Health system factors affecting implementation of integrated management of childhood illness (IMCI): qualitative insights from a South African province

Health system factors affecting implementation of integrated management of childhood illness... Abstract The Integrated Management of Childhood Illness (IMCI) strategy has been adopted by 102 countries including South Africa, as the preferred primary health care (PHC) delivery strategy for sick children under 5 years. Despite substantial investment to support IMCI in South Africa, its delivery remains sub-optimal, with varied implementation in different settings. There is scarce research globally, and in the local context, examining the effects of health system characteristics on IMCI implementation. This study explored key determinants of IMCI delivery in a South African province, with a specific focus on health system building blocks using a health system dynamics framework. In-depth interviews were conducted with 38 districts, provincial and national respondents involved with IMCI co-ordination and delivery, exploring their involvement in, and perceptions of, IMCI strategy implementation. Identified barriers included poor definition of elements of a service package for children and how IMCI aligned with this, incompetence of trained nurses exacerbated by inappropriate rotation practices, use of inappropriate indicators to track progress, multiple cadres coordinating similar activities with poor role delineation, and fragmented, vertical governance of programmes included within IMCI, such as immunization. Enabling practices in one district included the use of standardized child health records incorporating IMCI activities and stringent practice monitoring through record audits. Using IMCI as a case study, our work highlights critical health system deficiencies affecting service delivery for young children which need to be resolved to reposition IMCI within the broader child ‘survive, thrive and transform’ agenda. Recommendations for appropriate health system strengthening include the need for redefining IMCI within a broader PHC service package for children, prioritizing post-training supervision and mentoring of practitioners through appropriate duty allocation and rotation policies, strengthening IMCI monitoring with a specific focus on quality of care and building stronger clinical governance through workforce allocation, role delineation and improved accountability. IMCI implementation, child health, health systems, primary health care, South Africa Key Messages Failure to articulate and implement a preferred strategy for delivery of primary health care services for children led to idiosyncratic and often poor service provision. IMCI implementation was inadequately monitored, supported, co-ordinated and governed. Health professional training was prioritized, but insufficient subsequent supervision and mentoring, and limited adaptation of the work environment and routine allocation, compromised quality service delivery. Failure to collect and/or to meaningfully synthesize IMCI data, particularly related to the quality of its delivery, led to limited appreciation of the strategy’s successes or failures, and non-response to these. Ill-defined and overlapping responsibilities for child health managers, and higher priority for parallel child health programmes, affected IMCI coordination and implementation. Introduction The Integrated Management of Childhood Illness (IMCI) strategy has proven to be a cost-effective intervention to improve the quality of primary health care (PHC) service delivery in resource-limited settings. Its main objective is to reduce child mortality and morbidity by both preventive and promotive health care (Ahmed et al. 2010; Arifeen et al. 2004; WHO 2009; Tulloch 1999; Rakha et al. 2013). It targets improvements in the case management skills of health-care staff, in overall health systems and in community and family health practices. Within health facilities, the IMCI strategy seeks to promote the accurate identification of childhood illnesses with a comprehensive treatment response, strengthen the counselling of caretakers and speed up the referral of severely ill children (WHO 1999). Many countries have struggled to scale up IMCI and achieve high coverage since its introduction over two decades ago (Bryce et al. 2005; WHO 2004; Ahmed et al.2010). Effective IMCI implementation has been most challenging, unsurprisingly, in countries with weaker health systems (Armstrong et al. 2004). Currently, there is global discourse about the continued survival of IMCI as an integrated strategy amidst various vertical programmes. IMCI was specifically introduced to minimize vertical approaches to child health delivery (such as immunization or diarrhoeal disease control programmes) (WHO 1999). However, it has struggled to fulfil its potential in under-resourced health systems where vertical programmes continue to dominate (Ahmed et al. 2010). Recently, IMCI has elicited renewed interest at a global level, including a World Health Organization (WHO) strategic review re-examining IMCI in light of emerging priorities and new developments in the global child health agenda of ‘survive, thrive and transform’ (Costello and Dalglish 2016; United Nations 2015). A 2016 systematic review evaluating the effects of implementation of IMCI strategy on critical child health outcomes concluded that researchers should continue to explore how the IMCI strategy can best be delivered (Gera et al. 2016). South Africa adopted IMCI in 1998 and initial evaluations showed promising results (Chopra et al. 2005). By 2011, most PHC facilities were implementing the strategy, with two-thirds adequately saturated with IMCI trained health workers (Loening and Bamford 2012). However, even after two decades of sustained investment in training and infrastructure, IMCI implementation is far from optimal (Horwood et al. 2009a; 2009b; 2009c; Jonker and Stellenberg 2014; Mulaudzi 2015; Mazaleni 2009; District Department of Health 2012; National Department of Health 2016). Moreover, there is no evidence to demonstrate the impact of IMCI implementation on child mortality and morbidity outcomes in South Africa (Thandrayen 2010; Saloojee and Bamford 2006). Furthermore, IMCI evaluation data is either unavailable or not reliable. In South Africa, delivery of child PHC services is considered to be idiosyncratic, depending on the whims of individual practitioners and facilities, with IMCI regarded as an optional rather than prescribed practice in many districts (National Department of Health 2016; Thandrayen 2010). Documented reasons for poor IMCI implementation in South African settings include deficient practitioner competency, IMCI consultations being too time-consuming and poor clinical supervision of practitioners. This is consistent with reports from other countries (Bryce et al. 2005; Huicho et al. 2005; Kiplagat et al. 2014; Lange et al. 2014). However, there is limited description of health system’s influence, as a whole, on implementation of IMCI at national and indeed global level. Internationally, there is a lack of consensus on IMCI’s role within the delivery of an overall PHC service package for children and how it could best be integrated within broader health system strategies in future. This demands a relook at IMCI’s position in the larger scheme—particularly the future delivery of child ‘survival and thrive’ strategies and the associated health system strengthening required. Hence, the motivation for this study was two-fold—to analyse the failure of IMCI to deliver expected outcomes in South Africa through an in-depth analysis of health system factors affecting its implementation, and to contribute to the global discussion on the ongoing implementation and re-design of IMCI. Moreover, the objective was not only to identify barriers and enablers, but also to better understand the root causes for their existence in a local context. Although the study was based in a middle-income setting, the concerns identified and responses required are likely to be as relevant to all IMCI implementing settings. Methods Analytical/thematic framework for the study Health systems are complex and dynamic in the way they function (WHO 2007; De Savigny and Adam 2009). The WHO’s six building blocks of a health system are well known (WHO 2007). Van Olmen et al. proposed a health systems dynamics framework derived from the WHO model accentuating the dynamic relationships existing between the building blocks (Van Olmen et al. 2012). We utilized a simplified version of this framework to map and explain how a health system operates in the context of IMCI and factors associated with each building block affecting its implementation. According to our framework (Figure 1), the implementation of IMCI package at a facility level is considered synonymous to ‘service delivery’ as a process. The other building blocks, viz. human resources, medical technologies and health information system act as inputs which influence the process of service delivery. Leadership and governance serve as an overarching block affecting both the inputs and process. We assume that these building blocks, if properly operationalized, lead to improved implementation of IMCI and better child health outcomes. We focussed only on inputs and process which together constitute five building blocks. To stay focussed on our research question, we excluded study of IMCI outcomes. We also excluded examination of the sixth WHO building block of health financing since neither child PHC nor IMCI receive dedicated (ring-fenced) funding in South Africa. Their costs are immersed within a PHC budget. In addition, study participants had a limited ability to influence finances. Figure 1. View largeDownload slide Health systems dynamic framework for IMCI Figure 1. View largeDownload slide Health systems dynamic framework for IMCI Study setting The South African health system consists of one national, nine provincial and 52 district health structures. Districts are further divided into sub-districts and municipal wards as administrative units. Within each district, clinics and community health centres (CHCs) are the key delivery points for PHC services, including IMCI. Districts are responsible for operationalizing the IMCI strategy. Child PHC services are primarily delivered by professional nurses who receive pre-service and/or in-service training in IMCI. Additionally, some professional nurses undertake a 1-year primary clinical care diploma (PCCD) offered by the Nursing Council of South Africa. This supplementary training is meant to augment nurses’ ability to manage adults and sick children at a PHC level (Soweto trust for nurse clinical training 2012). It promotes traditional consultation with focus on chief complaint and use of equipment such as stethoscope and otoscope. The household and community component of IMCI are meant to be delivered though ward-based outreach teams (WBOTs) including community health workers who receive training on key family practices, but there has been variable implementation of this component (National Department of Health 2016; Mazaleni 2009). Study design and participant selection A qualitative, descriptive study was conducted from March 2015 to August 2016 in Gauteng province, South Africa. This is the wealthiest and most populous province in the country, and predominantly urban (Makgetla and Fotoyi 2016). Two of five districts in the province were purposively selected—subsequently referred to as district A (peri-urban/rural) and B (urban). We collected district IMCI reports with a list of all PHC facilities based on their IMCI implementation performance status (categorized as ‘good’ or ‘poor’ as reported by district child health coordinators). Next, we randomly selected two good and two poorly performing facilities, ensuring a selection from different sub-districts. Thus, the district sample consisted of four facilities including a CHC from each district. Only those facilities saturated with trained staff, i.e. a minimum of 60% of staff trained in IMCI (WHO 1999), were eligible for inclusion to minimize possible differences resulting from training status. To explore the status of IMCI implementation across all levels of health system, a range of participants were purposively selected including professional nurses (2–3 from each clinic depending on availability), PHC facility managers, and key coordinators and supervisors responsible for child health services at a sub-district, district, provincial and national level. All invited participants consented to involvement. Table 1 provides detail on participants and their designations. Table 1. Details of study participants   Level of health system where participant located  Participant’s designation  Number of participants  District A  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  District  District child health coordinator  1  District clinical specialist team (DCST) paediatrician  1  District B  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  Maternal Child and Women’s Health (MCWH) coordinatora  2  Districtb  District clinical specialist team (DCST) paediatrician  1  District clinical specialist team (DCST) paediatric nurse  1  Province    c  1  National    c  1  Total      38    Level of health system where participant located  Participant’s designation  Number of participants  District A  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  District  District child health coordinator  1  District clinical specialist team (DCST) paediatrician  1  District B  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  Maternal Child and Women’s Health (MCWH) coordinatora  2  Districtb  District clinical specialist team (DCST) paediatrician  1  District clinical specialist team (DCST) paediatric nurse  1  Province    c  1  National    c  1  Total      38  a District B had sub district MCWH coordinators who were not present in district A. Each such coordinator had 5-8 clinics under their supervision. Two coordinators responsible for the sampled clinics were interviewed. b MCWH coordinator of district B was not interviewed because she recently joined the post and did not have much knowledge about IMCI. The coordinator, whom she replaced, was posted to one of the sub-districts as an MCWH coordinator and was interviewed instead. c Designations of key informants at province and national level have been omitted to ensure anonymity. Data collection and analysis Thirty-eight in-depth, semi-structured interviews were conducted in English. Interviews were recorded via audiotaping (where consent was provided) and transcribed; otherwise detailed hand written notes were taken and typed. We designed an interview guide based on study objectives, with open ended questions (Table 2). It included questions on current status of IMCI implementation and the challenges faced specifically structured around health system building blocks. We also reviewed documents related to IMCI activities at various levels where available and permitted. These included clinic registers, child health visit records, supervisors’ reports and checklists, and provincial and national reports on IMCI. The document review was undertaken to supplement and validate data obtained from interviews on health information systems, explore the status of IMCI monitoring, and determine the use of indicators to track implementation progress. Table 2. Interview guide Describe how you provide services for children less than 5 years in this clinic.  What guidelines do you follow to manage children less than 5 years/how do you manage sick children at this facility?  What do you know about IMCI, what are your perceptions/opinions about the IMCI strategy?  How is IMCI governed/managed at various levels? (coordinators, directorates)  How do you monitor IMCI/under 5 child health services in this clinic/sub-district/district?  Do you receive/provide any supervision or mentoring support for IMCI? If yes, describe.  What challenges do you face, if any, in offering/implementing IMCI at your clinic/sub-district/district/province? (Human resources, equipment, supplies, monitoring, supervision, others)  Describe how you provide services for children less than 5 years in this clinic.  What guidelines do you follow to manage children less than 5 years/how do you manage sick children at this facility?  What do you know about IMCI, what are your perceptions/opinions about the IMCI strategy?  How is IMCI governed/managed at various levels? (coordinators, directorates)  How do you monitor IMCI/under 5 child health services in this clinic/sub-district/district?  Do you receive/provide any supervision or mentoring support for IMCI? If yes, describe.  