Relation of titers of antibodies to CMV in blood donors to the transmission of cytomegalovirus infection

Titers of antibody to cytomegalovirus (CMV) of 529 persons whose blood had been supplied to 51 selected patients who underwent open-heart surgery were determined by indirect hemagglutination (IHA) and IgM-specific indirect immunofluorescence (IFA). Twenty-eight patients showed evidence of active CMV infection after transfusion (seroconversion or a fourfold rise in titer by IHA), whereas 23 showed no serological change. Patients with active CMV infections had received, on average, a greater number of blood units (12.9 vs. 7.9), of which more were seropositive (6.9 vs. 3.5), than did patients who showed no serological change. Those seropositive units of blood that had been transfused into the group that showed evidence of active infection, however, had a lower geometric mean titer than did those transfused into the group that showed no serological change (1:654 vs. 1:1,360). Seven (1.3%) of the 529 blood donors had CMV-specific IgM titers (by IFA) of greater than or equal to 1:16; each of the seven recipients of their blood subsequently showed evidence of active CMV infection. This study suggests that donor blood with high IHA titers may prevent transmission of CMV infection, whereas blood from donors with IgM antibody to CMV may transmit CMV.     

Microbiology, time course and clinical characteristics of infection in critically ill patients receiving packed red blood cell transfusion

Background and Objectives  Packed red blood cell transfusion has been associated with increased infection in a variety of critically ill patient populations. We evaluated the microbiology and time course of infection in transfused patients in the intensive care unit (ICU) as no data exist on these parameters.
Materials and Methods  We performed a retrospective review of data for all patients admitted to a 24-bed medical-surgical ICU at Cooper University Hospital from July 2003 to September 2006 and entered in the Project Impact database.
Results  A total of 2432 patients were admitted during the study period, of which 609 underwent transfusion. Transfused patients were more likely to develop a nosocomial infection (10·5% vs. 4·9%, P  < 0·001). ICU and hospital length of stay were longer in the transfused group ( P  < 0·001 for both). Mortality was also greater (13·1% vs. 8·7%, P  = 0·001). Transfused patients had a shorter time from hospital admission to first infection ( P  < 0·001) and ICU admission to first infection ( P  < 0·001). Multivariate analysis confirmed transfusion as an independent risk factor for infection, mortality, hospital and ICU length of stay. Methicillin-resistant Staphylococcus aureus , vancomycin-resistant enterococcus and Acinetobacter occurred more often in transfused patients. Acinetobacter accounted for a disproportionate share of infections among transfused patients ( P  < 0·001).
Conclusions  Transfused ICU patients have a higher incidence of nosocomial infection and worse outcomes. Transfused patients had a shorter onset of infection. Acinetobacter infection appears to be particularly common among these patients. Further investigation is merited to better elucidate the mechanism for these findings and their therapeutic and clinical implications.     

A multicenter, prospective study of posttransfusion hepatitis in Milan

We studied the risk of posttransfusion hepatitis in recipients of blood collected from volunteer donors who tested negative for HBsAg and had serum ALT levels less than 1.5 times the upper limit of the normal range. Between October, 1983 and September, 1984, 676 consecutive patients who needed blood or plasma transfusions during or after elective surgery, who had no history of liver disease and had never received blood previously, were studied. The patients were given a total of 4,813 (mean = 7) units. Ninety-six patients developed posttransfusion hepatitis, which yielded a hepatic incidence of 20 cases per 1,000 units of transfused blood. Ninety-two patients had non-A, non-B hepatitis, 3 had hepatitis B and 1 had cytomegalovirus infection. The incubation periods for non-A, non-B hepatitis ranged from 2 to 26 (mean = 9.5 +/- 4) weeks. In 68 (73%) patients, the hepatitis was completely asymptomatic; only 24 (27%) patients developed symptoms, including jaundice and hepatomegaly. There were no cases of fulminant hepatitis. Sixty per cent of the patients still had elevated serum ALT levels 1 year after the onset of hepatitis. The 96 patients with hepatitis had received a mean of 9.6 blood units, as compared to a mean of 6.7 units for the unaffected patients (p less than 0.001). This study demonstrated that non-A, non-B hepatitis remains a common and important complication of blood transfusion despite screening of blood donors for HBsAg and elevated serum ALT levels.     

