Simple Summary Kenya has experienced an increased disease burden due to cholera outbreaks characterized by continuous transmission in the refugee camp settings of Garissa and Turkana, and sporadic transmission in cities such as Nairobi and Mombasa. However, while inquiry has been made regarding how the health sector has responded to and how populations have been impacted by epidemics such as cholera, not enough attention has been paid to causal relationships between the unique components of the health systems leading to an inadequate understanding of how the building blocks of the health systems interact in disease surveillance and epidemic response. Consequently, this research effort focuses on examining the health sector responses to these epidemics, developing a systems-theory-based description of the said responses, and based on the results, provide recommendations that may help break the epidemic's continuous and cyclical nature. The study will be conducted in Turkana and Garissa in Kenya, as well as in Nairobi. These are counties that have diverse levels of health system sophistication and recent experiences of cholera epidemics. Moreover, the study draws on several systems analytic frameworks to determine the actors and their decisions, as well as their inter-relationships, linkages and dependencies. Thus, the study appropriates systems-oriented research techniques and applies them in a novel fashion to the assessment of health service delivery. The findings of this study are expected to guide interventions aimed at improving disease surveillance and epidemic response. The study will also act as a primer for the research community by assessing the utility of systems-thinking approaches in such settings, and subsequently opening a related research agenda.

Health systems in low and middle income countries (LMIC) are increasingly pluralistic, involving a wide mix of public, not-for-profit and for-profit providers. Regulation should be a key foundation of the Government's stewardship role of these heterogeneous facilities, but performance of this function is generally weak, with serious consequence for patient safety and quality of care. There has been little evaluation of strategies to strengthen regulation in LMIC, a notable exception being the Kenya Patient Safety Impact Evaluation (KePSIE), a collaboration between the Kenyan Ministry of Health and the World Bank. This randomised controlled trial is assessing the impact of a set of innovative regulatory interventions in public and private facilities in 3 Kenyan counties. These comprise the use of the Joint Health Inspections Checklist (JHIC), which synthesises the areas covered by all the regulatory Boards and Councils; increased inspection frequency; risk-based inspections where warnings, sanctions and time to re-inspection depend on inspection scores; and display of regulatory results outside facilities. The KePSIE trial will provide a rigorous quantitative assessment of these regulatory strategies. However, such regulatory interventions are highly complex, requiring behaviour change by regulatory managers, front line inspection staff, health facilities, and clients. To understand the effectiveness of the intervention and why aspects do (or do not) work it is therefore essential to investigate the mechanisms and processes involved, the degree to which they are implemented effectively, and the reasons for the level of implementation observed. Other important dimensions include legitimacy, potential for corruption and regulatory costs. We will therefore conduct a companion study to enhance our understanding of the effectiveness of these regulatory innovations, and to consider their wider implications for the creation of a cost-effective, sustainable and equitable regulatory system. The research will begin with a review of key documents and records related to regulatory implementation, with the review updated periodically during the study. We will also systematically collate media articles from Kenyan newspapers and relevant social media concerning health facility regulation. Following a period of familiarisation with regulation by shadowing inspectors on their regular duties, we will undertake a set of in-depth interviews (IDIs) with a wide range of stakeholders including national regulators, county and sub-county managers, inspectors, facility owners/ staff, and Community Health Committee members. IDIs will cover their perceptions and experiences of regulatory implementation under the current regulatory system and the KePSIE regulatory innovations; their views on the legitimacy of regulatory systems, in terms of fairness and acceptability; perceptions of corruption; and perceptions of community views on facility regulation. We will also conduct patient exit interviews to assess community member understanding of the regulatory scores displayed outside facilities. Finally, we will assess the incremental costs of the KePSIE interventions compared to those of the current regulatory system, from the perspectives of both the regulating agencies and the health facilities. The results are expected to make an important contribution to the limited evidence base on regulation and regulatory reform. The findings will be of substantial benefit to those concerned with regulatory reform and the improvement of quality and safety more generally in Kenya and other LMIC settings.

