Insulin Pump Therapy: The Pros and Cons in Current Diabetes Management

Abstract

Continuous Subcutaneous Insulin Infusion (CSII), insulin pump therapy, is primarily recommended in the management of type 1 diabetes mellitus, whilst efficacy in type 2 diabetes has yet to be proven.  CSII facilitates flexible, precise and physiological basal and bolus insulin delivery via an external pump containing rapid-acting insulin.  Patients using CSII have shown favourable improvements in glycaemic control (mean decrease of 0.44% in glycated haemoglobin (HbA_1c) levels) and lower insulin requirements; however, effects on weight remain inconsistent.  Improvements in quality of life and enhanced treatment satisfaction have also been reported.  Although CSII is demonstrated to exhibit a range of attractive advantages, insulin pump therapy has yet to reach perfection.  CSII regimens require a lifetime commitment to intensive education, specialist care, frequent monitoring, in addition to the physical burdens on everyday life.  Insulin pumps present a greater risk of site infection compared with standard subcutaneous insulin injection.  Increased technical risks of insulin delivery failure are reported, which may result in hyperglycaemia and fatal ketosis in extreme cases.  Furthermore, the annual financial cost of CSII treatment is estimated to be 80% more than insulin injections.  The future of insulin pump therapy strives to provide a closed-loop system which may involve an implantable pump combined with continuous glucose monitoring – the ‘artificial beta-cell’.

Introduction

Insulin pump therapy, more accurately described as Continuous Subcutaneous Insulin Infusion (CSII) is a relatively novel method of insulin delivery.  First trialled in the 1970s as the ‘big blue brick’, the modern day open-loop pump involves a fine subcutaneous cannula connected via flexible tubing to a small external battery-operated device containing an insulin reservoir.  The cannula is typically situated in the abdomen and the pump worn around the waist [1,2].  Alternative infusion sites may include the buttocks, hips, thighs, or lateral aspect of the biceps [3].

As technology advances, the latest pump devices exhibit numerous additional functions including full memory of insulin doses, timings and totals; customised infusion programmes according to daily activity; databases of food carbohydrate content to facilitate patients with the calculation of insulin requirement; software to  download recordings; compatibility with blood glucose monitors so that the carbohydrate and insulin inputs can be compared with blood glucose level outcomes; and warnings when the insulin cartridge levels or batteries are low [4].  All of these features improve the ease of use and facilitate diabetes management; however, may also pose a technological challenge to some patients.

CSII aims to mimic the physiological non-diabetic insulin secretion pattern more closely using only monomeric insulin (rapid-acting insulin analogues such as aspart or lispro), in comparison with standard Multiple Daily Injection (MDI) regimens, which require a combination of both long and short-acting insulins (Figure 1a).  The CSII regimen combines a programmed background basal rate with bolus doses of insulin regulated by the patient according to their eating patterns (Figure 1b) [1,2,5].  Patients using CSII are taught to calculate their required mealtime dose according to carbohydrate intake and anticipated level of physical activity, a principle similar to MDI bolus doses of short-acting insulin.  CSII, however,  further enables delivery of an ‘extended bolus’ for meals that are metabolised more slowly, and so present a sustained increase in blood glucose levels, for example high fat and protein meals [6].

Both the advantages and disadvantages of using CSII as the chosen method of insulin delivery, thus replacing MDI should be assessed in relation to various diabetic patient groups.  The primary focus is applications of CSII in the management of type 1 diabetes mellitus (T1DM) in whom there is a complete lack of endogenous insulin.  In contrast, CSII has not proven to be superior in maintaining glycaemic control when compared with other pre-existing treatments for type 2 diabetes mellitus (T2DM) [8,9,10]; hence, the National Institute of Clinical Excellence (NICE) does not recommend CSII for this patient group [8].

The current NICE recommendations for CSII is in management of adults and children over 12 years-old with T1DM who experience “disabling hypoglycaemia” (defined as “repeated and unpredictable occurrence of hypoglycaemia that results in persistent anxiety about recurrence and is associated with a significant adverse effect on quality of life”) whilst using MDI regimens; and alternatively for cases in which glycaemic control (defined as HbA1c <8.5%) cannot be achieved despite intensive MDI therapy [8].  CSII is additionally recommended in management of children under 12 years-old with T1DM if MDI therapy is considered as “impractical or inappropriate”; however, these patients would be expected to trial MDI during their adolescence [8].

