Included articles varied widely in study design, making generalizable results difficult to isolate, and insurers may have instituted more than one cost-sharing mechanism simultaneously. Overall, for every 10% increase in cost sharing (via copayments or co-insurance) there was an associated 2%-6% decrease in prescription drug spending. Increasing consumer costs for medications clearly decreases usage.
Some studies demonstrated that the decrease in medication utilization was more pronounced for “nonessential” medications over “essential” medications. However, in specific chronic illnesses this is clearly associated with greater usage of inpatient and emergency medical services.
Cost sharing was also more likely to have adverse health consequences in vulnerable populations, particularly the elderly and poor. One in four Medicaid patients couldn’t fill at least one prescription in the past year, as opposed to one in 10 privately insured patients who couldn’t purchase one or more medications.
Further impact on healthcare consumption and outcomes may be masked because it is difficult to determine individual disease severity, and the effect on the more severely ill would be expected to be greater. These authors attempted to sort out a complex interaction between cost, consumption, and health, and they found important trends.
The goal of cost sharing is to align consumption more clearly with appropriate and economic products, thereby using cost sharing as a public health tool. The consequence of creating the incentives for ill patients to forego necessary treatments is a counterbalancing concern that is supported in some, but not all, of the literature.
Bottom line: Cost sharing for prescription medications decreases medication spending and utilization but disproportionately affects the disadvantaged and increases consumption of more costly healthcare services in patients with some chronic illnesses.
Citation: Goldman DP, Joyce GF, Zheng Y. Prescription drug cost sharing: associations with medication and medical utilization and spending and health. JAMA. 2007;298(1):61-69.
Background: Early administration of antibiotics in community-acquired pneumonia (CAP) improves patient outcomes. The Infectious Disease Society of America instituted guidelines that recommend initiation of antibiotics to all patients with suspected CAP within four hours of triage, and some payors are using this as a quality measure affecting reimbursement. However, this incentive may cause premature diagnosis of CAP and overuse of antibiotics.
Study design: Retrospective chart review
Setting: A large, high-volume teaching hospital with more than 500 beds and more than 112,000 annual emergency department (ED) visits
Synopsis: Charts of all patients with an admitting diagnosis of CAP were reviewed over two six-month periods. The initial review was prior to initiation of a four-hour antibiotics rule; the second was after a financial incentive to initiate antibiotics within four hours of triage was initiated.
After initiation of the four-hour rule, of the patients with an admitting diagnosis of CAP, significantly more patients received antibiotics within four hours of triage (66% versus 54%). However, the number of patients with abnormal chest X-ray findings associated with the diagnosis of CAP decreased from 28.5% to 20.6%, and the proportion of patients with a discharge diagnosis of CAP decreased from 75.9% to 58.9%.
The authors also used two diagnostic paradigms to make an independent diagnosis of CAP based on chart data. With the less rigorous independent analysis 44.7% of patients actually had CAP prior to the four-hour rule, and this fell to 36% after the four-hour rule. Using a more rigorous definition, only 32.7% of patients actually had CAP prior to initiation of the four-hour rule, and this fell to 27%.
There was no difference in length of stay or ICU transfers between the two analysis periods. The authors concluded that a four-hour rule increases premature diagnosis of CAP, presumably because providers felt compelled to initiate antibiotics before they had complete clinical data.