Continuous cytomegalovirus seroconversion in a large group of healthy blood donors

BACKGROUND AND OBJECTIVES: Transmission of cytomegalovirus (CMV) to seronegative, immunocompromised recipients can cause serious and fatal complications. Although the seroprevalence of CMV is high, the risk of primary CMV infection among healthy blood donors has not yet been analysed in a large population. MATERIALS AND METHODS: We developed an algorithm to determine the rate of CMV seroconversion in an overall cohort of 24,260 subjects who donated 176,474 blood units during an 11-year observation period. RESULTS: We detected CMV seroconversion in all relevant age groups (18-60 years) with an overall seroconversion rate of 0.55% per year. Both CMV seroconversion and seroprevalence occurred more frequently in female donors (P = 0.02 and P < 0.001, respectively). We identified 30-35-year-old blood donors as the group with the highest rate of CMV seroconversion per year (1.33% vs. 0.46%; P < 0.0001). CONCLUSIONS: We conclude that the risk of primary CMV infection is a continuous lifelong event and correlates with age and female gender.     

Transfusion-related acute lung injury: incidence and risk factors

Transfusion-related acute lung injury (TRALI) is the leading cause of transfusion-related mortality. To determine TRALI incidence by prospective, active surveillance and to identify risk factors by a case-control study, 2 academic medical centers enrolled 89 cases and 164 transfused controls. Recipient risk factors identified by multivariate analysis were higher IL-8 levels, liver surgery, chronic alcohol abuse, shock, higher peak airway pressure while being mechanically ventilated, current smoking, and positive fluid balance. Transfusion risk factors were receipt of plasma or whole blood from female donors (odds ratio = 4.5, 95% confidence interval [CI], 1.85-11.2, P = .001), volume of HLA class II antibody with normalized background ratio more than 27.5 (OR = 1.92/100 mL, 95% CI, 1.08-3.4, P = .03), and volume of anti-human neutrophil antigen positive by granulocyte immunofluoresence test (OR = 1.71/100 mL, 95% CI, 1.18-2.5, P = .004). Little or no risk was associated with older red blood cell units, noncognate or weak cognate class II antibody, or class I antibody. Reduced transfusion of plasma from female donors was concurrent with reduced TRALI incidence: 2.57 (95% CI, 1.72-3.86) in 2006 versus 0.81 (95% CI, 0.44-1.49) in 2009 per 10 000 transfused units (P = .002). The identified risk factors provide potential targets for reducing residual TRALI.     

Human herpesvirus-6 is associated with cytomegalovirus reactivation in liver transplant recipients

Human herpesvirus-6 (HHV-6) may be a risk factor for cytomegalovirus (CMV) disease in posttransplant patients, possibly through a direct interaction or through a general immunomodulatory effect. To examine this possibility, 88 liver transplant recipients were monitored with serial HHV-6 polymerase chain reaction (PCR), CMV antigenemia, and CMV plasma viral load. HHV-6 infection was defined by a positive PCR of peripheral blood lymphocytes. Forty-eight (54.4%) of 88 patients had at least 1 positive HHV-6 PCR. CMV recurrence was significantly more common in patients with HHV-6 infection (38/48 patients [79. 2%]), compared with recurrence in those without HHV-6 infection (18/40 patients [45%]; P=.001). Peak CMV viral load was 24, 147+/-6799 copies/mL in patients with HHV-6 infection versus 8391+/-4598 copies/mL in patients without HHV-6 infection (P=.001). Symptomatic CMV disease was more common in patients with HHV-6 infection than it was in those without infection (15/48 patients [31. 3%] vs. 4/10 patients [10.0%]; P=.013). In a multivariate analysis including other risk factors for CMV, HHV-6 infection remained an independent risk factor for CMV disease (P=.013; odds ratio, 7.26; 95% confidence interval, 1.52-34.72). HHV-6 is associated with CMV infection and disease, thus supporting an interaction between these viruses.     

