUSION Volume **, ** **
† List contained names of countries with high HIV prevalence.
1. If participant is male: In the last 12 months have you had male-to-male sex (that is, oral or anal sex) with or without a condom? a. Yes b. No 2. If participant is female: In the last 12 months have you had sex (with or without a condom) with a man who you think may have had anal or oral sex with another man? a. Yes b. No c. I do not know 3. Have you EVER “used drugs” by injection or been injected, even once, with drugs not prescribed by a doctor or dentist? a. Yes b. No 4. In the last 12 months have you engaged in sexual activity with someone you might think would answer “yes” to the question “Have you EVER ‘used drugs’ by injection or been injected, even once, with drugs not prescribed by a doctor or dentist?” a. Yes b. No c. I do not know 5. In the last 12 months have you been a male or female sex worker (e.g., received payment for sex in money, gifts, or drugs)? a. Yes b. No 6. In the last 12 months have you engaged in sexual activity with a male or female sex worker? a. Yes b. No 7A. In the 12 months before your last donation, did you have sex in an overseas country that is included in the list† below? a. Yes b. No If a donor answered “Yes” to question 7A then the following question (7B) was asked: 7B. Which of the following most appropriately describes the person you had sex with? a. A resident of that same overseas country b. A resident of another country in the list below c. A resident of an overseas country that is NOT on the list below d. A resident of Australia e. I do not know 8A. In the 12 months before your last donation, did you have sex with a new partner (i.e., someone you had not previously had sex with)? a. Yes b. No If a donor answered “Yes” to question 8A then the following question (8B) was asked: 8B. Which of the following most appropriately describes the person you had sex with? a. Someone who had lived overseas for 12 months or more during the previous 10 years in a country in the list below b. Someone who does NOT fit into the above category c. I do not know 9. Any of the above questions
Question and answer options
Noncompliers Compliers
Noncompliers Compliers Noncompliers Compliers
Noncompliers Compliers Noncompliers Compliers Noncompliers Compliers
Noncompliers Compliers
Overall noncompliers (to any)
Yes No/I do not know
Yes No Yes No/I do not know
Yes No Yes No Yes to q7A and either a. or b. to q7B Any other responses
Yes to q8A and a. to q8B Any other responses
Yes to q1, or Yes to q2, or Yes to q3, or Yes to q4, or Yes to q5, or Yes to q6, or Yes to q7A and either a. or b. to q7B, or Yes to q8A and a. to q8B Any other responses to Questions 1 to 8
Overall compliers
Noncompliers Compliers
Complier or noncomplier
Yes No
Donor response
Any of the above
Sex with a person from a HHPC (past resident)
Sex with a person from a HHPC (current resident)
Sex worker contact
Sex worker activity
Sex with an injecting drug user
IDU
Females who had sex with MSM
MSM
Noncompliance category
TABLE 1. Definition of self-reported compliance and noncompliance with selective donor deferral criteria (based on donor’s responses to relevant questions within the survey)
LUCKY ET AL.
0.25 (0.2-0.32) 0.17 (0.12-0.22) 0.37 (0.3-0.44) 0.29 (0.23-0.36) 0.05 (0.03-0.09) 0.31 (0.25-0.38) 0.22 (0.15-0.31) 0.09 (0.05-0.16) 0.27 (0.2-0.37) 0.01 (0.0002-0.04) 0.03 (0.01-0.07) 0.03 (0.01-0.07) * † ‡ §
Data are reported as % (95% CI). Refers to donors who were noncompliant with any of the selective deferral criteria, as defined in Table 1. Refers to donors who were noncompliant with selective deferral criteria for noncompliance, as defined in Table 1. Includes homosexual and bisexual.
