Guidelines Are Only as Good as Their Implementation: Closing the Diabetes Care Gap with GP Data 

Key Takeaways

• Diabetes care across England’s 6,000 GP practices suffers from significant unwarranted variation that has historically been difficult to measure in real-time.
• As of early 2026, the OpenSAFELY platform has expanded to non-COVID research, unlocking secure, whole-population analysis across 59 million electronic health records.
• Clinical guideline implementation on the ground remains patchy, frequently delaying life-saving therapies like SGLT2 inhibitors for eligible patients.
• Translating massive health data sets into everyday frontline interventions requires dedicated clinical capacity and robust primary care support systems.
• Utilising specialised medicines management services allows practices to deploy clinical pharmacists via the ARRS framework to easily operationalise these data insights.

Introduction

Clinical guidelines are designed to outline what “good” care looks like, yet their implementation on the front line remains one of primary care’s greatest modern challenges. Across the 6,000 GP practices spanning England, the care a patient receives can vary widely based on socioeconomic factors, practice capacity, and regional resource constraints. While the health system frequently analyses health inequalities by looking at downstream patient outcomes, it rarely moves upstream to investigate why equitable guideline adoption fails in the first place.

To explore this critical intersection of data science, policy, and clinical practice, this article reviews a recent lecture organised by The Medicines Management Team (TMMT). The session was led by Kareem Mohamed (MPharm, PGCert IP), an experienced primary care pharmacist and a DPhil candidate at the University of Oxford’s Nuffield Department of Primary Care Health Sciences. Based at the Bennett Institute for Applied Data Science, Mohamed’s current research focuses directly on the equitable implementation of clinical guidelines in primary care, using diabetes as a clinical lens. This comprehensive summary breaks down how population-scale data is being leveraged in 2026 to change how primary care networks audit, manage, and optimise chronic disease paths.

The Scale of Unwarranted Variation on the Frontline 

The pressure on the UK primary care framework is intensifying, driven largely by the rising prevalence of chronic metabolic conditions. Currently, the diabetes register in England has grown to include approximately 4.1 million individuals. This represents roughly 7.9% of the adult population and consumes an estimated 10% of the NHS’s entire £200 billion budget. 

Faced with this massive public health burden, the frontline primary care team is tasked with preventing avoidable downstream complications; such as cardiovascular events, renal failure, and amputations, by delivering consistent, evidence-based therapy. However, as Mohamed highlighted from his years of experience working within some of the most deprived primary care networks in Bradford, there is a substantial difference between a guideline on paper and a consultation in the real world.

The core challenge centers on distinguishing between justified clinical variation where a clinician alters a treatment target to suit an individual patient’s frailty or preference and unwarranted variation, where care gaps occur simply due to operational friction, systemic inequalities, or a lag in local protocol updates. Without clear data insights that look beyond a single practice or PCN, identifying and addressing these systemic shortfalls remains incredibly difficult.

Where Guidelines Meet the Real World: The SGLT2i Paradigm 

A clear example of the delay between clinical evidence and widespread practice adoption can be seen in medicines optimisation, specifically regarding SGLT2 inhibitors (SGLT2is). This class of medication offers excellent glycemic control alongside vital cardio-renal protection. While national guidelines emphasise their use for eligible type 2 diabetes profiles, their real-world uptake on the ground can be slow, patchy, and unequal.

Are these advanced therapies reaching the right patients at the right time? Conversely, are they being safely de-prescribed when a patient enters advanced stages of frailty?

To answer these questions systematically, Mohamed’s research at Oxford begins with a comprehensive scoping review focused on clinical guideline implementation in primary care. By evaluating the specific barriers and facilitators that dictate how guidelines move from paper to practice, this foundational research aims to map out exactly where health inequalities begin to develop in the standard consultation room.

“Clinical guidelines are only as good as their adoption and implementation. NICE can recommend a therapy, the evidence can be excellent, and yet uptake on the ground can be slow, patchy and unequal.” — Kareem Mohamed

OpenSAFELY: Whole-Population Analytics in 2026 

Historically, analysing prescribing habits and clinical variations across the entire country was slowed by data fragmentation and strict privacy guardrails. However, the data analytics landscape changed significantly in early 2026. OpenSAFELY; a highly secure data analytics platform developed by the Bennett Institute for Applied Data Science and enabled by NHS England officially expanded its scope beyond COVID-19 to facilitate non-COVID research projects. 

This expansion allows researchers to study first- and second-line type 2 diabetes management at an unprecedented scale. By analysing near real-time, population-scale data, the platform can quantify the exact gap between current clinical practice and best-practice care. For primary care clinicians, GP partners, and practice managers, several operational features make OpenSAFELY a trusted solution for population health management: 

• Zero Data Extraction: The underlying patient data never leaves the secure hosting environments of the major GP IT suppliers (TPP and EMIS/Optum). The analysis code travels to the data, rather than the data traveling to the researcher.
• Absolute Transparency: Every line of code utilised for data analysis is published openly in the public domain, creating a permanent, auditable log of what queries were run and by whom.
• Unrivaled Scale: The platform covers more than 95% of individuals registered with a GP in England, aggregating the records of roughly 59 million people.
• Institutional Backing: Due to its rigorous security architecture, the platform maintains the explicit support of the BMA, the RCGP, and prominent privacy advocacy groups.

By leveraging this whole-population dataset, the ultimate goal of this research is to provide robust, evidence-backed recommendations to inform national commissioning structures and incentive frameworks. This ensures that future policy updates are shaped by real-world clinical behavior, helping close the care gaps seen today. 

Operationalising Data Insights via ARRS Portfolios 

While population-scale research is essential for shaping national strategy, individual GP practices still face the immediate challenge of managing day-to-day operations. Frontline teams must balance workforce shortages and rising patient demand while trying to proactively audit their registers, identify eligible candidates for SGLT2 inhibitors, and address local care gaps.

This is where utilising structured medicines management services becomes an operational necessity. Many practices struggle to maximise their NHS Additional Roles Reimbursement Scheme (ARRS) allocations, often losing funding or lacking the internal infrastructure to manage specialised staff effectively. Partnering with a dedicated organisation like The Medicines Management Team (TMMT) provides an immediate, ready-made framework to help solve this capacity problem.

By integrating remote clinical pharmacists ARRS personnel into a practice’s workflow, primary care networks can easily hand over the time-consuming work of chronic disease auditing and medication reviews. These remote pharmacists can run targeted local audits that mirror national population data, identifying under-treated patients, correcting prescribing variations, and ensuring full compliance with the latest clinical guidelines. This clinical support unburdens GPs and practice managers, allowing them to focus on active clinics while ensuring their practice metrics are fully optimised.

Conclusion 

The insights from Mohamed’s research show that closing the diabetes care gap requires more than just publishing updated clinical guidelines; it demands an understanding of how those guidelines are implemented on the front line. By combining implementation science with secure, population-level health data, primary care can move away from uncoordinated care models and adopt proactive, equitable health management.

Turning these data insights into practical frontline care requires consistent clinical focus and capacity. If you are looking to optimise your chronic disease pathways, improve patient outcomes, and reduce your team’s clinical workload, consider exploring the tailored workforce solutions available on TMMT’s medicines management service page.

Frequently Asked Questions