Incidence of Post-Transfusion Hepatitis before and after Screening for Hepatitis C Virus Antibody

To evaluate the effectiveness of screening test for antibody to hepatitis C virus (anti-HCV), the incidence of acute post-transfusion HCV infection in patients who underwent cardiovascular surgery and received blood transfusion was studied. All patients were followed prospectively with serum biochemistry tests and viral hepatitis markers before and periodically for at least 6 months after cardiovascular surgery. None of them had history of liver disease and none tested positive for anti-HCV prior to blood transfusion. Before blood donors were screened for anti-HCV with a second-generation HCV diagnostic kit, 28 (12.4%) of 226 patients or 0.49% of 5,690 unit transfusion had seroconverted to anti-HCV during a 6-month follow-up. The incidence of post-transfusion hepatitis (PTH) C in 91 patients who had received 1–12 units transfusion was significantly lower than in 135 patients who had received more than 12 units transfusion (6.6 vs. 16.3%, p<0.05). However, none of the 87 transfused patients, since anti-HCV screening in July 1992, developed PTH C (p<0.05). The result demonstrates that screening for anti-HCV by a more sensitive second-generation HCV diagnostic assay may protect the patients studied from PTH C. It further provides a firm argument for the necessity of a nation-wide blood donor screening.     

Post-operative course of coronary artery bypass surgery patients who pre-donate autologous blood

BACKGROUND: Pre-operative autologous blood donation is used to reduce the need of allogeneic blood in patients undergoing coronary bypass surgery operations, but it is not clear what impact the blood donation has on the post-operative course of these patients. METHODS: We studied the post-operative course of 210 patients who pre-donated autologous blood before their coronary bypass operation (donors) and of 67 patients who were eligible to pre-donate but did not (controls). RESULTS: The clinical variables and the technical operative parameters of the patients in the two groups were similar. There was no significant difference between the duration of assisted ventilation post-operatively (756 +/- 197 vs. 802 +/- 395 min; P=0.54) or length of stay in the intensive care unit (1.8 +/- 1.1 vs. 1.7 +/- 0.9 days; P=0.52) of the two groups. The number of autologous units of packed red cells and of fresh frozen plasma (FFP) received by the donors was significantly higher than the number of units of allogeneic packed red cells (1.5 +/- 0.9 vs. 0.3 +/- 0.9; P=0.001) and the units of homologous FFP received by the controls (2.3 +/- 0.8 vs. 0.6 +/- 1; P=0.001). CONCLUSIONS: We found no evidence that autologous blood donation exerted a negative influence on the post-operative course of patients undergoing coronary bypass surgery. Patients who pre-donated blood received no allogeneic blood products, but the number of autologous blood products received by donors was higher than the number of blood products received by patients who did not pre-donate.     

The Incidence and Consequences of Cytomegalovirus Transmission via Blood Transfusion to Low Birth Weight, Premature Infants in North East Scotland

In a 2-year study involving 133 premature low birth weight (less than 1,500 g) infants, the impact of CMV infection via blood transfusion was assessed. 8.4% (7 out of 83) of transfused infants and 10% (7 out of 70) of those exposed to seropositive blood acquired CMV. In those less than 1,250 g the infection rate rose to 13.2% (7 out of 46). Seropositive infants were at a higher risk of acquiring CMV infection than seronegative ones. CMV infection did not give rise to specific immediate morbidity, and no deaths were attributed to CMV. The only source of nosocomial CMV infection was the transfused seropositive blood. Based on these findings, it was possible to formulate a CMV transfusion policy to premature infants in our region.     

