Advertisement

Diabetologia

, Volume 55, Issue 9, pp 2356–2360 | Cite as

The effect of deprivation and HbA1c on admission to hospital for diabetic ketoacidosis in type 1 diabetes

  • L. GovanEmail author
  • E. Maietti
  • B. Torsney
  • O. Wu
  • A. Briggs
  • H. M. Colhoun
  • C. M. Fischbacher
  • G. P. Leese
  • J. A. McKnight
  • A. D. Morris
  • N. Sattar
  • S. H. Wild
  • R. S. Lindsay
  • on behalf of the Scottish Diabetes Research Network (SDRN) Epidemiology Group*
Short Communication

Abstract

Aims/hypothesis

Diabetic ketoacidosis is a potentially life-threatening complication of diabetes and has a strong relationship with HbA1c. We examined how socioeconomic group affects the likelihood of admission to hospital for diabetic ketoacidosis.

Methods

The Scottish Care Information – Diabetes Collaboration (SCI-DC), a dynamic national register of all cases of diagnosed diabetes in Scotland, was linked to national data on hospital admissions. We identified 24,750 people with type 1 diabetes between January 2005 and December 2007. We assessed the relationship between HbA1c and quintiles of deprivation with hospital admissions for diabetic ketoacidosis in people with type 1 diabetes adjusting for patient characteristics.

Results

We identified 23,479 people with type 1 diabetes who had complete recording of covariates. Deprivation had a substantial effect on odds of admission to hospital for diabetic ketoacidosis (OR 4.51, 95% CI 3.73, 5.46 in the most deprived quintile compared with the least deprived). This effect persisted after the inclusion of HbA1c and other risk factors (OR 2.81, 95% CI 2.32, 3.39). Men had a reduced risk of admission to hospital for diabetic ketoacidosis (OR 0.71, 95% CI 0.63, 0.79) and those with a history of smoking had increased odds of admission to hospital for diabetic ketoacidosis by a factor of 1.55 (95% CI 1.36, 1.78).

Conclusions/interpretation

Women, smokers, those with high HbA1c and those living in more deprived areas have an increased risk of admission to hospital for diabetic ketoacidosis. The effect of deprivation was present even after inclusion of other risk factors. This work highlights that those in poorer areas of the community with high HbA1c represent a group who might be usefully supported to try to reduce hospital admissions.

Keywords

Deprivation Diabetes Diabetic ketoacidosis HbA1c Record linkage 

Abbreviations

ICD

International Classification of Diseases

ISD

Information Services Division

SCI-DC

Scottish Care Information – Diabetes Collaboration

SDRN

Scottish Diabetes Research Network

SIMD

Scottish Index of Multiple Deprivation

SMR01

Scottish Morbidity Records

Introduction

We have previously demonstrated that, among people with type 1 diabetes, HbA1c is an important indicator of risk of admission to hospital, independent of other predictors, with a particularly strong relationship to admissions coded as diabetic ketoacidosis [1]. Deprivation is a further important potential indicator of disease risk in the population, although previous analysis of data from a smaller Scottish population suggests that deprivation has a relatively weak relationship with HbA1c in people with type 1 diabetes [2].

This study examines how socioeconomic group affects likelihood of admission to hospital for diabetic ketoacidosis among people with type 1 diabetes, and whether this relationship may be explained by HbA1c.

Methods

Data

Information on diabetes was obtained from the Scottish Care Information – Diabetes Collaboration (SCI-DC) database – a dynamic national clinical information system of all diagnosed cases of diabetes in Scotland (www.diabetesinscotland.org.uk). This database contains records for over 99% of diagnosed cases of diabetes [3] with detailed clinical information including BMI, creatinine, age, sex and HbA1c. For the current analysis, HbA1c was measured using a variety of clinical methods, all of which were DCCT aligned. Type 1 diabetes was identified using an algorithm that incorporated age, drug prescription and clinical description of the type of diabetes. Those whose type diagnosis was not known were excluded from the analysis. Deprivation was defined using the Scottish Index of Multiple Deprivation (SIMD) [4], which provides a composite index of relative area deprivation across Scotland. Smoking status was defined as whether or not a person was a current or ex-smoker.

Information on hospital admissions was obtained using Scottish Morbidity Records (SMR01), national data on hospital admissions from the Information Services Division (ISD) of NHS National Services, Scotland. The SMR01 contain information on over 95% of Scotland’s hospital admissions and include administrative data and demographic information such as age, sex and postcode of the patient. Individual episodes of care are recorded within each admission entry with up to six International Classification of Diseases (ICD) diagnosis codes (1997–present: ICD-10 [www.who.int/classifications/icd/en/]).

