OBJECTIVE: A randomized controlled trial compared day care versus hospital care management of pneumonia.
METHODS: Children 2 to 59 months of age with severe pneumonia received either day care, with antibiotic treatment, feeding, and supportive care from 8:00 am to 5:00 pm, or hospital care, with similar 24-hour treatment.
RESULTS: In 2006–2008, 360 children were assigned randomly to receive either day care or hospital care; 189 (53%) had hypoxemia, with a mean ± SD oxygen saturation of 93 ± 4%, which increased to 99 ± 1% after oxygen therapy. The mean ± SD durations of day care and hospital care were 7.1 ± 2.3 and 6.5 ± 2.8 days, respectively. Successful management was possible for 156 (87.7% [95% confidence interval [CI]: 80.9%–90.9%]) of 180 children in the day care group and 173 (96.1% [95% CI: 92.2%–98.1%]) of 180 children in the hospital care group (P = .001). Twenty-three children in the day care group (12.8% [95% CI: 8.7%–18.4%] and 4 children in the hospital care group (2.2% [95% CI: 0.9%–5.6%] required referral to hospitals (P < .001). During the follow-up period, 22 children in the day care group (14.1% [95% CI: 9.5%–20.4%]) and 11 children in the hospital care group (6.4% [95% CI: 3.6%–11%]) required readmission to hospitals (P = .01). The estimated costs per child treated successfully at the clinic and the hospital were US$114 and US$178, respectively.
CONCLUSION: Severe childhood pneumonia without severe malnutrition can be successfully managed at day care clinics, except for children with hypoxemia who require prolonged oxygen therapy.
WHAT'S KNOWN ON THIS SUBJECT:
Severe and very severe pneumonia in children without comorbidities can be successfully managed on a day care basis at established day care clinics, if adequately trained and motivated staff members and logistic support can be made available.
WHAT THIS STUDY ADDS:
Severe childhood pneumonia without associated comorbidities, such as severe acute malnutrition, can be managed safely and effectively on a day care basis at established day care clinics, except for children with hypoxemia who require oxygen therapy for >6 hours.
The fourth Millennium Development Goal has concentrated efforts on addressing priority areas for improving child survival rates throughout the world, with an aim of reducing national child mortality rates by two-thirds in 1990–2015.1 Improvements in management strategies for major causes of child death, such as pneumonia and neonatal illnesses, should be a priority for improving child survival rates in developing countries.2,–,4 An estimated 150 million episodes of pneumonia occur throughout the world each year, and 11 to 20 million of those episodes require hospitalization in developing countries.5,6 Acute lower respiratory tract infections, particularly pneumonia, currently are the greatest single cause of deaths among <5-year-old children in developing countries, being responsible for 19% of 9.7 million deaths per year.7,8 Acute lower respiratory tract infections cause >2 million deaths per year throughout the world, mostly resulting from pneumonia, and 90% to 95% of the deaths occur in developing countries.9,–,13 In Bangladesh, acute lower respiratory tract infections account for 25% of deaths among <5-year-old children and 40% of deaths among infants.14 Depending on clinical presentation, pneumonia can be classified as very severe, severe, or nonsevere, according to the World Health Organization,15,–,18 which recommends that diagnosis of pneumonia should be based primarily on visible clinical parameters.15,19,20 Severe childhood pneumonia requires hospitalization for supportive treatment, such as suctioning, oxygen therapy, fluid and nutritional management, and close monitoring.15,–,18 In Bangladesh, there are not enough pediatric beds in hospitals for admission of all patients with severe pneumonia. Therefore, it is important to provide institutional care for children who cannot be hospitalized. Two prospective observational studies showed that day care facility-based, modified primary care management of severe acute malnutrition21 and severe pneumonia22 were successful and cost-effective as alternatives to hospitalization. On the basis of the findings of earlier studies indicating both efficacy and safety of day care-based management, we performed a randomized controlled trial (RCT) to compare day care management with hospital care management of severe childhood pneumonia.