Human T-lymphotropic virus type 1 and type 2 seroprevalence, incidence, and residual transfusion risk among blood donors in Brazil during 2007-2009

Abstract Human T-lymphotropic virus type 1/2 (HTLV-1/2) infection is endemic in Brazil but representative donor prevalence and incidence data are lacking. All blood donations (2007-2009) from three blood centers in Brazil were studied. Samples reactive on one HTLV screening test (EIA) were retested with a different EIA; dual EIA reactivity correlated strongly with a confirmatory Western blot. Prevalence, incidence, and residual transfusion risk were calculated. Among 281,760 first-time donors, 363 were positive for HTLV on both EIAs (135 per 10(5), 95% CI 122-150). Prevalence differed considerably by region, from 83 to 222 per 10(5). Overall incidence rate was 3.6/10(5) person-years and residual transfusion risk was 5.0/10(6) per blood unit transfused. The logistic regression model showed significant associations with: age [adjusted odds ratio (aOR)=5.23 for age 50+ vs. <20], female sex (aOR=1.97), black (aOR=2.70 vs. white), and mixed skin colors (aOR=1.78 vs. white), and inversely with education (aOR=0.49, college vs. less than high school). HTLV testing with a dual-EIA strategy is feasible and can be useful in areas with low resources. Incidence and residual risk of HTLV-1 transmission by transfusion were relatively high and could be reduced by improving donor recruitment and selection in high prevalence areas. Blood center data may contribute to surveillance for HTLV infection.     

Allogeneic red blood cell transfusion is an independent risk factor for the development of postoperative bacterial infection

BACKGROUND AND OBJECTIVES: Allogeneic red blood cell transfusions may exert immunomodulatory effects in recipients including an increased rate of postoperative bacterial infection. It is controversial whether allogeneic transfusion is an independent predictor for the development of postoperative bacterial infection. METHODS: We analysed a prospectively collected database of 1,349 patients undergoing colorectal surgery in 11 centres across Canada. The primary outcome was the development of either a postoperative wound infection or intra-abdominal sepsis in transfused and nontransfused patients. The effect of allogeneic transfusion on postoperative infection was evaluated with adjustment for all the confounding factors in a multiple regression analysis. RESULTS: The 282 patients who received a total of 832 allogeneic units had a significantly higher frequency of wound infections and intra-abdominal sepsis than the patients who were not transfused (25. 9 vs. 14.2%, p = 0.001). A significant dose-response relationship between transfusion and infection rate was demonstrated. Multiple regression analysis identified allogeneic transfusion as a statistically significant independent predictor for postoperative bacterial infection (OR 1.18, 95% CI 1.05-1.33, p = 0.007). Other independent predictors were anastomotic leak, repeat operation, patient age and preoperative haemoglobin level. The mortality rate was also significantly higher in the transfused group. CONCLUSION: These data support the hypothesis that allogeneic red cell transfusion is an independent risk factor for the development of postoperative bacterial infection in patients undergoing colorectal surgery. This association provides further reason to minimise exposure to allogeneic transfusions in the perioperative setting.     

Transmission of retroviruses from seronegative donors by transfusion during cardiac surgery. A multicenter study of HIV-1 and HTLV-I/II infections.

