ABSTRACT
Digital innovation has brought about huge paradigm shifts on how services are rendered to end- users in many sectors of the Kenya economy. Information and technology is seen to hold more potential within the healthcare sector by enhancing access to healthcare, cutting down operational cost and revolutionizing patient care. The purpose of this study was to develop a framework that will guide the implementation of the AI self-triage system within the healthcare sector in Kenya. Existing framework for the implementation of AI-based self-triage systems within healthcare sectors are meant for developed countries and are not adequate enough to be applied in developing country like Kenya. The study use embraced an epistemological view of things in undertaking grounded theory. The grounded theory approach was adopted in the collection and analysis of data for purpose of delivering a data driven model for implementation of AI-self-triage systems. Being a relative new field and topic, the study used snowball and saturation sampling to interview 10 respondents where aspects such as hindrances to quality healthcare, self-medication, experience in utilization of AI-based self-triage systems and recommendations on use and adoption of the self- triage tools were interrogated. Analysis of the data was by using NVivo software version 11 where codes, categories and themes were discovered from the underlying data. From data analysis where initial, axial and selective coding were embraced, the study uncovered five themes and 15 codes which formed the basis of memoing and the establishment of relationships. The identified themes include: hindrances to access to quality healthcare, prevalence of self-medication, inefficiencies at healthcare facilities, existing self-medication tools and their efficacy and requirement for adoption of AI self-triage systems. The developed framework highlights factors behind the preference of self-medication within the society and establishes the preference for AI-based self- medication. The framework takes into consideration end-user preferences for AI-self triage systems and how these preferences influence requirements for implementation of AI self-triage systems. From the framework prerequisites for the implementation of self-triage systems within the Kenyan health sector have been outlined. The model proposes key aspects such as policy, IT infrastructure, presence regulatory organizations and efficacy of AI self-triage systems as key requirements for their implementation. Findings of the study may in future influence policy formulation and shift the focus of digital innovation within healthcare from operational oriented to patient-centered systems and tools. Proper implementation of AI-based self-triage tools within the health sector may promote safe self-medication and enhance access to quality healthcare.
TABLE OF CONTENTS
Declaration ii
Acknowledgement iii
Dedication iv
List of Tables iv
List of Figures v
Abstract vi
CHAPTER ONE
INTRODUCTION
1.1 Introduction/Background 1
1.2 Statement of the Problem 2
1.3 Purpose of the Study 3
1.4 Objectives of the Study 3
1.5 Limitation of the Study 4
1.6 Research Questions 4
1.7 Justification 4
CHAPTER TWO
LITERATURE REVIEW
2.1 Introduction 6
2.2 Universal Health Coverage (UHC) Ambitions in Kenya 6
2.3 Barriers to Equitable Healthcare in Kenya 7
2.3.1 Self-Medication a Consequence of Inaccessible Healthcare 7
2.4 ICT Initiatives around Equitable Access to Healthcare 8
2.4.1 Telemedicine for Healthcare Accessibility 8
2.4.2 EHealth initiatives in Kenya 9
2.4.3 Effectiveness of Telemedicine, eHealth, and mHealth in meeting healthcare needs 10
2.5 AI Self Diagnostic Digital Health Technology, a component of mHealth 11
2.5.1 Cases of Adoption of Health Chatbots/Conversational Agents in Real-world Setting 12
2.5.2 Prerequisites for the Implementation of AI Self-triage systems in Kenya 13
2.6 Critical Analysis 14
CHAPTER THREE
RESEARCH METHODOLOGY
3.1 Introduction 16
3.2 Research Philosophy 16
3.3 Research Design 16
3.4 Research Area 17
3.5 Target Population 17
3.6 Sampling and Sampling Procedures 18
3.7 Data sources and collection 18
3.7.1 Interviews 19
3.7.2 Validity and reliability of the instruments 19
3.8 Data Analysis 19
3.9 Ethical Considerations 21
CHAPTER FOUR
DATA ANALYSIS AND INTERPRETATION
4.1 Introduction 23
4.2 Sampling 23
4.3 Data Collection Procedure 24
4.4 Data Collection Analysis and Memoing 25
