| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Beginning from the 1980s, the incidence of SMA has increased in parallel with the increasing prevalence of antimalarial drug resistance.10 Two major presentations of SMA occur; an acute episode of clinical disease or a consequence of the slow and insidious process of repeated, often asymptomatic, malaria infections.11 Severe anemia may develop rapidly in the course of a malaria illness especially in the presence of a high parasite density.12 Case management of SMA is complicated because of diagnostic difficulties in the absence of a supportive diagnostic laboratory, and the difficult decision on the use of blood transfusion. The mortality risk due to SMA is higher than that due to other causes of severe pediatric anemia13 and death often occurs soon after admission.1,14,15 Case fatality rates in children with SMA increases with young age and transmission intensity.7,16 Management of SMA often includes a blood transfusion, which has become an important risk factor for transmission of the human immunodeficiency virus (HIV) in sub-Saharan Africa.17
Over 30% of pediatric admissions to district hospitals in Kenya are due to malaria. The cost to the healthcare system of managing a case of severe malaria is substantial.18 To optimize survival, it is necessary that these facilities be adequately equipped with trained staff, drugs, and supplies to deal with the rising burden of SMA. Specifically, the availability and appropriate use of blood transfusion and antimalarial drugs could substantially impact patient survival.1,3,19 As part of a regional survey on the quality of clinical care available for severe pediatric malaria, we assessed the contribution of SMA and the effect of treatment practices to the overall malaria-related mortality in a typical Kenyan district hospital.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
During the study period, intravenous quinine was the drug of choice for severe malaria, whereas sulfadoxine/pyrimethamine (SP) was the recommended first-line therapy for uncomplicated falciparum malaria. The management of severe anemia (any cause) in the hospital is HIV-negative whole blood. During the study period, there was no change in the level of nursing and clinical care. A high prevalence of both chloroquine21 and SP resistance22 has been recorded in children with uncomplicated falciparum malaria in this hospital.
Data sources. This study was part of a larger study evaluating the quality of in-patient care for severe pediatric malaria in western Kenya. In October 2003, we conducted a retrospective review of a random sample of 100 case records per month for children aged between 1 month and 9 years who were admitted to the pediatric inpatient department at SDH with severe malaria January through December 2002. The children were identified through the admission register of the pediatric department. At this hospital, hospital records contain all information on a patient from admission to discharge or death recorded by nurses, laboratory staff, and clinicians. Data collected for the sample of children selected included the following: name, age, sex, inpatient number, date and time of admission and discharge (or death), admission hemoglobin and malaria smear result, diagnosis, treatments received (e.g., antimalarial, blood, IV fluids, oxygen), and the outcome of hospitalization (i.e., absconded, died, discharged alive).
At this hospital, hemoglobin was measured using the HemoCue® machine (Angelholm, Sweden). Thick and thin malaria smears were stained using Giemsa and read by experienced microscopists. Parasitemia was variably quantified using either the scant to heavy or the plus (+) system. Because of this variability, we used the presence or absence of parasitemia for this analysis.
Definitions.
Malaria was defined as the presence of asexual parasitemia of any non-zero density. Anemia was defined as hemoglobin (Hb) concentration < 11.0 g/dL and categorized as severe (Hb level 
5.0 g/dL), moderate (Hb 5.1–7.9 g/dL), and mild (Hb 8.0–10.9 g/dL) anemia. Severe malarial anemia (SMA) was defined as Hb level 5.0 g/dL in the presence of malaria parasitemia. Malaria-associated mortality was defined as death before discharge of a child admitted with severe anemia and having parasitemia.
Statistical methods.
