Transforming the Kenya Health Information System (KHIS) to an Early Warning and Real-Time Electronic Disease Notification System: Optimization for Epidemiology, Disease Surveillance and Response in Kenya.

Transforming the Kenya Health Information System (KHIS) to an Early Warning and Real-Time Electronic Disease Notification System: Optimization for Epidemiology, Disease Surveillance and Response in Kenya.
To develop a concept on transforming the Kenya Health Management Information System (KHIS) to an electronic disease early warning and real-time notification system through optimization of disease surveillance indicators: automatic and real-time notifications integrated into an informative standardised tool. The concepts will enhance and develop part of the list of diseases mandatorily reported within the stipulated period in disease occurrence depending on the case definition of the diseases, virulence and the degree of spread of new emerging diseases that have not been defined e.g. a new infectious disease.

1. Transforming the Kenya Health Information System (KHIS) to an Early
Warning and Real-Time Electronic Disease Notification System:
Optimization for Epidemiology, Disease Surveillance and Response in
Kenya.

Kenya Health Information System (KHIS) website(DHIS2, 2021)
Written by:
Stephen Olubulyera
Senior Public Health Specialist|Epidemiologist|Health Systems Strengthening
Specialist|Researcher|IT Enthusiast
Ministry of Health
Turkana County

2. Transforming the Kenya Health Information System (KHIS) to an Early
Warning and Real-Time Electronic Disease Notification System:
Optimization for Epidemiology, Disease Surveillance and Response in
Kenya.
Stephen Olubulyera

Objective
To develop a concept on transforming the Kenya Health Management Information System
(KHIS) to an electronic disease early warning and real-time notification system through
optimization of disease surveillance indicators: automatic and real-time notifications
integrated into an informative standardised tool. The concepts will enhance and develop
part of the list of diseases mandatorily reported within the stipulated period in disease
occurrence depending on the case definition of the diseases, virulence and the degree of
spread of new emerging diseases that have not been defined e.g. a new infectious disease.

Introduction
There have always been contrary opinions in the preliminary integration of Information,
Communication, and Technologies (ICT) in healthcare systems. Despite having the ability to
transform the quality and execution of healthcare services, the implementation of Health
Information Management Systems (HMIS) in healthcare organisations has experienced few
challenges that frustrated the acceptance and utilization of the service. The feasibility of
health management information system in effecting healthcare support nevertheless has
promoted the introduction and implementation of technological solutions in
healthcare (Dehnavieh et al., 2019).
Kenya is among seventy-three low and middle-income countries across the world using the
District Health Information System Software (DHIS2) in the management of Healthcare
information. The health system software is serving approximately 2.4 billion people (DHIS2,
2021). In the year 2010, Kenya opted to adopt the use and implementation of the District
Health Information System Software 2 in the healthcare of its country. The free, open-source
and web-based software application was effected in all health facilities across the forty-
seven counties in Kenya (Gathua, 2016). According to the report conducted on the impact of
DHIS2 in the healthcare system in Kenya: the software’s technical capabilities enhanced
prompt data analysis, availed report and timely presentable feedback. The analysis of the
potential benefits of DHIS2 on the healthcare system in most of the countries utilizing it
found out that flexibility of the software made the access to health data sufficient and robust
in terms of indicators across all areas of healthcare. Healthcare majorly depends on DHIS2

3. on proper management and storage as well as analysis and data visualization in form of
presentation (Dehnavieh et al., 2019).
The author of the article established that despite DHIS2 having technical capabilities, system
software robustness and tremendous strengths in the meta-analysis of healthcare data, there
are glaring operational challenges in Epidemiology, Disease Surveillance and Response in
regards to early warning system and real-time notification protocols and features. DHIS2
lacks the electronic disease early warning and real-time notification capabilities. One of the
mandates of Epidemiologist or Disease Surveillance and Response Coordinators working in
the Ministry of Health in Kenya is to scrutinize epidemiological data entered on DHIS
software by various health practitioners to ascertain and analyze anomalies or trends of
suspected or probable infectious diseases, which one-way or the other triggers further
actions. The ability of the DHIS2 to provide a real-time notification in form of email or text
message when and if thresholds or indicators on the system software surpass the set
thresholds will be fundamentally crucial in Epidemiology, Disease Surveillance and
Response. This technical feature will assist in the identification of the potential disease,
tracking and development of geospatial models of the spread of the probable infectious or
notifiable disease; this will be an electronic disease early warning.
An Electronic Early Warning System “facilitates collection of essential, minimal data on
prioritized epidemic-prone or selected diseases with significant public health consequences,
where rapid analysis of trends for outbreak or event detection for prompt response and
intervention is instrumental for mitigating the potential high morbidity and mortality that
may be attributed to the even”. Integrating real-time notification dictates that a built-in alert
capability feature will be assigned to the selected notifiable diseases and events coded for
monitoring which when reported or surpass the set-threshold will trigger an instant secure
and electronic alert in form of email or mobile text message to relevant parties for action or
response (Ahmed et al., 2019).

