Design and evaluation of a mobile application to assist the self-monitoring of the chronic kidney disease in developing countries

Abstract

BackgroundThe chronic kidney disease (CKD) is a worldwide critical problem, especially in developing countries. CKD patients usually begin their treatment in advanced stages, which requires dialysis and kidney transplantation, and consequently, affects mortality rates. This issue is faced by a mobile health (mHealth) application (app) that aims to assist the early diagnosis and self-monitoring of the disease progression.MethodsA user-centered design (UCD) approach involving health professionals (nurse and nephrologists) and target users guided the development process of the app between 2012 and 2016. In-depth interviews and prototyping were conducted along with healthcare professionals throughout the requirements elicitation process. Elicited requirements were translated into a native mHealth app targeting the Android platform. Afterward, the Cohen’s Kappa coefficient statistics was applied to evaluate the agreement between the app and three nephrologists who analyzed test results collected from 60 medical records. Finally, eight users tested the app and were interviewed about usability and user perceptions.ResultsA mHealth app was designed to assist the CKD early diagnosis and self-monitoring considering quality attributes such as safety, effectiveness, and usability. A global Kappa value of 0.7119 showed a substantial degree of agreement between the app and three nephrologists. Results of face-to-face interviews with target users indicated a good user satisfaction. However, the task of CKD self-monitoring proved difficult because most of the users did not fully understand the meaning of specific biomarkers (e.g., creatinine).ConclusionThe UCD approach provided mechanisms to develop the app based on the real needs of users. Even with no perfect Kappa degree of agreement, results are satisfactory because it aims to refer patients to nephrologists in early stages, where they may confirm the CKD diagnosis.

Computer-Aided Diagnosis of Chronic Kidney Disease in Developing Countries: A Comparative Analysis of Machine Learning Techniques

“…The DM (1), GFR (2), and albuminuria (3) attributes had the same prediction power for the experienced nephrologist and the J48 decision tree, considering a scale ranging from 1 (highest priority) to 8 (lowest priority); however, the nephrologist prioritized and used the creatinine (4), urea (5), gender (6), hypertension (7), and age (8) attributes. In Figure 4a, even if the decision tree is a simple yes/no evaluation that takes only a few steps to reach 95.00% accuracy (or global kappa of 0.9221), a simpler evaluation approach based on the KDIGO guideline without considering machine learning and data maintained over years may not reach the same performance, compared to the opinion of the most-experienced nephrologist [18]. The decision-making of the nephrologist considers age, gender, reference values of biomarkers, and risk factors, i.e., it is not restricted to the risk classification of the KDIGO guideline.…”

“…The Renal Disease Assistant App 3 allows users to record the results of laboratory tests and visualize their kidney function. Finally, the MultCare Android app provides features to help users conduct self-monitoring of their CKD risk in developing countries, based on medical guidelines [18].…”

“…An experiment focusing on CKD diagnoses in developing countries was conducted to improve the analysis and assist in addressing the research gap identified by the systematic review. Based on a previous study [18], the experiment considered a dataset containing subjects diagnosed or not diagnosed with CKD, collected from the University Hospital Prof. Alberto Antunes located at the Federal University of Alagoas (UFAL), Brazil. The learning techniques comprised a subset of the techniques identified when answering RQ1.1, i.e., RF, NB, J48 decision tree, KNN, SVM, and MLP neural network, were used to classify the CKD risk.…”

CKDNET, a quality improvement project for prevention and reduction of chronic kidney disease in the Northeast Thailand

“…YouTube, Facebook, website, digital kiosk and mobile applications were also approached as modern methods and found resourceful for awareness and self-monitoring. We feel strongly motivated from the finding that mobile application assisted self-checking of CKD has satisfactory outcome in developing countries [ 24 ]. Our salt reduction campaign was effective to reduce the salt content in 8.7% of food menu prepared in the university canteens.…”

Creating a Theoretically Grounded, Gamified Health App: Lessons From Developing the Cigbreak Smoking Cessation Mobile Phone Game

“…Several nongamified health apps use an iterative user-centered design process [9,13,79-136]. Most apps are designed by an expert multidisciplinary team.…”