Original Title: Efectividad, seguridad y coste-efectividad de los sistemas de monitorización continua de glucosa intersticial en tiempo real (SMCG-TR) para la diabetes mellitus tipo 1 y 2

Authors' objectives: The aim of this work is to review the effectiveness, safety and cost-effectiveness of CGMS-RT, compared to traditional self-monitoring, in patients with type-1 and type-2 diabetes mellitus.

Authors' results and conclusions: RESULTS: EFFECTIVENESS AND SAFETY: • [TYPE-1 DIABETES MELLITUS] The meta-analysis on HbA1c reduction produced a significantly favorable effect of the intervention at 3 (WMD = -0.31%; 95% CI: -0.39, -0.23; Q = 15.6, p = 0.21, I2 = 23%) and 6 months (WMD = -0.23%; 95% CI: -0.35, -0.11; Q = 65.3, p <0.001; I2 = 75%) although in both cases the effect was under the value considered clinically significant. These results are similar to those obtained in previous systematic reviews identified in this work. We did not observed a moderating effect of age, HbA1c level baseline or method of insulin administration. Adherence to the device is significantly related to greater effect, but we could not perform MA due to the variability in the reporting of results. No significant differences were obtained in the rate of severe hypoglycemia, defined as those that require the intervention of another person. These results, and those related to the reduction of HbA1c, were not affected by the methodological quality of the studies. Two studies included patients with hypoglycemic unawareness (and consequently an increased risk of severe hypoglycemia); both found a significant effect in favor of the intervention, although the two studies showed an imbalance in baseline in this outcome variable that questions the appropriateness of randomization performed, and therefore their results should be interpreted with caution. Time spent on biochemically documented hypoglycemia was significantly lower for intervention groups in the studies that analyzed this variable as the primary outcome, but this result was not consistent in other studies. No significant differences in quality of life (assessed with different instruments) were observed in most studies. No significant differences in the rate of diabetic ketoacidosis (OR = 0.63, 95% CI: 0.29, 1.37; I2 = 0%) was found, and other adverse effects associated with the device were mild, mainly skin reactions at the site insertion. • [TYPE-2 DIABETES MELLITUS] Only five studies provided data on the effectiveness of CGMS-RT and the quality of the evidence is low for all the variables assessed. The reduction in HbA1c was significantly in favor of intervention (WMD = -0.48%; 95% CI: -0.79, -0.17; Q = 8.74, p = 0.07, I2 = 54%) and consistent across studies when the one with the the worst methodological quality was excluded. Data on severe hypoglycemia and quality of life were scarce. ECONOMIC EVALUATION: The SR of economic evaluations identified four articles conducting a cost-effectiveness analysis of the CGM system against SMBG using decision models. No articles focusing on the Spanish context and for DM2 patients were founded. Three economic evaluations concluded that CGM for DM1 patients was cost-effective from a social perspective in the US and Sweden with some uncertainty. A fourth study compared a new insulin pump that also incorporate the CGM system. This study also found that CGM system together with this new insulin pump was cost-effective from the perspective of the Australian health system. The results of our de novo analysis show that the CGM is a more effective alternative but much more expensive than the SMBG system for glucose monitoring in DM patients in Spain. CGM system's health gains in terms of QALY were larger than those achieved by the SMBG system, but not enough to offset the increased cost. The incremental cost-effectiveness ratio (ICER) obtained was 375,272 € / QALY for all DM patients. These figures are well above the 30,000 € / QALY willingness to pay (WTP), frequently mentioned in the Spanish literature. This leads us to conclude that the CGM system is not a cost-effective alternative for glucose monitoring in DM patients at that WTP. If we split patients by DM type, DM1 and DM2, the results are still not in favor of the new technology. The ICER for DM2 patients decreased slightly to 305,823 € / QALY, while the ICER for DM1 patients rised up to 2,554,723 € / QALY. Both ICER values were well above the 30,000 €/ QALY. These results show that the CGM system is not cost-effective for neither of the DM patient types in Spain. The deterministic and probabilistic sensitivity analysis vary the ICER estimate but still remain well above the 30,000 € / QALY WTP. CONCLUSIONS: EFFECTIVENESS AND SAFETY • [TYPE-1 DIABETES MELLITUS] - There is high quality evidence that CGMS-RT, compared to traditional monitoring, produced a statistically significant reduction in HbA1c levels in the short term (3-4 months). However, the effect is under the level considered to be clinically significant (-0.40%). At 6-7 months, the confidence level of evidence decreases to moderate, and the effect size decreases, but remains statistically significant. More research on the maintenance of the effect in the longer term is needed. - The results of this study found no evidence of a significant effect of age of the participants, their HbA1c at baseline or insulin delivery method used on the efficacy of the RT-CGMS, although limitations of available data prevent definitive conclusions about the moderating effect of these variables. - The analysis of individual studies and a prior MA with individual data shows that increased adherence to the CGMS-RT improved results. - There is low quality evidence that CGMS-RT does not reduce the frequency of severe hypoglycemia events. For patients at high risk of such events, the two studies included show a significant imbalance in the incidence rate at baseline, with more severe patients in the intervention group, and therefore the observed improvements represents a low quality evidence that must be interpreted with caution. - There is low quality evidence that CGMS-RT do not improve the patients’ health-related quality of life. - There is low quality evidence that CGMS-RT do not increase the occurrence of diabetic ketoacidosis events. The occurrence of other adverse effects associated with the device is reduced to a few cases of minor skin problems. • [TYPE-2 DIABETES MELLITUS] - There is low quality evidence that CGMS-RT, compared to traditional self-monitoring, produced a statistically significant reduction of HbA1c levels in the short term (2-3 months), but in the limit of clinical significance. - There is no evidence of other moderating factors. - There is low quality evidence that CGMS-RT do not reduce the frequency of severe hypoglycemic events, do not improve patients’ health-related quality of life and do not cause significant adverse effects. ECONOMIC EVALUATION: - The CGMS-RT system is not a cost-effective alternative for glucose monitoring in DM patients from the perspective of the Spanish NHS. This conclusion still holds when both DM types, DM1 and DM2, are analyzed.

