The report by Maciejewski et al. (1) concerning the three quoted articles from our institution (2–4) is flawed. Our 1990 article is one of the first ever to attempt to assess custom-made “diabetic” footwear in a scientific manner. It was an observational study (not a trial), which could serve for generating hypotheses concerning technical features of diabetic footwear, the selection of patients for footwear studies, etc., rather than for settling the issue of the effectiveness of whatever kind of diabetic footwear.
Our second article is from 1994 (3) and was an observational study as well (as was stated explicitly several times, even in the title). It reported that patients who differed significantly in compliance toward both footwear and footcare (which Maciejewski et al. failed to mention) differed in outcome (e.g., the incidence of foot lesions). This article should have been included in Table 2 (multifactorial studies) rather than in Table 3 (studies of footwear) of the Maciejewski et al. (1) report.
Regarding our third article from 2003 (4), the internal validity was rated “poor” by Maciejewski et al., who criticized “uncertainty on whether insurance denial [of footwear coverage] was a proxy for insurance differences or lower incomes” (1). Had the authors of this article been asked for additional information, as is common practice of writers of metanalyses, Maciejewski et al. would have learned that in Germany, every citizen is obliged to be member of a health insurance, either legal or private. The vast majority of citizens belongs to one of the several hundreds of legal health insurances. While the legal insurances have to follow all the same governmental standards in terms of fees and benefices irrespective of the income of the insured person, the private insurances do not. In our article (4), patients belonging to a private health insurance were not included.
Finally, reproducing a picture of an unattractive diabetic shoe and claiming that this is an example of the “Chantelau study shoe” (Fig. 1, [1]) is highly unfair. The picture was taken from our 1990 study and shows an admittedly unattractive custom-made shoe. In 2003, we have been studying industrially manufactured LucRo shoes, which are depicted in our publication (4) and which appear even more attractive than the “Reiber study shoe” that Maciejewski et al. have chosen to depict for comparison (Fig. 1, [1]).
Anyone who wants to disprove our data are invited to repeat our study in a more sophisticated trial design (e.g., with step monitors incorporated into the footwear) (5). All relevant technical features of the LucRo shoe have been outlined in detail in our publication (4) to ease further studies in this field.
