MJA
MJA InSight
MJA Careers
Research

Dose adjustment for normal eating (DAFNE) — an audit of outcomes in Australia

H David McIntyre, Brigid A Knight, Dianne M Harvey, Marina N Noud, Virginia L Hagger and Kristen S Gilshenan
Med J Aust 2010; 192 (11): 637-640.
Abstract

Objective: To audit and describe the effects of participation in the Dose Adjustment for Normal Eating (DAFNE) course on clinical outcomes in people with type 1 diabetes mellitus (T1DM).

Design, setting and participants: Audit of clinical outcomes before and 1 year after DAFNE training for 145 people with T1DM who participated in courses at seven Australian diabetes centres between February 2005 and March 2007. Participants had been diagnosed with T1DM at least 1 year before and were beyond the “honeymoon phase”, with glycated haemoglobin (HbA1c) < 12% and no severe diabetes complications. They were aged over 17 years and able to understand written and spoken English.

Intervention: A 5-day structured education program covering T1DM management with an emphasis on unrestricted diet, precise carbohydrate estimation and prandial insulin dosing using insulin-to-carbohydrate ratios.

Main outcome measures: Glycaemic control (HbA1c levels), weight, severe hypoglycaemia, and quality of life scores on general (Hospital Anxiety and Depression) and diabetes-specific (Problem Areas in Diabetes) scales.

Results: Mean HbA1c fell from 8.2% to 7.8% (95% CI for change, 0.5% to 0.2%; P < 0.0001) and weight from 75.1 to 74.2 kg (95% CI for change, 1.6 to 0.2 kg; P = 0.012). Severe hypoglycaemia was less frequent after DAFNE training (P = 0.0001). Quality of life improved (P < 0.0001 for both scales).

Conclusions: One year after participation in the DAFNE program of structured education, people with T1DM showed improved glycaemic control, reduced incidence of severe hypoglycaemia, slightly reduced weight and improved quality of life. The DAFNE course offers one means of improving clinical outcomes in T1DM.

Many people with type 1 diabetes mellitus (T1DM) struggle with the everyday challenge of attempting to achieve near-normal glycaemia while avoiding hypoglycaemia. The Dose Adjustment for Normal Eating (DAFNE) program of structured education for people with T1DM,1 developed in the United Kingdom by adapting a German diabetes teaching and treatment program,2 offers one potential approach to managing this problem. Reported benefits of DAFNE training include improved overall glycaemic control, reduced hypoglycaemia and improved quality of life.1,2

DAFNE courses have been offered in Australia since February 2005. These involve a 5-day (Monday to Friday over 1 week) outpatient program of structured education in diabetes self-management for groups of six to eight people with T1DM, with an emphasis on insulin dose adjustment using a flexible dietary intake and detailed estimation of dietary carbohydrate intake. Insulin-to-carbohydrate ratios and corrective insulin doses form the basis of prandial insulin titration.

A national “OzDAFNE” collaborative coordinates training and accreditation of centres and ensures standardised, high-quality delivery of the DAFNE course in Australia. The dietary elements of DAFNE have been adapted to Australian food choices and nutritional analysis, but the core curriculum is consistent with the UK program. Collection of audit data (before and 1 year after DAFNE training) from participants is an agreed condition of membership of the OzDAFNE collaborative. This is funded through each centre’s existing resources.

This report presents outcome data for the cohort of people with T1DM who participated in DAFNE courses in Australia between February 2005 and March 2007.

Methods
Clinical data collection

Pre-specified inclusion criteria for DAFNE training included confirmation of T1DM at least 12 months previously; being in the post-“honeymoon phase”; glycated haemoglobin (HbA1c) level < 12%; age over 17 years; and ability to understand written and spoken English. To be included in this study, we required participants to have HbA1c levels recorded both before and after DAFNE training. De-identified clinical data from people undergoing DAFNE training were collected at seven specialist (secondary/tertiary) diabetes centres and community diabetes organisations and entered into a secure web-based database. The OzDAFNE database was locked on 31 March 2008.

Participants agreed to the use of their de-identified collective data for quality assurance purposes, but no formal consent form was used. The data collection and reporting process was reviewed by the Mater Health Services Human Research Ethics Committee and the study was exempted as a quality assurance activity, consistent with National Health and Medical Research Council guidelines.3

Data items included participants’ demographics, anthropometric data, history of diabetes and its complications, biochemical variables and scores from two measures of quality of life — the Hospital Anxiety and Depression Scale (HADS)4 and the Problem Areas in Diabetes (PAID) Scale5 — which have been validated in people with diabetes. The HADS is used to assess both anxiety and depressive symptoms. From a possible total score of 21 over both domains, scores < 7 are considered normal, 8–10 suggestive of anxiety or depression, and ≥ 11 indicative of a probable mood disorder. The PAID score covers a range of emotional states frequently reported in diabetes. These data were collected within 1 month of beginning and 1 year after participation in the DAFNE program. Biochemical variables were analysed using routine methods at each local DAFNE centre, without standardisation across centres.

