|Year : 2020 | Volume
| Issue : 2 | Page : 150-154
Postulating a positive predictive value for HIV, HBV, and HCV screening tests using electrochemiluminescent assay by comparing with ID-NAT results
Samrat Thapa, Ankit Mathur, Latha Jagannathan, RP Prashanth
Rotary Bangalore-TTK Blood Bank, Bangalore Medical Services Trust, Bengaluru, Karnataka, India
|Date of Submission||03-Jun-2020|
|Date of Decision||08-Aug-2020|
|Date of Acceptance||14-Sep-2020|
|Date of Web Publication||13-Nov-2020|
Rotary Bangalore-TTK Blood Bank, Bangalore Medical Services Trust, Bengaluru, Karnataka
Source of Support: None, Conflict of Interest: None
Background: In India, blood units reactive by the initial screening assays are discarded. A significant number of these are falsely positive, leading to wastage of precious blood. Also of importance is the dilemma created when it comes to donor disclosure. Hence, it is important to evaluate the positive predictive value (PPV) of the initial screening assay to identify false-positive screening results and implement a more balanced algorithm for donor disclosure and counseling. The study was undertaken with an objective of comparing signal-to-cutoff ratio (S/Co) of positive results in HBV, HCV, and HIV screening of electrochemiluminescent (ECI) assay with the results of ID-NAT so as to predict a S/Co ratio of indicative of true positives. Methods: The retrospective study was conducted in the transfusion-transmissible infection Laboratory in a Regional Blood Transfusion Centre in South India. Consecutive blood donors from January 2015 to December 2019 were evaluated simultaneously for HBV, HCV, and HIV by ECI (Vitros 3600, Ortho Clinical Diagnostics) and ID-NAT (Procleix Panther System by Grifols), and the positive results either by one or both were analyzed. Results: A total of 163,596 voluntary donor samples were evaluated. In case of HBV, a S/Co ratio of 1000 predicts a true positive more than 94% of the time. In case of HCV, a S/Co ratio of 8 predicts a true positive more than 95% of the time. Finally, a S/Co ratio of 30 predicts a true positive as well as true negative 100% of the time in case of HIV. Conclusions: It is challenging to propose a standard S/Co ratio because the performance of each assay varies across laboratories. However, the study throws light on the role of the S/Co ratio in reducing the perplexity during donor notification.
Keywords: HBV, HCV, HIV, positive predictive value, transfusion-transmissible infection screening
|How to cite this article:|
Thapa S, Mathur A, Jagannathan L, Prashanth R P. Postulating a positive predictive value for HIV, HBV, and HCV screening tests using electrochemiluminescent assay by comparing with ID-NAT results. Glob J Transfus Med 2020;5:150-4
|How to cite this URL:|
Thapa S, Mathur A, Jagannathan L, Prashanth R P. Postulating a positive predictive value for HIV, HBV, and HCV screening tests using electrochemiluminescent assay by comparing with ID-NAT results. Glob J Transfus Med [serial online] 2020 [cited 2020 Nov 27];5:150-4. Available from: https://www.gjtmonline.com/text.asp?2020/5/2/150/300628
| Introduction|| |
Screening for transfusion-transmissible infections (TTIs) to exclude blood donations at risk of transmitting infection from donors to recipients is a critical part of the process of ensuring that transfusion is as safe as possible. Effective screening for evidence of the presence of the most common and dangerous TTIs can reduce the risk of transmission to very low levels. Blood transfusion services have, therefore, established efficient systems to ensure that all donated blood is correctly screened for specific TTIs and that only nonreactive blood and blood components are released for clinical and manufacturing use.
The choice of specific markers to screen for depends on the infectious agent that we are screening for. Screening for specific antibody is usually most appropriate for HIV and HCV, whereas screening for specific antigen (HBsAg) is necessary for HBV. Three main kinds of primary screening assay available to detect infectious disease markers in donated blood are:
- Enzyme-linked immunosorbent assays/enzyme immunoassay (EIA)
- Particle agglutination assays
- Simple rapid assays.
The end result of an EIA is a set of numbers – the optical density (OD) values – which then have to be converted to positive and negative results. A further simple calculation is made which converts the individual OD values to a standard ratio. This standard ratio is called the signal/cutoff ratio (S/Co) and is calculated by dividing the individual sample OD value by the calculated cutoff value. A value below 1 indicates a nonreactive result and a value above 1indicates a reactive result.
