The Prognostic Value of Measures of Acid/Base Balance in Pediatric Falciparum Malaria, Compared with Other Clinical and Laboratory Parameters
Charles R. J. C. Newton,1,4 Clarissa Valim,6,8 Sanjeev Krishna,2,3 David Wypij,6,7,8,9 Christopher Olola,4 Tsiri Agbenyega,5 and Terrie E. Taylor,10 for the Severe Malaria in African Children Networka
1 3

Institute of Child Health and 2Department of Cellular and Molecular Medicine, St. George’s Hospital Medical School, London, United Kingdom; Medical Research Unit, Albert Schweitzer Hospital, Lambarene, Gabon; 4Centre for Geographic Medicine (Coast), Kenya Medical Research ´ ´ Institute, Kilifi, Kenya; 5Department of Physiology, School of Medical Sciences, University of Science and Technology, Kumasi, Ghana; 6Clinical Research Program and 7Department of Cardiology, Children’s Hospital Boston, 8Department of Biostatistics, Harvard School of Public Health, and 9Department of Pediatrics, Harvard Medical School, Boston, Massachusetts; and 10Department of Internal Medicine, College of Osteopathic Medicine, Michigan State University, East Lansing

Background. Identifying severe, life-threatening falciparum malaria in African children allows for the prompt institution of appropriate management. In the past 2 decades, hyperlactatemia and acidosis have been identified as being associated with mortality in patients with severe malaria, but measurement of blood lactate concentration and base excess is expensive and technically demanding. In this large, prospective study, we examined the prognostic value of acidosis and hyperlactatemia and compared these markers to clinically assessed variables. Methods. We examined several clinical and laboratory measurements as prognostic markers of mortality in 14,605 parasitemic children admitted to 3 hospitals in Africa. Whole-blood lactate concentration and acid/base status were used to identify subjects who had hyperlactatemia and acidosis. Results. Using cut-points established by sensitivity and specificity curves, the sensitivities and positive predictive values for both lactate concentration and base excess were low, the specificities were moderate, and the negative predictive values were high (197%). No reliable clinical surrogates for hyperlactatemia or acidosis were identified. Addition of lactate concentration and base excess to predictive models with previously identified clinical features (Blantyre Coma Score, deep breathing, prostration, and weight-for-age Z score) and 1 laboratory measure (blood glucose level) did not appreciably improve models to predict mortality. Conclusions. Measurements of lactate concentration and acid/base balance are expensive to perform, and performance of the latter can be problematic. Severe falciparum malaria may be readily recognized in children at admission to hospitals in sub-Saharan Africa with use of simple, inexpensive means and does not require knowledge of lactate concentration and base excess. Falciparum malaria is responsible for 11 million deaths per year, with 170% of the burden of infections, morbidity, and mortality occurring in children in subSaharan Africa [1]. Most of these children are cared for in health facilities that are underresourced, particularly in terms of laboratory investigations. However, whether laboratory investigations (beyond simple clinical observations) have significant prognostic value for identification of children with severe malaria is unclear. In 1986, the World Health Organization produced criteria for the definition of severe malaria that included laboratory investigations, such as determination of parasitemia, glucose level, and base excess (BE) value and assessment of renal function [2]. These criteria were based mainly on studies that involved semi-immune adults. The criteria were revised in 2000 [3], incorporating data from studies of African children that identified acidosis [4] and hyperlactatemia [5] as factors strongly associated with mortality. Although recent studies have confirmed that lactate concentration and acid/base status [6–11] are valuable prognostic indicators, health care providers have been slow to incor-

Received 19 March 2005; accepted 24 May 2005; electronically published 23 August 2005. a Members of the Severe Malaria in African Children Network are listed at the end of the text. Reprints or correspondence: Prof. Charles R. J. C. Newton, P.O. Box 230, Kilifi, Kenya ( Clinical Infectious Diseases 2005; 41:948–57 2005 by the Infectious Diseases Society of America. All rights reserved. 1058-4838/2005/4107-0004$15.00

948 • CID 2005:41 (1 October) • Newton et al.

