Subjective donor deferral as a tool for increased blood transfusion safety: A cross‐sectional observational study

Abstract Objectives This study aims at evaluating whether subjective donor deferral (SDD) has the potential for increasing blood transfusion safety. Background Appropriate donor selection via clinical and serologic screening is necessary to prevent transfusion‐transmissible infections (TTIs). One additional strategy adopted by some Brazilian blood transfusion centers (BTCs) is the rejection of a donation by the pre‐donation interviewer based on subjective factors. Methods/Materials We conducted a STROBE‐guided cross‐sectional study including 105 005 prospective donors who presented to our BTC between 1 January 2013, and 31 December 2015. Donors were evaluated for age, gender, education level, donation type and history, confidential unit exclusion, SDD, and results of serologic screening for TTIs. Results Even after controlling for potential confounding variables, subjectively deferred donors were more likely to have at least one reactive serology in the standard screening (OR: 2.80; 95% CI: 2.13‐3.69; P < .001). They also had a higher risk for testing positive for syphilis (OR: 4.47; 95% CI: 3.05‐6.55; P < .001), hepatitis B (OR: 5.69; 95% CI: 2.48‐13.08; P < .001), and HIV (OR: 6.14; 95% CI: 3.22‐11.69; P < .001). Conclusions Routine implementation of SDD in donor selection may be an effective additional measure to avoid TTIs, highlighting the importance of interviewer experience, perspicacity, and face‐to‐face contact with donors for blood safety assurance.


| INTRODUCTION
In Brazil, many tools ensure the safety of hemotherapy. Initially, candidates for blood donation are evaluated with a brief epidemiological survey. Then, donor selection is performed through three major steps: a pre-triage, in which weight, blood pressure, temperature, and heart rate are assessed; a hematological triage, in which hemoglobin and/or hematocrit levels are checked; and a pre-donation interview with a qualified health professional. In the latter, the clinical and epidemiological history of the candidate is, as well as their actual health state and habits, reviewed. 1 Subsequently, the donors' blood samples undergo a standard laboratory screening, where serologic and molecular biology tests are performed to identify potential transfusion-transmissible infections (TTIs). Currently, Chagas disease, human immunodeficiency virus (HIV) I and II, human T-lymphotropic virus (HTLV) I and II, syphilis, hepatitis B virus (HBV), and hepatitis C virus (HCV) infections are routinely evaluated. 2 The Confidential Unit Exclusion (CUE) system is also recognized by Brazilian law. It allows donors to indicate privately that their blood donation may be unsafe for transfusion; as a result, blood collection proceeds as usual, but the bag collected is then discarded. 3 Finally, an additional strategy is the subjective donor deferral (SDD), which means the possibility for discarding blood bags based on data perceived subjectively by the pre-donation interviewer (eg, signs given by the prospective donor during the interview that may indicate he/she is failing with the truth or trying to hide important health information). 4 Despite being empirically adopted by several Brazilian blood transfusion centers (BTCs), SDD is still an under-researched tool that has not been formally validated in the previous literature. In this scenario, this study primarily aims at evaluating whether the SDD strategy has the potential to improve blood transfusion safety.

| Study design and participants
This is an observational, monocentric, cross-sectional study including candidates for blood donation who presented to the Hematology and The inclusion criteria were age ≥18 years and presentation for blood donation between 1 January 2013, and 31 December 2015.
The exclusion criteria were intent for autologous blood donation and lack of Blood Bank System data on any of the parameters analyzed (eg, gender, age, years of education, donation history, CUE, SDD, or serologic test results). The flowchart of the study selection process is presented in Figure 1.
All blood samples were initially screened for the following: hepatitis B surface antigen (HBsAg) and anti-HIV1/2 antibodies by Elecsys electrochemiluminescence immunoassay (Roche Diagnostics, Penzberg, Germany); anti-hepatitis B core antigen (anti-HBc) and anti-HTLV1/2 antibodies by ARCHITECT chemiluminescence immunoassays (Abbott Laboratories, Abbott Park, Illinois, USA); and anti-HCV antibodies by VITROS chemiluminescence immunoassay (Ortho Clinical Diagnostics, Raritan, New Jersey, USA). In addition, Oxoid venereal disease research laboratory (VDRL) test kit (Thermo Fisher Scientific, Basingstoke, UK) was used for syphilis screening. For HIV nucleic acid testing (NAT), the Superscript III Platinum One-Step Quantitative RT-PCR kit (Thermo Fisher Scientific, Life Technologies, Carlsbad, CA) was used. Samples with positive results were retested by the same method and using the same equipment. Two positive tests defined a sample as reactive for a specific infectious agent. Individuals with at least one reactive serology after standard laboratory screening were considered as having "positive general serologic screening." Individuals with no reactive serology after standard laboratory screening were considered as having "negative general serologic screening." Despite being routinely performed in HEMOCE, serology for Chagas disease was not considered in this analysis due to the lack of association between risky sexual behavior and active infection.
During the analysis, we evaluated whether SDD was established or not for each of the study participants. SDD can be defined as an additional strategy that some BTCs use in the attempt to reduce transfusion of blood from individuals in the window period for TTIs. It consists of discard of the collected blood based on data that were not reported by the prospective donor during clinical triage but instead perceived subjectively by the pre-donation interviewer. The decision for SDD can be performed concomitantly, before, or after the clinical triage. Although serologic testing is still performed in blood samples F I G U R E 1 Flowchart of prospective blood donors included in the analysis after applying exclusion and inclusion criteria of subjectively deferred individuals, the blood bag collected is always discarded, even if initial serologic/molecular testing is negative.
Finally, all first-time donors submitted to SDD and with negative serologic/molecular testing were identified and, in those who returned for a second donation, the presence or absence of seroconversion for a sexually transmitted infection (STI) was evaluated.
The local Ethics Committee approved the study protocol (approval number 86154618.7.0000.8152) and provided exemption from the need for informed consent, as this study involved no risks to subjects, could not be carried out practicably without the waiver, and analyzed data already available in the Blood Bank System. In addition, the waiver would not adversely affect the rights or welfare of the subjects, who would still receive standard of care and have the privacy of their personal/health information guaranteed. Of note, we have conducted the present research according to the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) guidelines for cross-sectional studies.

