27 women were included at the Linköping study site. All were over 50 years of age (median 64, maximum 80), with minimal or no other health problems and a dorsally dislocated distal radial fracture that had required closed reduction but was deemed appropriate for nonoperative treatment. No strict radiographic inclusion criteria were defined, as this was dependent on clinical traditions of the different study sites, and the choice of nonoperative treatment should be based on the clinical situation rather than on radiographs. In Linköping, included patients had a pre-reduction dorsal angle roughly between 5 and 30 degrees and discontinuities of the radial joint surface of less than 2 mm. Main exclusion criteria were: concomitant injury, previous wrist fracture, malignant neoplasm during the previous 5 years, liver disease, high calcium levels, joint disease, or a disease affecting bone metabolism (Aspenberg et al. 2009
). The protocol for the multicenter study was followed rigorously. No patient was lost to follow-up. A site visit by the Swedish Medial Agency gave a good report.
Our evaluation was done after the last patient had been included, during the postoperative year, which had to pass before unblinding. After looking at series of digital radiographs for a number of patients, we decided only to use the pre-reduction images and those at the 5-week follow-up. We asked ourselves: “Assuming that PTH has a positive effect on fracture healing, do we think this patient received PTH?” and answered yes or no. After we had categorized all 27 patients, we went back and tried to describe the criteria that we had used.
This description was as follows. We (the authors) tried to get a general impression of the amount and density of callus, relative to what should be expected for that fracture type. We started with the lateral projection. In most cases, there is a sharp angle in the dorsal cortical contour at the fracture. At this angle there is usually an external callus. Most of our interest was focused on this callus. If absent, the case was rated as “placebo”. If there was a large or dense callus, it was rated as “PTH”. In cases of uncertainty, we looked for callus in other areas, i.e. at the volar cortex or at the sides in the anteroposterior projection. We paid no attention to the fracture line or to cortical continuity. We sometimes discussed the appearance of the cancellous bone, but generally concluded that this should not be included in the judgement. Fractures with little dislocation were expected to produce smaller external callus. In these cases, we looked more at callus density than at its size. We looked at the radiographs together and made consensus decisions.
We repeated the rating procedure 3 days later without seeing the previous results. 2 patients differed in rating between these time points. This latter rating was regarded as the definitive one and was reported to the sponsor who, at this time, did not inform us about whether or not the rating had been successful.
After the official outcome of the entire multicenter study was clear, but while we were still blinded, we were asked to do the same rating for all patients from all sites. This was not possible, however, because many patients at the other sites had had their 5-week radiographs taken before the plaster was taken off, and the images could not be evaluated. We therefore rated only the patients from Linköping again, almost a year after the first rating. This time, we rated the callus as rich, intermediate, or poor.