HLA alloimmunization in patients receiving multitransfusions of red blood cells.

HLA antibodies were studied in 88 patients with chronic hemolytic anemia who received multitransfusions of red blood cells prepared by conventional (PRC-C), inverted centrifugation (LR-I) and leukocyte filter (LR-F) techniques. Their mean age was 8 years and 4 months with a duration of transfusion of 8 years. The patients were divided into five groups: group 1, receiving PRC-C (n=20); group 2, receiving LR-I (n=33); group 3, receiving LR-F (n=11); group 4, subsequently receiving LR-I and LR-F (n=10); and group 5, receiving PRC-C followed by LR-I and LR-F (n=14). The HLA class I antibodies were found in 30 out of 88 patients (34%). All were against HLA antigens commonly found in the Thai population. The patients receiving PRC-C exhibited HLA antibodies of 65%, which was significantly higher than those of patients receiving LR-I (24%) and LR-F (0%). Consequently, the transfusion reactions of fever, chill, rash and urticaria were also commonly found in patients receiving PRC-C (13.4%), which was significantly higher than patients receiving LR-I (0.4%) and LR-F (0%). The leukocyte filter technique has been shown to be effective in preventing HLA alloimmunization and transfusion reactions but the price is rather high. For the inverted centrifugation technique, only transfusion reactions were effectively prevented and the HLA alloimmunization continued to develop. A more effective and low-cost method for the removal of leukocytes should be investigated for these multitransfusion patients.

Preliminary study of HLA-ABCDR antigens in CML, ANLL, thalassemia and severe aplastic anemia in Thais.

The objective of this study was to analyse human leukocyte antigen (HLA) and disease association in common blood diseases [chronic myelogenous leukemia (CML), acute nonlymphocytic leukemia (ANLL), thalassemia and severe aplastic anemia] in Thais. The subjects were patients from the Hematological Clinic, Departments of Medicine and Pediatrics, Ramathibodi Hospital who were referred for HLA typing for bone marrow transplantation (BMT) at the Histocompatibility Laboratory from March 1988 to September 1997. A total of 129 patients had complete HLA-ABC typing. The patients included 45 CML, 40 ANLL, 26 thalassemia (Thal) and 18 severe aplastic anemia (SAA). Of these, 88 patients were typed for HLA class II. The HLA class I (ABC) and II (DR, DQ) typings were performed by microlymphocytotoxicity test. It was found that HLA class I was associated with CML, ANLL and Thal, whereas, HLA class II was associated with SAA. HLA-B8 and HLA-B18 were increased in CML with R.R. values of 12.2 and 3.9, respectively, whereas, HLA-B18 was increased in ANLL with R.R. value of 4.5. In addition, HLA-DR2 and DR3 were increased in SAA with R.R. values of 3.8 and 4.8, respectively. For Thal, HLA-A2 and B46 were increased in Thal in Central Thais with R.R. values of 3.3 and 6.1, respectively, whereas, HLA-B13 was increased in Thal in Northern Thais with R.R. value of 8.5. On the other hand, HLA-B7 was absent in CML. HLA-Cw7 was decreased in CML and SAA, whereas, HLA-DR6 was decreased in ANLL and SAA. Furthermore, HLA-Cw6 was also decreased in CML, whereas, HLA-A33 and Bw4 were decreased in SAA. Although the sample size of each disease was small, the increase of HLA-DR2 was observed in SAA in Thais which was similar to other studies in different ethnic groups. These preliminary data may be useful for further study in HLA and blood disease association.

Lewis antigens on newborn red cells.

This study aimed to screen for Lewis antigens in Thai newborns. Although, these antigens are known to be weak or absent on the red cells of newborns, we encountered a case of a Le(a+) newborn baby when testing with monoclonal antibody and human anti-Le(a). Such a finding led us to conduct this study to explore further evidence of Lewis antigens in Thai newborn red cells. A total of 197 cord blood samples were tested with monoclonal anti-Le(a) and anti-Le(b). (Bioclone, Ortho Diagnostic Systems, USA). The tests were performed according to the manufacturer's recommendations. The results revealed that none of the cord red cells in this study group possessed Lewis antigens. This study showed that Lewis antigens were absent or were so extremely weak on the red cells of these newborn infants that they could not be demonstrated despite the use of potent monoclonal antibodies. However, further study should be done by using more cord blood samples, a more sensitive technique or even more potent antisera.

The organization of Kidney Transplantation Services at Ramathibodi Hospital: fourteen years experience on waiting list, kidney donors and kidney transplantation.

