The criterion for the selection of the two strategies appears to have been valid, in that they represented the two available surgical procedures for patients with SP.
The clinical evidence came from a prospective cohort study, the main drawback of which was the lack of random allocation of patients to treatment groups. In effect, the choice of the surgical procedure was based on patient and surgeon preferences. This may have introduced selection bias, which strongly affects the validity of the clinical comparison. The length of follow-up appears to have been adequate to capture the morbidity or mortality for the procedure. The evidence came from a single medical institution, which may not have represented the procedures carried out in other medical centres. The size of the patient sample was not justified on the basis of statistical calculation and some outcomes referred to smaller subgroups of patients, which further limited their power to detect statistically significant differences. Some differences in clinical endpoints were apparent and reached statistical significance. No statistical analysis was carried out to demonstrate the baseline comparability of the study groups. These issues should be kept in mind when assessing the validity of the clinical evaluation.
The economic viewpoint was not explicitly stated, but the types of costs and their source appear to indicate the perspective of the hospital. The costs appear to have been derived from the hospital accounting system although this was not explicitly stated. They were not broken down into single items and the only statement was that staff costs were excluded. Little information on the economic analysis was provided, with no unit costs, quantities of resources used, and price year reported. No statistical analyses on the costs were carried out.
Analysis and results:
The costs and benefits were not synthesised, given that a cost-consequences framework appears to have been used. The clinical and economic findings were clearly presented. The issues of uncertainty and generalisability were not addressed.
The study had several methodological limitations, which could make the authors’ conclusions less robust.