What challenges do you face, if any, in offering/implementing IMCI at your clinic/sub-district/district/province? (Human resources, equipment, supplies, monitoring, supervision, others)  Since the study did not involve any observational element, to elicit information on clinicians’ IMCI related practices and challenges, we probed using questions such as ‘how do you manage a sick child with diarrhoea or pneumonia’ and compared their responses to our benchmark, thereby interpreting good/poor IMCI adherence. In addition, a review of child health records where sick child consultations were captured, supplemented our interpretation. Transcripts were coded and analysed using the MaxQDA data analysis software (VERBI Software GmbH, Berlin)(MaxQDA 2016). Coded data were analysed using the framework analysis approach (Srivastava and Thomson 2009). The health system dynamic framework (Figure 1) acted as our thematic framework and formed basis of the analysis in which five building blocks were used as deductive themes. First, all text segments from transcripts, documents and written notes were coded to classify the data and identify emerging issues and concepts from interviews. Codes were predominantly inductively derived, and grouped within a framework building block, e.g. ‘poor IMCI adherence’ emerged as a code which was grouped under the theme of ‘service delivery’. A few deductive codes based on the literature of factors affecting IMCI were also added. Data analysis was initiated after conducting the first interview and concurrent analysis allowed us to revise questions to improve the richness and quality of information obtained. Benchmarks used for data analysis To objectively analyse the factors affecting IMCI implementation, we established benchmarks for actual performance. These were based on official guidelines (from WHO and South African national health department) and logical constructs. For service delivery, we used the South African IMCI guidelines (National Department of Health 2014) to assess the extent to which nurses followed or deviated from the IMCI package. For human resources, we set three benchmarks: (i) there should be adequate number of IMCI trained nurses in a clinic (> 60% saturation), (ii) only an IMCI trained nurse should be assigned to treat sick children, and (iii) IMCI trained nurses must retain their competence in delivering IMCI. For health information systems, the benchmark for ideal monitoring of IMCI was that activities related to health system elements were being regularly reviewed and improved. This had to be done by periodic collection of relevant data and include adequate feedback to implementers. For leadership and governance, three benchmarks were set: (i) policy guidance should be unambiguous and not in conflict with other existing guidelines, (ii) system design—new programmes (vertical or integrated) should not inhibit the effective functioning of existing strategies, and (iii) accountability—for effective implementation of IMCI, workforce planning ensured adequate staff numbers and assignment of responsibilities to appropriate staff with no overlap or omission of activities. Ethical clearance for the study was obtained from the authors’ institute. All participants signed informed consent forms for both interview participation and audio taping. Institutional approval for individual districts and clinics were also obtained from district and provincial research committees prior to study commencement. For the sake of retaining anonymity and confidentiality, we have not provided actual names of clinics or districts. Results Results are presented as per health system building blocks based on our thematic framework (Table 3). For each building block, an overview of the existing status is followed by key challenges reported by participants and how these affected IMCI implementation. Although our sample included a mix of good and poorly performing clinics, the challenges we identified existed across both categories. While most findings were directly reported by participants, some were derived indirectly from our synthesis of participant responses based on comparison with the established benchmarks. Table 3. Summary of key health system barriers and enablers affecting IMCI implementation in South Africa Key barriers obstructing IMCI implementation  Key enablers promoting IMCI implementation  Service delivery    Perceptions of primary clinical care trained nurses about IMCI  Use of standardized child health files with IMCI recording forms (compulsorily filled by nurses at every sick child consultation)  Lack of policy guidelines on what constitutes a primary care package  Long queues, high patient waiting times  Human resources    Frequent temporary staff shortages  High saturation of trained professional nurses at facilities  Poor allocation/rotation practices leading to IMCI trained nurses not being sufficiently competent  Medical technology    Stock outs of growth monitoring equipment, vitamin A and deworming drugs  Availability of a national Essential Drug List  Centralized drug dispensing system  Health information system    Non-systematic capturing of case management adherence  Internal and external record audits conducted by facility managers and Maternal Child and Women’s Health (MCWH) coordinators, respectively  Quality of care data not being utilized for decision making and monitoring of IMCI implementation  Leadership and governance    Poor supervision and mentoring of nurses due to shortage of Maternal Child and Women’s Health (MCWH) coordinators  Improved supervision and monitoring owing to availability of Maternal Child and Women’s Health coordinators at a sub-district level (5–10 clinics per coordinators)  Fragmented governance of vertical child health programmes (e.g. Expanded programme on Immunization) that are part of IMCI  Lack of clarity on roles and responsibilities amongst various stakeholders on IMCI related activities leading to duplication or omission of tasks  Poor accountability for child health outcomes  Key barriers obstructing IMCI implementation  Key enablers promoting IMCI implementation  Service delivery    Perceptions of primary clinical care trained nurses about IMCI  Use of standardized child health files with IMCI recording forms (compulsorily filled by nurses at every sick child consultation)  Lack of policy guidelines on what constitutes a primary care package  Long queues, high patient waiting times  Human resources    Frequent temporary staff shortages  High saturation of trained professional nurses at facilities  Poor allocation/rotation practices leading to IMCI trained nurses not being sufficiently competent  Medical technology    Stock outs of growth monitoring equipment, vitamin A and deworming drugs  Availability of a national Essential Drug List  Centralized drug dispensing system  Health information system    Non-systematic capturing of case management adherence  Internal and external record audits conducted by facility managers and Maternal Child and Women’s Health (MCWH) coordinators, respectively  Quality of care data not being utilized for decision making and monitoring of IMCI implementation  Leadership and governance    Poor supervision and mentoring of nurses due to shortage of Maternal Child and Women’s Health (MCWH) coordinators  Improved supervision and monitoring owing to availability of Maternal Child and Women’s Health coordinators at a sub-district level (5–10 clinics per coordinators)  Fragmented governance of vertical child health programmes (e.g. Expanded programme on Immunization) that are part of IMCI  Lack of clarity on roles and responsibilities amongst various stakeholders on IMCI related activities leading to duplication or omission of tasks  Poor accountability for child health outcomes  Service delivery Issues highlighted in this section include poor adherence with the IMCI strategy, omission of aspects of the consultation, lack of clarity about what constituted an IMCI consultation, and poor recording of the consultation. Poor IMCI adherence Despite claims about practicing IMCI, we identified deviations from our IMCI adherence benchmark at most facilities, especially in district A. Nurse practitioners preferred selectively focusing on the child’s main complaint and symptom-based components of IMCI consultation. They tended to make a single diagnosis instead of completing the entire process (algorithm) and providing treatment accordingly. Deviation was more frequently made by PCCD trained nurses who perceived the IMCI approach to be constraining and simplistic. The simple one is IMCI but the one that makes you feel like you have treated the patient is PCCD because you will be able to diagnose the actual thing. So with PCCD you go deep. With IMCI, you just do superficial (Nurse, participant 5, clinic 4, district A).Another explanation offered for non-adherence with IMCI was that IMCI consultations were more time consuming (responses ranged from 20 min to an hour for a single consultation) and extended patient waiting times. Hence, to compensate, activities such as nutritional status classification, immunization, feeding assessment and counselling of the caregiver were curtailed. There was a general under-appreciation of the need to complete all aspects of the IMCI consultation (particularly preventive and promotive aspects) and its associated benefits. The problem is when there are a lot of people in the clinic, there is no time for you to go through all those things for the child. You will be working fast fast fast to push the queue. Sometimes, sisters (nurses) will just be looking at conditions that the child has been brought with and then they leave the rest and you find that the child has not received immunizations and stuff like that (Nurse, participant 14, clinic 2, district B).Moreover, there was uncertainty among both nurses and coordinators as to the required level of engagement with the HIV, TB and malnutrition modules within IMCI (screening or classifying, and treatment) resulting in poor compliance with these sections. Many practitioner perspectives were contradicted by managers and programme coordinators. They disputed that long consultation times were inevitable, and believed these could be optimized (reduced) if IMCI was practiced regularly. They argued that PCCD trained nurses should supplement the IMCI approach rather than replace it. Presenting complaints not specifically covered by IMCI could be addressed under the section ‘assess other conditions’. Moreover, they also dismissed staff shortages and long queues as valid reasons for non-adherence to IMCI contending that these were excuses that could be overcome by better planning. We are saying that even though you have done PCCD training, whenever we manage children under 5 years, we have to use the IMCI strategy. We are not saying that don’t use your clinical skills from PCCD training. But we are saying, please, IMCI is the protocol that we are supposed to follow (Provincial key informant, participant 37). Recording of consultations Nurses’ understanding and adherence with IMCI principles was better in district B. This could be partly attributed to the compulsory requirement for a standardized IMCI recording form to be filled during every sick child encounter. The form promoted compliance with the entire IMCI clinical process, including completion of classifications, appropriate management and attention to promotive actions. In contrast, nurses in district A relied on free hand clinical notes captured on blank pages. These contained significantly less detail of the examination, diagnostic considerations and management plan. Because of the form, sisters are compelled to use IMCI, because if the child is sick, we don’t allow them to use extra filing papers. We need your classifications to be written on IMCI recording forms (Sub-district MCWH coordinator, participant 22, District B).A common misperception was that use of the prescribed form prevented the identification, recording and management of non-IMCI classifications. At the end of the day there are diagnoses that you cannot write in this IMCI form because they are not part of IMCI classification. If it is not included, then you refer. You cannot come up with that diagnosis as a sister (nurse) in that IMCI room (Nurse, participant 10, clinic 1, district B). Human resources for IMCI Themes that emerged were insufficient staff numbers, inappropriate staff allocation and rotation, and difficulties in maintaining competency. Adequate number and workload allocation of IMCI trained nurses Most facility respondents reported an inadequate staff complement for their patient load. However, this was a subjective judgement as they were unaware of what constituted an appropriate ratio and if such a standard existed. There are existing national norms on the number of nurses required for overall PHC clinical services according to patient load (Daviaud and Subedar, 2012). Additionally, the 2001 national PHC norms and standards state that at least one staff member in a clinic should take overall responsibility for managing children (National Department of Health 2001). However, neither of these norms offer any clarity on specific staff numbers and workload allocations required for child health generally and IMCI specifically. Our investigation revealed that due to the lack of clear guidelines, staff shortage was more of a perception than a reality. We observed that nurses (especially in district A) were overwhelmed with multiple responsibilities, e.g. the nurse allocated to IMCI services was also acting as a team leader of community health workers. Unplanned staff absenteeism, training commitments and nurses deputed for outreach service delivery contributed to temporary staff shortages, but since these were a regular occurrence, it led to the perceived notion of perpetual staff shortage. Clinics are busy places. We have done this exercise in quite a few clinics trying to figure out staff shortage because that is the first complaint that we get from everyone, but if you work out how many patients per day are being seen, then majority of clinics have enough staff (District clinical specialist team member, participant 35, district B).We also noted that nurses at CHCs in both districts did not practice IMCI after-hours (4 pm to 7 am) since only two nurses manned CHC during this period. Our findings highlight the need for a more suitable approach to manage IMCI specific staffing needs. Countries such as India and Tanzania have successfully utilized the WHO recommended Workload Indicators and Staffing Needs (WISN) tool to compute the surplus/gap in staffing levels needed to fulfil unmet needs for MCH services (Hagopian et al 2012; Nyamtema et al 2008; WHO 2010). Provincial and national respondents in our study revealed that districts and clinics were in the process of assessing staff workload and allocation using the WISN tool but no response has been implemented to any findings to date (October 2017). Poor competence of IMCI trained nurses To remain competent, IMCI trained practitioners require regular practice. Allocation of IMCI trained nurses to other tasks owing to poor rotation practices was common at all facilities. Respondents in both districts complained about not being allocated to managing sick children immediately after returning from IMCI training, and on a regular basis, which resulted in poor competence. The reported frequency of rotation through an IMCI duty station ranged from three months to two years which was mostly determined by the facility manager based on her/his own criteria. The question of not rotating is the main point, people stick in one service for a long time. We are not given the opportunity to practice what we have learnt. You forget. When you have to face a child then you have to start all over again or ask others because you haven’t been in touch (Nurse, participant 11, clinic 1, district B).This also made post training supervisory support difficult for supervisors since they did not find trained nurses at IMCI duty stations. Moreover, refresher IMCI trainings in both districts were infrequent. When I came back from IMCI training, I had to go to curative (adults). I think it had an impact on my IMCI practice because to be honest, there was never a chance for me to see a sick child (Nurse, participant 16, clinic 3, district B). Medical technology, supplies, drugs and equipment This was not reported as a significant barrier to IMCI implementation, although temporary stock-outs were reported from facilities in both districts. Non-availability of equipment such as mid upper arm circumference tapes and length mats/boards reportedly impacted growth monitoring and malnutrition assessment activities. Respondents identified the drug dispensing system as a barrier to effective IMCI implementation. Owing to a centralized pharmaceutical dispensing system, drugs were stored and dispensed only from the main pharmacy by a pharmacist or pharmacy assistant. Practitioners believed that this compromised counselling of caregivers on usage and storage of medication as recommended by IMCI guidelines. Also, administration of the first dose was not demonstrated by the practitioner potentially affecting caregiver compliance. When we were training for IMCI, we had everything in the cupboard in rooms, you would even mix the drug for mother and give it to the baby. Now they say you write the prescription and the mother has to collect from clinic’s pharmacy. They (pharmacists) are not going to explain to mothers the way we do (Nurse, participant 17, clinic 4, district B). Health information system Issues that were raised included poor monitoring of adequacy of IMCI case management and overall implementation. Monitoring of IMCI case management adherence IMCI consults were being captured more systematically in child health records in district B because of the use of compulsory IMCI forms. I used to photocopy and fill the IMCI forms for each child. But no one came to look for these IMCI forms, so I stopped filling them. Now when I do IMCI, I won’t fill out a form, I will just say the 4 things in my head, fever, cough blah blah. Then I start to follow PCCD approach and give treatment (Nurse, participant 6, clinic 3, District A).Forms also facilitated better monitoring and auditing of IMCI consultations which in turn led to some improvement in nurses’ adherence to IMCI. You do get into trouble when people come and check you or do an audit because they are going to ask you why you are not doing it correctly. You try to give them your reasons but they don’t take it so we have to do the correct classifications (Nurse, participant 16, clinic 3, district B). Our sisters (nurses) do adhere to IMCI guidelines. I know it because I do audit checks of files. We randomly take 5 or 10 files of children. You have to go and do rounds to support and supervise these sisters and then audit their files(Facility manager, participant 20, clinic 4, district B). Monitoring of overall IMCI implementation We identified three datasets with indicators used to monitor overall IMCI implementation (Table 4, Figure 2). Only one of these assessed the quality of care offered (supervisor’s report). The district health information system captured proxy indicators that were too distal to interpret IMCI performance, while the facility-based quarterly IMCI report only offered data on process indicators. The quality of care data (only available in district B) was collated at sub-district level but not shared with the district. The other two datasets were forwarded to district, provincial and national levels, but never resulted in any facility feedback or action. Table 4. Current indicators and data sources used to monitor IMCI implementation in South Africa Data source  District Health Information System (DHIS)  IMCI quarterly report  Supervisors’ report  Indicators included in the source  Children < 5 years pneumonia (new cases)  Is this facility implementing IMCI during this quarter? (yes/no)  Adherence to case management based on observation of IMCI consults and record audits in clinics (as per WHO checklist for IMCI supervision)  Children < 5 years diarrhoea with dehydration (new cases)  % saturation of health workers trained in IMCI available in a facility during this quarter  Children < 5 years severe acute malnutrition (new cases)  Number of professional health workers in PHC facilities (clinics and CHCs) during this quarter  Children < 2 years underweight (new cases) (between −2 and −3 SD)  Number of nurses trained in IMCI case management available at PHC facilities during this quarter  Children < 5 years on food supplementation  Total number of supervisors trained and active during this quarter  Infant exclusively breastfed at hepatitis-B third dose  Total number of IMCI facilitators  available and active during this quarter  (age—4 weeks)  Data source  District Health Information System (DHIS)  IMCI quarterly report  Supervisors’ report  Indicators included in the source  Children < 5 years pneumonia (new cases)  Is this facility implementing IMCI during this quarter? (yes/no)  Adherence to case management based on observation of IMCI consults and record audits in clinics (as per WHO checklist for IMCI supervision)  Children < 5 years diarrhoea with