Your browser is no longer supported

For the best possible experience using our website we recommend you upgrade to a newer version or another browser.

Your browser appears to have cookies disabled. For the best experience of this website, please enable cookies in your browser

We'll assume we have your consent to use cookies, for example so you won't need to log in each time you visit our site.
Learn more

The challenges of managing type 2 diabetes in primary care

  • Comment

The challenges of managing type 2 diabetes in primary care. This is an extended version of the article published in Nursing Times; 104: 7, 32-33.

  • This article has been double-blind peer reviewed
  • Figures and tables can be seen in the attached print-friendly PDF file of the complete article found under “related files”



Freda Mold, PhD, BSc, is research associate; Angus Forbes, PhD, MSc, BSc, RGN, RHV, DNCert, is senior lecturer; Alison While, PhD, MSc, BSc, RGN, RHV, is professor of community nursing; all at the Florence Nightingale School of Nursing and Midwifery, King’s College London.


Mold, F. et al (2008) The challenges of managing type 2 diabetes in primary care. This is an extended version of the article published in Nursing Times; 104: 7, 32-33.
BACKGROUND: Type 2 diabetes poses many challenges for primary care staff in preventing and managing complications. Effective management is important since the disorder is associated with serious complications which may lead to a reduction in both longevity and quality of life.
METHOD: A structured audit of patient electronic medical records (n=646) was undertaken in three general practices in an ethnically diverse outer London PCT.
RESULTS: Findings indicated that patients from minority ethnic groups were more likely to have poor glycaemic control, with Asian patients showing higher levels of overall BMI risk than other patient groups. A large proportion of patients with poor glycaemic control were aged 50 years or under, and this group had less frequent contact with primary care professionals.
DISCUSSION: This audit suggests there are significant relationships between diabetes outcomes, patient characteristics and the care process. These relationships are, in part, related to resource levels, management systems for delivering care and complexity of patient needs.
CONCLUSION: It is necessary to recognise these issues if future interventions are to be developed to better address patients’ diverse needs in metropolitan areas.


Type 2 diabetes affects at least two million people in the UK and its prevalence, linked to increasing levels of obesity, is rising rapidly. Type 2 diabetes is a heterogeneous metabolic disorder associated with serious complications including hypertension, stroke, kidney failure, heart disease and neuropathy. It leads to a significant reduction in both longevity and quality of life. While there is strong evidence to show that achieving certain clinical targets reduces the risk of these complications, most patients fail to achieve them (Strutton et al, 2000).

Most type 2 diabetes is managed in primary care settings. Increasingly, this care is led by practice nurses and/or GPs with an interest in diabetes, although there are large variations in the quality and range of services provided (Farooqi et al, 2004).

The most recent development in primary care has been the introduction of the quality and outcomes framework (QOF) as part of the general medical services contract (British Medical Association, 2006; Department of Health, 2004; 2003). This framework sets specific targets - linked to payments - for diabetes based on both care processes (assessment and treatment provision) and outcomes (improved glycaemic control).

However, delivering effective diabetes care is challenging, despite these incentives and guidelines. It is particularly challenging in areas of high socio-economic deprivation and where there are large populations of minority ethnic groups with a high prevalence of diabetes.

This article reports an audit of three general practices with high levels of ethnic diversity within a deprived North London borough. The audit explored the relationship between patient characteristics, the type of care provided and clinical outcomes.


An audit of patient electronic medical records was conducted using EMIS (Primary Care System Practice edition). Data was collected over an 11-month period from September 2004 to July 2005. The local research ethics committee approved the study.

Audit objectives
The objectives of the audit were to:

  • Describe the diabetes management systems in each practice;

  • Identify key practice characteristics;

  • Audit care processes and outcomes against best practice guidelines and targets;

  • Explore the relationship between practice characteristics, patient factors and clinical outcomes.

Inclusion criteria
Patients were selected using the following inclusion criteria:

  • Aged 18 years or over;

  • A diagnosis of type 2 diabetes (recorded in the electronic record).

Data extraction
An extraction tool was used to collect data from EMIS. The tool was based on established national audit systems (NHS Information Authority, 2004a; 2004b; Masding et al, 2003) and validated through consultation with medical staff in one practice. The tool content areas are presented in Fig 1.

