• We are sorry, but NCBI web applications do not support your browser and may not function properly. More information

Self-mixed or pre-mixed insulin therapy

StudyPopulationInterventionOutcomesResultsCommentsDesignEL
O’Hagan and Greene, 199320840 children with type 1 diabetes
Aged 7–16 years
UK
All children were on a twice-daily self-titrating regimen of short- and intermediate-acting insulins. 14 children were using insulin ratios other than 30:70 (range 10:90 to 50:50)
After a run-in period of 1 month patients were randomised. After 3 months the patients were swapped to the other treatment
  1. self titrating regimen (self- mixed) by conventional syringe
  2. pre-mixed regimen, disposable pen injector
  1. Glycated haemoglobin HbA1
  2. 6-point blood glucose profile
  3. Hypoglycaemic and ketoacidotic episodes
  4. Questionnaire discussing: convenience of use, ease and accuracy of dose settings, as well as in preference to conventional syringes. Completed after the study had ended
  1. No significant difference in the mean HbA1 at the beginning and the end of the treatment periods: start 13.1 ± 0.6%, self- titrating 11.8 ± 0.5%, pre-mixed 12.5 ± 0.4%
  2. No statistically significant difference between the self-mixed and pre-mixed treatments group children in their blood glucose profiles throughout the day
  3. No statistical difference in number of hypoglycaemic or ketoacidotic episodes reported during the self-titrating or pre-mixed treatments
  4. The disposable pen was preferred by 95% of the patients as regards convenience of use, ease and accuracy of dose setting as well as in preference to conventional syringes
No description of how randomisation took place
Different delivery systems used for each treatment group (syringes for self- mixed and pens for pre- mixed) may lead to bias if differences are due to delivery system not pre- mixed or self-mixed insulin, (especially in the patient preference questionnaire)
Novo Nordisk provided financial support, pen injectors and all insulin preparations
RCT crossoverIb
Arslanoglu et al, 200020920 patients with type 1 diabetes
Aged 8.2–19.6 years
Turkey
Duration of diabetes greater than 6 months, and they used conventional insulin therapy
Patients were divided randomly into 2 groups. After 2 months the patients were swapped to the other treatment
  1. self-mixed conventional insulin therapy by conventional syringe (12 mm long, 29 gauge)
  2. pre-mixed insulin therapy (available insulin ratios were 10:90, 20:80, 30:70, 40:60 and 50:50) by pen injector (8 mm long, 30 gauge)
  1. Hypoglycaemic and ketoacidotic episodes reported
  2. Blood glucose measurements to give mean glycaemia values
  3. Patient acceptance
  1. No significant difference in HbA1c values between the two groups at the end of the treatment periods: self-mixed 11.1 ± 2.6% vs. pre-mixed 10.9 ± 2.7%
  2. No significant difference in total number of hypoglycaemic episodes: 53 for self- mixed and 88 for pre-mixed periods. When divided by time, hypoglycaemic attacks were more common at 07:00 during pre-mixed treatment (p < 0.05) and at 03:00 during self- mixed treatment (p < 0.00001)
  3. No significant difference in mean glycaemia values between the two groups: self-mixed 8.3 ± 1.7 vs. pre-mixed 8.1 ± 1.7 mmol/l
  4. 100% of patients preferred the pre-mixed insulin delivered by pen to the self-mixed insulin delivered by syringe
No description of how randomisation took place
Different delivery systems used for each treatment group (syringes for self- mixed and pens for pre- mixed) may lead to bias
Novo Nordisk provided pen injectors and questionnaires
RCT crossoverIb
Dunbar et al, 199421227 adults with type 1 diabetes
Aged 18–63 years
Ireland
Patients had been receiving stable insulin regimens of short-acting and intermediate-acting insulin for 2 months
After a 1-month run-in period patients were randomised. After 2 months the patients were swapped to the other treatment
  1. continue previous treatment (self-mixed)
  2. pre-mixed insulin preparation that delivered short- and intermediate-acting insulin in the preparation closest fit to previous ratio from 10:90, 20:80, 30:70, 40:60 and 50:50 preparations
  1. Total glycated haemoglobin level
  2. Plasma glucose measurements
  3. Blood glucose measurements
  4. Hypoglycaemic episodes (classified in grades; I- mild, II- severe symptoms not requiring assistant from others and III- reduction in level of consciousness so assistant from others is required IV- severe reduction in consciousness necessitating parenteral treatment or treatment from a physician)
  5. Dose of insulin received
  6. Questionnaire discussing preference for treatment A or B after the study had ended Originally 32 patients, 5 dropped out due to a variety of reasons, 2 specially due to pen concerns
1.