with lower values for the subjects who died (figure 1A). there were statistically significant differences across sites. and determination of acid/base status can be problematic. This study was conducted through the Severe Malaria in African Children (SMAC) research network. Statistical analyses were performed using SAS software. Respiratory pattern was defined as deep “Kussmaul-type” breathing [13] or as an irregular pattern of breathing. Optimal cut-points for BE value and lactate concentration for predicting mortality were chosen as the points at which the sensitivity and specificity curves crossed. Instruments Ltd. Linear regression methods were used to assess the effects of multiple predictor variables. Clinical examination and laboratory tests. or absconded) were recorded for each patient. largely because measurement of these values with conventional equipment is relatively expensive. Outcomes (patient died. to predict malaria-related deaths was based on logistic regression methods. we added BE value and lactate concentration to the selected models. we estimated Spearman rank correlation coefficients for continuous predictors and mean levels for binary predictors (with P values for comparisons based on Student’s t test results). version 9 (SAS Institute). equivalent to the area under the receiver operating characteristic curve) by 0. hyperlactatemia was more common than it was elsewhere. All P values are 2-tailed.. Ghana. METHODS Subjects. temperature at the time of hospital admission. in part. Subjects were enrolled at the time of admission to the following hospitals: Queen Elizabeth Central Hospital (Blantyre. comparing the model with clinical predictors. subjects from Kumasi were the most ill. We examined established clinical and laboratory prognostic markers of mortality for 3 sites in Africa. Kilifi District Hospital (Kilifi. Blantyre Coma Score [12]. To assess the predictive ability of BE value and lactate concentration. To identify clinical markers of BE value and lactate concentration. Children (from birth to the age of 180 months) who had asexual Plasmodium falciparum parasites detected in peripheral blood samples at hospital admission were eligible for the study. As a result.E. Clinical details were recorded before the results of the laboratory investigations were available. in predicting mortality. and those from Blantyre were the least ill. and Komofo-Anokye Teaching Hospital (Kumasi. all analyses were site-specific. January 2001 through December 2003). in conjunction with other clinical and laboratory predictors. In particular. and spleen size. The distribution of BE values varied across sites. Because of heterogeneity across sites.05 for all sites or if they increased the c statistic (a measurement of the predictive ability or concordance of the fitted model probabilities to the observed outcomes. we examined the changes in the c statistic. respiratory rate. survived. Finally. Mortality rates were lowest in Blantyre and highest in Kumasi. prostration (inability to sit). After obtaining informed consent. Comparisons of proportions were based on Pearson x2 tests. Venous blood samples were collected at the time of admission to the hospital to determine parasite count (by thick and thin film stained with 10% Giemsa). unpublished data). Reliable means to measure lactate concentrations were available only in Blantyre (Arkay LactateProLT-1710.porate these measures into standard practice. Malawi. including blood glucose concentration. by 2300 Statplus analyzer [YSI Corporation] Kumasi. Tests for comparison of continuous variables across sites were based on analysis of variance and Kruskal-Wallis tests. and blood glucose concentration (table 1). temperature. June 2001 through December 2003). et al.605 subjects enrolled at the 3 sites were comparable with respect to demographic characteristics. subjects from Kilifi tended to have lower BE values than other subjects. and by Analox GL5 [Analox Instruments] in Kilifi). consistent with the trends in severity. we examined whether the expensive and difficult laboratory investigations—estimation of blood lactate concentration and BE value—added significantly to the prognostic value of a combination of simple clinical observations and readily available and affordable laboratory investigations. demographic variables. and vomiting. to maximize the adjusted R2 value.5 for Windows (SPSS). seizures. RESULTS The 14. and blood gas levels (by StatProfile pHOx Blood Gas/Oximeter [Nova Biomedical] in Blantyre and Kumasi. Using a cut-point BE value of less than or equal to 8 [7].) and Kumasi (2300 Stat plus analyzer) only. Model selection was performed with the all-possible subsets regression procedure. patient history and clinical examination findings were recorded. By most measures. and by IL 1620 [Instrument Laboratories] in Kilifi). version 11. Kenya. Identifying cut-points for base excess value and lactate concentration. hematocrit. hematocrit (by microhematocrit centrifuge in Blantyre. and standard laboratory investigations only with models in which BE value and/ or lactate concentration were added. and by Sysmex KX-21N [Sysmex Corporation] in Kumasi) or hemoglobin level (by Coulter MDII [Coulter Electronics] in Kilifi). Statistical methods. of the large sample sizes. and in Blantyre. April 2001 through December 2003).T. Acidosis in Severe Falciparum Malaria • CID 2005:41 (1 October) • 949 . glucose level (by ACCU-CHEK [Roche Diagnostics] in Blantyre. and SPSS software. Clinical predictors or demographic variables were retained in the model if the P values were !. including history of fever. We added readily available and affordable laboratory test measures to the model.01 for at least 2 sites. Details about these sites are described elsewhere (T. Assessment of the ability of BE value and lactate concentration.

