Frictions to treatment decisions at the end of life

Paluszynska, Aleksandra Joanna

doi:https://doi.org/10.21985/n2-2zsw-k265

Work

Frictions to treatment decisions at the end of life

Public

Download PDF

Download All Files (.zip)

``Hazy decisions: The effect of dementia on medical decision-making'' \\ I estimate the causal effect of having dementia on the course of treatment for unrelated diseases by leveraging differences in the relative time of onset of dementia and the other condition in a difference-in-differences event-study framework. To demonstrate this approach I look at heart attacks and show that after accounting for individual and calendar-time fixed effects the ``dementia first'' and ``heart attack first'' groups exhibit parallel trends in health care utilization before the heart attack. I find that health care cost is persistently lower following a heart attack for those who had dementia already and that this effect is driven by decreases in frequency of care. Dementia patients who experience a heart attack reduce the number of interactions with the health care system, while those who do not yet have dementia markedly increase their frequency of care in response to the health shock. I establish that this finding holds universally for a variety of other acute and chronic conditions and is not driven by higher mortality among dementia patients. Long-term reductions in medical care received are driven by care over which patients have the most discretion such as diagnostic tests and prescription drugs, which points to reduced follow-up care and treatment adherence as possible mechanisms behind my estimates. \\\\ ``Do procedure codes matter? A case study of Advance Care Planning'' \\ This project aims at understanding the way in which new procedure codes affect provision and cost of medical care. Primarily used for reimbursement purposes, new codes can generate additional revenue for existing services or lead to modifications in their insurance coverage that in turn affect physicians' financial incentives and can lead to changes in both quantity and kind of services provided. I study the case of Advance Care Planning (ACP), a 30 minute counseling service that became reimbursable by Medicare in 2016, to show that the new reimbursement lead to a 16% increase in the provision of all counseling services among physicians who used the new procedure code. Importantly, only 3pp of this increase can be attributed to the provision of ACP itself, and the remaining 13pp reflects an increase in the use of other services, such as optional screening for depression and alcohol abuse. Finally, I show that this increase is driven by recipients of ACP, suggesting that a more generous reimbursement of one service can substantially affect the provision of other services, which should be incorporated into projections of the potential cost of new reimbursement policies.

Creator