Abstract
Background:
In children in sub-Saharan Africa, severe anaemia (SA) is an important cause of mortality and malaria is a primary cause. The World Health Organization (WHO) recommends blood transfusion for all children with haemoglobin (Hb) <4 g/dL and for those with Hb 4–6 g/dL with signs of instability such as shock or respiratory distress. In sub-Saharan Africa, evidence of the effect on mortality of transfusion in children with SA with and without malaria is mixed.
Aim:
To determine in children with and without malaria whether receipt of transfusion was associated with lower mortality at WHO transfusion thresholds.
Methods:
This was a retrospective cohort study of 1761 children with SA (Hb ≤6 g/dL) admitted to Kamuzu Central Hospital in Malawi. In those whose Hb was 4–6 g/dL, mortality was compared by transfusion, stratified by haemoglobin, malaria status and signs of instability.
Results:
Children with profound anaemia (Hb <4 g/dL) and malaria were the only subgroup who had a significant decrease in the odds of in-hospital death if they received a transfusion (OR 0.43, p=0.01). Although children with Hb 4–6 g/dL and at least one sign of instability had higher mortality than children with none, there was no difference in odds of mortality between those who received a transfusion and those who did not (OR 1.16, p=0.62).
Conclusions:
This study suggests that transfusion for children with profound anaemia and malaria may confer increased in-hospital survival. An understanding of the factors associated with mortality from SA will allow for interventions to prioritise the provision of limited blood.
Keywords: Severe anaemia, profound anaemia, blood transfusion, malaria, sub-Saharan Africa, mortality
Introduction
Severe anaemia (SA) is a significant cause of morbidity and mortality in children in sub-Saharan Africa (SSA) [1–4]. Malaria is a key cause of SA, and severe malaria with anaemia (SMA) contributes significantly to hospital admissions and mortality in African children, with mortality estimates ranging from 2 to 40% [1,2,5–8]. In Malawi specifically, one study found that malaria and anaemia account for 26.1% and 2.6% of child deaths, respectively [9].
Because of limited blood supplies in low- and middle-income countries (LMIC), the World Health Organization (WHO) endorses a rational approach to blood transfusion in children with anaemia. Specifically, WHO recommends transfusion for all children with complicated SA, meaning either profound anaemia (haemoglobin (Hb) <4 g/dL) or Hb levels between 4 and 6 g/dL with associated clinical signs and symptoms of dehydration, shock, impaired consciousness, heart failure, respiratory distress or very high malaria parasitaemia [10].
Children under 5 years of age are the most vulnerable group affected by SA and SMA and are more likely to require a blood transfusion [5,11–14]. In SSA, there is mixed evidence of the beneficial effect of blood transfusion on mortality in children with SA [2,7,8,15–19], and the effect of blood transfusion on those with and without malaria has not been well studied. Furthermore, the transfusion thresholds for SA with and without malaria are inconsistent, are related to malaria endemicity, and are based largely on expert opinion [12].
This study describes the characteristics of children with SA and SMA admitted to Kamuzu Central Hospital (KCH), the main referral hospital in central Malawi. At KCH, a significant proportion of child deaths are associated with SA [9]. The study aimed to determine whether receipt of blood transfusion was associated with lower mortality in children with SA at different transfusion thresholds. In addition, it sought to determine whether the association between transfusion and mortality differed between children with and without malaria. Better understanding of the impact of blood transfusion on those with SA and SMA could allow for improved resource allocation where supply is limited.
Subjects and Methods
Setting
Malawi is a low-income country in SSA with a population of approximately 19 million people. KCH is the main referral hospital in central Malawi, and admits approximately 21,000–25,000 paediatric patients per year [20]. The paediatric department has approximately 300 beds, but in the rainy season there can be more than 500 inpatients per day in the department, largely because of the high malaria burden at that time of year [20]. In Malawi, malaria is transmitted throughout the year, but its incidence increases after rainfall during November to April [21]. Owing to plentiful rainfall, year-round high temperatures and high levels of humidity, malaria vectoral capacity is high in Malawi [21]. The intensity of malaria transmission varies geographically and areas along the lake have a higher prevalence [22].
Study design
This retrospective cohort study was undertaken at KCH. The data used were from an existing Research Electronic Data Capture (REDCap©) database [12] of children admitted to the paediatric ward at KCH from September 2017 to September 2019. From September 2017 to December 2018, this database included children aged 6–36 months. From January 2019 to September 2019, the database was expanded to include children aged 2 weeks to 60 months. Children with Hb levels ≤6 g/dL identified with the HemoCue system were included [24]. If patients had more than one Hb level measured during hospitalisation, the lowest value was used for inclusion and categorisation of severity. Patients admitted for trauma were excluded from this analysis, as well as patients with unknown or missing discharge data. All data were de-identified.
