Effect of a comprehensive clinical care program on disease course in severely ill children with sickle cell anemia in a sub-Saharan African setting

The clinical manifestations of sickle cell anemia or SS disease (SS) are notoriously variable.^1,2  The bases of this variability are poorly understood. Interplay of both genetic and environmental factors contribute to the variable clinical course of the disease. Alleviating factors include persistence of high levels of fetal hemoglobin through adulthood³  and coinheritance of α-thalassemia.⁴  There is a general belief that African patients exhibit a more severe phenotype compared with those from India or the eastern provinces of Saudi Arabia,^5,6  and the differences have been attributed to genetic factors.^7-9  However, economic and environmental conditions also differ in these parts of the world.

In developed countries, newborn screening for SS sustained by adequate medical comprehensive follow-up programs has been shown to significantly decrease the morbidity and mortality of SS.^10-13  In the sub-Saharan African countries, with the highest prevalence of SS, most of the affected children die in the early years of life.^14,15  In the Republic of Benin in West Africa, no reliable estimation of mortality rate in SS children is presently available, but more than 50% of the affected children seem not to reach their fifth birthday. Survivors suffer from repeated acute events and develop rapid progressive organ damage that reduces life expectancy. Cultural background, lack of (medical) education, and limited health care facilities are probably the major factors contributing to such high childhood mortality. In May 1993, we initiated a prospective study to evaluate the effect of an active and standardized comprehensive clinical care program on the course of the disease in children who presented with a severe clinical phenotype. The data presented here demonstrate that a relevant parental education program coupled with simple and cost-effective medical care can positively influence the clinical outcome of SS in an otherwise hostile African setting.

The study group consisted of SS children, referred to the Comprehensive Clinical Care Program (CCCP), and was conducted in the National Teaching Hospital at Cotonou, the largest city and the financial capital of the Republic of Benin. These children were referred by peripheral hospitals or the pediatric unit after the resolution of severe acute events. The hemoglobin phenotype of each patient was confirmed by citrate agar electrophoresis (Helena Laboratories, Beaumont, TX). A specially allotted sickle cell disease “(SCD) room” within the National Teaching Hospital was the focal point of CCCP for its clinical activities, which included follow-up/emergency visits, parental education, and management of acute events.

The study was an attempt to prospectively evaluate the effects of a comprehensive program of clinical care on the course of SS in young children. Given the severity of SS in Benin, it was considered unethical to withhold such care from a control group. As a result, we were forced to compare parents' reports of number and duration of painful crises, number of hospitalizations, and number of transfusions prior to joining the clinic, with those recorded after the initial clinic visit. Parental reports are not comparable with observations made by medical personnel, but as will be shown, results of such comparisons are consistent with observations of changes in children's growth during the study. The study was approved by the Ethical Committee of the Faculty of Health Sciences and was conducted between July 1, 1993, and December 31, 1999.

Taking into account the social, cultural, and economic backgrounds of the population, we focused our particular attention on the following: (1) provision of information and education about SS to the parents and to the affected older children at each medical visit, (2) education to improve the nutritional status of the child, (3) malarial prophylaxis and specific vaccination, and (4) encouragment of parents to keep appointments for scheduled medical visits.

The treatment team, devoted full-time to the CCCP, was composed of a pediatric hematologist (a civil servant, who has spent previously 10 years in Europe for training and practice) and 4 nurses specially trained—essentially on site by the pediatric hematologist. On entry into the study, an exhaustive history was taken for each patient and physical examination was performed. Height and weight were measured periodically. In the absence of a valid growth development chart applicable to children of West Africa, the data recorded were plotted on French growth charts to appreciate changes that occurred during the study period.

Detailed information about SS in a perceivable language (French and other local languages with which parents were familiar) was given to the parents with emphasis on factors that may precipitate acute events. The training included spleen palpation and proper use of a thermometer. The importance of drinking adequate amounts of water to keep urine volume high and ways to improve a locally adaptable clean and healthy lifestyle were explained. Emphasis was also placed on the advantages of regular medical follow-up and the necessity of a prompt presentation to the SCD room at the onset of any acute event, fever higher than 38.5°C, sudden spleen enlargement, or any other unusual symptoms. When the acute event required hospitalization, the child was admitted in the pediatric ward inside the hospital, and the CCCP team was responsible for the inpatient management.