Advantages of CSII Therapy

Hypoglycaemia

Hypoglycaemia is a known adverse effect of insulin therapy and occurs to varying degrees of severity and frequencies; however, as CSII offers greater precision and flexibility in insulin delivery, incidence and risk of extreme hypoglycaemic episodes is reported to be substantially reduced [2,11,12].

A comprehensive meta-analysis spanning literature published over the 10 years preceding 2006 by Pickup & Sutton [13] summarised findings that CSII significantly reduces the incidence of “severe hypoglycaemia” (defined as “requiring third-party assistance including unconsciousness, seizure, glucagon administration and emergency attendance or admission to hospital“) by an average rate ratio of 4.19, in comparison to MDI.  The greatest reductions were observed in those with the highest initial frequency of hypoglycaemia [13].

Furthermore, CSII in elderly (>60 years) patients with T2DM versus MDI, although statistically insignificant, indicated a trend of reduced hypoglycaemia incidence in the CSII group.  More MDI patients experienced severe hypoglycaemia in comparison to those using CSII [14].  Similarly, outcomes of CSII therapy in the paediatric age group also concur to a reduction in severe hypoglycaemia when using CSII, reaching a threefold decrease in one particular study [15]; yet, in another this significant reduction was not transposed in randomized controlled trials which have more power [7].

In a minority of studies, the incidence of hypoglycaemia with CSII was not significantly different to MDI regimens [14,16-19], also shown in a review of pregnant women [20].  However, incidence of hypoglycaemia was categorised as descriptive data and there were differences in recording events [17].

Blood Glucose

CSII therapy is associated with reduced blood glucose levels and glycaemic variability.  Meta-analysis of 21 studies found a significant reduction in average blood glucose levels of approximately 3.2mmol/L (9.8mmol/L pre-CSII to 6.5mmol/L post-CSII, converted from a change from 176.23mg/dL to 117.82 mg/dL, P<0.001) following commencement of CSII treatment [17].  A 4-month randomized crossover trial comparing CSII of lispro and MDI found not only lower blood glucose levels in the CSII group, but also less variability demonstrated by a tighter standard deviation (73 ±15 vs. 82 ±18 mg/dl, P<0.01) [16].  Similarly, daily blood glucose variability was significantly less in patients using CSII compared with those on MDI regimens, a finding which also correlated with the reduction in HbA1c observed [21].  Furthermore, overnight CSII combined with oral agents in the treatment of T2DM resulted in attenuation of fasting plasma glucose in those unable to achieve adequate glycaemic control on oral treatments alone [22].  In contrast, children and adolescents were found to show no significant difference between mean blood glucose and glycaemic stability when treated with CSII and MDI [19].

Development of superior long-acting soluble insulin analogues, such as glargine used as 24hr basal insulin in MDI could be a potential replacement of CSII as the “gold standard” of basal insulin delivery [12,23].  However, programmable CSII basal rates can account for circadian rhythms (Figure 1b), including the ‘Dawn Phenomenon’ during which blood glucose levels rise in the early morning due to fasting, consequently increasing the insulin requirement [24], and similarly during the ‘Dusk Phenomenon’ also.  CSII insulin infusions can also be suspended during exercise, during which blood glucose is anticipated to decline rapidly, again reducing the risk of hypoglycaemia.  In comparison, glargine release from subcutaneous stores cannot be modulated or suspended (Figure 1a), therefore indicating the superiority of CSII relative to MDI to mirror physiological insulin requirements.  Additionally, the precision of CSII is far greater, with a consistent minimum basal rate of only 0.025U/hr and mealtime boluses of 0.05U increments [6], thus facilitating tighter glycaemic control.

Glycated Haemoglobin (HbA1c)

A reduction in HbA1c as a measure of long-term glycaemic control is expected following the reported reductions in glycaemic variability and mean blood glucose.  Results of the large-scale Diabetes Control and Complications Trial (DCCT) indicated a 0.5% reduction in HbA1c in the intensive therapy group (which included CSII) [25].  More recently, meta-analysis by Pickup et al. reported a comparable standardized mean reduction in HbA1c of 0.44%. [26], and Nuboer et al., a reduction of 0.22% in their paediatric study [15] in comparison with MDI.