Statistical analysis

Data items were tested for normality of distribution using the Shapiro–Wilk test and results verified by examining Q-Q plots. Normally distributed data were analysed primarily using paired t tests and are reported as mean ± SD or mean difference (95% confidence interval). Non-normally distributed data were analysed using the Wilcoxon signed rank test and are reported as median (interquartile range). Statistical comparisons were by unpaired t tests for continuous variables. Categorical results are reported as frequencies (%) and were analysed using the χ2 test or Fisher exact test. Exact P values are noted unless P is less than 0.0001. Statistica, version 8.0 (StatSoft, Tulsa, Okla, USA) was used for all analyses.

Results
Participants

Baseline data for 446 DAFNE course participants were recorded in the system. Of these, 272 had completed DAFNE training at least 1 year before the locking of the database and were eligible for inclusion. One-hundred and fifty-five people had at least some recorded data 1 year after training, and 145 (53%) had HbA1c levels recorded both before and after DAFNE training. These 145 people form the cohort for this report. The number of participants with data recorded varies for different characteristics. Baseline characteristics of the people included and excluded from this study are presented in Box 1. Only the frequency of diabetic peripheral neuropathy varied significantly between these two groups. Nearly all of our final cohort (97%) were white. Participants’ mean body mass index (BMI) was in the overweight range. Twenty-eight per cent had known complications of diabetes.

Major clinical outcomes

Mean HbA1c levels for the whole cohort fell from 8.2% to 7.8% (95% CI for change, 0.5% to 0.2%; P < 0.0001), and the change in HbA1c varied according to baseline glycaemic control, being greater for participants with HbA1c levels in the highest quartile before DAFNE training (Box 2).

Thirty-four per cent of participants (48/143) reported severe hypoglycaemic events (defined as an episode requiring the assistance of a third party) in the year before they undertook DAFNE training (Box 1), with 13 people reporting more than four events. Overall, severe hypoglycaemic events were less frequent in the 12 months after DAFNE training (P < 0.0001). The distribution of severe hypoglycaemic events was markedly skewed at baseline, with the remaining 66% of participants (95/143) reporting no events in the previous year.

The change in number of severe hypoglycaemic events per year confirms an overall improvement after DAFNE training (Box 3). Eight per cent (11/137) of participants reported an increase in frequency of severe hypoglycaemia in the year after DAFNE training. This group had a longer duration of diabetes, but did not differ in any of the other baseline characteristics. Twenty-eight percent (39/137) reported a decrease in frequency of severe hypoglycaemia. Eighty-three of 87 people who reported no change in hypoglycaemia frequency had not experienced severe hypoglycaemia in the year before DAFNE training.

Participants’ mean weight had fallen by 0.9 kg 12 months after DAFNE participation (95% CI for change, 1.6 to 0.2 kg; P = 0.012) (Box 4).

Assessment using the PAID and HADS questionnaires showed that participants’ quality of life had improved 1 year after DAFNE training. The median PAID score before DAFNE training was 25, typical for an unselected outpatient population.5 This decreased to 16.25 (consistent with improved quality of life) 1 year after DAFNE training, corresponding to a small-to-moderate effect size (Box 4). Before training, participants’ PAID scores were highest for questions relating to hypoglycaemia, the future risk of complications, and guilt and anxiety for getting “off track” with diabetes management. The greatest reductions after training were seen in these areas and in concerns regarding food and eating.

The median scores for both HADS questionnaires showed small to moderate reductions after DAFNE training (Box 4). At the 12-month follow-up, a greater proportion of participants showed scores in the normal and mid ranges, with fewer having scores that are likely to indicate a mood disorder.

Discussion

Our audit of clinical outcomes suggests that the DAFNE program benefits people with T1DM in Australia. The results are broadly consistent with those reported in the UK DAFNE randomised controlled trial1 and in a large-scale German audit series.6 The UK trial reported a larger (about 1%) mean reduction in HbA1c levels, but it had recruited patients with poor glycaemic control and a mean HbA1c level of 9.3% at baseline. Consistent with our findings, the German audit described differences in HbA1c reduction according to baseline HbA1c quartile.

