Pediatric Diabetes 2006: 7 (Suppl. 4): 32–38 All rights reserved
#
2006 The Authors. Journal compilation # 2006 Blackwell Munksgaard
Pediatric Diabetes
Review Article
Hypoglycemia and ketoacidosis with insulin pump therapy in children and adolescents Hanas R, Ludvigsson J. Hypoglycemia and ketoacidosis with insulin pump therapy in children and adolescents. Pediatric Diabetes 2006: 7 (Suppl. 4): 32–38.
Abstract: This review deals with the two most serious side effects encountered with insulin pump therapy, severe hypoglycemia and diabetic ketoacidosis (DKA). Although clinical follow-up studies reported decreased rates of severe hypoglycemia, randomized studies have not confirmed this, showing no difference between the pump and injection groups. Less-severe hypoglycemia (mild/moderate/symptomatic hypoglycemia) was found to be more common with pump use. Some patients have inadvertently dosed or overdosed while awake or during sleep, causing fatal outcome in rare cases. Population-based or retrospective clinical studies reported a low rate of DKA in pump users that was still a higher rate than those using injection therapy, at least in some countries. In research settings and for patients with good compliance and adequate family support, the risk of DKA seems lower; many short-term studies report no DKA at all, possibly due to the increased attention given to participants. The use of continuous subcutaneous insulin infusion (CSII) seems to decrease the risk in patients who had recurrent DKA before pump start. Most episodes of DKA occur early after pump start, suggesting a learning curve occurs in all new forms of treatment. Increased teaching and awareness programs are vital to prevent severe hypoglycemia and DKA in children and adolescents using insulin pumps.
Good metabolic control can prevent vascular complications (1). The Linko¨ping Complication Study shows promising results in ‘real life’ with markedly decreased late complications in a non-selected patient population (2, 3). However, intensive treatment has been associated with increased risk of acute complications; a meta-analysis reported in 1997 (4) examined adverse events from intensified treatment in 14 trials of young and middle-aged persons with 1028 patients allocated to intensified [eight trials continuous subcutaneous insulin infusion (CSII), three trials multiple daily injections (MDIs), three trials both CSII and MDI] and 1039 allocated to conventional treatment (one to two injections/day). A total of 846 patients experienced at least one episode of severe hypoglycemia, 175 patients experienced diabetic ketoacidosis (DKA) and 26 patients died. The combined odds ratio [95% confidence interval (CI)] for hypoglycemia was 2.99 (2.45–3.64), for DKA 1.74 (1.27–2.38) and
32
Ragnar Hanasa and Johnny Ludvigssonb a Department of Pediatrics, Uddevalla Hospital; and bDivision of Pediatrics and Diabetes Research Center, Linko¨ping University Hospital, Sweden
Key words: adolescents – children – CSII – DKA – hypoglycemia – insulin pump – ketoacidosis – severe hypoglycemia Corresponding author: Ragnar Hanas, Uddevalla Hospital, S-451 80 Uddevalla, Sweden. Tel: þ46-522-92000; fax: þ 46-522-93149; e-mail:
[email protected]
for death from all causes 1.40 (0.65–3.01) when using intensified treatment. The DKA risk depended on the type of intensified treatment used. Odds ratios (95% CI) were 7.20 (2.95–17.58) for exclusive use of CSII, 1.13 (0.15–8.35) for MDI and 1.28 (0.90–1.83) for trials offering a choice between the two (p ¼ 0.004 for interaction). Mortality from acute metabolic causes was increased (five deaths from DKA and two sudden deaths that may have been caused by hypoglycemia); however, this was largely counterbalanced by a reduction in cardiovascular mortality (three deaths in intensified group, eight in the conventional group), which is not a consideration in children and adolescents. Serious hypoglycemia is often regarded as the limit for intensive treatment (1), even though there is not always a close correlation between severe hypoglycemia and low hemoglobin A1c (HbA1c) (5). For many clinicians, these experiences served as ‘once bitten, twice shy’ in that insulin pump treatment
Hypoglycemia and ketoacidosis using CSII in some countries like the United Kingdom and Denmark that were pioneers of this mode of therapy discontinued it, whereas for others that began later, improved technology helped to overcome these obstacles. In a Danish review from 2003, 20% of centers not using CSII reported that this was due to the risk of DKA and hypoglycemia with this form of treatment (6). Hypoglycemia and DKA are the major acute diabetes complications, which, beside other reasons, may limit the use of insulin pumps. There were some discouraging early experiences with pumps reported in the literature. In a study from 1984, Knight et al. (7) reported that 27 of 86 (32.6%) patients discontinued within 1 year, 23 doing so in the first 3 months of treatment. The reasons stated were pump too large (n ¼ 8), pump too limiting (n ¼ 6), could not cope (n ¼ 3), deterioration of control (n ¼ 3), excess hypoglycemia (n ¼ 3), abscess formation (n ¼ 3), and spouse disapproval (n ¼ 1). In a pediatric study from 1986, Brink et al. (8) from Boston reported that 7 of 24 (30%) patients discontinued. Side-effects were common: 25% experienced infections (abscess or nodule), 50% pump malfunction (mechanical or electrical), 25% battery problems, and 30% lipohypertrophy. They found no severe hypoglycemia and no change in the rate of DKA vs. the previous insulin regimen. Mecklenburg et al. (9) found 15.3 episodes of DKA/100 patient years in adults, pump malfunction (usually tubing problems) in 79%, and injection site infection rate of one episode/27 patient months. Teutsch et al. (10) reported 35 deaths among 3500 pump users, which was not different from non-pump users, but two deaths were caused by the pump. Steel (11) reported on an intrauterine death of a baby because of incorrect pump use in the mother. Thus, it is evident, that although insulin pump treatment has important advantages, serious acute complications have to be taken into account when prescribing such treatment to children.
Hypoglycemia Children with type 1 diabetes experience hypoglycemia with any type of insulin treatment (12). Taking into consideration the variations of insulin absorption and serum insulin profiles when using long-acting insulin, one would expect CSII to be a very useful tool to prevent hypoglycemia (13), and therefore, repeated severe hypoglycemia is, in fact, one of the clinical indications for use of pumps. However, to date, there is little solid evidence showing that CSII reduces hypoglycemia. Most studies reported are short-term studies of small groups of patients, usually non-randomized, and no or small differences are found regarding frequency of severe hypoglycemia. Pediatric Diabetes 2006: 7 (Suppl. 4): 32–38
Thus, reports from the United States, Germany, and France on non-randomized trials show a decrease in severe and moderate hypoglycemia, but still at a higher frequency than with MDI (14), an increase in severe hypoglycemia (15), and no difference regarding symptomatic hypoglycemia (16); and, three centers in the United States report reduced frequency of severe hypoglycemia in patients on pumps (17–21). However, because patients have themselves decided which regimen they prefer, or the comparison has been done before and after initiation of pump treatment, the scientific evidence is very weak. There are some randomized trials, mainly in very young children. The study by de Beaufort et al. (22) was done in newly diagnosed children during their first 2 years of diabetes, when residual insulin secretion contributes to less blood glucose fluctuations, and, consequently, only one severe case of hypoglycemia was observed in each group of 15 children on CSII or conventional insulin therapy. Three randomized studies on pre-school diabetic children have been reported, all showing no difference in severe hypoglycemia (23–25). Although DiMeglio et al. (24) found increased meter-detected hypoglycemia
