Contents
I / INTRODUCTION / 3
4
II / POPULATION NEED
2.1: Disease prevalence / 4
2.2: Premature disease / 5
2.3: Comorbidity / 5
2.4: Levels of service utilisation relative to need / 6
III / CLINICAL SERVICE DELIVERY & KEY OUTCOMES / 8
3.1: Coronary Heart Disease / 8
3.2: Stroke / 12
3.3: Type 2 Diabetes / 17
3.4: Discussion points / 22
Foreword
This paper provides some initial insights into the challenges and achievements of NHS GG&C’s care for patients with three major chronic diseases, Coronary Heart Disease (CHD), Stroke and Type 2 Diabetes through its programme of Local Enhanced Services. LES are contractual arrangements with primary care services, designed to augment the basic GMS Quality & Outcomes Framework (QOF) specification. Enhanced services for these three chronic diseases have now been established for over five years in Greater Glasgow and are being incrementally introduced into Clyde. The LES contracting process generates a rich quantity of clinical and process data, which is essential for programme delivery. However, these data are also enormously valuable in building a picture of the needs and characteristics of patients with chronic disease and in helping us to understand how our local healthcare system is responding to their needs. In this short introductory paper, we provide a brief overview of these themes, drawing on data from the LES to provide important insights into the following aspects of CHD, Stroke and Type 2 Diabetes:
- Estimates of population need
- Levels of service utilisation relative to need
- Variations in care processes
- Variations in clinical status
- Variations in clinical management
These themes will be further developed in a series of detailed thematic reports during 2010.
Anne Scoular,
Keep Well/Enhanced Services Data Group Chair
April 2010
IIntroduction
In common with many developed healthcare systems, NHS GG&C delivers an evidence-based programme of chronic disease management (CDM); randomised trials show that effective CDM programmes, incorporatingproactive case-finding, assessment, care and ongoing monitoring of people with chronic disease, are highly cost effective. Not only do they achieve significantly reductions in hospitalisation rates and bed days, but they also improve health status, increase participation in health-promoting behaviours, improve communication with physicians and reduce fatigue, disability and the social/role limitations associated with chronic disease. Although CDM operates at a basic level through the GMS Quality & Outcomes Framework (QoF) contract across the entire GG&C area, Local Enhanced Services (LES) contracts extend the clinically based QoFinto the all-important health related behaviour areas, using the practice nurse annualinterview to screen for problems and support onward referral to appropriate serviceswho can provide ongoing support for individuals. This includes compliance; up tohalf of all people with long term conditions do not take, or do not regularly take, theirprescribed medication, with a serious impact on their quality of life, clinical outcomesand cost effectiveness. Coverage of the GG&C population is, however, not yet complete, as summarised below:
- Diabetes: Established across entire GG&C area
- CHD: Established across most of the GG&C area, except Renfrewshire.
- Stroke: Old Greater Glasgow area only (not operational in Clyde).
The completeness of chronic disease registers allows some estimation of the prevalence of these diseases at community level in the CHCPs where the LESs are established. Caution should be used in interpreting population prevalence data in areas where there is only partial coverage by the LES. This applies in North Lanarkshire, South Lanarkshire and West Dunbartonshire. Data from North and South Lanarkshire have therefore been excluded from this paper and will be presented separately, as the NHS GG&C CDM programme only operates in a small proportion of the entire Local Authority area. Data from West Dunbartonshireare also complex to interpret, as the new CHCP incorporates some practices who were previously part of the old Argyll & Clyde Health Board area and others who were within Greater Glasgow. As the CHD and diabetes LESs are now established in West Dunbartonshire, valid prevalence estimates can be generated at CHCP level. However, the stroke LES is only operationalin practices within the old ‘Greater Glasgow Health Board’ locality, thus the numbers of patients on the stroke register has not been usedto estimate CHCP level stroke prevalence.
The information presented in this paper largely draws on data from the 2008/09 contracting year, with appropriate comparisons with previous years when relevant.