Weissberg-Benchell et al. revealed a 0.4% reduction in HbA1c levels in their meta-analysis; however, when subcategorised according to duration of CSII, using CSII for <1year duration were insignificant in effects on HbA1c, whereas those of CSII therapy >1year demonstrated a strong mean reduction in HbA1c of 1.2% (8.68 pre-CSII to 7.48 post-CSII, P<0.001) [17].  A general HbA1c reduction is similarly demonstrated in T2DM following CSII treatment, in a randomized cross-over trial of obese patients, CSII produced a reduction in HbA1c of 0.8±1.5% versus MDI therapy, which actually increased HbA1c by 0.4±1.39% [27].  In summary, CSII therapy is associated with a reduction in HbA1c, however evidence in short treatment durations and in T2DM is weaker.

Long-Term Complications

Tighter glycaemic control and maintenance of near normoglycaemia following CSII is associated with clinical benefits of a reduction in diabetes related long-term complications. The cardiovascular branch of the DCCT found that in the sample of T1DM patients, those using intensive therapy presented reduced risk of cardiovascular disease [28].  Similarly, a promising reduction in microvascular complications (retinopathy, nephropathy and neuropathy) were also demonstrated [25]; later reinforced by Pickup et al concluding a reduction in risk of retinopathy by about 25%, however, after extension to a 10-year expected outcome, the reduction was less marked [26].  Additionally, early worsening of retinopathy was observed in the intensive therapy arm of the DCCT due to the rapid improvements in glycaemic control; however, long-term benefits outweighed this risk [29].

Insulin Requirement & Weight

Numerous trials of CSII therapy demonstrated a reduction in the amount of insulin required to maintain glycaemic control [11,16,26,30]; a short-term average of 12 units less insulin (-11.90, 95% CI -18.16 to -5.63), but this reduction was less in the long-term [18].  Additionally, CSII was concluded to significantly reduce both the number of units per day (by ~10 units) as well as units per kilogram per day (0.12 units) when patients were using CSII instead of insulin injections [17].  A similar trend was also found in trials involving paediatric T1DM cases [15,31], with an average of 0.27 U/kg/day less insulin [15].

Weight gain is a concerning side-effect of insulin therapy [5], therefore the reduction in insulin required to maintain near normoglycaemia in CSII patients may correspond to less weight gain.  In the case of paediatric studies, a significant reduction in body mass index of approximately 0.6 was observed during the third and fourth years of a 4-year follow up [31], whereas another found no significant change [11].  Trials of CSII and MDI therapy in obese patients with T2DM also resulted in no significant weight change and a short-term reduction in insulin requirement [27].  However, unfortunately, meta-analysis of adult T1DM patients revealed that CSII has a significant effect on increasing mean body weight from 68.24kg before commencing CSII to 71.21kg after [17]; thus, concluding inconsistent effects of CSII therapy on body mass.

Quality of Life (QoL) & Patient Satisfaction

CSII is reported to significantly improve a range of patient quality of life (QoL) parameters, such as overall treatment satisfaction and lifestyle restrictions [32]; life expectancy (by a mean of 0.76±0.19 years as a quality-adjusted life expectancy) [33]; and increase satisfaction [30,34] in comparison with conventional injection regimens.  Upon assessment of children using CSII and their parents, a significant improvement in the paediatric QoL (PaedsQL) scores and decrease in impact of disease scores were also found [15].  Other studies noted that CSII produces less pain, burdens and interference to daily life when compared to MDI, thus resulting in greater convenience, flexibility and greatly increased patient preference [33,35].

The choice of insulin for infusion also affects perceived outcomes; patient feedback comparing two insulin analogues revealed greater comfort (Figure 2) [35], and overall preference and treatment satisfaction [36] for insulin aspart in comparison with lispro.

Disadvantages of CSII Therapy

Social & Psychological Implications

The initial commencement of CSII requires significant motivation and commitment to intensive education to adapt to a new regimen of diabetes therapy.  Greater input from healthcare professionals is required in management of CSII therapy in comparison with injection therapies; therefore a specific diabetes specialist nurse or diabetes educator plays an essential role in advising and educating patients and their carers.  Additionally, frequent self-monitoring of blood glucose levels, at least 4 times daily and using the result to adjust or calculate the insulin dose required is fundamental to ensure successful diabetes management [5,37].

Diabetes as a life-long disease itself poses various social and psychological concerns for patients, and the impact of pump therapy can often raise further issues.  Although technology has majorly advanced since the bulky pumps of the 1970s, the current pumps are the size and weight of a pager which needs to be closely strapped to the body and carried around at all times, as well as the prospect of a permanent cannula in the infusion site.  Some patients may find this socially uncomfortable, inconvenient, and the pump may interfere with their everyday activities, such as intensive sport during which the pump may be damaged or the cannula may dislodge, or swimming which exposes the device to water.  Consequently, CSII may not be a suitable therapy for all patients.

Technical Risks – Diabetic Ketoacidosis (DKA)

A range of technical problems may arise relating to CSII therapy, these may include pump malfunction due to battery failure or an empty insulin reservoir; or leakage or disruption to the cannula or infusion tube, all of which will prevent insulin delivery [38].  The most detrimental health implications arise when patients are unaware of these interruptions in their insulin delivery, thus leading to hyperglycaemia, which ultimately increases the risk of long-term complications and acute DKA.

Extreme hyperglycaemia, resulting in DKA is postulated to occur more frequently in patients using CSII in comparison to standard injected insulin regimens [32].  A key physiological reason for this observation is because patients lack a subcutaneous store of insulin as CSII only administers rapid-acting insulin; hence, in event of pump failure or interruption of insulin infusion, blood glucose levels will rise unchecked, consequently resulting in DKA, which in extreme cases may prove fatal.  This emphasises the importance of education, as well as frequent blood glucose monitoring to ensure correct insulin delivery [39].  Warning alarms which feature in latest pump devices may help patients to avoid hyperglycaemia, facilitate self-management and aid problem solving.

High rates of DKA were noted in the 1980s [38] and an increased frequency in research published prior to 1993 [17].  Other studies, however, did not identify an increased risk of ketoacidosis with the use of CSII [26,40], which the authors proposed explanations of a short study duration and possible poor reporting.  Similarly, a large matched-pair study revealed lower DKA rates in children using CSII over all stages of the 3-year trial when compared with MDI [11].

Infection Risk

As with all continuous insults to the skin and underlying subcutaneous tissue, infusion sites are prone to infection with common epidermal bacteria and lipohypertrophy.  Patients are recommended to change the cannula every 2-3days [41] and taught hygienic methods using sterile equipment to prevent infection with Staphylococcal bacteria [2].  To avoid lipohypertrophy, patients are advised to rotate infusion sites and relocate the cannula at least 2 inches away from the previous site every time it is changed [3].

Cost

Cost-effectiveness of CSII in comparison to conventional insulin injections is a key issue to be addressed.  The initial cost of a pump device is approximately £1000-2000, and additional infusion sets and other disposables amount to ~£1000 per year [5,18].  The greater healthcare professional input required also majorly increases costs of CSII therapy.  Employing 2003 value cost estimates, the total annual cost per patient of CSII is approximately 80% greater in comparison with MDI therapy (£2641 vs. £1482), with total lifetime costs accruing a difference nearing £20,000 [33].

Although the financial cost of CSII treatment is much greater compared with MDI; it has been justified by NICE to be balanced by greater reduction in long-term costs of management of diabetes related complications.  This is indicated by an incremental cost-effectiveness ratio (ICER) of £25,648 per quality-adjusted life year (QALY) gained, hence concluding that CSII is a cost-effective treatment choice compared with MDI as the threshold for NICE regulations is £30,000 for a new treatment to be deemed cost-effective [18].  However, in other countries such as Europe and the US, the financial burden of CSII poses a barrier although it has potential to fulfil the treatment goals [42].

Conclusions & the Future

The overall advantages and disadvantages of CSII therapy in the management of both type 1 and type 2 diabetes can be summarised in Table 1 below:

The majority of findings presented provide adequate significant evidence suggesting that CSII can be deemed an effective, safe and financially viable method of insulin delivery for use in the management of T1DM in all age groups.  It would also appear reasonable to conclude that CSII has the potential to supersede MDI regimens as CSII is flexible and compatible with insulin requirements of a range of lifestyles.  However, individual differences in patients’ attitudes to pump therapy and their own diabetes management may influence suitability and reliability of treatment; as commitment, training and self-management are essential factors for successful outcomes, thus limiting CSII to a select patient group.