dehydration (new cases)  % saturation of health workers trained in IMCI available in a facility during this quarter  Children < 5 years severe acute malnutrition (new cases)  Number of professional health workers in PHC facilities (clinics and CHCs) during this quarter  Children < 2 years underweight (new cases) (between −2 and −3 SD)  Number of nurses trained in IMCI case management available at PHC facilities during this quarter  Children < 5 years on food supplementation  Total number of supervisors trained and active during this quarter  Infant exclusively breastfed at hepatitis-B third dose  Total number of IMCI facilitators  available and active during this quarter  (age—4 weeks)  Figure 2. View largeDownload slide Process of monitoring IMCI implementation in South Africa at various levels of health system (this figure must be interpreted in combination with Table 4) Figure 2. View largeDownload slide Process of monitoring IMCI implementation in South Africa at various levels of health system (this figure must be interpreted in combination with Table 4) I think the IMCI report that I do, does not talk much of IMCI implementation because it is mainly about saturation of training for IMCI. They are only interested in figures, not implementation (Child health coordinator, participant 17, district A). WHO recommends that a health facility survey providing detailed information on quality of care and health workers’ adherence to IMCI be conducted once every 5 years at national and once every two years at provincial level (WHO 1999). We established that no such survey was conducted at either level since 2004. Leadership and governance Leadership and governance issues were identified as critical components of implementation failure. Themes that emerged were fragmented/vertical governance structures and lack of accountability among key stakeholders. Policy guidance IMCI was formally adopted as the strategy for delivering PHC services for children in South Africa in 1998 and forms part of the national norms and standards for primary health care (National Department of Health 2001). The strategy was also incorporated into the national maternal and child health strategic plan, as a key component of PHC service package for children (National Department of Health 2012). However, guidance on how to effectively operationalize IMCI implementation has always been lacking. As example, there are no specific guidelines on an optimal patient-staff ratio, on the allocation/rotation of nurses to IMCI duty station or on supervision (who should do it, how frequently, what should be done, using which tools?) System design We identified fragmentation in the governance structures for IMCI and other vertical programmes which were meant to be integrated into IMCI such as Expanded Programme on Immunization (EPI), HIV and tuberculosis (TB). For example, at a national, provincial and district level, while IMCI was governed by the maternal, child, women’s health and nutrition (MCWHN) directorate, the EPI programme was governed by communicable disease control directorate. As vertical programmes were prioritized, dedicated coordinators/supervisors for EPI, HIV and TB regularly visited facilities and offered supervision for programme specific components, while ignoring the IMCI aspects of those programmes. The EPI manager will push only her things that side and then even me this side in MCWH directorate. I think the main problem is because the EPI program manager is not IMCI trained (Provincial key informant, participant 37). Accountability of various stakeholders in IMCI implementation Districts had different management cadres assigned to perform similar IMCI clinical governance activities. For example, in district A, all monitoring and supervision of IMCI was performed by a district child health coordinator while in district B there were MCWH coordinators at sub-district level. As a consequence, the coordinator in district A supervised >40 clinics, while in district B, each sub-district coordinator was responsible for 8–10 clinics. IMCI related roles and responsibilities of different cadres of managers were not clearly defined with task performance left to individual discretion. For example, the same clinic would receive an IMCI supervisory visit by both a district clinical specialist team (DCST) member and MCWH coordinator on different days, using different checklists. On the other hand, another clinic in the same district would not receive any visits. This discordance was aggravated by poor communication between management cadres. We are not on the same page because they themselves (DCSTs and area managers) are interested in figures. It’s like the coordination between them and me is not that good so I think we need to sit down because they are doing their own thing, I’m doing my own thing. We are basically doing the same things (Child health coordinator, participant 17, District A). There is no post visit discussion between us (area manager and MCWH coordinator) because if you discuss it, you are offended, you are in my territory and it is not your business. Yet at the end of the day it is the PHC component (Area manager, participant 31, sub-district 2, district B). Discussion Considerable differences were observed between the two districts. District B offered superior IMCI implementation. Key enablers (in district B) included standardized recording forms, dedicated MCWH coordinators at sub-district level, regular, stringent monitoring and better support for staff. District A’s performance was hindered by negative perceptions of IMCI held by PCCD trained nurses, reduced competence due to unnecessary rotation, poor monitoring and supervision, fragmented governance and lack of accountability for programme outcomes. Many of the deficiencies identified mirror those identified globally in previous studies (Kiplagat et al. 2014; Huicho et al. 2005; Ahmed et al. 2010). This study, however, offers more detail about the root causes for these deficiencies. We recognized at the outset that confining the analysis to issues specific to IMCI without considering the broader health system deficiencies would restrict our ability to design a meaningful response. The absence of a coherent strategy to manage sick children was overt. Nurses felt obliged to select between IMCI and the PCCD approaches and lacked support to adequately implement key components such as HIV, TB and malnutrition. Negative perceptions of IMCI were fuelled by perceived, but non-validated, deficiencies such as staff shortages. Similar findings have been reported in Pakistan where district and provincial managers reported lack of appreciation of the IMCI strategy constraining implementation (Pradhan et al. 2013). The criticism that IMCI is too simplistic was shared by practitioners in Tanzania and Kenya (Kiplagat et al. 2014; Silali 2014). Implementers have placed substantial emphasis globally on IMCI training believing that this is the critical factor driving guideline adherence. This belief was reflected in our study in the IMCI quarterly report prioritizing submission of training related data. Our study confirms the well-accepted notion that training alone cannot ensure quality care and does not guarantee satisfactory implementation (Pariyo et al. 2005). Failure to define staffing needs for child health (and IMCI) delivery and poor workload allocation emerged as major barriers to IMCI implementation. Previous studies also confirmed issues such as staff shortages, rotation and frequent turnover as barriers to IMCI implementation (Huicho et al. 2005; Prosper et al. 2009; Tulloch 1999; Adam et al. 2005). Contrary to evidence from other countries, (Pariyo et al. 2005; Ahmed et al. 2010) we did not find the availability of medical technology, supplies, drugs and equipment as a major barrier to implementation in both districts. Global evidence indicates that quality of care is a key predictor of child mortality (Binkin et al. 2011; Rohde et al. 2008). It is well established that adherence to IMCI guidelines improves the quality of clinical care provided to sick children by changing case management practices (Armstrong et al. 2004; Chopra et al. 2005). Monitoring the quality of care should thus be prominent in any child health information system. Although WHO has set guidelines and standard indicators on monitoring IMCI, it does not provide details of how IMCI implementation (including quality of care) should be monitored and how results can be used for improving performance (WHO 1999). Hence, it is left to countries to decide how they will collect and utilize monitoring information. Our study highlights critical gaps in this process in South Africa where we identified that programmatic data collected at clinic and sub-district levels was not shared and utilized at higher levels (district, province and national) for strategic planning, decision making and improving implementation. This was best reflected in words of a national key informant stating ‘we don’t know what is happening with IMCI in this country, we are just speculating and assuming.’ The recently conducted IMNCI global strategic review echoes similar findings where only 33% of all IMCI implementing countries had a comprehensive monitoring and evaluation plan (Costello and Dalglish 2016). The benefits of an integrated approach for managing children are well-established (Tulloch 1999). Although IMCI was introduced as a response to uncoordinated vertical health programmes, it struggles to effectively coexist with them. Global studies have highlighted that vertical programmes are often prioritized and attract ring-fenced funding, dedicated governance cadres and monitoring systems (Ahmed et al. 2010). Evidence from Tanzania shows that well established and heavily funded vertical programmes such as EPI and malaria, considered IMCI as a threat to their existence (Prosper et al. 2009). We confirmed programme siloed governance at district, province and national levels. Dedicated managers and coordinators and clear monitoring and evaluation indicators existed for vertical programmes such as EPI and TB, while IMCI was ignored. The issue of co-existence of integrated and vertical approaches has been an ongoing international debate with few tangible solutions. Policy makers need to consider mechanisms at both strategic and operational (service delivery) levels to improve the links (Atun et al. 2008). While there are global studies that have explored integration in HIV, mental health and communicable diseases, there is a lack of evidence on integration in context of child health (Atun et al. 2008). Since IMCI is not a vertical programme, it is more likely to have management cadres (coordinators, supervisors, managers) with shared responsibilities. Each plays a critical role in ensuring that implementation activities such as monitoring, supervision, health facility visits, training etc. are conducted appropriately. More efficient execution of these activities in district B could mainly be attributed to better workforce planning and effective allocation of duties. The asymmetry between the two districts can be ascribed to lack of clear, consistent guidelines from national and provincial levels. Poor leadership and accountability at various levels of health system have previously been highlighted in the context of child health in South Africa—mainly in the failure of responsibility delegation, budget allocation and guideline setting (Saloojee 2009). In addition, there is no standardized organizational structure for district health services in South Africa, which adversely affects the capacity of health management cadres (Gilson et al. 2011). Our study adds to the global knowledge base on effect of poor governance and lack of a clear administrative structure at a district level on systematic planning and IMCI implementation activities (Huicho et al. 2005). Strengths and limitations Due to its qualitative nature, our study offers an in-depth analysis of various health system factors affecting IMCI implementation and answers the ‘why’ and ‘how’ aspects which have not been well described in previous research. Moreover, the use of health system dynamic framework allowed us to take a comprehensive view of enablers and barriers affecting IMCI implementation. Since the study sample included only two districts, the results reflect a local situation and cannot be extrapolated to other districts of South Africa or beyond since there is marked variability in how provinces and districts deliver child health services with no standardized approach. Nevertheless, we believe that systemic issues such as poor monitoring and supervision, limited governance and human resource challenges are replicated in many settings. We excluded health financing as a building block and acknowledge the need to explore it through dedicated research in future. Two integral components of IMCI implementation—community engagement and health worker training—were not specifically addressed. Finally, we recognize that the study findings are based on participant perception and may be biased. An observational study would have strengthened our findings. Nevertheless, many of the participants’ reflections strongly validated our own perspectives garnered through long-term engagement with the local and national health system. Key lessons for child health policy and practice As the global health community prepares to re-direct efforts to achieving the Sustainable Development Goals (SDGs), the quality of PHC services for children under 5 years remains a critical issue for many countries, including South Africa. Considering that there is no major alternate approach, investment in improving IMCI implementation is warranted. This study comes at a critical time when the global child health community is thinking about repositioning IMCI under the SDGs and the Global Strategy for Women, Children and Adolescents (2016–2030) (Costello and Dalglish 2016). It confirms that due to its integrated design, IMCI’s success is bound to depend on the strength of key components of a health system (building blocks), particularly in resource constrained environments. Taking the South African example, our study not only unpacked specific elements of these health system building blocks, which affect implementation, but also identified reasons behind their existence. Based on the study findings, we offer a few specific, actionable recommendations for child PHC (including IMCI) related policy, practice and research, which can be implemented at various levels of a health system to achieve biggest gains (Table 5). Considering that 20 years on, countries still face similar health system issues affecting IMCI implementation, we believe that these suggestions are of relevance to settings beyond our borders. However, they require testing of their feasibility, scalability and wider relevance before adoption. Table 5. Recommendations to strengthen health systems for effective IMCI implementation (in terms of policy and practice) WHO building blocks of health system  IMCI specific recommendations  Broader health system recommendations  Service delivery  More child health conditions need to be added to IMCI package (in alignment with country specific burden of disease, e.g. trauma and skin conditions)  Repackage IMCI under the umbrella of a broader primary health service delivery for children by defining the minimum package of services (including IMCI and PCCD) which a sick child should receive at every clinic visit and clearly communicate this to all implementers.  Human resources  Prioritize mentoring and support to practitioners immediately post training as well as on a regular basis.  Policy guidance on appropriate duty allocation and rotation of health workers in health facilities. This must ensure allocating adequate number of trained staff for effective execution of primary health services for children on a daily basis considering the patient load, clinic infrastructure and operating hours. (especially for community health centres)  Clinic managers need to ensure allocation of practitioners to IMCI duty station immediately post training as well as on a regular basis.  Conduct regular IMCI refresher trainings.  Additional options for building health workers’ capacity need to be considered such as self directed learning and peer learning.1  Auxiliary nursing cadres such as enrolled nurses can be utilized to deliver components of IMCI especially health prevention and promotion activities such as growth monitoring, immunization and nutritional counselling.  Medical technology  In centralized pharmacies dispensing IMCI drugs, pharmacists need to demonstrate first dose administration and counsel the caregiver on drug use and storage.    Health information system  Policy guidance on how to integrate IMCI monitoring with routine child health information system. This must include identification of a limited number of IMCI specific indicators to be tracked (including indicators suggested by WHO), who should collect them, how frequently, what data sources to be used at clinic, sub-district, province and national levels, data dissemination, interpretation and establishing a system of providing feedback to implementers.  Quality of care provided to sick children to be routinely monitored in order to assess practitioners’ adherence to IMCI guidelines.  This can be done by conducting regular audits of child health records in addition to observation of IMCI case management followed by providing feedback to child health practitioners on a regular basis.  Data sources reflecting quality of care and other indicators need to be regularly collated, disseminated and utilized optimally for decision making not only at district level but also at provincial and national levels.  Mobile phone and electronic devices for real time data capture can be considered for collecting data on monitoring1.  Standardized child health records with IMCI forms to be compulsorily filled by all practitioners on every sick child encounter.  Red flag poor performing provinces or districts that need additional support for improving IMCI implementation/child PHC services, based on routinely collected monitoring data.  Leadership and governance  Mapping of existing managerial cadres available for PHC/MCWH/child health on the basis of which, adequate number of MCWH coordinators/supervisors must be allocated at sub-district and district levels (according to catchment area and number of clinics) who will be responsible for coordinating and implementing IMCI activities such as supervision and monitoring on a regular basis.  Better coordination and communication between vertical child health programmes (such as HIV, TB and EPI) and IMCI, regarding strategic planning and implementation of activities.  Options such as integrated supervision and monitoring can be considered.  Clear and consistent policy guideline in terms of workforce planning for managerial/supervisory cadres and allocation of duties for child PHC services including IMCI activities. These cadres also need to be sensitized and trained on case management as well as supervision and monitoring elements.  