Article Text

This article has a correction. Please see:


Improving the quality of paediatric care in peripheral hospitals in developing countries
  1. T Duke1,
  2. G Tamburlini2,
  3. The Paediatric Quality Care Group
  1. 1Centre for International Child Health, Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Parkville, 3052, Victoria, Australia
  2. 2Unit for Health Services Research and International Health, Institute for Child Health IRCCS “Burlo Garofolo”, Trieste, Italy
  1. Correspondence to:
    Dr T Duke, Centre for International Child Health, Department of Paediatrics, University of Melbourne, Royal Children’s Hospital, Parkville, 3052, Victoria, Australia;

Statistics from

A coordinated and collaborative approach is required

Recent observational evidence suggests there is considerable scope for improving the quality of hospital care for severely ill children in many developing countries.1 In a study of 21 hospitals in Asia and Africa that evaluated the management of 131 children, more than half were under treated or inappropriately treated with antibiotics, fluids, feeding, or oxygen. Inadequate triage and assessment, poor treatment, and insufficient monitoring may adversely affect the outcome of a significant proportion of hospitalised children, and result in unnecessary suffering or avoidable death for many children each year. In some other settings over-hospitalisation, over-diagnosis of severe illness, and over-medication has adverse consequences for health outcomes and in wasted health expenditure.2–4 Until recently little attention was paid to these issues; the reasons are several. Firstly, because many more children in low income countries die before reaching any hospital for want of more basic interventions, such as measles vaccine or oral rehydration solution, and the legitimate concern that promoting hospitals may detract from primary level or community based care. Secondly, confronting potentially avoidable deaths or suffering in children is upsetting, to communities, to hospitals, and to health workers who struggle to provide the highest level of care they can. Furthermore, in most countries academic thought and energy has been predominantly directed towards clinical problems seen in tertiary university hospitals, with inadequate consideration of the health system problems in peripheral hospitals, which are often understaffed, poorly equipped and maintained, and with little or no ongoing staff education. Also, until recently there has been little evidence that outcomes could be improved without very major investment in staffing and technology.

While there are examples of inequitable and inefficient resource allocation where tertiary hospitals have consumed disproportionate resources and starved primary health care, good quality first referral level (or district hospital) care is crucial for reducing child mortality and for a credible and efficient primary health care system. Implementation of integrated management of childhood illness (IMCI) guidelines implies referral of up to 20% of patients in most settings. These children are the most severely ill and those at highest risk of death. It is important that case management in hospitals follows diagnostic criteria and treatment guidelines that are consistent with the IMCI outpatient or local primary care guidelines, otherwise the credibility of such guidelines may be undermined and the strategies ultimately abandoned.

“Good quality district hospital care is crucial for reducing child mortality”

There is now some evidence for improvement in outcomes for hospitalised children through support for health workers, structured clinical care, and better use of existing resources, rather than major financial or technological investment. Recent work from Brazil and Malawi suggests that emergency triage, assessment and treatment guidelines can be used by nurses to identify children needing high priority treatment.5,6 Case management of severe malnutrition7–9 and pneumonia,10 and neonatal care11 can be substantially improved with better ward organisation, clinical guidelines, active staff participation, and limited additional resources.

Components of quality improvement strategies are standards, assessment tools, participation, and driving forces. Standards of care must be specific to the level of the health facility and the needs of the country, reflect an appropriate balance between resources used for curative and public health, and emphasise low cost simple technology options and drugs. In collaboration with paediatricians worldwide, the World Health Organisation (WHO) has developed guidelines for clinical management of children with severe illnesses at hospitals of first referral.12 These require only limited laboratory services but essential drugs for the care of seriously ill children. Some countries have their own long traditions of treatment protocols for common illnesses that act as standards of care.13

