This is a pre-post intervention study involving pediatric patients presenting to the
study hospitals in seek of medical care for an acute illness. The study involves three
phases: (I) Baseline Period, (II) Interphase Period, (III) Intervention Period.
Phase I (3-6 months) will be a prospective observational cohort at Mbagathi County
Hospital in Nairobi, Kenya, and Jinja Regional Referral Hospital in Jinja, Uganda. During
this period, there will be no changes to healthcare delivery procedures in the study
hospitals. Triage will continue to be performed by hospital staff using Emergency Triage
and Treatment (ETAT) guidelines, the system that is currently in effect at the study
hospitals. Data collection will be undertaken in the triage waiting area. While
participants are waiting for their turn to be seen by the hospital triage nurses, our
trained study nurses will collect data on a pre-selected list of predictor variables.
These data will be used to develop a clinical prediction model based on the need for
hospital admission.
Control Site (Phase I, 12 months): Kiambu County Referral Hospital in Nairobi, Kenya will
serve as the control site and no intervention will be implemented. At this site, Phase I
will commence for a period of 12 months. There will be no Phase II or Phase III.
Phase II (1-3 months) will involve technology development, usability testing, and
training.
Phase IIa: Technology Development. A risk prediction model will be derived using the data
collected in Phase I and implemented in a Digital Triaging Platform, along with a
digitized version of the ETAT+ guidelines. The Digital Triaging Platform, including vital
sign measurement devices (PhoneOx and RRate and the mobile application and clinical
dashboard has already been developed and evaluated. Once the digital triage tool has been
developed, it will be evaluated in potential users using simulated patient scenarios and
a 'Think Aloud' method.
Phase IIb: Usability Testing and Training. The digital triage tool will be evaluated for
ease of interface navigation, functionality and basic workflow. A sample of 15 health
workers in the study hospitals to represent the primary user groups will be selected for
participation in the 60-minute-long usability testing initiatives. Participants will be
recruited through word of mouth as there is a very small cadre of potential participants.
The objective of the training is to (1) ensure healthcare workers understand how to
correctly collect and interpret patient information, and (2) to obtain feedback on the
digitization of the tool. Training will use a framework that meets key international
norms for testing digital tools, including, the think-aloud method and a questionnaire.
Each training session will be conducted by a moderator and observer. During the
evaluation, the observer will be seated next to the participant and will record user
interaction with each interface, comments, errors, and duration of each task.
Participants will be given 3-5 patient scenarios which will list hypothetical information
to be entered into the app. This information will be designed to represent routine data
collected during triage examination at the study hospitals. The moderator will provide
the fictional charts to participants and instruct them throughout the tasks. During the
simulated patient scenarios, participants will be asked to think aloud, in order to
assess their thought process as they used the app. Participants will be specifically
instructed to comment on the layout of the app screen, the dialogue on each interface,
the order of tasks, and any additional observations or opinions. After learning the
basics of the digital platform, the participants will be read the think aloud
instructions and asked to perform the list of tasks and answer questions. The observer
will complete a checklist to ensure that all tasks were completed, questions will be
asked to evaluate task comprehension, and notes will be taken about whether help was
needed in completing each task. At the end of the training session, participants will
complete a triage tool training questionnaire to provide an understanding of the
practical benefits and drawbacks of incorporating the digital triage tool into a clinical
context. The questionnaire will utilize open ended questions and comment responses. from
this evaluation. Responses from the survey will be anonymous. The data generated from the
training phase is fictitious and will not be linked to any individual subject.
Transcriptions and Think Aloud observations will be analyzed using the Framework Method
to assess attitudes of health workers. Responses will be transcribed and coded for the
identification, examination and interpretation of emerging themes and patterns. Results
from the analysis, feedback from the questionnaires, and comments on the observer
checklists will be used to generate a report with suggested improvements to be shared
with the quality improvement implementation team prior to Phase III.
Phase III (3-6 months) will be an interventional period involving routine use of the
digital triage tool by the hospital triage nurses at Mbagathi County Hospital in Nairobi,
Kenya, and Jinja Regional Referral Hospital in Jinja, Uganda. The digital triage tool
will not replace triage policies already in place at the study hospitals, but rather it
will supplement and strengthen existing triage systems. As done in Phase I, study nurses
will collect data on the pre-selected list of predictor variables in the triage waiting
area while participants are waiting to be seen by the hospital triage nurses (who will be
using the digital triage tool). Continued collection of predictor variables will allow
comparison of participant characteristics in the pre-intervention cohort and the
post-intervention cohort. (Funder: Wellcome Trust Innovator Award # 215695/Z/19/Z)