This should include clear directions on who should supervise/coordinate, how many clinics, which checklist to be used and frequency of supervisory and monitoring visits.  WHO building blocks of health system  IMCI specific recommendations  Broader health system recommendations  Service delivery  More child health conditions need to be added to IMCI package (in alignment with country specific burden of disease, e.g. trauma and skin conditions)  Repackage IMCI under the umbrella of a broader primary health service delivery for children by defining the minimum package of services (including IMCI and PCCD) which a sick child should receive at every clinic visit and clearly communicate this to all implementers.  Human resources  Prioritize mentoring and support to practitioners immediately post training as well as on a regular basis.  Policy guidance on appropriate duty allocation and rotation of health workers in health facilities. This must ensure allocating adequate number of trained staff for effective execution of primary health services for children on a daily basis considering the patient load, clinic infrastructure and operating hours. (especially for community health centres)  Clinic managers need to ensure allocation of practitioners to IMCI duty station immediately post training as well as on a regular basis.  Conduct regular IMCI refresher trainings.  Additional options for building health workers’ capacity need to be considered such as self directed learning and peer learning.1  Auxiliary nursing cadres such as enrolled nurses can be utilized to deliver components of IMCI especially health prevention and promotion activities such as growth monitoring, immunization and nutritional counselling.  Medical technology  In centralized pharmacies dispensing IMCI drugs, pharmacists need to demonstrate first dose administration and counsel the caregiver on drug use and storage.    Health information system  Policy guidance on how to integrate IMCI monitoring with routine child health information system. This must include identification of a limited number of IMCI specific indicators to be tracked (including indicators suggested by WHO), who should collect them, how frequently, what data sources to be used at clinic, sub-district, province and national levels, data dissemination, interpretation and establishing a system of providing feedback to implementers.  Quality of care provided to sick children to be routinely monitored in order to assess practitioners’ adherence to IMCI guidelines.  This can be done by conducting regular audits of child health records in addition to observation of IMCI case management followed by providing feedback to child health practitioners on a regular basis.  Data sources reflecting quality of care and other indicators need to be regularly collated, disseminated and utilized optimally for decision making not only at district level but also at provincial and national levels.  Mobile phone and electronic devices for real time data capture can be considered for collecting data on monitoring1.  Standardized child health records with IMCI forms to be compulsorily filled by all practitioners on every sick child encounter.  Red flag poor performing provinces or districts that need additional support for improving IMCI implementation/child PHC services, based on routinely collected monitoring data.  Leadership and governance  Mapping of existing managerial cadres available for PHC/MCWH/child health on the basis of which, adequate number of MCWH coordinators/supervisors must be allocated at sub-district and district levels (according to catchment area and number of clinics) who will be responsible for coordinating and implementing IMCI activities such as supervision and monitoring on a regular basis.  Better coordination and communication between vertical child health programmes (such as HIV, TB and EPI) and IMCI, regarding strategic planning and implementation of activities.  Options such as integrated supervision and monitoring can be considered.  Clear and consistent policy guideline in terms of workforce planning for managerial/supervisory cadres and allocation of duties for child PHC services including IMCI activities. These cadres also need to be sensitized and trained on case management as well as supervision and monitoring elements.  This should include clear directions on who should supervise/coordinate, how many clinics, which checklist to be used and frequency of supervisory and monitoring visits.  Firstly, the role of IMCI and other coexisting approaches (such as PCCD) needs to be clearly defined within a PHC service package for children. This will reduce the ambiguity and conflict which practitioners face while managing a sick child. Secondly, to retain competence of nurses in practicing IMCI, resources should be diverted from training to maintaining competence of already trained nurses. This can be achieved by a focus on post-training supervision and mentoring, duty allocation and rotation policies, considering the heterogeneity of clinic organization and needs in a particular setting. Thirdly, to compensate for human resource deficiencies, we recommend that auxiliary nursing cadres be utilized to deliver the prevention and promotion components of IMCI. This threatens the holistic single provider delivery model, but could optimize resource use, allowing trained practitioners to focus on diagnostic and management tasks while auxiliaries complete less skill-intensive and time-consuming components. Fourthly, compulsory completion of standardized child health records (IMCI forms) would enable process compliance and facilitate audit. Successful integration of different services and programmes remains elusive, but warrants innovation and research. Finally, strong clinical governance is required for optimal workforce deployment and accountability extraction. While district management has a critical role in effective operational planning and execution of activities at local level, optimal IMCI implementation cannot be achieved without appropriate strategic planning and policy direction from provincial and national levels. Funding Faculty Research Committee (FRC) Individual research grant awarded by the University of the Witwatersrand, Johannesburg. (Grant no. 0012548438104512110500000000000000004988). Footnotes 1 Costello AM, Dalglish SL. 2016. Towards a Grand Convergence for child survival and health: A strategic review of options for the future building on lessons learnt from IMNCI. Geneva: WHO. Online at http://apps.who.int/iris/bitstream/10665/251855/1/WHO-MCA-16.04-eng.pdf, accessed 20 Jan 2016. Acknowledgements We thank the national, provincial and district health departments and all interviewees for their participation in this research study. We acknowledge the financial assistance received from the University of the Witwatersrand, Johannesburg and recognize Simbareshe Tevera for his assistance in transcribing the interview audio tapes. Ethical approval for the study was obtained from the University of the Witwatersrand’s Human Research Ethics Committee (clearance number M141194). Conflict of interest statement. None declared. References Adam T, Amorim DG, Edwards SJ, Amaral J, Evans DB. 2005. Capacity constraints to the adoption of new interventions: consultation time and the Integrated Management of Childhood Illness in Brazil. Health Policy and Planning  20: i49– 57. Google Scholar CrossRef Search ADS PubMed  Ahmed HM, Mitchell M, Hedt B. 2010. National implementation of Integrated Management of Childhood Illness (IMCI): policy constraints and strategies. Health Policy  96: 128– 33. Google Scholar CrossRef Search ADS PubMed  Arifeen SE, Blum LS, Hoque DME. et al.   2004. Integrated Management of Childhood Illness (IMCI) in Bangladesh: early findings from a cluster-randomised study. The Lancet  364: 1595– 602. Google Scholar CrossRef Search ADS   Armstrong SJ, Bryce J, De Savigny D, Lambrechts T, Mbuya C. 2004. The effect of Integrated Management of Childhood Illness on observed quality of care of under-fives in rural Tanzania. Health Policy and Planning  19: 1– 10. Google Scholar CrossRef Search ADS PubMed  Atun RA, Bennett S, Duran A. 2008. When do vertical (stand-alone) programmes have a place in health systems ? Geneva: World Health Organization. Online at http://www.who.int/management/district/services/WhenDoVerticalProgrammesPlaceHealthSystems.pdf, accessed 10 April 2016. Binkin N, Chopra M, Simen-Kapeu A, Westhof D. 2011. Do improvements in outreach, clinical, and family and community-based services predict improvements in child survival? An analysis of serial cross-sectional national surveys. BMC Public Health  11: 456. Google Scholar CrossRef Search ADS PubMed  Bryce J, Victora CG, Habicht JP, Black RE, Scherpbier RW. 2005. Programmatic pathways to child survival: results of a multi-country evaluation of Integrated Management of Childhood Illness. Health Policy and Planning  20: i5– i17. Google Scholar CrossRef Search ADS PubMed  Callaghan M, Ford N, Schneider H. 2010. A systematic review of task- shifting for HIV treatment and care in Africa. Human Resources for Health  8: 1– 9. Google Scholar CrossRef Search ADS PubMed  Chopra M, Patel S, Cloete K, Sanders D, Peterson S. 2005. Effect of an IMCI intervention on quality of care across four districts in Cape Town, South Africa. Archives of Disease in Childhood  90: 397– 401. Google Scholar CrossRef Search ADS PubMed  Costello AM, Dalglish SL. 2016. Towards a Grand Convergence for child survival and health: a strategic review of options for the future building on lessons learnt from IMNCI . Geneva: WHO. Online at http://apps.who.int/iris/bitstream/10665/251855/1/WHO-MCA-16.04-eng.pdf, accessed 20 January 2016. Daviaud E, Subedar H. 2012. Staffing Norms for Primary Health Care in the context of PHC Re-engineering- Report to National Department of Health . Cape Town: Medical Research Council. Online at http://www.mrc.ac.za/healthsystems/StaffingNorms.pdf, accessed 4 October 2017. De Savigny D, Adam T. 2009. Systems thinking for health systems strengthening . Geneva: Alliance for Health Policy and Systems Research; World Health Organization. Online at http://apps.who.int/iris/bitstream/10665/44204/1/9789241563895_eng.pdf, accessed 5 May 2016. District Department of Health. 2012. Child health Situational Analysis . Johannesburg: DCST. Unpublished document. Gera T, Shah D, Garner P, Richardson M, Sachdev HS. 2016. Integrated management of childhood illness (IMCI) strategy for children under five. The Cochrane Database of Systematic Review  6: 1– 60. Gilson L, Daire J, Patharath A, English R. 2011. Leadership and governance within the South African health system. South African Health Review  2016: 69– 80. Hagopian A, Mohanty MK, Das A, House PJ. 2012. Applying WHO's ′workforce indicators of staffing need′ (WISN) method to calculate the health worker requirements for India's maternal and child health service guarantees in Orissa State. Health Policy and Planning  27: 11– 8. Google Scholar CrossRef Search ADS PubMed  Horwood C, Vermaak K, Rollins N. et al.   2009a. An evaluation of the quality of IMCI assessments among IMCI trained health workers in South Africa. PLoS One  4: e5937. Google Scholar CrossRef Search ADS   Horwood C, Vermaak K, Rollins N. et al.   2009b. Paediatric HIV management at primary care level: an evaluation of the integrated management of childhood illness (IMCI) guidelines for HIV. BMC Pediatrics  9: 59. Google Scholar CrossRef Search ADS   Horwood C, Voce A, Vermaak K, Rollins N, Qazi S. 2009c. Experiences of training and implementation of integrated management of childhood illness (IMCI) in South Africa: a qualitative evaluation of the IMCI case management training course. BMC Pediatrics  9: 62. Google Scholar CrossRef Search ADS   Huicho L, Dávila M, Campos M. et al.   2005. Scaling up integrated management of childhood illness to the national level: achievements and challenges in Peru. Health Policy and Planning  20: 14– 24. Google Scholar CrossRef Search ADS PubMed  Jonker L, Stellenberg EL. 2014. Missed opportunities in child healthcare. African Journal of Primary Health Care & Family Medicine  6: 8. Google Scholar CrossRef Search ADS   Kiplagat A, Musto R, Mwizamholya D, Morona D. 2014. Factors influencing the implementation of integrated management of childhood illness (IMCI) by healthcare workers at public health centers & dispensaries in Mwanza, Tanzania. BMC Public Health  14: 277. Google Scholar CrossRef Search ADS PubMed  Lange S, Mwisongo A, Mæstad O. 2014. Why don't clinicians adhere more consistently to guidelines for the Integrated Management of Childhood Illness (IMCI)? Social Science & Medicine  104: 56– 63. Google Scholar CrossRef Search ADS   Loening W, Bamford L. 2012. Integrated Management of Childhood Illness. In: Westwood T, Kibel M, Saloojee H(eds). Child Health for All . 5th edn. Cape Town: Oxford University Press, 200– 205. Makgetla N, Fotoyi A. 2016. Real Economy Bulletin 2016—Gauteng Provincial Review. Online at: https://www.tips.org.za/images/The_REB_Provincial_Review_2016_Gauteng.pdf, accessed 28 March 2017. MaxQDA. 2016. MaxQDA—The art of data analysis. Online at http://www.maxqda.com/legalinfo, accessed 20 November 2015. Mazaleni N. 2009. Strengthening community-based child health services in South Africa. South African Child Gauge. 71-76. Online at http://pmg-assets.s3-website-eu-west-1.amazonaws.com/docs/100811childgauge.pdf#page=73, accessed 26 May 2016. Mulaudzi MC. 2015. Adherence to case management guidelines of Integrated Management of Childhood Illness (IMCI) by healthcare workers in Tshwane, South Africa. South African Journal of Child Health  9: 89– 92. Google Scholar CrossRef Search ADS   National Department of Health (NDOH). 2014. Integrated management of childhood illnesses (IMCI) chart booklet . Pretoria: National Department of Health. Online at http://paediatrics.ukzn.ac.za/Files/2014%20IMCI%20CHART%20BOOKLET%20Final.pdf, accessed 6 April 2015. National Department of Health (NDOH). 2012. National strategic plan for maternal, newborn, child, women, adolescent health and nutrition (MNCWAH-N 2012-2016).  Pretoria: National Department of Health. National Department of Health (NDOH). 2016. Mid term review of National strategic plan for maternal, newborn, child, women, adolescent health and nutrition (MNCWAHN, 2012-2016) . Pretoria: National Department of Health. National Department of Health (NDOH). 2001. Primary health care package for South Africa - A set of norms and standards . Pretoria: National Department of Health. Online at https://www.westerncape.gov.za/general-publication/primary-health-care-package-sa-set-norms-and-standards, accessed 7 April 2016. Nyamtema AS, Urassa DP, Massawe S, Massawe A, Lindmark G. 2008. Staffing Needs for Quality Perinatal Care in Tanzania. African Journal of Reproductive Health  12: 113– 24. Google Scholar PubMed  Pariyo GW, Gouws E, Bryce J, Burnham G. 2005. Improving facility-based care for sick children in Uganda: training is not enough. Health Policy and Planning  20: i58– 68. Google Scholar CrossRef Search ADS PubMed  Pradhan NA, Rizvi N, Sami N, Gul X. 2013. Insight into implementation of facility-based integrated management of childhood illness strategy in a rural district of Sindh, Pakistan. Global Health Action  6: 20086. Google Scholar CrossRef Search ADS PubMed  Prosper H, Macha J, Borghi J. 2009. Implementation of integrated management of childhood illness in Tanzania: success and challenges. Consortium for Research on Equitable Health Systems. Online at http://citeseerx.ist.psu.edu/viewdoc/download? doi=10.1.1.690.9032&rep=rep1&type=pdf, accessed 13 May 2016. Rakha MA, Abdelmoneim ANM, Farhoud S et al.   2013. Does implementation of the IMCI strategy have an impact on child mortality? A retrospective analysis of routine data from Egypt. BMJ Open  3: e001852. Google Scholar CrossRef Search ADS PubMed  Rohde J, Cousens S, Chopra M et al.   2008. 30 years after Alma-Ata: has primary health care worked in countries? The Lancet  372: 950– 61. Google Scholar CrossRef Search ADS   Saloojee H. 2009. Managing resources and building capacity in the context of child health. Child Gauge  2009: 64– 70. Saloojee H, Bamford L. 2006. Key childhood health promotion and disease prevention programmes. South African Health Review  181– 202. Online at https://www.popline.org/node/196340, accessed 10 September 2017. Silali MB. 2014. Utilization of Integrated Management of Childhood Illnesses IMCI for Child health in Western Kenya. Journal of Biology, Agriculture and Healthcare  4: 102– 13. Soweto trust for nurse clinical training. 2012. Primary clinical care manual— A practical guide for primary health care personnel in the clinical setting. 7th ed. Johannesburg: Soweto trust for nurse clinical training. Srivastava A, Thomson SB. 2009. Framework analysis: a qualitative methodology for applied policy research. Journal of Administration & Governance  4: 72– 79 Online at http://research.apc.org/images/a/ad/Framework_analysis.pdf, accessed 4 June 2016. Thandrayen K. 2010. The quality of child health services offered at primary health care clinics in Johannesburg. South African Journal of Child Health  4: 73– 7. Tulloch J. 1999. Integrated approach to child health in developing countries. The Lancet  354: SII16– 20. Google Scholar CrossRef Search ADS   United Nations. 2015. The Global strategy for women's, children's and adolescents' health (2016-2030). Every woman every child. Online at: http://www.who.int/pmnch/media/events/2015/gs_2016_30.pdf, accessed 7 June 2016. Van Olmen J, Criel B, Van Damme W et al.   2012. Analysing health system dynamics–a framework. Studies in Health Services Organization and Policy  2: 1. WHO. 2004. The Analytic review of the Integrated Management of Childhood Illness strategy - Final Report . Geneva: World Health Organization. Online at: http://apps.who.int/iris/bitstream/10665/42965/1/9241591730.pdf, accessed 16 April 2016. WHO. 1999. IMCI Information—Monitoring and Evaluation . Geneva: World Health Organization. Online at http://apps.who.int/iris/bitstream/10665/65002/12/WHO_CHS_CAH_98.1K_eng.pdf, accessed 18 April 2016. WHO. 2007. Everybody's business–strengthening health systems to improve health outcomes: WHO's framework for action . Geneva: World Health Organization, p.3. Online at: http://apps.who.int/iris/bitstream/10665/43918/1/9789241596077_eng.pdf, accessed 12 January 2016. WHO. 2009. IMCI [Online]. Online at http://www.who.int/maternal_child_adolescent/topics/child/imci/en/, accessed 25th October 2014. WHO. 2010. Workload Indicators of Staffing Needs (WISN) - Users Manual . Geneva: World Health Organization. Online at http://www.who.int/hrh/resources/WISN_Eng_UsersManual.pdf? ua=1, accessed 14 August 2016. © The Author 2017. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Health Policy and Planning Oxford University Press