4.5 Initial Coding 25
4.6 Code Categorization 26
4.7 Memo writing 27
4.7.1 Access to Formal Healthcare 27
4.7.2 Self-Medication 33
4.7.3 Hospital Visit 39
4.7.4 AI Self-Medication Tools 41
4.7.5 Requirements for AI Self-Medication 50
CHAPTER FIVE
DISCUSSION
5.1 Theme 1: Access to Formal healthcare 57
5.2 Theme 2: Hospital Visit 58
5.3 Theme 3: Self Medication 58
5.4 Theme 4: AI Self Medication tools 61
5.5 Theme 5: Requirements for AI Self-Medication tools 63
5.6 Interrelationship between categories 65
CHAPTER SIX
SUMMARY, CONCLUSIONS AND RECOMMENDATIONS
6.1 Introduction 67
6.2 Summary findings 67
6.3 Conclusion 70
6.4 The implication of the study 71
6.4.1 Theoretical contribution 71
6.4.2 Contribution to knowledge 72
6.4.3 Methodological contribution 72
6.4.4 Contribution to Policy 72
6.4.5 Contribution to practice 73
6.5 Recommendations 73
6.6 Limitation of the study 73
6.7 Suggestions for further study 74
REFERENCES 75
APPENDICIES 81
Interview Guide 81
General Public 81
Medics 82
Informed Consent 83
Sample of Interview Scripts 86
extract from nvivo 98
LIST OF TABLES
Table 1: Respondents’ Database 23
Table 2 : Existing hindrances to access to healthcare 28
Table 3: Existing solutions around access to healthcare 30
Table 4: Factors that led to self-medication 34
Table 5: Existing Self Medication Options 37
Table 6: Assessment of eHealth at Health Facility 40
Table 7: Existing e-Diagnosis Tools 42
Table 8: Assessment of e-Diagnosis correctness 44
Table 9: Assessment of e-Diagnosis Effectivenss 45
Table 10: Recommendation on Adoption and Use of AI-based Self-triage 47
LIST OF FIGURES
Figure 1 Hindrances to quality healthcare 29
Figure 2 Healthcare options 33
Figure 3 Relationship between hindrances to healthcare and healthcare options 33
Figure 4 Enabling factors for self-medication 36
Figure 5 Self-medication options 39
Figure 6 Relationship between self-medication enabling factors and self-medication options 39
Figure 7 Perception on use of AI self-triage 50
Figure 8 Requirements for implementation of AI self-triage systems in Kenya 56
Figure 9: Enabling factors for Self-medication Source: (Yeika et al., 2021) 60
Figure 10 Framework for Implementation of AI self-triage system in the Kenyan Health sector 65 Figure 14 Project map extract from NVIVO 98
CHAPTER ONE
INTRODUCTION
1.1 Introduction/Background
Globally, access to healthcare services among the citizenry has been regarded as the “foremost human right.” The universally adopted resolution A/RES/70/1 of the 2014 UN General Assembly established the 2030 Agenda for Sustainable Development that envisions a world where universal and equitable access to healthcare is assured for all persons of all ages (Assembly, 2015). Consequently, governments globally acknowledged the need for provisioning healthcare services to reduce the global burden of morbidity and mortality. However, Peter O. Otieno et al. (2020, p. 2) state that access to primary healthcare services among the citizenry in Kenya’s low and middle- income levels has been challenging. As per the last census, the annual growth rate of Kenya’s population was set to 2.26%, corresponding to an increase in demands for healthcare services. Annually, the average rate of urbanization is 27.51% meaning more persons are migrating to towns, thus resulting in overpopulated cities and cities. Therefore, there is a need to formulate healthcare policies that are alive to population growth and demographics to improve healthcare accessibility. (Konstantina, Stella, Kleoparta, Aggeliki, & Alexandros, 2020, p. 39) states that governing bodies’ healthcare policies have not enhanced healthcare accessibility or promoted curative, preventive, and rehabilitative services. It is paramount that governments strengthen healthcare systems by improving the healthcare workforce, enhancing public health outcomes, and pushing for equitable access to quality healthcare among the people. However, the existing challenges of the Kenyan healthcare systems are systemic and crosscutting. Since devolution, the healthcare industry in Kenya has seen a reduction in funds, thus affecting resourcing of existing healthcare facilities and the construction new facilities to meet healthcare demands (Masaba, Moturi, Taiswa, & Mmusi-Phetoe, 2020). Underfunding has also contributed to the scarcity of medical supplies forcing patients to procure the medicine from private facilities.