IIPopulation need
2.1Disease prevalence
Crude CHD and stroke show twofold variations in prevalence within GG&C (Figure 1), with the highest prevalence in East Glasgow and the lowest inEast Renfrewshire. Crude rates of disease reflect the actual prevalence of disease.
Figure 1: Crude CHD, Stroke & Type 2 Diabetes prevalence, by CHCP
CHD, stroke and, to a lesser extent, Type 2 diabetes are all commoner among older people. South Asian populations have an approximately fourfold increased risk of Type 2 Diabetes. Thus, some of thisvariation could simply be due to their differing population structures. However, even after standardisation for age and sex, the steep internal gradient in disease prevalence within GG&C remains (Figure 2).
Figure 2: Age & sex standardised disease prevalence, by CHCP
As there are no current reliable estimates of the ethnic composition in each CHCP, it was not possible to standardisefor ethnicity.
2.2Premature disease
The GG&C population experiences more premature cardiovascular disease compared with other regions and developed nations. There are also large spatial variations in disease prevalence within GG&C, with afive-fold variation between East Renfrewshire and East Glasgow in CHD prevalence in those aged less than 65 years (Figure 3).Socioeconomic deprivation explains some of this effect, acting through a variety of different mechanisms.
Figure 3: CHD, Stroke & Type 2 Diabetes prevalence (per 1,000),males <65yrs
2.3 Comorbidity
87,554 individual patients appeared on one or more of the three disease registers during the financial year 2008/09; 16,818 (19%) had two diagnoses and 1,193 (1%) had all three (Figure4).
Figure 4: Comorbidity
In eight CHCPs, the LES programme is fully implemented across all three disease areas; in these areas,we can use the LES data to estimate the population prevalence of individuals with at least one chronic disease(ie diabetes, CHD and/or Stroke).Again, the highest prevalence (110.4 per 1,000) is observed in East Glasgow and the lowest (50.5 per 1,000) in East Renfrewshire(Figure 5,overleaf).
Figure 5: Crude chronic disease prevalence,* by CHCP, 2008-09
*Diagnosed with: Diabetes, CHD and/or Stroke as proportion of total CHCP population
2.4: Levels of service utilisation relative to need
The total volume of individual diagnosed on the diabetes, CHD and Stroke registers in each CHCP is shown in Figure 6. This includes those who are exception coded and also does not reflect the number of individuals involved, as many have more than one disease (see previous section).
Figure 6: Number of chronic disease registrations in 2008-09, by CHCP
Of thetotal cohort of patients on each disease register, the overall proportion who received disease monitoring in 2008/09 was 70.9% for stroke patients,77.2% for CHD and 83.5% for diabetes, with significant variations between CHCPs (Table1).Thebetween-CHCP variation was greatest for stroke (ranging from 56.9 to 75.0%) and least for diabetes (72.7 to 86.5%). As with all healthcare performance monitoring, there are many possible reasons for these variations and adjustment for age, sex, comorbidity and other determinantsof casemix within each CHCP would be required before drawing any substantive conclusions.