No significant difference in glycated haemoglobin values between the two groups

3.

No systematic differences were noted in the seven point blood glucose profiles

4.

No significant difference in hypoglycaemic episodes between the two groups. Total episodes: Grade I self-mixed 60 vs. pre-mixed 83; Grade II self-mixed 59 vs. pre-mixed 32; Grade III self-mixed 17 vs. pre-mixed 7; Grade IV self-mixed 1 vs. pre- mixed 0; total all grades self-mixed 137 vs. pre-mixed 122, p = 0.20

5.

No significant difference in dose of insulin received: pre-mixed 46.6 ± 18.9 units vs. self- mixed 46.9 ± 19.3 u, p = 0.62

6.

Pre-mixed regimens delivered by pen were preferred to self-mixed regimens delivered by syringe in 82% of patients

No description of how randomisation took place
Different delivery systems used for each treatment group (syringes for self- mixed and pens for pre- mixed) may lead to bias if differences are due to delivery system not pre- mixed or self-mixed insulin, (especially in the patient preference questionnaire)
RCT crossoverIb
Davies et al, 198821310 male adults with type 1 diabetes
Aged 21–40 years
UK
Patients had been receiving twice- daily mixtures of short- and intermediate-acting insulin
After a 4-week run-in period patients were allocated in random order to 3 regimens for 1 day each
  1. usual twice-daily mixtures of short- and intermediate- acting insulin
  2. pre-mixed insulin preparation 50:50 soluble (human actrapid), isophane (human protaphane). Administered by penject (Hypoguard). The total insulin dose was divided to give 30% at 07:30, 20% at 12:00, 30% at 17:30 and 20% at 21:30h
  3. pre-mixed insulin preparation 30:50 soluble and isophane (human Actraphane). Administered by NovoPen. The total insulin dose was divided to give 30% at 07:30, 20% at 12:00, 30% at 17:30, and 20% at 21:30
  4. prandial soluble and isophane insulin by NovoPen was divided to give 40% before breakfast, 20% before lunch and 40% before dinner. Evening crystalline insulin zinc human Ultratard) by disposable plastic syringe
  1. Total glycated haemoglobin level
  2. Plasma glucose measurements
  1. Total glycated haemoglobin level did not statistically significantly change during the study period, from 8.8 ± 1.6% at the start of the study to 8.3 ± 1.5% at the end of the study, this was not analysed by treatment group
  2. There were significant differences in the mean plasma glucose profiles of the basal and prandial regimen (D) compared with the other three regimens. Between 24:00 and 03:00, regimen D was significantly higher (p < 0.05), between 08:30 and 11:00, regimen D was significantly lower (p < 0.05)
Peak morning glucose levels were significantly lower on the basal and prandial regimen (D) than on the other three regimens: D 9.5 ± 4.3 vs. A 13.8 ± 2.8 vs. B 13.6 ± 5.3 vs. C 13.5 ± 4.2 mmol/l, p < 0.01
The mean 24-hour plasma glucose concentrations were similar for all 4 regimens but mean concentration between 07:30 and 12:00 was significantly lower on the basal and prandial regimen (D) than on the other three regimens: D 7.8 ± 3.6 vs. A 10.8 ± 3.3 vs. B 11.4 ± 5.1 vs. C 10.7 ± 3.6 mmol/l, p < 0.01
There were significant differences in the plasma free insulin profiles (as collected from 8 of the patients). The mean concentrations between 19:00 and 21:30 were significantly lower on the pre-mixed 30:70 regimen than the pre-mixed 50:50 regimen and the twice-daily regimen (p < 0.05). The mean concentrations between 22:30 and 01:00 were significantly lower on the basal and prandial regimen (D) than on the other three regimens (p < 0.05). Insulin concentrations were similar in the remainder of the 24-hour study period
No description of how randomisation took place
Different delivery systems used for each treatment group may lead to bias if
differences are due to delivery system not pre- mixed or self-mixed insulin
Novo Industries provided financial support
RCT crossoverIb
Pizzey et al, 199621065 adults with type 1 and type 2 diabetes who had been on insulin treatment for at least 3 months
Aged 19–72 years
5 centres in the UK
Patients were receiving Humulin S and Humulin I by self-mixture or fixed-mixture morning and evening
After a 3-month run-in period patients were randomly allocated by computer- generated numbers to treatment regimens for 9 months
  1. mixture (flexible – pre- mixed insulin adjusted for different ratios of soluble and intermediate insulin for both morning and evening dose)
  2. mixture (fixed – single fixed preparation of the same ratio both morning and evening)
  3. S + I (free mixing – self- mixed)
  1. Hypoglycaemic episodes
