D6.1 – Report including description and analysis of a theoretical model on incentives for GPs.pdf
The way family doctors are paid is rarely visible to patients, yet it quietly shapes the care they receive. A doctor paid a flat sum for each person on her list faces very different pressures from one paid for every treatment she performs, or one rewarded for keeping patients well. For people living with a chronic condition, these differences matter a great deal, because good chronic care depends on time and attention that no payer can easily observe or verify. Two decisions are especially sensitive to how money flows: how much preventive effort the doctor invests today, and whether the patient is sent to hospital tomorrow. This report builds a theoretical framework to understand both.
The report develops a two period model of a primary care physician caring for a chronically ill patient. In the first period the doctor exerts preventive effort that raises the probability the patient remains in a mild state of illness; in the second period she decides whether to treat the patient herself or refer her to hospital. Three payment components are compared, a capitation fee, a treatment fee and a pay for performance bonus, together with the case in which hospital capacity is unavailable, as during a pandemic. Doctors are assumed to care about their own income but also, to varying degrees, about patient benefit, and the analysis is extended to risk aversion and probability weighting within a rank dependent utility framework, using the Prelec weighting function. The main results are that pure capitation leads to all patients being referred to hospital, which is wasteful for the mildly ill, and to preventive effort below the efficient level even for highly altruistic doctors; that a treatment fee set between the costs of treating mild and severe cases restores appropriate referrals; and that performance pay can raise effort without distorting referral choices, provided the payer knows how altruistic the doctor is. Once probability weighting is introduced, however, responses become less predictable: effort can jump abruptly when the bonus crosses a critical value, and some effort levels cannot be induced at all. The propositions are then translated into an experimental design with five treatments and a set of testable hypotheses, which will be examined in the laboratory experiment of D6.4, while the model also provides the conceptual basis for the analysis of existing European payment schemes in D6.2 and D6.3.
What the deliverable contains:
- A two period theoretical model of preventive effort and hospital referral decisions in chronic care, with altruistic physicians
- A comparison of capitation, treatment fees and pay for performance, including the case of constrained hospital capacity
- An extension to risk aversion and probability weighting in a rank dependent utility framework, showing how behavioural traits alter responses to incentives
- Six propositions and five lemmas identifying which combinations of payments achieve efficient referrals and efficient effort
- An experimental design with five treatments and a set of qualitative and quantitative hypotheses for the laboratory experiment in D6.4
Download: D6.1 – Report including description and analysis of a theoretical model on incentives for GPs.pdf