Recent articles (1–3) have emphasized the need for real-time assessment of blood glucose levels. Real-time analysis is motivated by the belief that the random time series being analyzed manifests average properties that change with the passage of time. Results of such analyses are presented to patients as means (μ) ± SD. Farmer et al. (3) report that their statistical analysis is carried out using SPSS version 11; Garg et al. (1) and Deiss et al. (2) used other analysis methods.
A search for SPSS packages on the Internet reveals that there are several versions currently available, and all estimate means and SDs of time series data. Since none, however, make an “issue” of how they do so, it's not unreasonable to assume that all versions estimate means and SDs in the “traditional” way, i.e., as Eq. 1 illustrates:
where G(n) is the time series of individual glucose readings.
The unfortunate feature of Eq. 1 is that the μ and SD values generated are valid only when the G(n) is independent (4). Specifically, independence stipulates that the magnitude of any particular G(n) is not influenced by previous value(s). Time series data don’t meet that criterion. Liang and Zeger (5) are very specific on this, stating, “with repeated observations, the correlation among values for a given subject must be taken into account.” Harris and Boyd (6) and Matthews and Farewell (7) are equally explicit.
In short, averages of blood glucose generated by Eq. 1 produce unreliable values. This contradicts the conclusion of Garg et al. (1) “that [traditional] real-time continuous glucose monitoring for periods up to 72 h is accurate.” It also conflicts with findings of Deiss et al. (2) “that the use of [traditional] real-time CGM [continuous glucose monitoring] has considerable potential for the management of patients with diabetes.” The unreliability of traditional analyses of blood glucose data is perhaps one of the reasons that “overtreating hypoglycemia has resulted in a marginally significant increase in the frequency of hyperglycemic excursions” (3) and that “many patients still experience episodes of unrecognized hypo- and hyperglycemia” (2).
