When financial incentives were withheld from 35 outpatient facilities owned and operated by Kaiser Permanente, physicians were less likely to screen for four quality indicators. In particular, the screening rate for cervical cancer dropped even lower than before the incentives were instituted, according to a study in the British Medical Journal.

The indicators were screenings for diabetes retinopathy and for cervical cancer; glycemic control in diabetes; and control of hypertension. Initially, incentives to screen for diabetic retinopathy and cervical cancer were removed. Payments were then reinstated for five years. During that period, screening for diabetic retinopathy rose from 84.9 percent to 88.1 percent. After those five years, the incentives were removed again, this time for four years, and resulted in a rate of 80.5 percent — lower than before the incentives were reintroduced.

Helen Lester, a professor of primary care at the University of Manchester in the United Kingdom and lead author, surmises that the physicians might have been focused on other clinical indicators that did have incentives attached to them.

She says that there was “a year-on-year fall-off in performance of about 3 percent in screening for diabetic retinopathy and a 1.6 percent fall-off for cervical cancer screening. Although it doesn’t seem like too much of a drop-off, keep in mind that these are real people who were not having cervical screenings or not having eye examinations if they had diabetes.”

She points out that the Kaiser incentives were given to the organization, not to the individual physicians, so the “incentives did not financially benefit the physician directly.” She says the findings show the need to continue to monitor performance, especially after incentives are removed.

Performance levels fluctuate as payment is instituted

Source: Lester H, Selby J, Fireman B, Campbell S, et al. “The impact of removing financial incentives from clinical quality indicators: longitudinal analysis of four Kaiser Permanent indicators.” BMJ; 2010;340:c1898