OBJECTIVE: To evaluate the effectiveness of serologic testing of blood donors for human immunodeficiency virus type 1 (HIV-1) and human T-cell lymphotropic virus types I and II (HTLV-I/II) infections and to estimate the risk for transmission of HIV-1 and HTLV-I/II by transfusion of seronegative blood from screened donors. DESIGN: A prospective multicenter cohort study of cardiac surgery patients who received multiple transfusions between 1985 and 1991. SETTING: Cardiac surgery services of three large tertiary care hospitals. PATIENTS: The study included 11,532 patients in three hospitals who had cardiovascular surgery. MEASUREMENTS: Incident HIV-1 and HTLV-I or HTLV-II infection. RESULTS: We detected two new HIV-1 infections among patients transfused with 120,312 units of blood components from seronegative donors. In each case a donor was detected on follow-up who had seroconverted since the donation. The HIV-1 infection rate was 0.0017% with an upper limit of the 95% CI of 0.0053%. Before donor screening for HTLV-I, transfusion of 51,026 units resulted in two HTLV-I infections (0.0039%) and four HTLV-II infections (0.0078%). After HTLV-I screening was instituted, one recipient was infected with HTLV-II among participants exposed to 69,272 units, a rate of 0.0014%. A corresponding HTLV-I/II-infected donor was found for this patient. CONCLUSION: Serologic screening of donors for antibodies to HIV-1 and HTLV-I coupled with exclusion of donors from groups having a relatively high risk for infection has led to a low incidence of transfusion-transmitted HIV-1 and HTLV-I/II infection in the United States. A small risk remains, however, despite these measures. We estimate the residual risk for HIV-1 and HTLV-II infection from transfusion of screened blood during the time of this study to be about 1 in 60,000 units.     

High incidence of ganciclovir-resistant cytomegalovirus infection among lung transplant recipients receiving preemptive therapy.

Preemptive antiviral therapy in transplant patients is thought to be less likely to lead to antiviral resistance than is routine prophylaxis. Cytomegalovirus (CMV)-seropositive lung transplant patients (R+) were assigned to receive pp65 antigen-guided ganciclovir therapy, and seronegative recipients of organs from seropositive donors (D+/R-) were assigned to receive initially preemptive and then routine ganciclovir prophylaxis. The incidence of infection with ganciclovir-resistant (ganR) CMV was assessed retrospectively. GanR CMV infection developed in 4 (9%) of 45 patients, at a median of 4.4 months (range, 3.1-6.6 months) after transplantation, and was more common among D+/R- patients than among R+ patients (3 of 11 vs. 1 of 34; P =.04). The incidence among patients who received preemptive therapy was similar to that among patients who received routine prophylaxis. All ganR isolates contained a UL97 mutation. GanR CMV infection occurs in nearly 10% of lung transplant recipients, despite preemptive antiviral therapy, and is more common among D+/R- patients.     

Preoperative aspirin therapy and reoperation for bleeding after coronary artery bypass surgery

We performed a case-control study to estimate the relative risk of reoperation for bleeding in coronary artery bypass graft patients who had taken aspirin within the 7 days preceding surgery. Comparison of 90 cases of reoperation with 180 matched control subjects gave an estimated odds ratio for reoperation of 1.82 (95% confidence interval, 1.23 to 3.32). Although their preoperative coagulation values were similar, cases used significantly more whole blood (cases, 9.5 +/- 5.2 units; control subjects, 3.0 +/- 2.0 units; median +/- interquartile range), packed red blood cells (cases, 2.1 +/- 4.0 units; control subjects, 0.9 +/- 2.0 units), and platelets (cases, 12.2 +/- 12.0 units; control subjects, 2.9 +/- 4.0 units) than control subjects. Cases had intensive care unit stays of 4.7 +/- 5.7 days (mean +/- SD) vs 2.1 +/- 1.9 days for control subjects and postoperative hospitalizations of 10.9 +/- 8.2 days vs 7.0 +/- 3.2 days for control subjects. We conclude that aspirin exposure within 7 days before coronary bypass surgery is associated with an increased rate of reoperation for bleeding and that reoperation is associated with large increases in transfusion requirements and intensive care unit and hospital stays.     

