CSII

StudyPopulationInterventionOutcomesResultsCommentsDesignEL
Colquitt et al, 20031726 RCTsRapid-acting insulin analogue
versus
soluble insulin
Trial length: at least 10 weeks
HbA1c
Hypoglycaemic episodes
Body weight
Insulin dosage
5 crossover design RCTs investigated the use of insulin lispro compared with soluble insulin
1 parallel design RCT has 3 groups and investigated the use of insulin lispro, insulin aspart and soluble insulin
HbA1c level was found to be significantly improved with insulin lispro (WMD −0.26, 95% CI −0.47 to −0.06)
Some studies reported fewer hypoglycaemic episodes with analogue insulin but this varied with the definition of hypoglycaemia used
No differences in body weight or insulin dosage were seen
RCTs included Bode 2002178, Melki 1998173, Raskin 2001174, Renner 1999175, Schmauss 1998176 Zinman 1997 177Systematic reviewIa
Bode et al, 2002178146 adult patients with type 1 diabetes
Aged 18–71 years
USA
After 4-week run-in period using an external CSII, patients were randomly assigned to receive either:
  1. rapid-acting insulin analogue (insulin aspart) (n = 59)
  2. soluble human insulin (n = 59)
  3. rapid-acting insulin analogue (insulin lispro) (n = 28)
Trial length:16 weeks
  1. HbA1c change from baseline
  2. Hypoglycaemic episodes (p value from wilcoxon rank-sum test relative to insulin aspart group)
  1. 0.00 ± 0.51% vs. 0.15 ± 0.63% vs. 0.18 ± 0.84%
  2. All: 6.7 ± 5.4 vs. 10 ± 8.9 (p = 0.034) vs. 10.5 ± 8.1 (p = 0.044)
    With blood glucose < 50 mg/dl: 3.7 ± 3.6 vs. 4.8 ± 4.2 (p = 0.175) vs. 4.4 ± 4.7 (p = 0.841)
    Nocturnal < 50 mg/dl: 0.5 ± 0.83 vs. 0.9 ± 0.97 (p = 0.004) vs. 0.6 ± 0.61 (p = 0.189)
In systematic reviewmulticentre parallel RCTIb
Raskin et al, 200117458 adult patients with type 1 diabetes
Aged 13–60 years
USA
After 2–4 week run-in period using an external CSII, patients were randomly assigned to receive either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 58)
  2. soluble human insulin (n = 58)
Trial length: two 12-week periods
  1. 7.41 ± 0.97% vs. 7.65 ± 0.85%, p = 0.004
  2. 7 patients reported 8 episodes vs. 7 patients reported 11 episodes
  3. 12 patients reported 16 episodes due to occlusion vs. 12 patients reported 23 episodes
In systematic reviewCrossover RCTIb
Schmass et al, 199817611 adult patients with type 1 diabetes
Mean age 30 ± 2.5 years
Germany
After 4-week run-in period using an external CSII, patients were randomly assigned to receive either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 11)
  2. soluble human insulin (n = 11)
Trial length: two 3-month periods
  1. Blood glucose
  2. HbA1c in first period
  3. Adverse effects
  1. Fasting: 6.5 ± 0.4 vs. 7.5 ± 0.4 mmol/l, NS
    2-hour postprandial: 6.8 ± 0.3 vs. 8.3 ± 0.3 mmol/l, p = 0.03
  2. 5.7 ± 0.3% vs. 6.5 ± 0.3%
  3. No significant difference
  4. No significant difference
In systematic reviewCrossover RCTIb
Zinman et al, 199717730 adult patients with type 1 diabetes
Aged 35.1 ± 1.5 years
Canada
Patients were all treated using CSII system
After 1-month run-in period using an external CSII, patients were randomly assigned to receive either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 30)
  2. soluble human insulin (n = 30)
After 3 months the patients were switched to the other treatment for a further 3 months
  1. Blood glucose levels