Data were linked within the ISD using probabilistic methods based on name, sex, date of birth and postcode, as previously described [5]. No personal identifiers were released to researchers and all subsequent analyses were conducted on anonymised datasets.

Statistical analysis

The outcome of interest was recording of diabetic ketoacidosis as a reason for admission to hospital at least 180 days after the diagnosis of diabetes. Outcome was determined if the ICD codes for diabetic ketoacidosis were present in any position of diagnosis. Parametric relationships between mean HbA1c (as measured between 2005 and 2007) and admission to hospital for diabetic ketoacidosis were investigated using the fractional polynomial methods in Stata, version 11 (Stata, College Station, TX, USA). Natural log of mean HbA1c was found to provide the best fit to the data. Logistic regression models were used to estimate the association between admissions to hospital for diabetic ketoacidosis, log mean HbA1c, deprivation quintiles (referent quintile 1, least deprived) and history of smoking. Analysis was adjusted for potential confounding factors including age, sex, previous vascular disease (ICD-9 [www.icd9data.com/2007/Volume1/240-279/250-259/250/default.htm]: 410–414, 430–438, 443; ICD-10: I20–I25, I60–I69, I73), creatinine, BMI and diabetes duration.

Results

Between January 2005 and December 2007, we identified 24,750 people with type 1 diabetes (Scottish population 5.1 million) of whom 23,479 had complete recording of all covariates; 64% were either a current or ex-smoker. There were a total of 4,577 admissions to hospital coded for diabetic ketoacidosis during the study period (79% of these admissions were single admissions for different people and 21% were multiple admissions). The number of admissions per 1,000 persons per year was significantly higher in the most deprived fifth of socioeconomic groups (175 admissions per 1,000 persons per year compared with 60 admissions per 1,000 persons per year in the least deprived fifth). Figure 1 shows that some of this effect may be explained by HbA1c, but the effect of deprivation is evident in all thirds of HbA1c (range of HbA1c in: tertile 1 = 4.4–8.15% [24.6–65.6 mmol/mol]; tertile 2 = 8.16–9.26% [65.7–77.7 mmol/mol]; tertile 3 = 9.27–18.3% [77.8–176.5 mmol/mol]). Compared with those who did not have an admission mentioning diabetic ketoacidosis, the HbA1c values in those with an admission mentioning diabetic ketoacidosis were greater for all socioeconomic groups
Fig. 1

Number of admissions to hospital for diabetic ketoacidosis per 1,000 persons per year in groups defined by tertiles of HbA1c (range of HbA1c in tertile 1 = 4.4–8.15% [24.6–65.6 mmol/mol]; tertile 2 = 8.16–9.26% [65.7–77.7 mmol/mol]; tertile 3 = 9.27–18.3% [77.8–176.5 mmol/mol]) and quintiles of deprivation

Deprivation was an independent predictor of admission to hospital with an increase in odds of admission of 4.51 (95% CI 3.73, 5.46) in the most deprived fifth compared with the least deprived. After adjustment for HbA1c (expressed as loge[HbA1c in %]/0.09531, where one unit increase is equivalent to 10% increase in HbA1c), the increase in odds of admission in all deprivation quintiles was reduced (for most deprived compared with least deprived: OR 3.41; 95% CI 2.54, 3.71). After adjusting for the other covariates, younger patients, women (not including maternity admissions) and people with a history of vascular disease had a higher risk of admission (Table 1). History of smoking also had a substantial effect, increasing the odds of admission by 1.55 (95% CI 1.36, 1.78). The size of the effect of deprivation was reduced by the inclusion of other covariates; however, there remained a substantial increase in odds of admission in the most deprived fifth compared with the least deprived fifth (OR 2.82, 95% CI 2.33, 3.42).
Table 1

Coefficients, ORs and 95% CIs obtained from the logistic regression model for admissions to hospital for diabetic ketoacidosis

Risk factor

Coefficient

OR (95% CI)

Sex, male

−0.348

0.71 (0.63, 0.79)

Age (years)a

−0.033

0.97 (0.96, 0.97)

Previous admission for vascular reasons

0.961

2.61 (1.96, 3.49)

Creatinine (μmol/l)a, b

0.003

1.003 (1.002, 1.004)

BMI (kg/m2)a, b

−0.059

0.94 (0.93, 0.95)

Diabetes duration (years)a, b

−0.015

0.98 (0.98, 0.99)

Smokers

0.439

1.55 (1.36, 1.78)

Loge(HbA1c)/0.09531a, b

0.458

1.58 (1.53, 1.63)

Deprivationc

  

  Quintile 1 – least deprived

Reference

Reference

  Quintile 2

0.266

1.31 (1.05, 1.62)

  Quintile 3

0.515

1.67 (1.38, 2.03)

