Identifying Patient Smoking History for Cessation and Lung Cancer Screening through Mining Electronic Health Records

Title:

Identifying Patient Smoking History for Cessation and Lung Cancer Screening through Mining Electronic Health Records

Link:

https://aacrjournals.org/cebp/article/26/3/437/71229/Perceptions-of-Cervical-Cancer-Risk-among-Ohio

Abstract:

OBJECTIVE: To describe how Ohio Appalachian women are over- or under-estimating risk for cervical cancer by comparing an objective measure of cervical cancer risk to subjective interpretations of risk. METHODS: A cervical cancer risk index was computed for 530 female participants recruited from clinics in Ohio Appalachia (2006–2008). Risk factors for cervical cancer, including human papillomavirus infection, number of sexual partners, parity, age at first intercourse, age at first pregnancy, oral contraceptive use and cigarette smoking were used to create a cervical cancer risk index. This index was used to categorize participants into low, average or high risk for cervical cancer. Subjective risk was assessed by asking women to indicate whether they believe their risk of cervical cancer to be below, about average, or higher than other women. Objective risk categories were compared to subjective risk to determine if women were accurately perceiving risk, or over- or under-estimating risk. A multinomial logistic regression model was used to determine correlates of risk perception accuracy. The outcome was a categorical variable consisting of over-estimating risk, underestimating risk and accurate risk perception, with accurate risk perception serving as the reference level in the model. RESULTS: Out of 530 women, 33.6% (n = 178) over-estimated risk and 24.9% (n = 132) under-estimated risk. Women with household incomes of <$25 k were more likely to under-estimate cervical cancer risk, as opposed to being accurate in their risk perception, than women in households earning over $50 k [OR = 2.06, 95% CI (1.04–4.07)] as were women with less than a high-school education [OR = 2.52, 95% CI (1.06–5.96)] compared to women with a college education. Also, women who reported recent alcohol use were more likely to under-estimate risk [OR = 1.94, 95% CI (1.18–3.19)] compared to women with no self-reported recent alcohol use. No variables were found to be associated with risk over-estimation. CONCLUSIONS: Many women in Ohio Appalachia do not accurately perceive their cervical cancer risk. Low income and education are associated with under-estimating risk. Future studies should determine if risk misperceptions are associated with cervical cancer screening and other preventive behaviors.

Citation:

Tracy L. Onega , Ellen L. Nutter, James D. Sargent, Jennifer A. Doherty, Saeed Hassanpour, “Identifying Patient Smoking History for Cessation and Lung Cancer Screening through Mining Electronic Health Records”, American Society of Preventive Oncology Annual Meeting (ASPO), Seattle, WA, 2017.