0.21 (0.14-0.31) 0.1 (0.05-0.16) 0.26 (0.18-0.36) 0 0.02 (0.005-0.06) 0.02 (0.005-0.06) 0.27 (0.1-0.58) 0.09 (0.01-0.32) 0.36 (0.15-0.7) 0.04 (0.001-0.25) 0.09 (0.01-0.32) 0.09 (0.01-0.32) 0.25 (0.17-0.35) 0.21 (0.14-0.31) 0.41 (0.31-0.53) 0.58 (0.45-0.72) 0.08 (0.04-0.14) 0.6 (0.47-0.75) 0.7 (0.32-1.32) 0.54 (0.22-1.11) 1.09 (0.59-1.81) 0.85 (0.43-1.52) 0.08 (0.002-0.43) 0.85 (0.43-1.52)
0.29 (0.21-0.39) 0.24 (0.17-0.34) 0.47 (0.36-0.6) 0.6 (0.48-0.74) 0.08 (0.04-0.14) 0.62 (0.5-0.76)
0.54 (0.46-0.62) 0.36 (0.29-0.43) 0.86 (0.75-0.97) 0.73 (0.61-0.88) 0.33 (0.25-0.43) 1.03 (0.87-1.2) 0.72 (0.58-0.87) 0.32 (0.23-0.43) 1 (0.84-1.18) 0.84 (0.51-1.31) 0.4 (0.18-0.76) 1.2 (0.79-1.74) 0.3 (0.21-0.4) 0.36 (0.27-0.48) 0.62 (0.49-0.77) 0.54 (0.22-1.11) 0.62 (0.27-1.22) 1.16 (0.65-1.91)
0.32 (0.23-0.42) 0.39 (0.29-0.5) 0.67 (0.54-0.82)
0.23 (0.16-0.33) 0.18 (0.12-0.26) 0.21 (0.16-0.27) NA 0.18 (0.12-0.26) 0.18 (0.12-0.26) NA 0.16 (0.1-0.24) 0.16 (0.1-0.24) NA 0.36 (0.15-0.7) 0.36 (0.15-0.7) 0.24 (0.17-0.34) NA 0.24 (0.17-0.34) 0.16 (0.02-0.56) NA 0.16 (0.02-0.56)
0.23 (0.16-0.33) NA 0.23 (0.16-0.33)
Total 1.65 (1.51-1.8) Total 1.48 (1.3-1.68) Female Repeat donors 1.4 (1.21-1.62) First-time donors 1.95 (1.42-2.61) Total 1.83 (1.62-2.06) Male Repeat donors 1.73 (1.51-1.97) First-time donors 2.87 (2.03-3.93)
Noncompliance category Overall noncompliance† Noncompliance (by deferral criteria)‡ MSM§ Females who had sex with MSM Male to male sex (MSM/female with MSM partner) Sex with an injecting drug user IDU Injecting drug user (donor/sex partner) Sex with a person from a HHPC Current resident Past resident Past/current resident Sex worker contact Sex worker activity Sex work (activity/contact)
TABLE 2. Prevalence of self-reported noncompliance with any and to selective donor deferral categories by sex and donor status*
NONCOMPLIANCE WITH HIGH-RISK BEHAVIOR DEFERRALS
A total of 29 of 15,798 female donors (0.18%) disclosed having sex with MSM in the 12 months before their last donation with no significant difference in the prevalence between first-time and repeat donors. A history of IDU (ever) was reported by 108 survey respondents, a noncompliance prevalence of 0.36%. A total of 162 donors (0.54%) disclosed having sex with an injecting drug user in the 12 months before their last donation. While noncompliance with the deferral for IDU did not differ significantly by sex, the proportion of donors who had sex with an injecting drug user was significantly greater among females (0.73%; 0.61%-0.88%) compared to males (0.32%; 0.23%-0.42%). In general, noncompliance with sex work–related deferrals was low (0.31%). The prevalence of noncompliance with deferral for sex worker activity (0.05%) was the lowest among the eight deferral criteria evaluated. A total of 88 donors (0.29%) reported sex worker contact in the 12 months before their most recent donation. Noncompliance with deferral for sex worker contact differed significantly between male (0.6%; 0.48%-0.74%) and female donors (0.01%; 0.0002%-0.04%). A total of 111 donors (0.37%) reported having sex with a new partner in the 12 months before their last donation where the sexual partner was either a current (0.25%) or past (0.17%) resident from a HHPC. Overall, noncompliance to this category did not vary by sex or donor status. Nonetheless, although not significant, the prevalence was greater in first-time compared to repeat donors among both male and female respondents. Approximately three-quarters (78/108) of the respondents with a history of ever injecting illicit drugs reported that the most recent episode occurred more than 5 years before their last donation. All else being equal, if deferral for IDUs had been 5 years or 1 year or 6 months at that time, the predicted level of noncompliance would have been 0.10, 0.05, and 0.02%, respectively. A majority of the donors who were noncompliant with sexual activity–based deferrals (113 of 162 who reported sex with an injecting drug user, 13 of 16 who reported sex worker activity, 63 of 88 who reported sex worker contact) disclosed that the last sexual contact occurred within 6 months before their donation. If a 6-month (instead of the current 12-month) exclusion was applicable during the survey and all else being equal, the levels of noncompliance for sex with an injecting drug user, sex worker activity, and sex worker contact are predicted to have been 0.37, 0.04, and 0.20%, respectively. The proportion of blood donors with selected deferrable risks was substantially less than the proportion of the age-eligible general population with those risks (Table 3). In comparison with the general population, the highest relative reduction was observed for sex worker activity (14 times lower) followed by MSM (11 times lower) and females who had sex with MSM (9 times lower). Table 4 summarizes key sociodemographic characteristics of the overall compliers and noncompliers. A total Volume **, ** **
TRANSFUSION
5
LUCKY ET AL.
