Eric C. Wong
Advanced analytics leads to lifesaving results
Palo Alto Medical Foundation zeros in on disease with SAS® Visual Data Discovery, leading to better patient care
With its diverse Asian population, the San Francisco Bay Area provides medical researchers a unique setting in which to study disease trends among an ethnic group for which few studies exist. Using SAS Visual Data Discovery, the research institute of a local health care organization analyzes medical data from the local Asian-American population – often with lifesaving results.
The Palo Alto Medical Foundation Research Institute's Department of Health Policy Research conducts studies in the areas of cardiovascular disease and epidemiology. Focused on 660,000 active patients throughout the San Francisco Bay Area, the team's research is aimed at disease trends among the Asian population in particular.
One thing I like about SAS is that it's integrated. It performs data management and analytics well, which is helpful because it streamlines the process and reduces errors. I also like the visualizations, especially the statistical graphics.
Eric C. Wong
"There haven't been many studies on the Asian-American population," says Eric Wong, Senior Statistician in the Department of Health Policy Research. "And those that have been done did not look at specific Asian subgroups. We have a unique opportunity in the San Francisco Bay Area to study the six largest subgroups – comprising Asian-Indian, Chinese, Filipino, Japanese, Korean and Vietnamese – in one geographic setting and in one health care organization."
Not-for-profit leader in cardiac care
Part of the Sutter Health family of not-for-profit hospitals and physician organizations, it serves more than 100 communities in Northern California. The organization is also a regional leader in cardiac care, and a pioneer in advanced patient safety technology and the use of electronic health records.
Under the leadership of researcher Latha Palaniappan, Wong's team conducts a variety of studies, such as research into the trends and risk factors for diabetes and whether they vary among racial/ethnic subgroups. Findings inform internal policies that lead to improvements in patient care, Wong explains.
"If we find that one population is at a higher risk than other racial/ethnic groups, we can develop culturally relevant content on risk factors, suggestions on nutrition, or recommend lifestyle changes," Wong says. "We also disseminate our findings back to clinical staff so they can implement changes and policies that result in higher-quality patient care. We share everything we learn with the medical community at large, and we make it a priority to publish our findings."
Wong points to a number of interesting studies that clearly illustrate how disease trends are different among Asian ethnic subgroups. One found that Asian-Americans have a greater prevalence of metabolic syndrome – a cluster of conditions that increases the risk of heart disease, stroke and diabetes – even though they present at lower levels of traditional risk factors.
In 2004, Wong says, the World Health Organization published a research article suggesting that Asian populations should be examined at body mass index (BMI) cut points lower than traditional values. So Palaniappan, Wong and their collaborators studied the association of BMI to the risk of metabolic syndrome in Asian-Americans, compared to non-Hispanic whites. For every Asian subgroup, the researchers learned that the risk of metabolic syndrome was higher than non-Hispanic whites across all BMI values.
"We also looked at the burden of cardiovascular disease in Asian-Americans, specifically focused on stroke and coronary heart disease," he explains. "What we found was a similar discovery. Compared to non-Hispanic whites, we found that stroke rates were elevated in some subgroups, though not statistically significant. We also found that coronary heart disease rates were higher in Asian-Indian and Filipino men, compared to non-Hispanic whites, which was statistically significant."
In yet another interesting study, Wong describes how his team analyzed the effectiveness of shared medical visits versus one-to-one patient and provider appointments.
"One of our physicians was conducting shared appointments as a way of promoting weight loss in a clinical setting," Wong says. "In a shared medical appointment, one physician meets with a group of patients for a longer period of time, for example, 6 to 12 patients for 90 minutes. The approach borrows from forums that involve a group of individuals facing similar circumstances, where there's group camaraderie – they learn together and share information. The question we had was whether the approach affected health outcomes. Did it really promote greater weight loss compared to individual medical appointments? What we found was that it did. On average, patients in the group setting lost two pounds during the observation period, compared to one-to-one patients who actually gained weight on average. And while a decrease of a pound or two might not be that dramatic, if we can at least use shared medical appointments to maintain weight, then that is a victory as well."
According to Wong, analytics from SAS also support proactive medical intervention practices and policies throughout the health network.
"When there's an observation of a trend, we try to figure out the immutable risk factors associated with the probability of having the disease," he explains. "Based on that, we can think about whether there's something we can do about it, like education, screenings or preventative measures. If we identify a specific population at a higher risk, it's natural to design specific intervention or screening methods that target that group, as well as design communication methods to reach the at-risk population."
To perform the myriad studies it conducts each year, the research team derives its de-identified data from a variety of sources, including: electronic administration, registration and billing data; electronic medical records; and Census, state and national health surveys.
"One thing I like about SAS is that it's integrated," Wong says. "It performs data management and analytics well, which is helpful because it streamlines the process and reduces errors. I also like the visualizations, especially the statistical graphics. Being able to visualize the information and present it to decision makers, while looking for and discovering trends, is very important.
"Our patients and health network have benefited from the work we've been doing here, and SAS has been a valuable tool," he concludes. "We are striving to be leaders in multispecialty care and at providing the full spectrum of primary care services. Having a better understanding of the patient population and disease risk factors is integral to becoming a real pioneer and leader in the industry."
Required an integrated data management and analytics solution to consolidate electronic health records to perform disease trend studies among demographic segments.
- Supports studies that inform policies and provide better patient care.
- Enables shared medical visits versus one-to-one appointments.
- Identify specific populations at higher risk of disease.
- Develop education, screenings and preventative measures for high-risk populations.
- Streamline data management and analytical processes, reduced errors in data quality.