The Kidney Transplantation Program at Ramathibodi Hospital was established in 1985. By the end of 1998, there were 1,614 patients on the cumulative waiting list. The first kidney transplantation (KT) was started in 1986 by using kidney from living-related donor (LD) while cadaveric KT (CD-KT) was started in 1987. A total of 528 KT were done, 278 cases (52.7%) were CD-KT and 250 cases (47.3%) were LD-KT. Six patients had two kidney transplants. 278 kidneys were donated from 189 cadaveric donors. Fifty cadaveric donors (26.4%) were from Ramathibodi Hospital while the rest were from other hospitals and the Organ Donation Center, Thai Red Cross Society. For LD, 233 out of 250 (93.2%) were from living-related, more than 50 per cent of these donors were from siblings. 17 spousal donors have been accepted for KT at Ramathibodi Hospital since 1997. Concerning the recipient pools, 522 patients (32.3%) were transplanted, 123 patients (7.6%) died without KT, 111 patients (6.9%) underwent KT at other hospitals, and 78 patients (4.8%) changed to waiting lists at other hospitals. The rest were lost to follow-up. At present, only 265 patients are still actively waiting (send serum every month). The number of KT and living donors has gradually increased, whereas, the number of cadaveric donors has decreased. However, cooperation with the "Organ Donation Center" has improved the number of cadaveric donation in the last two years. Sufficient organ donations and an active working team will provide a good kidney transplant service for the patients.

Low cost locally prepared fibrin glue for clinical applications: reported of 145 cases. Committee of Bangkok International Hemophilia Training Center.

Fibrin glue (FG) is one of the blood products known to be very useful for local hemostatic measure and as a medically valuable tool for adhesion, sealing, anastomosis, repair microvascular and nerve grafts in medical and surgical procedures. Before 1996, FG was used to a limited extent in Thailand due to the high cost. Technology for locally prepared FG was transferred to Bangkok International Hemophilia Training Center of the World Federation of Hemophilia (IHTC-WFH) in July 1996 by Prof. Uri Martinowitz and the late Prof. Henri Horoszowski. Since then FG has been widely used and proved to be very useful in Thailand. This paper reports 145 cases using low cost locally prepared FG at Ramathibodi Hospital during November 1996 to December 1997. A total of 145 cases with age range from 5 months to 73 years, which included 55 pediatrics and 90 adults, 100 males and 45 females. The amount of FG used was 1-80 ml per case. Clinical procedures included dental surgery (46), open heart surgery (35), ENT (28), orthopedic (13) including 2-3 joint correction in one session in 2 hemophiliacs, neurology (11), plastic repair (7), liver (2) and severe bleeding in dengue hemorrhagic fever (3). Forty-seven cases had hemostatic disorders. The result of local hemostatic, adhesive and sealant effect of FG was satisfactory with no complications. In open heart surgery, the amount of content in chest drain decreased and none required reopen-surgery to stop bleeding. Dental surgery was performed in 43 patients with bleeding disorders i.e. hemophilia, idiopathic thrombocytopenic purpura, leukemia, severe thrombocytopenia, patients on anticoagulant, etc. Only 3 cases (7%) required blood component compared to all of the 50 no-FG controlled cases (100%) that required blood component therapy. FG has proved to be very useful in many aspects i.e. minimizing blood product usage, decreasing medical workload, reducing medical cost and increasing patients' convenience and satisfaction in particular.

Detection of HBV genome by gene amplification method in HBsAg negative blood donors.

The incidence of post-transfusion hepatitis has been reduced greatly by screening blood donors for hepatitis B surface antigen (HBsAg). However, hepatitis B virus infection still accounts for a certain number of cases of post-transfusion hepatitis. The purpose of this study was to detect HBV DNA in the HBsAg negative blood samples by using nested PCR with two primer pairs specific to core region. Two hundreds blood samples from HBsAg negative donors, and 14 samples from HBsAg positive donors were provided by the blood bank of Ramathibodi Hospital. The results showed that HBV DNA was detected in all 14 HBsAg positive blood samples and in 7 (3.5%) of 200 HBsAg negative blood samples. This study showed that the absence of HBsAg in otherwise apparently healthy individuals may not be enough to ensure lack of circulating HBV. The more sensitive ELISA technique is still in need. Otherwise, the safety of blood transfusion can be enhanced by careful selection of blood donors and careful consideration of risks and benefit of the patients who need blood transfusions.

