Quality of hospital care for seriously ill children in less-developed countries

Transcription

1 Quality of hospital care for seriously ill children in less-developed countries Terry Nolan, Patria Angos, Antonio J L A Cunha, Lulu Muhe, Shamim Qazi, Eric A F Simoes, Giorgio Tamburlini, Martin Weber, Nathaniel F Pierce Summary Background Improving the quality of care for sick children referred to hospitals in less-developed countries may lead to better outcomes, including reduced mortality. Data are lacking, however, on the quality of priority screening (triage), emergency care, diagnosis, and inpatient treatment in these hospitals, and on aspects of these potential targets that would benefit most from interventions leading to improved health outcomes. Methods We did a qualitative study in 13 district hospitals and eight teaching hospitals in seven less-developed countries. Experienced paediatricians used a structured survey instrument to assess initial triage, emergency and inpatient care, staff knowledge and practices, and hospital support services. Findings Overall quality of care differed between countries and among hospitals and was generally better in teaching hospitals. 14 of 21 hospitals lacked an adequate system for triage. Initial patient assessment was often inadequate and treatment delayed. Most emergency treatment areas were poorly organised and lacked essential supplies; families were routinely required to buy emergency drugs before they could be given. Adverse factors in case management, including inadequate assessment, inappropriate treatment, and inadequate monitoring occurred in 76% of inpatient children. Most doctors in district hospitals, and nurses and medical assistants in teaching and district hospitals, had inadequate knowledge and reported practice for managing important childhood illnesses. Interpretation Strengthening care for sick children referred to hospital should focus on achievable objectives with the greatest potential benefit for health outcome. Possible targets for improvement include initial triage, emergency care, assessment, inpatient treatment, and monitoring. Priority targets for individual hospitals may be determined by assessing each hospital. Lancet 2001; 357: See Commentary page??? Department of Paediatrics, University of Melbourne and Murdock Children s Research Institute, Royal Children s Hospital, Parkville, Australia (Prof T Nolan FRACP); Vicente Soto Memorial Medical Center, Cebu City, Philippines (P Angos MD); Department of Pediatrics, IPPMG, Universidade Federal do Rio de Janeiro, Brazil (A J L A Cunha MD); Department of Pediatrics and Child Health, Addis Ababa University, Addis Ababa, Ethiopia (L Muhe MD); Department of Child and Adolescent Health, WHO, Geneva, Switzerland (S Qazi MD, M Weber MD); Department of Pediatrics, Section of Infectious Diseases, University of Colorado Health Sciences Centre and the Children s Hospital, Denver CO, USA (E A F Simoes MD); Unit for Health Services Research and International Health, IRCCS Burlo Garofolo, Trieste, Italy (G Tamburlini MD); and Department of International Health, Johns Hopkins University School of Hygiene and Public Health, Baltimore MD, 21205, USA (Prof N F Pierce MD) Correspondence to: Prof Nathaniel F Pierce ( npierce@jhsph.edu) Introduction Each year about 11 million children die before reaching their fifth birthday, mostly from pneumonia, diarrhoea, malaria, or measles, often with underlying malnutrition. 1 More than 99% of these deaths occur in less-developed countries. 2 Recent efforts to develop simple, effective, lowcost methods to diagnose and treat these problems have achieved substantial success. The WHO and the United Nations International Children s Emergency Fund (UNICEF) have used these findings to develop and promote Integrated Management of Childhood Illness (IMCI), an evidence-based strategy for assessing and treating sick children in ambulatory care facilities. 3,4 Although most sick children can be successfully managed at these sites, 12 34% are seriously ill and require referral to hospital for further assessment and possible admission. 5,6 In more-developed countries, clinical outcome for specific conditions, including the risk of death, is correlated with quality of hospital care. 7,8 There is little information, however, on the quality of hospital care available for sick young children in less-developed countries or its relation to outcome. It is known, however, that many die at home soon after being treated for their illness as inpatients or outpatients. 9,10 Reports of poorly organised triage and emergency care, 11 and of increased mortality associated with non-standardised management of malnourished children, 12 suggest that these and other defects may contribute to substantial avoidable morbidity and mortality. Possible targets for strengthening of hospital care include triage, emergency care, initial and follow-up assessment, in-patient management, and support services. It is unclear, however, which aspects of these potential targets would benefit most from intervention, and what specific changes would most improve health outcomes. We have carried out an observational study in seven lessdeveloped countries to describe the quality of care of seriously ill infants and young children in public hospitals, and to identify potentially correctable problems in the organisation and delivery of hospital care, especially those that might contribute most to poor outcomes. Methods Selection of sites and evaluators The study involved 21 hospitals in Bangladesh, Dominican Republic, Ethiopia, Indonesia, Philippines, Tanzania, and Uganda. Countries were chosen to ensure broad geographical representation and a wide range of infant and child mortality; reported infant and under-five mortality rates in the seven countries ranged from 32 to 113 per 1000 and 38 to 177 per 1000, respectively. 