This was a collaborative study involving the International Centre for Diarrhoeal Disease Research, Bangladesh (ICDDRB), the Radda Maternal and Child Health-Family Planning Centre, the Institute of Child Health and Shishu Saystha Foundation Hospital (ICHSH), and the University of Basel (Basel, Switzerland). The research and ethical review committees of the ICDDRB approved the study. We conducted the study from September 2006 through November 2008 at the Radda clinic and the ICHSH. The Radda clinic is a nongovernmental organization that has provided maternal and child health services since 1974 in Mirpur, Dhaka, from 8:00 am to 4:00 pm every day except for weekends and holidays. It provides care for common childhood illnesses, including pneumonia, on an outpatient basis. We had a ward consisting of 5 beds and a study team comprising 1 physician, 2 nurses, and 4 health care workers.22
The ICHSH was the primary site for hospital care; it has provided health care to children through outpatient facilities, attended by ∼450 children per day, and 125-bed inpatient facilities for pediatric medicine, pediatric surgery, and neonatology since 1994. We had a ward consisting of 5 beds and a study team comprising 2 physicians, 4 nurses, 1 research assistant, and 4 health care workers. Provisions for oxygen therapy and electric suction therapy, a pulse oximeter, a nebulizer, a glucometer, and a weighing scale were made available at both the clinic and the hospital.
We estimated the sample size for severe pneumonia on the basis of the success rate of 93% in our recently published day care pneumonia study,22 with the assumption of a maximal expected success rate of 99% with hospital management. With 80% power and 5% significance, the required sample size was estimated to be 165 for each of the 2 treatment groups. To allow 10% rates of discontinuation for any reason, 368 children (184 at each site) with severe pneumonia needed to be studied at both the clinic and the hospital.
Children of either gender, 2 to 59 months of age, with symptoms of severe pneumonia, defined as cough or difficulty breathing with lower chest wall indrawing, according to World Health Organization criteria,15,–,18 were enrolled after parental consent was obtained. Children with a history of receiving injectable antibiotics for treatment of pneumonia during the illness and those with associated comorbidities, such as tuberculosis, congenital heart disease, bronchiolitis, bronchial asthma, severe acute malnutrition (SAM) (weight-for-age z score of <3), sepsis, hypoglycemia, convulsions, or meningitis, were not eligible. Children were not screened for HIV exposure through history or laboratory investigation, because the prevalence of HIV infection is very low (<1%) in our setting.
After the eligibility criteria were fulfilled and written informed consent was obtained, children were assigned randomly to receive either day care at the clinic or hospital care at the ICHSH. We used block randomization and produced computer-generated random numbers assigned to 1 of the 2 interventions. The random numbers, corresponding to specific interventions, were sealed in sequentially numbered envelopes. Until the envelopes were opened, neither patients nor physicians were aware of the treatment groups. Treatment was assigned at enrollment through opening of the sequentially numbered envelopes, starting with the lowest number and moving toward higher numbers. One statistician not involved in the study prepared the randomization list.
Parents brought their children to the clinic at 8:00 am and took them home at 5:00 pm each day. Oxygen saturation was measured routinely, by using pulse oximetry, for every child after enrollment and before administration of oxygen therapy. Because oxygen saturation measurements depend on patients' skin pigmentation and a higher target value of 95% resulted in a satisfactory level of oxygenation for black patients, we used this cutoff value for our study.23,24 Routine monitoring of hypoxemic children was performed at intervals of 30 minutes to 2 hours, depending on the patient's condition, until oxygen saturation remained stable at ≥95% with room air. Oxygen saturation was checked again during oxygen therapy, as well as after removal of administered oxygen for 2 to 5 minutes. Children who remained hypoxemic at 5:00 pm were referred to the ICHSH or Dhaka Hospital of the ICDDRB. All children received once-daily injections of ceftriaxone, in doses of 75 to 100 mg/kg, for ≥5 days. Cotrimoxazole was not considered for the treatment of pneumonia attributable to Pneumocystis jiroveci (previously known as Pneumocystis carinii), because the prevalence of HIV infection is very low in our setting. Feeding of children receiving day care was described in our recent report.22
The management described above continued on each day of the week until there was clinical improvement, defined as no fever, no fast breathing, no lower chest wall indrawing, no danger signs, no rales on auscultation, and no hypoxemia. Children were considered afebrile when the rectal temperature became <38°C and remained so for ≥24 hours. No fast breathing was defined as age-specific respiratory rates of <50 breaths per minute for infants 2 to 12 months of age and <40 breaths per minute for children 12 to 59 months of age. No lower chest wall indrawing was indicated if the lower chest wall did not go in when child inhaled. No danger signs referred to the absence of all of the following danger signs: not able to drink/feed, central cyanosis, head nodding, stridor in a calm child, abnormal sleepiness, convulsion, and severe clinical malnutrition. There was no hypoxemia when the oxygen saturation was ≥95% with room air, as recorded by the pulse oximeter. After successful treatment, children were discharged from the clinic with advice for follow-up care. Children were referred to the ICHSH or Dhaka Hospital of the ICDDRB if they failed to experience improvement with the day care management.
