General Comments

In 2010 – ANDIAB2 collected and benchmarked data that were more Education Centre and Patient Self-care focused than the more ‘Medically’ focused ANDIAB. Twenty seven NADC member Diabetes Centres participated. Data were provided for assessment on 2131 individuals [with similar demographic findings to those in ANDIAB 2009 [4.2 & Table 2 In Findings / Results > Demographic Data] – although there were some differences with [for instance] one quarter initial visit patients (25.6% versus 13.2% ANDIAB 2009) and 3.7% (versus 2.1%) having GDM.

Missing Data

There were minimal missing data [4.10 Page 17 & Table 15 Page 18]. In ANDIAB2 53.7% of the data items were less than 5% missing. Whilst this is not as impressive as was ANDIAB2 2005 [85.5%], it is still far superior to ANDIAB 2009, where only 20.0% of the data items were less than 20% missing. Whilst much of the ANDIAB 2009 missing data relate to Fields such as Eye Data, which may not be readily available to the Clinician, there would appear no doubt that the ANDIAB2 completeness of data collection is a testament to the diligence of those who participated [including the individuals themselves in completing the EQ-5D and DDS components]. Sites were given an opportunity to supply any missing data and to validate questionable data. Table 15[a] In Findings / Results > Missing Data shows the Missing ‘Vital’ Data items obtained by requesting their provision from sites – with quite substantial improvements noted.

Limitations of ANDIAB2

High numbers of Initial Visit Individuals

There were a quarter on the patients [25.6%] for whom this was an initial visit and the possibility exists that the reduced prevalence findings for many of the items assessed may in some part be related to no [or minimal] previous diabetes education [or possibly no recent educational contact for individuals with longer diabetes duration] – even though overall 67.7% of individuals reported having seen a Diabetes Educator within the last 12 months. To assess this we looked at duration of diabetes, initial visit data and whether individuals had seen a Diabetes Educator in the last 12 months - Table 17[a] {complete data available for 2051 of the 2131 individuals [96.2%]}.Overall two thirds or more of the individuals had seen an Educator [DE] in the last 12 months irrespective of duration [decreasing slightly with increasing duration]. As regards the recently diagnosed, where analysis shows this represented 17.0% of individuals, that for 46.8% of them this was an initial visit, and that only half [50.3%] had already seen an Educator. However this represents only 4.0% of the total number of individuals who may have had no or minimal education. Of those with longer diabetes duration – half or more initial visit individuals {48.4% only for 2-5 years duration} and 71.3 to 77.7% of non-initial visit individuals - had seen an Educator in the last 12 months. It is thus more likely that the reduced prevalence findings for the self-care items assessed, are true ‘deficiencies’ in self-care and less likely due to no or minimal education, or to education undertaken some time ago.

Table 17[a] 2010 Duration - Initial Visit - Seen by Educator last 12 Months

Duration
2010 n
2010%
2010 DE last 12 mth
2010 Initial
Visit = Yes
2010 DE last
12mth
2010 Initial
Visit = No
2010 DE last
12mths
<1
348
16.9%
71.3%
46.8%
50.3%
53.2%
89.7%
1-2
84
4.1%
72.6%
27.4%
65.2%
72.6%
75.4%
2-5
212
10.3%
68.9%
30.2%
48.4%
69.8%
77.7%
5-10
411
20.0%
69.3%
23.6%
58.8%
76.4%
72.6%
10+
1001
48.7%
69.2%
16.5%
58.8%
83.5%
71.3%
The same analysis in ANDIAB2 2005 is presented in Table 17[b], with very similar findings.

Table 17[b] 2005 Duration - Initial Visit - Seen by Educator last 12 Months

Duration
2005 n
2005%
2005 DE last 12 mth
2005 Initial
Visit = Yes
2005 DE last
12mth
2005 Initial
Visit = No
2005 DE last
12mths
<1
288
28.7%
78.1%
41.0%
51.6%
59.0%
96.5%
1-2
64
4.7%
70.3%
20.3%
69.2%
79.7%
70.6%
2-5
162
12.0%
66.7%
19.1%
41.9%
80.9%
72.5%
5-10
281
20.8%
67.3%
16.4%
54.3%
83.6%
69.8%
10+
458
33.9%
64.6%
16.2%
50.0%
83.8%
67.4%
A similar analysis to that in Table 17[a] is presented in Table 17[c], showing data re ‘seen by Dietitian in the last 12 months’. The patterns are similar across duration categories, although fewer individuals had seen a Dietitian compared with a Diabetes Educator.