Health system factors affecting implementation of integrated management of childhood illness (IMCI): qualitative insights from a South African province

Loading next page...
 
/lp/ou_press/health-system-factors-affecting-implementation-of-integrated-Di0ugcNony
Publisher
Oxford University Press
Copyright
© The Author 2017. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
0268-1080
eISSN
1460-2237
D.O.I.
10.1093/heapol/czx154
Publisher site
See Article on Publisher Site

Abstract

Abstract The Integrated Management of Childhood Illness (IMCI) strategy has been adopted by 102 countries including South Africa, as the preferred primary health care (PHC) delivery strategy for sick children under 5 years. Despite substantial investment to support IMCI in South Africa, its delivery remains sub-optimal, with varied implementation in different settings. There is scarce research globally, and in the local context, examining the effects of health system characteristics on IMCI implementation. This study explored key determinants of IMCI delivery in a South African province, with a specific focus on health system building blocks using a health system dynamics framework. In-depth interviews were conducted with 38 districts, provincial and national respondents involved with IMCI co-ordination and delivery, exploring their involvement in, and perceptions of, IMCI strategy implementation. Identified barriers included poor definition of elements of a service package for children and how IMCI aligned with this, incompetence of trained nurses exacerbated by inappropriate rotation practices, use of inappropriate indicators to track progress, multiple cadres coordinating similar activities with poor role delineation, and fragmented, vertical governance of programmes included within IMCI, such as immunization. Enabling practices in one district included the use of standardized child health records incorporating IMCI activities and stringent practice monitoring through record audits. Using IMCI as a case study, our work highlights critical health system deficiencies affecting service delivery for young children which need to be resolved to reposition IMCI within the broader child ‘survive, thrive and transform’ agenda. Recommendations for appropriate health system strengthening include the need for redefining IMCI within a broader PHC service package for children, prioritizing post-training supervision and mentoring of practitioners through appropriate duty allocation and rotation policies, strengthening IMCI monitoring with a specific focus on quality of care and building stronger clinical governance through workforce allocation, role delineation and improved accountability. IMCI implementation, child health, health systems, primary health care, South Africa Key Messages Failure to articulate and implement a preferred strategy for delivery of primary health care services for children led to idiosyncratic and often poor service provision. IMCI implementation was inadequately monitored, supported, co-ordinated and governed. Health professional training was prioritized, but insufficient subsequent supervision and mentoring, and limited adaptation of the work environment and routine allocation, compromised quality service delivery. Failure to collect and/or to meaningfully synthesize IMCI data, particularly related to the quality of its delivery, led to limited appreciation of the strategy’s successes or failures, and non-response to these. Ill-defined and overlapping responsibilities for child health managers, and higher priority for parallel child health programmes, affected IMCI coordination and implementation. Introduction The Integrated Management of Childhood Illness (IMCI) strategy has proven to be a cost-effective intervention to improve the quality of primary health care (PHC) service delivery in resource-limited settings. Its main objective is to reduce child mortality and morbidity by both preventive and promotive health care (Ahmed et al. 2010; Arifeen et al. 2004; WHO 2009; Tulloch 1999; Rakha et al. 2013). It targets improvements in the case management skills of health-care staff, in overall health systems and in community and family health practices. Within health facilities, the IMCI strategy seeks to promote the accurate identification of childhood illnesses with a comprehensive treatment response, strengthen the counselling of caretakers and speed up the referral of severely ill children (WHO 1999). Many countries have struggled to scale up IMCI and achieve high coverage since its introduction over two decades ago (Bryce et al. 2005; WHO 2004; Ahmed et al.2010). Effective IMCI implementation has been most challenging, unsurprisingly, in countries with weaker health systems (Armstrong et al. 2004). Currently, there is global discourse about the continued survival of IMCI as an integrated strategy amidst various vertical programmes. IMCI was specifically introduced to minimize vertical approaches to child health delivery (such as immunization or diarrhoeal disease control programmes) (WHO 1999). However, it has struggled to fulfil its potential in under-resourced health systems where vertical programmes continue to dominate (Ahmed et al. 2010). Recently, IMCI has elicited renewed interest at a global level, including a World Health Organization (WHO) strategic review re-examining IMCI in light of emerging priorities and new developments in the global child health agenda of ‘survive, thrive and transform’ (Costello and Dalglish 2016; United Nations 2015). A 2016 systematic review evaluating the effects of implementation of IMCI strategy on critical child health outcomes concluded that researchers should continue to explore how the IMCI strategy can best be delivered (Gera et al. 2016). South Africa adopted IMCI in 1998 and initial evaluations showed promising results (Chopra et al. 2005). By 2011, most PHC facilities were implementing the strategy, with two-thirds adequately saturated with IMCI trained health workers (Loening and Bamford 2012). However, even after two decades of sustained investment in training and infrastructure, IMCI implementation is far from optimal (Horwood et al. 2009a; 2009b; 2009c; Jonker and Stellenberg 2014; Mulaudzi 2015; Mazaleni 2009; District Department of Health 2012; National Department of Health 2016). Moreover, there is no evidence to demonstrate the impact of IMCI implementation on child mortality and morbidity outcomes in South Africa (Thandrayen 2010; Saloojee and Bamford 2006). Furthermore, IMCI evaluation data is either unavailable or not reliable. In South Africa, delivery of child PHC services is considered to be idiosyncratic, depending on the whims of individual practitioners and facilities, with IMCI regarded as an optional rather than prescribed practice in many districts (National Department of Health 2016; Thandrayen 2010). Documented reasons for poor IMCI implementation in South African settings include deficient practitioner competency, IMCI consultations being too time-consuming and poor clinical supervision of practitioners. This is consistent with reports from other countries (Bryce et al. 2005; Huicho et al. 2005; Kiplagat et al. 2014; Lange et al. 2014). However, there is limited description of health system’s influence, as a whole, on implementation of IMCI at national and indeed global level. Internationally, there is a lack of consensus on IMCI’s role within the delivery of an overall PHC service package for children and how it could best be integrated within broader health system strategies in future. This demands a relook at IMCI’s position in the larger scheme—particularly the future delivery of child ‘survival and thrive’ strategies and the associated health system strengthening required. Hence, the motivation for this study was two-fold—to analyse the failure of IMCI to deliver expected outcomes in South Africa through an in-depth analysis of health system factors affecting its implementation, and to contribute to the global discussion on the ongoing implementation and re-design of IMCI. Moreover, the objective was not only to identify barriers and enablers, but also to better understand the root causes for their existence in a local context. Although the study was based in a middle-income setting, the concerns identified and responses required are likely to be as relevant to all IMCI implementing settings. Methods Analytical/thematic framework for the study Health systems are complex and dynamic in the way they function (WHO 2007; De Savigny and Adam 2009). The WHO’s six building blocks of a health system are well known (WHO 2007). Van Olmen et al. proposed a health systems dynamics framework derived from the WHO model accentuating the dynamic relationships existing between the building blocks (Van Olmen et al. 2012). We utilized a simplified version of this framework to map and explain how a health system operates in the context of IMCI and factors associated with each building block affecting its implementation. According to our framework (Figure 1), the implementation of IMCI package at a facility level is considered synonymous to ‘service delivery’ as a process. The other building blocks, viz. human resources, medical technologies and health information system act as inputs which influence the process of service delivery. Leadership and governance serve as an overarching block affecting both the inputs and process. We assume that these building blocks, if properly operationalized, lead to improved implementation of IMCI and better child health outcomes. We focussed only on inputs and process which together constitute five building blocks. To stay focussed on our research question, we excluded study of IMCI outcomes. We also excluded examination of the sixth WHO building block of health financing since neither child PHC nor IMCI receive dedicated (ring-fenced) funding in South Africa. Their costs are immersed within a PHC budget. In addition, study participants had a limited ability to influence finances. Figure 1. View largeDownload slide Health systems dynamic framework for IMCI Figure 1. View largeDownload slide Health systems dynamic framework for IMCI Study setting The South African health system consists of one national, nine provincial and 52 district health structures. Districts are further divided into sub-districts and municipal wards as administrative units. Within each district, clinics and community health centres (CHCs) are the key delivery points for PHC services, including IMCI. Districts are responsible for operationalizing the IMCI strategy. Child PHC services are primarily delivered by professional nurses who receive pre-service and/or in-service training in IMCI. Additionally, some professional nurses undertake a 1-year primary clinical care diploma (PCCD) offered by the Nursing Council of South Africa. This supplementary training is meant to augment nurses’ ability to manage adults and sick children at a PHC level (Soweto trust for nurse clinical training 2012). It promotes traditional consultation with focus on chief complaint and use of equipment such as stethoscope and otoscope. The household and community component of IMCI are meant to be delivered though ward-based outreach teams (WBOTs) including community health workers who receive training on key family practices, but there has been variable implementation of this component (National Department of Health 2016; Mazaleni 2009). Study design and participant selection A qualitative, descriptive study was conducted from March 2015 to August 2016 in Gauteng province, South Africa. This is the wealthiest and most populous province in the country, and predominantly urban (Makgetla and Fotoyi 2016). Two of five districts in the province were purposively selected—subsequently referred to as district A (peri-urban/rural) and B (urban). We collected district IMCI reports with a list of all PHC facilities based on their IMCI implementation performance status (categorized as ‘good’ or ‘poor’ as reported by district child health coordinators). Next, we randomly selected two good and two poorly performing facilities, ensuring a selection from different sub-districts. Thus, the district sample consisted of four facilities including a CHC from each district. Only those facilities saturated with trained staff, i.e. a minimum of 60% of staff trained in IMCI (WHO 1999), were eligible for inclusion to minimize possible differences resulting from training status. To explore the status of IMCI implementation across all levels of health system, a range of participants were purposively selected including professional nurses (2–3 from each clinic depending on availability), PHC facility managers, and key coordinators and supervisors responsible for child health services at a sub-district, district, provincial and national level. All invited participants consented to involvement. Table 1 provides detail on participants and their designations. Table 1. Details of study participants   Level of health system where participant located  Participant’s designation  Number of participants  District A  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  District  District child health coordinator  1  District clinical specialist team (DCST) paediatrician  1  District B  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  Maternal Child and Women’s Health (MCWH) coordinatora  2  Districtb  District clinical specialist team (DCST) paediatrician  1  District clinical specialist team (DCST) paediatric nurse  1  Province    c  1  National    c  1  Total      38    Level of health system where participant located  Participant’s designation  Number of participants  District A  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  District  District child health coordinator  1  District clinical specialist team (DCST) paediatrician  1  District B  Clinic (n = 4)  Facility manager  4  Professional nurse  9  Sub-district (n = 3)  Area managers/PHC supervisor  3  Maternal Child and Women’s Health (MCWH) coordinatora  2  Districtb  District clinical specialist team (DCST) paediatrician  1  District clinical specialist team (DCST) paediatric nurse  1  Province    c  1  National    c  1  Total      38  a District B had sub district MCWH coordinators who were not present in district A. Each such coordinator had 5-8 clinics under their supervision. Two coordinators responsible for the sampled clinics were interviewed. b MCWH coordinator of district B was not interviewed because she recently joined the post and did not have much knowledge about IMCI. The coordinator, whom she replaced, was posted to one of the sub-districts as an MCWH coordinator and was interviewed instead. c Designations of key informants at province and national level have been omitted to ensure anonymity. Data collection and analysis Thirty-eight in-depth, semi-structured interviews were conducted in English. Interviews were recorded via audiotaping (where consent was provided) and transcribed; otherwise detailed hand written notes were taken and typed. We designed an interview guide based on study objectives, with open ended questions (Table 2). It included questions on current status of IMCI implementation and the challenges faced specifically structured around health system building blocks. We also reviewed documents related to IMCI activities at various levels where available and permitted. These included clinic registers, child health visit records, supervisors’ reports and checklists, and provincial and national reports on IMCI. The document review was undertaken to supplement and validate data obtained from interviews on health information systems, explore the status of IMCI monitoring, and determine the use of indicators to track implementation progress. Table 2. Interview guide Describe how you provide services for children less than 5 years in this clinic.  What guidelines do you follow to manage children less than 5 years/how do you manage sick children at this facility?  What do you know about IMCI, what are your perceptions/opinions about the IMCI strategy?  How is IMCI governed/managed at various levels? (coordinators, directorates)  How do you monitor IMCI/under 5 child health services in this clinic/sub-district/district?  Do you receive/provide any supervision or mentoring support for IMCI? If yes, describe.  What challenges do you face, if any, in offering/implementing IMCI at your clinic/sub-district/district/province? (Human resources, equipment, supplies, monitoring, supervision, others)  Describe how you provide services for children less than 5 years in this clinic.  What guidelines do you follow to manage children less than 5 years/how do you manage sick children at this facility?  What do you know about IMCI, what are your perceptions/opinions about the IMCI strategy?  How is IMCI governed/managed at various levels? (coordinators, directorates)  How do you monitor IMCI/under 5 child health services in this clinic/sub-district/district?  Do you receive/provide any supervision or mentoring support for IMCI? If yes, describe.  