Expert calls for improved access to healthcare database for big data research
While big data have been used to promote better healthcare services in Hong Kong, access to local healthcare database has to be improved to fully unleash the power of big data research.
“Big data are often ‘found’ data, meaning that they are readily available for analysis. Examples include data from patients’ electronic health records, nursing home assessment data, and administrative data from non-governmental organizations,” said Professor Terry Lum of the Department of Social Work and Social Administration, The University of Hong Kong. “Big data research is not new to aging research and has been used to shape policies and interventions to promote healthy aging.”
“For example, big data research may help evaluate and improve clinical care,” he suggested.
Lum and his team conducted an observational study to investigate whether presence of depressive symptoms and antidepressant use predict subsequent decline in physical functioning and number of hospitalization in nursing home residents. They analyzed data from 1,076 residents of six government-subvented long-term care facilities in Hong Kong. [J Am Med Dir Assoc 2015;16:1048-1054]
Results demonstrated correlations between presence of depressive symptoms and faster deterioration of physical functioning as well as increase in the number of hospitalizations. The use of antidepressants was found to alleviate decline in activities of daily living in individuals with depressive symptoms.
“We found that the presence of depressive symptoms is associated with more utilization of healthcare services, and the use of antidepressants may play a significant role in altering the trajectory,” reported Lum. “In other words, depression is worrisome but treatable. However, the condition is often left untreated.”
“Big data research is also useful in studying patient’s behaviour,” he noted.
In 2016, Lum and his team examined the neuropsychological and clinical profile of 1,005 cognitively- impaired help-seekers in a local early-detection community dementia programme. Results indicated that a longer waiting time from symptom recognition to early assessment was associated with worse cognitive function upon presentation. [J Am Geriatr Soc 2016;64:584-589]
Previously, Lum also participated in several big data researches in the US, which had provided insights on US’s healthcare policy and system. [Gerontologist 2013;53:334-344; J Am Geriatr Soc 2014;62:71-78; Health Serv Res 2014;49:778-797]
“These studies would not be made possible without the provision of data from the US federal and state governments,” he pointed out.
“In contrast, conducting big data research in Hong Kong is faced with difficulties,” Lum continued. “It is extremely difficult to gain access to some of the local database, such as medical reimbursement data. If these databases can be utilized by big data researchers, it may be possible to innovate the healthcare system in Hong Kong in response to the rapidly aging society.”