TABLE 3. Comparison between donor and age-eligible general population for the presence of deferrable risk
Deferrable risk MSM† Female who had sex with MSM IDU (ever) Sex with an injecting drug user Sex worker activity Sex worker contact Sex with a new partner from a HHPC Past resident‡ Current resident§
Proportion (%) of blood donors with deferrable risk 0.23 (of male) 0.18 (of female) 0.36 0.54 0.05 0.29 0.17 0.25
Proportion (%) of age-eligible general population with deferrable risk* 1.9 (of male) 1.6 (of female) 3.00 1.86 0.70 1
Ratio (age-eligible general population/donor population) 8 9 8 3 14 3
1.12
7
* Population data on: • MSM from the ASHR;24 • IDU from the ASHR;22 • Sex with an injecting drug user estimated based on data from the Australian national survey of people who inject drugs;25 • Sex worker activity/contact from the ASHR;23 • New sex partner from a HHPC (past resident) was estimated based on new partner acquisition rate in Australia;26 • Females who had sex with MSM was estimated based on data from the ASHR. † Includes homosexual and bisexual men who reported having sex with another man in the past 12 months. ‡ Past resident is someone who had lived overseas for 12 months or more during the previous 10 years in any of the HHPCs. § Population data unknown.
TABLE 4. Sociodemographic characteristics of self-reported compliant and noncompliant* donors, overall and by sex; number (%)† Variables Sex Female Male Status First-time Repeat Age group (years) 50 years) were 40% less likely to be noncompliant. Non– English-speaking donors (donors who reported English not being the main language spoken at home) were 2.84 times more likely to be noncompliant. Male donors who reported lack of privacy during assessment, those who identified some questions on the DHQ as being too personal, and those who reported discomfort raising questions to donation center staff were at approximately two times, five times, and four times more likely to be noncompliant overall, respectively. Having multiple sexual partners in the 12 months before donation had the strongest association with noncompliance among male respondents. Compared with donors with one or no partner, male donors who had two to four partners were approximately 12 times and donors who had five or more partners were approximately 20 times more likely to be noncompliant. Finally, testseeking donors (i.e., those disclosing that they donated blood to be tested for infection) also had significantly greater likelihood of nondisclosure during predonation assessment (AOR, 2.39; 1.14-5.04). As with male respondents, having multiple sexual partners in the 12 months before their donation was also the most significant factor in females. When compared with younger donors (
BACKGROUND: Using a predonation screening questionnaire, potential blood donors are screened for medical or behavioral factors associated with an increased risk for transfusion-transmissible infection. After disclosure of these risks, potential donors are deferred from donating. Understanding the degree of failure to disclose full and truthful information (termed noncompliance) is important to determine and minimize residual risk. This study estimates the prevalence of, and likely reasons for, noncompliance among Australian donors with the deferrals for injecting drug use, sex with an injecting drug user, male-to-male sex, sex worker activity or contact, and sex with a partner from a highHIV-prevalence country. STUDY DESIGN AND METHODS: An anonymous, online survey of a nationally representative sample of Australian blood donors was conducted. Prevalence of noncompliance with deferrable risk categories was estimated. Factors associated with noncompliance were determined using unadjusted and adjusted odds ratios. RESULTS: Of 98,044 invited donors, 30,790 donors completed the survey. The estimated prevalence of overall noncompliance (i.e., to at least one screening question) was 1.65% (95% confidence interval CI, 1.51%-1.8%). Noncompliance with individual deferrals ranged from 0.05% (sex work) to 0.54% (sex with an injecting drug user). The prevalences of the disclosed exclusionary risk behaviors were three to 14 times lower than their estimated prevalence in the general population. CONCLUSION: The prevalence of noncompliance is relatively low but our estimate is likely to be a lower bound. The selected high-risk behaviors were substantially less common in blood donors compared to the general population suggesting that self-deferral is effective. Nevertheless, a focus on further minimization should improve the blood safety.