Table1: Proportion of patients who received monitoring in 2008/09, by CHCP
Proportion of patients on each disease register who received disease monitoring in 2008/09Type 2 Diabetes / CHD / Stroke
CHCP / % (95% ci) / CHCP / % (95% ci) / CHCP / % (95% ci)
Inverclyde / 86.5 (85.1 to 87.7) / West Dun / 80.3 (79.0 to 81.4) / West Dun / 75.0 (72.4 to 77.4)
West Dun / 83.0 (81.7 to 84.2) / Inverclyde / 77.4 (76.1 to 78.7) / SW Glasgow / 71.4 (69.4 to 73.3)
Renfrewshire / 80.3 (79.3 to 81.2) / SW Glasgow / 75.6 (74.2 to 76.8) / W Glasgow / 70.2 (68.5 to 71.8)
SW Glasgow / 78.8 (77.5 to 80.1) / W Glasgow / 73.4 (72.2 to 74.5) / East Dun / 65.0 (63.1 to 66.9)
E Glasgow / 77.8 (76.7 to 78.9) / East Dun / 72.1 (70.7 to 73.4) / E Glasgow / 65.0 (63.5 to 66.6)
East Ren / 77.6 (75.5 to79.5) / E Glasgow / 71.9 (70.9 to 72.9) / N Glasgow / 60.6 (58.5 to 62.8)
W Glasgow / 77.0 (75.8 to78.1) / East Ren / 66.8 (64.8 to 68.7) / East Ren / 56.9 (54.1 to 59.7)
East Dun / 76.8 (75.3 to 78.2) / N Glasgow / 65.3 (63.8 to 66.8) / SE Glasgow / 56.9 (55.0 to 58.7)
SE Glasgow / 73.7 (72.4 to75.0) / SE Glasgow / 62.9 (61.6 to 64.2)
N Glasgow / 72.7 (71.1 to 74.3)
Overall / 83.5 (83.2 to 83.9) / Overall / 77.2 (72.2 to 74.5) / Overall / 70.9 (70.2 to 71.6)
III:Clinical service delivery & key outcomes
3.1: Coronary Heart Disease
Of the 44,635 patients on the disease register, 7,867 (17.6%) were exception coded. Results presented from this point onwards relate to the remaining 36,768 patients. 20,839 (57%) were male and 15,929 (43%) female. Women were, on average, slightly older (mean age 73.6 years) than men (mean age 69.5 years). Ethnicity information was available in 29,822 (81%) of patients. Overall, 28,844 (96.7%) were white (Table 2)
Table 2: Non-exception coded CHD patients, by ethnic subgroup
Ethnic subgroup / Number (%)White / 28,844 (96.7%)
South Asian / 746 (2.5%)
Black / 21 (0.07%)
Chinese / 12 (0.04%)
Other ethnic subgroups / 150 (0.5%)
TOTAL (known ethnicity) / 29,822 (100%)
There were significant variations across GG&C in the ethnic composition of the CHD register; non-white patients represented a very small minority (0.7%)of CHD patients in West Dunbartonshire, but10.3% in SE Glasgow (Figure 7). In SE Glasgow, non-white subgroups were virtually all South Asian, whereas North Glasgow, Inverclyde and West Dunbartonshire were more diverse.
Figure 7: Spatial variations in ethnic distribution of diagnosed CHD patients
The patient’s postcode of residence was known in 34244 (93%), allowing estimation of small area deprivation. The vast majority of GG&C’s CHD patients live in highly deprived areas, with an inverse relationship between ageand deprivation (Figure 8).
Figure 8: Mean age by SIMD deprivation quintile of diagnosed CHD patients
Cholesterol monitoring:27,766 (76%) patients had total cholesterol recorded in the past year, of whom 20,187values were considered to be valid (2-15 mmol/l). Of these, 15,706 (78%) met the target cholesterol value (<5 mmol/l). There was considerable variation by CHCP in the proportion of patients monitored, ranging from 63% in SE Glasgow to 88% in SW Glasgow, however the proportion of patients who met the target cholesterol value was similar across the system (Figure 9).
Figure 9: Cholesterol monitoring activity & outcomes in CHD patients, by CHCP
Support with stopping smoking:30,988 (84%)patients had smoking status recorded in the past year. Overall, 7,476 (24%) of CHD patients were current smokers; however this ranged from 7.4% in the least deprived areas to 28.5% in the most deprived (Figure 10).
Figure10: Smoking in CHD patients, by deprivation quintile
The proportion of CHD patients in whom smoking status was recorded ranged from 69% in SE Glasgow to 96% in SW Glasgow (Figure 11, overleaf). Around one third of smokers were recorded as wanting to stop, although again this varied from 17% in Inverclyde to 42%in North Glasgow. The proportion of all smokers referred to smoking cessation services ranged from 13% in Inverclyde to 61% in East Renfrewshire; in four CHCPs, the number of smokers referred to smoking cessation services exceeded those wanting to stop.