Children assigned randomly to receive hospital care remained at the ICHSH for the entire duration of the illness and received similar treatment, with once-daily injections of ceftriaxone, in doses of 75 to 100 mg/kg, for ≥5 days and supportive care such as suctioning, oxygen therapy, and diet. Infants and children >6 months of age received milk-suji (a mixture of milk and rice powder boiled together, with 281 kJ and 1.4 g of protein per 100 mL) at 10 mL/kg per feeding, every 2 hours. Breastfed children continued breastfeeding, and infant formula (286 kJ and 1.5 g of protein per 100 mL) was given to nonbreastfed infants 2 to 6 months of age. The same treatment continued for 24 hours every day (compared with 9 hours/day at the clinic) until improvement of the child's condition, as defined earlier, was observed. Parents were asked to bring their children to their primary sites of management (clinic or hospital) every 2 weeks for a period of 3 months, for examination by a physician, as described in our recent report.22
The main outcome measures of the study were the success of day care/hospital care, defined as improvement in clinical condition without referral to another hospital, full compliance with the day care/hospital care management until recovery, without premature discontinuation of the study for any reason, and not dying during the study. All data were collected on pretested case report forms, edited, entered into a personal computer, and analyzed by using SPSS 11.5 (SPSS, Chicago, IL). The main outcome measures were expressed as the proportions (with 95% confidence intervals [CIs]) of successes and failures of day care/hospital care management. Other outcome measures were the proportions (with 95% CIs) of children who required hospital referrals and who discontinued the study prematurely without fulfilling the criteria for success.
The cost implications for treating severe childhood pneumonia were assessed for an urban day care health facility. In addition to the assessment of clinical effectiveness, analysis of cost-effectiveness of the day care clinic-based approach would be useful for determining the real value of these day care facilities for treating severe childhood pneumonia. The analysis was performed from a societal perspective, as recommended by current standards for cost-effectiveness analyses. The costs incurred at 2 alternative health care settings for treatment of children with severe pneumonia were recorded over a period of 1 year (January through December 2007). The components of provider or facility costs included costs related to personnel, diagnostic tests, drugs, diet, food preparation, and supportive care, and the average costs per patient were calculated. Estimates of treatment success were determined from the study for both health care facilities. The cost-effectiveness analysis involved calculation of the average cost per patient treated successfully; the estimation involved computing the total cost estimate and then dividing that by the number of patients who were treated successfully.
We screened 913 children from the Radda clinic and the ICHSH, 553 of whom were not enrolled in the study for various reasons (Fig 1). Three hundred sixty children with severe pneumonia were enrolled either from the day care clinic (51%) or from the hospital (49%) and were equally (180 in each) assigned randomly to day care or hospital care (Table 1). Tables 1 and 2 show the comparison of baseline characteristics of the study children. A statistically significant difference between the 2 groups was observed for none of the parameters except hypoxemia and hepatomegaly. On examination, 189 (53%) of 360 children had hypoxemia, with a mean ± SD oxygen saturation of 91 ± 5% (Table 3), which increased to 99 ± 1% with oxygen therapy (Table 4). The mean ± SD duration of oxygen therapy for correction of hypoxemia was 340 ± 555 minutes (Table 2). Of 113 clinic children with hypoxemia, 20 (17.7% [95% CI: 11.8%–25.8%]) were referred to the hospital because they remained hypoxemic for prolonged periods, even up to the end of the first day of enrollment (5:00 pm) (Fig 1). A significantly greater number of children with hypoxemia were treated at the clinic, compared with the hospital (P < .001) (Table 1).
