iDentistry The Journal identistry_may_aug2019 | Page 7

The Journal
Statistical methods Discussion
The linear variables was expressed in
percentage repeatability. The sensitivity and
specificity were calculated by comparison with
the test kits and clinical judgement at baseline
which classified the sites as diseased or
healthy. After treatment, each site group as;
"better", "no change" and "worse". The collection of data carried out in this study
from 8 sites in each patient which took
approximately 30-40 min. 15 min allowed for
incubation of the Periocheck and Perioscan
test. On this basis this is time consuming to
provide a full mouth evaluation with these tests
for any patient.
Results
Thirty subjects (18 females,12 males) entered
the study with 240 potential sites for data. Two
patient were decreased during the trial reducing
the available sites for complete data to 224. At
baseline, 3 periocheck sites and 7 perioscan
sites were discarded due to blood
contamination of the test strips, yielding final
totals of 221 and 217 useable sites respectively.
The intra examiner reproducibility of the
variables like plaque index, colour score and
bleeding index expressed good agreement with
unweighted K scores ranging from 0.69 to 0.78
and inter-examiner comparisons from 0.65-
1.00.
Both tests Perioscan and Periocheck had a high
sensitivity (Perioscan 97% and Periocheck
87%) where sensitivity measured the
probability that a site diagnosed as clinically
positive would yield a positive test result, but a
low specificity (Perioscan 54% Periocheck
58%) where specificity measured the
probability that a site diagnosed as clinically
negative would yield a negative test result.
7
Vol. 15
No. 2
May-Aug 2019