Were all possible hypotheses examined? Often the report of an observational study has as a focus the most dramatic finding, e.g. a difference between teaching and non-teaching hospitals, but when comparing services in which there are differences in the same outcome measure more than one explanation is possible.
• Who did the operation, e.g. specialist or generalist?
• When was it done, e.g. day or night, weekday or weekend?
• Where was it done, e.g. in a large-volume or small-volume service?
• How was it done?
• Which operation was done?