Eluding transfusion-transmitted HIV infection: report of two cases.

Two hemophilia A boys (FVIII: C < 1% and 2.2%), whose ages were 12 and 14 years. old, received fresh frozen plasma of 140 ml and 210 ml, respectively, in 1989. It was the 27th and 13th donation for each regular donor who was negative for anti-HIV testing. However, both donors had HIV seroconversion within 95 to 110 days after the last donation. They might have contracted HIV infection shortly after the last donation. Luckily, the two hemophiliac recipients are still in good health and negative for anti-HIV and HIV-antigen testings for 7 years.

Effectiveness of leukocyte removal by Imugard III.

Study was conducted to evaluate the effectiveness of Imugard III-RC (4B) to remove leukocyte from packed red cells (PRCs). The leukafiltration set, Imugard III (4B), bedside use for PRCs or whole blood was converted to laboratory use by connecting the distal end of filter tubing to the transfer bag and the proximal end of tubing to the primary blood pack by sterile connecting device to ensure that the closed system is intact. Twenty units of PRCs were prepared from CPD whole blood by removal of platelet rich plasma. All units were stored at 4 degrees C for 1 day before filtration. Volumes of PRC/Bag ranged from 240-340 ml, mean = 276.75 ml and SD = 26.57. Prefiltration WBC/Bag ranged from 464 x 10(6)-5910 x 10(6) cells, mean = 2862.05 x 10(6) cells and SD = 1280.87 x 10(6). The filtration time/Bag ranged from 18-45 min, mean = 27.3 min and SD = 7.9. The WBC removals ranged from 99.99 per cent-100 per cent, mean = 99.99 per cent and SD = 0. The residual WBC in PRC/Bag ranged from 0-0.24 x 10(6) cells, mean = 0.134 x 10(6) cells and SD = 0.063 x 10(6). The RBC loss from PRCs with no additive solution was approximately 15 per cent. There was no need for priming and purging the filtration set with saline before and after filtration. CONCLUSION: The reduction of WBC by 20 sets of Imugard III-RC (4B) was 99.99 per cent or higher with the residual WBC < 1 x 10(6) which is able to reduce nonhemolytic transfusion reaction, CMV transmission and HLA alloimmunization.

Panel reactive antibodies in post-kidney transplantation.

This study was aimed to evaluate the clinical relevance of the panel reactive antibodies (PRA) post kidney transplantation (KT). A total number of 90 KT recipients consisted of 71 male and 19 female patients. Thirty-two haploidentical and 3 HLA-identical pairs for living related KT and 55 cadaveric KT with 3-6 mismatched antigens were included in this study. The analysis revealed that there were 2 out of 69 (2.89%) patients with no episode of rejection who had Pre-KT PRA-T and or PRA-B > 80 per cent while they were 5.79 per cent and 23.19 per cent for Post-KT. No patient in 21 cases with KT rejection had Pre-KT-PRA-T and -B > 80 per cent. There was significant increase of antibodies in Post-KT rejections which were 28.57 per cent and 33.3 per cent for Post-KT-PRA-T and -B respectively. None of 3 cases with graft failure (GF) from chronic rejection had Pre-KT-PRA-T and -B > 20 per cent and only one of them had Post-KT-PRA-T = 80 per cent. No donor specific HLA antibody was found among this group of patients. Although antibody to donor HLA antigens was not observed in these patients, the increase of PRA-T and -B in Post-KT may indicate the immunological reaction resulting in GF.

A*02 in southern Thai Muslims and central Thais.

The HLA-A*02 subtyping in Thais was conducted and included in the 12th International Histocompatibility Workshop (12WS). A total of 81 randomized individuals previously serologically or DNA typed as A2 were studied for A2 subtypings. The subjects consisted of 32 Southern Thai-Muslims (STM) and 49 Central Thais (CT). The 12WS HLA-A*02 subtyping DNA typing kit was employed. The most common A*02 subtypes in STM were A*0203,*0201 and *0207 while they were A*0203, *0207 and *0201 in CT. A*0202, *0204, *0208, *0209, *0212, *0213, *0214, *0215, *0216 and *0217 were not found in both STM and CT. The 12WS data indicated that A*0201 was also the most frequent allele of A*2 among North-East Asians. A2 subtype study in 32 STM revealed that 2 in 8 of A*0201 showed the absence of bands at 813 bp and 705 bp with primer mix number 03A and 517A and weak reaction band with primer mix number 33A. In addition, 3 subjects with A*0201 variations have one nucleotide difference in exon 2 by sequence base typing (by MGJ. Tilanus) which will be reported separately. CONCLUSION: More variations of A*02 were observed among STM compared to CT. The variations of reactions with the set of primer mix should be carefully observed and subjected to further analysis.