Cytomegalovirus infection after bone marrow transplantation: an association with acute graft-v-host disease

Among 181 patients undergoing allogeneic bone marrow transplantation over a five-year period (1978 through 1982), cytomegalovirus (CMV) infection was a frequent and often lethal complication. Recipient pretransplant serology was the most important predictor of posttransplant CMV infection. CMV infection occurred in 26/137 seronegative recipients and in 28/44 seropositive recipients (P less than .001). Among patients who developed CMV infection, the time to infection was identical in seronegative and seropositive patients (median, 71 days post transplant). Bone marrow donor CMV serology did not significantly influence CMV infection rate. CMV infection was strongly associated with acute graft-v-host disease (AGVHD), occurring in 34/81 patients with AGVHD and 20/100 without GVHD (P less than .001). AGVHD preceded CMV infection by 33.7 days (mean) in patients developing both complications. Patients who developed CMV infections had also received more cellular blood products post transplant. These data suggest that CMV infection may occur through reactivation of latent virus (in seropositive recipients) or through exogenous exposure, possibly through transfused blood products, but that duration of immunoincompetence may be more critical than route of exposure in timing of clinically evident CMV infection. Prophylaxis tailored to the likely infectious source and more effective GVHD prevention both may be critical in preventing CMV infection after bone marrow transplantation.     

SEN virus infection and its relationship to transfusion-associated hepatitis

SEN virus (SEN-V) is a recently identified single-stranded, circular DNA virus. Two SEN-V variants (SENV-D and SENV-H) were assayed by polymerase chain reaction (PCR) to investigate their role in the causation of transfusion-associated non-A to E hepatitis. The incidence of SEN-V infection after transfusion was 30% (86 of 286) compared with 3% (3 of 97) among nontransfused controls (P < .001). Transfusion risk increased with the number of units transfused (P < .0001) and donor-recipient linkage for SEN-V was shown by sequence homology. The prevalence of SEN-V in 436 volunteer donors was 1.8%. Among patients with transfusion-associated non-A to E hepatitis, 11 of 12 (92%) were infected with SEN-V at the time of transfusion compared with 55 of 225 (24%) identically followed recipients who did not develop hepatitis (P < .001). No effect of SEN-V on the severity or persistence of coexistent hepatitis C virus (HCV) infection was observed. In 31 infected recipients, SEN-V persisted for greater than 1 year in 45% and for up to 12 years in 13%. SEN-V-specific RNA (a possible replicative intermediate) was recovered from liver tissue. In summary, SENV-D and -H were present in nearly 2% of US donors, and were unequivocally transmitted by transfusion and frequently persisted. The strong association of SEN-V with transfusion-associated non-A to E hepatitis compared with controls raises the possibility, but does not establish that SEN-V might be a causative agent of posttransfusion hepatitis. The vast majority of SEN-V-infected recipients did not develop hepatitis.     

Human immunodeficiency virus testing in acute leukemia patients transfused between 1978 and 1985: a retrospective study on 91 cases

Human immunodeficiency virus (HIV-1) can be transmitted by blood transfusions. A recent report focused on the relativey high risk of HIV-1 infection in American patients treated for leukemia and multiply transfused as a consequence of therapy. We therefore conducted a retrospective study on the presence of HIV-1 antibodies among 91 acute leukemia patients diagnosed between 1978 and 1985, before the onset of routine tests for HIV-1 contamination of blood products. The transfusion requirement (platelet units, red blood cell concentrates) involved almost 7,000 donors. We did not find any case of seropositivity in patients transfused with units from the donor pool. The only case of HIV-1 seropositivity was due to a bone marrow transplant donor, retrospectively found to be HIV-1 seropositive. These results differ from the American data previously cited. This is probably due both to differences in diffusion of the HIV-1 infection in the two countries and to differences in the selection of the two donor populations. We conclude that the risk of contracting HIV-1 infections before 1985 through multiple transfusions from registered donors in our Italian area was very low, if not absent, not only for leukemia patients but reasonably for other categories of heavily transfused groups.     