In-hospital mortality was analysed as the primary outcome and was compared between those who did and did not receive a blood transfusion. Differences between the two groups were also compared, stratified by anaemia severity, malaria status and the presence of clinical signs of instability. Malaria status was defined as either positive or negative based on a malaria rapid diagnostic test (MRDT) [25].
Identification of anaemia
Data were analysed using two Hb cut-offs: severe anaemia (4–6 g/dL) and profound anaemia (<4 g/dL). The WHO guidelines for the care of hospitalised children with anaemia recommend transfusing all children with Hb <4 g/dL, and 4–6 g/dL in children with clinical signs of instability such as shock or respiratory distress [10]. Although the Malawian guidelines state the same, given the severe shortage of blood, many clinically unstable children do not receive a transfusion until their haemoglobin levels are ≤5 g/dL [26].
Clinical signs of instability
In-hospital mortality was calculated by receipt of transfusion and clinical signs of instability in children with Hb 4–6 g/dL, as this is the group in whom clinical signs influence receipt of transfusion. Of the signs and symptoms stated by WHO to indicate clinical instability, this database included delayed capillary refill (yes or no), impaired consciousness (Blantyre coma scale score, normal or abnormal) and convulsions (yes or no), and respiratory distress (yes or no). The database did not include reliable information on heart failure or levels of parasitaemia.
Statistical analysis
Demographics and patient characteristics of children admitted to KCH with SA were summarised using descriptive statistics by receipt of blood transfusion. Differences in these statistics were assessed using t-tests for continuous variables and χ2 tests for nominal variables. Hospital mortality was compared between those who were and were not transfused. These groups were then stratified according to Hb (<4 vs 4–6 g/dL), MRDT status (positive vs negative), Hb <4 g/dL plus MRDT status, and Hb 4–6 g/dL plus MRDT status. Differences within each subgroup were compared using χ2 tests. Logistic regression models were used to estimate odds ratios and 95% confidence intervals. Models were first unadjusted, then adjusted for age and sex. To assess the impact of clinical signs of instability on the relationship between in-hospital mortality and transfusion, the analysis was repeated in children with Hb 4–6 g/dL using the clinical signs as subgroups. All analyses were performed using SAS version 9.4 (Cary, NC, USA) [27].
Ethics
Institutional review board approval was obtained from the University of Utah (IRB_00116377) and the National Health Sciences Research Committee in Malawi. As this was a retrospective database, the ethics committees waived the requirement for subject consent.
Results
Demographics
During the study period, 1761 children with SA were admitted (Table 1), 989 (56.2%) of whom received a blood transfusion during their hospitalisation. The mean age of the children who received a transfusion was 22.4 months and the mean age of those who did not was 20.2 months (p<0.01). Those who received a blood transfusion had a mean (SD) MUAC of 13.8 cm (1.3) and those who did not had a mean (SD) MUAC of 13.7 cm (1.4) (p=0.21). Since a normal MUAC is >13.5 cm, the majority of the patients were well nourished. Those who received a blood transfusion had a significantly longer length of hospital stay (4.3 days) than those who did not (3.7) days (p<0.01). In addition, those who received a blood transfusion had a significantly lower Hb level (4.1 g/dL) than those who did not (4.6 g/dL) (p<0.01).
Table 1.
Demographics of 1761 patients with severe anaemia (Hb ≤6 g/dL).
| Missing data (n) | Transfusiona n=989 | No transfusion n=772 | p-valueb | |
|---|---|---|---|---|
| Average age, mths, mean (SD) | 4 | 22.4 (12.6) | 20.3 (11.1) | <0.01 |
| <1 year, n (%) | 232 (23.5) | 217 (28.1) | 0.03 | |
| Male, n (%) | 8 | 529 (53.7) | 398 (51.9) | 0.46 |
| MUAC, cmc, mean (SD) | 90 | 13.8 (1.3) | 13.7 (1.4) | 0.21 |
| MRDT+d, n (%) | 80 | 800 (84.6) | 608 (82.7) | 0.31 |
| Average length of stay, days, mean (SD) | 0 | 4.3 (2.1) | 3.7 (2.1) | <0.01 |
| Lowest Hb, mean (SD) | 0 | 4.1 (1.2) | 4.6 (1.1) | <0.01 |
| <4 | 400 (40.4) | 203 (26.3) | <0.01 | |
| ≥4 | 589 (59.6) | 569 (73.7) | ||
| Total transfusions | 0 | |||
| 0 | 0 | 772 (100.0) | ||
| 1 | 852 (86.1) | 0 (0.0) | ||
| 2 | 121 (12.2) | 0 (0.0) | ||
| ≥3 | 15 (1.5) | 0 (0.0) | ||
| BCS scoree | 369 | |||
| Normal (5) | 600 (78.3) | 499 (79.7) | 0.53 | |
| Abnormal (<5) | 166 (21.7) | 127 (20.3) | ||
| Delayed capillary refill, sec | 347 | |||
| No (≤3) | 512 (66.7) | 484 (74.9) | <0.01 | |
| Yes (>3) | 256 (33.3) | 162 (25.1) | ||
| Convulsions | 307 | |||
| No | 661 (84.5) | 585 (87.1) | 0.17 | |
| Yes | 121 (15.5) | 87 (12.9) | ||
| Respiratory distress | 252 | |||
| No | 673 (81.7) | 571 (83.4) | 0.39 | |
| Yes | 151 (18.3) | 114 (16.6) | ||
| Presence of ≥1 clinical sign | 82 | |||
| No | 405 (43.4) | 356 (47.8) | 0.07 | |
| Yes | 529 (56.6) | 389 (52.2) |
Received a blood transfusion either on arrival or during hospitalisation;
p-values in bold are statistically significant;
mid-upper-arm circumference is a measurement of nutritional status;
MRDT positive either on arrival or during hospitalisation;
Blantyre coma scale is a modification of the paediatric Glasgow coma scale and is designed to assess malarial coma in children.
