6.2. Methodological aspects

The findings in these five studies were based on a variety of observations and sampling methods. In study I the harvested cores were measured so that their dimensions, weights and volumes could be determined. The fact that the two investigators performed all of the surgical procedures together and that one investigator made all the measurements, helped to minimize the problems of interexaminer variability. There was good agreement between the measurements of the CC obtained from cadaver subjects and from actual patients with no statistical differences between the two.

Intra-operative observations, a chart review and survey questionnaire were used as the 3 components of study II. There was good agreement with study I with respect to CC dimensions, although the actual total volume of bone ultimately harvested in study II was determined by the reconstructive requirements of the patient. There was a 98.8% patient satisfaction rate with this procedure and generally the patients were quite happy to complete the survey questionnaire as evidenced by a 100% completion rate. The questionnaire was modelled after the well known SF-36 Health Survey (Ware & Sherbourne 1992). This type of questionnaire was designed to be administered in person by an individual trained in interviewing techniques or by telephone (Ware & Sherbourne 1992). The SF-36 has been tested in over 1700 patients with evidence of the responsiveness of the questionnaire to changes in perceived clinical status (Ware 1993, Garratt et al. 1994, 1995).

A prospective case-control format was used in study III with closely matched for age and sex treatment groups or cohorts. Analysis of variance (ANOVA) failed to show any statistically significant differences between the two groups. Scheduling issues required the participation of two investigators in data collection and interexaminer variability may remain an unknown entity in this study. However variables such as time to first ambulation and length of hospital stay are precisely defined objective outcome measures that have no subjective component and should not be affected by interexaminer variability. The main survey tool to measure pain was a visual analogue scale, which was self-administered. The visual analogue scale is an established, validated, self-report measure usually consisting of a 10 cm line on paper with verbal anchors clearly labelling the ends (Gracely & Kwilosz 1988, Gracely 1990, 1999, Eliav & Gracely 1988, Jamison et al. 2002). There are a number of variations on the use of the scale such as with palm top computers or personal digital appliances instead of paper, but the concept of a measurable line which connects two clearly defined anchors at either end is its major feature. This recording method has been validated using a multivariate analysis (Jamison et al. 2002). Such a scale can be quite useful in measuring pain experiences in the post surgical patient. Since the visual analogue scale was self-administered, it too should be minimally affected by interexaminer variability.

The format of study IV is that of a case series with no control group, in which there was excellent compliance by patients in terms of follow-up attendance. The short follow-up period of 1 to 3 years could have been extended and this may have revealed further problems with the resorption of the CDG material. Although only 5 of 54 (9.3%) of sites showed clinically or radiographically apparent resorption of the xenograft, it is possible that serial volumetric CT scans may have shown otherwise.

In study V closely matched for age and sex treatment groups were used and analysis of variance (ANOVA) failed to show any statistically significant differences between the two groups. While an attempt was made in study V to choose two very similar wound types to test the behaviour of CDG in the dento-alveolar region, the fact is that the biology of traumatically induced wounds may be quite different from elective surgical wounds. This could account for the dramatic difference in success of the CDG in the elective surgical wounds when compared to the traumatized wounds. The small sample size of 21 patients with 48 surgical sites may also have failed to disclose other differences or possible complications, which may occur with the xenograft. Finally this study was purposely carried out in a young population, but growth itself may also be a confounding variable in this study.