Authors' recommendations: • [TYPE-1 DIABETES MELLITUS] - Based on the results of effectiveness and safety, the score using the GRADE system leads to a weak pro-CGMS-TR recommendation. However, from the economic and budgetary impact assessment the use of CGMS-TR in patients with DM1 it is not recommended. • [TYPE-2 DIABETES MELLITUS] - Due to the low quality of the evidence obtained and the results of the economic evaluation, the implementation of the CGMS-RT is not recommended in patients with DM2.

Authors' methods: EFFECTIVENESS AND SAFETY: Electronic databases Medline, PreMedline, Embase, Cochrane Central and Systematic Reviews, and Web of Science were consulted from 2010 to February 2015. Systematic reviews (SR) and primary studies were selected if they included patients of any age diagnosed with DM1 and/or DM2 in outpatient care, and analyzed the impact of CGMS-RT compared to traditional monitoring. For primary studies, additional criteria were established: a follow up of eight weeks or more, and that the method of insulin delivery was the same in either groups (or similar percentages for each method). From the SR identified, we selected the primary studies published before 2010. Effectiveness outcome variables were the reduction of HbA1c and severe hypoglycemia events, time spent in biochemically documented hypoglycemia and quality of life. Safety variables were the rate of diabetic ketoacidosis and other adverse effects. Risk of bias of the studies included was assessed with the Oxman’s scale for SR and by means of the Cochrane Collaboration criteria for primary studies. Global quality of evidence was assessed with the GRADE system, from which the strength (weak or strong, depending of the quality of evidence) and direction (favoring the evaluated technology or not) of recommendations are established. Where data were available, a quantitative synthesis of results was performed by means of a meta-analysis. The weighted (WMD) or standardized (DMS) mean difference for continuous variables and the odds ratio (OR) for dichotomous ones were calculated. Heterogeneity was calculated using the Cochran Q and Higgins I2 statistics. A significance value of Q = 0.10 or a I2 value greater than 30% were set to consider the existence of heterogeneity, in which case a random effects model was applied. To explore sources of heterogeneity subgroup analyzes were performed according to the following confounders established a priori: age of participants (children, adolescents or adults), HbA1c at baseline, method of insulin delivery (pump, multiple injections) and adherence to the device. Subgroup analysis was performed using the formulae QINT = QTOT - (Q1 + Q2 + ... + Qn) where QTOT represents the overall between- study variability and Qn represents the between-study variability for each subgroup in comparison. The statistical QINT is then compared with a χ2 distribution with J-1 degrees of freedom, where J is the number of subgroups. ECONOMIC EVALUATION: A SR of economic evaluations on the CGM system against SMBG system for monitoring glucose in patients with DM was firstly conducted. Search strategies applied were the same as the review of effectiveness and safety. In addition, a further search was made in the NHS EED database of the University of York CRD through the Cochrane Library without date restrictions. Full economic evaluations were included, i.e., those in which both costs and results from at least two alternatives were compared. Both economic models and economic evaluations conducted alongside clinical trials were included. We conducted a full economic evaluation in which health outcomes and costs from two alternative for glucose monitoring in DM patients, CGM against SMBG, were compared. A cost-effectiveness analysis (CEA) considering the life-years gained (LYG) as health outcome, and a cost-utility analysis (CUA) considering quality adjusted life years (QALY), were conducted. Both analysis were based on a Markov model from the perspective of the NHS for a lifetime time horizon using annual cycles. The DM progression model was based on McQueen et al. design. Additionally, deterministic and probabilistic sensitivity analysis were performed to show the model's robustness and the uncertainty on parameters and results.

Project Status: Completed

Year Published: 2015

URL for published report: https://sescs.es/efectividad-de-los-smcg-tr-para-la-diabetes-mellitus-tipo-1-y-2/

English language abstract: An English language summary is available

Publication Type: Full HTA

Country: Spain

Organisation Name: Canary Health Service

Contact Address: Dirección del Servicio. Servicio Canario de la Salud, Camino Candelaria 44, 1ª planta, 38109 El Rosario, Santa Cruz de Tenerife

Contact Name: sescs@sescs.es

Contact Email: sescs@sescs.es