With large numbers of people in slums seeking care through the private sector, it is important to develop tools to help providers in these clinics improve the quality of their services. For example, documenting the need for an antibiotic helps reduce the number of cases of unnecessary antibiotic prescriptions. Clinical practice guidelines (CPGs) are tools developed to help doctors and nurses give evidence-based care. These are however not easy to use in a patient-facing scenario (e.g. a doctor looking through a manual when the patient is seated in front of him/her). They also need to be tweaked to be relevant to the local context (e.g. is the first line drug available or affordable?). For CPGs to be relevant in low-resource settings it is important to address multiple, if not all, barriers to using guidelines, but in a manner that does not strain limited resources. Our intervention involves working with clinical care providers and developing templates (think checklist) that can be used while they are seeing the patient. The templates take the form of rubber stamps that can be printed into the paper case sheet (e.g. if a woman presents with increased frequency of urination, the clinician stamps the Urinary Tract Infection template into her case sheet). This template is both a guide to what questions to ask the patient and how to manage UTIs. Other illnesses get other templates, but a set of 6-8 templates covers the majority of patients walking into primary care clinics. Importantly these templates are easy to digitise and analyse. They are in the form of a multiple choice exam paper where bubbles need to be shaded. A cell phone image of a filled-in template can quickly give us data on how the case was managed without revealing patient identity. Being rubber stamps we avoid the need to keep track of multiple printed sheets of paper. We also avoid wastage when guidelines change because changing the rubber engraving on the stamp is simple and cheap. We now have a tool to easily monitor the clinician's work, check for quality, and work with them if there are reasons to deviate from the guidelines by during regular feedback sessions. The intervention is being used in two slum clinics in Nairobi with great initial responses. We now want to study this intervention in a set of 10 different private sector clinics in Nairobi's slums. We would like to test if this intervention is: a) Usable - different clinics have different priorities and attitudes and we need to be sure that the intervention poses no big challenges b) Effective - does the intervention actually improve clinical practice (e.g. by reducing unnecessary antibiotic prescription)? c) Sustainable - how much does it cost for us to support these clinics with tools and feedback? Can the clinics afford to pay us for this service? d) Scalable - is there a realistic chance for us to roll this intervention out at national (or even provincial) level? If successful the intervention has the potential to change how healthcare is delivered in low-resource settings. More and more people are seeking care in the private sector, but very few regulations, services and tools exist to ensure that care in the private sector is of high quality. We hope to make a significant impact in the quality of care that is delivered to the poor.

In 2009, there were 42,000 newborn deaths and they accounted for 40% of all deaths among children under 5 Kenya. This high neonatal mortality is a major reason why Kenya is not succeeding in its battle to reduce child deaths in line with stated targets. Recognising this, the Ministry of Health has started to focus on improving newborn (and maternal) health with strategies aimed at communities and small clinics. However, sick or vulnerable newborns will often require inpatient care in referral facilities from skilled workers with access to basic technologies. Interventions typically delivered at this level include, for example, fluids or feeds for those unable to suck or oxygen for respiratory support. Such interventions require carers to perform the same, time-consuming tasks multiple times per day for many days. Shortage of skilled health workers often means these services are inadequately delivered, potentially delaying or preventing recovery. We are planning research that will establish: the potential burden of severe neonatal illness; what existing infrastructure and human resource capacity is available supporting access for this population; utilisation of these services; and the quality of existing nursing care services. We will do this focusing on Nairobi's population of 5 million, many of whom are very poor. With a focus on universal coverage and neonatal care meeting agreed standards, this work will provide the basis for estimating the gap between available and needed services (Gap 1) and the quality gap between existing and desired services (Gap 2). In partnership with important stakeholders, we will explore how a low-income country might best tackle health workforce challenges to close these gaps and improve provision of essential nursing care to all sick newborn babies in an affordable and efficient way. This ultimate aim of research is driven by the fact that salary costs are a major proportion of total health care costs. One option will therefore be to explore alternatives to employing professional nurses if necessary interventions can be effectively