Abstract:
Open-EMR is an electronic free data storage that is used by many healthcare centers and clinics
that includes Rwandan healthcare centers. Data recorded in these software platforms form various
formats of patient data, diagnosis, financial transactions, and many other forms of data. The
availability of these data has great and useful benefits to deal with the managerial, financial, and
planning departments' issues to develop the clinic /healthcare centers if the proper analysis were
performed.
The research “Need Assessment for Patient Data Analysis Tools in the KIBAGABAGA
hospital Healthcare system” used 2015-2022 malaria data cases from Open-EMR of the
KIBAGABAGA hospital, and it performed analysis of these data. Malaria data cases forms time
series data and the methods applied in the modeling were ARIMA and SARIMA statistical
machine learning techniques.
For the 7 years of malaria-data cases; 2016-2017 had a high variation of positive malaria cases and
it gets reduced in 2018-2019. The rate of malaria cases at KIBAGABAGA Hospital is decreasing
and becoming more stable in the near 2 coming years. This shows that the danger of this disease
is not so much in the future but the hospital can maintain its strategies of the prevention of the
malaria disease
The results of analysis for both methods indicate that the SARIMA model performed well in
comparison with the ARIMA model. By considering the same observations in the training phase,
the root mean square errors (RMSE) in SARIMA went to 34.07 while ARIMA got 80.03. The
same behaviour to indicate SARIMA as the best model was reproduced in the testing phase which
is the prediction for the 2 coming years.
