Now here is a weird little study. We would like people to avoid so-called low-value care, but get plenty of that high-value care, especially preventive care. People are working on various interventions to lead to this outcome. One such intervention is reported in Health Affairs and had a somewhat strange outcome. (HA Article) The test bed for the study was a benefit design for a state’s employees that reduced cost-sharing for receipt of high-value preventive care, which included cancer screenings and chronic disease medications. The researchers looked at overall use of both what they defined as high-value care and low-value care in the year before introduction of the new benefit design and in the two years after introduction. A comparison group was created of employees in other states who were not covered by a similar benefit design. Specific services analyzed as high-value were colon cancer screening, cervical cancer screening, mammography and lipid screenings. Low value services were derived from the Choosing Wisely list and included vitamin D screening, PSA testing (how that is low-value when new research is indicating that men are now dying more from prostate cancer when they don’t get screened is beyond my comprehension), cervical cancer screening in young women and cardiac screening in low-risk patients, as well as several imaging services.
For some reason, in the baseline year, about one-third of the intervention population received a “low-value” service, compared to only a fourth of the control group. In that baseline year, 60% of the intervention group and 55% of the control one received a high-value service. After the intervention, the study group got more high-value care, but also more low-value care. Relative to the control group, intervention employees increased high-value service use by 11% in the first year and 5% in the second year. But low-value use also rose relative to the control group, by 85 in the first year and 3% in the second year. There was an overall increase in office visits, which appeared to be correlated with much of the rise in both high-value and low-value services. Hey, you go see the doctor, of course they are going to order more follow-up services. From my perspective this study is telling us two, potentially reconcilable things. One is that doctors may not yet have good education on what services are really needed by a patient and when it is appropriate to order certain services. That could be fixed by better real-time information system guidance. The second is that medicine is still significant art and medical judgment and doctors may know better than academic researchers what services a patient could most benefit from.