The landmark Diabetes Control and Complications Trial (DCCT)7 offered the first clear evidence that improved glycaemic control in T1DM improves clinical (in particular, microvascular) outcomes, but both the DCCT8 and a subsequent meta-analysis9 suggested that HbA1c reduction must inevitably be accompanied by an increase in severe hypoglycaemia. This, in addition to the weight gain associated with intensive therapy,7 has served to discourage many people with diabetes, and their treating doctors, from attempting more intensive glycaemic control.

However, both our data and the UK and German reports1,6 demonstrate that it is possible to lower HbA1c levels without these unwanted effects using the DAFNE structured education program. We suggest that these benefits may derive from the very detailed 5-day program of patient education that comprises the DAFNE course and from the less stringent glycaemic targets used in DAFNE. While the DCCT program was largely doctor-, dietitian- and nurse-driven, with prescription of dietary regimens, and insulin doses adjusted by clinicians, DAFNE promotes self-management with dietary flexibility and self-titration of insulin doses.

Type 1 diabetes is a profoundly vexatious chronic disease for many people, as it involves complex juggling of insulin doses according to food intake, glucose concentrations, exercise and other, often apparently random, factors on an unremitting daily basis. This clearly can impair quality of life. Although quality of life was not markedly abnormal in our cohort at baseline, all measures did improve after participation in the DAFNE course. The reasons for this may include reduced hypoglycaemia,10 improved self-management skills and confidence, and building a relationship with the health professional team, as well as the supportive environment and sharing of experiences of the “diabetic life” which, anecdotally, are always part of the DAFNE group dynamic.

Some limitations of our data should be noted. All data items were collected in a routine clinical environment, without standardised assays, dedicated staff or uniform data collection systems.

People with T1DM participating in DAFNE may not be representative of the population of people with T1DM as a whole. However, the mean baseline HbA1c level and BMI for our cohort were very similar to reported values for the cohort of T1DM patients who participated in the 2006 Australian National Diabetes Information Audit and Benchmarking survey of patients attending specialist diabetes clinics in Australia (HbA1c, 8.1% ± 1.6%; BMI, 26.2 ± 4.5 kg/m2).11 The distribution of HADS scores in our cohort at baseline was also similar to that reported in a UK cohort of adults with T1DM.4 The reasons why some participants did not attend for data collection 1 year after DAFNE training are not known, but these participants do not appear to differ in terms of baseline clinical characteristics from those who did attend. This audit reports outcomes on a pre- and post-intervention basis, with no control group to assess changes over time without the intervention. However, such data are available from the UK DAFNE study1 and our results are consistent with other published reports.2,6

Structured patient education has been endorsed as a routine part of diabetes management by the National Institute for Clinical Excellence in the UK,12 with DAFNE being recognised as one program suitable for people with T1DM. Cost modelling in the UK has suggested that DAFNE is cost-saving, due to reduced diabetic complications, rather than just cost-effective.13 Other structured education courses, such as the Empowerment program developed and conducted in Newcastle, Australia,14 and flexible insulin therapy teaching courses in Basel, Switzerland,15 have also shown benefits, including reduced hypoglycaemia and improved quality of life, although reductions in HbA1c levels have not always been shown, perhaps due to differing participant populations.

Despite the positive outcomes of DAFNE in Australia, funding for this type of intensive, structured education program remains difficult to secure. Limited support for group education is available to people with type 2 diabetes.16 However, the current Medicare Benefits Schedule rebate of $16.00 per group service (Item 81105) is clearly inadequate to fund an intensive education program, and T1DM is excluded. Many OzDAFNE centres have been charging no or minimal fees for the DAFNE course. This allows some people with T1DM to access the program, but limits the number of courses that centres are able to provide.

In summary, our audit of people with T1DM undergoing the DAFNE course demonstrates clinical benefits similar to those reported in other health care settings. DAFNE provides one potential means of improving glycaemic control and other important health outcomes in people with T1DM. We believe that OzDAFNE merits consideration for more widespread availability, predicated on more systematic funding.