A
fter the emergence of human immunodeficiency virus (HIV)/AIDS in the 1980s, blood services around the world introduced questions about high-risk lifestyle behaviors associated with transfusion-transmissible infections (TTIs) including HIV and hepatitis.1 The importance of deferral measures focusing on preventing donation by “high-risk” individuals was first illustrated in the early 1980s before universal HIV antibody screening when Busch and colleagues2 demonstrated a marked decline in the “per-unit” risk of HIV infection in the United States coinciding with the progressive implementation of donor selection measures. ABBREVIATIONS: ACASI = audio computer-assisted structured interview; AOR = adjusted odds ratio; ASHR = Australian Study of Health and Relationships; DHQ = donor health questionnaire; HHPC = high-HIV-prevalence country; IDU = injecting drug use; MSM = men who have sex with men; STI = sexually transmitted infection; TTI(s) = transfusiontransmissible infection(s). From 1The Kirby Institute, The University of New South Wales, Sydney, New South Wales; 2Australian Red Cross Blood Service, Perth, Western Australia; 3Australian Red Cross Blood Service, Sydney, New South Wales; and 4Australian Red Cross Blood Service, Brisbane, Queensland, Australia. Address reprint requests to: David P. Wilson, The Kirby Institute, The University of New South Wales, Sydney, NSW 2052, Australia; e-mail: [email protected]. The authors acknowledge funding jointly from the Australian Red Cross Blood Service and the Kirby Institute. The Australian Red Cross Blood Service is fully funded by the Australian Government for the provision of blood products and services to the Australian community. The Kirby Institute is funded by the Australian Government Department of Health and Ageing and is affiliated with the Faculty of Medicine at the University of New South Wales. Received for publication September 27, 2013; revision received November 27, 2013, and accepted November 27, 2013. doi: 10.1111/trf.12554 © 2014 AABB TRANSFUSION **;**:**-**. Volume **, ** **
TRANSFUSION
1
LUCKY ET AL.
Importantly, as tests for TTIs are unable to detect donors with very recently acquired infection (known as the “window period”) the donor health questionnaire (DHQ) remains important for maximizing the safety of blood transfusions even where universal testing for a TTI is undertaken.3,4 In Australia, each donor is required to self-complete a comprehensive DHQ every time he or she donates. The DHQ is reviewed at a private and confidential interview with the donor and a legally binding declaration form is signed before donation. The Australian Red Cross Blood Service (Blood Service) is highly reliant on the donor’s complete and truthful answers to all interview questions (i.e., “compliance”). This is particularly important for questions relating to TTI risk behavior given the existence of the testing window period. Should donors in the window period fail to answer truthfully a question that would normally result in their deferral from donation, they may place recipients at risk because a potentially infectious unit of blood that testing will not identify could be transfused. The risk has been substantially reduced as a result of state-of-the-art TTI testing including nucleic acid testing (NAT) for HIV, hepatitis C virus (HCV), and hepatitis B virus in Australia.5 Despite this, the risk is real as evidenced by occasional cases of TTI transmission associated with noncompliance, which continue to occur even where NAT is implemented.6 Comprehensive data on noncompliance with highrisk donor deferral criteria are limited. In Australia, noncompliance among test-positive donors ranges between 12.9 and 24.4%.5 However, before this study no data were available for noncompliance among TTI test–negative donors. Internationally, studies conducted in Canada,7,8 Germany,9 the United Kingdom,10 the United States,11-14 and Hong Kong15 have examined various aspects of risk nondisclosure during predonation assessment. The majority focused on permanent deferral for male-to-male sex with a range of noncompliance of 0.8% to 2.3% among male donors.7,12,15 A few studies assessed other high-risk nonsexual behaviors such as injecting drug use (IDU).8,9,13 However, there is a paucity of data concerning noncompliance with high-risk heterosexual deferrals (e.g., sex workers and their sexual contacts). In Australia, a history of IDU (even one episode) results in permanent deferral while sex with an injecting drug user defers a person temporarily (12 months since last sexual contact).16 Similarly, a 12-month exclusion also applies for men who have sex with men (MSM), females who have sex with MSM, sex worker activity or contact, and sexual contact with a person from a high-HIVprevalence country (HHPC). These policies are reviewed annually and revised where appropriate. The current 12-month exclusion for the previously defined sexual activity–based deferrals became nationally applicable during the late-1990s when the Blood Service was estab2
TRANSFUSION Volume **, ** **
lished as a national service. The latest change to the duration of deferral occurred in 2000 when the current 12-month deferral for MSM was implemented nationwide.17 Notably, the deferral policy for IDU has not been revised since the 1980s, when the original 5-year deferral was changed to permanent deferral. A previous Blood Service study of HIV-positive donors highlighted the importance of compliance with the current 12-month deferral for male-to-male sex on HIV residual risk.17 Subsequent modeling studies have supported the critical role of compliance as a risk determinant.18,19 Understanding the level of, and possible motivators for, noncompliance are important both in the context of minimizing TTI risk and underpinning evidence-based review of deferral policies. The lack of reliable estimates of noncompliance in TTI test–negative Australian donors has been a barrier to comprehensive review of existing deferral policies. An independent expert committee has reviewed sexual activity–based deferrals and recommended that before shortening the deferral duration, the Blood Service consider data on the level of compliance with the existing deferral and possible impacts of any reduction in deferral duration.20 This study aimed to address the knowledge gap and underpin evidence-based policy review. We conducted a large, online, anonymous national survey to estimate the prevalence of, and factors associated with, noncompliance with the existing deferrals for IDU; sex with an injecting drug user; MSM; and females with MSM partners, sex worker activity or contact, and new sexual partner from a HHPC, among TTI test–negative Australian donors. Using a subset of data from this survey we have previously reported results on noncompliance with our MSM deferral.21 Here we report data associated with the other highrisk sexual and nonsexual deferral criteria for which information was collected in the survey.