6 29 Characteristic or finding.6 1.7 7.2 4.4 0. and laboratory test characteristics of and outcomes for African children with malaria.2 1. interquartile range. and 67% and 76% in Kumasi.000 parasites 103/mL. with specificities of 84%. With use of a cut-point BE value of less than or equal to 12 [14]. Defined as a hematocrit of !15% or a hemoglobin level of !5 g/dL.0 6.5 3 1 4.001.0 1.7 28 5.3 25 5.5 1. and analysis of variance was used to compare other continuous variables.Table 1.4 38.6 2.9 …b 8 5.1 in Blantyre. the sensitivities were 21% in Blantyre. P values for each comparison were ! . unless indicated otherwise.0 28 Kumasi.9 7. and 7.7 5.5 2 1. mean C SD Spleen size. by site.4 2.5 1. they were 60% and 98% in Blantyre.4 5 7. Malawi (n p 4362) 36 54 SD 12. .9 43 (3–214) 8 24.3 in Kilifi.1 in Kumasi (figure 2A).6 e 75 (17–265) 12 20. Lactate concentration. 9. Data are percentage of patients. 1500. IQR. Demographic. for a BE value of less than or equal to 12.4 5 8. 78% and 57% in Kilifi. a b c d e f Parasite load.7 10.7 28 5 9 1.4 4.8 18 42 5 2 4 2 4 38.6 32 Kilifi. mean cm SD Laboratory test results at admission Parasite count.3 0. the Kruskal-Wallis test was used to compare parasitemia levels.8 1.5 1.7 44 56 16 19 6 47 13 25 1. mean mmol/L Hyperlactatemiaf Mortality 81 (33–177) 4 26. Hemoglobin level was not measured in Blantyre. Pearson x2 tests were used to compare proportions. 57% in Kilifi.4 18 31 11 14 3 26 10 20 1. mean kg SD Weight-for-age Z score. clinical.10). median parasites 103/mL (IQR) a Hyperparasitemia Hematocrit.9 SD 4.5 3.7 3.8 2. With use of the cut-point for 950 • CID 2005:41 (1 October) • Newton et al. 7% and 99% in Kilifi. The respective positive and negative predictive values for a BE value of less than or equal to 8 were 30% and 98% in Blantyre.8 12 5.8 7. mean months SD Male sex Weight. mean g/dL SD Anemiac Glucose. and 45% in Kumasi. except for sex (P p . Ghana (n p 4969) 32 55 5. and 13% and 98% in Kumasi. the respective sensitivities and specificities were 36% and 98% in Blantyre.7 11. To compare sites. 12% and 98% in Kilifi. Blantyre.5 2. Glucose level.2 1. The distribution of lactate concentrations was similar in Blantyre and Kumasi (figure 1B). The optimal cut-points to predict mortality were 0. 15 mmol/L.1 44 17 …e … 3.0 NOTE.2 1.9 30 2.7 37.2 mmol/L. mean mmol/L SD Hypoglycemiad Base excess value Mean SD Less than or equal to 8 Less than or equal to 12 Lactate concentration. and 22% and 97% in Kumasi. suggesting that optimal BE cut-point selection can be heterogeneous across sites.5 26 10 4. Lactate concentration was not measured in Kilifi.4 2. mean % SD Hemoglobin.9 5. 2. by class Demographic Age.21) and glucose level (P p . Kenya (n p 5274) 34 53 4. mean Medical history Seizures Vomiting Examination findings at admission Deep breathing Indrawing Irregular breathing Unable to sit Blantyre coma score Blantyre coma score Temperature.1 2.

and hematocrit in Blantyre.Figure 1. Surrogate clinical markers for BE value and lactate concentration. whiskers extend to the most extreme observation within 1. BE value was significantly correlated (P ! .1 mmol/L in Blantyre and 4.1 mmol/ L in Kumasi (figure 2B). and temperature in Kilifi.5 interquartile range units of the 25th and 75th percentiles. However. lactate concentration of 15 mmol/L [5. except for vomiting in Blantyre. the lower boundary of the box represents the 25th percentile. the Spearman Acidosis in Severe Falciparum Malaria • CID 2005:41 (1 October) • 951 . More extreme values are plotted with small circles. The optimal cut-points to predict mortality were 5. except for hemoglobin level.001) for all binary risk factors. spleen size.001) with continuous clinical and laboratory test predictors. according to outcome status and by site. The solid bar within each box represents median value. Similarly. 16]. Box plots of base excess value (A) and lactate concentration (B). the upper boundary of the box represents the 75th percentile. mean BE values were significantly lower (P ! . At all sites. 15. hematocrit. The positive predictive values for a lactate concentration of 15 mmol/L were low (6% in Blantyre and 12% in Kumasi). although the negative predictive values were extremely high (99% in Blantyre and 97% in Kumasi). and temperature in Kumasi. the sensitivities and specificities were moderate in magnitude: 72% and 71% in Blantyre and 60% and 77% in Kumasi. weight-for-age Z score. respectively. seizures in Kilifi. and vomiting and seizures in Kumasi (table 2).