1 Comparison of baseline characteristics of participants in the DAFNE program who were included and excluded from this study

With follow-up HbA1c level* (n = 145)


Without follow-up HbA1c level (n = 127)


Characteristic

No. with data recorded

Mean (SD)

No. with data recorded

Mean (SD)

P


Age (years)

133

43.3 (14.3)

125

41.2 (13.5)

0.22

Duration of diabetes (years)

134

16.7 (11.5)

127

15.6 (10.3)

0.39

HbA1c (%)

145

8.2% (1.2%)

121

8.3% (1.3%)

0.66

Weight (kg)

142

75.1 (13.8)

124

74.4 (13.7)

0.75

BMI (kg/m2)

141

26.5 (4.7)

122

25.9 (3.9)

0.23

No. (%)


No. (%)


Female

145

95 (66%)

127

86 (68%)

0.62

Known diabetes complications

137

38 (28%)

121

29 (24%)

0.49

Diabetic retinopathy

115

33 (29%)

104

28 (27%)

0.77

Diabetic neuropathy

113

11 (10%)

102

3 (3%)

0.04

Microalbuminuria

113

13 (12%)

125

36 (29%)

0.09

Severe hypoglycaemia in previous year

143

48 (34%)

125

36 (29%)

0.26

Ketoacidosis in previous year

145

8 (6%)

124

8 (6%)

0.31


BMI = body mass index. DAFNE = Dose Adjustment for Normal Eating. HbA1c = glycated haemoglobin.
* Participants for whom HbA1c levels were recorded before and 1 year after DAFNE training (included in this study). † Participants for whom HbA1c levels were only available before DAFNE training (excluded from this study). ‡ Defined as an episode requiring the assistance of a third party.

2 Change in participants’ HbA1c levels 1 year after DAFNE training*


DAFNE = dose adjustment for normal eating. HbA1c = glycated haemoglobin. * Absolute change in mean HbA1c (95% CI) (HbA1c post DAFNE HbA1c pre DAFNE), categorised by baseline HbA1c quartile group. Quartile ranges: Q1, 6.2%–7.4%; Q2, 7.5%–8.1%; Q3, 8.2%–9.0%; Q4, 9.1%–11.4%).

3 Change in self-reported frequency* of severe hypoglycaemia among DAFNE participants (n = 137)


DAFNE = dose adjustment for normal eating. * Comparing the year following DAFNE training with the year before DAFNE training. † Defined as an episode requiring the assistance of a third party.

4 Major clinical outcomes for participants 1 year after DAFNE training (n = 145)

Before DAFNE training


1 year after DAFNE training


Normally distributed variables

Mean (SD)

Mean (SD)

Mean difference (95% CI)*

P


HbA1c (%)

8.2% (1.2%)

7.8% (1.3%)

0.4% ( 0.5% to 0.2%)

< 0.0001

Weight (kg)

75.1 (13.8)

74.2 (13.4)

0.9  ( 1.6 to 0.2)

0.012

Non-normally distributed variables

Median (IQR)

Median (IQR)

P


Severe hypoglycaemic episodes previous year

0 (0–1)

0 (0–0)

0.0001

Self-glucose tests/week

28 (20–35)

30 (25–38)

0.013

HADS — anxiety score

5 (3–9)

4 (2–6)

< 0.0001

HADS — depression score

5 (2–8)

3 (1–6)

0.0003

PAID — total score

25 (15–45)

16.25 (10–30)

< 0.0001


DAFNE = dose adjustment for normal eating. HADS = Hospital Anxiety and Depression Scale.4 HbA1c = glycated haemoglobin. IQR = interquartile range. PAID = Problem Areas in Diabetes Scale.5
* Mean and 95% CIs shown for change in HbA1c and weight (post DAFNE pre DAFNE). † P values are for comparison of pre- versus post-DAFNE results.

Received 
28 May 2009
accepted 
3 Feb 2010
H David McIntyre, MB BS, FRACP, Head of Mater Clinical School and Director1
Brigid A Knight, BSc, GradDipNutrDiet, Dietitian and Diabetes Educator2
Dianne M Harvey, BSc, GradDipDietetics, Dietitian and OzDAFNE Coordinator3
Marina N Noud, MNurs, DipEd, CDE, Clinical Nurse Consultant, Diabetes and Endocrinology2
Virginia L Hagger, MPH, RN-CDE, GradDipVet, Health and Education Services Manager3
Kristen S Gilshenan, BMaths(Hons), BInfoTech, Statistician, Mater Research Support Centre2
1 Endocrinology and Obstetric Medicine, University of Queensland and Mater Health Services, Brisbane, QLD.
2 Mater Health Services, Brisbane, QLD.
3 Diabetes Australia, Victoria, Melbourne, VIC.
Competing Interests: 

David McIntyre has received speaker fees and travel assistance from companies involved in provision of insulin/delivery systems for type 1 diabetes care, including Novo Nordisk, Eli Lilly, sanofi-aventis and Medtronic. He is a previous President of the DAFNE Association of Australia Inc (now delisted) and Director of a diabetes centre involved in the provision of DAFNE courses, but has derived no personal profit from this.

