Applied Health Economics and Health Policy

, Volume 9, Issue 6, pp 351–365

Cost effectiveness of self-monitoring of blood glucose (SMBG) for patients with type 2 diabetes and not on insulin

Impact of modelling assumptions on recent Canadian findings
Original Research Article

DOI: 10.2165/11594270-000000000-00000

Cite this article as:
Tunis, S.L. Appl Health Econ Health Policy (2011) 9: 351. doi:10.2165/11594270-000000000-00000



Canadian patients, healthcare providers and payers share interest in assessing the value of self-monitoring of blood glucose (SMBG) for individuals with type 2 diabetes but not on insulin. Using the UKPDS (UK Prospective Diabetes Study) model, the Canadian Optimal Prescribing and Utilization Service (COMPUS) conducted an SMBG cost-effectiveness analysis. Based on the results, COMPUS does not recommend routine strip use for most adults with type 2 diabetes who are not on insulin. Cost-effectiveness studies require many assumptions regarding cohort, clinical effect, complication costs, etc. The COMPUS evaluation included several conservative assumptions that negatively impacted SMBG cost effectiveness.


Current objectives were to (i) review key, impactful COMPUS assumptions; (ii) illustrate how alternative inputs can lead to more favourable results for SMBG cost effectiveness; and (iii) provide recommendations for assessing its long-term value.


A summary of COMPUS methods and results was followed by a review of assumptions (for trial-based glycosylated haemoglobin [HbA1c] effect, patient characteristics, costs, simulation pathway) and their potential impact. The UKPDS model was used for a 40-year cost-effectiveness analysis of SMBG (1.29 strips per day) versus no SMBG in the Canadian payer setting. COMPUS assumptions for patient characteristics (e.g. HbA1c 8.4%), SMBG HbA1c advantage (−0.25%) and costs were retained. As with the COMPUS analysis, UKPDS HbA1c decay curves were incorporated into SMBG and no-SMBG pathways. An important difference was that SMBG HbA1c benefits in the current study could extend beyond the initial simulation period. Sensitivity analyses examined SMBG HbA1c advantage, adherence, complication history and cost inputs. Outcomes (discounted at 5%) included QALYs, complication rates, total costs (year 2008 values) and incremental cost-effectiveness ratios (ICERs).


The base-case ICER was $Can63 664 per QALY gained; approximately 56% of the COMPUS base-case ICER. SMBG was associated with modest risk reductions (0.10–0.70%) for six of seven complications. Assuming an SMBG advantage of −0.30% decreased the current base-case ICER by over $Can10 000 per QALY gained. With adherence of 66% and 87%, ICERs were (respectively) $Can39231 and $Can54349 per QALY gained. Incorporating a more representative complication history and 15% complication cost increase resulted in an ICER of $Can49 743 per QALY gained.


These results underscore the importance of modelling assumptions regarding the duration of HbA1c effect. The current study shares several COMPUS limitations relating to the UKPDS model being designed for newly diagnosed patients, and to randomized controlled trial monitoring rates. Neither study explicitly examined the impact of varying the duration of initial HbA1c effects, or of medication or other treatment changes. Because the COMPUS research will potentially influence clinical practice and reimbursement policy in Canada, understanding the impact of assumptions on cost-effectiveness results seems especially important. Demonstrating that COMPUS ICERs were greatly reduced through variations in a small number of inputs may encourage additional clinical research designed to measure SMBG effects within the context of optimal disease management. It may also encourage additional economic evaluations that incorporate lessons learned and best practices for assessing the overall value of SMBG for type 2 diabetes in insulin-naive patients.

Supplementary material

40258_2012_90603511_MOESM1_ESM.pdf (200 kb)
Supplementary material, approximately 205 KB.

Copyright information

© Adis Data Information BV 2011

Authors and Affiliations

  1. 1.Independent Health Economics Research ConsultantIndianapolisUSA
  2. 2.IndianapolisUSA

Personalised recommendations