Article Text

Integrated management of childhood illness: what have we learned and how can it be improved?
  1. Mickey Chopra1,
  2. Nancy J Binkin2,
  3. Elizabeth Mason3,
  4. Cathy Wolfheim3
  1. 1School of Public Health, University of the Western Cape, Cape Town, South Africa
  2. 2Division of Global Health, San Diego State University, San Diego, California, USA
  3. 3Department of Maternal, Newborn, Child and Adolescent Health, WHO, Geneva, Switzerland
  1. Correspondence to Mickey Chopra, UNICEF, 3 UN Plaza, New York, NY 10017, USA; mchopra{at}unicef.org

Abstract

Integrated management of childhood illness (IMCI) was developed by the WHO and UNICEF in the mid-1990s as a strategy to reduce under-five mortality. Aimed at countries with mortalities >40/1000 live births, it has been adopted by more than 100 countries. Repeated evaluations have demonstrated that IMCI can improve the quality of clinical care for sick children, but coverage has rarely reached high enough levels to achieve the expected reductions in mortality. In this paper, we discuss the reasons why IMCI has fallen short of its full potential, what has already been done to address some of these issues, and how it could be repositioned to achieve its original goals.

Statistics from Altmetric.com

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.

Introduction

In the mid-1990s, the WHO and UNICEF developed and promulgated Integrated Management of Childhood Illness (IMCI), which was aimed at reducing morbidity and mortality from common and potentially serious childhood illnesses or conditions that were at the time contributing to more than 70% of childhood deaths.1 The strategy, which was ultimately adopted by more than 100 countries, remains a central element of many child health programmes, although the extent to which it is implemented varies from country to country (see online supplementary appendix 1).

Provision of quality clinical services remains an important predictor of child mortality,2 3 and there has been a renewed interest in IMCI as countries strive to attain the fourth Millennium Development Goal of a two-thirds reduction in child mortality between 1990 and 2015. In this paper, we summarise the history of IMCI, its successes and limitations, and suggest how it should be revised and improved to meet the needs of countries still facing high child mortality.

Historical background

Since its inception, IMCI has been seen as an essential response to what had been a fragmented system of vertical programmes that had emphasised the treatment of specific diseases or conditions without taking into account the frequent overlapping of symptoms and possible presence of more than one illness. In many countries, it built on and even replaced the Control of Diarrheal Diseases (CDD) and Acute Respiratory Infections programmes4 that had been criticised for being too narrow and vertical.5

The original focus of IMCI was threefold: (1) to improve health worker performance; (2) to improve health systems; and (3) to improve family and community practices. Initially it consisted of a set of evidence-based treatment guidelines for countries with an under-five mortality of at least 40/1000, where infectious conditions accounted for the large majority of deaths.6 Under IMCI, health workers were trained to recognise and treat pneumonia, diarrhoea, malaria, measles, meningitis/sepsis, otitis media, malnutrition and anaemia; to immunise; and to provide counselling and assistance with breastfeeding and with nutrition for the older infant. The approach also included guidelines for referral of children who required additional evaluation and treatment at higher level facilities (see online supplementary appendix 2).7

What is already known on this topic

  • Integrated management of childhood illness (IMCI) has improved the quality of care for children in developing countries.

  • The impact of IMCI on child mortality has remained limited as a result of poor population coverage.

What this study adds

  • Lack of local adaptation of guidelines, insufficient trained staff, supervision and referral mechanisms and poor utilisation of health services have limited IMCI coverage and impact.

  • Better focus and targeting of guidelines, innovative training and supervision methods, community health worker and private sector involvement and improved coordination can improve IMCI coverage.

Guidance was developed to facilitate the adaptation of the generic guidelines by individual countries based on their local disease profile and other considerations,8 and technical assistance was made available.6 To ensure that healthcare workers were familiar with the guidelines and referral mechanisms, an innovative skills-based training programme lasting 11 days was developed and successfully piloted; the training plan included a substantial number of hours of hands-on practice and a supervisory visit 4–6 weeks later to further reinforce skills.