22 in Blantyre and to 0. Blantyre. rank correlation coefficients between BE and continuous predictors were only moderate in magnitude.22 in Blantyre.15) and Kilifi (r p 0. Kenya. demographic. suggesting that these variables (alone or in combination) cannot serve as good surrogate measures for BE value. the maximum adjusted R2 increased to 0. and laboratory candidate predictors. Kilifi. When lactate concentration was included as a covariate in the model selection. the strongest correlations with BE value were positive associations with Blantyre Coma Score (r p 0. by site. Ghana. Sensitivity curves (solid lines) and specificity curves (dashed lines) for prediction of mortality as a function of base excess value (A) and lactate concentration (B). KU.Figure 2. basis of clinical.20 in Kilifi.40 in Kumasi.19) and Kumasi (r p 0. B.19. at most.26.16). and Kumasi were 0. In multiple linear regression models to predict BE on the 952 • CID 2005:41 (1 October) • Newton et al. At all sites. The factors that . Initials for the sites are positioned at the quartiles of the site-specific distributions of BE value and lactate concentration. and r p 0. respectively.42. ∼40% of the variability in BE value was accounted for by the other covariates. and 0. Kumasi. 0.22 in Kumasi) and with glucose level in Kilifi (r p 0. the maximum adjusted R2 values attained in Blantyre. Therefore. KI. with inverse associations between BE value and parasitemia in Blantyre (r p 0. r p 0. Kilifi. Malawi.21).

3 2. seizures in Kilifi (P 1 .7 0.3 5.6 4.3 4.2 mmol/L).2 8.3 3.6 5.8 0.8 1. Ghana 5.3 4.7 4.8 3.8 5.4 8.5 5.4 5.7 5.0 5.9 4. by site.7 4.8 3.8 2.8 4.4 1.2 3.0 4.0 10.5 4.8 2.8 5.8 4.Table 2.0 3. irregular breathing.1 13.7 5.6 5.7 0.1 9.001.1 2.4 5.0 4.9 4.5 5.3 7.3 5.1 3.1 4.5 6.6 5.6 5.10). We assessed the relationships between acidosis (BE value.9 3.4 3.4 3.1 5.3 8. suggesting Acidosis in Severe Falciparum Malaria • CID 2005:41 (1 October) • 953 .3 3.3 4.0 8.4 5.9 4.2 4.1 4.4 5.0 7. less than or equal to 8) and deep breathing. Malawi 4.9 7.8 5.3 4.10 for each). except for vomiting in Blantyre (P p .2 6.3 3. 81% in Kilifi.9 … … 3.6 7.0 4.2 4.6 … … … … … … 10.6 6.7 5.4 4.8 3.1 7.8 3.5 Kumasi.9 4.5 2.4 6.8 5.8 4.6 5.6 1.7 6. Mean value Base excess value Risk factor Vomiting Yes No Seizures Yes No Deep breathing Yes No Indrawing Yes No Irregular breathing Yes No Unable to sit Yes No Anemia Yes No Hyperparasitemia Yes No Hypoglycemia Yes No Hyperlactatemia Yes No Blantyre coma score 2 Yes No 4 Yes No BE value less than or equal to 8 Yes No Blantyre. were consistently selected in models that had the highest adjusted R2 at the 3 sites included deep breathing.6 4.8 3.1 4. hematocrit. Base excess value and lactate concentration across levels of risk factors in African children with malaria.1 5.1 10.003).5 4. age.4 1.5 6.1 1. and glucose level.7 7.5 7.0 8.9 4.8 4.1 5.9 4.3 7. !2.6 5.5 10.5 4.6 4.02) and seizures in Kumasi (P 1 .4 6.7 0.9 9.5 7.2 5.3 4.3 3.0 4.8 4.3 5.7 5.9 7.9 6.5 7.8 … … 5.6 5. Malawi 1.5 4.6 4. and 89% in Kumasi) were high at all sites.6 1.4 9.6 6.0 SD Lactate concentration.8 4. except for vomiting in Blantyre (P p .7 4. Kenya 9.8 10.5 4.2 3. mmol/L Blantyre.1 7. P values for each BE comparison were !.3 3.4 6.1 4.8 5.6 2.2 7.4 3. and 68% in Kumasi) or irregular breathing (71% in Blantyre.4 7.4 4.8 6.8 Kumasi.0 0.1 4. parasitemia.7 2.4 8.8 6. P values for each lactate comparison were !.5 4.6 … … 4.5 4. The proportions of acidotic subjects who did not have deep breathing (63% in Blantyre.4 5.5 5.3 3. temperature.6 3.8 NOTE.1 Kilifi.5 5.3 6.0 3.1 3.9 4.9 4.5 2.7 1.1 3.8 4.4 3.3 2. Ghana 4.3 11.6 6.0 4.2 1.6 5.5 4. and vomiting and seizures in Kumasi (P 1 . t Tests were used to compare base excess (BE) values and lactate concentrations across levels of risk factors.8 8.0 1.2 1.6 4.1 3.3 4.6 2.9 3.5 4.3 4.5 4.5 4.5 4.9 4.9 4.3 3.7 8.1 7.6 5.6 9.7 4.7 5.10).4 4.9 10.0 5.6 5.0 7.0 7.0 4.9 5.1 7.0 4.3 4.4 8.2 6. hyperlactatemia.3 5.3 4.8 2.6 5.8 6.1 4.1 4.0 6.1 3. by site.4 4.8 12.1 3. and hypoglycemia (blood glucose level.7 1.5 4. 96% in Kilifi.9 13.001.7 3.3 4.2 5.2 3.2 5.