| Statistical analysis
Descriptive statistics of categorical variables comprised the calculation of absolute frequency (n) and relative frequency (%). The association between sociodemographic factors and clinical triage information (explanatory variables) and the primary outcome variable (ie, the result of the general serologic screening of blood donors) was evaluated using univariate logistic regression analysis.
The strength of this association was measured by determining the crude odds ratio (OR), as well as the accuracy (95% confidence interval; 95% CI) and significance (Wald test) of the estimate. In order to evaluate each disease individually, the same approach was repeated with HIV, HBV, and syphilis serologic results as separate outcome variables.

| General serologic screening
Univariate analysis of the association between sociodemographic/clinical triage data and a positive general serologic screening is shown in

| Donor seroconversion
During the study period, the frequency of return of the subjectively deferred donors to the BTC was also evaluated. Only 60 (5.4%) of the 1115 individuals submitted to SDD returned for a second donation.
Of those, two tested positive for HIV, four for syphilis, and one for anti-HBC, corresponding to 11.7% of seroconversion.

| DISCUSSION
SDD was demonstrated to be a helpful donor triage tool in this crosssectional study, remaining a statistically significant risk factor for all the analyzed outcomes even when other variables' influence was controlled. Subjectively deferred donors were 2.80 times more likely to have at least one positive screening test, 4.47 times more likely to have positive syphilis serology, 5.69 times more likely to have positive HBV serology, and 6.14 times more likely to have positive HIV serology. As the rates of positive infectious disease markers can correlate with the incidence of window-period infections, 5 the data herein presented suggest that the SDD strategy has the potential to improve blood transfusion safety.
Although serologic testing is still performed in subjectively  12,13 Our results indicated that no specific age range is associated with higher odds of HIV infection, contrasting with previous studies suggesting that younger age is a risk factor. 14,15 The current rapid increase in the prevalence of HIV-infected people aged 50 years or older may have contributed to such an outcome. 16 Lower education level was associated with a higher risk for positive general serological screening and positive syphilis screening, in consonance with previous literature data. 17 However, a similar association was not found for a positive screening for HIV or HBV. Similar to gender, donation type had no significant association with any of the outcomes described, which contrasted with the results of Jaques et al, who showed replacement donations were more likely to be HBV-or HIVinfected than spontaneous donations. 13 In all the analyses conducted, subjectively deferred donors had higher odds of a positive screening test than donors who confidentially excluded their blood. Of note, when multivariate regression models were performed, opting for CUE was not a statistically significant risk factor for positive general serologic screening, HIV serology, HBV serology, or syphilis serology. Although other studies found an association of CUE with a higher prevalence of reactive markers for HIV, HBV, and syphilis, they also showed CUE to have a limited effect on reducing TTIs secondary to window-period donations. 5,18,19 Also, CUE may be a confusing method, with donors often opting for CUE by mistake, 20 and rarely prevents the transfusion of units from donors with undisclosed high-risk deferrable behavior. 21 Besides, while SDD highlights the importance of the healthcare worker's experience and perspicacity for blood safety assurance, CUE use may reduce the perceived responsibility of staff on eliciting a history of high-risk behavior. 22 Finally, SDD represents no additional costs for BTCs. 4 With that in mind, the SDD strategy, which is only used in some Brazilian BTCs, could be a better tool for blood safety improvement than CUE, which is already implemented in many BTCs worldwide. blockade in the BTC system due to a reactive serology, and collateral information from a third-party acquainted with the donor. 4 However, due to its subjective nature, the SDD tool is difficult to standardize, creating the need for further studies to define the criteria used by the interviewers more clearly.
Despite the large sample size and robust statistics, the present study has several limitations. Due to its strong regional focus, the results may be too population-specific. Thus, a multicentric study is needed to assess further the potential of SDD for preventing TTIs. In addition, direct interference with the amount of collected bags used for transfusion may be a major drawback. Considering that CUE has been shown to cause a low but consistent discard rate of safe units, 21 it is essential to analyze whether SDD can cause a similar effect in future studies. Finally, there are some ethical standpoints to consider: Does the method reflects the staff's inability to develop empathy and confidentiality with the donor? In order to reduce bias, should more than one healthcare professional decide if the donor's blood ought to be discarded? As the current SDD policy does not demand informing donors of the interviewers' decision, is it ethical to subject the individuals to the illusion and risks of donating if the blood will not be transfused?
Nevertheless, in an era in which computerized questionnaires and computer-assisted self-interviews are increasingly discussed and adopted as substitutes for the traditional pre-donation interviews, 23,24 the present study highlights the importance of the healthcare worker experience and face-to-face contact with the prospective donor for assurance of blood safety, especially in low-income regions where associated costs and technical complexity may limit the use of modern molecular testing methods for the prevention of TTIs secondary to window-period donations. 25

| CONCLUSION
The implementation of SDD as an adjunct tool for clinical triage of blood donors may be an effective way to increase the detection of individuals at risk of transmitting infectious diseases, such as HIV, HBV, and syphilis.
The possibility of detecting donors in the window period for the cited infections may reduce TTIs in the recipients. However, further studies on SDD are needed to confirm this assertion.