Is this study valid Definition of the tool outcome:   Don't confuse this with the search outcome, discussed above.  the outcome of the tool (which is really a diagnostic test here) is the presence or absence of an ectopic pregnancy.   Our paper clearly defines these outcomes and displays the criteria for diagnosis or exclusion of an ectopic in a easy to follow table Blind assessment of outcome:  The article does not specifically state that persons determining outcome (ectopic or not) were blinded to findings on clinical examination Prospective assesment of predictors: Yes.  In fact a template form including all potential predictor variables studied was used as the medical record. Are the potential predictor variables clearly defined: Yes.  An appendices briefly defines criteria for each variable Are the potential predictors reliable?   They did not assess inter-rater reliability for physical examination findings.  They note in their discussion that subjecting patients to two, potentially painful examinations, may not be ethical. Were the mathematical techniques used to define the rule explained: yes.  They used recursive partitioning and described the method. Did they include too many variables in the final rule? 2 variables define a high risk group.  4 additional variables differentiate the remaining patients into moderate and low risk groups.  Thus 6 variables.  To avoid over fitting they should have had at least 60 ectopic gestations in their sample.  They had only 39.  Thus the potential for data over fitting exists. Description of the derivation population:  The demographic characteristics are displayed in a table Validation: This is a derivation study. data splitting was used but we will have to search for an independent validation study.