covariate in the model selection.03) 3. 1. seizures. c p 0.05–5. Lactate concentration was significantly correlated with continuous clinical and laboratory test predictors (P ! .42 in Kumasi.24 in Kumasi (95% CI. c p 0.07–1.71–1. irregular breathing.75 (0.76) and 1.15) 0.22 in Kumasi (95% CI. Hyperlactatemia and acidosis were dissociated in a moderate number of subjects in Blantyre and Kumasi.14 (1.63) 2.78). The crude ORs per unit decrease in BE value were 1. Lower BE value and higher lactate concentration were associated with mortality.74).64 (1. c p 0. 156 deaths in Kilifi.07) 1.85) 2. and 2% in Kumasi) were relatively low. The P values for all BE value and lactate concentration ORs were !. P p . a Glucose level. Mean lactate concentrations were typically 1.17) 2. The absence of hyperlactatemia misclassified as false negatives 33% of acidotic subjects in Blantyre and 44% in Kumasi.21) 0. 1. Kilifi. and P p .18–1. the presence of hyperlactatemia misclassified as false positives 29% of nonacidotic subjects in Blantyre and 15% in Kumasi.17 in Blantyre and 0. or hypoglycemia (1% in Blantyre. and indrawing in Kilifi. and 0.29) 1.93 (1.84 (0.83 (1.10 (1. and 4% in Kumasi). and irregular breathing and seizures in Kumasi. 2% in Kilifi.81). Predictors of mortality via multiple logistic regression in African children with malaria. In Blantyre and Kumasi.14) 2. Therefore. inability to sit. and 11% in Kumasi). 1% in Kilifi.84.78 (1.09–1. Prognostic models for mortality. BE value and lactate concentration had the highest predictive abilities (i. and weight-for-age Z score.48 (1.26 (1.92) … 1.17 in Blantyre and r p 0.06 for deep breathing in Kilifi.08–1.01 at all sites: deep breathing. 1.99) 2. tests other than those for BE value and . There were a total of 59 deaths in Blantyre. Also.76–8.90 (0.81–0. and Kumasi were 0. that not all subjects with acidosis present with either deep or irregular breathing.53 in Kumasi. c p 0.36) 1.18–4. except for weight-for-age Z scores and temperature in Blantyre and for weight-for-age Z scores and glucose level in Kumasi.15 in Blantyre and 0.18–1.11 (1.25. The crude ORs per unit increase in lactate concentration were 1.29 in Blantyre (95% CI.13 (1.76).85.27) 3. Other predictor variables that were statistically significantly associated with mortality included vomiting in Blantyre. P p .20–1.14–1.23–1. when present. the maximum adjusted R2 values were 0.31 for lactate concentration in Kumasi. 4% in Kilifi.05 (0.25–2. 1.2 mmol/L.14 for hypoglycemia. 0.41 in Kumasi).82) 1.e. Characteristic Weight-for-age Z score (per unit increase) Deep breathing Blantyre coma score Unable to sit Hypoglycemiaa 2 Blantyre.02 (0..12) 1.87 Kumasi.25.22 in Kilifi (95% CI.56) 1. and both variables were approximately linearly associated with the log odds of the probability of death.20 in Blantyre and r p 0. Ghana (n p 3730) 0. Blantyre Coma Score of 2.17) 0.17–2.68–4. Kenya (n p 4670) 0.Table 3. except for vomiting in Blantyre and seizures in Kumasi.98–1. When BE value was allowed as a 954 • CID 2005:41 (1 October) • Newton et al. irregular breathing (4% in Blantyre.99) 2.73–7.14 for lactate concentration in Blantyre. and 183 deaths in Kumasi. nor did they appreciably confound the other covariates in the model.57–3. P values were !.e.88 Kilifi.05 whenever an OR of 1. In multiple linear regression models to predict lactate concentration that were based on all candidate predictors.15 (2.21 in Blantyre and r p 0. 1. 1. c statistics) among all univariate models considered.80.28.01) 3. and P p . unless otherwise indicated. because the optimal cut-points for BE value differed across sites.25. we added affordable laboratory tests to the logistic regression model (i.17 (0.001.65 (1. with the largest correlations with Blantyre Coma Score (r p 0. compared with subjects who did not have such risk factors (table 2)..01.19 in Blantyre (95% CI.99–1.0 was excluded from the 95% CI. In the multiple logistic regression analysis of predictors of mortality with candidate clinical and demographic variables. c p 0.53 (0. The proportions of subjects without acidosis but with deep breathing (3% in Blantyre.34–3. Spearman rank correlations were only moderate in magnitude. these values increased to 0.06 for weight-for-age Z score. We chose to assess the predictive ability of acidosis and hyperlactatemia using BE value and lactate concentration as continuous covariates.001).77–6.5–2-fold higher in subjects who had a risk factor for severe disease. by site.001) for all binary risk factors. the addition of these variables to the model did not increase the c statistic by 10.12) 1. these markers are highly specific and rarely misclassify subjects as having false-positive results. However.25 in Kumasi) and hematocrit (r p 0.99–2. and 1.34 in Kumasi).68–0.83 Lactate (per mmol/l increase) Base excess (per unit decrease) c Statistic NOTE. Malawi (n p 2159) 0. 