Reference Text: 
DAFNE Study Group. Training in flexible, intensive insulin management to enable dietary freedom in people with type 1 diabetes: dose adjustment for normal eating (DAFNE) randomised controlled trial. BMJ 2002; 325: 746.
Reference Order: 
1
PubMed ID: 
12364302
Reference Text: 
Mühlhauser I, Bruckner I, Berger M, et al. Evaluation of an intensified insulin treatment and teaching programme as routine management of type 1 (insulin-dependent) diabetes. The Bucharest-Düsseldorf study. Diabetologia 1987; 30: 681-690.
Reference Order: 
2
PubMed ID: 
3123298
Reference Text: 
National Health and Medical Research Council. When does quality assurance in health care require independent ethical review? Canberra: NHMRC, 2003.
Reference Order: 
3
PubMed ID: 
Reference Text: 
Shaban MC, Fosbury J, Kerr D, Cavan DA. The prevalence of depression and anxiety in adults with Type 1 diabetes. Diabet Med 2006; 23: 1381-1384.
Reference Order: 
4
PubMed ID: 
17116192
Reference Text: 
Welch G, Weinger K, Anderson B, Polonsky WH. Responsiveness of the Problem Areas In Diabetes (PAID) questionnaire. Diabet Med 2003; 20: 69-72.
Reference Order: 
5
PubMed ID: 
12519323
Reference Text: 
Sämann A, Mühlhauser I, Bender R, et al. Glycaemic control and severe hypoglycaemia following training in flexible, intensive insulin therapy to enable dietary freedom in people with type 1 diabetes: a prospective implementation study. Diabetologia 2005; 48: 1965-70.
Reference Order: 
6
PubMed ID: 
16132954
Reference Text: 
Diabetes Control and Complications Trial Research Group. The effect of intensive treatment of diabetes on the development and progression of long-term complications in insulin-dependent diabetes mellitus. N Engl J Med 1993; 329: 977-986.
Reference Order: 
7
PubMed ID: 
8366922
Reference Text: 
Diabetes Control and Complications Trial Research Group. Hypoglycemia in the Diabetes Control and Complications Trial. Diabetes 1997; 46: 271-286.
Reference Order: 
8
PubMed ID: 
9000705
Reference Text: 
Egger M, Davey Smith G, Stettler C, Diem P. Risk of adverse effects of intensified treatment in insulin-dependent diabetes mellitus: a meta-analysis. Diabet Med 1997; 14: 919-928.
Reference Order: 
9
PubMed ID: 
9400915
Reference Text: 
Davis RE, Morrissey M, Peters JR, et al. Impact of hypoglycaemia on quality of life and productivity in type 1 and type 2 diabetes. Curr Med Res Opin 2005; 21: 1477-1483.
Reference Order: 
10
PubMed ID: 
16197667
Reference Text: 
National Association of Diabetes Centres. ANDIAB 2006 — Australian national diabetes information audit and benchmarking. Final report. Canberra: Australian Government Department of Health and Ageing, 2007.
Reference Order: 
11
PubMed ID: 
Reference Text: 
National Institute for Clinical Excellence. Guidance on the use of patient-education models for diabetes. (Technology appraisal guidance 60.) London: NICE, 2003.
Reference Order: 
12
PubMed ID: 
Reference Text: 
Shearer A, Bagust A, Sanderson D, et al. Cost-effectiveness of flexible intensive insulin management to enable dietary freedom in people with Type 1 diabetes in the UK. Diabet Med 2004; 21: 460-467.
Reference Order: 
13
PubMed ID: 
15089791
Reference Text: 
Lowe J, Linjawi S, Mensch M, et al. Flexible eating and flexible insulin dosing in patients with diabetes: results of an intensive self-management course. Diabetes Res Clin Pract 2008; 80: 439-443.
Reference Order: 
14
PubMed ID: 
18353485
Reference Text: 
Bendik CF, Keller U, Moriconi N, et al. Training in flexible intensive insulin therapy improves quality of life, decreases the risk of hypoglycaemia and ameliorates poor metabolic control in patients with type 1 diabetes. Diabetes Res Clin Pract 2009; 83: 327-333.
Reference Order: 
15
PubMed ID: 
19128852
Reference Text: 
Australian Government Department of Health and Ageing. Medicare Benefits Schedule. Allied Health Services. 1 May 2010. Part 4. Allied health group services for patients with type 2 diabetes. http://www.health.gov.au/internet/mbsonline/publishing.nsf/Content/FD65645DA9682F63CA2576F60002194F/$File/201005-Allied.pdf (accessed 20 Apr 2010).
Reference Order: 
16
PubMed ID: