Patients' views on pay for performance in France: a qualitative study in primary care

Lessons learned from a pay-for-performance scheme for appropriate prescribing using electronic health records from general practices in the Netherlands

INTRODUCTION

A nationwide pay-for-performance (P4P) scheme was introduced in the Netherlands between 2018 and 2023 to incentivize appropriate prescribing in general practice. Appropriate prescribing was operationalised as adherence to prescription formularies and measured based on electronic health records (EHR) data. We evaluated this P4P scheme from a learning health systems perspective.

METHODS

We conducted semi-structured interviews with 15 participants representing stakeholders of the scheme: general practitioners (GPs), health insurers, pharmacists, EHR suppliers and formulary committees. We used a thematic approach for data analysis.

RESULTS

Using EHR data showed several benefits, but lack of uniformity of EHR systems hindered consistent measurements. Specific indicators were favoured over general indicators as they allow GPs to have more control over their performance. Most participants emphasized the need for GPs to jointly reflect on their performance. Communication to GPs appeared to be challenging. Partly because of these challenges, impact of the scheme on prescribing behaviour was perceived as limited. However, several unexpected positive effects of the scheme were mentioned, such as better EHR recording habits.

CONCLUSIONS

This study identified benefits and challenges useful for future P4P schemes in promoting appropriate care with EHR data. Enhancing uniformity in EHR systems is crucial for more consistent quality measurements. Future P4P schemes should focus on high-quality feedback, peer-to-peer learning and establish a single point of communication for healthcare providers.

Risk selection reduces efficiency of value-based healthcare

Value-based healthcare aims to improve efficiency and value for patients. Value-based payment models are a form of provider reimbursement to achieve this. Studies on these models have found positive results, but may be biased by unintended consequences, such as risk selection. Risk selection is a multi-dimensional phenomenon that occurs at the patient, hospital, and system level, and is a source of inefficiency and inequality in healthcare. Risk selection may occur because of selection bias in the outcomes that are evaluated and rewarded, or due to the selection of lower cost patients. Risk selection may also stem from professional reputation. The motivation to engage in risk selection may also arise from differences in the meaning of value. To mitigate these unintended consequences, several strategies can be adopted. These include making value-based payment models attractive, but not mandatory, as well as incentivising transparent reporting of best practices, using adequate risk adjustment, expanding performance metrics, and including patient-reported experience measures. Other mitigation strategies could include adopting a mixture of performance measures, using mixed methods of paying physicians, and implementing monitoring and evaluation mechanisms. However, such approaches are not flawless, and the problem may never be fully solved. This perspective serves as a warning for the constant presence of risk selection, as well as informing policy makers, politicians, and organisations implementing VBP models on ways to minimise the possibility of risk selection.

The effect of 'paying for performance' on the management of type 2 diabetes mellitus: a cross-sectional observational study

Background

The ‘cycle of care’ (COC) pay for performance (PFP) programme, introduced in 2015, has resourced Irish GPs to provide structured care to PCRS eligible patients with type 2 diabetes mellitus (T2DM).

Aim

To investigate the effect of COC on management processes.

Design &setting

Cross-sectional observational study undertaken with two points of comparison (2014 and 2017) in participating practices (Republic of Ireland general practices), with comparator data from the United Kingdom National Diabetes Audit (UKNDA) 2015–2016.

Method

Invitations to participate were sent to practices using a discussion forum for Health One clinical software. Participating practices provided data on the processes of care in the management of patients with T2DM. Data on PCRS eligible patients was extracted from the electronic medical record system of participating practices using secure customised software. Descriptive analysis, using IBM SPSS Statistics for Windows (version 25), was performed.

Results

Of 250 practices invited, 41 practices participated (16.4%), yielding data from 3146 patients. There were substantial improvements in the rates of recording of glycosylated haemoglobin ([HbA1c] 53.1%–98.3%), total cholesterol ([TC] 59.2%–98.8%), urinary albumin:creatinine ratio ([ACR] 9.9%–42.3%), blood pressure ([BP] 61.4%–98.2%), and body-mass index ([BMI] 39.8%–97.4%) from 2014 to 2017. For the first time, rates of retinopathy screening (76.3%), foot review (64.9%), and influenza immunisation (69.9%) were recorded. Comparison of 2017 data with UKNDA 2015–2016 was broadly similar.

Conclusion

The COC demonstrated much improved rates of recording of clinical and biochemical parameters, and improved achievement of targets in TC and BP, but not HbA1c. Results demonstrate substantial improvements in the processes and quality of care in the management of patients with T2DM.