Case study: Our work with the National Institute for Health and Care Excellence (NICE)
You can read the peer-review publication in the Lancet's eClinicalMedicine by clicking here.
NICE guidance development utilises a number of evidence sources. Potential gaps arise in the evidence where studies have not been conducted or are not yet published, or more holistic health impacts have not been quantified. This is especially likely for patients with comorbidities as they are often excluded from clinical trials. Real world evidence (RWE) can provide a valuable source of evidence for assessing groups of people who are excluded from RCTs (randomised control trials) owing to the presence of comorbidities. RWE can also cover a wider set of factors and outcomes, thus better reflecting the totality of patients’ health needs and the impacts on health systems.
Multimorbidity, living with two or more chronic conditions, is increasing in the UK and globally, including for patients living with type 2 diabetes (T2DM). Multimorbidity impacts both the health needs of patients and the resource implications for health systems, yet there is little evidence detailing patterns of multimorbidity, estimating impacts, and identifying patient groups most likely to benefit from more holistic approaches.
What we did
We worked with the National Institute for Health and Care Excellence (NICE) to investigate multimorbidity in patients with T2DM. Our specific aim was to identify comorbidity patterns, how they vary across patients with T2DM and over time, and the impacts of this on health outcomes and resource utilisation.
We used a RWE dataset - the Discover dataset is accessible via Discover-NOW Health Data Research Hub for Real World Evidence through their data scientist specialists and IG committee-approved analysts, hosted by Imperial College Health Partners. Discover-NOW covers approximately 2.5 million people across North West London and in which primary care is linked to secondary care records. We identified 35 comorbidities of interest. A mix of traditional epidemiological approaches and more innovative machine learning methods were used to investigate how comorbidities clustered, how this varied across patient groups according to risk factors and comorbidities at diagnosis, and the implications of this for patient outcomes and healthcare costs.
Multimorbidity is common in patients with T2DM and increasing. More than 75% of patients with T2DM had at least one comorbidity when their T2DM was diagnosed, with around 50% having two additional comorbidities or more (Figure 1). The number of comorbidities increased over time, with nearly a three-fold increase from 2000 to 2019; and a doubling from the point of diagnosis to 10 years later (Figure 1).
Figure 1: Proportion of T2DM patients with 0-10 comorbidities at diagnosis and 2, 5 and 10 years after diagnosis of T2DM
Some common comorbidities were expected, others less so. Hypertension was the most common comorbidity of those studied. In 2000, 1 in 3 patients had hypertension, increasing to 1 in 2 by 2019. Other vascular-related conditions with strong associations with T2DM were also common, such as retinopathy and ischemic heart disease. However, some of the most common comorbidities are not traditionally associated with T2DM. Back pain (40%) and depression (16%) are the second and fourth most common comorbidities in those with T2DM, while osteoarthritis (16%) and asthma are more prevalent than many vascular diseases (Figure 2).
Figure 2: Age standardised comorbidity prevalence between 2000 and 2019 in T2DM patients. 7 most common comorbidities included.
Comorbidities vary substantially across patients with T2DM and so do outcomes. We found comorbidities varied substantially across patients with T2DM according to age, sex and ethnicity, as well as risk factors and comorbidities at diagnosis. For example, those who had obesity, hypertension or renal disease had very different comorbidity profiles and large differences in health care usage, which has implications both for the patients and for those providing care.
You can explore our findings in our interactive Type 2 Diabetes Comorbidity Explorer by clicking here.
Our full analysis, including analysis of these sub-populations and impacts on healthcare usage, will be submitted to a peer-reviewed journal in due course. Please get in touch if you would like to be among the first to receive this information when available.
More broadly, our findings and similar approaches could be used to:
- forecast the burden of T2DM more holistically, incorporating comorbidities and healthcare usage to inform resource planning;
- inform prioritisation of T2DM guidelines that should be updated in the context of comorbidities and impacts on patients;
- inform realigning care pathways for patients with T2DM according to the most common groups of comorbidities and health needs in order to improve patient care and reduce resource utilisation;
- develop a framework to inform personalised risk prediction models that are more holistic and more granular; and
- estimate the impacts of interventions from a broader health perspective compared with traditional approaches.