Fig 1. Extraction tool contents

Data was extracted on all patients with type 2 diabetes. The initial identification of patients was undertaken by practice managers who had knowledge of the coding structures in each practice. Once the patient ID number had been provided, a researcher was able to review each individual patient record and extract the relevant clinical data. At this stage, the accuracy of the coded diagnosis was checked in the fuller clinical record. Frequency of contact data was retrieved by counting the number of diabetes-related consultations (face-to-face or telephone contact) over the audit period.

The extraction tool was used to collect data retrospectively over a 14-month period. This time frame was adopted because many clinical objectives in diabetes are based on a 12-month cycle incorporating at least one comprehensive assessment or annual review.

Data analysis
Two stages of data analysis were undertaken. The first involved analysing the audit data against standards for care delivery (for example, whether the patient has had an eye or foot check within the last 14 months) and clinical outcome targets for glycaemic and metabolic control (NICE, 2002a-d). In the second stage, the data was modelled to explore the relationship between independent factors such as patient characteristics (sex, age, ethnicity) and care processes (interventions, screenings, diabetes outcomes) in relation to glycaemic and metabolic outcomes. Frequency of contact was also explored in relation to independent factors such as patient age. The associations between these variables were tested using chi-square tests and, where appropriate, binary logistic regression. The analysis was performed in SPSS version 14.


Practice characteristics
Variations were found between the three practices in relation to the number of practice nurses and GPs providing care and the overall size of the practice, in terms of both the total number of patients registered and those with diabetes (Table 1).

Table 1. Practice characteristics

Practice codeMedical staff (whole-time equivalent)GP with special interest in diabetesPractice nurses (whole-time equivalent)Patient list size (approx)Patients with type 2 diabetes
A7 (5.5)P2 (1.5)9,500310
B5 (4.5)P3 (1.5)5,500163
C7 (5.0)P3 (2.08)7,500173

The structure and resourcing of diabetes clinics differed between practices. Practice A held a weekly clinic led by a GP with a special interest in diabetes and a practice nurse, with the support of an on-site podiatrist and dietitian. It also had access to a practice-based counsellor and psychotherapist. Practices B and C held monthly and fortnightly clinics respectively. These were also run by a GP with a special interest in diabetes and practice nurse, with podiatry and dietetic services available on referral.

While these practices were well resourced in terms of access to different professional disciplines, each encountered varying levels of difficulty in responding to patients’ healthcare needs. These difficulties reflected the low socio-economic status of the area and the high proportion of minority ethnic patients, many of whom had English as a second language. These factors increased both the level of demand and the complexity of care needed (Morgan and Baker, 2006; Farooqi et al, 2004; Census, 2001; Goddard and Smith, 2001).

Patient characteristics
A total of 646 patients were included in the audit; their characteristics are summarised in Table 2. At the time of the audit, recording of ethnicity was not mandatory, so recording of ethnic status varied between practice groups. This resulted in only one practice consistently recording this data.

Table 2. Patient characteristics

Sexn (%)
Male351 (54.3)
Female295 (45.7)
Under 50 years106 (16.4)
50-59 years118 (18.3)
60-69 years201 (31.1)
70 years and over221 (34.2)
White British/Irish/other103 (15.9)
Indian/Pakistani/Bangladeshi/other Asian91 (14.1)
Black Caribbean/African181 (28.0)
Not known271 (42.0)
Years since diagnosisMean 7 (SD 6.102)
Medication/dietary control
Insulin/oral hypoglycaemic agents/combination553 (85.6)
Diet only93 (14.4)
Hypertensive therapy
No hypertensive therapy315 (48.8)
Receiving one or more hypertensive drugs331 (51.2)
Overall total cholesterolMean 4.6 (SD 1.09645)
Ex-smokers117 (18.1)
Current smokers123 (19.0)

Care processes
Table 3 details the number of patients who received specific diabetes assessments and interventions during the audit period.

Table 3. Care processes

Lifestyle issuesn (%)
Smoking management/action227 (35.1)
Seen dietitian208 (32.2)
General health education518 (80.2)
Diabetes consultations and medication review 
Number of annual diabetes consultations in 14 monthsMean 7.7 (SD 5.047)
Medication reviews in previous 14 months509 (78.8)
Feet and eye examination 
Feet examination516 (79.9)
Eye examination549 (85)
Retinopathy screening500 (77.4)

There were differences in the type of diabetes care received by the different patient groups.