  1. Postprandial blood glucose levels were significantly lower with insulin lispro compared with soluble human insulin: (from graph lispro vs. soluble human insulin) post- breakfast; 8.3 vs. 9.8 mmol/l, post-lunch 7.5 vs. 8.6 mmol/l, post-supper 7.7 vs. 8.8 mmol/)
  2. The decrease in HbA1c after treatment was greater in the insulin lispro treatment group than the soluble human insulin group (7.66 ± 0.13% vs. 7.00 ± 0.16%, p = 0.0041, difference of 0.66%)
No description of how randomisation took place
Sponsored by Eli Lilly
In systematic review
RCT crossover double blindIb
Melki et al, 199817339 adults with type 1 diabetes
Aged 39.4 ± 1.5 years
France
After 4-week run-in period using an external CSII patients were randomly assigned to receive either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 39)
  2. soluble human insulin (n = 39)
After 3 months the patients were switched to the other treatment
  1. HbA1c (measured only during the first period of treatment, so no carry-over effect)
  2. Mean daily blood glucose levels
  3. 2-hour postprandial blood glucose level
  4. Preprandial blood glucose level
  1. The reduction in HbA1c level was significantly more pronounced with insulin lispro than soluble human insulin (lispro: from 7.74 ± 0.20% to 7.11 ± 0.15%, a difference of −0.62 ± 0.13% vs. soluble human insulin: from 7.97 ± 0.13% to 7.88 ± 0.16%, a difference of −0.09 ± 0.15%, p = 0.01, difference of 0.71%)
  2. There was significantly lower mean daily blood glucose levels in the lispro group compared with the soluble human insulin group (7.93 ± 0.15 vs. soluble human insulin 8.61 ± 0.18 mmol/l, p < 0.0001)
  3. There was significantly lower 2-hour postprandial blood glucose level in the lispro group compared with the soluble human insulin group (8.26 ± 0.19 vs. 9.90 ± 0.20 mmol/l, p < 0.0001)
  4. There was no significant difference in the preprandial blood glucose level between the two groups (7.70 ± 0.17 vs. 7.75 ± 0.21 mmol/l, p > 0.05)
No description of how randomisation took place
Sponsored by Lilly France
RCT crossover multicentreIb
Renner et al, 1999175113 adults with type 1 diabetes
Mean age 37 ± 12 years
Germany
Patients were all treated with CSII for 6 months or more prior to study
After 4-week run-in period patients were randomly assigned to receive treatment through CSII either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 113)
  2. soluble human insulin (n = 113)
After 4 months the patients were switched to the other treatment for 4 months
  1. Daily and postprandial blood glucose profiles
  2. Adverse events
  3. Hypoglycaemic events (< 3.5 mmol/l)
  4. Treatment satisfaction
  1. The mean HbA1c at the end of the period was significantly lower after insulin lispro treatment than after soluble human insulin treatment (6.77 ± 0.88% vs. 6.90 ± 0.97%, p < 0.02, difference of 0.13%)
  2. Insulin lispro compared with the soluble human insulin treatment significantly reduced the breakfast, lunch and dinner postprandial blood glucose excursions and the 22:00 blood glucose level (breakfast 7.0 ± 1.9 vs. 8.6 ± 2.6 mmol/l, p < 0.001; lunch 7.6 ± 1.9 vs. 8.7 ± 2.4 mmol/l, p < 0.001; dinner 7.2 ± 1.9 vs. 8.3 ± 1.9 mmol/l, p < 0.001 22:00 7.6 ± 1.8 vs. 8.3 ± 2.0 mmol/l, p < 0.001)
    There was no significant difference in the 02:00 or breakfast, lunch or dinner preprandial blood glucose results between the treatment groups (p > 0.05)
  3. There were 6 serious adverse events during treatment with soluble human insulin and 1 during treatment with insulin lispro; these were not identified to be related to the study drug.