Overall mortality rates
The overall in-hospital mortality rate for this cohort was 7.2%; 126 of 1761 patients died. The in-hospital mortality rate did not differ by sex or age. Of those who died, 50.8% were male (p=0.63); the mean (SD) age of patients who died was 20.4 months (13.7) vs 21.5 months (11.6) in survivors (p=0.38). The in-hospital mortality rate for those with severe anaemia (Hb 4–6 g/dL) was 6.0% (70 of 1158 patients). The in-hospital mortality rate for patients with profound anaemia (Hb <4 g/dL) was 9.3% (56 of 603 patients). The overall in-hospital mortality rate for patients who received a blood transfusion was 7.8% (60 of 772 patients) and for those who did not receive a blood transfusion was 6.7% (66 of 989 patients). The mortality rate in patients who received a single transfusion was 7.4% (63 of 852 patients) and in those who received several transfusions it was 2.2% (3 of 137 patients).
Effect of blood transfusion by malaria status and severity of anaemia
Sixty-six (52.38%) of the 126 patients who died received a transfusion and 923 (56.45%) of those who lived received a transfusion [OR 0.85 (0.59–1.22), p=0.38]. Of the 126 patients who died, 18 (14.3%) had a transfusion recorded as ordered but died before receiving it. In subgroup analyses of all children with SA who did or did not receive a blood transfusion during hospitalisation, the only subgroup with a significant difference in in-hospital mortality were those who had both malaria and profound anaemia (Hb <4 g/dL) (Table 2). Children with malaria and profound anaemia who received a transfusion were 57% less likely to die in the hospital than those who did not receive a transfusion [unadjusted OR 0.43 (0.22–0.85), p=0.01; adjusted OR 0.40 (0.20–0.81), p=0.01]. While there was a difference in mortality with transfusion in all patients with profound anaemia [unadjusted OR 0.55 (0.32–0.96), p=0.04; adjusted OR 0.54 (0.31–0.95), p=0.03], subgroup analysis of those with profound anaemia without malaria showed no detectable impact on mortality according to whether they did or did not receive blood [unadjusted OR 0.95 (0.26–3.41), p=0.93; adjusted OR 0.93 (0.23–3.83), p=0.92]. In children with Hb 4–6 g/dL, the odds of in-hospital death did not differ by receipt of transfusion [unadjusted OR 1.02 (0.63–1.66), p=0.92; adjusted OR 1.01 (0.62–1.65), p=0.97], and this held true despite a positive [unadjusted OR 1.03 (0.55–1.94), p=0.93; adjusted OR 1.03 (0.55–1.95), p=0.92] or negative [unadjusted OR 1.32 (0.57–3.06), p=0.52; adjusted OR 1.21 (0.51–2.88), p=0.66] malaria status.
Table 2.
