Fig. 28.1
Use of the tumorimeter. (a) The nodule is encircled, (b) the loop is placed over the nodule, whereafter the surface area can be read off
28.2.3 Procedure
All measurements were done by two observers. One of the observers was a medical doctor with extensive experience in diagnosing Dupuytren Disease. The other was a human movement scientist, who was trained to recognize Dupuytren Disease (Broekstra et al. 2015).
First, measurements were taken only by the first observer. To determine the intra-observer agreement, the participants returned 2–4 weeks later for the second measurement by the first observer. To determine the interobserver agreement, the participants were measured by the second observer immediately after the measurements of the first observer. The same procedure and measurement instrument were used in all measurements.
28.2.4 Statistical Analyses
To determine the agreement on the continuous variables (TPED, area of nodules and cords), using a one-way random effect model was used, whereafter the ICC was calculated. Only fingers with agreed positive diagnosis were used in these analyses.
The agreement on diagnosis and Tubiana stage was determined by calculating the ICC, using a latent variable underneath the binary or ordinal outcome. Detailed information on the statistical analyses is found in Broekstra et al. (Broekstra et al. 2015)
28.3 Results
A number of 54 patients (33 men and 21 women) agreed to participate, having 78 primary affected hands. Their mean age was 65.8 ± 9.2 years. Agreed positive diagnosis of Dupuytren Disease was found in 194 fingers, while in 8 fingers there was no consensus between the observers about the presence of Dupuytren Disease.
The agreement for diagnosis was very good (Altman 1991), ranging from 95.5 to 99.9 % for the intra- and interobserver agreement.
The agreement on Tubiana stage ranged from 73.5 to 98.9 %. Specified results are reported elsewhere (Broekstra et al. 2015).
For the other outcome measures, the agreements were good overall (Table 28.1). Measurements of TPED in the left middle fingers were lower than average. This was also the case for measurements of area of nodules and cords in the left middle finger. The intra-observer agreement was higher on average than the interobserver agreement.
Table 28.1
Intraclass correlations and 95 % CI for each outcome variable, presented for each hand and finger separately