MATERIALS AND METHODS Data source This study assesses data from the DHQ accuracy survey, an online survey of Australian blood donors conducted between November 2012 and March 2013. Detailed methodology of the survey is described elsewhere.21 Briefly, donors who had donated within the previous 6 weeks, had successfully completed predonation “assessment,” and had tested negative for all mandatory TTI tests were eligible for invitation. A total of 98,044 whole blood donors (randomly selected from a sample representative of national donors) were invited nationwide. E-mails with secure links to the survey questionnaire were sent on a monthly basis. Participation was anonymous and voluntary. Participants were required to read an information statement and had to signal agreement to participation before gaining access to the survey. Ethics approval was
NONCOMPLIANCE WITH HIGH-RISK BEHAVIOR DEFERRALS
obtained from the human research ethics committees of both the Blood Service and the University of New South Wales. Information on key sociodemographics including age, state or territory of donation, country of birth, main language spoken at home (termed “language”), highest education level completed, and number of blood donations in Australia was collected. Questions on perception of donors about their latest predonation assessment experience and test-seeking behavior were also included. A donor was defined to have ever donated blood in Australia for the purpose of test-seeking (i.e., lifetime prevalence of test-seeking) if he or she replied “yes” to the question “Have you ever donated blood in Australia just to be tested for a possible infection?” Information on the reasons for not getting testing done elsewhere, and the infections they were most interested in, were also asked using two separate questions; these questions allowed survey respondents to select multiple answers as relevant (Appendix S1, available as supporting information in the online version of this paper). Finally, donors were asked to disclose if they had ever injected illicit drugs or if they had male-to-male sex (male donors only), sex with MSM (female donors only), sex with an injecting drug user, sex worker activity or contact, and a new sexual partner from a HHPC. Information on the timing of the noncompliant behavior in relation to their most recent donation was also collected. Noncompliance with any (termed “overall noncompliance”) and to each of the selective donor deferral criteria was defined according to the donor’s responses to the relevant questions in the survey (Table 1).
Statistical analysis A detailed analysis on noncompliance with the deferral for MSM among male donors has been described by Seed and colleagues.21 Here we present a similar analysis for other TTI-associated risk behaviors including MSM among male donors, other sexual activity–based deferrals, and IDU. We estimated prevalence and 95% confidence interval (CI) of overall noncompliance and noncompliance with each of the TTI-risk categories. Estimates were presented by sex and donor status (first-time vs. repeat donors) for the permanent deferral for IDU and 12-month deferral for sex with an injecting drug user, MSM, females with MSM partner, sex worker activity or contact, and new sexual partner from a HHPC. The prevalences of the selected high-risk behaviors were compared between blood donors and the age-eligible general population (persons aged 16-80 years) to understand the degree of self-deferral. Population data on the estimated prevalence of IDU,22 sex worker activity or contact,23 and MSM24 were obtained from the Australian Study of Health and Relationships (ASHR). Population data on sex with an injecting drug user
were estimated (Appendix S2A, available as supporting information in the online version of this paper) based on behavioral data from the Australian national survey of people who inject drugs.25 Population data on new sexual partners from a HHPC were estimated (Appendix S2C, available as supporting information in the online version of this paper) based on a published report.26 Population data on females with MSM partners were estimated (Appendix S2B, available as supporting information in the online version of this paper) based on findings from the ASHR.23 Descriptive statistics compared sociodemographic characteristics of overall self-reported compliers and noncompliers in the survey. Analyses were stratified by sex and included data on age, state or territory of donation, country of birth, language, and education. Chi-square and Fisher’s exact tests were used to compare the categorical variables. Factors associated with overall noncompliance were determined using unadjusted and adjusted odds ratios (AORs). A multivariate logistic regression model was built using a stepwise backward elimination approach where the base model contained all variables from the univariate analyses. p values less than 0.05 were considered significant. Analyses were conducted using statistical analysis software (SAS, Version 9.3, SAS, Inc., Cary, NC).