In-hospital mortality by receipt of blood transfusion in subgroups
| In-hospital mortality n (%) | Discharged alive n (%) | p-valuea | In-hospital mortality n (%) | Discharged alive n (%) | p-value | |
|---|---|---|---|---|---|---|
| Patients with Hb <4 g/dL | Patients with Hb 4–6 g/dL | |||||
| Received transfusion | 30 (53.57) | 370 (67.64) | 36 (51.43) | 553 (50.83) | ||
| No transfusion | 26 (46.43) | 177 (32.36) | 34 (48.57) | 535 (49.17) | ||
| Unadjusted odds ratio (95% CI) | 0.55 (0.32–0.96) | 0.04 | 1.02 (0.63–1.66) | 0.92 | ||
| Adjusted odds ratiob (95% CI) | 0.54 (0.31–0.95) | 0.03 | 1.01 (0.62, 1.65) | 0.97 | ||
| Patients MRDT+ | Patients MRDT− | |||||
| Received transfusion | 38 (50.00) | 762 (57.21) | 22 (57.89) | 124 (52.77) | ||
| No transfusion | 38 (50.00) | 570 (42.79) | 16 (42.11) | 111 (47.23) | ||
| Unadjusted odds ratio (95% CI) | 0.75 (0.47–1.19) | 0.22 | 1.23 (0.62–2.46) | 0.56 | ||
| Adjusted odds ratio (95% CI) | 0.74 (0.47–1.18) | 0.21 | 1.20 (0.59–2.45) | 0.61 | ||
| Patients with Hb <4 g/dL and MRDT+ | Patients with Hb <4 g/dL and MRDT− | |||||
| Received transfusion | 17 (47.22) | 308 (67.69) | 9 (69.23) | 50 (70.42) | ||
| No transfusion | 19 (52.78) | 147 (32.31) | 4 (30.77) | 21 (29.58) | ||
| Unadjusted odds ratio (95% CI) | 0.43 (0.22–0.85) | 0.01 | 0.95 (0.26–3.41) | 0.93 | ||
| Adjusted odds ratio (95% CI) | 0.40 (0.20–0.81) | 0.01 | 0.93 (0.23–3.83) | 0.92 | ||
| Patients with Hb 4–6 g/dL and MRDT + | Patients with Hb 4–6 g/dL and MRDT − | |||||
| Received transfusion | 21 (52.50) | 454 (51.77) | 13 (52.00) | 74 (45.12) | ||
| No transfusion | 19 (47.50) | 423 (48.23) | 12 (48.00) | 90 (54.88) | ||
| Unadjusted odds ratio (95% CI) | 1.03 (0.55–1.94) | 0.93 | 1.32 (0.57–3.06) | 0.52 | ||
| Adjusted odds ratio (95% CI) | 1.03 (0.55–1.95) | 0.92 | 1.21 (0.51–2.88) | 0.66 | ||
p-values in bold are statistically significant;
adjusted odds ratio adjusted for age and sex.
Effect of blood transfusion in children with clinical signs of instability
In patients with SA (Hb 4–6 g/dL), children with at least one clinical sign of instability had a mortality rate of 9.69% compared with 4.07% in children with no clinical signs of instability (p<0.01).
In the same subgroup of patients, there was no significant difference in the odds of in-hospital mortality between those who received a transfusion and those who did not [unadjusted OR 1.16 (0.65–2.05), p=0.62; adjusted OR 1.13 (0.63–2.02), p=0.68]. There was no significant difference in the odds of mortality in patients with no clinical sign of instability who did and did not receive a transfusion [unadjusted OR 0.69 (0.24–1.97), p=0.49; adjusted OR 0.72 (0.25–2.07), p=0.54]. When broken down into different clinical signs of instability, transfusion did not significantly influence mortality in any subgroup (Table 3).
Table 3.
In-hospital mortality by receipt of blood transfusion and clinical signs of instability in children with Hb 4–6 g/dL.
| In-hospital mortality n (%) | Discharged alive n (%) | p-value | In-hospital mortality n (%) | Discharged alive n (%) | p-value | |
|---|---|---|---|---|---|---|
| Delayed capillary refill (>3 sec) | No delayed capillary refill (≤3 sec) | |||||
| Received transfusion | 11 (64.71) | 137 (54.80) | 14 (42.42) | 297 (47.29) | ||
| No transfusion | 6 (35.29) | 113 (45.20) | 19 (57.58) | 331 (52.71) | ||
| Unadjusted odds ratio (95% CI) | 1.51 (0.54–4.22) | 0.43 | 0.82 (0.41–1.67) | 0.59 | ||
| Adjusted odds ratioa (95% CI) | 1.43 (0.50–4.06) | 0.51 | 0.84 (0.41–1.70) | 0.62 | ||
| Patients with convulsions | Patients with no convulsions | |||||
| Received transfusion | 10 (62.50) | 71 (50.00) | 22 (51.16) | 361 (48.01) | ||
| No transfusion | 6 (37.50) | 71 (50.00) | 21 (48.84) | 391 (51.99) | ||
| Unadjusted odds ratio (95% CI) | 1.67 (0.58–4.83) | 0.35 | 1.14 (0.61–2.10) | 0.69 | ||
| Adjusted odds ratio (95% CI) | 1.65 (0.57–4.81) | 0.36 | 1.11 (0.60–2.09) | 0.73 | ||
| Patients with respiratory distress | Patients with no respiratory distress | |||||
| Received transfusion | 17 (53.13) | 75 (51.72) | 15 (50.00) | 385 (49.30) | ||
| No transfusion | 15 (46.88) | 70 (48.28) | 15 (50.00) | 396 (50.70) | ||