  1. During the run-in period HbA1c showed a statistically significant improvement in groups B (from 11.0 ± 4.2% to 9.0 ± 3.0%, p < 0.005) and C (from 10.5 ± 4.2% to 8.6 ± 3.0%, p < 0.005) but not in group A (from 9.3 ± 3.6% to 8.9 ± 4.2%, p > 0.05)
    There was a statistically significant improvement in HbA1c levels during the 9- month trial in group C (from 8.6 ± 3.0% to 7.6 ± 3.6%, p < 0.05) but not in groups A (from 8.9 ± 4.2% to 8.9 ± 5.3%, p > 0.05) or B (from 9.0 ± 3.0% to 8.5 ± 3.6%, p > 0.05)
  2. No significant difference in the number of hypoglycaemic episodes among the groups
Financial support from Lilly Industries UK
26 of the 91 patients originally recruited dropped out: this was mainly due to failure to attend clinic. Similar number of patients dropped out of each group
RCTIb
Cucinotta et al, 199121520 adults with type 1 diabetes
Aged 19–72 years
Italy
Patients had been receiving soluble and isophane human insulin injected before breakfast and dinner
After a 2-week run-in period patients were randomised. After 4 weeks the patients were swapped to the other treatment
  1. continue previous treatment (self-mixed)
  2. human pre-mixed insulin
  1. 6-point blood glucose profile
  2. Hypoglycaemic episodes (graded I – mild, II – moderate, III – severe and IV – severe including hospitalisation)
  3. Preference for future treatment
  1. Blood glucose profiles did not change during the study compared with the run-in period and did not show any difference between the pre-mixed insulin and the self- mixed insulin
    The mean daily glycaemia with pre-mixture (139.7 ± 2.4 mg/dl) was similar to that of the self-mixture (137.5 ± 2.9 mg/dl)
  2. No difference was observed in the number or severity of the hypoglycaemic episodes between the two treatments. No adverse events were reported during the study
  3. 18/20 patients declared their preference for the pre-mixture for safety and use simplicity reasons
No description of how randomisation took place
Unknown methods of delivery
RCT crossoverIb
Corcoran and Yudkin, 198621112 patients with type 1 diabetes
Unknown age
UK
Porcine soluble and isophane insulin twice daily
versus
pre-mixed Porcine soluble and isophane insulin twice daily
versus
porcine soluble and insulin zinc suspension twice daily
versus
beef ultralente (crystalline insulin zinc suspension) in morning and three purified porcine soluble before meals
Trial length: four 8-weeks periods
Mean blood glucose
HbA1c
No significant difference (numbers not given – presented as a graph)No description of how randomisation took placeRCT crossoverIb
Kinsley and McKenna, 1999214600 patients with type 1 diabetes
Mean age: pre- mixed 42 ± 1.6, separate insulin preparations 40 ± 0.7 years
Ireland
Survey
Pre-mix insulin
versus
separate insulin preparations (2, 3 and 4 injections a day)
HbA1cPatients aged < 35:
7.8 ± 0.2% (n = 62) vs. (2 separate insulin preparations) 6.9 ± 0.2% (n = 85), p < 0.001
7.8 ± 0.2% (n = 62) vs. (3 separate insulin preparations) 7.6 ± 0.2% (n = 38), p = NS
7.8 ± 0.2% (n = 62) vs. (4 separate insulin preparations) 7.3 ± 0.2% (n = 83), p < 0.05
Patients > 35:
7.5 ± 0.2% vs. 7.5 ± 0.1% (all numbers of insulin injection)
CohortIIb

From: Evidence tables

Cover of Type 1 Diabetes
Type 1 Diabetes: Diagnosis and Management of Type 1 Diabetes in Children and Young People.
NICE Clinical Guidelines, No. 15.2.
National Collaborating Centre for Women's and Children's Health (UK).
London: RCOG Press; 2004 Sep.
Copyright © 2004, National Collaborating Centre for Women’s and Children’s Health.

Apart from any fair dealing for the purposes of research or private study, criticism or review, as permitted under the Copyright, Designs and Patents Act, 1988, no part of this publication may be reproduced, stored or transmitted in any form or by any means, without the prior written permission of the publisher or, in the case of reprographic reproduction, in accordance with the terms of licences issued by the Copyright Licensing Agency in the UK. Enquiries concerning reproduction outside the terms stated here should be sent to the publisher at the UK address printed on this page.

The use of registered names, trademarks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant laws and regulations and therefore for general use.

The rights of National Collaborating Centre for Women’s and Children’s Health to be identified as Author of this work have been asserted by them in accordance with the Copyright, Designs and Patents Act, 1988.

PubMed Health. A service of the National Library of Medicine, National Institutes of Health.