Furthermore, due to poor remuneration, healthcare workers have continually been on a go-slow or strikes, leading to the closure of healthcare facilities. The WHO has stipulated that globally, the recommended ratio between skilled health workers and the population is 1.74 per 1000 population. This is equivalent to 17.4 per 10,000 in Kenya (Organization, 2019). Muthuri, Senkubuge, and Hongoro (2021, p. 2) highlight that the decline in the health worker-to-population ratio in Kenya has been attributed to neglect of healthcare system development, brain drain, deplorable working conditions, inadequate human resource practices, low and delayed salary, and recurrent industrial strikes. As a result, well-equipped and established public healthcare facilities in both urban and rural areas in Kenya lack qualified healthcare staff, forcing patients to wait in long queues or seek alternatives. Qualified medical personnel have opted for private practice or joined well-established private hospitals with dignified pay and a better working environment. The availability of medical supplies among all healthcare facilities in Kenya is critical in optimizing healthcare accessibility by presenting an opportunity for the citizenry to access healthcare when needed. As stated in (Kiplagat & Musyoka, 2021, p. 3), statistically, counties have invested considerably in availing critical equipment and essential medicine in healthcare facilities. However, some crucial medical supplies essential to primary healthcare are still inadequate. The existing gaps in funding coupled with supply chain challenges at KESMA have contributed to existing gaps in quality care. Generally, existing challenges in the healthcare system in Kenya have primarily contributed to the inaccessibility of healthcare services by making healthcare an expensive and scarce commodity (Peter O. Otieno et al., 2020, p. 9).
1.2 Statement of the Problem
Universal healthcare in Kenya is the core of the government’s effort to strengthen health systems, improving accessibility and availability of healthcare services to all regardless of their financial status. The government, through UHC, aims to ensure the healthcare system is well funded, facilities are established, fully equipped, medical personnel is well remunerated, and primary healthcare is accessible to all people without discrimination. Otieno, Kiroro, Runyenje, and Kamau (2021, p. 1) highlight a huge unmet and unexpressed demand for healthcare among the citizenry, which stems from healthcare gaps related to accessibility and availability and acceptability. Currently, existing gaps have contributed to delays in seeking appropriate care, enhanced severity of illnesses, and reduced prognosis. Additionally, the cost of healthcare services has ballooned over time, thus being a financial burden to all citizens’ predominantly low-income households. Existing ICT-related efforts in enhancing access to quality primary healthcare have proved not practical and sustainable in meeting the demand for healthcare. The ever-increasing primary healthcare deficit has resulted in mushrooming of unorthodox means of acquiring healthcare services, for instance, self-medication and the existence of un-approved medical facilities and practitioners.
Public and private healthcare providers have incorporated eHealth and telemedicine to streamline healthcare processes, reduce cost, enhance quality, and reach out to persons with unmet healthcare needs (Christie, 2020). Telemedicine has gained traction among providers who render specialized healthcare services, meeting patients with chronic diseases. Existing Telemedicine initiatives have been through online video conferencing and call centers. Both eHealth and telemedicine have not been effective in meeting the considerable healthcare deficit among the population due to underlying systemic challenges within the healthcare sector (Mbugua, 2016).
Densely populated countries such as the US, UK, and China have streamlined their healthcare systems, national telemedicine infrastructure, and self-triage options to enhance access to healthcare among the citizenry. Self-triage or symptoms checkers were embraced to reduce over- reliance on medical professionals, reduce medical visits, and reduce the cost of healthcare (Montenegro, da Costa, & da Rosa Righi, 2019). Unlike telemedicine, the adoption of AI-led conversational agents triage systems has proved cost-effective as they leverage mobile applications accessible to all citizens.