| Unadjusted odds ratio (95% CI) | 1.06 (0.49–2.28) | 0.89 | 1.03 (0.50–2.13) | 0.94 | ||
| Adjusted odds ratio (95% CI) | 1.09 (0.50–2.36) | 0.83 | 0.97 (0.46–2.05) | 0.94 | ||
| Abnormal BCS score (<5) | Normal BCS score (5) | |||||
| Received transfusion | 9 (52.94) | 97 (52.72) | 15 (45.45) | 338 (49.20) | ||
| No transfusion | 8 (47.06) | 87 (47.28) | 18 (54.55) | 349 (50.80) | ||
| Unadjusted odds ratio (95% CI) | 1.01 (0.37–2.73) | 0.99 | 0.86 (0.43–1.74) | 0.67 | ||
| Adjusted odds ratio (95% CI) | 1.08 (0.40–2.95) | 0.88 | 0.82 (0.40–1.68) | 0.58 | ||
| At least one clinical sign of instability | No clinical signs of instability | |||||
| Received transfusion | 28 (55.90) | 291 (51.32) | 6 (40.00) | 231 (49.15) | ||
| No transfusion | 23 (45.10) | 276 (48.68) | 9 (60.00) | 239 (50.85) | ||
| Unadjusted odds ratio (95% CI) | 1.16 (0.65–2.05) | 0.62 | 0.69 (0.24–1.97) | 0.49 | ||
| Adjusted odds ratio (95% CI) | 1.13 (0.63–2.02) | 0.68 | 0.72 (0.25–2.07) | 0.54 | ||
| Patients with two or more clinical signs | Patients with less than two clinical signs | |||||
| Received transfusion | 14 (63.64) | 75 (55.15) | 20 (45.45) | 447 (49.61) | ||
| No transfusion | 8 (36.36) | 61 (44.85) | 24 (54.55) | 454 (50.39) | ||
| Unadjusted odds ratio (95% CI) | 1.42 (0.56–3.62) | 0.46 | 0.85 (0.46–1.55) | 0.59 | ||
| Adjusted odds ratio (95% CI) | 1.53 (0.6–3.92) | 0.38 | 0.81 (0.44–1.51) | 0.52 | ||
Adjusted odds ratio adjusted for age and sex.
Discussion
This study assessed whether receipt of blood transfusion was associated with lower odds of mortality in children with SA with and without malaria who met WHO-defined global transfusion criteria. In this cohort, the only subgroup with a statistically significant difference in odds of in-hospital mortality by transfusion status were children with both profound anaemia (Hb <4 g/dL) and malaria. While this analysis detected an effect of transfusion for all children with profound anaemia in aggregate, subgroup analysis demonstrates that this difference was attributed to those with malaria since children with profound anaemia in the absence of malaria had no difference in the odds of mortality related to receipt of blood. Similarly, children with Hb 4–6 g/dL had no detectable difference in the odds of mortality based on receipt of blood transfusion, regardless of malaria status. Although children with Hb 4–6 g/dL and at least one clinical sign of instability had a significantly higher mortality rate than children with no clinical signs of instability, receipt of blood transfusion in these children did not significantly affect the odds of mortality.
While this current study supports previous studies which found that blood transfusion was protective in patients with profound anaemia [1,2], the findings contradict those of a study in Kenya which did not observe lower mortality rates in children with SA who were transfused compared with those who were not [5]. However, the Kenyan study defined children with severe anaemia as Hb <5 g/dL and did not perform stratified analyses for different haemoglobin levels. This study suggests that in children with profound anaemia the benefit of transfusion may be related to malaria status. This differential effect of transfusions on children with profound anaemia with and without malaria suggests that patients with malaria may have a more acute presentation of anaemia and may not have had time to make physiological adaptations. This is supported by a study which showed that, in children with severe anaemia, malaria caused a more acute drop in haemoglobin than in children without malaria [28]. Anaemia secondary to malaria involves both acute haemolysis and increased clearance of red blood cells that is an ongoing process during malaria disease, resulting in a rapid drop in haemoglobin [29]. In more chronic causes of anaemia, patients often have a slower haemoglobin decline, giving their bodies a chance to adjust. Another possible explanation for the different effects of blood transfusions in children with profound anaemia with and without malaria is that the transfused blood may contain immunoglobins which protect against malaria since the blood comes from surrounding malarious areas.