2. respectively. The c statistics for models including each of these variables for Blantyre. P p . the following covariates attained P values of .57–3. Data are OR (95% CI).29 (1. Next. BE value and lactate concentration had a weak-to-moderate inverse association (r p 0.34. mean lactate values were significantly higher (P ! .

which increased the c statistic to 0. When BE value was included with lactate concentration in multivariable models.03 in Blantyre. appeared to predict death accurately. or hypoglycemia at admission captured 84% of 432 deaths. and 0. Addition of acidosis (BE value. and 0. Among subjects who died and who had hyperlactatemia. respectively. the P value for the coefficient of lactate was 1. Malawi (A). and lactate concentration (dashed lines) for Blantyre. When added separately to this multivariable model. Kenya (B).81 in Blantyre. and 0. These tests did not appreciably add to Figure 3. base excess value (BE). inability to sit. Ghana (C).001) at each site. In all sites. The c statistic of this model was high (0. respectively). Kilifi.87. the addition of BE value and lactate concentration only slightly improved the predictive ability of the model that included the clinical predictors and hypoglycemia (c statistic increases of 0. compared with use of additional information on BE value and lactate concentration. deep breathing. The receiver operating characteristic curves of these modelbased criteria. with and without BE value and lactate concentration.01. 0. less than or equal to 8) to these criteria increased this percentage by 6%. However. 92% satisfied at least 1 of the clinical or hypoglycemic criteria. acidosis and hyperlactatemia were captured by at least 1 of the clinical or hypoglycemic criteria in most subjects who died. to 90% (389 of 432 deaths). 91% met at least 1 of the clinical or hypoglycemic criteria. Table 3 presents the ORs for the logistic regression models including both BE value and lactate concentration. and Kumasi. 15 mmol/L) instead of acidosis would have increased this percentage in Blantyre and Kumasi by 5%.1 for each site. hypoglycemia. neither of these laboratory tests on venous blood.86. and the cut-points for BE value varied with site.01 in Kilifi. Kilifi.1 or increased the c statistic by more than 0. We investigated the number of deaths that would have been misclassified as being of low risk at hospital admission on the basis of simplified clinical criteria. The adjusted associations of individual predictors and mortality tended to be stronger in Blantyre. both BE value and lactate concentration reached statistical significance (P ! .01 ) was hypoglycemia. and Kumasi.83 in Blantyre. Neither hyperparasitemia nor severe anemia achieved P values !. DISCUSSION Both acidosis and hyperlactatemia were significantly associated with mortality in children with severe falciparum malaria admitted to these hospitals in sub-Saharan Africa. Although BE value remained a statistically significant predictor of mortality at all sites.88. the only statistically significant laboratory test (P ! . 0.lactate concentration). Therefore. Acidosis in Severe Falciparum Malaria • CID 2005:41 (1 October) • 955 . used alone. 0. Among subjects who died and who had acidosis. Clinical implications. from 87% (223 of 257 deaths) to 92% (236 of the 257 deaths). Kilifi. Receiver operating characteristic curves and c statistics of the predicted probability of mortality from models including only clinical predictors and hypoglycemia (solid lines) and models including clinical predictors.85. Classification of a child as being at high risk if he or she had coma. Addition of hyperlactatemia (lactate concentration. and Kumasi.02 in Kumasi). were comparable (figure 3).