After an initial phase of piloting and testing, IMCI was launched in 1996 in Tanzania and Uganda.6 The initial evaluations in the two countries showed promising results,9 especially in improving the quality of care,9 with a subsequent detailed study in Tanzania demonstrating improvements in observed healthcare worker practices, including assessment and referral of sick children in IMCI districts when compared with districts that had not implemented IMCI.10

Early implementation in these and other countries focused mainly on training physicians and nurses in clinical management. The other two components, improving health systems and improving family and community practices, received considerably less attention. In 1998, WHO and UNICEF agreed on a list of 12 key family and community practices, and a review of the evidence of their impact on child health, growth and development was performed and disseminated to countries implementing IMCI.11

The response to IMCI was highly positive on the part of many countries and partners, including the World Bank and numerous bilateral donors. WHO and its partners put together a detailed blueprint for a three-phase roll-out in the countries that wished to adopt IMCI,6 and within less than a decade, it was being implemented in over 100 countries worldwide.12

An important innovation was the design and funding of a robust evaluation strategy that was introduced early to maximise learning. The Multi-Country Evaluation of IMCI (MCE-IMCI) set out to review at least one or two countries per WHO region which were considered likely to succeed at implementing the three IMCI components and to achieve a mortality reduction.6 13 The MCE-IMCI oversaw a series of retrospective, prospective and mixed studies that were conducted in five countries.6 13

IMCI successes and shortcomings

To illustrate what went well and what was less successful in the implementation of IMCI, we examine the strategy according to the four elements that drive the health impact of programmes on populations: the efficacy of the intervention, the quality of its delivery, the utilisation of the services and the coverage achieved.14

Efficacy of IMCI

The evidence-based approach to the diagnostic and treatment guidelines was a key feature in the development of IMCI, and the efficacy of the individual interventions to reduce mortality was clear.7 Furthermore, these guidelines, which were initially based on a combination of findings from the literature and expert opinion, have not remained static but have been regularly updated based on an increasing body of scientific evidence. Examples include the addition of zinc supplementation in the treatment of diarrhoea, care for the HIV-affected child, improved guidelines on infant and young child feeding, and care for the very young infant.

While evidence has been available to support the guidelines for the diagnosis and treatment of the individual diseases and conditions, evidence supporting the efficacy of integrated approaches at reducing mortality was not available when IMCI was developed and has remained elusive. A 2007 Cochrane review on strategies for integrating primary care at the point of service delivery, which included two studies on IMCI that compared it with routine care and were deemed to be of reasonable methodological quality, concluded that IMCI increased utilisation of services and the quality of services among those who got care, but the decline in mortality in the IMCI areas in the Tanzanian study was not statistically significant.15 Mortality data from the Bangladeshi study, not yet available in 2007, also demonstrated a lower rate than in the control areas, but this difference was again not statistically significant.16

Why the logical packaging of efficacious interventions in IMCI has not had the anticipated impact on mortality is matter of speculation, although a number of factors may have contributed. One of the factors was that the primary diseases that IMCI targeted were not always the major causes of mortality in the countries where it was implemented. Although guidance and technical assistance were available to adapt the guidelines to different country settings,6 this did not always occur, especially during the early years of rapid scale-up. The problem was particularly acute in those countries where deaths from the key infectious killers had already declined and deaths during the early neonatal period, initially not included in the guidelines, were responsible for a substantial proportion of deaths.

An additional efficacy element was that the guidelines were meant for the provision of local care for children with less severe diseases and conditions, but referral was needed for those who were more severely ill. Even when diagnosis and treatment at local level appeared to be functioning well, the referral mechanisms were often more problematic.12 15 Some of the reasons included the perception by providers that they could manage many of the severe cases, the lack of facilities to which severely ill children could be referred, financial and transport barriers, as well as lack of skilled providers, equipment and supplies at the facilities.

Quality of delivery

Improved quality of services for sick children has been the greatest success of IMCI. A number of studies, including the MCE-IMCI, showed that health worker performance, patient management and antibiotic prescribing practices had indeed improved.6 9 10 Furthermore, results from some of the detailed studies demonstrated that these improvements persist13; in the case of Tanzania, effects were seen up to 3 years after initial training,10 and that although overall costs were similar for IMCI and conventional treatment strategies, the costs per child treated correctly were sixfold lower with IMCI.17 In addition, even though most countries directed training to nurses and physicians, who were the initial target of IMCI efforts, it was possible to train lower-level community health workers to successfully follow a simplified version of the guidelines.9 In busy clinics, issues emerged about the perception of the greater amount of time that it took to carry out the IMCI approach compared with previous clinical practices,12 but this did not seem to be a critical point in most countries in its further scale-up.

At the same time, however, adherence to the guidelines was often an issue.18 This situation is amply illustrated in a detailed study in Tanzania, which demonstrated a number of problems, including failure to recognise and treat more than one illness in the same child and inadequate referral for severe illness. Although such can often be corrected with adequate supervision, the difficulties of maintaining regular and high quality supervision of service providers are numerous19: poor coordination, lack of skilled managers, competing priorities, lack of financial incentives and logistical difficulties.

