Selective outcome-reporting bias refers to the selective reporting of a subset of study findings. This methodological limitation may occur in cancer-related acupuncture studies, where valid empirical studies on psychometric performance are still lacking. We assessed the risk of selective outcome reporting bias in studies published in English that were included in a systematic review on acupuncture for preventing cancer chemotherapy-induced nausea and vomiting. For each study, we searched for registry availability and, if present, assessed its validity. We described each study outcome (nausea, vomiting, or both) according to the following seven items: type of outcome, domain, specific measurement, specific metric, type of data, methods of aggregation, and timepoint unit and time. Eleven studies published between 1987 and 2019 in English were evaluated. Only four (36%) had a registry, of which only two were prospective and therefore considered valid. Discrepancies were found in the specific measurement of the outcome in two studies and in the specific metric. In many other cases, discrepancies were not evaluable due to missing information. No study reported complete outcomes as planned in the published protocol. Communication about the importance of prospective trial registration, including outcome details, should be enforced to reduce the risk of selective outcome reporting bias in oncology acupuncture studies.
Penati et al. (Fri,) studied this question.