In this study, we recorded 477 blood donors and 83 recipients. Volunteer donors accounted for 10.5% and family donors for 89.5%. There was statistically significant difference in blood donation between volunteer donors and family donors (p = 0.03). We noticed a higher proportion of volunteer donors compared to family donors attending the LHD blood center. The reasons for increased voluntary donation within the study population can be explained by various factors, including previous loss of a relative due to lack of blood, or the quest for their HIV status. Family donors are those who donate blood to acquaintances (relatives and friends) in urgent need. Our results are similar to those reported in other studies conducted in several countries [7, 18]. Contrarily to our study, some studies conducted over long periods in France and Turkey have shown that blood supply is mainly based on volunteer donation [6, 19]. With the presidential decree signed on the 5th April 2013, recognizing the utility of the “National Organization of volunteer blood donors, we can expect in the near future that voluntary blood donation becomes more frequent in Cameroon.
We reported a statistically significant difference between mean ages in the two donor groups (p = 0.017), with volunteer donors dominating the older age group. The number of blood donation was higher in the 23–33 years old group, which is consistent with a study conducted in 2010 at the LHD . This number decreased considerably with age and volunteer donors represented the majority of donors. We can hypothesize that many persons of more than 30 years old and probably having professional activities, voluntarily make blood donation, unlike young people who may be forced to make blood donation for family reasons.
Most donors had a secondary-school education, consistent with a similar study in Côte d'Ivoire . Male/female sex ratio was 4/1 for all donors in our study. It was 2.8/1 from voluntary donors and 4.1/1 from family donors. In 2010, Dikosso  also found a higher frequency of men among both donors groups; male/female sex ratio of 8/1 for family donors and 2/1 for volunteer donors at the HLD. Our findings are consistent with previous observations in Cameroon [7, 20, 22], Tanzania , and in a multicenter study involving seven sub-Saharan African countries  that showed less representation of women (< 30%) as blood donors, whether volunteer or not. This could be explained by physiological differences between men and women, who are not allowed to make blood donation during menstruation, lactation or in pregnancy.
In our donor population, men are more infected than women, but this difference was not significant. This trend has also been reported in Cameroon  and in Côte d’Ivoire . HIV seroprevalence among blood donors was 1.77%. This prevalence is lower than that found in other localities of Cameroon (Edéa, Douala or Yaoundé) with prevalences between 4.1% and 7.9% [20, 23, 24]. This prevalence is also lower than the estimated national HIV prevalence of 2.6% among blood donors in 2011 . The same year, the overall HIV seroprevalence in Cameroon was estimated to be 4.3% and 4.6% in the coastal region . In our study, HIV infection was found only in 8/406 (~2.0%) family donors. The relative low prevalence as observed in this study could be explained by the relatively short recruitment period of 4 months compare to > 6 months in others studies. Additionally, this could result from a change in people's attitudes following the efforts of the National Committee for AIDS Control to prevent HIV transmission. Another parameter that could explain this difference is our sample size that is lower than in similar studies. HIV prevalence in our study and in all studies conducted in blood donors from 2003 was higher than that found (0.4%) in Yaoundé in the 1990 . This could be explained by the introduction of 4th generation tests that are more sensitive .
HBV infection was found in 3.5% of the donor population. It was 2.1% (1/48) among volunteer donors and 3.7% (15/413) among family donors. These prevalence values are lower than the prevalence reported in Cameroon in 2003, 2004, and 2012, that were 10.7%, 9.9% and 10.1%, respectively [23, 24, 27].
HCV seroprevalence in our donor population was 1.35%. It was 4.2% (2/47) among volunteer donors and 1.0% (4/396) among family donors. The overall prevalence was lower than that reported (2.3%) in 2003 by Koanga et al. in the LHD, but identical to the 1.6% prevalence reported in Tanzania in 2006 by Mecky et al.. This difference could result from differences in the used diagnostic tests. In the study conductyed by Mogtomo, HCV diagnosis was performed only by ELISA, whereas a combination of RDT and ELISA were employed in our studies. National HCV seroprevalence was estimated to 13.8% in 2002, with regional variations .