Kathryn Maitland. given the standardized methods employed. interobserver variability in detecting these clinical signs may reduce their sensitivity and specificity for acidosis or hyperlactatemia Thus. Janet Fullah. Alfred Njobvu. 956 • CID 2005:41 (1 October) • Newton et al. model-based decision rule that includes clinical features and blood glucose level. Pamela Esangbedo.. However. the clinicians occasionally knew the results of some laboratory investigations before assessing the clinical findings. can be used to identify children with acidosis or hyperlactatemia reliably. Ghana). Krishna S. Dina Kayaya. Acidosis and hyperlactatemia are likely to have a number of different causes [17]. but this does not appear to have improved the relationship between the clinical features and the results of these tests. The optimal cut-points for venous BE value varied across sites. The measurements of blood gas and lactate concentrations. and T. and the University of Science and Technology. and Augustine Ebonyi (Banjul. Pierre Blaise Matsiegui. differences related to measurements of these parameters are possible but are less likely.. either alone or in combination.N. Severe falciparum malaria. blood glucose. Molyneux ME. which prompts immediate therapy. because identical Lactate Pro machines were used in Blantyre and Kumasi. Daniel Ansong. Arnaud Dzeing. to C. All authors: no conflicts. Although deep breathing was reported to be specific and sensitive for detection of acidosis in a small study of well-trained clinicians [13]. Kalifa Bojang. Clinical features and outcome of severe malaria in Gambian children. however. and hypoglycemia were associated with hyperlactatemia and acidosis across all 3 sites.V. Saadou Issifou. . Bertrand Lell. Cynthia Donkor. so these results should be extrapolated with caution to other settings in which the staff have had less training. 5. Findings from this large prospective comparison suggest that these measures are not essential laboratory investigations for the management of severe falciparum malaria in sub-Saharan African hospitals. Betty Wamola. may be useful in identifying children for trials of interventions targeted specifically against acidosis and/or hyperlactatemia. Tim Planche. 4. Maryvonne Kombila. Institute of Allergy and Infectious Diseases (AI45955.). 3. The Gambia). but a cut-point of 5 mmol/L for hyperlactatemia was nearly optimal at the 2 sites that assessed this parameter.W. The differences between sites could reflect differences in the types of patients admitted to each hospital or different methods of processing the sample. Acknowledgments Financial support. we did not find that it was a sensitive marker of acidosis in this larger study. less expensive. Justice Sylverkyn. Kingsley Osei-Kwakye. Q J Med 1993. and the appropriate quality controls were used for all 3 blood gas machines. C. David Ameh.J. Steffen Borrmann. Trans R Soc Trop Med Hyg 2000. The variability may represent differences in the patient populations at each hospital. 80(Suppl):3–50. none of these features. Malawi).. although the measurement of levels of blood gases and lactate help identify children admitted to African hospitals who are likely to die of severe falciparum malaria. The high negative predictive values for these tests are unlikely to influence clinical management.C. Taylor TE. St. which was more representative of practice in Africa. the Wellcome Trust. The predictive ability of some clinical features varied across sites. Mbort Atan Ayibo. 86:99–109. Jalli Mori. World Health Organization. World Health Organization Malaria Action Programme. Sadik Mithwani. and Norbert Peshu (Kilifi. Lloyd Bwanaisa. Potential conflicts of interest. and their clinical manifestations are probably influenced by other pathophysiological perturbations in falciparum malaria. will effectively predict malaria deaths at any cut-point.T. Emmanuel Asafo-Adjei. Rachael Odhiambo. Borgstein A. Efforts to standardize observations within the SMAC network were made before the start of the study. Timothy Mnalemba. Acid-base status in paediatric Plasmodium falciparum malaria. 