Structured assessments of quality of care in hospitals have taken place in several countries, including Peru, Brazil, Angola, and several newly independent states of Central Asia.2–4,14,15 These have included assessment of infrastructure, drugs, equipment, human resources, laboratory services, records, and observed quality of clinical care. Assessment models combine self reported data by hospitals, and experienced observers working with local paediatricians and health workers. The aim is to describe the characteristics of hospitals and the health service, their major problems and strengths, and the resources available. By targeting a few crucial areas that are limiting steps in quality of clinical care, strategies for improvement can be defined.14,15 As part of informing health policy these hospital assessments provide feedback for clinicians and prompt discussion about ideas for change where it is necessary. Such assessments may be an important entry point in planning improvements to hospital clinical care,15 but it is important that they are linked to long term implementation strategies and not exercises in data gathering. Identifying solutions to high priority problems, and highlighting key structural or political problems that require longer term input, are both important if quality improvement is to be sustained, and if clinicians are to have confidence in the process. It is important also that hospital based interventions do not take precedence when there are major deficiencies in public health such as unsafe drinking water, poor immunisation rates, or a lack of primary care strategy. The issues are not confined to developing countries. Transition countries and developed countries need to improve the quality of care as well, although priorities and limiting factors may be substantially different.

“Documented experience with paediatric mortality and morbidity audit is scarce”

Development of audit and critical incident monitoring may be an initial step when quality improvement is being generated within small hospitals where staff have limited time and resources. Although there is a strong tradition of effective perinatal and surgical auditing from developed and developing countries, documented experience with paediatric mortality and morbidity audit is scarce. A problem solving approach to paediatric audit, based on comparison of actual management against standards and evidence based clinical guidelines can have positive benefits. If such audits are conducted in a sensitive way they can serve the purpose of team building and be a focus of continuing education.16

The success of hospital assessments and internally driven quality improvement processes depend on having sufficient resources to effect and sustain gradual changes, and local staff having ownership over the process.17 In Peru a maternal and neonatal quality improvement programme has built a wide variety of improvements to clinical practice and education around self identification of problems by health facility clinical teams. The problems identified relate to drug and equipment supplies, lack of standardised care or training for specific diseases, and issues of client satisfaction. The clinical teams attempt to identify and solve problems by themselves. In only about 20% of cases this is not possible, and they then request help from the local level authorities, referral hospitals, or the central ministries of health. Health centres participating in this project report an increased demand for care, and a reduction in case fatality among mothers delivering, compared with non-participating health centres.18

There is a need for more research on strategies that work and those that do not, and the levels of support required. The evidence base to WHO hospital guidelines12 should be carefully documented if we are to encourage widespread usage and incorporation into teaching programmes. In addition the evidence for and against treatments for common illnesses that are given in some countries but not recommended need to be documented. This will be necessary if effective local adaptations, which merge current treatments and WHO guidelines are to occur. Research is needed to field test the effectiveness of a range of approaches to quality improvement (external assessment, internal audit, etc).

Standards, assessment tools, and evidence are essential, but not alone sufficient for the process to succeed. There is a financial cost to running a service that is good enough to prevent avoidable morbidity and deaths. In many settings much can be done with very little if one builds on existing human resources; health workers generally embrace initiatives to improve care as long as such initiatives are appropriately supported, adding value and a sense of fulfilment to their work. In some settings there will be substantial cost shifting from inefficient practices, and a better quality clinical service may be provided for not much more than current costs. Rationalisation of over-hospitalisation and over-medication using standardised clinical guidelines would free up considerable funds in some health systems. However, in most lower income countries resources are so limited and manpower so over-extended that problems highlighted by external or internal assessments will be demoralising to health workers if sufficient financial, technical, and human resources are not made available to affect change. In the chain where good data inform effective implementation the links are made up of on-ground motivation, funding, and political will.

Quality improvement should be integrated into a variety of clinical and management initiatives. The overall approach should not be limited to paediatric medical care, as there is evidence of a need for improvement in perinatal care,19,20 pre- and post-surgical care and anaesthesia management,21–23 and general hospital practices,23 to further lower hospital case fatality rates in less developed countries.