Regarding Kenya, the realization of AI-led self-triage systems is a possibility. According to Mureithi and Nyaguthii (2021, p. 5), digital penetration and literacy in Kenya have been on the rise, meaning the public has access to digital technical technology and better understands their operation. Telemedicine initiatives have proved to be costly, as it requires the deployment of ICT infrastructure within the country. Leveraging digital penetration and literacy among citizens in delivering AI-Led conversational self-triage systems will enhance access to primary healthcare. Accessible to all citizenry on the web or smartphone, the AI-self-diagnostic conversation agents will provide medical consultation to the citizenry and make referrals.
1.3 Purpose of the Study
This study explored implementing and using a mHealth AI-based self-triage conversational agent to provide cost-effective primary healthcare in Kenya. The study emphasized implementation, adoption, and acceptability issues within the Kenyan population.
1.4 Objectives of the Study
The study’s broad objective was to explore the implementation and use of AI self-triage digital mHealth solutions for primary healthcare diagnosis to increase access and reduce the existing healthcare deficit. Specific objectives include:
• To evaluate the effectiveness of mHealth and eHealth in enhancing access to primary healthcare among the Kenyan population.
• To evaluate requirements that necessitate implementing and adopting the AI self-triage system for primary healthcare.
• To assess the public’s readiness and perception regarding the utilization of AI self-triage systems for primary healthcare.
• To establish a model for implementing and rolling out a mHealth based AI self-triage system for primary healthcare.
1.5 Limitation of the Study
This study encountered the following challenges:
• There was little understanding among the Kenyan medical professionals and the public knowledge around AI self-triage conversational agents.
• There was limited study around AI self-triage conversational agents in an African context.
1.6 Research Questions
In regards to the objective of the study and the problem statement, the research sought to answer the following questions:
1. How does the Kenyan population presently use eHealth and telemedicine applications to access primary healthcare?
2. What is required for seamless implementation and use of “an automated, online” self-triage system in the Kenya Health sector?
3. How ready is the Kenyan health system for the implementation and use of AI Self-triage systems to deliver primary healthcare?
1.7 Justification
This study aimed to explore aspects implementation of AI health Chabot in the Kenyan healthcare system to enhance equitable access to quality and primary healthcare services. Access to direct healthcare services among the citizenry has been highly challenging due to healthcare infrastructure gaps, inadequate resourcing of healthcare facilities, insufficient staffing, and poor quality of healthcare services. Population growth has led to increased demand for healthcare services, but existing challenges have resulted in a considerable healthcare deficit.
In the meeting, the healthcare gaps and ensuring healthcare services are accessible to all; the government commenced implementing a universal healthcare coverage program. According to Okech and Lelegwe (2016, p. 219), UHC in Kenya is founded on enhancing financial protection among the citizenry, increasing the quality of health services, enhancing coverage in the informal sector, and enhancing access among the underserved population. Consequently, the government has implemented eHealth initiatives supporting UHC, i.e., mobile access to NHIF, comprehensive national implementation of ERM systems, and adoption of telemedicine in remote delivery of specialized healthcare services. There is a massive opportunity to upscale existing eHealth initiatives to enhance equitable access to healthcare services by leveraging digital penetration. Therefore, the study explores the position of mHealth in supporting UHC and further assesses the place of AI health Chabot’s in providing personalized healthcare services, specifically self- diagnosis of common diseases. Tremendous opportunities lie within the utilization of AI, specifically conversational agents, in replacing doctor-patient interactions, cutting down hospital visits, pushing for correct self-medication practices, and cutting costs for healthcare.
This study’s audience included stakeholders & policy makers in the ministry of health within the Kenyan government wither national or county. The Kenyan government encourages investors and development partners to establish ICT-related health initiatives to enhance healthcare access in Kenya through its eHealth policy. Therefore, insights acquired from this study shall provide foundation knowledge on the prospects around implementation and use of eHealth initiatives. This study highlighted the prevalence of self-medication and assessed how AI self-triage systems could minimize associated risks.
Login To Comment