The prevalence of T. Pallidum infection is high in our study. The prevalence in volunteer donors was 2.1% while it was 8.88% among family donors. The overall seroprevalence among all donors was 8.1%, similar to the 2003 prevalence of 7.9% in the same hospital . It was higher than the 5.7% reported in the regional hospital of Edea in 2012 , and lower than the 9.1% found at the University Teaching Hospital of Yaoundé in 2003 . In other studies conducted in Cameroon and elsewhere [12, 13], high prevalence of T. pallidum infection have been reported. Syphilis seems to be overlooked in Cameroon and there is no awareness campaign for this infection. Most are unaware of their T. pallidum status until probably at the moment of blood donation when the infection is diagnosed. Another problem is the absence of serological results given to family donors in LHD. This suggests that family donors with positive T. pallidum test do not know their status and could therefore continue to serve as reservoirs for the bacteria, which can lead to infertility. Taken together, these findings indicate that bacterial contamination is the major risk of infection during blood transfusion. In addition, we report a single case of co-infection HCV/T. pallidum in the study population. In a study conducted in China, HIV/T. pallidum and HBV/T. pallidum co-infections were noticed in addition to HCV/T. pallidum infected cases . The prevalence of T. pallidum infections and co-infections was probably underestimated in our study given that the HIV positive participants were not screened for T. pallidum infection.
None of the assessed risk factors was significantly associated with infection in our study. This is consistent with the study conducted in Australia between 2005 and 2010 in a population of infected blood donors . However, Kra et al., in 2001 had found that having multiple partners, unprotected sex, or history of hazardous behavior (injections, blood transfusions) were risk factors for positive HBsAg among blood donors in Côte d'Ivoire .
We recorded 83 blood recipients (male/female sex ratio of 3.6/1) in the Pediatric, Gyneco-obstetrical and emergency wards. These wards were chosen because they presented the highest blood requests as noted in blood bank records. The high number of female recipient is due to the high frequency of anemia in Gyneco-obstetrical ward. Various complications have been identified from medical records of patients during and after transfusion, and 31.32% of recipients had at least one complication. This value is lower than that obtained by Mbanya et al. (> 50%) between 1994 and 1998 in Cameroon . Complications were mainly found in Gyneco-obstetrical ward, as well as the number of blood transfusions. This could be explained by frequent bleeding in women during childbirth. Studies have shown that transfusion occurs when blood loss exceeds 500 ml [29, 30]. Chills and hematuria were the most frequent complications (14.5% of recipients) and accounted for 46.1% of all complications. Mbanya et al. reported febrile reactions and urticaria in 40.1% and 19.4% recipients respectively . Chills and hematuria are signs and symptoms of acute or delayed hemolytic reactions. Chills may be due to the fact that some transfused patients did not receive blood of the same group. Due to lack of blood, several patients received O group blood although not belonging to this group. The O group was predominant among all blood donors in our study (52.2%), and blood transfusion in this group was generally for emergency purposes. A study conducted by Mandengue et al. (2003) in the LHD showed the existence of several blood systems and also, that the distribution among donors and recipients is not always the same . Hematuria due to non-immune hemolysis of red blood cells compatibility was particularly observed in Sickle Cell Disease (SCD) in our study. Among the recipients, 10 SCD patients agreed to participate in this study. SCD patients are generally subject to blood transfusions. A study conducted in Yaoundé in 2012 by Ngo Sack et al. showed that SCD patients may receive more than 10 transfusions and that the risk of infection increases with the number of transfusion . Another study conducted in Libreville in 2002 by Moussaoui et al. showed the benefits of these transfusions used in systematic prophylaxis .
Among the recipients, 57 (68.7%) were transfused for the first time and 26 (31.3%) had received previous transfusions. The number of previous transfusion was higher in Gyneco-obstetrical and Pediatric wards. In the Pediatric ward, anemia due to malaria was the main indication for transfusion, followed by sickle cell vaso-occlusive crisis. In a study conducted in Benin, malarial anemia was the main indication for blood transfusion in Pediatric wards, followed by malnutrition, viral and bacterial infections (pneumonia, meningitis, measles, typhoid fever), and SCD . Our results, however, do not reflect the morbidity associated with sickle cell disease.
Limitations and strengths
The weaknesses of this study include: 1) the small sample size (83 recipients) and the lack of funding which limited our ability to explore seroconversion of viral and bacterial infections in blood recipients and 2) the low representation of all sub-regional hospitals in this study that potentially could impose significant bias in the obtained data; for instance, if a patient visited another hospital with an acute complication, our institution database may not have a record of that visit.
Compared to previous reports in this area, our study is one of the few researches to investigate transfusion-related complications in blood recipients attending the HLD, while past research has largely focused on blood donors. This study provides a comprehensive analysis of the prevalence of both viral and bacterial infections among blood donors in Douala, providing additional insight into TTI-associated disease burdens and opportunities for prevention from a local standpoint.