Crossmatching technique facilitating kidney transplantation.

Accelerated acute cellular rejection (AR) continues to be a serious problem in kidney transplantation (KT), suggesting that undetected presensitization may be encountered. The purpose of this study was to determine the most sensitive crossmatching (XM) technique to detect the preformed antibody (Ab) which may cause AR. One hundred and twenty two sera from 98 patients, on the waiting list for KT at Ramathibodi Hospital were XMed with 23 cadaveric splenic lymphocytes including 2 living related KT (LR-KT). The XM was performed by 3 different techniques namely, standard microlymphocytotoxicity test (standard NIH), antihuman globulin microlymphocytotoxicity test (AHG) and flow cytometric XM (FCXM). The XM results revealed that 8 out of 75 (10.7%) tests were negative by standard NIH, i.e., 5 tests were positive by AHG only and 1 test was positive by FCXM only and 2 tests were positive by both AHG and FCXM. In addition, the patients who had the AHG technique were not done, 5 out of 47 (10.7%) tests were also negative by standard NIH but were positive by FCXM. The sensitivity of the techniques was done by titrations of anti HLA-A2. It was found that FCXM was the most sensitive technique, followed by AHG and standard NIH, consecutively. In the retrospective study of LR-KT, case #1, the standard NIH for XM using pre-KT blood sample was negative while AHG and FCXM were strongly positive. The patient had AR at day 2 post-KT which confirmed by needle biopsy. The serum at day 11 and day 116 post-KT were tested again and were positive by the 3 techniques. Case #2, pre-KT blood sample showed negative T-XM by the 3 techniques while auto-B and B-XM were positive by standard NIH and AHG but negative by FCXM. This patient had rejection at day 16 after KT. The post-KT blood sample at day 30 showed positive auto T/B and T/B-XM by standard NIH and AHG whereas it was still negative by FCXM. It was also noted that Ab to donor B cell was better detected by standard NIH and AHG than FCXM. In conclusion, FCXM is more sensitive than standard NIH and AHG, however this technique is limited in detecting IgM T and B cell Ab. AHG technique can detect both IgG and IgM antidonor T and B cell Abs. In addition, AHG technique is more sensitive than standard NIH and does not require sophisticated equipment. AHG technique should be appropriate for routine XM, especially, in LR-KT and sensitized patients.

Increasing the chance to accept HLA-ABDR mismatched donor in kidney transplantation.

Two hundred and fifty-three kidney transplantations (KT) which included 68 (26.9%) living-related (L) and 185 (73.1%) cadaveric (C) KT with 0-6 HLA-ABDR mismatches (MM) were studied for the association of HLA-ABDR-MM specificities and the occurrence of graft rejection (GR). It was found that the incidence of acute and chronic rejection in CKT was significantly higher than that of LKT (42.1% vs 22.1%, p < 0.005). It was also observed that the number of ABDR-MM, AB-MM and BDR-MM which is important in GR were 2 times in CKT compared with LKT. The analysis revealed that HLA-A11, B16, B22, B35, B5, B17 and DR3 were good responders, whereas, HLA-A30, A2, B62, B18, B40, B44, B46 and DR10 were good stimulators for KT. GR were significantly increased with p < 0.01 and < 0.05, respectively. Specific HLA-MM specificities played a significant role in GR, i.e., some HLA-MM specificities were permissible, whereas, some were immunogenic. Careful selection of donor and recipient for KT by avoiding immunogenic HLA-MM and/or accepting permissible HLA-MM will improve graft survival and reduce the demand of kidney for retransplantation.

HLA in southern Thai-Muslims.