Utilisation of services

One underlying assumption of IMCI was that utilisation would improve as the quality of services increased.6 This has not happened universally. Three major factors contributed to poor utilisation in the countries studied. The first was that the focus of IMCI was on government facilities, but in reality only a minority of children were taken to government facilities in some study sites. The second was the assumption that the community and family component of IMCI would encourage families to seek care at government facilities, but this component was rarely or inadequately implemented. The third was the lack of efforts to involve the private sector, which indeed provides a great deal of care even in the poorest countries. It was hoped that this sector could be successfully involved in the 2002 Bangladeshi IMCI implementation, and although efforts were made to identify and train village practitioners, many children were found to have received care from non-IMCI-trained providers.16

Coverage achieved

Even the best quality care will have little impact on mortality unless population coverage is high, and coverage remains probably the major limiting factor for the success of IMCI. Although by the end of 2009, 36 countries reported implementing IMCI in more than 75% of their districts,20 this has not necessarily translated into high population coverage. A number of factors have contributed to low coverage, including the inability to train, supervise and retain a sufficient number of care providers, the low level of implementation of the health system strengthening and family and community components and a complex set of issues related to the coordination of donors and stakeholders.6 12

While the rigour and quality of the IMCI clinical training were high, they may have inadvertently contributed to the inability to train a sufficient number of providers. The 11-day training sessions were relatively costly, and there were logistical difficulties in freeing up health staff to attend. Thus, after an initial burst of training activity during which a large number of healthcare workers were trained, budgetary and other constraints led to a decrease in the number of courses. As a result, only a fraction of the workforce in most countries received IMCI training.12 Huicho noted that in Peru, for example, only 10% of the eligible public sector workers had been trained after 7 years, and that at that rate it would take decades to produce full coverage.21 In response, attempts were made in several countries to decrease the length of training. A meta-analysis that compared the effectiveness of the standard 11-day IMCI in-service training with shortened training suggested that the standard in-service IMCI training course is more effective than short training, although the magnitude of the difference varied among countries.22

Other human resource issues also played a role in the lack of trained and well-performing staff. Staff rotation or promotions meant that facilities were often left without an IMCI-trained provider.12 Because of the cost and logistical factors mentioned above, however, training new workers was often not feasible. In addition, low salaries and high workloads contributed to low motivation and performance in many countries.

The second major factor contributing to low coverage was the lack of implementation of the health systems and the family and community components.4 6 12 The intent of IMCI was to have an integrated and coordinated approach involving care providers at multiple levels, as well as key prevention activities that would result in synergy between these elements. In most countries, these two additional components were never implemented, or implementation was weak.6 As a result, functional referral systems and supervision were often lacking, and lack of continuous drug supplies due to problems with funding, delays in ordering, and lack of timely transport, was common. Furthermore, the community component, which was meant to decrease the burden of illness as well as to make sure sick children sought care, never gained much ground. Interventions such as breastfeeding tended to fall between the cracks in the absence of programmes responsible for health promotion.12

Problems with donor and stakeholder coordination played a role in poor coverage, as reported in both the MCE-IMCI and DFID-led evaluations.4 6 There was often a lack of investment of government funds, which resulted in IMCI in some places being donor-driven and often led by ministry of health staff without the necessary rank or authority.4 In addition, difficulties in coordination among UN agencies also may have played a role. WHO and UNICEF were partners in IMCI, with WHO assuming a larger role in the clinical aspects and UNICEF serving as the leader in developing and helping implement the community level aspects, often partnering with non-governmental organisations at country level.12 However, coordination was not always optimal, with some areas of a country implementing the clinical component, while others implemented the community component. Furthermore, harmonisation with existing entities involved at country and local level in the prevention activities, including those responsible for immunisation, water and sanitation, and nutrition, was inconsistent and the delivery of consistent messaging also suffered.4 12

Where do we go from here? Repositioning IMCI

Although the initial expectations for IMCI have not been achieved, the strategy remains intellectually sound and provides an important approach to lowering child mortality. Given the renewed interest in improving health systems and in ensuring better prevention and treatment at the community level, as well as the extensive learning that has occurred through the experience of implementing the programme in 100 countries, prospects are good that IMCI can be scaled up. For this to happen successfully we would strongly recommend the following 10 actions:

  1. Concentrate effort. IMCI needs to be focused on the countries, or areas within countries, with the highest disease burden and/or highest numbers of deaths. Going smaller will allow for better adjustment of guidelines to fit country situations, and will permit greater resources for adapting guidelines as the epidemiological picture changes.