What challenges do you face, if any, in offering/implementing IMCI at your clinic/sub-district/district/province? (Human resources, equipment, supplies, monitoring, supervision, others)  Since the study did not involve any observational element, to elicit information on clinicians’ IMCI related practices and challenges, we probed using questions such as ‘how do you manage a sick child with diarrhoea or pneumonia’ and compared their responses to our benchmark, thereby interpreting good/poor IMCI adherence. In addition, a review of child health records where sick child consultations were captured, supplemented our interpretation. Transcripts were coded and analysed using the MaxQDA data analysis software (VERBI Software GmbH, Berlin)(MaxQDA 2016). Coded data were analysed using the framework analysis approach (Srivastava and Thomson 2009). The health system dynamic framework (Figure 1) acted as our thematic framework and formed basis of the analysis in which five building blocks were used as deductive themes. First, all text segments from transcripts, documents and written notes were coded to classify the data and identify emerging issues and concepts from interviews. Codes were predominantly inductively derived, and grouped within a framework building block, e.g. ‘poor IMCI adherence’ emerged as a code which was grouped under the theme of ‘service delivery’. A few deductive codes based on the literature of factors affecting IMCI were also added. Data analysis was initiated after conducting the first interview and concurrent analysis allowed us to revise questions to improve the richness and quality of information obtained. Benchmarks used for data analysis To objectively analyse the factors affecting IMCI implementation, we established benchmarks for actual performance. These were based on official guidelines (from WHO and South African national health department) and logical constructs. For service delivery, we used the South African IMCI guidelines (National Department of Health 2014) to assess the extent to which nurses followed or deviated from the IMCI package. For human resources, we set three benchmarks: (i) there should be adequate number of IMCI trained nurses in a clinic (> 60% saturation), (ii) only an IMCI trained nurse should be assigned to treat sick children, and (iii) IMCI trained nurses must retain their competence in delivering IMCI. For health information systems, the benchmark for ideal monitoring of IMCI was that activities related to health system elements were being regularly reviewed and improved. This had to be done by periodic collection of relevant data and include adequate feedback to implementers. For leadership and governance, three benchmarks were set: (i) policy guidance should be unambiguous and not in conflict with other existing guidelines, (ii) system design—new programmes (vertical or integrated) should not inhibit the effective functioning of existing strategies, and (iii) accountability—for effective implementation of IMCI, workforce planning ensured adequate staff numbers and assignment of responsibilities to appropriate staff with no overlap or omission of activities. Ethical clearance for the study was obtained from the authors’ institute. All participants signed informed consent forms for both interview participation and audio taping. Institutional approval for individual districts and clinics were also obtained from district and provincial research committees prior to study commencement. For the sake of retaining anonymity and confidentiality, we have not provided actual names of clinics or districts. Results Results are presented as per health system building blocks based on our thematic framework (Table 3). For each building block, an overview of the existing status is followed by key challenges reported by participants and how these affected IMCI implementation. Although our sample included a mix of good and poorly performing clinics, the challenges we identified existed across both categories. While most findings were directly reported by participants, some were derived indirectly from our synthesis of participant responses based on comparison with the established benchmarks. Table 3. Summary of key health system barriers and enablers affecting IMCI implementation in South Africa Key barriers obstructing IMCI implementation  Key enablers promoting IMCI implementation  Service delivery    Perceptions of primary clinical care trained nurses about IMCI  Use of standardized child health files with IMCI recording forms (compulsorily filled by nurses at every sick child consultation)  Lack of policy guidelines on what constitutes a primary care package  Long queues, high patient waiting times  Human resources    Frequent temporary staff shortages  High saturation of trained professional nurses at facilities  Poor allocation/rotation practices leading to IMCI trained nurses not being sufficiently competent  Medical technology    Stock outs of growth monitoring equipment, vitamin A and deworming drugs  Availability of a national Essential Drug List  Centralized drug dispensing system  Health information system    Non-systematic capturing of case management adherence  Internal and external record audits conducted by facility managers and Maternal Child and Women’s Health (MCWH) coordinators, respectively  Quality of care data not being utilized for decision making and monitoring of IMCI implementation  Leadership and governance    Poor supervision and mentoring of nurses due to shortage of Maternal Child and Women’s Health (MCWH) coordinators  Improved supervision and monitoring owing to availability of Maternal Child and Women’s Health coordinators at a sub-district level (5–10 clinics per coordinators)  Fragmented governance of vertical child health programmes (e.g. Expanded programme on Immunization) that are part of IMCI  Lack of clarity on roles and responsibilities amongst various stakeholders on IMCI related activities leading to duplication or omission of tasks  Poor accountability for child health outcomes  Key barriers obstructing IMCI implementation  Key enablers promoting IMCI implementation  Service delivery    Perceptions of primary clinical care trained nurses about IMCI  Use of standardized child health files with IMCI recording forms (compulsorily filled by nurses at every sick child consultation)  Lack of policy guidelines on what constitutes a primary care package  Long queues, high patient waiting times  Human resources    Frequent temporary staff shortages  High saturation of trained professional nurses at facilities  Poor allocation/rotation practices leading to IMCI trained nurses not being sufficiently competent  Medical technology    Stock outs of growth monitoring equipment, vitamin A and deworming drugs  Availability of a national Essential Drug List  Centralized drug dispensing system  Health information system    Non-systematic capturing of case management adherence  Internal and external record audits conducted by facility managers and Maternal Child and Women’s Health (MCWH) coordinators, respectively  Quality of care data not being utilized for decision making and monitoring of IMCI implementation  Leadership and governance    Poor supervision and mentoring of nurses due to shortage of Maternal Child and Women’s Health (MCWH) coordinators  Improved supervision and monitoring owing to availability of Maternal Child and Women’s Health coordinators at a sub-district level (5–10 clinics per coordinators)  Fragmented governance of vertical child health programmes (e.g. Expanded programme on Immunization) that are part of IMCI  Lack of clarity on roles and responsibilities amongst various stakeholders on IMCI related activities leading to duplication or omission of tasks  Poor accountability for child health outcomes  Service delivery Issues highlighted in this section include poor adherence with the IMCI strategy, omission of aspects of the consultation, lack of clarity about what constituted an IMCI consultation, and poor recording of the consultation. Poor IMCI adherence Despite claims about practicing IMCI, we identified deviations from our IMCI adherence benchmark at most facilities, especially in district A. Nurse practitioners preferred selectively focusing on the child’s main complaint and symptom-based components of IMCI consultation. They tended to make a single diagnosis instead of completing the entire process (algorithm) and providing treatment accordingly. Deviation was more frequently made by PCCD trained nurses who perceived the IMCI approach to be constraining and simplistic. The simple one is IMCI but the one that makes you feel like you have treated the patient is PCCD because you will be able to diagnose the actual thing. So with PCCD you go deep. With IMCI, you just do superficial (Nurse, participant 5, clinic 4, district A).Another explanation offered for non-adherence with IMCI was that IMCI consultations were more time consuming (responses ranged from 20 min to an hour for a single consultation) and extended patient waiting times. Hence, to compensate, activities such as nutritional status classification, immunization, feeding assessment and counselling of the caregiver were curtailed. There was a general under-appreciation of the need to complete all aspects of the IMCI consultation (particularly preventive and promotive aspects) and its associated benefits. The problem is when there are a lot of people in the clinic, there is no time for you to go through all those things for the child. You will be working fast fast fast to push the queue. Sometimes, sisters (nurses) will just be looking at conditions that the child has been brought with and then they leave the rest and you find that the child has not received immunizations and stuff like that (Nurse, participant 14, clinic 2, district B).Moreover, there was uncertainty among both nurses and coordinators as to the required level of engagement with the HIV, TB and malnutrition modules within IMCI (screening or classifying, and treatment) resulting in poor compliance with these sections. Many practitioner perspectives were contradicted by managers and programme coordinators. They disputed that long consultation times were inevitable, and believed these could be optimized (reduced) if IMCI was practiced regularly. They argued that PCCD trained nurses should supplement the IMCI approach rather than replace it. Presenting complaints not specifically covered by IMCI could be addressed under the section ‘assess other conditions’. Moreover, they also dismissed staff shortages and long queues as valid reasons for non-adherence to IMCI contending that these were excuses that could be overcome by better planning. We are saying that even though you have done PCCD training, whenever we manage children under 5 years, we have to use the IMCI strategy. We are not saying that don’t use your clinical skills from PCCD training. But we are saying, please, IMCI is the protocol that we are supposed to follow (Provincial key informant, participant 37). Recording of consultations Nurses’ understanding and adherence with IMCI principles was better in district B. This could be partly attributed to the compulsory requirement for a standardized IMCI recording form to be filled during every sick child encounter. The form promoted compliance with the entire IMCI clinical process, including completion of classifications, appropriate management and attention to promotive actions. In contrast, nurses in district A relied on free hand clinical notes captured on blank pages. These contained significantly less detail of the examination, diagnostic considerations and management plan. Because of the form, sisters are compelled to use IMCI, because if the child is sick, we don’t allow them to use extra filing papers. We need your classifications to be written on IMCI recording forms (Sub-district MCWH coordinator, participant 22, District B).A common misperception was that use of the prescribed form prevented the identification, recording and management of non-IMCI classifications. At the end of the day there are diagnoses that you cannot write in this IMCI form because they are not part of IMCI classification. If it is not included, then you refer. You cannot come up with that diagnosis as a sister (nurse) in that IMCI room (Nurse, participant 10, clinic 1, district B). Human resources for IMCI Themes that emerged were insufficient staff numbers, inappropriate staff allocation and rotation, and difficulties in maintaining competency. Adequate number and workload allocation of IMCI trained nurses Most facility respondents reported an inadequate staff complement for their patient load. However, this was a subjective judgement as they were unaware of what constituted an appropriate ratio and if such a standard existed. There are existing national norms on the number of nurses required for overall PHC clinical services according to patient load (Daviaud and Subedar, 2012). Additionally, the 2001 national PHC norms and standards state that at least one staff member in a clinic should take overall responsibility for managing children (National Department of Health 2001). However, neither of these norms offer any clarity on specific staff numbers and workload allocations required for child health generally and IMCI specifically. Our investigation revealed that due to the lack of clear guidelines, staff shortage was more of a perception than a reality. We observed that nurses (especially in district A) were overwhelmed with multiple responsibilities, e.g. the nurse allocated to IMCI services was also acting as a team leader of community health workers. Unplanned staff absenteeism, training commitments and nurses deputed for outreach service delivery contributed to temporary staff shortages, but since these were a regular occurrence, it led to the perceived notion of perpetual staff shortage. Clinics are busy places. We have done this exercise in quite a few clinics trying to figure out staff shortage because that is the first complaint that we get from everyone, but if you work out how many patients per day are being seen, then majority of clinics have enough staff (District clinical specialist team member, participant 35, district B).We also noted that nurses at CHCs in both districts did not practice IMCI after-hours (4 pm to 7 am) since only two nurses manned CHC during this period. Our findings highlight the need for a more suitable approach to manage IMCI specific staffing needs. Countries such as India and Tanzania have successfully utilized the WHO recommended Workload Indicators and Staffing Needs (WISN) tool to compute the surplus/gap in staffing levels needed to fulfil unmet needs for MCH services (Hagopian et al 2012; Nyamtema et al 2008; WHO 2010). Provincial and national respondents in our study revealed that districts and clinics were in the process of assessing staff workload and allocation using the WISN tool but no response has been implemented to any findings to date (October 2017). Poor competence of IMCI trained nurses To remain competent, IMCI trained practitioners require regular practice. Allocation of IMCI trained nurses to other tasks owing to poor rotation practices was common at all facilities. Respondents in both districts complained about not being allocated to managing sick children immediately after returning from IMCI training, and on a regular basis, which resulted in poor competence. The reported frequency of rotation through an IMCI duty station ranged from three months to two years which was mostly determined by the facility manager based on her/his own criteria. The question of not rotating is the main point, people stick in one service for a long time. We are not given the opportunity to practice what we have learnt. You forget. When you have to face a child then you have to start all over again or ask others because you haven’t been in touch (Nurse, participant 11, clinic 1, district B).This also made post training supervisory support difficult for supervisors since they did not find trained nurses at IMCI duty stations. Moreover, refresher IMCI trainings in both districts were infrequent. When I came back from IMCI training, I had to go to curative (adults). I think it had an impact on my IMCI practice because to be honest, there was never a chance for me to see a sick child (Nurse, participant 16, clinic 3, district B). Medical technology, supplies, drugs and equipment This was not reported as a significant barrier to IMCI implementation, although temporary stock-outs were reported from facilities in both districts. Non-availability of equipment such as mid upper arm circumference tapes and length mats/boards reportedly impacted growth monitoring and malnutrition assessment activities. Respondents identified the drug dispensing system as a barrier to effective IMCI implementation. Owing to a centralized pharmaceutical dispensing system, drugs were stored and dispensed only from the main pharmacy by a pharmacist or pharmacy assistant. Practitioners believed that this compromised counselling of caregivers on usage and storage of medication as recommended by IMCI guidelines. Also, administration of the first dose was not demonstrated by the practitioner potentially affecting caregiver compliance. When we were training for IMCI, we had everything in the cupboard in rooms, you would even mix the drug for mother and give it to the baby. Now they say you write the prescription and the mother has to collect from clinic’s pharmacy. They (pharmacists) are not going to explain to mothers the way we do (Nurse, participant 17, clinic 4, district B). Health information system Issues that were raised included poor monitoring of adequacy of IMCI case management and overall implementation. Monitoring of IMCI case management adherence IMCI consults were being captured more systematically in child health records in district B because of the use of compulsory IMCI forms. I used to photocopy and fill the IMCI forms for each child. But no one came to look for these IMCI forms, so I stopped filling them. Now when I do IMCI, I won’t fill out a form, I will just say the 4 things in my head, fever, cough blah blah. Then I start to follow PCCD approach and give treatment (Nurse, participant 6, clinic 3, District A).Forms also facilitated better monitoring and auditing of IMCI consultations which in turn led to some improvement in nurses’ adherence to IMCI. You do get into trouble when people come and check you or do an audit because they are going to ask you why you are not doing it correctly. You try to give them your reasons but they don’t take it so we have to do the correct classifications (Nurse, participant 16, clinic 3, district B). Our sisters (nurses) do adhere to IMCI guidelines. I know it because I do audit checks of files. We randomly take 5 or 10 files of children. You have to go and do rounds to support and supervise these sisters and then audit their files(Facility manager, participant 20, clinic 4, district B). Monitoring of overall IMCI implementation We identified three datasets with indicators used to monitor overall IMCI implementation (Table 4, Figure 2). Only one of these assessed the quality of care offered (supervisor’s report). The district health information system captured proxy indicators that were too distal to interpret IMCI performance, while the facility-based quarterly IMCI report only offered data on process indicators. The quality of care data (only available in district B) was collated at sub-district level but not shared with the district. The other two datasets were forwarded to district, provincial and national levels, but never resulted in any facility feedback or action. Table 4. Current indicators and data sources used to monitor IMCI implementation in South Africa Data source  District Health Information System (DHIS)  IMCI quarterly report  Supervisors’ report  Indicators included in the source  Children < 5 years pneumonia (new cases)  Is this facility implementing IMCI during this quarter? (yes/no)  Adherence to case management based on observation of IMCI consults and record audits in clinics (as per WHO checklist for IMCI supervision)  Children < 5 years diarrhoea with dehydration (new cases)  % saturation of health workers trained in IMCI available in a facility during this quarter  Children < 5 years severe acute malnutrition (new cases)  Number of professional health workers in PHC facilities (clinics and CHCs) during this quarter  Children < 2 years underweight (new cases) (between −2 and −3 SD)  Number of nurses trained in IMCI case management available at PHC facilities during this quarter  Children < 5 years on food supplementation  Total number of supervisors trained and active during this quarter  Infant exclusively breastfed at hepatitis-B third dose  Total number of IMCI facilitators  available and active during this quarter  (age—4 weeks)  Data source  District Health Information System (DHIS)  IMCI quarterly report  Supervisors’ report  Indicators included in the source  Children < 5 years pneumonia (new cases)  Is this facility implementing IMCI during this quarter? (yes/no)  Adherence to case management based on observation of IMCI consults and record audits in clinics (as per WHO checklist for IMCI supervision)  Children < 5 years diarrhoea with dehydration (new cases)  % saturation of health workers trained in IMCI available in a facility during this quarter  Children < 5 years severe acute malnutrition (new cases)  Number of professional health workers in PHC facilities (clinics and CHCs) during this quarter  Children < 2 years underweight (new cases) (between −2 and −3 SD)  Number of nurses trained in IMCI case management available at PHC facilities during this quarter  Children < 5 years on food supplementation  Total