Consultation frequency over the audit period varied between age and ethnic groups. Patients aged 70 and over (n=221) had a mean consultation rate of 8.4 (95% CI 7.7-9.1) compared with 6.7 (95% CI 6.03-7.41) in those aged 50 and under (n=106; f=2.4, 4df, p=0.047). Asian patients (n=91) also had a significantly lower consultation rate of 6.7 (95% CI 5.9-7.53) compared with an overall mean of 8.4 (95% CI 7.8-9.1) (f=4.1; 3df; p=0.006). Low consultation rates for Asian patients could, however, be related to age since a higher proportion of this group were aged under 50 (28.6%, n=26).

There was a low level of referral to dietetic services across all practices, with almost 70% of patients not having seen a dietitian (67.8%; n=438). While there were no statistically significant differences between age or ethnic groups, black patients had proportionally the lowest level of dietetic assessment, with 72.4% (n=131) not having seen a dietitian compared with 59.2% (n=61) of white patients.

The majority of patients received some form of general health education (80.2%; n=518) (lifestyle, dietary, smoking cessation, medication and/or exercise). However, provision varied between patient groups, with a higher proportion of white patients receiving general health information (88.3%; n=91) than black patients (77.9%; n=141).

Clinical outcomes
Three main markers of glycaemic and metabolic control were examined: glycated haemoglobin (HbA1c); body mass index (BMI) (obesity is a key determinant of metabolic health); and blood pressure (BP is a strong predictor of cardiovascular morbidity and mortality in diabetes). The standards adopted for the study are summarised in Fig 2.

Fig 2. Standard guideline

Glycaemic control
Nearly half of patients (48.3%; n=301) had poor glycaemic control with HbA1c levels over 7.6%. Only 12% (n=80) had good control (HbA1c <6.5%). Black patients (n=181) were significantly more likely to have poor glycaemic control (54%; n=95) than white patients (48%; n=49) (c213.02; 6df; p=0.043). Glycaemic control also differed between age groups, with a higher proportion of patients with poor control being aged 50 or under (58%; n=58) or aged 50-59 years (57.9%; n=66) (c216.08; 6df; p=0.013). After adjusting for age and duration of illness, patients in practice B had significantly better glycaemic control than patients in practice A (f=5.9; 2df; p=0.0003), with a mean HbA1c of 7.4% (95% CI 7.2-7.7) compared with practice A with a mean HbA1c of 8.2% (95% CI 8.02-8.4). Level of service use over a 14-month period was also associated with level of glycaemic control. Patients with good glycaemic control (HbA1c <6.5%) had a mean of nine consultations (95% CI 7.8-10.5), while those with poor control had a mean of 7.5 consultations (f=3.2; 2df; p=0.036; 95% CI 6.9-8.1). Glycaemic control by mean number of consultations is summarised in Table 4.

Table 4. Glycaemic control

HbA1c (%)n (%)Mean no. of consultations (SD)95% confidence intervals for mean
<6.4 (good)73 (12.3)9.2 (SD 5.833)7.8-10.5
6.5-7.5 (borderline)232 (39.1)7.5 (SD 4.637)6.9-8.1
>7.6 (poor)288 (48.6)7.5 (SD 5.094)6.9-8.1

Body mass index
Almost half of all patients had a BMI >30 (46%; n=285). A higher proportion of those with a BMI of 30 and over were female and/or aged over 60.

Ethnic differences around BMI are more complex as different populations have lower risk thresholds. Hence all Asian patients with BMI recorded (n=85) had a suboptimal BMI (BMI>23) while fewer than half white (46.9%; n=46) and black (42.2%; n=70) patients had suboptimal (BMI>25) BMIs (c2=85.2; 2df; p<0.001). BMI risks by ethnicity are summarised in Table 5.