  Quintile 4

0.691

2.00 (1.64, 2.42)

  Quintile 5 – most deprived

1.038

2.82 (2.33, 3.42)

aPer one unit increase. For loge(HbA1c)/0.09531, equivalent to 10% increase in HbA1c (expressed as %)

bMean of any values recorded between January 2005/date of diagnosis and December 2007/date of death

cMultivariate analysis results for model including all factors in the table

Discussion

There was a strong association between deprivation and odds of admission to hospital for diabetic ketoacidosis, with people in more deprived areas having an odds of admission 4.5 times higher than those in the least deprived areas. Higher HbA1c also increased the odds of admission to hospital. Although the inclusion of HbA1c and other risk factors reduced the size of the effect of deprivation, a substantial effect remained. This suggests that only some of the effect of deprivation can be explained by these other risk factors in people with type 1 diabetes. Those who had a history of smoking, previous vascular admissions and women also had an increased risk of admission associated with diabetics ketoacidosis.

We have taken advantage of the linked data for both hospital admissions (SMR01) and clinical information (SCI-DC), providing almost 100% coverage of all data for people with diagnosed type 1 diabetes in Scotland between 2005 and 2007. This allowed us to avoid under-reporting in hospital discharge information as found in other studies [6, 7]. Some limitations of our study have also been identified. First, an algorithm that incorporated age, drug prescription and clinical description of the type of diabetes was used to determine the type of diabetes. As with nearly all cohorts of this size and type, there is potential for misclassification of diabetes type. However, we believe those misclassified will be an extremely small percentage. Second, we used the annual average of HbA1c, chosen as the best measure of prevailing HbA1c for each person. This yearly average does not capture the variability in HbA1c and a measurement taken before admission for diabetic ketoacidosis could be argued to be more appropriate. However, HbA1c was measured locally in diabetes clinics during routine clinic visits. Using the last measurement of HbA1c instead of the average did not change the conclusions of our study. In addition, the number of recordings of HbA1c per person was not significantly associated with odds of admission to hospital for diabetic ketoacidosis. Third, in determining the reason for an admission to hospital we relied on hospital admission data, which are dependent on the accuracy (around 90%) of coding of ICD in hospital diagnosis codes [8]. However, hyperglycaemia codes are often underused in discharge coding [6]. The codes used do not include people admitted with coma (ICD10 E10.0, E14.0) as these categories include, but do not distinguish between, those with ketoacidotic, hyperosmolar or hypoglycaemic coma. We also did not include those with multiple complications potentially including ketoacidosis (ICD10 E10.7, E14.7) since ketoacidosis could not be separately identified within these codes.

Our results show that people in more deprived areas are at a greater risk of admission to hospital for diabetics ketoacidosis, and it is important to discuss why this effect exists. We can speculate that those in more deprived areas have poorer control of their diabetes resulting in increased risk of hospital admission for diabetics ketoacidosis. As discussed, however, deprivation score is not a strong predictor of HbA1c in our population, and much of the effect of deprivation score appears independent of HbA1c.

Despite universal health coverage that is free at point of care, social deprivation is also associated with reduced engagement with health services, such as retinal screening [9], and deprivation is in turn associated with higher risk of complications [10]. Admission to hospital for diabetic ketoacidosis will also relate to other comorbidities, social and medical support and relevant education in diabetes management during intercurrent illness and other crises. In that context, the relationship between smoking and admission to hospital for diabetic ketoacidosis is of interest. This relationship is unlikely to be causal but smoking may act as a marker of other health behaviours and thus increased risk of admission to hospital for diabetic ketoacidosis.

There is an increasing emphasis on structured education programmes as a means of improving health outcomes in type 1 diabetes. Our analysis shows that socioeconomic group carries a risk of admission to hospital for diabetic ketoacidosis over and above that predicted by HbA1c. An important implication of our work is that where programmes are directed towards prevention of expensive and distressing complications, such as diabetic ketoacidosis, consideration of their relevance to those patients with diabetes who are most at risk (including consideration of socioeconomic group) is vital.

Notes

Acknowledgements

These data were available for analysis by members of the SDRN thanks to the hard work and dedication of NHS staff across Scotland who enter the data and also the people and organisations (the SCI-DC Steering Group, the Scottish Diabetes Group, the Scottish Diabetes Survey Group, the managed clinical network managers and staff in each Health Board) involved in setting up, maintaining and overseeing the SCI-DC. The SDRN receives core support from the Chief Scientist’s Office at the Scottish Government Health Department.

Funding

The costs of data linkage were covered by the Scottish Government Health Department. This work was funded by the Wellcome Trust through the Scottish Health Informatics Programme (SHIP) Grant (Ref. WT086113). SHIP is a collaboration between the Universities of Aberdeen, Dundee, Edinburgh, Glasgow and St Andrews and the ISD of NHS Scotland.

