Lecture 10 - Paying Providers: Incentives and Quality

Question 1

What are the 3 conceptual models of service delivery?

Answer 1

1. Trust model of public service delivery
2. Mistrust model of public service delivery
3. Voice model of public service delivery
4. Choice model of public service delivery

Question 2

What is the trust model of public service delivery?

Answer 2

Physician is given autonomy to do what is in the best interest of the patient.

Government sets budget, professionals decide how to spend it, then submit bill to third party payer (government, insurance)

Used by Scandanavian countries

Will only work if the government and the provider have the same aim (e.g. patient welfare)

Question 3

What is the voice model of public service delivery?

Answer 3

Giving patients a voice in shapign how healthcare is delivered

Done via face to face talks (patients/providers/consultants), boards, forums, complaints procedures, petitions

Will work if user is well informed, providers have the same aim as users, it is designed to be low cost, and the ability to use voice is spread equally across groups

Question 4

What is the mistrust model of public service delivery?

Answer 4

Creates incentives that motivate providers to do what the government wants. It is based on performance, incentive or a high degree of control

Will only work if the government has the right aims and can pick the right targets

Government needs to understand the providers motivation to design incentive structures. Also need to design it so it doesn’t drive out instrinsic motivation

Government needs to be able to monitor activities accurately

Reporting needs to be manageable

Question 5

What are 2 examples of the mistrust model in practice?

Answer 5

1. Managed care in the US - insurance only contracts with certain providers and they have direct managerial control over the providers
2. Targets in the UK - providers have to report on a lot of metrics, and have lots of targets they need to meet

Question 6

What is the choice model of public service delivery?

Answer 6

Patient can pick their provider

Providers are independent so they can keep any surplus on their budget

Will only work if users are well informed, competition is possible, there are few opportunities for cream-skimming, and the ability to use choice is spread equally

Problems: people may not want choice, they just want good local service. Wealthy/more educated may make better choices, competition may make providers more concerned about their market share instead of patients, elderly or multimorbid patients may not be able to exert choice

Question 7

Why are teaching hosptials in the UK seen as predatory?

Answer 7

Large hospitals can be a threat to the public realm since they win big contracts and can shove smaller hospitals out of the market

Question 8

Which conceptual model of service delivery is P4P grounded in?

Answer 8

Mistrust model of service delivery

Question 9

What is the theoretical basis for P4P? Hint: there are 3 seperate theories

Answer 9

Agency theory: designing incentives will align the interests of the agent and the principal so they act in the best interest of the principal

Microeconomic theory: providers actions are a function of incentive, constraints, opportunities and preferences

Cognitive psychology: providers have intrinsic motivation and extrinsic motivation so they can be driven by external reward

Question 10

According to Bokhour et al 2006, what are the 5 different mechanisms for the distribution of incentives (who payment is made to)?

Answer 10

1. Equal distribution to all providers
2. Dependent on individual provider performance on payer’s quality targets
3. Dependent on individual provider performance on practice-based incentive schemes (pay practice, not individuals)
4. Money retained wholly by organisation
5. Hybrid approach; money to individual physicians and organisation

Question 11

According to Bokhour et al 2006, what is the potential quality impact of paying incenties equally to all providers?

Answer 11

This might encourage collaboration and teamwork among providers as everyone benefits equally.

However, the individual incentive to improve performance might be weaker as everyone gets the same reward regardless of their individual effort.

Question 12

According to Bokhour et al 2006, what is the potential quality impact of paying incentives based on the peformance of individual providers on the payer’s quality targets?

Answer 12

This approach has the strongest potential to directly influence a doctor’s behavior.

If they know their incentive depends on meeting specific quality goals set by the payer (insurance company), they’re more likely to adjust their practice to achieve those goals.

Question 13

According to Bokhour et al 2006, what is the potential quality impact of paying incentives based on the peformance of the entire practice?

Answer 13

It could encourage better collaboration within the practice to achieve the overall goals.