2 All hospitals belonged to the national health system. Those selected included both first-level (district) hospitals and tertiary (teaching) hospitals because both are used as primary care and referral facilities, and teaching hospitals are the source of clinical routines used throughout the country. Teaching hospitals provided clinical training for medical and paramedical graduates, but were not necessarily university hospitals. In six countries, one teaching hospital and two district hospitals were selected; in one country two of the 106 THE LANCET Vol 357 January 13, 2001

2 three hospitals visited were teaching hospitals. Showcase hospitals and those where staff had received IMCI training were excluded; in all countries, however, staff had been exposed directly or indirectly to the WHO guidelines for management of diarrhoea and, to a lesser extent, management of acute respiratory infections. The countries selected, and the types of hospitals studied, were intended to reflect the range of quality of hospital care provided for children in less-developed countries. No country declined our invitation to take part in the assessment. No formal sample size calculation or power analysis was done. The investigation excluded those born in the neonatal unit. Paediatricians with extensive experience managing sick children in less-developed countries were trained to act as evaluators. All had extensive experience using the WHO/UNICEF guidelines for managing the most common and important childhood illnesses. Hospitals were visited during One or two evaluators spent 3 days in each hospital; in most hospitals, assessments were also done in the evening and, in several, at night. During the visit, between 2 and 2 5 days were devoted to directly assessing patient care, emergency and inpatient facilities, and interviewing medical staff. Ministry of Health officials accompanied evaluators to the study hospitals, but their assessments and interviews within the hospitals were done unaccompanied by these individuals. Study instrument and assessment methods The study team developed a structured audit instrument. Hospital bed capacity, staffing levels, and the most common serious illnesses and causes of death by age-group were obtained from hospital officials. Quality and organisation of initial triage, and of emergency care facilities and practices, were determined by observation in the triage and emergency areas. Quality and organisation of inpatient care facilities and practices were determined by observation on inpatient wards, examination of patients, and detailed review of patient records. The views of emergency and inpatient ward staff on the most important factors limiting the quality of care, and how care might best be improved, were obtained by interview. Knowledge and reported practices of clinical staff were assessed by an evaluator-administered questionnaire that focused on assessment and management of five important clinical problems: pneumonia, diarrhoea with dehydration, sepsis, severe malnutrition, and hypoglycaemia. Radiology, pharmacy, and laboratory services were assessed by structured interviews with the persons in charge and confirmed by interviews with clinical staff. An item-by-item inventory was done to identify drugs, fluids, and clinical supplies in the paediatric ward, emergency area, and pharmacy. The evaluators summary conclusions and recommendations were recorded using semi-structured instruments. The complete survey instrument can be found at For the detailed reviews of clinical cases, evaluators held discussions with house staff, examined the patients, and reviewed the medical records before making judgments as to whether adverse factors had occurred at any point in the clinical care pathway, including initial triage, emergency care, assessment, inpatient treatment, monitoring, and reassessment. Adverse factors were defined as actions or omissions that might have had an important negative effect on the patient s outcome. Frequently, individual patients were seen during more than one day, beginning in the emergency area and continuing to the inpatient ward. The reference standard of care was good evidence-based clinical practice that was considered possible and appropriate in the country. This generally reflected, but was not restricted to, the WHO/UNICEF guidelines for assessing and treating sick young children who are referred to hospital This type of semi-retrospective assessment involved varying degrees of interpretation and interpolation, that is, clinical judgment. To do a truly prospective assessment for this purpose was not feasible given the logistic and ethical constraints of the study, and even if this were possible, inference about the impact of adverse factors on outcome would remain largely