Impact of gender representativeness in online symptom survey and clinical trial participation among patients with myeloproliferative neoplasms.

ABSTRACT

Patients with myeloproliferative neoplasms (MPNs) face chronic symptom burden. Online symptom assessment studies allow for recruitment of large numbers of motivated patients, but patient self-selection can lead to sampling bias. This study evaluated how gender representativeness in MPN symptom surveys and trials impacted symptom score mean estimates, using data from 4825 survey respondents and 291 trial participants with MPNs. The survey data showed that men participated at a rate roughly 50% less than what would be expected based on prevalence, and women reported higher scores than men on average for six of 10 symptoms. Together, this led to potential over estimation in six of 10 symptom score means (ranging from 5.8% to 15.3% overestimated). The trial data showed less gender-based sampling bias compared to the survey data. Studies utilizing online symptom surveys should implement study design features to recruit more men, assess for gender participation imbalances, and provide weighted estimates where appropriate.