Methodology
Since optimization of the Kenya Health Information System (KHIS) has never been
conducted to incorporate early warning system, this section will highlight ways and
mechanisms of integration on the health system while providing various strategies of
upscaling and improving the health information system. Development and integration of
electronic reporting tools used in different settings and situation are vital for the success of
the transformation. The electronic reports will not only adopt the context of the manual
reports but also integrate a real-time data entry capture, resonate feedback and statistical
analysis of the health data including thorough data authentication. Monthly disbursement of
the data in form of presentation to the relevant authorities through email and alerts or
alarms: summarized into short texts on mobile phones that are easily consumed and
comprehended. Integration of the real-time notification features on the thirty-six notifiable

4. epidemiological diseases indicators on the DHIS2 system as soon as data entry is done on the
system will be fundamental in the surveillance of disease in the country.
Given that DHIS2 is an open-source software; programmers, information technologists,
epidemiologist and health specialist will work together to reprogram, revamp and revitalize
the health information system; concentrating on the crucial indicators, specific areas and
reporting tools including incorporating additional abilities to the health information system
to implement a statistical analysis by use of artificial intelligence capabilities.
The Kenya Health Information System (KHIS) will undergo reprogramming and
customization to accommodate mobile applications running on Android, IOS, and Windows
operating systems. This will promote accessibility and feasibility on its workability for quick
system interrogation after an alarm or warning on disease or event is triggered. Its
customization will also ease up data entry on an instance or notifiable diseases by field
epidemiologist. Having the ability to access the health information system via the mobile
application will grant geolocation (latitude and longitudes) capture capabilities: the ability
to capture coordinates especially of emerging diseases or rarely unique events and linking it
to the patterns of their occurrence across the country.
Integration of Artificial Intelligence (AI) analysis will be fundamental: linking of emerging or
reemerging diseases to geographical location, analysis of patterns or trends of a cluster of
signs and symptoms to connect to probable disease affecting the geographical area. The AI
will also calculate the rate of spread of disease: by analysis of the number of cases reported
per minute and geologically link to specific areas reporting diseases of similar symptoms
while factoring in the distance, behaviour of host and environment.
There is a need for the incorporation of a section to capture qualitative data on the Kenya
Health Information System (KHIS). This section will fundamentally capture views and
suggestions of users of the system specifically on the respective data reporting tools and the
importance of the collection of the data in regards to utilization, making decision and action.
Irrelevant or outdated data sets collecting information that is neither consumed nor
beneficial during action or decision-making will be reanalysed and deactivated.

Discussions
Advanced informatics in Public Health Surveillance could resonate early detection of
diseases. This has been as a result of proper designing, incorporation and integration of well-
defined reporting electronic tools on the Health Information Systems tasked with analysis of
crucial indicators at different sectors of health thus ensured timely capture, analysis and
real-time technological feedback of epidemiological information (Ahmed et al., 2019).
Implementation of early warning system in health information system will improve on
timely reporting and detection of infectious diseases. Designing, incorporating and
integration of the early warning system on the Health Information System (HIS) promotes

5. the use of an already available resource that will neither overburden the system nor the
health professionals tasked with data entry (Bieh et al., 2020).
Optimization of Kenya Health Information System through the integration of Artificial
Intelligence feature with the robust capability of deeper statistical data interrogation, trend
and pattern analysis including incorporating the location feature to link disease spread,
suitability of environment and behaviour of the host will improve the functionality of
Epidemiology, Disease Surveillance and Response.
Integration and use of mobile applications (android and IOS) and other mobile features on
Kenya Health Information Systems will initiate real-time reporting of syndromic and event-
based surveillance data thus providing an early warning opportunity that will promote
prompt risk analysis, early response and development of mitigation, control and prevention
measures. The use of mobile applications for data entry and interrogation will be convenient
to the field epidemiologist that will access and input data on the health information system
on the go. The aim of implementing the use of mobile applications to access, data entry and
interrogation on the Kenya Health Information Systems is for convenience and an
opportunity to capture more information in terms of photos, video and geographical
locations (geocodes). This will be fundamental in analyzing the pattern of disease,
environmental factors influencing transmission and behaviour, traits or characteristics of
the host promoting the rapid spread of the disease, especially of infectious diseases. By
capturing and storing photos, videos and geocodes on the health information system:
provides an opportunity for further interrogation of the collated data through linkage to
attached pictorial evidence including coordinates to link the disease cases or events to
geographical areas depicting the possibility and rate of transmission by gauging on the real-
time reports vis a vis the similarity of symptoms.
Despite having the ability to timely and promptly alert the relevant health authorities on the
real-time reporting of the thirty-six notifiable diseases, the Kenya Health Information System
will have an automatic generation, analysis and submission of indicator monthly summary
data feature. This will be initiated on the 16th of every month just a few hours after the
deadline of monthly data entry, which is the 15th of every month. The summary will be
submitted via email for monthly consumption or action: this may include a summary of
multiple dashboards, analysis of the occurrence of various diseases and any abnormal spike
of diseases as well as the summary of the thirty-six notifiable disease that was reported in
the consecutive weeks of that month.
Integration of the Artificial Intelligence (AI) feature on the Health Information System will
promote machine learning, which will upscale specificity and sensitivity in terms of
enhancing accuracy and linkage to disease or events. In addition, AI will be fundamental in

