OBJECTIVES : To evaluate the prevalence and spending of low-value healthcare services under a National Health Insurance program.

METHODS : Data used in the study was obtained from the National Health Insurance claims data in 2009-2013. Use of seven services that population with minimal clinical benefit were measured for 406,162 beneficiaries. Selected services included PSA testing for men aged over 75 years old, bone mineral density testing at frequent intervals, preoperative chest radiography, preoperative echocardiography, preoperative pulmonary function test, preoperative stress tests, and screening for carotid artery disease in asymptomatic adults.

Individual-level characteristics (age, gender, comorbidity score), hospital-level characteristics (ownership, accreditation level, annual ambulatory visit, total LOS), area-level characteristics (ratio of the specialist to primary care physicians, ratio of low-income, percentage of >12 years schooling and burden of disease) were also collected.

We assessed the proportion of beneficiaries receiving services, the number of services and total spending devoted with broad and narrow definitions. We also estimated the variation and association between the use of different services within hospitals, adjusting for hospital and area-level characteristics.

RESULTS : The annual prevalence of selected services ranged from 0.01% (preoperative pulmonary function test) to 36.63% (screening for carotid artery disease in asymptomatic adults). In the seven services, screening for carotid artery disease had the highest proportion (34.30%-36.63%) during 2009-2013, followed by preoperative chest radiography (13.74%-14.20%). Total spending of these seven services increased from US$136,260 to US$179,654. PSA testing for men aged over 75 years old contributed the most spending of low-value healthcare service (42.50%), and the following was screening for carotid artery disease in asymptomatic adults (41.84%).

CONCLUSIONS : Services detected by the limited number of measures of low-value healthcare contributed modest proportions of overall spending but the proportions of beneficiaries and total spending sustained increasing. It may reflect the overuse of low-value healthcare services more broadly.