One hundred and two Southern Thai-Muslims (STM) from Nakhon Si Thammarat province were studied for HLA class I and II by SSP ARMS-PCR and PCR-SSO, respectively. The allele frequencies, haplotype frequencies, delta value and linkage disequilibrium between alleles were expressed. The most frequent alleles for HLA-A, HLA-B and HLA-C were A*24(02,03), A*11 (01,02), A*02(01,03,05-07,11): B*15(01,04-07,12,19,20), B*07(02-05), B*51(01-05)/B*52 (011,012); and Cw*07(01-03), Cw*04(01,02), Cw*08(01-03), respectively. The HLA class II alleles frequently found were DRB1*1202, DRB1*15021, DRB1*0701; DRB3*0301; DRB5* 0101; DQA1*0101, DQA1*0103, DQA1*0601; DQB1*0301, DQB1*0501, DQB1*0201; and DPB1*1301, DPB1*2301 and DPB1*0501. Two common HLA class I and II haplotypes with significant linkage disequilibrium were A*24 (02,03)-Cw*08 (01-03)-B*15 (01,04-07,12,19,20) -DRB1*1202 and A*33 (01,02)-Cw*0302-B*5801-DQB1*0201. The absence of B*27 and DRB1 *1401, the presence of A*2301 and high frequency of A*68 were observed in STM. Conclusion: Certain level of genetically distinction among STM, CT and NET existed. However, the genetic diversity of STM was relatively closer to CT than NET.

HTLV-I antibody screening in donated blood and thalassemic patients.

The HTLV-I antibody screening by passive partial agglutination test, Serodia HTLV-I (Fuji-rebio Inc) was performed in a total of 1,225 male and 468 female blood donors from September 1992 to March 1993. In addition, a total of 64 multitransfused thalassemic patients were also studied. All of the sera from donors and patients in both study groups showed no reactive results for HTLV-I antibody. This finding indicated that the HTLV-I infection is not yet a problem in blood transfusion in Thailand. However, the scattered reports of HTLV-I infections from our region should urge the national authority on the surveillance of this infection by epidemiological survey and occasional screening in blood donors.

Transmission of HIV infection by seronegative blood in Thailand.

In Thailand, the prevalence of HIV infection in the general population and in donor blood (DB) has sharply increased since 1987. The HIV seropositive rate in DB was increased from 0.0065 per cent in 1987 to 0.95 per cent in 1993 (150 times in 6 years). Heterosexual transmission is the major route of spreading. Therefore, HIV seronegative blood (SNB) poses significant hazard to the recipients because of the risk of viraemia during the window period of early HIV infection. In Thailand HIV Ab screening in all units of blood was started in 1987 and was compulsory nationwide in early 1989. Donor self exclusion (DSE) has been implemented since 1990. It is not fully effective in the prevention of transfusion associated AIDS (TAA) because of many limiting factors. However, DSE should be promoted to practice in every blood bank particularly those that can not do HIV Ag screening. During 1990-1992, there were 30 reported cases of TAA by SNB. The study of risk figure (HIV Ag positive-neutralization with HIV Ab negative) in DB was 1:3,400 and 1:10,000 in two reports in 1991. Under all these circumstances, the national AIDS committee has documented the policy to do HIV Ag screening in every unit of blood from August 1991 and allocated a 10 million baht budget (year 1992) for Ag testing. Several hospitals and NBC reported the risk figures which varied from 1:3,400 to 1:25,000. A certain amount of blood is processed to 2-4 blood components given to 2-4 patients which will increase the number of TAA by SNB.(ABSTRACT TRUNCATED AT 250 WORDS)

Response of hemophilia A with bleeding to fresh dry plasma.

Fresh dry plasma (FDP) is a lyophilized form of fresh frozen plasma (FFP) which can be stored at 4 degrees C for one year. One bottle of FDP is prepared from 220 ml of FFP and contains FVIII: C 0.75 +/- 0.3 U/ml (mean +/- 1 SD). This study describes the clinical and laboratory response in 7 severe and 4 moderate hemophilia A patients. The age ranged from 7-17 years (mean +/- SD = 11.7 +/- 2.9 years). Either 12.4 ml/kg. FDP or 12.2 ml/kg FFP was transfused to the patients when they had bleeding episodes such as hemarthrosis, hematoma. 16 episodes of FDP and 12 episodes of FFP transfusion were studied. The bleeding could be effectively controlled by FDP or FFP except one patient who had physical therapy at 8 hours post FDP transfusion. The increment of FVIII:C was 14.1 +/- 5.3% at 30 minutes after FDP transfusion and 12.1 +/- 3.7% at 30 minutes after FFP transfusion. The recovery rate was 83.2 +/- 32.6% in FDP and 65.3 +/- 22.7% in FFP transfusion. The FVIII:C was decreased to 78.9 +/- 12.3%, 55.6 +/- 13% and 16.3% of the initial level at 2, 8, 24 hours after FDP transfusion respectively which were not statistically significantly different from FFP transfusion. No serious complication was found. FDP will replace FFP for the treatment of coagulation disorders such as hemophilia A. It is an useful alternative therapy which can be provided to the hemophiliac patients in the rural area in developing countries.