subjective. The same type of process was used to make overall judgments about the potential benefits for patients of possible improvements in clinical care standards, health worker knowledge and skills, facilities, and resources. With this approach, staff performance, quality of care, resources and facilities were classified as either: 1, poor, with substantial potential for avoidable mortality and morbidity; 2, many problems, considerable potential for improvement; 3, quite good, but some potential for improvement; or 4, outstanding, little or no avoidable morbidity or mortality. In the analysis, categories 3 and 4 were combined. Coding and analysis Data were coded by a single investigator (TN) and analysed with Stata (version 5). Given the design of the study and the sampling strategy, no formal statistical inferences have been drawn. Results Hospital characteristics Summary data on hospital admissions, paediatric beds, and staffing levels are given in table 1. Written guidelines were available for managing referred patients with pneumonia and diarrhoea in ten and 11 hospitals, respectively, but guidelines for the treatment of malaria, measles, meningitis, sepsis, or severe malnutrition were available in only two to four hospitals. Two hospitals in one country had guidelines for managing all of these conditions. Mortality rates for inpatient children, excluding neonates born in the neonatal unit, ranged up to 15%. Organisation of triage and emergency care In four of eight teaching hospitals, and three of 13 district hospitals, the quality of triage was considered to be good (six hospitals) or excellent (one hospital): patients were assessed promptly, emergencies were recognised, and treatment begun without avoidable delay. Only one teaching hospital, however, used a formal algorithm to promote a uniform triage process. In the remaining 14 hospitals triage quality was judged to be poor because of avoidable delays, poorly organised facilities, or inadequate assessment of patients. In only four teaching hospitals and one district hospital was the availability of supplies for Teaching hospitals District hospitals (n=8) (n=13) Number of paediatric admissions per year* Number of paediatric beds Median number of paediatric beds per doctor (range) Day 8 (5 36) 13 Night 58 (5 255) 19 Median number of paediatric beds per nurse (range) Day 20 (4 64) 10 (5 26) Night 25 (10 255) 19 *Age 0 14 years, excluding those infants born in the neonatal unit. Ratios are not adjusted for occupancy rate. One district hospital had no doctors; another had only one during the daytime. The same nurses usually rotated between day, evening, and night shifts. Table 1: Hospital staffing and case load THE LANCET Vol 357 January 13,

3 Principal diagnosis Number of patients with adverse factors* Number Inappropriate Inadequate Appropriate Inappropriate treatment Treatment given Inadequate of patients or late triage assessment treatment Antibiotics Fluids Feeding Oxygen incorrectly monitoring or given late not reassessed Pneumonia Diarrhoea Sepsis Malaria Meningitis Other Total Total patients affected (%).. 8% 41% 19% 35% 20% 21% 14% 3% 30% *Up to four adverse factors were coded for each patient. Adverse factors were identified in 100 (76%) of 131 patients. 13 adverse factors classified as other are not shown. Table 2: Assessment or treatment factors considered likely to increase the risk of poor outcome treatment in the emergency area considered good or excellent. Delays in admission to the inpatient ward following triage and any required emergency care occurred frequently in four hospitals. Four hospitals had no organised triage system or staff assigned to this task; in each, the quality of triage was considered poor. In the remainder, triage was done by doctors or paediatric residents (six hospitals), nurses (seven hospitals), or less qualified staff (four hospitals). There was no clear link between the assessed quality of triage and the type of staff doing it: triage was considered good or excellent in three of six hospitals where it was done by doctors, including paediatric residents, and in four of 11 hospitals where it was done by nurses or other staff. The quality of triage was generally better in teaching hospitals, which had somewhat higher staff to bed ratios than district hospitals (table 1). This difference seemed to be related not to staffing numbers, but rather to the presence of an organised system, however basic, for recognising and promptly dealing with the sickest infants. Adverse factors in case management Evaluators studied the management of 131 children in emergency departments or inpatient wards (table 2). 