Successful management was possible for 156 (87.7% [95% CI: 80.9%–90.9%]) of 180 children who received day care and 173 (96.1% [95% CI: 92.2%–98.1%]) of 180 children who received hospital care (P = .001). Of the remaining 24 children (13.3% [95% CI: 9.1%–19.1%]) in the day care group, 23 (12.8% [95% CI: 8.7%–18.4%]) were referred to the hospital and 1 (0.6% [95% CI: 0.1%–3.1%]) discontinued treatment (Table 5). Of the remaining 7 children (3.9% [95% CI: 1.9%–7.8%]) in the hospital care group, 4 (2.2% [95% CI: 0.9%–5.6%]) were referred to other specialized hospitals and 3 (1.7% [95% CI: 0.6%–4.8%]) discontinued treatment (Table 5). There were no deaths during the day care and hospital care study period; however, 2 children (1.1% [95% CI: 0.2%–4%]) died during the 3-month follow-up period and 1 child (0.6% [95% CI: 0.1%–3.1%]) died after the follow-up period, all 3 in the hospital care group, with none in the day care population (Tables 5 and 6). During the follow-up period, 22 children (14.1% [95% CI: 9.5%–20.4%]) in the day care group and 11 children (6.4% [95% CI: 3.6%–11%]) in the hospital care group required hospital readmission (P = .01) (Table 5).
The total annual costs incurred for treatment at the clinic and the hospital were US$9951 and US$17 516, respectively. The average cost per patient at the ICHSH (US$175) was much higher than that at the clinic (US$99.5). The estimated costs per patient treated successfully at the clinic and the hospital were US$114 and US$178, respectively. For 87 patients who were treated successfully at the clinic, a total amount of US$6581 (difference in costs) was saved, compared with hospitalized children. If those 87 patients had been treated at the ICHSH, then additional costs of US$6581 (the savings) would have been incurred.
The main purposes of this RCT were to assess whether it would be possible to treat severe childhood pneumonia safely and effectively on a day care basis and whether this modality would be more cost-effective than hospital care. The results showed that children with severe pneumonia could be treated safely and effectively on a day care basis, as effectively as in the hospital. These results may have a great impact on the treatment of severe childhood pneumonia, particularly in resource-poor countries where hospital beds are scarce. The findings should be easily replicated in most urban and rural outpatient clinics in developing countries, provided proper training and motivation of staff members and provision of logistic support are guaranteed. The additional funds needed would be well invested, given the lower costs of day care versus hospital care (US$114 vs US$178). However, policy and program changes would be necessary to add such components to outpatient clinics, and this would require additional human and financial resources, which would not be easy to obtain.
The reason for a significantly greater number of children in the day care group, compared with the hospital care group, requiring referral to a hospital during the follow-up period was the greater number of children with severe hypoxemia in the day care group. Twenty (17.7%) of 113 children in the day care group who presented with hypoxemia were referred to the hospital for prolonged oxygen therapy, and thus day care management was a failure. In addition, 20 (87%) of 23 children who experienced failure of day care management had hypoxemia. Moreover, 12 children were readmitted to a hospital because of hypoxemia that developed during the follow-up period. All of these findings clearly indicate that hypoxemia is an important risk factor for failure of day care management, as well as the need for follow-up admissions. Our results are consistent with earlier reports that hypoxemia in children with pneumonia is a predictor of severe disease and is a risk factor for death.25,26 There is now evidence that ensuring ample supplies of oxygen and promoting a routine, systematic approach to screening for hypoxemia by using pulse oximetry are associated with improved quality of care and reduced mortality rates and that the required technology is affordable and sustainable in district-level hospitals and outpatient clinics in developing countries.25,27,–,31 However, a significant proportion of the children in the current study had hypoxemia, which clearly indicates the importance of accurate diagnosis of hypoxemia in determining the safety of outpatient treatment.32 Pulse oximetry would enable accurate identification of hypoxemia and might increase the safety and cost-effectiveness of this recommendation,33 and this diagnostic tool should be included. This was demonstrated in our recent study that noted successful day care management of severe childhood pneumonia with the use of pulse oximetry as an important part of the treatment algorithm.22,34
A reviewer identified the total duration of parenteral ceftriaxone therapy as the reason for prolonged duration of day care and hospital care, which was attributable to the study design of keeping all children until full recovery and completion of antibiotic therapy. We are now planning investigations of a modified day care concept with reduction of the total duration of the day care stay, with proper arrangement for administration of injections in daily clinic visits after most of the end points are met. We also can consider the possibility of prescribing suitable oral antibiotic therapy as home therapy after improvement, as shown in a multicenter study in which parenteral penicillin therapy for 48 hours was replaced by oral amoxicillin therapy after improvement.35
The fact that none of the children who were treated on either a day care or hospital care basis died is very reassuring. However, this may be attributable to exclusion of children with associated comorbidities such as SAM, sepsis, hypoglycemia, convulsions, meningitis, congenital heart disease, and tuberculosis, according to protocol guidelines. Children were treated by adequately trained research staff members who remained under greater supervision than normal and thus were likely to have been more motivated and to have worked with greater devotion. A better staff member/patient ratio than normal, in a country where the ratio is usually suboptimal, also might have played a role.