After a thorough discussion with the parents about their current nutritional trends, ways of improving nutritional status were suggested, in particular by drawing the parents' attention to affordable local products that have a high nutritional value but are often ignored. For example, in Cotonou, exclusive use of corn flour for cooking is the common practice, whereas other cereals with complementary nutritional value are readily available at an equivalent price. Use of mixed cereals to increase the nutritional value was stressed and for each family, a weekly nutritional program was formulated.

The first (inclusion) visit lasted approximately 2 hours. During the initial phase of the study, children were brought by their mother, but later on, we succeeded in persuading fathers to participate during the visits. Patients were evaluated at monthly intervals during the first 6 months of follow-up and then every 3 months. The cost of each medical visit was limited to a token payment (ie, the tenth of the actual cost).

Basic treatment consisted of antipneumococcal and antimalarial prophylaxis (oral penicillin 20 000 IU/kg twice a day, chloroquine 5 mg/kg every 2 days) and a dietary supplement of folic acid (5 mg/d). Regular use of an insecticide-impregnated bed-net was advised to reduce the risk of mosquito bites. In addition to the 6 vaccines recommended by WHO, hepatitis B, Haemophilus influenzae B, antipneumococcal, and anti–Salmonella typhi vaccines were strongly recommended and offered at a low cost (half of actual cost).

Basic laboratory studies included a complete blood cell count, including reticulocyte count, and blood smears for Plasmodium falciparum asexual parasites at each follow-up visit. As these patients do also form part of a second study program aimed to explore the influence of genetic and environmental interactions in modulating the clinical expression of SCD in an African setting, genetic data including the β-globin haplotypes (Drs J. Elion and R. Krishnamoorthy, INSERM U458, Hôpital Robert Debré, Paris) and α-globin gene status (Dr T. Vulliamy, Department of Hematology, Imperial College School of Medicine, London) for these patients were available including the status of G6PD A^- 202G>A mutation and level of fetal hemoglobin (measured by high-performance liquid chromatography [HPLC] in Paris).

Clinical data were collected using a specific form. Since patients had not been regularly followed prior to this study, it was difficult to obtain a precise medical history prior to their joining the clinic. Using concordant testimony of both parents (when both were available), we estimated frequency and duration of recorded attacks of severe pain and frequency of hospitalizations and transfusions, prior to clinic entry.

Specific data recorded included growth parameters, vaso-occlusive painful crises and other SCD-related events (acute chest syndrome, stroke, anemia, bacterial infections), malarial crises, number of hospitalizations, and days spent at the hospital. The major clinical parameters assessed were (1) the frequency and severity of SCD-related events (Table 1), (2) changes in the slope of the height/weight curves, (3) the frequency of P falciparum malarial attacks, (4) the causes of worsening anemia requiring transfusion, and (5) the causes of death.

Of the parents, 15% declined the proposed medical follow-up program. Table 2 summarizes the characteristics of patients enrolled in this study. A total of 236 children (age range, 8 months to 12 years), most of them resident in Cotonou and nearby suburbs, were followed for 983 patient-years; the length of follow-up varied from 18 to 78 months. Among them, 152 (64.4%) were younger than 5 years at entry into the study, and the median age of patients at entry was 35 months. Fetal hemoglobin was deemed low if it was assumed that most of the patients had yet to reach their steady-state level (mean age at the time of assessment, 82.2 ± 44.6 months). Among the 236 chromosomes (50% of the study group) analyzed for the β-globin gene cluster haplotype, the frequency of the Benin haplotype was 94.9%, and 91.5% were homozygous for this haplotype. During the study period, all enrolled patients were completely managed by the CCCP team. Of the patients, 74% completed the vaccination program.

Table 3 lists the acute events that were recorded during the study period. Of the patients, 19% were transfused and 23% were hospitalized. The mean duration of hospitalization was 3.2 ± 1.2 days. Children presenting with “chest syndrome” had significantly elevated neutrophil counts (compared with baseline values) that returned to prior values after antibiotic treatment, and were considered to be infected with bacterial pneumonia (a very frequent infection in young children in Cotonou). None of them had chest pain, respiratory distress, or a significant drop in hemoglobin level; they were managed as outpatients according to our protocol for outpatient management of fever,¹⁶  with rapid resolution following administration of antibiotics alone.