64% were less than 12 months old, and 5% were 5 years of age or older. Principal diagnoses included: pneumonia (53 children), diarrhoea (24), malaria (18), sepsis (14), and meningitis (nine). Among these patients, pre-referral factors judged to adversely affect outcome (delayed presentation, overwhelming disease) were identified in 56 (43%) of them. In 100 (76%) patients, one or more factors that were considered to increase the risk of adverse outcome were identified during hospital care (table 2). Inappropriate or late triage, and inadequte initial assessment following triage, were identified in 8% and 41% of cases, respectively. Inappropriate treatment with antibiotics, Dehydration Pneumonia Malnutrition Hypoglycaemia Sepsis Doctors: teaching 7% 32% 32% 19% 19% hospitals (n=16) Doctors: district 48% 72% 72% 80% 80% hospitals (n=25) Nurses and 29% 58% 86% 86% 100% assistants: teaching hospitals (n=7) Nurses and 50% 57% 44% 63% 88% assistants district hospitals (n=16) Table 3: Proportion of doctors, nurses, and medical assistants with inadequate knowledge or reported practice for assessing and managing specific conditions fluids, feeding, or oxygen, which accounted for 44% of all adverse factors, occurred in 80 (61%) patients. This category included both failure to give an indicated treatment and treatment given unnecessarily. Delay in giving appropriate treatment occurred in 24 (18%) patients, and inadequate monitoring, or failure to re-assess adequately during treatment, occurred in 39 (30%) patients. Staff knowledge and reported practice The questionnaire on assessing and managing five common and serious paediatric problems was administered to 41 doctors and 23 nurses or medical assistants. The results are summarised in table 3. Inadequate knowledge or reported practice occurred much less frequently among doctors in teaching hospitals than doctors in district hospitals, and nurses and medical assistants in district or teaching hospitals. Among all staff in both types of hospitals inadequate knowledge and reported practices for detecting and managing dehydration caused by diarrhoea occurred less frequently than for the other four conditions. Overall, 30 (73%) of 43 doctors assessed from all hospitals were inadequate in at least one area compared with 21 (91%) of 23 nurses and medical assistants. Laboratory and radiology services Basic laboratory diagnostic tests (blood glucose, malaria smear, haemoglobin/haematocrit, cerebrospinal-fluid microscopy, blood typing and cross matching) were each provided in at least 18 of 21 hospitals. In five teaching hospitals and seven district hospitals, however, staff reported that these services were not always available owing to lack of supplies or equipment failure. In four teaching hospitals and eight district hospitals qualitycontrol procedures were poor or absent, leading to low levels of staff confidence in test results. Chest radiographs could be done in 19 of 21 hospitals, but in only ten hospitals (four teaching hospitals, six district hospitals) were results available within 4 h; in three hospitals the delay was more than 24 h. Pharmacy and drug supplies The organisation of pharmacies was considered satisfactory or excellent in 19 of 21 hospitals. Problems with the supply of essential drugs, however, were common. Examples of the availability of several supplies required for emergency treatment are shown in table 4. Most items were currently unavailable in at least one of the eight teaching hospitals and in higher proportions within district hospitals. Moreover, even when supplies were found in the hospital they were often unavailable in the area where emergency treatment was given. In 19 of 21 hospitals parents of patients were required to pay for medications and intravenous fluids before the hospital s pharmacy 108 THE LANCET Vol 357 January 13, 2001

4 Supplies in emergency Supplies anywhere in hospital* treatment area Teaching District Teaching District hospitals (%) hospitals (%) hospitals (%) hospitals (%) Oxygen Suction device Adrenaline Salbutamol Gentamicin Chloramphenicol ORS packets Normal saline Intravenous drip sets *Proportion of hospitals with supply seen in emergency area, ward, or pharmacy. Parenteral formulation. ORS=oral rehydration salts. Table 4: Supplies found in emergency treatment area or entire hospital would provide them, even for emergencies, or to procure the medications themselves from the local market. Potential for reducing morbidity and mortality A semi-quantitative summary of the evaluators assessment of the contributions of specific hospital factors and staff performance to potentially avoidable mortality and morbidity is shown in table 5. There were important interhospital and inter-country differences in the evaluators judgments about the potential for reducing mortality and morbidity through targeted improvement in hospital services. Nevertheless, potentially correctable problems were often similar between institutions and countries, with the greatest potential for improvement being in district hospitals. The most important problems were poorly organised and delayed triage and emergency care, inappropriate inpatient or emergency treatment, poor monitoring of patients during inpatient treatment, inadequate training of physicians and nurses, a lack of job aids and guidelines for standard case management, sporadic lack of essential drugs and supplies, and understaffing, particularly at night. In some settings, these deficiencies were exacerbated by the large number of patients being seen. Country Staff performance Facilities and Support services and quality of care supplies Triage Emergency Ward Triage and Ward Pharmacy Laboratory emergency Teaching hospitals A B C D E F G District hospitals A B C D E F G =poor: substantial avoidable morbidity and mortality. 2=fair: many problems, considerable potential for improvement. 