A major strength of our study is that this was a RCT; therefore, the efficacy of day care management was compared directly with that of hospital care. This RCT also included a component for assessment of the cost-effectiveness of interventions, which indicated that day care management was less expensive. This would be important for selection of the intervention for wider implementation in national programs.
A major limitation of our study was the exclusion of children with SAM, which is a common underlying condition among children with pneumonia in developing countries. We are now conducting another study comparing the effectiveness and safety of day care-based management with those of hospital-based management of severe childhood pneumonia with SAM, at the Radda clinic and the ICHSH.
The results of this RCT identified a way to use scarce hospital beds in developing countries more efficiently, by selecting for day care treatment children with severe pneumonia who, on the basis of existing guidelines, would have been identified as requiring hospitalization. This would be a practical approach in developing countries. A smaller investment in upgrading day care facilities through development of trained resources and procurement of supporting equipment could provide a much greater return. Our results demonstrated that a select group of children with severe pneumonia, without associated comorbidities such as SAM, could be treated safely and effectively on a day care basis, as effectively as with hospital management, except for children with hypoxemia who required oxygen therapy for >6 hours. Identification of these children with hypoxemia is necessary, because they are at risk of death and therefore need to be hospitalized for support and care for a longer period. Such an ambulatory treatment option also is less expensive than hospital care.
This study was funded by the Eagle Foundation (Geneva, Switzerland) (grant GR-00485); the ICDDRB acknowledges with gratitude the commitment of the Eagle Foundation to the research efforts of the ICDDRB.
- Accepted June 18, 2010.
- Address correspondence to Hasan Ashraf, MD, International Centre for Diarrhoeal Disease Research, Bangladesh, Clinical Sciences Division, 68 Shaheed Tajuddin Ahmed Sharani, Mohakhali, Dhaka 1212, Bangladesh. E-mail:
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.
- ICDDRB =
- International Centre for Diarrhoeal Disease Research, Bangladesh •
- ICHSH =
- Institute of Child Health and Shishu Sasthya Foundation Hospital •
- RCT =
- randomized controlled trial •
- SAM =
- severe acute malnutrition •
- CI =
- confidence interval
- 1.↵United Nations General Assembly. United Nations Millennium Declaration. New York, NY: United Nations; 2000. Available at: www.un.org/millennium/declaration/ares552e.pdf; Accessed March 2, 2009
- Duke T,
- Tamburlini G
- 4.↵World Health Organization. Life in the 21st Century: A Vision for All. Geneva, Switzerland: World Health Organization; 1998:66
- Wardlaw TM,
- Johansson EW,
- Hodge M
- 15.↵World Health Organization. Acute Respiratory Infections in Children: Case Management in Small Hospitals in Developing Countries: A Manual for Doctors and Other Senior Health Workers. Geneva, Switzerland: World Health Organization; 1990. Publication WHO/ARI/90.5
- 17.↵World Health Organization. Management of the Child With a Serious Infection or Severe Malnutrition: Guidelines for Care at the First-Referral Level in Developing Countries. Geneva, Switzerland: World Health Organization; 2000. Publication WHO/FCH/CAH/00.1
- 18.↵World Health Organization. Management of the Young Child With an Acute Respiratory Infection: Programme for Control of Acute Respiratory Infections. Geneva, Switzerland: World Health Organization; 1991
- 19.↵World Health Organization. Technical Basis for the WHO Recommendations on the Management of Pneumonia in Children at First-Level Health Facilities. Geneva, Switzerland: World Health Organization; 1991. Publication WHO/ARI/91.20
- Ashraf H,
- Ahmed T,
- Hossain MI,
- et al
- Ashraf H,
- Jahan SA,
- Alam NH,
- et al
- Onyango FE,
- Steinhoff MC,
- Wafula EM,
- Wariua S,
- Musia J,
- Kitonyi J
- Duke T,
- Campbell H,
- Graham S
- Copyright © 2010 by the American Academy of Pediatrics