Increased prevalence of iron-overload associated endocrinopathy in thalassaemia versus sickle-cell disease

Iron-overload associated endocrinopathy is the most frequently reported complication of chronic transfusion therapy in patients with thalassaemia (Thal). This study compared iron-overloaded subjects with Thal (n = 142; 54%M; age 25.8 +/- 8.1 years) and transfused sickle-cell disease (Tx-SCD; n = 199; 43%M, 24.9 +/- 13.2 years) to non-transfused SCD subjects (non-Tx-SCD; n = 64, 50%M, 25.3 +/- 11.3 years), to explore whether the underlying haemoglobinopathy influences the development of endocrinopathy. Subjects were recruited from 31 centres in the USA, Canada and the UK. Subjects with Thal had more evidence of diabetes (13% vs. 2%, P < 0.001), hypogonadism (40% vs. 4%, P < 0.001), hypothyroidism (10% vs. 2%, P = <0.001) and growth failure (33% vs. 7%, P < 0.001), versus Tx-SCD. Fifty-six per cent of Thal had more than one endocrinopathy compared with only 13% of Tx-SCD (P < 0.001). In contrast, Tx-SCD was not different from non-Tx-SCD. Multivariate analysis indicated that endocrinopathy was more likely in Thal than SCD [Odds Ratio (OR) = 9.4, P < 0.001], with duration of chronic transfusion a significant predictor (OR = 1.4 per 10 years of transfusion, P = 0.04). Despite iron overload, endocrinopathy was not increased in Tx-SCD versus non-Tx-SCD, suggesting that the underlying disease may modulate iron-related endocrine injury. However, because transfusion duration remained a significant predictor of endocrinopathy, these data should be confirmed in SCD subjects that have been chronically transfused for longer periods of time.     

Predictors of outcome in massive upper gastrointestinal hemorrhage

We reviewed 100 consecutive cases of massive upper gastrointestinal hemorrhage (UGIH). The criteria for inclusion were a decrease in hematocrit greater than or equal to 6%, unstable vital signs, and greater than or equal to units of blood transfused (16 +/- 18 units, mean +/- SD). A multiple regression analysis of 96 variables was employed to determine the most accurate predictors of outcome. The overall mortality was 35%. Hospital status (whether the patient was an inpatient or outpatient when the UGIH began) showed a striking association with mortality (70% for inpatients vs. 22% for outpatients, p less than 0.001). Nonsurvivors also had a greater number of life-threatening diseases than survivors (1.4 +/- 1.1 vs. 0.3 +/- 0.5, p less than 0.001) and greater transfusion requirements (27 +/- 20 units vs. 10 +/- 13 units, p less than 0.001). Age, the presence of cirrhosis, and recent excessive alcohol intake were not important risk factors. At presentation, the most reliable predictor of a fatal outcome was the brevity of the interval between the onset of bleeding and the initiation of a medical work-up. The primary predictor when considering the entire hospitalization was the number of life-threatening diagnoses. Our data indicate that stratification for hospital status and for other potentially predictive risk factors should be incorporated in future trials of therapy for UGIH.     

Prevalence and risk factors for red blood cell alloimmunization in 175 children with sickle cell disease in a French university hospital reference centre

Patients with sickle cell disease (SCD) show a high prevalence of red blood cell (RBC) alloimmunization, but few studies have focused on children. We aimed to study the prevalence and risk factors of RBC alloimmunization in SCD children. We retrospectively analysed the medical and transfusion files for 245 SCD children hospitalized in our centre in 2014 and included 175 patients who had received at least one RBC unit in their lifetime. The main clinical and immuno‐haematological characteristics of alloimmunized and non‐alloimmunized patients were compared. The prevalence of alloimmunization was 13·7% [95% confidence interval (CI) (8·6–18·6)], and 7·4% [95% CI (3·5–11·3)] after excluding the probable irregular natural antibodies (anti‐M, anti‐Le^a, anti‐Le^b, anti‐Le^x). Main risk factors for alloimmunization were increased number of RBC units received (median of 65 vs. 10 units per patient; P = 0·01) and the presence of one or more red cell autoantibodies (46·2% vs. 4·7%; P < 0·0001). The alloimmunization rate was higher for episodically transfused than chronically transfused patients (1·43 vs. 0·24/100 units received; P < 0·001). The presence of red cell autoantibodies appears to be a major risk factor for alloimmunization in SCD children and could justify specific transfusion guidelines.