Data was entered, cleaned, and validated using EPI Info v. 6.04d (Centers for Disease Control and Prevention, Atlanta, GA) and analyzed using SPSS for Windows (version 12.0, SPSS Inc., Chicago, IL). Descriptive statistics were used to compute the frequencies, proportions, means, and confidence intervals. For data not conforming to normal distribution, medians and inter-quartile ranges (IQR) were computed. P values < 0.05 were considered statistically significant. We compared the mortality rates between children with Hb 5.0 versus Hb > 5.0 g/dL and transfused versus not transfused, using relative risk measures, and present the results as the point estimate together with the 95% confidence intervals (CI). When the 95% confidence interval for a relative risk included 1, then the value was considered statistically not significant.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
shows the characteristics of the 1,116 children whose records we sampled for this study. There were 592 (53%) male children and the mean age of study population was 16 months. A total of 497 (44.5%) children were aged below 1 year, 792 (71%) below 2 years, and 960 (86%) below 3 years. The mean age, body weight, duration of hospitalization, and hemoglobin were similar between the male and female children (data not shown). Blood smears were taken for 1,062 (95%) of the children and 1,032 (92.5%) had a hemoglobin measurement taken at admission. Only 54 (4.8%) and 84 (7.5%) children had no malaria smear or Hb results, respectively. Body weight measurements were taken for 869 (77.9%) children. A clinical diagnosis of malaria was the leading cause of admission (98.4%). The outcome of hospitalization was known for 1,104 (99%) of the children.
|
shows the age-specific parasite prevalence, which was not statistically different between the ages (
2 for trend = 3.26, P = 0.660). The highest parasite prevalence was among those aged 6–11 months (84.4%) and the lowest among those aged 48–59 months (76%). The prevalence of parasitemia was high throughout the year with significant differences between the months (2 for trend = 21.8, P = 0.026). The peak prevalence was between March and June with smaller peaks between September and January, corresponding to the long and short rains, respectively (Figure 2).
|
|
4.0g/dL) was present in 141 of 1,032 (13.7%) of all admissions. Overall, 206 of 233 (88.4%) children with severe anemia were below 3 years of age. The mean age of severely anemic children was significantly lower than that of children with Hb > 5.0 g/dL (14 versus 17 months, P = 0.003).
A total of 191 of 1,067 (18%) had SMA. Of these, 171 of 191 (89.5%), were below 3 years of age. The age-specific prevalence of SMA is shown in Figure 1
. Children aged 1–5 months had the highest SMA prevalence (59/227 [26.0%]) and the lowest prevalence was in those 48–59 months of age (5/52 [9.6%]). There was a negative correlation between SMA prevalence and age. The mean age of those with SMA was significantly lower than those without SMA (13.8 versus 17 months, P = 0.002).
The mean Hb was significantly lower in parasitemic compared with aparasitemic children (7.01 versus 7.74 g/dL, P < 0.001). Severe anemia was associated with parasitemia in 191 of 226 (84.5%) of the admissions. Among children with Hb level between 5.1 and 7.9 g/dL, 86.9% were parasitemic; among those with Hb 
8.0 g/dL, 78.4% were parasitemic.
Mortality due to severe malarial anemia.
Overall, the outcome of hospitalization was known for 1,104 (99%) children. Of these, 12 (1.1%) absconded and 83 (7.5%) are known to have died. Death was significantly associated with a diagnosis of anemia (RR = 1.99, 95% CI 1.32–3.01), severe anemia (RR = 2.04, 95% CI 1.32–3.16), or diarrhea (RR = 2.40, 95% CI 1.56–3.69). There was no difference in the risk of inhospital mortality among children with or without a diagnosis of malaria (RR = 0.44, 95% CI 0.15–1.77). Most deaths occurred early during hospitalization: 21 of 83 (40.4%) of the deaths occurred on the day of admission and 44 of 83 (53%) had occurred by 24 hours after admission. Children who died were hospitalized for a significantly shorter duration compared with those who survived (2.4 ± 2.8 versus 3.6 ± 3.0 days, P < 0.0001). In total, 33 of 83 (45.8%) and 74 of 83 (89%) of those who died were aged below 1 and below 3 years, respectively. The mean age of the children who died was not significantly different from that of those who survived: 15.2 versus 16.8 months, F-statistic = 1.04, P = 0.307. Figure 3 shows the age-specific mortality rates. Infants had the highest risk of dying in-hospital. Thereafter, the risk of in-hospital mortality decreased with increasing age.
|
Compared with weekdays (Monday through Friday), mortality over the weekends and out of working hours (between 5 PM and 7 AM) was substantially high: 11 of 53 (20.8%) versus 72 of 1,051 (6.9%), RR = 3.03 (95% CI 1.71–5.36), P < 0.0001.
Treatment practices.
