STUDY DESIGN: Interrupted time series (ITS).
METHODS: We used the monthly aggregated data of the number of PEG procedures in older adults with dementia (both broad and narrow definitions), between 2012 and 2018, from the claims data in Fukuoka Prefecture, Japan. A single ITS design was used to estimate changes in the outcome following each intervention (i.e., first, second, and third interventions performed in 2014, 2015, and 2016, respectively). A controlled ITS design was applied to estimate the effects after the sequence of interventions (pre-intervention: 2012-2014; post-intervention: 2016-2018). The control group comprised patients with malignant head and neck tumors who underwent PEG procedures outside the scope of this policy restriction.
RESULTS: The number of PEG procedures decreased significantly only in the month wherein the third intervention was introduced (broad definition: IRR = 0.11, CI = 0.03-0.49; narrow definition: IRR = 0.15, CI = 0.03-0.75). No significant difference was observed between the treatment and control groups during the post-intervention phase.
CONCLUSIONS: The impact of fee-revision policy for PEG on the decrease in PEG procedures among older adults with dementia is remarkably minimal. It is difficult to reduce unnecessary PEG procedures by relying on this financial incentive alone. Policy decision-makers should consider methods to prevent inappropriate use of artificial nutrition for older adults at their end-of-life stage by reforming the health delivery system.