Improving the Reporting of Clinical Cae erie

Purpoe To decribe common error in the analye and data preentation of a clinical cae erie and to ugget imple olution. Deign Intructional example. Method Problem with commonly ued data analyi and reporting technique in clinical cae erie are decribed uing both theoretical example and thoe from the literature. Reult An analyi reporting the proportion of a erie of patient with variable follow-up doe not adequately account for the differential follow-up among patient and i a potentially mileading way to preent data. Intead, the proportion of patient at preentation (or tudy entry) and the rate during follow-up hould be reported. imilarly, an analyi in which the final viual acuity of a erie of patient with variable follow-up i reported doe not adequately account for the effect of time and alo may be mileading. Reporting of the rate of viual acuity event during follow-up (e.g., falling below a pecified threhold, uch a 20/50 or wore) i preferred. Alternatively, when there i nearly complete follow-up, reporting the ditribution of viual acuity at pecified time point (e.g., 1 year after tudy preentation) i appropriate. mall cae erie hould not be overinterpreted becaue of the effect of chance, and appropriate tatitical analye hould be employed. Concluion Clinical cae erie often uffer from everal potential reporting flaw. Correction of thee flaw would permit the proper interpretation of the data and allow for the ability to combine data from everal cae erie to aemble more meaningful and reliable concluion.