The findings of this study support current WHO guidelines to transfuse children with profound anaemia [10]. However, current WHO guidelines on transfusion for severe anaemia in LMIC suggest transfusing children with Hb 4–6 g/dL if they have very high malaria parasitaemia, but do not consider malaria status (positive versus negative) [10]. In LMIC settings such as Malawi which may not have rapid access to malaria smears, malaria status may be determined by MRDT positive or negative status. Decisions and guidelines influencing the use of this critical and limited resource should be informed by data showing a demonstrable positive effect. The association observed in this study between survival and receipt of blood in children with profound anaemia and malaria suggests that this group should be considered for transfusion when blood is a limited resource. Furthermore, these findings raise important questions for subsequent studies with more rigorous designs and analyses.
In this study, children with Hb 4–6 g/dL had no statistically significant difference in the odds of mortality by receipt of blood transfusion, regardless of malaria status. This is consistent with recent findings by a study in Uganda and Malawi which showed that, in children with uncomplicated severe anaemia with haemoglobin levels of 4–6 g/dL, there was no significant difference in clinical outcome in children who received an immediate transfusion and those who did not [30]. Furthermore, this is supported by a Cochrane Database Systematic Review which showed that there are insufficient data to show that giving blood in clinically stable children with severe anaemia in areas with malaria reduces death [31].
Current WHO recommendations advise transfusion not only in all children with Hb <4 g/dL but also for those with Hb 4–6 g/dL with signs of clinical instability. In this study, mortality in children with at least one clinical sign of instability was significantly greater than in children with no clinical signs of instability. These findings are supported by a study in Kenya which showed that, in children with either malaria or anaemia, respiratory distress and neurological impairment were both associated with a greater risk of death [1]. Furthermore, another study also in Kenya found that the presence of respiratory distress increased mortality rates in children with severe anaemia [2]. However, in this study, receipt of blood transfusion in children with Hb 4–6 g/dL and clinical signs of instability did not significantly affect the odds of mortality. Thus, further studies are needed in this group of children in order to guide the rational use of blood in the context of limited supplies.
This study was limited by the post-hoc nature of the database in which some entries were incomplete because the data were limited to what was documented by clinicians. In addition, the data were collected by data clerks and so some entries for clinical signs could be inaccurate. Specifically, capillary refill may have been inaccurate owing to potential data entry errors related to interpretation of the clinical charts by the data clerks, so conclusions relating to this clinical sign must be guarded. The database did not include information on time of transfusion, so conclusions could not be drawn regarding time to transfusion or transfusion delays. However, as 14.3% of the deaths occurred after a transfusion was ordered but before it was given, it is likely that transfusion delays owing to limited blood supply did play a role. Also, the database did not contain clinical information on dehydration, heart failure, shock or levels of parasitaemia which could have assisted in determining the clinical severity of anaemia or malaria in these patients. Furthermore, since an MRDT can remain positive for up to 2 months after treatment [32], it is possible that, even in a patient with a positive MRDT, acute malaria might not have been the sole cause of SA. Apart from malaria, studies in other regions of Malawi have found that other infections, G6PD deficiency, alpha-thalassaemia, sickle cell disease and nutritional deficiencies are large contributors of anaemia in Malawian children [15,33]. Thus, it is possible that, in this study, there were several causes of anaemia, but, given the lack of such laboratory tests at KCH, other aetiologies could not be sufficiently investigated. In addition, in this setting there is no rapid access to malaria parasite smears, and thus MRDT is the standard of care test for malaria. Hence, its use for this analysis was justified as this is what is used at KCH. Furthermore, in some of the sub-analyses, owing to small numbers, some of the results should be interpreted with caution. However, sub-cohorts in whom a clear benefit of blood transfusion was demonstrated could inform systems-based process improvement for timely provision of blood to these high-risk cohorts. Finally, the primary outcome of this study was in-hospital mortality which inherently does not include children who may have died after discharge. Future studies should include additional follow-up of these patients to establish long-term outcome.
In conclusion, the study suggests that blood transfusion in children with profound anaemia and malaria may improve in-hospital survival. There was no improvement in in-hospital survival in children with Hb 4–6 g/dL who received blood regardless of malaria status and regardless of whether they had clinical signs of instability. Further randomised controlled studies should be conducted to determine which children may benefit most from blood transfusion. These findings raise important questions for subsequent prospective studies with more rigorous designs and analyses. Understanding which children with severe anaemia could benefit from blood transfusion should lead to more focused interventions for timely transfusion and a more rational approach to the use of a critical and limited resource.
Acknowledgments
The authors wish to thank the data clerks who collected all relevant patient data, the KCH paediatric staff for their diligent care of the patients, and all the patients and caregivers for their involvement in the study.