    The most frequently reported adverse events in both groups were infections (lispro 19.4%, soluble 21.1%; mainly common cold) and rhinitis (lispro 15.8%, soluble 13.8%)
    Among adverse events possibly related to study drug, the most frequent was ketosis, reported by 5 (4.5%) patients receiving lispro and 4 (3.7%) patients receiving soluble human insulin
    4 (3.7%) patients during insulin lispro treatment and 2 (1.8%) patients during soluble human insulin treatment reported injection site reactions
    No difference between insulin lispro and soluble human insulin in the rate and number of catheter occlusions
  4. There were no significant differences between groups in the average number of hypoglycaemic episodes per patient (12.4 ± 13.9 vs. 11.0 ± 11.2, p > 0.05)
  5. The questionnaire used to assess the satisfaction of patients with each therapy (maximum score 48) showed a statistically significant result in favour of insulin lispro treatment (35.16 ± 4.25 vs. 32.36 ± 5.87, p < 0.001)
No description of how randomisation took place
Sponsored by Lilly
RCT open label crossoverIb
Guerci et al, 199917910 adults with type 1 diabetes
Mean age 41.1± 7.1 years
France
Patients were all treated using CSII system
After 1-month run-in period patients were randomly assigned to receive through CSII either:
  1. rapid-acting insulin analogue (insulin lispro) (n = 10)
  2. soluble insulin (n = 10)
After 1 month the patients were switched to the other treatment for 1 month
HbA1c
Mean blood glucose levels
Mean standard deviation of blood glucose levels
Postprandial blood glucose level
Hypoglycaemia
There was no significant difference in HbA1c, mean blood glucose levels or mean standard deviation of blood glucose between the two treatment groups ((mean ± 1SD) HbA1c 7.07 ± 0.51% vs. 6.67 ± 0.67%, mean blood glucose 9.04 ± 0.89 vs. 9.32 ± 1.17 mmol/l, mean standard deviation of blood glucose 4.44 ± 0.49 vs. 4.82 ± 0.83 mmol/l)
There was a significant decrease in the postprandial blood glucose level and the hypoglycaemia incidences in the lispro compared with the soluble insulin treatment group (postprandial blood glucose level 9.43 ± 1.39 vs. 10.49 ± 2.05 mmol/l, p < 0.05; hypoglycaemia incidence 7.1 ± 4.6 vs. 12.6 ± 10.2, p < 0.05)
No description of how randomisation took place
Small number of participants
Patients used CSII – can this be generalised to general population?
RCT crossoverIb
Johansson et al, 200018041 adults with type 1 diabetes
Mean age 40.2 ± 10.0 years
Sweden
Some patients were treated using CSII system
After 2–4 week run-in period patients were randomly assigned to receive (through CSII or injection treatment) either:
  1. rapid-acting insulin analogue (insulin lispro) (5 min before meal) (n = 40)
  2. soluble insulin (30 min before meal) (n = 30)
After 2 months the patients were switched to the other treatment for 2 months
  1. Mean glycaemia and SD of all blood glucose levels
  2. Mean postprandial glycaemia and SD of postprandial blood glucose levels
  3. Mean preprandial glycaemia and SD of preprandial blood glucose levels
  4. Hypoglycaemic events
  1. There was a significantly higher level of HbA1c in the soluble insulin treatment group than the insulin lispro treatment group: mean HbA1c (lispro vs. soluble insulin) 7.4 vs. 7.6%, p = 0.047, difference of 0.2%
  2. There was a significantly higher mean glycaemia and SD of all blood glucose levels in the soluble insulin treatment group than the insulin lispro treatment group: (lispro vs. soluble insulin) mean glycaemia 7.4 vs. 7.6 mmol/l, p < 0.001; SD of all blood glucose levels 3.6 vs. 3.9 mmol/l, p = 0.012
  3. There was a significantly higher mean postprandial glycaemia and SD of postprandial blood glucose levels in the soluble insulin treatment group than the insulin lispro treatment group: (lispro vs. soluble insulin) mean postprandial glycaemia 8.1 vs. 9.6 mmol/l, p < 0.001; SD of postprandial blood glucose levels 3.6 vs. 4.0 mmol/l, p = 0.006
  4. There was no significant difference in mean preprandial glycaemia and SD of preprandial blood glucose levels between the soluble insulin treatment group and the insulin lispro treatment group: (lispro vs. soluble insulin) mean preprandial glycaemia 8.5 vs. 8.4 mmol/l, p = 0.86; SD of preprandial blood glucose levels 3.4 vs. 3.6 mmol/l, p = 0.86
  5. There was no significant difference in the hypoglycaemic events between the soluble insulin treatment group and the insulin lispro treatment group: (lispro vs. soluble insulin) 9.7 vs. 8.0 per 30 days, p = 0.23
No description of how randomisation took place
Sponsored by Lilly
Some patients used CSII – can this be generalised to general population?
RCT crossoverIb

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.