This can still motivate doctors, but the direct financial reward might be less powerful than paying based on individual provider performance.

Question 14

According to Bokhour et al 2006, what is the potential quality impact of paying incentives to the organisation which retains the funds?

Answer 14

This offers the least financial incentive for individual doctors.

However, the organization might use the money to implement system-wide changes that improve quality of care, such as investing in new equipment, infrastructure or training programs.

Question 15

According to Bokhour et al 2006, what is the potential impact of paying incentives via a hybrid approach (to individual physicians and organisation)

Answer 15

This combines the potential benefits of the other approaches.

Doctors might be motivated by individual incentives while the organization can also make system-level improvements to enhance quality of care.

Question 16

What are 3 criteria that performance measures should meet?

Answer 16

1. Measurable
2. Clinically relevant
3. Data is valid/reliable

Question 17

What are some examples of performance measures?

Answer 17

Clinical quality/effectiveness, efficiency, utilisation, access, safety, patient satisfaction

Question 18

What are the 3 boxes in the Donabedian Quality Assurance Model and what is contained within them?

Answer 18

1. Structure: material resources, operational characteristics, organisational characteristics
2. Process: clinical care, policy and procedure, adherence to standards
3. Outcome: health status of patients, clinical measures

Question 19

What are some arguments FOR measuring outcomes of care (against process measures)?

Answer 19

• provides a standardised indicator of success
• determines “what works” in healthcare
• universal
• clinical attention moves towards improving health
• harder to manipulate than process measures
• process measures may become dated
• provides a wholistic view while process measures are less relevant when considered alone
• incentivising processes means you will get a lot more of that process, not nessecarily better outcomes

Question 20

What are some challenges with measuring the outcomes of care?

Answer 20

• confounding variables; patients may receive care from multiple providers, each patient has unique needs, multimorbid conditions complicate
• timing of measurement; e.g. if using hospital mortality instead of 30 day mortality then hospitals may discharge patient if they think they are going to die
• using absolute versus relative measures give different perceptions of situation

Question 21

What happened at Papworth Hospital in the UK?

Answer 21

• Papworth Hospital was notified by the Care Quality Commission that they triggered a mortality outlier alert
• Asserted that they treated the oldest and sickest patients in the country and based on their risk-adjusted survival metrics, they are above the national average
• Argued that Hospital Episode Statistics (HES) data was innacurate due to data entry errors, inadequately trained staff, and incompleteness of data. Argue that HES data is not designed for assessing clinical effectiveness.
• Argued that the analysis did not consider patient risk factors. Similar procedures can lead to very different outcomes based on the patient. Using clinical databases would be more accurate.
• Grouping diagnoses makes mortality comparisons irrelevant since patients are diverse.

Question 22

Are readmissions indicative of poor quality?

Answer 22

Mixed evidence

Also have to consider treating sicker patients

Question 23

What is one way to prevent crowding out instric motivation as per Young and Conrad 2007?

Answer 23

Modest-sized incentives in the early phase of a P4P programme minimises the risk of crowding out and incentive to game

Use a phased approach

Question 24

What are some unintended consequences of P4P?

Answer 24

Distortion of priorities

May neglect non-targeted activities

“Squeezing excellence”: high performing providers might shift towards simply meeting the target, potentially leading to a “good enough” mentality instead of continuous improvement. Could stifle innovation.

Question 25

What is the case for measuring processes instead of outcomes?

Answer 25

• certain aspects of processes (such as wait times) are often valued by patients
• certain process measures are associated with desired health outcomes
• measuring outcomes can be difficult, time consuming and costly
• process measures are almost instantaneous and therefore can be acted upon quickly
• processes are attributable to the provider so more easily interpretable than outcomes which are influenced by the patient and the provider

Question 26

What 2 categories of processes can be used in incentive design?