94(Suppl):S1–90. Ismaela Abubakar. George’s Hospital Medical School. D.A. Geneva: WHO. 2001. taken alone. World Health Organization (WHO). Abdou Bah. The world health report. Crawley J. Victor Degenu.). Kenya). our analyses suggest that neither of these tests. London. There was variability in measurements of acid/base balance. it was specific in Blantyre and Kilifi.). Collins Qongwane. US National Institutes of Health. Brett Lowe. Furthermore. Trans R Soc Trop Med Hyg 1986. and Sophie Kazembe (Blantyre. Emmanuel Onyekwelu. United Kingdom (050533. coma. 2.O. and Esther Esumming (Kumasi. Frankie Mbadinga. Kumasi. Evelyn Anane-Sarpong. Severe and complicated malaria. However. References 1. Maganizo Chagomerana. Ghana (to T. Although deep and/or irregular breathing. these parameters add little to the prognostic significance of a simpler. James Mwenechanya. and Nestor Obiang (Lambarene. Emmanuel Owusu-Ansah. 21:577–87. Gabon). Stanley Usen. and Peter Kremsner. Both tests were strong univariate predictors of malaria mortality. et al. David Sambian.R. Joshua Ngala. which includes contributions from the following individuals: Dominic Kwaitkowski. Vida Asante. None of the standard clinical observations or laboratory investigations were reliable surrogate markers for acidosis or hyperlactatemia. The clinical predictors suggested that children admitted to Kumasi were more ill than were those at the other 2 sites. Waller D.E. Sampson Antwi. Alex Owusu-Ofori. Mariatou Jallow. Osei Yaw Akoto. Beatrice Mkondiwa. In Kumasi. United Kingdom (to S. MEMBERS OF THE SMAC NETWORK This article is published on behalf on the SMAC network.K. The latter is less likely.). Clin Infect Dis 1995.the predictive ability of a model based on clinical features and a relatively simple laboratory test. to C.

et al. Levin M. Planche T. et al. 7. Stacpoole PW. Malar J 2005. 37:890–7. Krishna S. Mwangi I. Pamba A. 22:221–45. 10. et al. Schwenk A. 1:e18. 43:386–96. falciparum malaria in Kenyan children: evidence for hypovolaemia. Amukoye E. 17. et al. Perturbations in electrolyte levels in Kenyan children with severe malaria complicated by acidosis.6. Population kinetics. et al. 4:1. 72:334–6. Planche T. Clin Infect Dis 2005. Nze Obiang PC. Borgstein A. Taylor TE. Bedu-Addo G. Newton CR. Newton CR. Severe falciparum malaria in Gabonese children: clinical and laboratory features. Marsh K. Assessment of volume depletion in children with malaria. Tchoua R. 71:441–59. and prognostic indicators in paediatric cerebral malaria: a study of 131 comatose Malawian children. Arch Dis Child 1995. Planche T. Murphy S. Maitland K. Lactic acidosis. Severe P. Am J Trop Med Hyg 1996. Waruiru C. Molyneux ME. Crawley J. 14. 40:9–16. Maitland K. Maitland K. Bedu-Addo G. Fegan G. et al. English M. 4:426–31. Endocrinol Metab Clin North Am 1993. Dzeing-Ella A. Q J Med 1989. Agbenyega T. A prospective comparison of malaria with other severe diseases in African children: prognosis and optimization of management. Pharmacol Ther 1998. QJM 2003. English M. Pamba A. Wirima JJ. 11. 15. Interobserver variation in respiratory signs of severe malaria. Levin M. 8. Agbenyega T. 79:1–53. Severe falciparum malaria in children: current understanding of pathophysiology and supportive treatment. Deep breathing in children with severe malaria: indicator of metabolic acidosis and poor outcome. PloS Med 2004. efficacy. 12. Response to volume resuscitation in children with severe malaria. Acidosis in Severe Falciparum Malaria • CID 2005:41 (1 October) • 957 . English M. Peshu N. J Clin Pharmacol 2003. 9. and safety of dichloroacetate for lactic acidosis due to severe malaria in children. 55:521–4. 16. 96:427–34. Onanga M. Pediatr Crit Care Med 2003. Clinical features 13. et al. Clin Infect Dis 2003.

Sign up to vote on this title
UsefulNot useful