Blood transfusions were given to 19.5% (217) of all the admissions. The transfusion rate varied greatly throughout the year with peaks in February and July, corresponding to the period after the short and long rains, respectively (Figure 2). Transfused children were significantly younger than non-transfused children (median age 11 versus 13 months, P = 0.020). Overall, 193 of 217 (88.9%) transfusions were given to children below 3 years of age (Figure 4A). At admission, transfused children had significantly lower mean Hb compared with those not transfused (4.3 ± 1.9 versus 7.9 ± 2.3 g/dL, P < 0.0001). About 151 of 217 (70%) transfusions were given to children with severe anemia (Figure 4B). The proportion of severely anemic children transfused was 151 of 233 (64.8%). Transfusion prescriptions in this hospital did not always adhere to the guidelines: 51 of 798 (6.4%) children with Hb > 5.0 g/dL were transfused, as were 15 of 84 (17.9%) others without hemoglobin measurement. Of the 191 children with SMA, 127 (66.5%) were transfused.
|
Antimalarial drugs were prescribed for 1,062 (98.2%) children, and of these, 870 (81.9%) had positive smears for malaria. A total of 174 (16.4%) children with negative malaria smears also received an antimalarial drug. Intravenous (IV) quinine was the most commonly prescribed antimalarial received by 903 (81%) of all admissions. A total of 105 (9.9%) children with negative smears also received IV quinine. Only 6 children with positive malaria smears received no antimalarial drug.
Overall, children treated with IV quinine had a non-significant increased risk of mortality compared with those treated with other antimalarial drugs: 70 of 892 (7.8%) versus 10 of 192 (5.2%), RR = 1.51, 95% CI 0.79–2.87, P = 0.204. A similar effect was observed in children with Hb 5 g/dL: 26 of 181 (14.4%) versus 2 of 44 (4.5%), RR = 3.16, 95% CI 0.78–12.8, P = 0.077. Among children with Hb > 5 g/dL, administration of IV quinine had no effect on the risk of in-hospital mortality (P = 0.97).
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our study confirms that in areas of high malaria transmission, severe anemia in young children is the most common presentation of P. falciparum.23,24 Malaria transmission in this western Kenya setting is intense and stable throughout the year, and malaria was the leading cause of admission and pediatric mortality. A high proportion of the children were already anemic at admission: 86% were anemic and 21% had severe anemia. Severe anemia occurred almost always in young children below 3 years of age and in association with malaria parasitemia. It is not clear whether the high prevalence of severe malarial anemia was due to resistant parasites that had failed therapy or from an acute-on-chronic, asymptomatic parasitemia. The rising prevalence of malaria-associated severe anemia in areas of high malaria transmission has been attributed to the emergence and intensification of drug resistance.21,25–28 Consequently, the frequency with which blood transfusions are administered has increased17 and similarly, malaria-specific mortality has increased.29 In Zaire, the Democratic Republic of Congo (ex-Zaire), when drug-resistant parasites emerged, the requirements for pediatric transfusions increased by 250%.30 Other possible explanations for the rising prevalence of severe anemia may include the changing epidemiology of pediatric HIV type 1. Recent studies in western Kenya have documented a high prevalence of severe malarial anemia in association with HIV infection or exposure.31,32 We could not assess the contribution of HIV to the high prevalence of severe anemia because information about the HIV status of the children in our study was not available.
Asymptomatic severe anemia per se is associated with low risk of mortality.1,8,33 The risk of mortality in a child with SMA increases with the presence of respiratory distress,1,8,14,15 impaired consciousness,8,15 or bacteremia.3,34 In our study, 23% of children with severe malaria-related anemia died. This is consistent with previous studies that found rates between 2.3 and 18%.1,15,33 Over half of the malaria-related deaths in our study were due to severe anemia, consistent with findings from Malawi over a decade ago, which highlighted the significant contribution of severe anemia to malaria-specific mortality in areas of high transmission.7 Death in severely anemic children occurred soon after admission. This highlights the need for early decision-making regarding the need for a transfusion and the availability of screened banked blood.19 Previous studies at our study site showed that severely anemic children have a high risk of inhospital mortality that extends up to 8 weeks in the post-discharge period.35 It is noteworthy that in our study, children hospitalized over the weekends and outside normal working hours were three times more likely to die compared with those admitted during weekdays and normal office hours. A possible explanation could be the compromised quality of care probably arising from understaffing. Similar findings have been found at Kilifi district hospital on the Kenyan coast.36
Transfusions were given to 20% of all admissions. However, in 70% of the cases transfusion was given to severely anemic children, the mortality rate among severely anemic children who were transfused versus non-transfused children was comparable. This could be explained by the case mix of the children we sampled that did not distinguish children at high risk of mortality. Indeed, there was a non-significant increase in the risk of mortality among transfused severely anemic children. We speculate that these children were sicker or were hospitalized too late so the available interventions could not save them. We did not record information on when transfusions actually occurred in relation to time of admission. This would have helped to confirm whether there was a benefit in transfusing children early during their hospitalization. Our findings confirm the results of a review that found poor correlation between transfusion rates and case fatality rates.37 Consistent with previous studies, not all severely anemic children were transfused: we did not evaluate the reasons why they were not transfused, but we suspect that similar factors found in the early 1990s at this hospital may still apply.19 Transfusion is an intervention of undetermined benefit that can transmit HIV, and should be limited to situations when it can improve survival.