Duality of interest

H. Colhoun has financial relationships with Pfizer, Roche, Astra Zeneca and Eli Lilly. The authors declare that there is no duality of interest associated with this manuscript.

Contribution statement

All authors contributed to the data provision and/or quality. LG and EM carried out the data analysis. LG, EM, BT, RL, OW and AB contributed to study design and analysis. CF, GL, JM, AM, NS, SW contributed to the interpretation of results. LG wrote the initial draft of the manuscript. All authors reviewed and edited the manuscript and approved the final version.

References

  1. 1.
    Govan L, Wu O, Briggs A et al (2011) Achieved levels of HbA1c and likelihood of hospital admission in people with type 1 diabetes in the Scottish population. Diabetes Care 34:1992–1997PubMedCrossRefGoogle Scholar
  2. 2.
    Wild SH, McKnight JA, McConnachie A, Lindsay RS (2010) Socioeconomic status and diabetes-related hospital admissions: a cross-sectional study of people with diagnosed diabetes. J Epidemiol Commun Health 64:1022–1024CrossRefGoogle Scholar
  3. 3.
    Colhoun H, SDRN Epidemiology Group (2009) Use of insulin glargine and cancer incidence in Scotland: a study from the Scottish Diabetes Research Network Epidemiology Group. Diabetologia 52:1755–1765PubMedCrossRefGoogle Scholar
  4. 4.
    Scottish Government (2009) Scottish Index of Multiple Deprivation 2009: general report. www.scotland.gov.uk/Publications/2009/10/28104046/0, accessed October 2011
  5. 5.
    Wild S, MacLeod F, McKnight J et al (2008) Impact of deprivation on cardiovascular risk factors in people with diabetes: an observational study. Diabet Med 25:194–199PubMedCrossRefGoogle Scholar
  6. 6.
    Anwar H, Fischbacher CM, Leese GP et al (2011) Assessment of the under-reporting of diabetes in hospital admission data: a study from the Scottish Diabetes Research Network Epidemiology Group. Diabet Med 28:1514–1519PubMedCrossRefGoogle Scholar
  7. 7.
    Leslie PJ, Patrick AW, Hepburn DA, Scougal IJ, Frier BM (1992) Hospital in-patient statistics underestimate the morbidity associated with diabetes mellitus. Diabet Med 9:379–385PubMedCrossRefGoogle Scholar
  8. 8.
    ISD Scotland. SMR01 and associated data quality assurance 2004–2006. www.isdscotlandarchive.scot.nhs.uk/isd/2737.html. Accessed 21 May 2012 (archived)
  9. 9.
    Leese GP, Boyle P, Feng Z, Emslie-Smith A, Ellis JD (2008) Screening uptake in a well-established diabetic retinopathy screening program. Diabetes Care 31:2131–2135PubMedCrossRefGoogle Scholar
  10. 10.
    Bachmann MO, Eachus J, Hopper CD et al (2003) Socio-economic inequalities in diabetes complications, control, attitudes and health service use: a cross-sectional study. Diabet Med 20:921–929PubMedCrossRefGoogle Scholar

Copyright information

© Springer-Verlag 2012

Authors and Affiliations

  • L. Govan
    • 1
    Email author
  • E. Maietti
    • 2
  • B. Torsney
    • 2
  • O. Wu
    • 1
  • A. Briggs
    • 1
  • H. M. Colhoun
    • 4
  • C. M. Fischbacher
    • 5
  • G. P. Leese
    • 6
  • J. A. McKnight
    • 7
  • A. D. Morris
    • 4
  • N. Sattar
    • 3
  • S. H. Wild
    • 8
  • R. S. Lindsay
    • 3
  • on behalf of the Scottish Diabetes Research Network (SDRN) Epidemiology Group*
  1. 1.Health Economics and Health Technology Assessment, Institute of Health and WellbeingUniversity of GlasgowGlasgowUK
  2. 2.School of Mathematics and StatisticsUniversity of GlasgowGlasgowUK
  3. 3.Institute of Cardiovascular and Medical SciencesUniversity of GlasgowGlasgowUK
  4. 4.Medical Research InstituteUniversity of DundeeDundeeUK
  5. 5.Information Services DivisionNHS National Services ScotlandEdinburghUK
  6. 6.Ninewells Hospital and Medical SchoolUniversity of DundeeDundeeUK
  7. 7.Western General HospitalUniversity of EdinburghEdinburghUK
  8. 8.Centre for Population Health SciencesUniversity of EdinburghEdinburghUK

Personalised recommendations