Discussion

Compared to other countries, Bahrain’s SCD alloimmunization rate was intermediate, lower than Oman but higher than Saudi Arabia, Nigeria, Tanzania, and India (Table 3).

Study location | Alloimmunized patients/total | Alloimmunization rate (%) | P-value
Current study | 41/203 | 20.2% | 
Oman [12] | 42/133 | 31.6% | 0.018
Al-Ahsa, Saudi Arabia [13] | 13/78 | 16.7% | 0.50
Central India [10] | 3/140 | 2.15% | <0.001
Tanzania [11] | 17/200 | 8.5% | <0.001
Abuja, Nigeria [14] | 26/205 | 17.6% | 0.50

However, the alloimmunization rate among patients with thalassemia in Bahrain was lower than the rates reported in Oman, Saudi Arabia, Egypt, and India (Table 4).

Study location | Alloimmunized patients/total | Alloimmunization rate (%) | P-value
Current study | 5/108 | 4.6% | 
Oman [12] | 26/129 | 20% | <0.001
Al-Ahsa, Saudi Arabia [13] | 34/284 | 11.8% | 0.03
North India [15] | 17/255 | 6.6% | 0.46
Cairo, Egypt [16] | 36/200 | 18% | 0.001

A notable finding of this study was the significant difference in alloimmunization rates between the two disease groups. Patients with SCD had a significantly higher prevalence of 20.2% compared to only 4.6% in patients with thalassemia. This disparity is consistent with the findings from several other centers. Kuriri et al. (2023) reported alloimmunization rates of 16.7% in SCD versus 11.97% in thalassemia in the Al-Ahsa region of Saudi Arabia [13]. A study from Oman documented even more pronounced differences, with rates of 31.6% in SCD and 20% in thalassemia [12]. The higher alloimmunization rate in patients with SCD has been attributed to the pro-inflammatory state inherent to SCD, which can make the immune system more likely to produce alloantibodies. Additionally, the intermittent and episodic nature of transfusions in SCD compared to the regular, scheduled transfusions in thalassemia may further increase the risk of alloimmunization [17].

The slightly higher prevalence of alloimmunization among males (58.7%) compared to females (41.3%) in our alloimmunized group is an interesting observation, though the difference should be interpreted cautiously, given the overall male predominance in the study population (59.5%). The published literature is inconsistent on the role of sex as a risk factor. Some studies have reported higher rates in females, potentially due to previous pregnancy-related sensitization, while others, including Al-Allawi et al. (2025), identified female sex as a significant risk factor in several Middle Eastern studies [8,18]. The male predominance observed in our cohort may simply reflect the sex distribution of the underlying study population rather than a true biological predisposition. Moreover, this study did not account for other confounding risk factors such as cumulative transfusion burden, age at first transfusion, splenectomy status, or pregnancy history. Therefore, no definitive conclusions can be drawn regarding sex as an independent risk factor for alloimmunization.

Regarding blood group distribution, blood group O was the most prevalent in our cohort (54.3%), and the overwhelming majority (95.5%) were Rh-positive. This distribution is reflective of the blood group frequencies reported in Arab populations. In settings where the donor pool is ethnically similar to the recipient population, the antigenic mismatch that drives alloimmunization may be reduced. Gader et al. (2008) demonstrated this in a Saudi Arabian study, where none of the 13 patients who received blood exclusively from Arab donors developed alloantibodies, compared to a 22% alloimmunization rate when multi-ethnic donor blood was used [19].

The overall alloimmunization rate of 14.8%, and particularly the rate of 20.2% in patients with SCD, highlights the ongoing challenge of alloimmunization at SMC despite the existing matching protocols. Currently, SMC employs extended phenotypic crossmatching for patients with SCD and provides Rh and Kell antigen-matched transfusions for patients with thalassemia. The lower alloimmunization rate observed in patients with thalassemia (4.6%) compared to patients with SCD (20.2%) may, in part, reflect the protective effect of Rh and Kell matching in this group. Franchini et al. (2019) estimated that matching for Rh and Kell antigens alone could reduce alloimmunization by approximately 80% [9]. Similarly, a meta-analysis by Sugrue et al. (2024) found that adding c, E, and K matching to standard ABO/RhD matching reduced alloimmunization rates from 6.2% to 1.9% [20].

The persistently high alloimmunization rate among patients with SCD despite extended phenotypic matching is a concern that warrants further investigation. The chronic inflammatory state inherent to SCD and Rh antigen variant alleles, which are common in populations of African and Middle Eastern descent, can lead to alloimmunization even when conventional serologic matching appears compatible [17,21]. Al-Allawi et al. (2025) specifically highlighted the need for more research on Rh variants in the Middle East, as these may contribute to residual alloimmunization despite phenotypic matching [8]. The adoption of molecular genotyping for Rh variants, in addition to the current serologic phenotyping, could help identify patients at risk for alloimmunization due to variant antigen expression and further improve donor-recipient matching [5].

Strengths

This study has several notable strengths. It is the first to report the prevalence of RBC alloimmunization among patients with transfusion-dependent thalassemia and SCD in Bahrain, addressing an important gap in the regional literature. The study also included a relatively large cohort of 311 patients drawn from the largest tertiary care hospital in Bahrain. Moreover, the study covered a broad nine-year period (2016-2025), allowing for a more robust assessment of alloimmunization prevalence. In addition, the use of a standardized IAT-based gel column agglutination method with a clearly defined alloimmunization threshold ensured methodological consistency across all patients throughout the study period. Finally, the inclusion of both patients with SCD and thalassemia enabled a direct statistical comparison of alloimmunization rates between the two disease groups.

Limitations

This study has several limitations. First, data on alloantibody types (e.g., anti-K, anti-E, anti-C) and their frequencies were not available for the majority of alloimmunized patients and were therefore excluded from this study. This is a significant gap, as knowledge of the predominant antibody types is essential for guiding transfusion matching strategies. Second, data on risk factors of alloimmunization, such as cumulative transfusion exposure, age at first transfusion, splenectomy, pregnancy history, and Rh variant status, were not collected or analyzed. As a result, we were unable to assess the contribution of these risk factors to alloimmunization. Third, the retrospective design of the study relied on electronic medical records, which may have contained missing or incomplete information. In addition, the frequency of IAT testing differed across patients, as some underwent testing only once, whereas others had repeated IAT tests. This may have introduced variability in the detection of alloimmunization over time. Finally, patients with weak (1+) IAT reactions were not classified as alloimmunized because most of these reactions were transient and became negative on repeat testing, suggesting false-positive results. Excluding weakly positive IAT reactions likely improved specificity but may have slightly reduced sensitivity, leading to an underestimation of the true prevalence of alloimmunization.