Since the seminal Rand study several decades ago it has become apparent that cost-sharing for patients affects utilization. Whether the utilization foregone or added is necessary or unnecessary is less clear. A new piece of research published at the National Bureau of Economic Research uses a natural variation in cost-sharing in Japan to further ascertain the impacts and what kind of care seems to be most affected. (NBER Paper) One interesting thing about the paper is that it uses services for children aged 7 to 14, which seems to me could introduce complexities, since decisions about seeking health care for children are typically made by adults, and a level of emotionality may be present beyond that that adults experience when making decisions regarding their own care. So you might expect less sensitivity to cost-sharing. During the time period used in the study, local governments in Japan made over 5000 changes in the amounts of cost sharing and who was eligible for them. The researchers attempted to tie these changes to utilization variations. In addition to studying absolute changes in utilization, the researchers sought to identify whether changes were for good-value or low-value services. Low-value care included more after-hours visits and inappropriate use of antibiotics, beneficial care was that that would lower avoidable hospital admissions, result in lower mortality or otherwise improve health outcomes.
When the cost of an outpatient visit to the consumer was reduced from 30% cost-sharing to zero, utilization went up significantly. The likelihood of an outpatient visit in any month increased by 6 to 8 percentage points, or over 20% on an absolute basis. Monthly outpatient spending increased by over 20%. The greatest sensitivity to price appeared to be around the zero point; that is eliminating any cost-sharing had the strongest effect on more utilization and adding any copayment amount when their previously was none had a strong tendency to reduce utilization. Along the range of cost-sharing, the effects were roughly the same whether the changes were to add cost-sharing or reduce it. There was some evidence of increased outpatient utilization for ambulatory care sensitive conditions when cost-sharing was reduced, which was deemed to be higher value care, but that increased utilization for these conditions was not tied to lower hospitalization for them. And there was no other evidence of an increase in receipt of beneficial care. There was also no evidence of cost offsets in other care categories from more spending on outpatient visits. There was more low-value care receipt when cost-sharing was lowered, including more after-hour visits and more inappropriate use of antibiotics. The study suggests that even a small amount of cost-sharing can be useful in avoiding unnecessary care and spending.