RESULTS Of 98,044 donors invited, 30,790 donors (31.4%) completed the survey. The response rate was similar among male (30.7%) and female (31.1%) donors but slightly differed by donor status (20.9% in first-time compared to 33% in repeat donors). A total of 516 responses were excluded from analyses (including 455 where responses came from links that were not matched with any of the unique links created and distributed). The remaining 30,274 responses were included in final analyses. Study participants comprised 11.7 and 88.3% first-time and repeat donors, and 47.82 and 52.18% male and female donors, respectively. Table 2 presents summarized data on prevalence of noncompliance, overall and to individual deferral category included in the study. Overall noncompliance among Australian TTI test–negative donors was low with a prevalence of 1.65% (95% CI, 1.51%-1.8%). Although not significant, male respondents (1.83%) showed a greater rate of noncompliance compared to females (1.48%). Among males, overall noncompliance varied significantly by donor status; first time 2.87% (2.03%-3.93%) versus repeat 1.73% (1.51%-1.97%). The same trend was also evident among female respondents. Overlap of noncompliance between deferral categories was low. Just over 10% (n = 55) of all (n = 499) noncompliers (to any deferral category) disclosed having more than one deferrable risk behavior. Volume **, ** **
TRANSFUSION
3
4
TRANSFUSION Volume **, ** **
† List contained names of countries with high HIV prevalence.
1. If participant is male: In the last 12 months have you had male-to-male sex (that is, oral or anal sex) with or without a condom? a. Yes b. No 2. If participant is female: In the last 12 months have you had sex (with or without a condom) with a man who you think may have had anal or oral sex with another man? a. Yes b. No c. I do not know 3. Have you EVER “used drugs” by injection or been injected, even once, with drugs not prescribed by a doctor or dentist? a. Yes b. No 4. In the last 12 months have you engaged in sexual activity with someone you might think would answer “yes” to the question “Have you EVER ‘used drugs’ by injection or been injected, even once, with drugs not prescribed by a doctor or dentist?” a. Yes b. No c. I do not know 5. In the last 12 months have you been a male or female sex worker (e.g., received payment for sex in money, gifts, or drugs)? a. Yes b. No 6. In the last 12 months have you engaged in sexual activity with a male or female sex worker? a. Yes b. No 7A. In the 12 months before your last donation, did you have sex in an overseas country that is included in the list† below? a. Yes b. No If a donor answered “Yes” to question 7A then the following question (7B) was asked: 7B. Which of the following most appropriately describes the person you had sex with? a. A resident of that same overseas country b. A resident of another country in the list below c. A resident of an overseas country that is NOT on the list below d. A resident of Australia e. I do not know 8A. In the 12 months before your last donation, did you have sex with a new partner (i.e., someone you had not previously had sex with)? a. Yes b. No If a donor answered “Yes” to question 8A then the following question (8B) was asked: 8B. Which of the following most appropriately describes the person you had sex with? a. Someone who had lived overseas for 12 months or more during the previous 10 years in a country in the list below b. Someone who does NOT fit into the above category c. I do not know 9. Any of the above questions
Question and answer options
Noncompliers Compliers
Noncompliers Compliers Noncompliers Compliers
Noncompliers Compliers Noncompliers Compliers Noncompliers Compliers
Noncompliers Compliers
Overall noncompliers (to any)
Yes No/I do not know
Yes No Yes No/I do not know
Yes No Yes No Yes to q7A and either a. or b. to q7B Any other responses
Yes to q8A and a. to q8B Any other responses
Yes to q1, or Yes to q2, or Yes to q3, or Yes to q4, or Yes to q5, or Yes to q6, or Yes to q7A and either a. or b. to q7B, or Yes to q8A and a. to q8B Any other responses to Questions 1 to 8
Overall compliers
Noncompliers Compliers
Complier or noncomplier
Yes No
Donor response
Any of the above
Sex with a person from a HHPC (past resident)
Sex with a person from a HHPC (current resident)
Sex worker contact
Sex worker activity
Sex with an injecting drug user
IDU
Females who had sex with MSM
MSM
Noncompliance category
TABLE 1. Definition of self-reported compliance and noncompliance with selective donor deferral criteria (based on donor’s responses to relevant questions within the survey)
LUCKY ET AL.