“Principles of quality improvement will need to be integrated into training”

For these issues to be part of the health culture, principles of quality improvement will need to be integrated into undergraduate and postgraduate training, and incorporated into regional hospitals providing second referral level and tertiary teaching hospital care. Larger hospitals are where medical students, interns, and nurses get their value systems about good medical practice, and set patterns for a lifetime of work. For effective local adaptation of clinical guidelines, evidence based capacity must be developed among paediatricians in lower income countries. This may be done by partnerships with Western universities that offer appropriate evidence based medicine and public health training, with the assistance of international partners who could provide training fellowships.

Basic principles of quality improvement need also to be incorporated into training for child health nurses, clinical assistants, health extension officers, and other paramedical health workers, as these clinicians often work in the least supported and resourced environments. In such settings empowerment of and support for nurses to make management decisions will require skills not previously taught in undergraduate training, and a cultural shift in roles, hierarchies, and expectations for both nurses and doctors.

There is a need for international and local organisations to support these issues. A worldwide effort for improving quality of paediatric hospital care requires collaboration from WHO, other international agencies, and non-government organisations with large international coverage, and major donors such as the World Bank. Initially support may help define strategies that are effective and acceptable in a variety of resource poor settings. There are pivotal roles for paediatricians, paediatric nurses, and their professional organisations to play; the International Paediatric Association has the potential to encourage this work at national and regional levels. Effective partnerships between the IPA and WHO have the potential for extensive coverage, offering technical support to clinical paediatricians and to those in management positions within Ministries of Health. National paediatric societies can collaborate with health authorities in deciding what approaches are appropriate locally, adapting standards of care to local resources and circumstances, training and supervision, assessing clinical care, providing feedback and support, and evaluating outcomes. Countries without formal paediatric societies or paediatricians need the most support. There are also roles for collaborative involvement of non-government organisations; however, the processes decided on must be owned internally, within health ministries, paediatric societies, and health facilities. The equity of approaches within the overall health care system, and what are appropriate for an individual country’s circumstances, will require careful consideration, and can only be decided at a local level.


The quality of medical and nursing care provided in peripheral hospitals in developing countries has an impact on the health and lives of millions of children each year. With recent evidence that such care is poor in many places there is a need for a serious coordinated approach. Improvements in triage, diagnosis, treatment guidelines, monitoring, and follow up may reduce case fatality and iatrogenic complications. These can be seen as public health as well as clinical problems, and demand approaches that can be brought to large scale. Structured hospital assessments might provide an entry point for deciding priorities for improvement, and the effective implementation and integration of evidence based standardised clinical guidelines will be important for many countries. Simple ways to evaluate the quality of clinical care and strategies for making improvements need to be included in undergraduate medical, nursing, and paramedical curricula and in postgraduate training. Research is needed to document the evidence behind standard treatment guidelines and investigate the acceptability and effectiveness of different quality development approaches in smaller rural and remote health systems. Developing skills in evidence based practice among paediatricians in lower income countries will be important for local adaptation and longer term sustainability of quality improvement strategies. Academic thought needs to be applied to addressing the problems of poorly functioning district hospitals; the problems in these hospitals should not be seen as out of sight and out of mind. A coordinated and collaborative approach is required that may involve ministries of health, national paediatric associations, the WHO, the International Paediatric Association, Western and developing country university departments, and non-government organisations.


Many people contributed ideas to this paper. The Paediatric Quality Care Group is an email discussion group that has included: Ganapati Bhat, Harry Campbell, Andrew Clark, Simona Di Mario, Mike English, Angelika Krug, Claudio Lanata, Michael Gracey, Lena Grahnquist, Elmarie Malek, Sue Nicholson, Steve Rogers, Jacques Schmitz, TL Soo, David Southall, Abner Targoola, and Martin Weber. Many of the ideas in the paper were discussed at an International Workshop at the University of Pretoria, South Africa in November 2001.

A coordinated and collaborative approach is required


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.

Linked Articles

    BMJ Publishing Group Ltd and Royal College of Paediatrics and Child Health