Research into CSII therapy in children and adults with T1DM is generally comprehensive, whereas trials of CSII for T2DM and during pregnancy are relatively sparse and less conclusive.  Meta-analyses provide a strong evidence base; however, additional limitations of research include short study durations; small sample sizes; variations in reporting and outcome definitions, incidence of hypoglycaemic or ketoacidotic events and severity being one example; and finally, the employment of descriptive outcome measures, for example questionnaire feedback used to assess quality of life, which additionally produces selection bias, therefore indicating a requirement for further studies of CSII outcomes.

The future of CSII pump therapy encompasses pioneering of an ‘artificial beta-cell’ (Figure 4); the combination of an implanted insulin pump and intravenous glucose monitor, which facilitates closed-loop diabetes management [2,43,44].  This system incorporates algorithms of glucose feedback control to regulate direct peritoneal insulin delivery, hence requiring minimal patient input.  However, at present, the main challenges are ensuring efficiency of glucose monitoring and addressing the issue of feedback delay [43,44]; therefore, development of fully automated insulin delivery systems remains an area of prime research focus.

References

1. Williams G, Pickup J, Handbook of Diabetes. Oxford: Blackwell Science 2004 3^rd edition: pp103-105.
2. Pickup J, ‘Alternative forms of insulin delivery‘, in Pickup J, Williams G, Textbook of Diabetes. Oxford: Blackwell Science 2002 3^rd edition, Volume 2.
3. Rice D, Sweeny K, Choosing and Using an Insulin Pump Infusion Set [online], updated 17 March 2008. Available: http://www.diabetesselfmanagement.com/articles/Insulin/Choosing_and_Using_an_Insulin_Pump_Infusion_Set/2 (accessed 12/10/08).
4. Animas^TM, Animas^TM 2020 Pump Specifications [online], updated 15 July 2008. Available: http://www.animascorp.co.uk/2020-specifications.aspx (accessed 06/10/08).
5. Watkins P, Amiel S, Howell S, Turner E, Diabetes and its Management. Oxford: Blackwell Science 2003; 6^th edition: pp66-78.
6. Animas^TM, The Animas^TM 2020 Insulin Pump [online], updated 15 July 2008. Available: http://www.animascorp.co.uk/2020-diabetes-insulin-pumps.aspx (accessed 06/10/08).
7. Shalitin S, Phillip M. The Use of Insulin Pump Therapy in the Pediatric Age Group. Hormone Research 2008; 70: 14-21.
8. NHS NICE. Continuous subcutaneous insulin infusion for the treatment of diabetes mellitus, Review of technology appraisal guidance 57, [online] published July 2008.  Available: http://www.nice.org.uk/Guidance/TA151 (accessed 08/10/08).
9. Jeitler K, Horvath K, Berghold A, Gratzer TW, Neeser K, Pieber TR, Siebenhofer A. Continuous subcutaneous insulin infusion versus multiple daily insulin injections in patients with diabetes mellitus: systematic review and meta-analysis. Diabetologia 2008; 51: 941-951.
10. Raskin P, Bode BW, Marks JB, Hirsch IB, Weinstein RL, McGill JB, Peterson GE, Mudaliar SR, Reinhardt RR. Continuous subcutaneous insulin infusion and multiple daily injection therapy are equally effective in type 2 diabetes: a randomized, parallel-group, 24-week study. Diabetes Care 2003; 26(9): 2598-603.
11. Jakisch BI, Wagner VM, Heidtmann B, Lepler R, Holterhus P-M, Kapellen TM, Vogel C, Rosenbauer J, Holl RW. Treatment Comparison of continuous subcutaneous insulin infusion (CSII) and multiple daily injections (MDI) in paediatric Type 1 diabetes: a multicentre matched-pair cohort analysis over 3 years. Diabetic Medicine 2008; 25: 80-85.
12. Pickup J, Keen H. Continuous Subcutaneous Insulin Infusion at 25 Years. Diabetes Care 2002; 25: 593-598.