provided by other groups under the supervision of professionals - an approach known as task-shifting. Although task-shifting sounds a simple solution, it may not always be. Failure to consider national regulations, the opinions of important professionals, managers or parents may lead to the approach being rejected or failing. Taking account of the local situation may be particularly important when those being cared for are sick, newborn babies and when day to day care has traditionally been given by professional, even specialist nurses. First, therefore, we will define with the major groups what forms of care should be available to all, learn what regulations exist on providing care, and consider the concerns of major groups with respect to task-shifting. We will examine carefully all the things that nurses have to do in a range of different facilities, explore with experts which tasks may be simple enough for others to do, and examine whether there is time to do all the essential care tasks. We will estimate how much need there is for neonatal nursing care in Nairobi and the gap between what is available and what is needed. Using all these data we will explore how many new staff might be needed to improve the delivery of essential care for all newborns in need. We will also undertake preliminary work to explore the costs of meeting this need using extra professional nurses or if tasks were shifted to other, lower cost staff. Possible roles for lower cost staff will be informed by work examining what tasks to shift and how they might fit within existing patterns of providing care. All this work will be conducted with the major decision makers in health, health professionals and parents to develop options sensitive to local conditions. Based on this body of work we aim to develop a task-shifting approach that can be tested in Kenya in the future.

Artificial Intelligence has altered and in future going to influence the way our economies and societies are organised and function. The economic systems are increasingly being digitalised which has generated the growing chorus around digital economy. Digital data-fuelled devices be it in our homes or workplaces, are shaping human interactions. The machines of today are going to define the human life of tomorrow. While the bold proclamations about digital technologies and their development impacts have received extensive coverage, it is rarely acknowledged that behind most contemporary digital transformations and advanced digital technologies is human labour. Just like the profitable commodities of earlier decades (oil, diamond, gold), contemporary digital products (autonomous vehicles, machine learning systems, next-generation search engines) are sourced and developed by workers in the low and middle-income regions. It is this behind-the-scene human labour that faces uncertain future. The centrality of labour from the Global South in some of these technologies raises critical questions around the new division of labour, developmental impacts for workers, and what the future would look like for workers on the continent and also other low and middle-income regions. The project explores these issues about our rapidly changing world of work and the implications on humans and regions that motors the global digital economy. The four-year project uses mixed methods to conduct a comprehensive empirical and theoretical assessment of behind the scene human labour in the Global South and leverage cross-country project partnership to develop analytical insights into the data work value chains of AI. By bringing theoretical sophistication and grounded empirical insights, the project's overall contribution is to unravel the geographies of data work and its implications for LMIs. The Planetary AI project will: (1) develop conceptual frameworks for studying data work value chains and labour market transformation, (2) generate empirical data on the scope of data work across the four case study countries and its developmental impacts (e.g. access to decent work), (3) produce research outcomes useful for academia, policy and practice. By combining discourse analysis, surveys, in-depth interviews, it captures the socio-political and economic transformations associated with the rise of data work across the Global South. Hence, the project contributes not only to the academic and policy debates surrounding AI, employment, and poverty reduction but will also be crucial in shaping the future rounds of digital-related development projects in low- and middle-income (LMI) countries. It also addresses policy-level changes required to protect the vulnerable sections of the society who are affected by the changing dynamics of technology and work. To make sure the project reaches a wider audience, it has developed multi-stakeholder networks of project partners. The ILO, business owners, policy makers, and data workers will be closely involved. Their expertise and networks will ensure it reaches actors who can influence the world of work. This is urgently needed as the risks of AI use has exposed the need for adequate regulatory reforms so that workers in the Global South are protected in their everyday lives. This project provides tools and evidence to ensure that such reforms are designed to strengthen policies related to labour standards, employment law, and social protection in the Global South and beyond.