October 17, 2017
Although the prevalence of tuberculosis (TB) has more than halved in China since 1990, early diagnosis remains a challenge. A new study led by a researcher from the UNC Gillings School of Global Public Health provides, for the first time, data on the quality of care found at three different levels of the Chinese rural health system.
Sean Y. Sylvia, PhD, assistant professor of health policy and management, is first author of an article titled, “Tuberculosis Detection and the Challenges of Integrated Care in Rural China: A Cross-sectional Standardized Patient Study,” published online Oct. 17 in the journal PLOS Medicine. His co-authors include researchers from four universities in China as well as Stanford University, McGill University and The World Bank.
“In China, TB remains a major public health threat with just under one million new cases each year,” Sylvia said. “Further progress against the disease requires improved detection, particularly in rural areas where prevalence rates are high. Before this study, we had no information on the ability of primary care providers in rural areas to diagnose and properly manage presumptive cases of TB, whether in village clinics, township health centers or county hospitals.”
Information on how the quality of care changes when moving up the tiers of China’s rural health system is critical to an ongoing debate around health reforms, Sylvia explained. The reforms, in part, seek to relieve stress at upper tiers of the health system by encouraging patients to seek care at lower tiers first.
The study collected these data and combined them with a new methodology to determine how patients would be treated if, for instance, they could visit county hospitals only with referrals from lower tiers. To conduct the research, investigators sent unannounced standardized patients (people hired from the local community and trained to present a consistent case of presumptive TB to health-care providers) to more than 270 clinics at all levels of the health system.
Two main findings suggest that improving the quality of primary care available in rural areas is critical in China.
First, 90 percent of county hospital visits were managed correctly, compared to only 38 percent of visits in township health centers and 28 percent of visits in village clinics. Incorrect antibiotic use remained high at all levels of care, ranging from 24 percent in county hospitals to 65 percent in township health centers.
Second, the researchers found that these deficits did not reflect a lack of provider knowledge or facility equipment. Instead, they discovered a significant `know-do’ gap, in which the clinical practice of providers fell below their level of knowledge.
In light of these findings, Sylvia and colleagues concluded that policies encouraging patients to go first to village clinics before seeking care at higher levels would increase these deficits, unless there is significant improvement in quality at the lower tiers of the system.
“The assumption behind `gate-keeping’ systems and similar policies is that if providers do not know how to manage a case, they will refer it,” said Sylvia. “This assumes, however, that providers will correctly recognize which cases need to be referred. This assumption did not hold true in our study – only 28 percent of village doctors and 18 percent of township providers correctly referred the case onward.”
This study is part of a larger initiative in China and globally that aims to measure the quality of primary care and evaluate strategies to improve it. Sylvia currently is working with colleagues in China using methods similar to this study’s but targeting other diseases; he also is helping to develop other, complementary approaches to measuring quality and is leading a randomized trial testing approaches to improve the quality of primary care provided by clinicians in rural China.
An article from The Economist, titled, “In poor countries it is easier than ever to see a medic… But it is still hard to find one who will make you better,” discussed other recent work by Sylvia and his co-authors using standardized patients.