Answer 26

1. Clinical (e.g. beta blockers prescribed at discharge for acute myocardial infarction)
2. Non-clinical (e.g. wait times)

Question 27

What 3 design features should P4P programs have?

Answer 27

1) incentives are large enough to motivate hospitals to make sizeable investments in improving care
2) focus on a small number of high value measures that will motivate clinicians to change their practice
3) simple design that will enable clinicians and organisations to know how they are doing and how to improve

Question 28

What are value-based payments (VBP)?

Answer 28

Focuses on the overall value of care delivered, considering both quality and cost-effectiveness. It rewards providers for achieving positive patient outcomes while managing costs efficiently.

Question 29

What are the benefits of VBP?

Answer 29

Improved Quality: Incentivizes providers to prioritize patient well-being and positive outcomes.

Cost Savings: Reduces unnecessary spending on ineffective interventions.

Focus on Prevention: Encourages preventative care that can prevent more expensive future treatments.

Question 30

What are the challenges with VBP?

Answer 30

Data Collection and Measurement: Developing reliable metrics to assess quality and value of care can be complex.

Administrative Burden: Implementing and managing VBP programs can add administrative tasks for providers.

Standardization: Ensuring fair comparisons across different healthcare settings can be challenging.

Question 31

According to Figueroa et al 2016, what were the outcomes of the US’s VBP program?

Answer 31

Medicare Hospital Value-Based Purchasing (VBP) Program: This program adjusts payments to hospitals based on the quality of care provided to Medicare beneficiaries. Hospitals are evaluated on various measures, including patient readmission rates, mortality rates, and patient experience.

Targeted conditions: acute myocardial infarction, congestive heart failure, pneumonia

• 3 years after the initiation of their program, there was no impact on patient outcomes, may have even had a detrimental one
• Before the program the mortality rate of targeted conditions were declining at a faster rate that after the incentives kicked in

Question 32

What is the Hospital Readmission Reduction Program in the US?

Answer 32

• Created in 2012. Hospitals are financially penalized if they have higher than expected risk standardized 30 day readmission rates for acute myocardial infarction, heart failure and pneumonia
  o In 2015 expanded to include acute exacerbation of COPD, patients admitted for elective total hip arthroplasty and total knee arthroplasty
• Before 2012, hospitals received payment using the inpatient prospective payment system which was based on DRGs. Did not include post-discharge care or interventions that would potentially reduce the incidence of readmission
• Readmission rates are adjusted for age, sex, and co-existing conditions
• Savings are added to the Medicare Hospital Insurance Trust Fund
  o Protects guaranteed benefits
  o Providing new benefits and services for all Medicare beneficiaries
  o Lowering the cost of Part B premiums
• Assumes that hospital readmissions are associated with unfavourable patient outcomes

Question 33

What did Ross 2017 find about the Hospital Readmission Redution Program in the US?

Answer 33

• Large percentage in the reduction of readmissions was attributable to the way hospitals were describing patients. By describing them as sicker, they can increase their risk adjustment, allowing them to avoid financial penalties
• Some evidence that 30 day mortality increased after implementing HRRP
• Studies found that hospital differences only marginally accounted for patient’s readmission risk
• Certain conditions may warrant exclusion from the program

Question 34

What are some pros and cons of HRRP as per McIlvennan et al 2015?

Answer 34

PROS
* Focus on the care coordination across silos of care, focus on improving the overall patient experience through hospitalization and beyond
* Emphasis on patient outcomes rather than a few care processes
* If properly adjusted for case mix can better reflect the overall performance of a health system
* Improved communication between inpatient and outpatient providers, more seamless transition form hospital to home
* Focus on all-cause readmission rates incentivizes hospitals to focus not only on the primary medical problem, but also a patient’s comorbid, psychological, social, and environmental conditions.
* Since balanced with DRG based reimbursement, there is a balance to limit unnecessarily long hospital stays while discouraging unnecessary readmissions due to premature discharge

