Development of a difficulty score for spinal anaesthesia

Background. Multiple attempts at spinal puncture may be hazardous. Accurate preoperative prediction of difficulty adds to the delivery of high quality care. This clinical trial was designed to: (i) determine the predictive performance of difficulty variables; (ii) compare senior and junior anaesthetists; (iii) develop a score to predict difficulty during the performance of spinal anaesthesia. Methods. A total of 300 patients subjected to urological procedures and scheduled for spinal anaesthesia were independently assessed and stratified according to the categories of the difficulty predictors of spinal anaesthesia into one of nine grades (0-8) and randomized according to the experience of the anaesthetist into two groups (group A, staff with more than 15 yrs' experience; group B, resident with more than 6 months but less than 1 yr in training). The number of attempts and levels, and success rate of the technique were the outcome variables. Data were analysed by multivariate analysis and receiver operating characteristic (ROC) curves. Results. The bony landmarks of the back and the radiological characteristics of the lumbar vertebrae were two independent predictors of difficulty. Multivariate analysis indicated differences between junior and senior staff but ROC curves indicated no difference. Grade 4 was the difficulty score at or above which difficulty was expected whether or not radiological characteristics of the vertebrae were included. Conclusions. Spinal bony landmarks and radiological characteristics of the lumbar vertebrae are independent predictors of difficulty during spinal anaesthesia. There is no difference between senior and junior anaesthetists. Grade 4 is the difficulty score at or above which difficulty is expected.