Date of Award


Document Type


Degree Name

Doctor of Philosophy (PhD)



Research Advisor

Wendy M. Likes, PhD, DNSc


Michael A. Carter, DNSc, DNP Ann K. Cashion, PhD Jeffry S. Gordon, PhD Cheryl Cummings Stegbauer, PhD


Cervical Colposcopy Dysplasia Papanicolaou Progesterone Intrauterine device


Abstract Introduction: Papanicolaou screening is the standard of care for the detection of abnormal cervical cells. Early detection and treatment of cervical neoplasia depends on the accurate interpretation of the cytologic sample. Diagnosis of dysplasia is made by histology interpretation to confirm the results of cytology screening. Cervical cells are susceptible to change as a result of hormonal exposure, and this has the potential to affect cytologic interpretation. Cervical cells may be exposed to exogenous progesterone if the woman uses injectable depot medroxyprogesterone acetate or a progesterone intrauterine device. Previous study results have been inconsistent regarding the effect of progesterone on relationship between cytology and histology interpretation. The purpose of this exploratory study was to examine the accuracy of cervical cytologic findings in women who were using either one of two forms of progesterone only contraception, depot medroxyprogesterone acetate injection or the progesterone IUD, as compared to women who either used a non-hormonal form of contraception or no contraception. Accuracy, for the purposes of this study, was defined as the association between cytology interpretation and histology interpretation. A secondary purpose of the exploratory study was to investigate whether the length of time of progesterone exposure had an effect on the association between cytologic and histologic findings. Methods: A de-identified synthetic derivative database was used for this study in which cytology and histology results from 180 women were analyzed. Three groups were mined from the database: a group using the progesterone releasing intrauterine device (IUD, N = 35), a group using the injectable depot medroxyprogesterone (N = 73), and a control group of women who used either no contraception or non-hormonal contraception (N = 72). Cytology and histology were analyzed for associations as well as for sensitivity and specificity. Chi square analysis and logistic regression were the statistical methods used to test the associations. Results: Cytology and histology results were collapsed into low grade/high grade categories. There was a significant association between cytology and histology findings (p = .008) among the total group (N = 180). Among subgroups analyzed from low grade/high grade perspective, a significant relationship was found between cytology and histology among women who had no progesterone exposure (p = .019) and those using the progesterone IUD (p = .019). However, there was not a significant association between cytology and histology findings in progesterone injection users (p = .790). Thus a new outcome variable was then applied for more precise analysis in which each cytology/histology relationship was labeled as agree, false negative or false positive. Using the new outcome variable (false negative, false positive or agree) there was no difference in the association of cytology and histology by progesterone exposure. Logistic regression revealed no relationship between length of time of progesterone exposure and accuracy of cytologic interpretation. Conclusion: Although our exploratory study was significantly underpowered, findings suggest that the subjects in this study who used either the progesterone IUD or injection had no difference in association of their cytology to histology findings compared to women who had no progesterone exposure. What can be derived from the study is a basis for comparing varied progesterone delivery systems, as well as format for repetition of a similar study in a multi center, multi site study. There are several other forms of progesterone (implant and oral progesterone formulations) which were not included in this study which may affect the association between cytology and histology results. Further investigation is warranted.



Included in

Diagnosis Commons