Our study had several limitations. We reviewed retrospectively hospital discharge records, which has the advantage that data is readily available at low or no cost and contain routinely collected data, which may be useful for evaluating health services and for epidemiologic research. However, because multiple health workers collected the data for different purposes, the quality of the data may be questionable. Because of the cross-sectional design, we cannot make causality inferences. We included mainly children with admission hemoglobin measurements, so we may have missed some children with emergency transfusions or deaths that occurred on arrival before diagnostic laboratory investigations were undertaken. There were 84 children without hemoglobin measurements, of whom 5 (6%) died and 4 (5%) absconded. We analyzed mortality in the context of routine medical care, therefore we did not control for many factors that could impact the outcome (e.g., the use of supportive care). Clinicians at our study site did not always record information on whether the children were symptomatic or asymptomatic, the presence of respiratory distress, or impaired consciousness, which are known prognostic factors for mortality in children with severe anemia. Our estimates of malaria-specific mortality estimates may be a gross under-estimation because it is common for children to die at home or before arriving at a health facility.
Prognostic factors in a case of severe malaria-associated anemia include a delayed diagnosis, hyperparasitemia, respiratory distress, impaired consciousness, hypoglycemia, and a delayed transfusion.1,8,14,15 Death from severe falciparum malaria can be prevented by early diagnosis and prompt institution of effective antimalarial treatment. However, the condition is often recognized late and not all cases present to health facilities. It would be important to educate communities on early recognition of severe malaria, to strengthen referral systems, and to make pre-referral treatment available (e.g., rectal artesunate) or pre-transfusion interventions (e.g., intravenous fluids).38 The need for standardization in the definition and the training of health workers in what constitutes respiratory distress or impaired consciousness is paramount.39,40
Our study has identified severe anemia as a major cause of admission, which carries a high fatality rate among young children (< 3 years) who have malaria parasitemia. These young children should be the target of life-saving interventions.
Received August 21, 2006. Accepted for publication March 7, 2007.
Acknowledgments: This study was financially supported by an Operational Research grant from World Health Organization, Regional Office for Africa. The authors thank the Director, Kenya Medical Research Institute for permission to publish these results.
* Address correspondence to Charles O. Obonyo, Centre for Vector Biology & Control Research, Kenya Medical Research Institute, P.O. Box 1578, Kisumu 40100, Kenya. E-mail: Cobonyo{at}kisian.mimcom.net 
Authors’ addresses: Charles O. Obonyo and John M. Vulule, Centre for Vector Biology & Control Research, Kenya Medical Research Institute, P.O. Box 1578, Kisumu, Kenya 40100. E-mails: Cobonyo{at}kisian.mimcom.net and JVulule{at}kisian.mimcom.net. Willis S. Akhwale, Division of Malaria Control, Kenyan Ministry of Health, P.O. Box 20750, Nairobi, Kenya, E-mail: wilak4{at}yahoo.com. Diederick E. Grobbee, Julius Center for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, The Netherlands, E-mail: D.E.Grobbee{at}umcutrecht.nl
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  J. G. Breman, M. S. Alilio, and N. J. White Defining and Defeating the Intolerable Burden of Malaria III. Progress and Perspectives Am J Trop Med Hyg, December 1, 2007; 77(6_Suppl): vi - xi. [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
