CCG initiates a scheme to reduce variation in care and prevent complications for diabetic patients, achieving improvements in uptake of NICE recommended care processes

Outline:

Barking and Dagenham (B&D) is an area of marked deprivation and an escalating ethnic population. The local health economy is challenged by high care costs for diabetes, and there had historically been significant inter-practice variation in the quality of care.

In 2016, B&D CCG initiated a Local Improvement Scheme (LIS) to improve diabetes diagnosis and outcomes and reduce variation. An external consultancy was employed to extract and evaluate practice-level data, providing lead clinicians and administrators with training calls and a website regularly updated to show achievements.

At the outset, a data-gathering exercise was undertaken to highlight inter-practice variations and achievement was compared with national best practice. Improvement was driven by clinician engagement, robust discussions at network level and peer support.

Financial incentives were offered to practices to meet LIS targets. Results after 12 months showed significant improvements in uptake of NICE recommended care processes (from 28% to 67%).

More patients were offered structured education and there was an increase in the number of patients diagnosed with pre-diabetes (from 0.62% to 4.7%). Diagnosis of diabetes also improved substantially and the estimated number of undiagnosed diabetes cases fell from 1,642 in 2012-13 to 624 in 2017-18 (more than 60% reduction).

Challenges:

Socio-economic deprivation and ethnicity have long been associated with poor health outcomes. However, emerging evidence has indicated there is no statistically significant correlation between diabetes outcomes and locality deprivation. Variability in care may be reduced by better organisation of services.

This premise formed the foundation of our improvement scheme. Diabetes care processes are the cornerstone of good diabetes care and help prevent complications, thereby improving patient experience and outcomes.High uptake of care processes requires practices to maintain proactive patient recall which in turn relies on good organisation of services, patient engagement, clinical leadership and teamwork.

The National Diabetes Audit showed uptake of care processes in B&D was well below national average (28.6% against 49.6%) and the rate of complications was high. The LIS challenged practices to achieve better levels of care by improving uptake of care processes and achieve better control of blood pressure, cholesterol and diabetes levels.

A health economics model was used to estimate the impact of improving diabetes care on the health economy. Business intelligence verified that the scope of activity existed and a benefits stream was developed for five years reflecting the profile expected from improved diabetic care over more than a one-year period.

Outcomes:

All 37 member practices participated (12,500+ diabetes patients). Results after 12 months showed significant improvements. Patients receiving eight NICE diabetes care processes increased (from 24% to 60%). The highest increase was in urine albumin:creatinine ratio testing (from 36% to 70%).

More patients were offered structured education (40.7% to 63.1%), mean HbA1c reduced (56 to 53 mmol/mol), and patients achieving HbA1c ≤58 mmol/mol increased from 45.7% to 53.6% (P = 0.000052). More than half of the practices gained 3% improvement in the number of patients achieving a BP <5 mmol/mol; thirty gained a 3% improvement in patients achieving an HbA1c ≤58 mmol/mol.

The percentage of patients achieving an 11 mmol/mol (1%) improvement in HbA1c in 12 months rose to 12.0% (from 6.9% the previous year). Across the CCG, patients diagnosed with pre-diabetes increased (0.62% to 4.7%). With improved screening, the estimated number of undiagnosed diabetes cases fell from 1,642 in 2012-13 to 624 in 2017-18 (60% reduction). Improvements were seen across most practices, with inter-practice variability reduced.

The challenge was to keep practices engaged and drive the stragglers to reduce variability. Regular data feedback, networking sessions, and sharing good practice helped overcome this.

Spread:

The success of the initiative was down to coordinated efforts by the CCG and GP practices. Learning from the LIS has been shared with primary care colleagues and GP practices in Havering and Redbridge where similar work is under way to improve diabetes outcomes.

It has also been shared with colleagues nationwide, via an article published in the British Journal of General Practice. The project included practice networking and sharing good practice. Its success has boosted the confidence and morale of local clinicians and has led to improved patient care and outcomes. Furthermore, the project proves that better organisation of care, clinical leadership and meaningful use of clinical data can overcome the challenges posed by socioeconomic deprivation and achieve high quality care.

This founding principle can be used to improve the care and outcomes of most long term conditions. The scheme has now been extended to cover atrial fibrillation and plans are underway to widen it further, to cover all the long term conditions across the three neighbouring CCGs (Barking and Dagenham + Havering and Redbridge). During the scheme, GP feedback and quality improvement initiatives led to further systems innovation to track patients and their eight care processes completion.

Value:

The project used a proxy model to estimate the benefits to the health economy. This was commissioned by Healthy London Partnerships (HLP) for the purpose of assisting London CCGs to form robust business cases for change programmes aligned to Right Care analysis. This method is clinically evidenced at calculating an estimate of health economy benefits rather than relying on SUS data which could be inaccurate, incomplete or over-coded by providers (please see Appendix 1 for details of the model developed by Optimity Consulting for HLP).

This model asserts: “For every 100 diabetic patients who have their care improved there is a five year saving of £68,600”. The key data to demonstrate impact of our scheme is: Source: National Diabetes Audit (Type 2 data) Register Patients with completed eight care processes Start point/baseline: March 2016 10,287 2,920 28.4% Scheme end point: March 2018 12,610 8,470 67.2%

The increase in patients completed is therefore 5,550 which, if sustained, accrues value to the health economy of £3,807,300 over five years. This was achieved with an investment of £1,200,00 – approximately £5 per registered patient and £100,000 for technological innovation and clinical facilitation.

Involvement:

Better patient care and outcomes were at the heart of this project, with local practices’ teams it’s key stakeholders. All 37 practices were involved from the outset. For individual practices, data sharing with a third party consultancy and the CCG was a novel idea and was met with some apprehension. Practices were wary of data being used for performance management.

This was addressed with a considered and empathic approach, supported by CCG practice improvement liaison, by the third party consultancy and by the CCG’s clinical directors. Once practices were signed up to the scheme, they were regularly contacted and areas requiring improvement were highlighted. Network data was shared in project meetings and this led to a sense of comradery and healthy competition, with each practice keen to do better than its neighbour.

Within the practices, clinical and administrative leads for diabetes were identified. These were central to the scheme’s success. The administrative lead focused on call-recall of the patients and helped address the gaps in care processes, while the clinical lead championed patient-centred care.

Other stakeholders, including the local community provider, the acute trust and the neighbouring CCGs were also kept abreast of the scheme and its potential impact.