All episodes of acute splenic sequestration were recorded in the first year of follow-up, leading to death in 4 cases. One patient presented with Streptococcus pneumoniae meningitis and died; his parents did not comply with oral penicillin prophylaxis and he had not received antipneumococcal vaccine. During the study period, 183 patients (77.5%) experienced vaso-occlusive painful crises with a mean of 3.3 episodes per patient. These crises were mild in all cases, were treated essentially with aspirin and acetaminophen, and did not require narcotics or hospitalization.

Except for the stroke, the number of each kind of SS-related acute events recorded within the first year was higher than the number recorded during the following years of the study.

Effect of our program on growth and weight was ascertained from the evolution of the slope of the height/weight curves. Among the 236 patients followed during the study period, 184 (78%, group 1: 94 males and 90 females) showed a remarkable recovery phase with the slope of the height/weight curves shifted upward, indicating an increased velocity of their linear physical growth. In contrast, there was no evidence of improvement in the remaining 52 patients (22%, group 2) with slopes of the height/weight curves either unchanged or shifted downwards. In group 1, the mean absolute gain in weight during the first year of follow-up was 3430 ± 1140 g (range, 2000-8800 g); the mean absolute gain in height was 9.9 ± 4.3 cm (range, 1-25 cm). This improvement was sustained over the study period irrespective of the age of the child at inclusion. In this group of patients the hemoglobin level increased significantly from 78 ± 11 g/L at inclusion to 83 ± 11 g/L (P < .001 Student t test) at the end of the first year of follow-up, while the reticulocyte counts significantly decreased from .118 ± .065 to .105 ± .059 (P < .05 Student t test). No relationship was found between these features and either the number of α-globin genes, the level of fetal hemoglobin, or the amount of family income. Among the 11 girls older than 9 years at inclusion, 6 started puberty during the second year of follow-up and 4 of them reached menarche within 3 years.

Overall, improvement includes regression of jaundice, partial or total regression of spleen enlargement, and general health status.

In group 2, the mean absolute gain in weight was only 1260 ± 480 g (range, -200 g to 1950 g) during the first year of follow-up, and the mean absolute gain in height was 6.3 ± 3.3 cm (range, 0-14 cm). Parameters such as age, sex, clinical severity at inclusion, level of fetal hemoglobin, β-globin haplotype, α-globin and G6PD genotype, and parents' income were not different in this group compared with group 1.

We questioned whether, in these patients, lack of improvement results from lack of understanding of the educational program by the parents. To explore this possibility, we intensified the provision of information to further motivate these parents. After a year of follow-up, in 22 patients (42.3%) the mean increase in weight (3080 ± 680 g vs 1100 ± 510 g) and in height (6.5 ± 3.2 cm vs 5.1 ± 2.9 cm) was noted as well as reduction in the frequency and the severity of SS-related acute events (Table 4). The number of other SS-related acute events recorded in the second year was less than that during the first year.

Given these findings, we focused on parental education (repeated provision of information and subsequent verification of parental understanding), and when the study was subdivided into 2 2.5-year periods, the proportion of patients with improvement was significantly higher during the second period than in the initial stages of the program (P < .01, Fischer exact test): 92 (86.0%) of 107 versus 92 (71.3%) of 129.

P falciparum malaria attacks were frequent despite antimalarial prophylaxis (12.8 episodes per 100 patient-years), and 27.8% of cases were not associated with fever (a usual major sign of an attack) despite a significant fall in the hemoglobin level.

The causes of the 56 blood transfusions are shown in Table 5; it is noteworthy that P falciparum malaria was the cause for transfusion in 12 (21.4%) cases.

Of the 10 deaths, 7 occurred in patients younger than 5 years. The number of deaths was significantly higher (P < .01, Fisher exact test) in patients with poor improvement during the first year of follow-up: 6 (11.1%) of 52 compared with 4 (2.2%) of 184. There were 4 deaths that resulted from acute splenic sequestration; in 2 of these cases, the mother had made the diagnosis at home and the child was rushed to the hospital where blood products were not available for immediate blood transfusion; in the other 2 cases, the children reached the hospital too late to get transfused.

Within the limits of our ability to collect prestudy data, disease severity appeared to be lower during CCCP than prior to the program (Tables 4, 6). The frequency of hospitalization decreased by more than half. The average cost charged to the parents for each hospitalization was approximately US $100. The CCCP annual cost charged to the parents was approximately US $40, including medical visit bills, drugs for basic treatment, and basic laboratory measurements, less than half the cost of a single hospitalization.