Funding: The statistical analysis in this investigation was supported by the University of Utah Population Health Research (PHR) Foundation, with funding in part from the National Center for Research Resources and the National Center for Advancing Translational Sciences, National Institutes of Health, through Grant 5UL1TR001067-05 (formerly 8UL1TR000105 and UL1RR025764). In addition, the database described was supported by the National Center for Advancing Translational Sciences (NCATS), National Institutes of Health, through Grant Award Number UL1TR001111. The content is solely the responsibility of the authors and does not necessarily represent the official views of the NIH.
Abbreviations:
- BCS
Blantyre coma scale
- Hb
haemoglobin
- KCH
Kamuzu Central Hospital
- LMIC
low- and middle-income countries
- MRDT
malaria rapid diagnostic test
- SA
severe anaemia
- SMA
severe malaria with anaemia
- SSA
sub-Saharan Africa
- WHO
World Health Organization
Biographical Note:
Elizabeth M. Keating is a fellow in Pediatric Emergency Medicine and Global Health at University of Utah, USA.
Msandeni Chiume is head of the Department of Paediatrics, Kamuzu Central Hospital, Malawi.
Elizabeth Fitzgerald is an Assistant Professor of Pediatric Emergency Medicine, University of North Carolina Chapel Hill School of Medicine, USA, and Director of UNC Pediatric Global Health.
Yamikani Mgusha is the Deputy Head of the Department of Paediatrics, Kamuzu Central Hospital, Malawi.
Tisungane Mvalo is a Senior Clinical Research Investigator and Pediatrician at the University of North Carolina Project Malawi and Research Assistant Professor at the Department of Pediatrics at the University of North Carolina at Chapel Hill, North Carolina, USA.
Nora Fino is a biostatistician in the Division of Epidemiology at the University of Utah, USA.
Heather L. Crouse is an Associate Professor of Pediatrics in Pediatric Emergency Medicine, Baylor College of Medicine, USA.
Michelle Eckerle is an Assistant Professor of Pediatrics in Pediatric Emergency Medicine, Cincinnati Children’s Hospital, USA.
Kathleen Gorman is a student of public health at the London School of Hygiene and Tropical Medicine, UK.
Emily J. Ciccone is an Infectious Diseases fellow, University of North Carolina Chapel Hill School of Medicine, USA.
Gladstone Airewele is is an Associate Professor of Pediatrics in Pediatric Hematology and Oncology, Baylor College of Medicine, USA.
Jeff A. Robison is an Associate Professor of Pediatrics in Pediatric Emergency Medicine, University of Utah, USA.
Footnotes
Conflict of interest: None.
Data Availability Statement: The data for the findings of this study are available from Kamuzu Central Hospital at msandeni@gmail.com on reasonable request.
References
- 1.English M, Ahmed M, Ngando C, et al. Blood transfusion for severe anaemia in children in a Kenyan hospital. Lancet. 2002;359:494–495. [DOI] [PubMed] [Google Scholar]
- 2.Lackritz EM, Campbell CC, Ruebush TK, et al. Effect of blood transfusion on survival among children in a Kenyan hospital. Lancet. 1992;340:524–528. [DOI] [PubMed] [Google Scholar]
- 3.Koram KA, Owusu-Agyei S, Utz G, et al. Severe anemia in young children after high and low malaria transmission seasons in the Kassena-Nankana district of northern Ghana. Am J Trop Med Hyg. 2000;62:670–674. [DOI] [PubMed] [Google Scholar]
- 4.Newton CR, Warn PA, Winstanley PA, et al. Severe anaemia in children living in a malaria endemic area of Kenya. Trop Med Int Health. 1997;2:165–178. [DOI] [PubMed] [Google Scholar]
- 5.Obonyo CO, Vulule J, Akhwale WS, et al. In-hospital morbidity and mortality due to severe malarial anemia in western Kenya. Am J Trop Med Hyg. 2007;77:23–28. [PubMed] [Google Scholar]
- 6.Schellenberg D, Menendez C, Kahigwa E, et al. African children with malaria in an area of intense Plasmodium falciparum transmission: features on admission to the hospital and risk factors for death. Am J Trop Med Hyg. 1999;61:431–438. [DOI] [PubMed] [Google Scholar]
- 7.Thomas J, Ayieko P, Ogero M, et al. Blood transfusion delay and outcome in county hospitals in Kenya. Am J Trop Med Hyg. 2017;96:511–517. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Maitland K, Pamba A, English M, et al. Pre-transfusion management of children with severe malarial anaemia: a randomised controlled trial of intravascular volume expansion. Br J Haematol. 2005;128:393–400. [DOI] [PubMed] [Google Scholar]
- 9.Fitzgerald E, Mlotha-Mitole R, Ciccone EJ, et al. A pediatric death audit in a large referral hospital in Malawi. BMC Pediatr. 2018;18:75. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.World Health Organization. Pocket Book of Hospital Care for Children: Guidelines for the Management of Common Childhood Illnesses. Geneva: WHO; 2013. Available from: http://www.who.int/iris/handle/10665/81170. [PubMed] [Google Scholar]
- 11.Austin N, Adikaibe E, Ethelbert O, et al. Prevalence and severity of malaria parasitemia among children requiring emergency blood transfusion in a tertiary hospital in Imo State, Nigeria. Ann Med Health Sci Res. 2014;4:619–623. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.World Health Organization. Hospital Care for Children: Guidelines for the Management of Common Illnesses with Limited Resources. Geneva: WHO; 2005. Available from: https://www.who.int/maternal_child_adolescent/documents/9241546700/en/ [Google Scholar]
- 13.World Health Organization. Malaria 2018. Available from: http://www.who.int/malaria/areas/high_risk_groups/children/en/.