CONS
- Potential to disproportionately penalize hospitals serving poorer populations. Initial risk adjustment models did not account for socioeconomic status. Several analyses have shown that caring for patients with a lower socioeconomic status puts a hospital at higher risk of incurring penalties
- Potential to avoid readmissions and increase mortality. Hospitals with higher admission rates will likely also have higher readmission rates which is not fully accounted for in the risk-standardization process
- Hospitals with higher mortality rates will have lower readmissions
- Institutions and providers cite that many readmissions are due to disease progression and patient behaviour. Need to be able to determine whether the readmissions are due to preventable events
- Arbitrary time window. Readmissions after a few days may be more indicative of poor care than 4+ weeks which could be from patient disease or events out of control of the hospital
- Potential to overlook the impact of hospitalization
- Limitations on existing admin data and concerns for coding manipulation
dardized readmission rates between institutions is at face value a strong argument that many readmissions are preventable
* Arbitrary time window
* Readmissions after a few days may be more indicative of poor care than 4+ weeks which could be from patient disease or events out of control of the hospital
o Some have suggested weighting HRRP’s penalties according to timing of readmissions
* Controversial inclusions and exclusions
* Initial algorithm penalized hospitals for any other planned admission including procedures such as implantable cardioverter-defibrillators in heart failure patients even though that represents high quality care. In response the CMS accounted for a wider range of planned readmissions in 2014.
* Potential to overlook the impact of hospitalization
* A focus not only on transitional care but hospitalization itself may help reduce post-discharge syndrome (stress from admission) and it’s potential to increase readmissions
* Limitations on existing admin data and concerns for coding manipulation
* Forced to lump necessary and unnecessary readmissions together and rely on aggregate rates to reflect potentially preventable event

Question 35

What is the Quality Outcomes Framework?

Answer 35

Introduced in 2004

Provides incentives to GPs in the UK for providing quality care.

Original scheme had 76 clinical indicators covering 10 conditions

Doctors could exclude patients if they were not deemed appropriate

Question 36

What was the problem with the Quality Outcomes Framework in the UK?

Answer 36

• Created significant additional income for GPs which ran the risk of crowing out intristic motivation
• GPs had already been doing these things but now they were being paid for it
• very expensive, cost over £1 billion
• no impact on mortality
• larger and more affluent practices received higher payments due to the payment formulas
• diagnostic codes were sometimes incorrectly specified
• lack of regulation of exception reporting
• did not accomodate for multimorbidity
• increased admin load

Question 37

What was the success of the Quality Outcomes Framework in the UK?

Answer 37

Reduced readmissions
Minimised inequalities in care
Some improvements in chronic disease outcomes

Question 38

Explain how P4P was implemented in Rwanda. Why was it a success?

Answer 38

Rwanda: Rwanda has implemented a well-evaluated P4P scheme since 2002.
* Rwanda started experimenting with pay-for-performance schemes in 2002 to address underlying poor performance by public sector providers who were paid on salary and had little motivation to increase outputs.
* Providers who meet the targets and demonstrate high-quality care are eligible for financial bonuses as part of the P4P scheme.
* The scheme includes a set of priority services, such as vaccination, ante-natal care, and deliveries in health centers. Each service is assigned a unit weight, similar to a fee schedule, which determines the payment amount for the provider.
1. P4P increased deliveries in health centres; increased prevention interventions in children such as immunisation. It also led to reduced child mortality and taller children.
* There were also several other reforms occurring in Rwanda at the same time, including the rapid expansion of community-style insurance which fueled demand for health services
* The evaluation of the P4P scheme in Rwanda was conducted by taking into account the rollout of the national program. The first healthcare facilities to implement the scheme were chosen through random assignment, allowing for a comparison between those that implemented the P4P scheme and those that had not yet implemented it.
* The lesson from the Rwanda P4P is not the type of incentives used, but rather the proper approach to evaluation which allows the isolation of the effect of P4P from other reforms.