In India, blood screening involves a single test with the resultant action, which means the release or discard of the donation arising from that single test alone. The blood units reactive by the initial screening assays are discarded, and a significant number of these are falsely positive, leading to wastage of precious blood. Despite financial constraints, a large number of blood bank in India supplement their testing algorithm with nucleic acid amplification test (NAT) since a single test result alone is not relied upon to determine infection or subsequent actions. As of now, initial reactive units must be discarded, taking recipient's safety into consideration. However, in case of false positives, it creates a state of perplexity during donor notification. Thus, an accurate interpretation of TTI results is needed to implement adequate counseling.
Drawing inspiration from the Centers for Disease Control and Prevention (CDC) recommendation of cutoff values for HCV screening assay for donor counseling, and based on our own experience and observations, an attempt has been made to identify cutoff value for HIV and HBV as well, in addition to HCV. We performed TTI screening both by primary assay and NAT testing in case of HBV and HCV. In case of HIV, beside primary assay and NAT testing, screening was done additionally by rapid assay as well. On the basis of the combined results of all these assays, positive predictive value (PPV) was formulated.
In this study, PPV refers to a S/Co ratio that represents the cutoff value for positivity and negativity of the infection based on ECI as well as ID-NAT results.
Aims and objectives
The study was undertaken with an objective of comparing S/Co of positive results in HBV, HCV, and HIV screening of electrochemiluminescent (ECI) assay with the results of ID-NAT so as to predict a S/Co ratio of indicative of true positives.
| Materials and Methods|| |
A retrospective analysis of 163,596 voluntary donors' data recorded from 2015 to 2019 was conducted in the TTI Laboratory of a Regional Blood Transfusion Centre in Karnataka, South India. Blood donors were screened for HBV, HCV, and HIV by ECI (Vitros 3600, Ortho Clinical Diagnostics) and simultaneously by ID-NAT (Procleix Panther System by Grifols). As per our institution's policy, an external control: Virotol-1 (Biorad) is run along with the kit controls on the day of run. Furthermore, the institution takes part in proficiency testing programs such as BEQAS (SDMH, Jaipur) and TEQAS (CMC, Vellore).
The study was approved by the institutional ethical committee.
The positive results either by one or both were recorded on Microsoft Excel and analyzed for all three TTIs for the following parameters:
- A total number of serology reactive: samples reactive by ECI (regardless of ID-NAT result)
- ID-NAT yield: Samples reactive by ID-NAT and nonreactive by ECI
- Serology yield: Samples reactive by ECI and nonreactive by ID-NAT
- ECI-NAT concordance: Samples reactive by both ECI and ID-NAT OR total ECI/serology reactive – ECI/serology yield
- ECI-NAT discordance: ECI/serology yield together with ID-NAT yield OR Serology yield + ID-NAT Yield.
The PPV was calculated by the total number of ECI-NAT concordance and ECI-NAT discordance result based on S/Co ratio as follows:
- >1000 and <1000, in HBV
- >8 and <8, in HCV
- >50, 30–50, and <30, in HIV.
For HIV, we also carried our supplementary testing using rapid kit (HIV TRI-DOT Diagnostic Enterprises) and compared the results with the abovementioned assays.
| Results|| |
For HBV, out of 163,596 units, 721 (0.44%) were reactive for ECI and NAT as well, of which 35 (0.02%) were NAT yield and 61 (0.04%) sero yield. ECI-NAT was concordant for 625 units, of which 594 (95.04%) units had S/Co ratio >1000, and of the 96 units which were discordant, 44 (72.13%) units had S/Co ratio <1000 (see [Table 1]).
|Table 1: Distribution of hepatitis B virus reactive samples assayed with electrochemiluminescent and nucleic acid amplification test|
Click here to view
For HCV, out of 163,596 units, 670 (0.40%) were reactive for ECI and NAT as well, of which 2 (0.001%) were NAT yield and 648 (0.39%) sero yield. ECI-NAT was concordant for 20 units, of which 19 (95%) units had S/Co ratio >8, and of the 650 units which were discordant, 592 (91.4%) units had S/Co ratio <8 (see [Table 2]).
|Table 2: Distribution of hepatitis C virus reactive samples assayed with electrochemiluminescent and nucleic acid amplification test|
Click here to view
For HIV, a total of 423 (0.25%) units were found to be reactive for ECI and NAT as well, of which two (0.001%) were NAT yield and 374 (0.22%) sero yield. ECI-NAT-rapid was found to be concordant for 47 units, of which 42 (95.2%) units had S/Co ratio >50 (89.36%) and 5 (10.63%) units had S/Co ratio in between 30 and 50 and not a single unit had S/Co <30.