  2. Tackle the most relevant problems. IMCI has already been adapted to include the full neonatal period and some countries, such as India, have already incorporated community-delivered newborn interventions. Many other countries will need to incorporate these adaptations as deaths from infectious causes continue to decline. In addition, consideration should be given to better understanding what can be done about injury deaths, which become increasingly important as mortality continues to decline.

  3. Maintain and improve the syndromic approach and guidelines. The guidelines that form the basis of syndromic diagnosis and treatment must consider not only new diagnostic techniques and treatments but also the positive predictive value of symptoms such as fever in countries with a rapidly decreasing burden of malaria and determinants of treatment success such as antibiotic resistance.13

  4. Ensure regular updates and adaptation of guidelines to reflect local epidemiology and programmatic progress. For example, insecticide-treated bed nets for the prevention of malaria have been a game-changer in those countries that once experienced a large burden of the disease, and the predictive value of fever for malaria has changed considerably. Two other game-changing interventions on the immediate horizon include the wide-scale introduction of the rotavirus and pneumococcal vaccines, which will not only change the contributions of diarrhoea and pneumonia to mortality but will also alter treatment strategies; as bacterial pneumonia decreases, viral pneumonia will account for an increasing percentage of pneumonia cases.

  5. Make care and services available closer to home. To improve coverage of interventions, countries must be supported to develop, strengthen and sustain a system of community health workers to ensure care close to where people live, and to improve the linkages between the population and local health facilities.

  6. Strengthen health systems and supervision. Improvements in the health system that provide for a motivated and trained work force, periodic face-to-face supervision, a steady and reliable supply of drugs and diagnostics, financial access and transport for patients who need treatment and referral, and quality care at secondary level are all essential if IMCI is to have a dramatic effect on mortality. New efforts by UNICEF, WHO, the World Bank, USAID and others to improve district-level management and to address barriers to access through vouchers and other financing schemes provide hope that such improvements can be made. IMCI supervision also requires re-thinking since it is hard to maintain quality and correct problems in its absence. The Bill and Melinda Gates Foundation and others are funding efforts to explore the use of innovations such as mobile phones and text messaging to conduct at least some routine supervision and data collection, a method that has become feasible as mobile phone ownership becomes increasingly ubiquitous, even in the poorest countries. Improvements in the health system, however, will be necessary to truly fix the issue of supervision.

  7. Invest in methods to efficiently train more people in IMCI. The inclusion of IMCI principles in preservice professional training is vital. WHO developed an extensive guideline for such training in 2001,23 and has subsequently developed model chapters for textbooks and e-learning materials. However, alternate and innovative methods of in-service training, including computer-based training or distance learning, need additional investment. ICATT (IMCI Computerised Adaptation and Training Tool) is a platform that has already proved successful in numerous countries. Distance learning packages are being evaluated, as are courses of shorter duration.

  8. A constructive way must be found to involve the private sector and improve the quality and accessibility of governmental services. The involvement of the private sector can be problematic – their numbers are great and it is not always easy to identify which of these providers are caring for the greatest numbers of patients.14 Furthermore, it may be difficult to convince them of the value of spending substantial time on training and that the algorithmic approach to diagnosis and provision of a limited number of drugs can result in better outcomes. With respect to the public sector, efforts need to be made to reduce waiting times and identify incentives to health providers that will ensure patients are treated with respect and dignity.

  9. Increase coordination among various programmes in each country and harmonise donor effortsto maximise the use of resources and improve effectiveness. Valuable experience has been gained in successful coordination across programmes within ministries of health and between ministries. A careful examination of best practices from those countries with successful efforts should be compiled. The process of coordinating donors is also maturing, with the United Nations agencies, including WHO, UNICEF and the World Bank increasingly working as ‘One UN’, and the overall donor community in many countries is striving toward the coordination of its efforts.

  10. Maintain a strong approach to evaluation. It remains essential to continue learning and making modifications if aspects of IMCI are found to be functioning poorly. A staged evaluation strategy that optimises the use of resources and rapidly identifies issue and problems,13 exemplified by the evaluation in Bangladesh,14 will be vital to improving programs and maximising the impact of the strategy.

Conclusions

IMCI is a technically sound, internationally recognised strategy that remains central to the child health programming of more than 100 countries. We are confident that the 10 actions proposed in this paper, many of which are already being undertaken, will help countries and partners use IMCI in the way it was intended, and increase its contribution towards reaching Millennium Development Goal 4 and beyond.

Acknowledgments

The authors wish to thank Jose Martines and Rajiv Bahl for their thoughtful review and comments on this manuscript.

References

Supplementary materials

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

    Files in this Data Supplement:

    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
    • Web Only Data - This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

Footnotes

  • Competing interests None.

  • Provenance and peer review Commissioned; externally peer reviewed.