6. collating, analysis and timely submission of complete, accurate and relevant monthly-
analysed summary presentations of various indicators via email. Moreover, the Early
Warning System and the capability of real-time technological alert of any notifiable disease
reported across any health facility on Kenya Health Information System will be efficient and
effective because of the implementation of Artificial Intelligence. The ability of AI system to
accommodate and integrate the Health Information System with Health Facility Management
Systems (HFMS) such as AfyaEHMS will be beneficial in facilitating smooth transition of data
sets computations and submission of data from Health Facility Management Systems (HFMS)
to Kenya Health Information System (KHIS).
Incorporation of a system application feature for capturing qualitative data will be useful
and informative: qualitative data provided on the system will be beneficial in improving the
implementation of mitigation measures and assist in upscaling the system itself. The system
will have the ability to capture and analyse qualitative data scribed on the reporting tools:
using qualitative analytical tools or software such as Nvivo and AI.

Limitations
The process and transformation of the Kenya Health Information Systems to an early
warning and real-time electronic disease notification systems has limitations:
The capacity of User and Machine Learning for AI: For Artificial Intelligence to be
effective there has to be a machine-learning period, which will depend on the skilful use of
the Health Information System by the Health professionals.
District Health Information System: DHIS is an open-source application, therefore there
are limitations when it comes to systems upgrades and updates. For customization and
integration of new features and change of some of the protocols on the systems may
malfunction depending on the degree of manipulation of the system; reprogramming might
be costly.
System Maintenance and Upgrades: During pandemics in low-and middle-income
countries, the underperformance of surveillance systems is experienced and overburden,
especially in areas where regular surveillance of health indicators are disrupted because of
other competing phenomena that need immediate attention or response.

Conclusions
For the establishment of an effective and efficient early warning system, the Kenya Health
Information System has to undergo an upgrade to render the system robust in terms of
analytical features, evolve, and customize the data capture, integrate Artificial Intelligence
on the HIS platform, threshold alarms or alerts on notifiable diseases according to case
definitions, automatic analysis and timely submission of summarized monthly data.

7. References
Ahmed, K., Bukhari, M. A. S., Altaf, M. D., Lugala, P. C., Popal, G. R., Abouzeid, A., & Lamunu, M.
(2019). Development and Implementation of Electronic Disease Early Warning
Systems for Optimal Disease Surveillance and Response during Humanitarian Crisis
and Ebola Outbreak in Yemen, Somalia, Liberia and Pakistan. Online J Public Health
Inform, 11(2), e11. doi:10.5210/ojphi.v11i2.10157
Bieh, K. L., Khan, A., Yezli, S., El-Ganainy, A., Asiri, S., Alotaibi, B., . . . Jokhdar, H. (2020).
Implementing the Health Early Warning System based on syndromic and event-based
surveillance at the 2019 Hajj. East Mediterr Health J, 26(12), 1570-1575.
doi:10.26719/emhj.20.129
Dehnavieh, R., Haghdoost, A., Khosravi, A., Hoseinabadi, F., Rahimi, H., Poursheikhali, A., . . .
Aghamohamadi, S. (2019). The District Health Information System (DHIS2): A
literature review and meta-synthesis of its strengths and operational challenges
based on the experiences of 11 countries. Health Inf Manag, 48(2), 62-75.
doi:10.1177/1833358318777713
DHIS2. (2021). The District Health Information System Software 2. Retrieved from
dhis.org/about
Gathua, P. W. (2016). Assessment of Data Use of the District Health Information
System(DHIS2): A Case Study of Nairobi County. (Masters of Art in Monitoring and
Evaluation), University of Nairobi, Nairobi.