Evaluation of arthroscopy and macroscopic scoring

Abstract

Introduction: Arthroscopy is a minimally invasive technique for retrieving synovial biopsies in rheumatology during the past 20 years. Vital for its use is continual evaluation of its safety and efficacy. Important for sampling is the fact of intraarticular variation for synovial markers. For microscopic measurements scoring systems have been developed and validated, but for macroscopic evaluations there is a need for further comprehensive description and validation of equivalent scoring systems.

Methods: We studied the complication rate and yield of arthroscopies performed at our clinic between 1998 and 2005. We also created and evaluated a macroscopic score set of instructions for synovitis.

Results: Of 408 procedures, we had two major and one minor complication; two haemarthrosis and one wound infection, respectively. Pain was most often not a problem, but 12 procedures had to be prematurely ended due to pain. Yield of biopsies adequate for histology were 83% over all, 94% for knee joints and 34% for smaller joints. Video printer photographs of synovium taken during arthroscopy were jointly and individually reviewed by seven raters in several settings, and intra and inter rater variation was calculated. A macroscopic synovial scoring system for arthroscopy was created (Macro-score), based upon hypertrophy, vascularity and global synovitis. These written instructions were evaluated by five control-raters, and when evaluated individual parameters were without greater intra or inter rater variability, indicating that the score is reliable and easy to use.

Conclusions: In our hands rheumatologic arthroscopy is a safe method with very few complications. For knee joints it is a reliable method to retrieve representative tissue in clinical longitudinal studies. We also created an easy to use macroscopic score, that needs to be validated against other methodologies. We hope it will be of value in further developing international standards in this area.

Figure 1

Arthroscopy map of the knee joints. The patella is folded sideways. Areas of hypertrophy (villi, granulations) are marked as indicated, areas of increased vascularity are encircled and biopsy sites are mapped immediately following each procedure by the performing arthroscopist. Dx, right side of patient; SIN, left side of patient.

Figure 2

Median scores. Scores (%) sorted by median. Fifty images were scored for three parameters twice by seven raters, producing a total of 2100 scores. For S no images had a maximum median score of 4. H = hypertrophy; S = synovitis; V = vascularity.

Figure 3

Intra-rater variation. (a) Percentage of perfect matches (score 1 = score 2) for individual raters. (b) Percentage of perfect matches (score 1 = score 2) at different scoring levels. (c) Rater consistency (sum of score 2 minus score 1). 0 indicates a high consistency between scoring sessions. A positive value indicates an increase in scores at the second scoring session. A negative value indicates a decrease in scores at the second scoring session.

Figure 4

Inter-rater variation. (a) Distance to the median. Percentage of scores (from the first rating session) with an absolute distance from the median of one point or less. (b) Score range in percentage of total scores. A range of 0 is a perfect match of all seven scores for an image (seven raters). A range of one means that all scores are within two values. H = hypertrophy; S = synovitis; V = vascularity.

Figure 5

Control-rater variation (%). (a) Perfect matches (score of raters A to E = median score of raters 1 to 7). *Median score of raters A to E = median score of raters 1 to 7. (b) Distance to the median. Scores of raters A to E with an absolute distance from the median of raters 1 to 7 of one point or less (compare with Figure 4a). *Median scores of raters A to E with an absolute distance from the median of raters 1 to 7 of one point or less. H = hypertrophy; S = synovitis; V = vascularity.