Table 5. BMI risk by ethnicity (missing n=27)

High-risk BMI >23/>25n (%)
White46 (46.9)
Asian85 (100)
Black70 (42.2)
Not stated136 (50.4)

Blood pressure
While most patients met the general standards for systolic (140mmHg) (74.3%; n=462) and diastolic (80mmHg) blood pressure (72.7%; n=452) (Table 6), two-thirds (63%; n=392) failed to achieve the Joint British Societies’ (2005) more stringent systolic target of 130mmHg, and one-fifth (19.7%; n=123) were outside the diastolic target of 85. Poorer systolic control was found among black patients and those aged 60 and over. A higher proportion of patients with poor diastolic BP control were aged under 50. Systolic blood pressure outcomes by age are summarised in Table 7.

Table 6. Blood pressure outcomes (missing n=24)

Systolicn (%)
140mmHg462 (74.3)
>140mmHg160 (25.7)
80mmHg452 (72.7)
81-90mmHg138 (22.2)
91mmHg32 (5.1)

Table 7. Systolic blood pressure by age (missing n=24)

Systolic<50 n (%)50-59 n (%)60-69 n (%)70 n (%)
140mmHg91 (89.2)85 (73.9)139 (70.9)147 (70.3)
>140mmHg11 (10.8)30 (26.1)57 (29.1)62 (29.7)


The audit suggested there are significant variations in primary care services for diabetes. These variations are in part related to the resources and level of demand that practices have to manage. This was confirmed by feedback provided by the clinical teams in reviewing their findings. For example, the explanation given for the low level of referrals to dietetic services was that this service was in short supply.

To compensate for limited local dietetic, district nursing and podiatrist services, GPs and practice nurses at practices B and C adopted a highly structured approach and provided educational advice themselves (such as dietary, activity and foot care advice) at each consultation. Educational advice was, therefore, provided routinely at each consultation rather than patients being referred to outside agencies.

The way practices adapt their care system to the resources available may be important in maximising the effectiveness of the care they provide. For example, although practice B had fewer GPs and practice nurses and poorer access to support services (dietetic, district nursing and podiatry services), patients in this practice had significantly better glycaemic control.

Variations emerged in terms of different systems of data recording and retrieval at each practice, which may affect their ability to shape future service development. For example, practice B had comprehensive recording of socio-demographic data and coverage of clinical data. This may be essential in undertaking audit studies and informing future service delivery in terms of identifying and recalling specific patient groups at risk of complications.

The audit also revealed a relationship between level of service use at each practice over a 14-month period and glycaemic control, suggesting that care processes such as frequency of diabetes clinic and diabetic review/follow-up affects clinical outcomes. Service use was lower in the younger age groups, which may be related to lifestyle factors. Future studies might explore these further to identify ways of encouraging greater service use in these populations.

The audit confirmed that one of the continuing challenges in diabetes in primary care is to provide better care to marginalised groups. It suggested that the younger black populations in particular experience underprovision. This is a worrying trend as these patients will be accumulating significant macro- and micro-vascular damage.

In conclusion, the audit indicated that many patients in primary care have suboptimal glycaemic control with the associated risks of complications and mortality. While the new QOF system has had some impact on glycaemic control (Gulliford et al, 2007), there is still much work needed if the majority of patients are to achieve better clinical outcomes.

Finally, the audit also highlighted the continuing challenge of managing obesity within existing healthcare systems. Clearly this problem extends beyond diabetes - strategies to promote weight loss in primary and other care settings are a priority.

Implications for practice

  • The continued provision of general health education (lifestyle, diet and exercise) by primary care staff may be important in preventing further BMI risk for specific patient groups.

  • Good glycaemic control is associated with greater service use, suggesting the importance of regular medical and nursing contact for patients with poorer glycaemic control.

  • Monitoring younger patients’ frequency of service use and exploring possible reasons for non-attendance may be necessary to better manage poorer glycaemic control.

  • The comprehensive recording of ethnicity data is essential if audit studies are to be undertaken to inform future service delivery.

  • There is a need to develop primary care support services, such as district nursing care, podiatry and dietetic services to facilitate care already being provided by general practices.

We would like to thank the three general practices for allowing us access to their practice data and for their assistance throughout this project.

  • Comment

Have your say

You must sign in to make a comment

Please remember that the submission of any material is governed by our Terms and Conditions and by submitting material you confirm your agreement to these Terms and Conditions. Links may be included in your comments but HTML is not permitted.