13. Pickup JC, Sutton AJ. Severe hypoglycaemia and glycaemic control in Type 1 diabetes: meta-analysis of multiple daily insulin injections compared with continuous subcutaneous insulin infusion. Diabetic Medicine 2008; 25(7): 765-74.
14. Herman WH et al. A Clinical Trial of Continuous Subcutaneous Insulin Infusion Versus Multiple Daily Injections in Older Adults with Type 2 Diabetes. Diabetes Care 2005; 28: 1568-1573.
15. Nuboer R, Borsboom GJJM, Zoethout JA, Koot HM, Bruining J. Effects of insulin pump vs. injection treatment on quality of life and impact of disease in children with type 1 diabetes mellitus in a randomized, prospective comparison. Pediatric Diabetes 2008: 9(Part I): 291-296.
16. Hanaire-Broutin H, Melki V, Bessières-Lacombe S, Tauber JP. Comparison of Continuous Subcutaneous Insulin Infusion and Multiple Daily Injection Regimens Using Insulin Lispro in Type 1 Diabetic Patients on Intensified Treatment. Diabetes Care 2000; 23: 1232-1235.
17. Weissberg-Benchell J, Antisdel-Lomaglio J, Seshadri R. Insulin pump therapy: a meta-analysis. Diabetes Care 2003; 26: 1079-1087.
18. Colquitt JL, Green C, Sidhu MK, Hartwell D, Waugh N. Clinical and cost-effectiveness of continuous subcutaneous insulin infusion for diabetes. Health Technology Assessment 2004; 8(43).
19. Alemzadeh R, Palma-Sisto P, Parton EA, Holzum MK. Continuous Subcutaneous Insulin Infusion and Multiple Dose of Insulin Regimen Display Similar Patterns of Blood Glucose Excursions in Pediatric Type 1 Diabetes Diabetes Technology & Therapeutics 2005; 7(4): 587-96.
20. Farrar D, Tuffnell DJ, West J. Continuous subcutaneous insulin infusion versus multiple daily injections of insulin for pregnant women with diabetes. Cochrane Database of Systmatic Reviews 2007; 18(3): CD005542.
21. Pickup JC, Kidd J, Burmiston S, Yemane N. Determinants of glycaemic control in type 1diabetes during intensified therapy with multiple daily insulin injections or continuous  subcutaneous insulin infusion: importance of blood glucose variability. Diabetes/Metabolism Research Reviews 2006; 22: 232-237.
22. Parkner T, Møller MK, Chen JW, Laursen T, Jørgensen C, Smedegaard JS, Lauritzen T, Christiansen JS. Overnight CSII as supplement to oral antidiabetic drugs in type 2 diabetes. Diabetes, Obesity and Metabolism 2008;10(7): 556-63.
23. Lepore M, Pampanelli S, Fanelli C, Porcellati F, Bartocci L, Di Vincenzo A, Cordoni C, Costa E, Brunetti P, Bolli GB. Pharmacokinetics and pharmacodynamics of subcutaneous injection of long-acting human insulin analog glargine, NPH insulin, and ultralente human insulin and continuous subcutaneous infusion of insulin lispro. Diabetes 2000; 49: 2142-2148.
24. Radermecker RP, Scheen AJ. Continuous subcutaneous insulin infusion with short-acting insulin analogues or human regular insulin: efficacy, safety, quality of life, and cost-effectiveness. Diabetes/Metabolism Research and Reviews 2004; 20: 178-188.
25. The Diabetes Control and Complications Trial Research Group. The Effect of Intensive Treatment of Diabetes on the Development and Progression of Long-Term Complications in Insulin-Dependent Diabetes Mellitus. The New England Journal of Medicine 1993; 329: 977-986.
26. Pickup J, Mattock M, Kerry S. Glycaemic control with continuous subcutaneous insulin infusion compared with intensive insulin injections in patients with type 1 diabetes: meta-analysis of randomised controlled trials. British Medical Journal 2002; 324; 705.
27. Wainstein J, Metzger M, Boaz M, Minuchin O, Cohen Y, Yaffe A, Yerushalmy Y, Raz I, Harman-Boehm I. Insulin pump therapy vs. multiple daily injections in obese Type 2 diabetic patients. Diabetic Medicine 2005; 22(8): 1037-46.