0.25 (0.2-0.32) 0.17 (0.12-0.22) 0.37 (0.3-0.44) 0.29 (0.23-0.36) 0.05 (0.03-0.09) 0.31 (0.25-0.38) 0.22 (0.15-0.31) 0.09 (0.05-0.16) 0.27 (0.2-0.37) 0.01 (0.0002-0.04) 0.03 (0.01-0.07) 0.03 (0.01-0.07) * † ‡ §
Data are reported as % (95% CI). Refers to donors who were noncompliant with any of the selective deferral criteria, as defined in Table 1. Refers to donors who were noncompliant with selective deferral criteria for noncompliance, as defined in Table 1. Includes homosexual and bisexual.
0.21 (0.14-0.31) 0.1 (0.05-0.16) 0.26 (0.18-0.36) 0 0.02 (0.005-0.06) 0.02 (0.005-0.06) 0.27 (0.1-0.58) 0.09 (0.01-0.32) 0.36 (0.15-0.7) 0.04 (0.001-0.25) 0.09 (0.01-0.32) 0.09 (0.01-0.32) 0.25 (0.17-0.35) 0.21 (0.14-0.31) 0.41 (0.31-0.53) 0.58 (0.45-0.72) 0.08 (0.04-0.14) 0.6 (0.47-0.75) 0.7 (0.32-1.32) 0.54 (0.22-1.11) 1.09 (0.59-1.81) 0.85 (0.43-1.52) 0.08 (0.002-0.43) 0.85 (0.43-1.52)
0.29 (0.21-0.39) 0.24 (0.17-0.34) 0.47 (0.36-0.6) 0.6 (0.48-0.74) 0.08 (0.04-0.14) 0.62 (0.5-0.76)
0.54 (0.46-0.62) 0.36 (0.29-0.43) 0.86 (0.75-0.97) 0.73 (0.61-0.88) 0.33 (0.25-0.43) 1.03 (0.87-1.2) 0.72 (0.58-0.87) 0.32 (0.23-0.43) 1 (0.84-1.18) 0.84 (0.51-1.31) 0.4 (0.18-0.76) 1.2 (0.79-1.74) 0.3 (0.21-0.4) 0.36 (0.27-0.48) 0.62 (0.49-0.77) 0.54 (0.22-1.11) 0.62 (0.27-1.22) 1.16 (0.65-1.91)
0.32 (0.23-0.42) 0.39 (0.29-0.5) 0.67 (0.54-0.82)
0.23 (0.16-0.33) 0.18 (0.12-0.26) 0.21 (0.16-0.27) NA 0.18 (0.12-0.26) 0.18 (0.12-0.26) NA 0.16 (0.1-0.24) 0.16 (0.1-0.24) NA 0.36 (0.15-0.7) 0.36 (0.15-0.7) 0.24 (0.17-0.34) NA 0.24 (0.17-0.34) 0.16 (0.02-0.56) NA 0.16 (0.02-0.56)
0.23 (0.16-0.33) NA 0.23 (0.16-0.33)
Total 1.65 (1.51-1.8) Total 1.48 (1.3-1.68) Female Repeat donors 1.4 (1.21-1.62) First-time donors 1.95 (1.42-2.61) Total 1.83 (1.62-2.06) Male Repeat donors 1.73 (1.51-1.97) First-time donors 2.87 (2.03-3.93)
Noncompliance category Overall noncompliance† Noncompliance (by deferral criteria)‡ MSM§ Females who had sex with MSM Male to male sex (MSM/female with MSM partner) Sex with an injecting drug user IDU Injecting drug user (donor/sex partner) Sex with a person from a HHPC Current resident Past resident Past/current resident Sex worker contact Sex worker activity Sex work (activity/contact)
TABLE 2. Prevalence of self-reported noncompliance with any and to selective donor deferral categories by sex and donor status*
NONCOMPLIANCE WITH HIGH-RISK BEHAVIOR DEFERRALS
A total of 29 of 15,798 female donors (0.18%) disclosed having sex with MSM in the 12 months before their last donation with no significant difference in the prevalence between first-time and repeat donors. A history of IDU (ever) was reported by 108 survey respondents, a noncompliance prevalence of 0.36%. A total of 162 donors (0.54%) disclosed having sex with an injecting drug user in the 12 months before their last donation. While noncompliance with the deferral for IDU did not differ significantly by sex, the proportion of donors who had sex with an injecting drug user was significantly greater among females (0.73%; 0.61%-0.88%) compared to males (0.32%; 0.23%-0.42%). In general, noncompliance with sex work–related deferrals was low (0.31%). The prevalence of noncompliance with deferral for sex worker activity (0.05%) was the lowest among the eight deferral criteria evaluated. A total of 88 donors (0.29%) reported sex worker contact in the 12 months before their most recent donation. Noncompliance with deferral for sex worker contact differed significantly between male (0.6%; 0.48%-0.74%) and female donors (0.01%; 0.0002%-0.04%). A total of 111 donors (0.37%) reported having sex with a new partner in the 12 months before their last donation where the sexual partner was either a current (0.25%) or past (0.17%) resident from a HHPC. Overall, noncompliance to this category did not vary by sex or donor status. Nonetheless, although not significant, the prevalence was greater in first-time compared to repeat donors among both male and female respondents. Approximately three-quarters (78/108) of the respondents with a history of ever injecting illicit drugs reported that the most recent episode occurred more than 