The Impact of Continuity of Care on Cervical Cancer Screening: How Visit Pattern Affects Guideline Concordance

• Published in: Journal of cancer education Vol. 37; no. 1; pp. 30 - 36
• Main Authors: Li, Tao, Yoon, Jangho, Luck, Jeff, Bui, Linh N., Harvey, Marie
• Format: Journal Article
• Language: English
• Published: New York Springer US 01.02.2022; Springer Nature B.V
• Subjects: Adult; Biomedical and Life Sciences; Biomedicine; Cancer; Cancer Research; Cervical cancer; Continuity of care; Continuity of Patient Care; Early Detection of Cancer; Female; Guidelines; Humans; Mass Screening; Medical screening; Pap smear; Papanicolaou Test; Papillomavirus Infections - prevention & control; Pharmacology/Toxicology; Pregnancy; Screening Tests; United States; Uterine Cervical Neoplasms - diagnosis; Uterine Cervical Neoplasms - prevention & control; Vaginal Smears; Young Adult; United States; Continuity of care; Medicaid; Cervical cancer; Guideline concordance; Cancer screening
• ISSN: 0885-8195; 1543-0154
• DOI: 10.1007/s13187-020-01777-3

Cervical cancer can be prevented and highly curable if detected early. Current guidelines recommend women to receive cervical cancer screening starting at age 21. Our study aims to investigate how improving continuity of care (COC) may influence guideline concordance of cervical cancer screening. Using the eligibility and claims data, we created a person-month panel data set for women who were enrolled in Oregon Medicaid for at least 80% of the period from 2008 to 2015. We then selected our study cohort following the cervical cancer screening guidelines. Our dependent variable is whether a woman received cervical cancer screening concordant with guidelines in a given month, when she did not receive Pap test in the past 36 months and did not receive co-testing of HPV test plus Pap test in the past 60 months. We used both population-averaged logit model and conditional fixed-effect logit model to estimate the association between the guideline concordance and the COC index, after controlling for high risk, pregnancy, age, race, and ethnicity. A total of 466,526 person-month observations were included in our main models. A 0.1 unit increase of the COC score was significantly associated with a decrease in the odds of receiving guideline-concordant cervical cancer screening (population-averaged logit model: OR = 0.988, p  < .001; conditional fixed-effect logit model: OR = 0.966, p  < .001). Our findings remain robust to a series of sensitivity analyses. A better COC may not be necessarily beneficial to improving cervical cancer prevention. Educations for both physicians and patients should be supplemented to assure quality of preventive care.