Trial of screening invitation for patients over 15% QDiabetes risk

Today we started the third thread of the Project, inviting patients with > 15% 10 yr risk of Diabetes via QDiabetes to have screening.  As a trial in one practice (my own, Falkland Surgery) we have today written to the 266 patients with a QDIabetes score of 15% and above, offering them one of two screening options.

Group 1 will get a letter inviting then to a drop in clinic next Saturday where we will offer them an on-the-spot HbA1c and random blood sugar test and advice including enrollment in our Eat4Health sessions.

Group 2 will get a similar letter but inviting them to get a routine HbA1c and Glucose blood test performed by phlebotomy in the usual way.

Although not statistically significant in size we hope to compare the costs and response rates of the two groups to guide us when we repeat this across the patch (potentially inviting about 1500 patients for screening).

Based on our Newbury Show Screening where we tested 111 patients with risks above 15% and picked up 5 Diabetic HbA1c levels, and another 5 borderline results, we might expect a similar pickup rate.  Depending on the response rate to the letters we could potentially identify 12-15 new Diabetics, and the same number of borderline (IGT) results in this process.

Initial results for the first arm  (The drop in clinic) will be posted next weekend

2nd Cohort PreDM results just in

The second cohort of patients has now completed their Eat4Health course.  They acheived a fantastic 4.5kg average weight loss, or the equvalent of 5% of their 96Kg starting weight through the course of the sessions.  Congratulations, that is a massive effort.

Distribution of QDiabetes Results in a random GP practice


We looked at the distribution of QDiabetes scores in an urban practice of 14500 in semi-rural Berkshire, so an area of mainly white/caucasian patients and found this distribution.  Clearly a higher incidence of higher scores in men, and mainly in the 50-60yr old age group

We are shortly going to target those with a higher risk score for focussed screening.

 QDScore > 20% Female   0 0 0 0 11 26 51 6 0 0 0
  Male   0 0 1 2 16 40 70 9 0 0 0
QDScore > 30% Female   0 0 0 0 1 11 19 2 0 0 0
  Male   0 0 0 1 4 16 23 5 0 0 0
QDScore > 40% Female   0 0 0 0 0 4 1 0 0 0 0
  Male   0 0 0 1 2 5 3 0 0 0 0
QDScore > 50% Female   0 0 0 0 0 1 0 0 0 0 0
  Male   0 0 0 0 1 1 2 0 0 0 0

QDiabetes compared with other risk tools

QDiabetes appears to fare well in comparison with other Diabetes risk assessment tools

“The QDScore (QDiabetes) was one of the best discriminatory models across countries. It is based on more predictors than other models are and so uses more information, and was developed from a large sample, providing precise estimates of the associations between predictors and outcomes”

PreDM Project on tour!!!

I’m currently stopped over at Nottingham on my way to the RCGP Conference tomorrow in Harrogate for a Poster Presentation, then back down for Friday at the Emis NUG Conference and a lunchtime presentation there.

Come and see me at either (or both) conference!

Costings for PreDM stand at Newbury Show

These are the costings for the PreDM Stand at Newbury Show


6*6m erected tent = 890
3*2m trestle table = 43
Car Passes = 15
Power Supplies * 2 = 228
Work tables * 3 = 51
Passes * 24 = 229
iPads * 4 = FOC (loan)
Public Liability Insurance = 71
Laptop for data entry = FOC
Banners for tent 6m * 2 = 175
Banners for tables 4*2m = 222
Table cloths = 58
Board covering = 10
Thermal Rolls = 11
Pins, tape etc = 6
Velcro for banners etc = 25
t-Shirts for E4H * 4 = 110
Mileage for E4H
Leaflets from Diabetes UK = 15
Chairs * 10 = 100
Projector and screen = FOC
Display boards/privacy screens * 5 = FOC
HbA1c test strips * 110 = 440
HbA1c machines * 2 = Loan
Gloves and lancets etc = from stock from Dept of PH
Sweatshirts * 5 =120

Misc catering etc circa 100


E4H staff * 4 = 500
PH Nurses * 3 = (approx 500) funded by PH
PreDM staff * 3 for front of house daily = 1200 (paid at 15pph)

Total cost for the 2 days, (excluding planning and setting up the previous day), approximately 5000 (of which half was staffing, split between the PreDM staff doing the QDiabetes screening, the E4H staff promoting exercise, and the nursing staff running the HbA1c tests and counselling)

It is impossible to break down the individual value of each component but this “bought”

310 QDiabetes screening and lifestyle advice
111 HbA1c tests and lifestyle advice
80 Body compositions analyses and lifestyle advice
5 new diagnoses of Diabetes
Approximately 10 borderline results
At least 30 people signed up for weight loss program, Eat4Health
General raising of awareness of diabetes risk in the community (approximately 60,000 people at the show) and both the wider public and local clinicians

Newbury Show Pre-diabetes Project Results

A few days on and I have the final results from our Newbury Show stand.

We performed

310 QDiabetes assessments and risk counselling conversations.

111 HbA1c point of care tests on those with a QDiabetes Risk of 10% or more.

80 Body Composition Analyses as part of healthy lifestyle discussions.

We provisionally diagnosed 5 people as diabetic who were not previously diagnosed.

There were 8 people with HbA1c 42-47 (at risk of DM) who were advised to get further assessment.

We have subsequently booked 30 people into the Eat4Health Sessions as a result of the stand


Notes.  Many of the QDiabetes calculations were prompted by people requesting them (particularly at busy times) though we did invite passer-by people to be tested (and we recognise a slight bias to inviting those people who might have a higher BMI!)

The threshold for offering HbA1c was set at 20% QDiabetes risk on Saturday, (plus clinically driven exceptions) but having assessed the day one data we revised our threshold down to 10% risk on the second day.  This was based on the experience of identifying someone with DM with a risk of 16% on day one.  Overall there is a balance to be struck between cost, workload and pickup rate, but informally we would suggest a 15% QDiabetes risk would have caught all of the “hidden” diabetics in our small cohort.

There was a fairly “generous” approach to applying the rules, with clinicians asking for HbA1c according to their judgement at times.

Some people slipped through the net and only told us they were diabetic after QDiabetes assessment (!) but gratifyingly their risks were generally over 30%.  Similarly we tested a couple of people with very low apparent QDiabetes scores, due to symptoms, BMI or suspicions.

In our overall small sample there was a small bias of risk score to be higher in the male subjects.  There appears to be an overall trend of higher HbA1c results in those tested, correlating to the QDiabetes risk score.

Risk vs score graph

End of Newbury Show – immediate thoughts

We are exhausted after 2 full-on days of PreDM work at the Newbury Show.  We are still counting the numbers but know we did in excess of 300 QDiabetes assessments, about 120 targetted HbA1c tests and advice sessions (based on their QDiabetes score), and about 80 Body Composition Measurements (full body fat analyses).  We have provisionally diagnosed 5 people with Diabetes (based on combined readings of HbA1c up to 64, Raised random blood sugar and symptoms) and a similar number of borderline results (generally HbA1c 42-47).  The biggest impact, albeit un-measureable, is likely to be the general raising of awareness in those just passing, or visiting the stand.

We did have one lady with a Blood Sugar of 28, until we washed the doughnuts off her hands!

Further details and analysis to follow

Many thanks to the whole team for a fantastic weekend

The Pitch during a quiet time!




Some members of the team