Pulmonary disease (including primarily asthma, Chronic Obstructive Pulmonary Disease (COPD), pneumonia, lung cancer, and tuberculosis) is an increasingly large portion of the global health burden. Developing countries with large low-income populations are disproportionately affected, due to increased risk factors (e.g. biomass cooking stoves) as well as poor access to health care and lack of affordable screening tools for early detection. Chronic Obstructive Pulmonary Disease (COPD) alone is currently the third leading cause of death in the world and second leading cause of death in India after ischemic heart disease. In the younger population, pneumonia is a particular concern, being the leading cause of death for children under 5 years of age. Tuberculosis (TB) has also reached alarming proportions in India (24% of all cases worldwide). Despite this great prevalence of pulmonary disease in India, access to modern diagnostics instruments is not possible; furthermore, approximately 60% of general practice (GP) clinic doctors in India are primarily trained in Ayurvedic medicine with little or no training for diagnosing respiratory disease. As a result, many of the patients with lung disease are underdiagnosed or misdiagnosed (often confused with cardiovascular disease). As a result, there is a great need to provide health workers in India with simple tools that can be used to diagnose or screen for respiratory disease in the primary care setting. Addressing this need, our team has been developing a mobile diagnostic platform consisting of a digital stethoscope, peak flow meter, and mobile phone that can be used to screen for symptoms of lung disease and provide a guide for diagnosis. The present study extends this work and has the following aims: (1) To validate and test a low-cost mobile diagnostic platform for the purpose of identifying symptoms of lung disease and providing diagnostic guidance. (2) To assess the acceptance and usability of the mobile diagnostic platform by the local general practitioner (GP) doctors in rural India. We propose to deploy and test a low-cost mobile diagnostic platform, making use of machine learning algorithms that will detect specific symptoms of lung disease and help guide diagnosis. In the first year of the project, we shall create and train the mobile software algorithm using data collected in the field from patients (N=250) that have been previously diagnosed with specific lung diseases. Then year 2, we shall evaluate and test our mobile platform with Indian patients (N=250) recruited from four GP clinic sites in the Pune, India region. The automated mobile phone diagnosis result shall then be compared with the diagnosis from trained pulmonologists, using a traditional stethoscope as well as standard lung function testing instruments. A preliminary diagnosis and qualitative feedback shall also be collected from non-trained GP doctors using the mobile tools in order to ascertain usability and diagnostic value.
Pulmonary disease is a very large public health concern in India, as well as in most developing countries. Relatively few general practitioner (GP) doctors are trained to properly diagnose pulmonary disease, and affordable tools for diagnostic support simply do not exist. This proposal seeks to validate and test the use of low-cost mobile phone-based diagnostic tools for pulmonary disease.
