Page 70 - Glucose Monitoring Devices

68 CHAPTER 4 Consequences of SMBG systems inaccuracy

Cost compared to an ideal
Compared to an “ideal” BGM system, that is, one with no error and no bias, the
average additional cost of inaccuracy associated with the entire group of BGM
systems was £155 per patient year (PPY) (£95 to £219, depending on the popula-
tion). The average additional cost of not meeting the ISO 15197:2003 standard
[34] is approximately £306 PPY (£169 to £446), which is £178 PPY (£85 to
£270) more than the additional cost for the average ISO 15197:2003 compliant
BGM system. It is also £235 PPY (£113 to £357) more expensive than systems
compliant with the ISO 15197:2013 [34]. Complying with the ISO 15197:2003
but failing to comply with the ISO 15197:2013 results in an additional cost of
£216 PPY (£123 to £311). Splitting BGM systems according to 10/10 performance
shows that the bottom results in a higher additional cost of £79 compared to the
top-performing group (£64). The left panel of Fig. 4.12 illustrates these effects.
Worst-case costing
Alternative, and perhaps as a measure of risk, one can focus on the worst BGM
system in each group. The result illustrated in the right panel of Fig. 4.12, shows
the same relative effects. The overall worst system, that is, the most expensive not
compliant to ISO 15197:2003, has an additional cost of £597 compared to the ideal
system. Complying with ISO 15197:2003, but not with the 2013 version results in a
maximum of £440 in additional costs. The worst system complying with ISO 15197:
2013 has an additional cost of £278 compared to the ideal BGM system.

Lower Upper
50% 50%

11 11 Compliant ISO 15197:2013

7

14 Compliant Not

Not ISO 15197:2003
Compliant Compliant
FIGURE 4.11
Classification of 43 commercial BGM systems according to their compliance with
accuracy standards.