Similarly, ECI-NAT discordance was found in 376 units, of which 374 (100%) units had S/co ratio <30 (see [Table 3]).
|Table 3: Distribution of HIV reactive samples assayed with electrochemiluminescent, nucleic acid amplification test, and rapid kit|
Click here to view
| Discussion|| |
Blood screening along with supplementary testing enables the identification of infected donors or donors with nonspecific reactivity or inconclusive results. However, the blood transfusion service has a duty of care to donors, their families, and the general population to ensure that infected individuals are referred for appropriate counseling, treatment, and further management as they may infect other individuals if they are not aware of their status. The Blood Transfusion service (BTS) and relevant authorities should have a clear policy and systems for communicating with these donors and informing them of their status to minimize any risk of further transmission. Donors who test negative for TTIs should be encouraged to donate regularly and lead low-risk lifestyles.
Informing donors that they are confirmed positive for an infection clearly poses sensitive issues, and donors need to be counseled on the results and the actions that should subsequently be taken. Donor notification of nonspecific reactivity or false positive is highly problematical and should be undertaken with care because this reactivity often varies and usually does not have any impact on the actual health of the individuals. Clear policies on the handling of nonspecifically reactive donors are essential. The permanent deferral of these donors is sometimes considered to be unnecessary but may be unavoidable unless policies and procedures are in place that recognize variable nonspecific reactivity and facilitate the appropriate management of such donors.,
Based on our experience with Vitros 3600, it was observed that, for each TTI assay, there was a cutoff which predicted the result congruent to ID-NAT. The cutoff figures (i.e., 1000 for HBV, 8 for HCV, and 30 for HIV) were agreed upon purely based on our observations for years. For example, in case of HBV, we noticed that ID-NAT results were the positive maximum number of times when the S/Co ratio in ECI was above 1000 and negative when the S/Co ratio was below 1000. Furthermore, the CDC on its website had proposed a S/Co ratios for Ortho CIA predictive of a true positive ≥ 95% of the time. With these information along with several other literatures available on similar concept of predicting course of infection based on screening test results, we investigated and analyzed the data present in our laboratory to evaluate PPV of screening assays so that false-negative results could be identified and an efficient algorithm can be developed for donor notification and counseling.
According to our study, in case of HBV, a S/Co ratio of 1000 predicts a true positive more than 95% of the time and a true negative more than 72% of the time. Similarly, in case of HCV, a S/Co ratio of 8 predicts a true positive more than 95% of the time, as suggested by the CDC, and a true negative more than 91% of the time. Finally, a S/Co ratio of 30 predicts a true positive as well as true negative 100% of the time in case of HIV.
| Conclusions|| |
It may not be possible at present to conclusively determine an infection by NAT only or by the S/Co ratio. Furthermore, it is challenging to propose a standard S/Co ratio because the performance of each assay varies across laboratories. However, this study is significant because it sought to throw light on the role of the S/Co ratio by comparing it to the ID-NAT.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
World Health Organization. Screening Donated Blood for Transfusion-Transmissible Infections: Recommendations. Geneva, Switzerland: World Health Organization; 2009.
World Health Organization. Safe Blood and Blood Products. Module 2. Screening for HIV and Other Infectious Disease. Geneva, Switzerland: World Health Organization; 2009.
Sommese L, Iannone C, Cacciatore F, De Iorio G, Napoli C. Comparison between screening and confirmatory serological assays in blood donors in a region of South Italy. J Clin Lab Anal 2014;28:198-203.
Centers for Disease Control and Prevention. Hepatitis C information for Health Professionals: Laboratory Testing. Atlanta, GA: Centers for Disease Control and Prevention; 2011. Available from: https://www.cdc.gov/hepatitis/hcv/index.html
[Last accessed on 2020 Oct 31].
Choi MS, Lee K, Hong YJ, Song EY, Kim DS, Song J. The role of the signal-to-cutoff ratio in automated anti-hcv chemiluminescent immunoassays by referring to the nucleic acid amplification test and the recombinant immunoblot assay. Ann Lab Med 2018;38:466-72.
[Table 1], [Table 2], [Table 3]