3=good: few or no significant problems. Table 5: Ratings of avoidable morbidity and mortality in study hospitals Discussion This study has at least two limitations. First, care given to sick children was assessed in only 21 hospitals in seven countries. Although selected to represent a wide range of countries and levels of development, these may not have reflected conditions in some less-developed countries, especially those in South America and the Middle East, regions that were not visited. Second, the study was largely observational. Assessments were based in part on the judgment of eight observers; unavoidable differences in their assessments might have reduced the ability to compare results between countries. Despite these potential limitations, we believe this study provides information that can guide efforts to improve the quality of care, and outcome, for sick children in hospitals. The reviewers were expert paediatricians with extensive experience in less-developed countries who had developed and used a common assessment instrument. Their assessments of quality of care, and of potential for improving outcome, were based on the WHO/UNICEF guidelines and on what they considered possible and optimally effective under local conditions. Wherever possible, observations were supported by objective information. For example, knowledge and reported practices of doctors and nurses were compared with care given to children in hospital who were examined and whose records were reviewed by the evaluators. This led to substantial agreement on areas where strengthening of hospitals would have the greatest impact on morbidity and mortality. In this study, 90% of children assessed by the study team had severe forms of common childhood illnesses, especially pneumonia, diarrhoea, malaria, sepsis, and meningitis. These conditions, often with underlying malaria, are also the most common disorders for which children are referred to hospital from ambulatory clinics. 6,16 Although their risk of mortality is high, most children referred with these problems can be treated successfully, provided they are promptly and correctly assessed, and treatment of proven effectiveness is given without delay. Strengthening hospital services so these goals may be achieved should substantially reduce severe morbidity and mortality among young children. Such efforts require the sustained support of hospital administrators and should include both small district hospitals, often located in the periphery, and large urban teaching hospitals, because both are used as primary referral sites for children seen at ambulatory clinics; in addition, teaching hospitals play a major part in establishing case management practices within a country. Our study suggests several priority areas for improving hospital services to achieve reduced morbidity and mortality. Initial triage and emergency treatment were poor in most district hospitals and in many teaching hospitals. Staff were not specifically trained and assigned to these activities, facilities were not well organised, guidelines for standard assessment and treatment were not provided, basic drugs and supplies were not available in the emergency area, and inordinate delays were frequent. These problems can be addressed by training nurses, paramedics, or doctors to do triage and provide emergency treatment following standard protocols, ensuring they are available around the clock, and stocking the triage area with essential drugs and supplies for which prepayment is not required. The WHO has developed guidelines for triage and emergency treatment of sick children younger than age 5 years, focusing especially on common emergencies, including severe dehydration from diarrhoea, severe pneumonia, severe malaria, severe malnutrition, and THE LANCET Vol 357 January 13,

5 sick infants younger than 2 months of age. 11 Field testing in Brazil has shown that after a short training period, nurses using these guidelines can assess and treat these other emergency conditions with a high level of sensitivity and specificity when compared with experienced paediatricians. 17 Important, often multiple, deficiencies were noted in the assessment, treatment, and ongoing monitoring of most children admitted to hospital as inpatients and these were judged likely to contribute to poor outcomes, including increased mortality. Deficient case management was paralleled by the reviewers assessment that most hospital staff, with the exception of doctors in teaching hospitals, had inadequate knowledge or reported practice for managing at least four of the five common paediatric problems assessed. Widespread efforts by the WHO, UNICEF, and other organisations to improve the management of children with diarrhoea may explain the better knowledge and reported performance of staff for this condition. During interviews, hospital staff frequently stated that case management could be strengthened by increasing levels of staffing and by providing them training in the management of important clinical conditions. Although increasing the number of staff may be difficult because of economic constraints, training that stresses the management of the most common paediatric problems by methods that are affordable, appropriate for local conditions, and of proven efficacy should be possible. The better knowledge and reported practices of staff for management of diarrhoeal dehydration suggests this training is also effective. Such training, including supervised practice, should be a prominent part of medical, paramedical, and nursing school curricula, and of postgraduate training of resident physicians, and should also be given as in-service training to hospital staff, including nurses. The WHO has developed guidelines for the inpatient care of children with the most common medical problems. 