number of supervisors trained and active during this quarter  Infant exclusively breastfed at hepatitis-B third dose  Total number of IMCI facilitators  available and active during this quarter  (age—4 weeks)  Figure 2. View largeDownload slide Process of monitoring IMCI implementation in South Africa at various levels of health system (this figure must be interpreted in combination with Table 4) Figure 2. View largeDownload slide Process of monitoring IMCI implementation in South Africa at various levels of health system (this figure must be interpreted in combination with Table 4) I think the IMCI report that I do, does not talk much of IMCI implementation because it is mainly about saturation of training for IMCI. They are only interested in figures, not implementation (Child health coordinator, participant 17, district A). WHO recommends that a health facility survey providing detailed information on quality of care and health workers’ adherence to IMCI be conducted once every 5 years at national and once every two years at provincial level (WHO 1999). We established that no such survey was conducted at either level since 2004. Leadership and governance Leadership and governance issues were identified as critical components of implementation failure. Themes that emerged were fragmented/vertical governance structures and lack of accountability among key stakeholders. Policy guidance IMCI was formally adopted as the strategy for delivering PHC services for children in South Africa in 1998 and forms part of the national norms and standards for primary health care (National Department of Health 2001). The strategy was also incorporated into the national maternal and child health strategic plan, as a key component of PHC service package for children (National Department of Health 2012). However, guidance on how to effectively operationalize IMCI implementation has always been lacking. As example, there are no specific guidelines on an optimal patient-staff ratio, on the allocation/rotation of nurses to IMCI duty station or on supervision (who should do it, how frequently, what should be done, using which tools?) System design We identified fragmentation in the governance structures for IMCI and other vertical programmes which were meant to be integrated into IMCI such as Expanded Programme on Immunization (EPI), HIV and tuberculosis (TB). For example, at a national, provincial and district level, while IMCI was governed by the maternal, child, women’s health and nutrition (MCWHN) directorate, the EPI programme was governed by communicable disease control directorate. As vertical programmes were prioritized, dedicated coordinators/supervisors for EPI, HIV and TB regularly visited facilities and offered supervision for programme specific components, while ignoring the IMCI aspects of those programmes. The EPI manager will push only her things that side and then even me this side in MCWH directorate. I think the main problem is because the EPI program manager is not IMCI trained (Provincial key informant, participant 37). Accountability of various stakeholders in IMCI implementation Districts had different management cadres assigned to perform similar IMCI clinical governance activities. For example, in district A, all monitoring and supervision of IMCI was performed by a district child health coordinator while in district B there were MCWH coordinators at sub-district level. As a consequence, the coordinator in district A supervised >40 clinics, while in district B, each sub-district coordinator was responsible for 8–10 clinics. IMCI related roles and responsibilities of different cadres of managers were not clearly defined with task performance left to individual discretion. For example, the same clinic would receive an IMCI supervisory visit by both a district clinical specialist team (DCST) member and MCWH coordinator on different days, using different checklists. On the other hand, another clinic in the same district would not receive any visits. This discordance was aggravated by poor communication between management cadres. We are not on the same page because they themselves (DCSTs and area managers) are interested in figures. It’s like the coordination between them and me is not that good so I think we need to sit down because they are doing their own thing, I’m doing my own thing. We are basically doing the same things (Child health coordinator, participant 17, District A). There is no post visit discussion between us (area manager and MCWH coordinator) because if you discuss it, you are offended, you are in my territory and it is not your business. Yet at the end of the day it is the PHC component (Area manager, participant 31, sub-district 2, district B). Discussion Considerable differences were observed between the two districts. District B offered superior IMCI implementation. Key enablers (in district B) included standardized recording forms, dedicated MCWH coordinators at sub-district level, regular, stringent monitoring and better support for staff. District A’s performance was hindered by negative perceptions of IMCI held by PCCD trained nurses, reduced competence due to unnecessary rotation, poor monitoring and supervision, fragmented governance and lack of accountability for programme outcomes. Many of the deficiencies identified mirror those identified globally in previous studies (Kiplagat et al. 2014; Huicho et al. 2005; Ahmed et al. 2010). This study, however, offers more detail about the root causes for these deficiencies. We recognized at the outset that confining the analysis to issues specific to IMCI without considering the broader health system deficiencies would restrict our ability to design a meaningful response. The absence of a coherent strategy to manage sick children was overt. Nurses felt obliged to select between IMCI and the PCCD approaches and lacked support to adequately implement key components such as HIV, TB and malnutrition. Negative perceptions of IMCI were fuelled by perceived, but non-validated, deficiencies such as staff shortages. Similar findings have been reported in Pakistan where district and provincial managers reported lack of appreciation of the IMCI strategy constraining implementation (Pradhan et al. 2013). The criticism that IMCI is too simplistic was shared by practitioners in Tanzania and Kenya (Kiplagat et al. 2014; Silali 2014). Implementers have placed substantial emphasis globally on IMCI training believing that this is the critical factor driving guideline adherence. This belief was reflected in our study in the IMCI quarterly report prioritizing submission of training related data. Our study confirms the well-accepted notion that training alone cannot ensure quality care and does not guarantee satisfactory implementation (Pariyo et al. 2005). Failure to define staffing needs for child health (and IMCI) delivery and poor workload allocation emerged as major barriers to IMCI implementation. Previous studies also confirmed issues such as staff shortages, rotation and frequent turnover as barriers to IMCI implementation (Huicho et al. 2005; Prosper et al. 2009; Tulloch 1999; Adam et al. 2005). Contrary to evidence from other countries, (Pariyo et al. 2005; Ahmed et al. 2010) we did not find the availability of medical technology, supplies, drugs and equipment as a major barrier to implementation in both districts. Global evidence indicates that quality of care is a key predictor of child mortality (Binkin et al. 2011; Rohde et al. 2008). It is well established that adherence to IMCI guidelines improves the quality of clinical care provided to sick children by changing case management practices (Armstrong et al. 2004; Chopra et al. 2005). Monitoring the quality of care should thus be prominent in any child health information system. Although WHO has set guidelines and standard indicators on monitoring IMCI, it does not provide details of how IMCI implementation (including quality of care) should be monitored and how results can be used for improving performance (WHO 1999). Hence, it is left to countries to decide how they will collect and utilize monitoring information. Our study highlights critical gaps in this process in South Africa where we identified that programmatic data collected at clinic and sub-district levels was not shared and utilized at higher levels (district, province and national) for strategic planning, decision making and improving implementation. This was best reflected in words of a national key informant stating ‘we don’t know what is happening with IMCI in this country, we are just speculating and assuming.’ The recently conducted IMNCI global strategic review echoes similar findings where only 33% of all IMCI implementing countries had a comprehensive monitoring and evaluation plan (Costello and Dalglish 2016). The benefits of an integrated approach for managing children are well-established (Tulloch 1999). Although IMCI was introduced as a response to uncoordinated vertical health programmes, it struggles to effectively coexist with them. Global studies have highlighted that vertical programmes are often prioritized and attract ring-fenced funding, dedicated governance cadres and monitoring systems (Ahmed et al. 2010). Evidence from Tanzania shows that well established and heavily funded vertical programmes such as EPI and malaria, considered IMCI as a threat to their existence (Prosper et al. 2009). We confirmed programme siloed governance at district, province and national levels. Dedicated managers and coordinators and clear monitoring and evaluation indicators existed for vertical programmes such as EPI and TB, while IMCI was ignored. The issue of co-existence of integrated and vertical approaches has been an ongoing international debate with few tangible solutions. Policy makers need to consider mechanisms at both strategic and operational (service delivery) levels to improve the links (Atun et al. 2008). While there are global studies that have explored integration in HIV, mental health and communicable diseases, there is a lack of evidence on integration in context of child health (Atun et al. 2008). Since IMCI is not a vertical programme, it is more likely to have management cadres (coordinators, supervisors, managers) with shared responsibilities. Each plays a critical role in ensuring that implementation activities such as monitoring, supervision, health facility visits, training etc. are conducted appropriately. More efficient execution of these activities in district B could mainly be attributed to better workforce planning and effective allocation of duties. The asymmetry between the two districts can be ascribed to lack of clear, consistent guidelines from national and provincial levels. Poor leadership and accountability at various levels of health system have previously been highlighted in the context of child health in South Africa—mainly in the failure of responsibility delegation, budget allocation and guideline setting (Saloojee 2009). In addition, there is no standardized organizational structure for district health services in South Africa, which adversely affects the capacity of health management cadres (Gilson et al. 2011). Our study adds to the global knowledge base on effect of poor governance and lack of a clear administrative structure at a district level on systematic planning and IMCI implementation activities (Huicho et al. 2005). Strengths and limitations Due to its qualitative nature, our study offers an in-depth analysis of various health system factors affecting IMCI implementation and answers the ‘why’ and ‘how’ aspects which have not been well described in previous research. Moreover, the use of health system dynamic framework allowed us to take a comprehensive view of enablers and barriers affecting IMCI implementation. Since the study sample included only two districts, the results reflect a local situation and cannot be extrapolated to other districts of South Africa or beyond since there is marked variability in how provinces and districts deliver child health services with no standardized approach. Nevertheless, we believe that systemic issues such as poor monitoring and supervision, limited governance and human resource challenges are replicated in many settings. We excluded health financing as a building block and acknowledge the need to explore it through dedicated research in future. Two integral components of IMCI implementation—community engagement and health worker training—were not specifically addressed. Finally, we recognize that the study findings are based on participant perception and may be biased. An observational study would have strengthened our findings. Nevertheless, many of the participants’ reflections strongly validated our own perspectives garnered through long-term engagement with the local and national health system. Key lessons for child health policy and practice As the global health community prepares to re-direct efforts to achieving the Sustainable Development Goals (SDGs), the quality of PHC services for children under 5 years remains a critical issue for many countries, including South Africa. Considering that there is no major alternate approach, investment in improving IMCI implementation is warranted. This study comes at a critical time when the global child health community is thinking about repositioning IMCI under the SDGs and the Global Strategy for Women, Children and Adolescents (2016–2030) (Costello and Dalglish 2016). It confirms that due to its integrated design, IMCI’s success is bound to depend on the strength of key components of a health system (building blocks), particularly in resource constrained environments. Taking the South African example, our study not only unpacked specific elements of these health system building blocks, which affect implementation, but also identified reasons behind their existence. Based on the study findings, we offer a few specific, actionable recommendations for child PHC (including IMCI) related policy, practice and research, which can be implemented at various levels of a health system to achieve biggest gains (Table 5). Considering that 20 years on, countries still face similar health system issues affecting IMCI implementation, we believe that these suggestions are of relevance to settings beyond our borders. However, they require testing of their feasibility, scalability and wider relevance before adoption. Table 5. Recommendations to strengthen health systems for effective IMCI implementation (in terms of policy and practice) WHO building blocks of health system  IMCI specific recommendations  Broader health system recommendations  Service delivery  More child health conditions need to be added to IMCI package (in alignment with country specific burden of disease, e.g. trauma and skin conditions)  Repackage IMCI under the umbrella of a broader primary health service delivery for children by defining the minimum package of services (including IMCI and PCCD) which a sick child should receive at every clinic visit and clearly communicate this to all implementers.  Human resources  Prioritize mentoring and support to practitioners immediately post training as well as on a regular basis.  Policy guidance on appropriate duty allocation and rotation of health workers in health facilities. This must ensure allocating adequate number of trained staff for effective execution of primary health services for children on a daily basis considering the patient load, clinic infrastructure and operating hours. (especially for community health centres)  Clinic managers need to ensure allocation of practitioners to IMCI duty station immediately post training as well as on a regular basis.  Conduct regular IMCI refresher trainings.  Additional options for building health workers’ capacity need to be considered such as self directed learning and peer learning.1  Auxiliary nursing cadres such as enrolled nurses can be utilized to deliver components of IMCI especially health prevention and promotion activities such as growth monitoring, immunization and nutritional counselling.  Medical technology  In centralized pharmacies dispensing IMCI drugs, pharmacists need to demonstrate first dose administration and counsel the caregiver on drug use and storage.    Health information system  Policy guidance on how to integrate IMCI monitoring with routine child health information system. This must include identification of a limited number of IMCI specific indicators to be tracked (including indicators suggested by WHO), who should collect them, how frequently, what data sources to be used at clinic, sub-district, province and national levels, data dissemination, interpretation and establishing a system of providing feedback to implementers.  Quality of care provided to sick children to be routinely monitored in order to assess practitioners’ adherence to IMCI guidelines.  This can be done by conducting regular audits of child health records in addition to observation of IMCI case management followed by providing feedback to child health practitioners on a regular basis.  Data sources reflecting quality of care and other indicators need to be regularly collated, disseminated and utilized optimally for decision making not only at district level but also at provincial and national levels.  Mobile phone and electronic devices for real time data capture can be considered for collecting data on monitoring1.  Standardized child health records with IMCI forms to be compulsorily filled by all practitioners on every sick child encounter.  Red flag poor performing provinces or districts that need additional support for improving IMCI implementation/child PHC services, based on routinely collected monitoring data.  Leadership and governance  Mapping of existing managerial cadres available for PHC/MCWH/child health on the basis of which, adequate number of MCWH coordinators/supervisors must be allocated at sub-district and district levels (according to catchment area and number of clinics) who will be responsible for coordinating and implementing IMCI activities such as supervision and monitoring on a regular basis.  Better coordination and communication between vertical child health programmes (such as HIV, TB and EPI) and IMCI, regarding strategic planning and implementation of activities.  Options such as integrated supervision and monitoring can be considered.  Clear and consistent policy guideline in terms of workforce planning for managerial/supervisory cadres and allocation of duties for child PHC services including IMCI activities. These cadres also need to be sensitized and trained on case management as well as supervision and monitoring elements.  This should include clear directions on who should supervise/coordinate, how many clinics, which checklist to be used and frequency of supervisory and monitoring visits.  WHO building blocks of health system  IMCI specific recommendations  Broader health system recommendations  Service delivery  More child health conditions need to be added to IMCI package (in alignment with country specific burden of disease, e.g. trauma and skin conditions)  Repackage IMCI under the umbrella of a broader primary health service delivery for children by defining the minimum package of services (including IMCI and PCCD) which a sick child should receive at every clinic visit and clearly communicate this to all implementers.  