Figure 11: Documented support to CHD patients who smoke, by CHCP
Recording of Body Mass Index (BMI):23,940 (65%) patients had a BMI recorded in the past year. Of these, 2,236 (9.3%) were excluded from the analyses as their BMI fell outside aplausible range of 10-150. Of the remainder, the mean BMI was 28.5 and only 5,575 (26%)CHD patients wereof healthy weight (BMI 18.5 – 25); this proportiondecreased from 20 to 14% with increasing deprivation (Table 3).
Table3:Proportion (%) CHD patients with healthy weight
SIMD Quintile / Number (%)1 Most deprived / 2228 (14.2%)
2 / 1114(16.0%)
3 / 576 (14.0%)
4 / 684 (17.2%)
5 Least deprived / 686 (19.6%)
TOTAL (known postcode) / 5,288 (100%
Figure 12: BMI in CHD patients, by CHCP
Assessment of mental health:Assessment of patients’ mental wellbeing, usingHADS scoring for anxiety and depression, is an important component of the annual review. Valid reasons for excluding patients from this component of the review includeongoing psychological care, receipt of psychoactive medication or unsuitability for HADS scoring. During 2008/2009, 9,501 (25.8%) CHD patients were excluded for one of these reasons. Of those eligible, 12,873 patients (47%) had a recorded HADS score for anxiety and 12,825 (47%) for depression.
Figure 14: % Eligible patients assessed for anxiety & depression, by CHCP
Of the eligible patients with valid values, 514 (4%) had HADS scores suggesting significant anxiety (>12) and 263 (2%) had equivalent scores for depression.
A substantial number of patients were reported to have declined HADS; 1,488 (12%) for anxiety and 1,475 (12%) for depression. Conversely, 41% of the patients excepted from HADS actually did have scoring performed (3,923 for anxiety and 3,937 for depression). There was some variation in these proportionsacross CHCPs, particularly in the proportions reported to have declined HADS (Figure 15).
Figure 15:HADS scores for depression in CHD patients, by CHCP
Finally, there was a strong association between residential deprivation and poor mental health, as evidenced by HADS scores over 12 (Figure 16).
Figure 16: % patients with significant anxiety & depression, by SIMD quintile
3.2: Stroke
Of the 19,952 patients on the disease register, 4,653 (30.4%) were exception coded. Results presented below relate to the remaining 15,299 patients. 7,557 (49%) were male and 7,742 (51%) female. Women were, on average, slightly older (mean age 73.4 years) than men (mean age 70.8 years). Ethnicity information was available in 12,543 (82%) of patients. Overall, 12,219 (97.2%) were white (Table 4)
Table 4: Non-exception coded stroke patients, by ethnic subgroup
Ethnic subgroup / Number (%)White / 12,219 (97.2%)
South Asian / 227 (1.8%)
Black / 10 (0.08%)
Chinese / 39 (0.3%)
Other ethnic subgroups / 48 (0.4%)
TOTAL (known ethnicity) / 12,543 (100%)
As with CHD, there were significant variations across GG&C in the ethnic composition of the stroke register(Figure 17)
Figure 17: Spatial variations in ethnic distribution of diagnosed strokepatients
The patient’s postcode of residence was known in 14,675(96%) patients, allowing estimation of small area deprivation. The vast majority of GG&C’s stroke patients live in highly deprived areas, with an inverse relationship between age and deprivation (Figure 18).
Figure 18: Mean age by SIMD deprivation quintile of diagnosed strokepatients
Cholesterol monitoring:11,093 (73%) patients had total cholesterol recorded in the past year, of whom 8,530values were considered to be valid (2-15 mmol/l). Of these, 6,323(74%) met the target cholesterol value (<5 mmol/l). There was considerable variation by CHCP in the proportion of patients monitored, ranging from 58% in SE Glasgow to 85% in SW Glasgow, however the proportion of patients who met the target cholesterol value was similar across the system (Figure 19).
