Published online: 9 August 2019
Note: The VDR algorithm was updated in 2017 resulting in a small reduction in the total number of people estimated to have diabetes. It is important that the updated numbers are not compared with numbers derived using any previous version of the VDR. Any comparison may result in artificial and inaccurate trends.
Each year the New Zealand Ministry of Health releases national estimates of the prevalence of diabetes based on the VDR. The VDR is an important tool to monitor prevalence of diabetes and support national and local clinical quality improvements.
The VDR contains data about people suspected as having diabetes, identified through their use of diabetes health services. The VDR uses an algorithm to identify these people in data extracted from hospital inpatient and outpatient, laboratory test, and pharmaceutical dispensing data collections. The VDR is collated annually at the end of March and national and regional diabetes prevalence estimates are calculated based on the number of people on the VDR as at 31 December of the previous year. People with diabetes who were deceased and those not enrolled in a PHO were excluded from the totals.
Virtual Diabetes Register 2018 revision
In 2016 the algorithm used to create the VDR was assessed against the Auckland TestSafe repository of actual glycaemic test results. (For more information read the published article: Can administrative health utilisation data provide an accurate diabetes prevalence estimate for a geographical region? - ScienceDirect website).
As a result, improvements to the algorithm were made in early 2017 to create the latest version of the VDR. A comparison of the numbers from the old VDR and the latest version (v686) is shown in the chart below.
Number of people estimated as having diabetes, on the old VDR and the new VDR, 2005–2018
|Old VDR||New VDR|
Note: People who were either not enrolled in a PHO or deceased, as at 31 December of the relevant year, have been excluded.
It is important to note that numbers produced by the algorithm in the old VDR are now obsolete and should not be used, nor compared to the numbers produced by the new VDR algorithm. From now onwards, please use the numbers published here or request updated numbers or datasets firstname.lastname@example.org.
|DHB of domicile||2010||2011||2012||2013||2014||2015||2016||2017||2018|
|Bay of Plenty||9507||9917||10,241||10,433||10,537||10,330||10,391||10,722||11,169|
|Capital and Coast||9393||10,788||11,899||12,436||12,805||13,119||13,350||13,370||13,823|
Note: People that were either not enrolled in a PHO or were not alive, as at 31 December of the relevant year, have been excluded.
Source: VDR Dec 2010–2018 (v686)
Notes about VDR 2018:
- People who weren't enrolled in a PHO, or were dead as at 31/12/2018, have been excluded.
- Data was extracted on 05/04/2019.
- NNPAC data from Canterbury DHB was removed for Oct, Nov, and Dec 2018 due to data quality concerns.
- Any unknown ethnicity information on the NHI was bulk updated from ethnicity data on PHO enrolments in 2018. As a result this has changed the ethnicity distribution in the VDR 2018 when compared to previous years.
- Chemical ID 4103 Vildagliptin and 4101 Vildagliptin with metformin hydrochloride were added to the list of pharmaceuticals included in the algorithm for the VDR 2018. These pharmaceuticals were not availalble prior to 2018, so do not affect the data in previous VDR years.