Table 17[c] 2010 Duration - Initial Visit - Seen by Dietitian last 12 Months

Duration
2010 n
2010%
2010 Dietitian
last 12mth
2010 Initial
Visit = Yes
2010 Dietitian
last 12mth
2010 Initial
Visit = No
2010 Dietitian
last 12mths
<1
348
17.0%
62.6%
47.1%
42.7%
52.9%
80.4%
1-2
84
4.1%
59.5%
27.4%
47.8%
72.6%
63.9%
2-5
210
10.3%
46.2%
30.0%
38.1%
70.0%
49.7%
5-10
409
20.0%
51.6%
23.7%
43.3%
76.3%
54.2%
10+
993
48.6%
47.1%
16.4%
45.4%
83.6%
47.5%

Some sites provided data on small numbers of individuals

This fact is always of concern regarding the reliability of the data provided being representative of the individuals seen at a particular Diabetes Centre. The Mean and Median number of individual forms provided by sites in ANDIAB2 [76 and 79 respectively (range 8-220)] suggests a reasonable spread – and indeed 9 sites had 100 or more forms. Eight had less than 50 however - with six of these less than 40. Notwithstanding this, the pooled data are similar in demographic characteristics to the last ANDIAB collection in 2009 as noted above [Page 7 and Table 2]. It is therefore not considered that small ‘non-representative’ numbers have adversely affected interpretation of the pooled data as reported here.

Strengths of ANDIAB2

Geographical spread; significant patient numbers

A total of twenty seven sites participated from across the country, but predominantly from the east coast: [NSW 10; VIC 7; QLD 6; TAS 2; WA 1; ACT 1]. De-identified data were provided from 2131 individuals.

Data completeness and correctness

‘Validation Reports’ generated for each Centre requesting missing data and correction of questionable or potentially invalid data were addressed and returned by all but one of the twenty seven sites. This means that sites have done their upmost to ensure data completeness and correctness, enhancing the reliability of the findings.

Presentation of the Data

The following Abstract [Figure 2] was submitted and accepted as a Poster Presentation at the 2010 ADS/ADEA Annual Scientific Meeting and was published in the Proceedings: Australian Diabetes Educators Association Meeting, Sydney, Sept 2010; Abstract 513, page 224. Further presentations of the results are planned in 2011.

Figure 2
DEVELOPMENT AND IMPLEMENTATION OF A DIABETES DISTRESS SCALE CALCULATOR: FREE-STANDING AND WEB BASED COMPUTER APPLICATIONS
Jeff R Flack, Brian Sandiforth, William H. Polonsky*
Diabetes Centre, Bankstown-Lidcombe Hospital, NSW, on behalf of the NADC
*Department of Psychiatry, University of California, San Diego
Background: The Diabetes Distress Scale [DDS] is a validated 17-item questionnaire, yielding four reliable subscales targeting different areas of potential diabetes-specific distress to help clinicians and patients identify areas where interventions might be helpful: emotional burden; physician-related distress; regimen-related distress; and interpersonal distress.1 The Australian National Diabetes Information Audit and Benchmarking2 [ANDIAB2] initiative is an audit of patient characteristics, self-care practices, quality of life and outcomes. The DDS instrument was chosen to assess individuals in specialist diabetes services in ANDIAB2, to be undertaken in April/May 2010. Aim: To develop, test and implement a DDS Calculator to automate the scoring of the DDS and its subscales in ANDIAB2.
Methods: Utilising an advanced Flash™ creation tool, we developed an executable program that provides a total and four-subscale scores, with the option to graph the output. Various rules were implemented: Total-Score calculations do not proceed if fewer than 14 responses are provided; Subscale calculations do not proceed if more than one item response in that subscale is missing. On-screen pop-up messages explain this to the user if invoked. Multiple test scenarios were developed to ensure outputs were correct.
Results: A free-standing Shockwave Flash-based executable application for Windows XP© was developed which performed correctly in all test scenarios, and was distributed to ANDIAB2 participants, and a web-based version loaded onto our Diabetes Centre website.2 Value-added components include on-screen explanatory notes, hyperlinks to relevant publications, ability to see results on-screen in graphical format, and ability to print the computed results and/or the graph.
Conclusions: We believe that this easy to use application will assist clinicians utilising the DDS to calculate the relevant scores and display them in a format suitable for feedback of results to individuals with diabetes. If opinion from ANDIAB2 participants is favourable, we intend to offer this application for wider distribution.
1. Development of the Diabetes Distress Scale. Assessing psychosocial distress in diabetes. Polonsky WH, Fisher L, Earles J, Dudl RJ, Lees J, Mullan J, Jackson RA (2005). : Diabetes Care, 28, 626-631.
2. Bankstown-Lidcombe Hospital Diabetes Centre
Acknowledgement: ANDIAB2 2010 was funded by the Commonwealth Department of Health and Ageing. Presented on behalf of the National Association of Diabetes Centres [NADC].

Top of page