- 14.Barreiro P, Tiziano G, Fano H, et al. Malaria and severe anemia over eight years at Gambo Rural Hospital, southern Ethiopia. Pathog Glob Health. 2017;111:195–199. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Calis JC, Phiri KS, Faragher EB, et al. Severe anemia in Malawian children. N Engl J Med. 2008;358:888–899. [DOI] [PubMed] [Google Scholar]
- 16.Kiguli S, Maitland K, George EC, et al. Anaemia and blood transfusion in African children presenting to hospital with severe febrile illness. BMC Med. 2015;13:21. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Lackritz EM, Hightower AW, Zucker JR, et al. Longitudinal evaluation of severely anemic children in Kenya: the effect of transfusion on mortality and hematologic recovery. Aids. 1997;11:1487–1494. [DOI] [PubMed] [Google Scholar]
- 18.Maitland K, Kiguli S, Opoka RO, et al. Mortality after fluid bolus in African children with severe infection. N Engl J Med. 2011;364:2483–2495. [DOI] [PubMed] [Google Scholar]
- 19.Mpoya A, Kiguli S, Olupot-Olupot P, et al. Transfusion and treatment of severe anaemia in African children (TRACT): a study protocol for a randomised controlled trial. Trials. 2015;16:593. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Eckerle M, Crouse HL, Chiume M, et al. Building sustainable partnerships to strengthen pediatric capacity at a government hospital in Malawi. Front Public Health. 2017;5:183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Bennett A, Kazembe L, Mathanga DP, et al. Mapping malaria transmission intensity in Malawi, 2000–2010. Am J Trop Med Hyg. 2013;89:840–849. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Chirombo J, Ceccato P, Lowe R, et al. Childhood malaria case incidence in Malawi between 2004 and 2017: spatio-temporal modelling of climate and non-climate factors. Malar J 2020;19:5. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap) – a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42:377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Medina Lara A, Mundy C, Kandulu J, et al. Evaluation and costs of different haemoglobin methods for use in district hospitals in Malawi. J Clin Pathol. 2005;58:56–60. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Abba K, Deeks JJ, Olliaro P, et al. Rapid diagnostic tests for diagnosing uncomplicated P. falciparum malaria in endemic countries. Cochrane Database Syst Rev. 2011:Cd008122. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Queen Elizabeth Central Hospital, The College of Medicine. Electronic protocols for the management of common childhood illnesses in Malawi. Blantyre, Malawi, 2018. Available from: https://www.openguidelines.net/data/set_1000/html/_chapters.htm. [Google Scholar]
- 27.SAS software. Cary, NC, USA: SAS Institute Inc.; 2002–2012. [Google Scholar]
- 28.Sumbele IU, Sama SO, Kimbi HK, et al. Malaria, moderate to severe anaemia, and malarial anaemia in children at presentation to hospital in the Mount Cameroon area: a cross-sectional study. Anemia. 2016;2016:5725634. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Kiggundu VL, O’Meara WP, Musoke R, et al. High prevalence of malaria parasitemia and anemia among hospitalized children in Rakai, Uganda. PLoS One. 2013;8:e82455. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Maitland K, Kiguli S, Olupot-Olupot P, et al. Immediate transfusion in African children with uncomplicated severe anemia. N Engl J Med. 2019;381:407–419. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 31.Meremikwu M, Smith HJ. Blood transfusion for treating malarial anaemia. Cochrane Database Syst Rev. 2000;1999:CD001475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Dalrymple U, Arambepola R, Gething PW, et al. How long do rapid diagnostic tests remain positive after anti-malarial treatment? Malar J 2018;17:228. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.McGann PT, Williams AM, Ellis G, et al. Prevalence of inherited blood disorders and associations with malaria and anemia in Malawian children. Blood Adv. 2018;2:3035–3044. [DOI] [PMC free article] [PubMed] [Google Scholar]