Human resources  Prioritize mentoring and support to practitioners immediately post training as well as on a regular basis.  Policy guidance on appropriate duty allocation and rotation of health workers in health facilities. This must ensure allocating adequate number of trained staff for effective execution of primary health services for children on a daily basis considering the patient load, clinic infrastructure and operating hours. (especially for community health centres)  Clinic managers need to ensure allocation of practitioners to IMCI duty station immediately post training as well as on a regular basis.  Conduct regular IMCI refresher trainings.  Additional options for building health workers’ capacity need to be considered such as self directed learning and peer learning.1  Auxiliary nursing cadres such as enrolled nurses can be utilized to deliver components of IMCI especially health prevention and promotion activities such as growth monitoring, immunization and nutritional counselling.  Medical technology  In centralized pharmacies dispensing IMCI drugs, pharmacists need to demonstrate first dose administration and counsel the caregiver on drug use and storage.    Health information system  Policy guidance on how to integrate IMCI monitoring with routine child health information system. This must include identification of a limited number of IMCI specific indicators to be tracked (including indicators suggested by WHO), who should collect them, how frequently, what data sources to be used at clinic, sub-district, province and national levels, data dissemination, interpretation and establishing a system of providing feedback to implementers.  Quality of care provided to sick children to be routinely monitored in order to assess practitioners’ adherence to IMCI guidelines.  This can be done by conducting regular audits of child health records in addition to observation of IMCI case management followed by providing feedback to child health practitioners on a regular basis.  Data sources reflecting quality of care and other indicators need to be regularly collated, disseminated and utilized optimally for decision making not only at district level but also at provincial and national levels.  Mobile phone and electronic devices for real time data capture can be considered for collecting data on monitoring1.  Standardized child health records with IMCI forms to be compulsorily filled by all practitioners on every sick child encounter.  Red flag poor performing provinces or districts that need additional support for improving IMCI implementation/child PHC services, based on routinely collected monitoring data.  Leadership and governance  Mapping of existing managerial cadres available for PHC/MCWH/child health on the basis of which, adequate number of MCWH coordinators/supervisors must be allocated at sub-district and district levels (according to catchment area and number of clinics) who will be responsible for coordinating and implementing IMCI activities such as supervision and monitoring on a regular basis.  Better coordination and communication between vertical child health programmes (such as HIV, TB and EPI) and IMCI, regarding strategic planning and implementation of activities.  Options such as integrated supervision and monitoring can be considered.  Clear and consistent policy guideline in terms of workforce planning for managerial/supervisory cadres and allocation of duties for child PHC services including IMCI activities. These cadres also need to be sensitized and trained on case management as well as supervision and monitoring elements.  This should include clear directions on who should supervise/coordinate, how many clinics, which checklist to be used and frequency of supervisory and monitoring visits.  Firstly, the role of IMCI and other coexisting approaches (such as PCCD) needs to be clearly defined within a PHC service package for children. This will reduce the ambiguity and conflict which practitioners face while managing a sick child. Secondly, to retain competence of nurses in practicing IMCI, resources should be diverted from training to maintaining competence of already trained nurses. This can be achieved by a focus on post-training supervision and mentoring, duty allocation and rotation policies, considering the heterogeneity of clinic organization and needs in a particular setting. Thirdly, to compensate for human resource deficiencies, we recommend that auxiliary nursing cadres be utilized to deliver the prevention and promotion components of IMCI. This threatens the holistic single provider delivery model, but could optimize resource use, allowing trained practitioners to focus on diagnostic and management tasks while auxiliaries complete less skill-intensive and time-consuming components. Fourthly, compulsory completion of standardized child health records (IMCI forms) would enable process compliance and facilitate audit. Successful integration of different services and programmes remains elusive, but warrants innovation and research. Finally, strong clinical governance is required for optimal workforce deployment and accountability extraction. While district management has a critical role in effective operational planning and execution of activities at local level, optimal IMCI implementation cannot be achieved without appropriate strategic planning and policy direction from provincial and national levels. Funding Faculty Research Committee (FRC) Individual research grant awarded by the University of the Witwatersrand, Johannesburg. (Grant no. 0012548438104512110500000000000000004988). Footnotes 1 Costello AM, Dalglish SL. 2016. Towards a Grand Convergence for child survival and health: A strategic review of options for the future building on lessons learnt from IMNCI. Geneva: WHO. Online at http://apps.who.int/iris/bitstream/10665/251855/1/WHO-MCA-16.04-eng.pdf, accessed 20 Jan 2016. Acknowledgements We thank the national, provincial and district health departments and all interviewees for their participation in this research study. We acknowledge the financial assistance received from the University of the Witwatersrand, Johannesburg and recognize Simbareshe Tevera for his assistance in transcribing the interview audio tapes. Ethical approval for the study was obtained from the University of the Witwatersrand’s Human Research Ethics Committee (clearance number M141194). Conflict of interest statement. None declared. References Adam T, Amorim DG, Edwards SJ, Amaral J, Evans DB. 2005. Capacity constraints to the adoption of new interventions: consultation time and the Integrated Management of Childhood Illness in Brazil. Health Policy and Planning  20: i49– 57. Google Scholar CrossRef Search ADS PubMed  Ahmed HM, Mitchell M, Hedt B. 2010. National implementation of Integrated Management of Childhood Illness (IMCI): policy constraints and strategies. Health Policy  96: 128– 33. Google Scholar CrossRef Search ADS PubMed  Arifeen SE, Blum LS, Hoque DME. et al.   2004. Integrated Management of Childhood Illness (IMCI) in Bangladesh: early findings from a cluster-randomised study. The Lancet  364: 1595– 602. Google Scholar CrossRef Search ADS   Armstrong SJ, Bryce J, De Savigny D, Lambrechts T, Mbuya C. 2004. The effect of Integrated Management of Childhood Illness on observed quality of care of under-fives in rural Tanzania. Health Policy and Planning  19: 1– 10. Google Scholar CrossRef Search ADS PubMed  Atun RA, Bennett S, Duran A. 2008. When do vertical (stand-alone) programmes have a place in health systems ? Geneva: World Health Organization. Online at http://www.who.int/management/district/services/WhenDoVerticalProgrammesPlaceHealthSystems.pdf, accessed 10 April 2016. Binkin N, Chopra M, Simen-Kapeu A, Westhof D. 2011. Do improvements in outreach, clinical, and family and community-based services predict improvements in child survival? An analysis of serial cross-sectional national surveys. BMC Public Health  11: 456. Google Scholar CrossRef Search ADS PubMed  Bryce J, Victora CG, Habicht JP, Black RE, Scherpbier RW. 2005. Programmatic pathways to child survival: results of a multi-country evaluation of Integrated Management of Childhood Illness. Health Policy and Planning  20: i5– i17. Google Scholar CrossRef Search ADS PubMed  Callaghan M, Ford N, Schneider H. 2010. A systematic review of task- shifting for HIV treatment and care in Africa. Human Resources for Health  8: 1– 9. Google Scholar CrossRef Search ADS PubMed  Chopra M, Patel S, Cloete K, Sanders D, Peterson S. 2005. Effect of an IMCI intervention on quality of care across four districts in Cape Town, South Africa. Archives of Disease in Childhood  90: 397– 401. Google Scholar CrossRef Search ADS PubMed  Costello AM, Dalglish SL. 2016. Towards a Grand Convergence for child survival and health: a strategic review of options for the future building on lessons learnt from IMNCI . Geneva: WHO. Online at http://apps.who.int/iris/bitstream/10665/251855/1/WHO-MCA-16.04-eng.pdf, accessed 20 January 2016. Daviaud E, Subedar H. 2012. Staffing Norms for Primary Health Care in the context of PHC Re-engineering- Report to National Department of Health . Cape Town: Medical Research Council. Online at http://www.mrc.ac.za/healthsystems/StaffingNorms.pdf, accessed 4 October 2017. De Savigny D, Adam T. 2009. Systems thinking for health systems strengthening . Geneva: Alliance for Health Policy and Systems Research; World Health Organization. Online at http://apps.who.int/iris/bitstream/10665/44204/1/9789241563895_eng.pdf, accessed 5 May 2016. District Department of Health. 2012. Child health Situational Analysis . Johannesburg: DCST. Unpublished document. Gera T, Shah D, Garner P, Richardson M, Sachdev HS. 2016. Integrated management of childhood illness (IMCI) strategy for children under five. The Cochrane Database of Systematic Review  6: 1– 60. Gilson L, Daire J, Patharath A, English R. 2011. Leadership and governance within the South African health system. South African Health Review  2016: 69– 80. Hagopian A, Mohanty MK, Das A, House PJ. 2012. Applying WHO's ′workforce indicators of staffing need′ (WISN) method to calculate the health worker requirements for India's maternal and child health service guarantees in Orissa State. Health Policy and Planning  27: 11– 8. Google Scholar CrossRef Search ADS PubMed  Horwood C, Vermaak K, Rollins N. et al.   2009a. An evaluation of the quality of IMCI assessments among IMCI trained health workers in South Africa. PLoS One  4: e5937. Google Scholar CrossRef Search ADS   Horwood C, Vermaak K, Rollins N. et al.   2009b. Paediatric HIV management at primary care level: an evaluation of the integrated management of childhood illness (IMCI) guidelines for HIV. BMC Pediatrics  9: 59. Google Scholar CrossRef Search ADS   Horwood C, Voce A, Vermaak K, Rollins N, Qazi S. 2009c. Experiences of training and implementation of integrated management of childhood illness (IMCI) in South Africa: a qualitative evaluation of the IMCI case management training course. BMC Pediatrics  9: 62. Google Scholar CrossRef Search ADS   Huicho L, Dávila M, Campos M. et al.   2005. Scaling up integrated management of childhood illness to the national level: achievements and challenges in Peru. Health Policy and Planning  20: 14– 24. Google Scholar CrossRef Search ADS PubMed  Jonker L, Stellenberg EL. 2014. Missed opportunities in child healthcare. African Journal of Primary Health Care & Family Medicine  6: 8. Google Scholar CrossRef Search ADS   Kiplagat A, Musto R, Mwizamholya D, Morona D. 2014. Factors influencing the implementation of integrated management of childhood illness (IMCI) by healthcare workers at public health centers & dispensaries in Mwanza, Tanzania. BMC Public Health  14: 277. Google Scholar CrossRef Search ADS PubMed  Lange S, Mwisongo A, Mæstad O. 2014. Why don't clinicians adhere more consistently to guidelines for the Integrated Management of Childhood Illness (IMCI)? Social Science & Medicine  104: 56– 63. Google Scholar CrossRef Search ADS   Loening W, Bamford L. 2012. Integrated Management of Childhood Illness. In: Westwood T, Kibel M, Saloojee H(eds). Child Health for All . 5th edn. Cape Town: Oxford University Press, 200– 205. Makgetla N, Fotoyi A. 2016. Real Economy Bulletin 2016—Gauteng Provincial Review. Online at: https://www.tips.org.za/images/The_REB_Provincial_Review_2016_Gauteng.pdf, accessed 28 March 2017. MaxQDA. 2016. MaxQDA—The art of data analysis. Online at http://www.maxqda.com/legalinfo, accessed 20 November 2015. Mazaleni N. 2009. Strengthening community-based child health services in South Africa. South African Child Gauge. 71-76. Online at http://pmg-assets.s3-website-eu-west-1.amazonaws.com/docs/100811childgauge.pdf#page=73, accessed 26 May 2016. Mulaudzi MC. 2015. Adherence to case management guidelines of Integrated Management of Childhood Illness (IMCI) by healthcare workers in Tshwane, South Africa. South African Journal of Child Health  9: 89– 92. Google Scholar CrossRef Search ADS   National Department of Health (NDOH). 2014. Integrated management of childhood illnesses (IMCI) chart booklet . Pretoria: National Department of Health. Online at http://paediatrics.ukzn.ac.za/Files/2014%20IMCI%20CHART%20BOOKLET%20Final.pdf, accessed 6 April 2015. National Department of Health (NDOH). 2012. National strategic plan for maternal, newborn, child, women, adolescent health and nutrition (MNCWAH-N 2012-2016).  Pretoria: National Department of Health. National Department of Health (NDOH). 2016. Mid term review of National strategic plan for maternal, newborn, child, women, adolescent health and nutrition (MNCWAHN, 2012-2016) . Pretoria: National Department of Health. National Department of Health (NDOH). 2001. Primary health care package for South Africa - A set of norms and standards . Pretoria: National Department of Health. Online at https://www.westerncape.gov.za/general-publication/primary-health-care-package-sa-set-norms-and-standards, accessed 7 April 2016. Nyamtema AS, Urassa DP, Massawe S, Massawe A, Lindmark G. 2008. Staffing Needs for Quality Perinatal Care in Tanzania. African Journal of Reproductive Health  12: 113– 24. Google Scholar PubMed  Pariyo GW, Gouws E, Bryce J, Burnham G. 2005. Improving facility-based care for sick children in Uganda: training is not enough. Health Policy and Planning  20: i58– 68. Google Scholar CrossRef Search ADS PubMed  Pradhan NA, Rizvi N, Sami N, Gul X. 2013. Insight into implementation of facility-based integrated management of childhood illness strategy in a rural district of Sindh, Pakistan. Global Health Action  6: 20086. Google Scholar CrossRef Search ADS PubMed  Prosper H, Macha J, Borghi J. 2009. Implementation of integrated management of childhood illness in Tanzania: success and challenges. Consortium for Research on Equitable Health Systems. Online at http://citeseerx.ist.psu.edu/viewdoc/download? doi=10.1.1.690.9032&rep=rep1&type=pdf, accessed 13 May 2016. Rakha MA, Abdelmoneim ANM, Farhoud S et al.   2013. Does implementation of the IMCI strategy have an impact on child mortality? A retrospective analysis of routine data from Egypt. BMJ Open  3: e001852. Google Scholar CrossRef Search ADS PubMed  Rohde J, Cousens S, Chopra M et al.   2008. 30 years after Alma-Ata: has primary health care worked in countries? The Lancet  372: 950– 61. Google Scholar CrossRef Search ADS   Saloojee H. 2009. Managing resources and building capacity in the context of child health. Child Gauge  2009: 64– 70. Saloojee H, Bamford L. 2006. Key childhood health promotion and disease prevention programmes. South African Health Review  181– 202. Online at https://www.popline.org/node/196340, accessed 10 September 2017. Silali MB. 2014. Utilization of Integrated Management of Childhood Illnesses IMCI for Child health in Western Kenya. Journal of Biology, Agriculture and Healthcare  4: 102– 13. Soweto trust for nurse clinical training. 2012. Primary clinical care manual— A practical guide for primary health care personnel in the clinical setting. 7th ed. Johannesburg: Soweto trust for nurse clinical training. Srivastava A, Thomson SB. 2009. Framework analysis: a qualitative methodology for applied policy research. Journal of Administration & Governance  4: 72– 79 Online at http://research.apc.org/images/a/ad/Framework_analysis.pdf, accessed 4 June 2016. Thandrayen K. 2010. The quality of child health services offered at primary health care clinics in Johannesburg. South African Journal of Child Health  4: 73– 7. Tulloch J. 1999. Integrated approach to child health in developing countries. The Lancet  354: SII16– 20. Google Scholar CrossRef Search ADS   United Nations. 2015. The Global strategy for women's, children's and adolescents' health (2016-2030). Every woman every child. Online at: http://www.who.int/pmnch/media/events/2015/gs_2016_30.pdf, accessed 7 June 2016. Van Olmen J, Criel B, Van Damme W et al.   2012. Analysing health system dynamics–a framework. Studies in Health Services Organization and Policy  2: 1. WHO. 2004. The Analytic review of the Integrated Management of Childhood Illness strategy - Final Report . Geneva: World Health Organization. Online at: http://apps.who.int/iris/bitstream/10665/42965/1/9241591730.pdf, accessed 16 April 2016. WHO. 1999. IMCI Information—Monitoring and Evaluation . Geneva: World Health Organization. Online at http://apps.who.int/iris/bitstream/10665/65002/12/WHO_CHS_CAH_98.1K_eng.pdf, accessed 18 April 2016. WHO. 2007. Everybody's business–strengthening health systems to improve health outcomes: WHO's framework for action . Geneva: World Health Organization, p.3. Online at: http://apps.who.int/iris/bitstream/10665/43918/1/9789241596077_eng.pdf, accessed 12 January 2016. WHO. 2009. IMCI [Online]. Online at http://www.who.int/maternal_child_adolescent/topics/child/imci/en/, accessed 25th October 2014. WHO. 2010. Workload Indicators of Staffing Needs (WISN) - Users Manual . Geneva: World Health Organization. Online at http://www.who.int/hrh/resources/WISN_Eng_UsersManual.pdf? ua=1, accessed 14 August 2016. © The Author 2017. Published by Oxford University Press in association with The London School of Hygiene and Tropical Medicine. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com

Journal

Health Policy and PlanningOxford University Press

Published: Mar 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 12 million articles from more than
10,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Unlimited reading

Read as many articles as you need. Full articles with original layout, charts and figures. Read online, from anywhere.

Stay up to date

Keep up with your field with Personalized Recommendations and Follow Journals to get automatic updates.

Organize your research

It’s easy to organize your research with our built-in tools.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

Monthly Plan

  • Read unlimited articles
  • Personalized recommendations
  • No expiration
  • Print 20 pages per month
  • 20% off on PDF purchases
  • Organize your research
  • Get updates on your journals and topic searches

$49/month

Start Free Trial

14-day Free Trial

Best Deal — 39% off

Annual Plan

  • All the features of the Professional Plan, but for 39% off!
  • Billed annually
  • No expiration
  • For the normal price of 10 articles elsewhere, you get one full year of unlimited access to articles.

$588

$360/year

billed annually
Start Free Trial

14-day Free Trial