5 years before their last donation. All else being equal, if deferral for IDUs had been 5 years or 1 year or 6 months at that time, the predicted level of noncompliance would have been 0.10, 0.05, and 0.02%, respectively. A majority of the donors who were noncompliant with sexual activity–based deferrals (113 of 162 who reported sex with an injecting drug user, 13 of 16 who reported sex worker activity, 63 of 88 who reported sex worker contact) disclosed that the last sexual contact occurred within 6 months before their donation. If a 6-month (instead of the current 12-month) exclusion was applicable during the survey and all else being equal, the levels of noncompliance for sex with an injecting drug user, sex worker activity, and sex worker contact are predicted to have been 0.37, 0.04, and 0.20%, respectively. The proportion of blood donors with selected deferrable risks was substantially less than the proportion of the age-eligible general population with those risks (Table 3). In comparison with the general population, the highest relative reduction was observed for sex worker activity (14 times lower) followed by MSM (11 times lower) and females who had sex with MSM (9 times lower). Table 4 summarizes key sociodemographic characteristics of the overall compliers and noncompliers. A total Volume **, ** **
TRANSFUSION
5
LUCKY ET AL.
TABLE 3. Comparison between donor and age-eligible general population for the presence of deferrable risk
Deferrable risk MSM† Female who had sex with MSM IDU (ever) Sex with an injecting drug user Sex worker activity Sex worker contact Sex with a new partner from a HHPC Past resident‡ Current resident§
Proportion (%) of blood donors with deferrable risk 0.23 (of male) 0.18 (of female) 0.36 0.54 0.05 0.29 0.17 0.25
Proportion (%) of age-eligible general population with deferrable risk* 1.9 (of male) 1.6 (of female) 3.00 1.86 0.70 1
Ratio (age-eligible general population/donor population) 8 9 8 3 14 3
1.12
7
* Population data on: • MSM from the ASHR;24 • IDU from the ASHR;22 • Sex with an injecting drug user estimated based on data from the Australian national survey of people who inject drugs;25 • Sex worker activity/contact from the ASHR;23 • New sex partner from a HHPC (past resident) was estimated based on new partner acquisition rate in Australia;26 • Females who had sex with MSM was estimated based on data from the ASHR. † Includes homosexual and bisexual men who reported having sex with another man in the past 12 months. ‡ Past resident is someone who had lived overseas for 12 months or more during the previous 10 years in any of the HHPCs. § Population data unknown.
TABLE 4. Sociodemographic characteristics of self-reported compliant and noncompliant* donors, overall and by sex; number (%)† Variables Sex Female Male Status First-time Repeat Age group (years) 50 years) were 40% less likely to be noncompliant. Non– English-speaking donors (donors who reported English not being the main language spoken at home) were 2.84 times more likely to be noncompliant. Male donors who reported lack of privacy during assessment, those who identified some questions on the DHQ as being too personal, and those who reported discomfort raising questions to donation center staff were at approximately two times, five times, and four times more likely to be noncompliant overall, respectively. Having multiple sexual partners in the 12 months before donation had the strongest association with noncompliance among male respondents. Compared with donors with one or no partner, male donors who had two to four partners were approximately 12 times and donors who had five or more partners were approximately 20 times more likely to be noncompliant. Finally, testseeking donors (i.e., those disclosing that they donated blood to be tested for infection) also had significantly greater likelihood of nondisclosure during predonation assessment (AOR, 2.39; 1.14-5.04). As with male respondents, having multiple sexual partners in the 12 months before their donation was also the most significant factor in females. When compared with younger donors (