Figure 19: Cholesterol monitoring & outcomes in strokepatients, by CHCP
Support with stopping smoking:12,314 (80%)patients had smoking status recorded in the past year. Overall, 3,361 (27%) of CHD patients were current smokers; however this ranged from 9% in the least deprived areas to 30% in the most deprived (Figure 20, overleaf).
Figure 20: Smoking in strokepatients, by deprivation quintile
The percentage of patients in whom smoking status was recordedranged from62% in SE Glasgow to 94% in SW Glasgow (Figure 21). The proportion of smokers recorded as wanting to stop varied from 21% in East Dunbartonshire to 41% in East Renfrewshire. The proportion referred to smoking cessation services ranged from 13% in SE Glasgow to 61% in East Renfrewshire; in four CHCPs, the number of smokers referred to smoking cessation services exceeded those wanting to stop.
Figure 21: Documented support to strokepatients who smoke, by CHCP
Recording of Body Mass Index (BMI):9,463 (62%) patients had a BMI recorded in the past year. Of these, 842 (8.9%) were excluded from the analyses as their BMI fell outside aplausible range of 10-150. Of the remainder, the mean BMI was 27.7 and only 2,734 (31.6%)of strokepatients wereof healthy weight (BMI 18.5 – 25), decreasing from 24 to 16% with increasing deprivation (Table 5).
Table5: Proportion (%) strokepatients with healthy weight
SIMD Quintile / Number (%)1 Most deprived / 1,090 (16.4%)
2 / 523 (17.6%)
3 / 302 (17.7%)
4 / 356 (20.0%)
5 Least deprived / 367 (23.6%)
TOTAL (known postcode) / 2,734 (100%
Figure 22: BMI in strokepatients, by CHCP
Assessment of mental health:During 2008/2009, 4,262 (27.9%) stroke patients were excluded from HADS assessment. Of those eligible, 4,908 patients (44.5%) had a recorded HADS score for anxiety and 4,895 (44.4%) for depression.
Figure 23: % Eligible stroke pts assessed for anxiety & depression, by CHCP
Of the eligible patients with valid values, 203 (4%) had HADS scores suggesting significant anxiety (>12) and 117 (2%) had equivalent scores for depression.
A substantial number of patients were reported to have declined HADS; 678 (13.8%) for anxiety and 670 (13.7%) for depression. Conversely, 40% of the patients excepted from HADS actually did have scoring performed (1,707 for anxiety and 1,693 for depression). There was some variation in these proportions across CHCPs, particularly in the proportions reported to have declined HADS (Figure 24).
Figure 24:HADS scores for depression in stroke patients, by CHCP
Finally, there was a strong association between residential deprivation and poor mental health. Figure 25 shows the proportion of stroke patients with HADS scores of greater than 12.
Figure 25: % patients with significant anxiety & depression, by SIMD quintile
3.3: Type 2 Diabetes
Of the 42,152 patients on the disease register, 7,647 (18.1%) were exception coded. Results presented below relate to the remaining 34,505 patients. 18,780 (54%) were male and 15,725 (46%) female. Women were, on average, slightly older (mean age 66.1 years) than men (mean age 63.2 years). Ethnicity information was available in 29,966 (87%) of patients. A higher proportion of diabetic patients (8.7%) were of non-white ethnic subgroups compared with 2.7 and 3.3 for stroke and CHD respectively (Table 6).
Table 6: Non-exception coded Type 2 diabetic patients, by ethnic subgroup
Ethnic subgroup / Number (%)White / 27,365 (91.3%)
South Asian / 1,939 (6.5%)
Black / 127 (0.4%)
Chinese / 197 (0.7%)
Other ethnic subgroups / 535 (1.2%)
TOTAL (known ethnicity) / 29,966 (100%)
In contrast to patients on the CHD and stroke registers, non-white ethnic subgroups accounted for a substantial proportion of diabetic patients in some CHCPs, accounting for up to 28% (Figure 26). In some areas, patients of South Asian ethnicity were clearly predominant, whereas others were more diverse.
Figure 26: Spatial variations in ethnic distribution of diabetic patients