The Republic of Benin, in West Africa, is a country with a β^S-gene frequency of about 25%. Cotonou is a swampy area with endemic P falciparum infection, and the setting is intrinsically hostile to a child with SCD. Benin's annual gross domestic product per capita was estimated to be US $287 in 1994, with less than 5% of its budget dedicated to health services.

A more important problem, however, is that attitudes toward SCD among our population are not rational; people believe that the disease is caused by an evil spirit gnawing on their child's bones and, for treatment, seek the help of “Marabou.” Prior to the children's entry into the study, most of their health care providers were not aware of contemporary procedures of management of SS-related acute events. Treatment had consisted essentially of traditional preparations of unknown composition, vasodilators (hydergine and pervincamine), pentoxyphylline, and nonsteroidal anti-inflammatory drugs.

In a context of extreme poverty, inadequate education, and scanty health care facilities, parental understanding, faith, and compliance are vital to the success of any comprehensive clinical care program. In the present study, 85% of parents seen at the inclusion visit accepted and complied with the proposed medical follow-up. The positive outcome of our study strongly suggests that delivery of information in a perceivable manner could change the parent's irrational attitudes about the disease.

We cannot rule out a possible selection bias in the type of patient referred to us, or in the willingness of parents to participate in the clinic. However, if there were bias, it would probably result from the selection of more severe cases. Our patients were almost all homozygous for the Benin β-globin haplotype (91.5%) with a low level of fetal hemoglobin; only half of them had associated α-thalassemia. All had a severe clinical course. In developed countries, they might be eligible for hydroxyurea treatment, regular exchange transfusion, or bone marrow transplantation. But in sub-Saharan Africa it is presently illusory to contemplate such therapeutic options. In contrast, with the application of a relatively simple care program, a relative reduction in the number of hospitalizations and transfusions was observed, growth accelerated, and the frequency and severity of SS-related acute events appeared to decrease when compared with reports of illness before the opening of the clinic. Furthermore, the number of events recorded within the first year was higher than that during the following years, suggesting that once the parents complied and got used to the recommendations of the program, the SS-related acute events became less frequent and less severe.

We could not ethically include a control group in our study, but the subgroup of 22 patients is close to a cross-over control. The observed improvement in these patients after remotivation of their parents reflects the effectiveness of the CCCP, as did regression of jaundice and diminution or total regression of splenomegaly. These results clearly suggest that cultural background and lack of adequate medical care greatly contribute to the severity of SS in children in sub-Saharan Africa.

Attendance at our clinic did not solve all problems. The incidence of Plasmodium falciparum malaria was high in these subjects despite our preventive measures. Even though they had reached the hospital promptly, 2 children died from acute splenic sequestration because of the unavailability of suitable blood products.

Delayed growth and maturation, and poor nutritional status are well known in SS children; the reason for this is not clearly understood but is likely to involve increased caloric requirements.¹⁷  Several studies in developed countries have documented increased resting energy expenditure in children and adults with SCD,^18-21  and recently, low activity energy expenditure has been noted in children with SCD.²²  It has been suggested that a voluntary reduction in physical activity may be a compensatory mechanism, but the best approach to overcome undernutrition in SCD children is yet to be determined.²²  We do not have the laboratory facilities to accomplish these measures, but the sustained positive change in physical growth and nutritional status observed in 87.3% of patients (206 of 236) suggests that the metabolic requirements are sufficiently satisfied in these patients.

In conclusion, we have shown that, in an African setting, a sustained positive change could be achieved in severely ill children with SS if they were provided adequate clinical care and tenacious parental education. The program described did not require significant increase in health care expenditures and could therefore be readily implemented in communities with limited resources in which SCD incidence is quite high. The need for a committed and trained team is instrumental in the outcome of these patients. The benefit of such an active management in an African setting was recently exemplified in the outcome of pregnancy in women with SCD.²³  In recognition of the achievement of the CCCP, the Ministry of Health of Benin has come forward to finance a separate building construction within our hospital complex that would be entirely devoted to SCD.

We would like to thank Drs J. Elion, R. Krishnamoorthy, and T. Vulliamy for their help in the determination of patients' genetic characteristics and in the preparation of this manuscript; and Dr S. Charache for his comments, suggestions, and substantial editorial assistance.