The value of HLA-ABC common typing tray in relation to bone marrow transplantation.

Bone marrow transplantation (BMT) is one of the most effective procedures to cure the previously uncured hematologic diseases. However, it is costly and HLA typing to select the compatible donors contributed to its cost. A total of 53 prospective patients for BMT and their 114 siblings were analyzed to evaluate the use of locally prepared HLA-ABC common typing tray (ABCCT) during Mar 1988-Mar 1992. The 16, 9, 7, 5, 5 and 12 patients were diagnosed as aplastic anemia, CML, thalassemia, ALL, ANLL and other blood diseases, respectively. It was found that 18 patients were HLA-identical (HLA-ID) with one of their siblings except one patient had 2 HLA-ID sibs. All of those who appeared to be HLA-ID were further tested for the HLA-ABCDR typings. It was observed that 16 (88.89%) of 18 patients and 17 (89.47%) of 19 sibs were confirmed as HLA-ID. After careful clinical screening, only 13 HLA-ID pairs were able to proceed to the mixed lymphocyte culture and confirmed their status of HLA-ID by this test. Finally, only 6 (46.15%) of 13 patients received BMT with a high rate of success, ie all patients have survived with bone marrow engraftment. Thus, ABCCT is very useful for related BMT. It was highly efficient to exclude HLA-non-ID and haplo-ID yet the cost and workload were greatly reduced.

HIV seroconversion in Thai hemophiliacs up to 1991.

In Thailand, the anti-HIV screening in the donor blood was started in 1987 and was compulsory nationwide in February 1989. Sixty-six hemophilia A and 10 hemophilia B patients who received approximately six million units of factor VIII and IX in the form of fresh frozen plasma, frozen cryoprecipitate, cryoprecipitate removed plasma, fresh dry plasma and factor concentrate during 1976 to 1991 were tested for anti-HIV since 1987. The age ranged from 1-39 year (mean +/- SD = 15 +/- 7.3). The anti-HIV test was performed by ELISA and/or gel agglutination and confirmed by Western blot analysis. The patients would be checked 1-2 times per year and as necessary. A total of 174 tests for the first, second, third, fourth, fifth and sixth tests were studied in 76, 49, 27, 14, 5 and 3 patients respectively during 1987 to 1991. The prevalence of HIV seroconversion in the year 1987, 1988, 1989, 1990 and 1991 was 2.2% (1/45), 1.9% (1/53), 1.6% (1/63), 1.5% (1/67) and 3.9% (3/76) respectively. Three HIV seroconversion were found in the first, fourth and fifth anti-HIV test in 3 hemophilia A patients who received massive infusion of blood components during orthopedic corrective surgery. One case of HIV seroconversion found in 1987 was transmitted by HIV unscreened blood while 2 cases in 1991 by anti-HIV seronegative blood whose donors were in the window period of HIV infection. The prevalence of HIV seroconversion in Thai hemophiliacs is much lower than those in western countries.(ABSTRACT TRUNCATED AT 250 WORDS)

Evaluation of donor self exclusion program.

Donor self exclusion is a low technology procedure to avoid blood donations from the high-risk groups for HIV infection. This strategy has been widely used in western countries to reduce the risk of transfusion associated AIDS. At Ramathibodi Hospital, we conducted a study on donor self exclusion program during March-December 1991. It was found 2.60%-6.55% (mean = 4.59%) of a total of 4,286 units of blood that were from the donors who indicated that their blood may not be safe for transfusion while the rest of them declared that their blood was safe based on sexual behavior during the past 3 months and the history of intravenous drug use. Among 202 units of unsafe blood, there were 1 (0.49%) positive for HIV-Ag, 7 (3.46%) for anti-HIV, 5 (2.48%) for anti-HCV, 10 (4.95%) for HBsAg and 6 (2.97%) for VDRL while there was no HIV-Ag detected in 4,084 units of safe blood but 19 (0.46%) were positive for anti-HIV (p < 0.05), 65 (1.59%) for anti-HCV, 219 (5.36%) for HBsAg and 56 (1.37%) for VDRL. It was clearly demonstrated in this study that confidential self-exclusion or HIV-Ag testing would have eliminated this HIV-Ag reactive unit in the "window period" from transfusion, while the syphilis screening would not have had any value as a surrogate marker. However, self-exclusion programs are likely to prove too non-specific and need more time to educate the donors.(ABSTRACT TRUNCATED AT 250 WORDS)