28. The Diabetes Control and Complications Trial/Epidemiology of Diabetes Interventions and Complications (DCCT/EDIC) Study Research Group. Intensive Diabetes Treatment and Cardiovascular Disease in Patients with Type 1 Diabetes. The New England Journal of Medicine 2005; 353: 2643-53.
29. The Diabetes Control and Complications Trial Research Group. Early Worsening of Diabetic Retinopathy in the Diabetes Control and Complications Trial. Archives of Ophthalmology 1998; 116: 874-886. [Ret1]
30. Jeitler K, Horvath K, Berghold A, Gratzer TW, Neeser K, Pieber TR, Siebenhofer A. Continuous subcutaneous insulin infusion versus multiple daily insulin injections in patients with diabetes mellitus: systematic review and meta-analysis. Diabetologia 2008; 51(6): 941-951.
31. Sulli N, Shashaj B. Long-term benefits of continuous subcutaneous insulin infusion in children with Type 1 diabetes: a 4-year follow-up. Diabetic Medicine 2006; 23: 900-906.
32. Linkeschova R, Raoul M, Bott U, Berger M, Spraul M. Less severe hypoglycaemia, better metabolic control, and improved quality of life in Type 1 diabetes mellitus with continuous  subcutaneous insulin infusion (CSII) therapy; an observational study of 100 consecutive patients followed for a mean of 2 years. Diabetic Medicine 2002; 19: 746-751.
33. Roze S, Valentine WJ, Zakrzewska KE, Palmer AJ. Health-economic comparison of continuous subcutaneous insulin infusion with multiple daily injection for the treatment of Type 1 diabetes in the UK. Diabetic Medicine 2005; 22(9): 1239-45.
34. Hammond P, Liebl A, Grunderc S. International survey of insulin pump users: Impact of continuous subcutaneous insulin infusion therapy on glucose control and quality of life. Primary Care Diabetes 2007; 1(3): 143-146.
35. Bode BW. Use of Rapid-Acting Insulin Analogues in the Treatment of Patients with Type 1 and Type 2 Diabetes Mellitus: Insulin Pump Therapy Versus Multiple Daily Injections. Clinical Therapeutics 2007; 29 Suppl D: S135-44.
36. Wittlin SD, Marcus AO, Weng CS, Howard CP, Schorr AB.  Evaluation of Treatment Satisfaction Associated with the Use of Insulin Aspart in Continuous Subcutaneous Insulin Infusion.  Diabetes Technology & Therapeutics 2008; 10(1): 1-10.
37. Wittlin SD. Treating the Spectrum of Type 2 Diabetes: Emphasis on Insulin Pump Therapy. The Diabetes Educator 2006; 32: 39S.
38. Guilhem I, Leguerrier AM, Lecordier F, Poirier JY, Maugendre D. Technical risks with subcutaneous insulin infusion. Diabetes & Metabolism 2006; 32: 279-284.
39. Ponder SW, Skyler JS, Kruger DF, Matheson D, Brown BW. Unexplained Hyperglycemia in Continuous Subcutaneous Insulin Infusion: Evaluation and Treatment. The Diabetes Educator 2008; 34: 327-333.
40. Doyle EA, Weinzimer SA, Steffen AT, Ahern JA, Vincent M, Tamborlane WV. A randomised, prospective trial comparing the efficacy of continuous subcutaneous insulin infusion with multiple daily injections using insulin glargine. Diabetes Care 2004; 27: 1554-1558.
41. BMJ Publishing Group. BMJ Best Treatments: Insulin Pumps [online], last amended 03/09/08.  Available: http://besttreatments.bmj.com/btuk/conditions/1000277108.html (accessed 06/10/08).
42. Danne T et al. Establishing glycaemic control with continuous subcutaneous insulin infusion in children and adolescents with type 1 diabetes: experience of the PedPump Study in 17 countries. Diabetologia 2008; 51: 1594-1601.
43. Hovorka R. Continuous glucose monitoring and closed-loop systems. Diabetic Medicine 2006; 23(1): 1-12.
44. Renard E. Clinical Experience with an Implanted Closed-Loop Insulin Delivery System. Arquivos brasileiros de endocrinologia e metabologia 2008; 52(2): 349-354.