17 These may provide a basis for increasing the relevance of teaching in medical, paramedical, and nursing schools, and for strengthened inservice training of hospital staff. Other strategies may also be needed to ensure that guidelines and training are reflected in daily practice. These might include involving national medical colleges and professional associations as implementers of guidelines and standards, linking the use of guidelines to hospital accreditation, use of incentive schemes tied to provision of financial assistance from national or international agencies, and participation of hospital staff in regular assessment of quality of care. Despite the general patterns discussed above, differences in the quality and organisation of facilities, and in the knowledge and performance of staff, were seen among hospitals both within and between countries. This suggests that efforts to improve patient care in hospitals may need to be tailored to address problems identified in a specific country and, possibly, in individual hospitals. Problem identification would be facilitated by use of a structured survey instrument designed especially to assess patient care, organisation of facilities, and staff performance during triage, emergency treatment, and inpatient management of sick children. We believe an instrument could be developed that would permit a hospital to be adequately assessed, and its priority needs defined, during a 1 to 2 day visit by a trained paediatrician. Contributors All investigators participated in planning the study and developing the study instrument. All except N F Pierce took part in country visits to assess hospital care. T Nolan did the analysis. T Nolan and N F Pierce wrote the report. N F Pierce coordinated the study. Acknowledgments Financial support for this study was provided by the Department of Child and Adolescent Health of the WHO, Geneva, Switzerland. References 1 Murray CJL, Lopez AD. The global burden of disease: comprehensive assessment of mortality and disability from diseases, injuries and risk factors in 1990 and projected to Geneva: WHO, UNICEF. The state of the world s children New York: Oxford University Press, Gove S. Integrated management of childhood illness by outpatient health workers: technical basis and overview. Bull World Health Organ 1997; 75 (suppl 1): S Tulloch J. Integrated approach to child health in developing countries. Lancet 1999; 354 (suppl II): SII Simoes EAF, Desta T, Tessema T, Gerbresellassie T, Dagnew M, Gove S. Performance of health workers after training in integrated management of childhood illness in Gondar, Ethiopia. Bull World Health Organ 1997; 75 (suppl 1): S Kalter HD, Schillinger JA, Hossain M, et al. Identifying sick children requiring referral to hospital in Bangladesh. Bull World Health Organ 1997; 75 (suppl 1): S Meehan TP, Hennen J, Radford MJ, Petrillo MK, Elstein P, Ballard DJ. Process and outcome of care for acute myocardial infarction among Medicare beneficiaries in Connecticut: a quality improvement demonstration. Ann Intern Med 1995; 122: Meehan TP, Fine MJ, Krumholz HM, et al. Quality of care, process, and outcomes in elderly patients with pneumonia. JAMA 1997; 278: Sodemann M, Jakobsen MS, Molbak K, Alvarenga IC Jr, Aaby P. High mortality despite good care-seeking behaviour: a community study of childhood deaths in Guinea-Bissau. Bull World Health Organ 1997; 75: Islam MA, Rahman MM, Mahalanabis D, Rahman AK. Death in a diarrhoeal cohort of infants and young children soon after discharge from hospital: risk factors and causes by verbal autopsy. J Trop Pediatr 1996; 42: Gove S, Tamburlini G, Molyneaux E, Whitesell P, Campbell H. Development and technical basis of simplified guidelines for emergency assessment and treatment in developing countries. Arch Dis Child 1999; 81: Ahmed T, Ali M, Ullah MM, et al. Mortality in severely malnourished children with diarrhoea and use of a standardised management protocol. Lancet 1999; 353: WHO. Acute respiratory infections in children: case management in small hospitals in developing countries a manual for doctors and other senior health workers. Geneva: WHO, Programme for Control of Acute Respiratory Infections, WHO. The treatment of diarrhoea a manual for physicians and other senior health workers. Geneva: WHO Division of Diarrhoeal and Acute Respiratory Disease Control, WHO. Management of the child with a serious infection or malnutrition. Guidelines for care at the first referral level in developing countries. Geneva: WHO, WHO Division of Child Health and Development and WHO Regional Office for Africa. Integrated management of childhood illness: field test of the WHO/UNICEF training course in Arusha, United Republic of Tanzania. Bull World Health Organ 1997; 75 (suppl 1): S Tamburlini G, Di Mario S, Maggi RS, Vilarim JN, Gove S. Evaluation of guidelines for emergency triage assessment and treatment in developing countries. Arch Dis Child 1999; 81: THE LANCET Vol 357 January 13, 2001