Contraceptive Methods and the Impact of Menstruation on Daily Functioning in Women With Sickle Cell Disease

Melissa E. Day, BS; Sarah-Jo Stimpson, MD; Mark Rodeghier, PhD; Djamila Ghafuri, MBBS; Michael Callaghan, MD; Ahmar Urooj Zaidi, MD; Bryan Hannan, MS; Adetola Kassim, MD, MS; Andra H. James, MD, MPH; Michael R. DeBaun, MD, MPH; Deva Sharma, MD, MS


South Med J. 2019;112(3):174-179. 

In This Article


Study Design and Populations

We obtained approval from the institutional review board for study protocols at Vanderbilt University Medical Center and Wayne State University Medical Center. In a cross-sectional study conducted across two urban academic medical centers in the United States, we administered paper surveys to adolescent and adult women with SCD after obtaining informed consent. We recruited participants from outpatient SCD clinics and inpatient wards. Women recruited from the inpatient wards were admitted because of acute vaso-occlusive pain episodes. The study coordinators reviewed the weekly clinic patient list and the inpatient census to determine who could be approached for the study and was available throughout the week to recruit and consent participants. Inclusion criteria were girls and women aged 10 to 55 years with a confirmed diagnosis of SCD (of any phenotype) who had reached menarche. Exclusion criteria included women with primary amenorrhea or secondary amenorrhea caused by pregnancy, lactation, or menopause. Women using menstrual-suppressing contraception methods were eligible to participate, as well as women with irregular menstrual periods. Data were collected from February 2016 to August 2017.

Surveys Administered

Participants completed four surveys: a demographic survey, the validated Menstrual Symptom Questionnaire (MSQ),[9] the validated Sickle Cell Disease Pain Burden Interview-Youth (SCPBI-Y) survey,[10] and a modified version of the Sickle Cell Disease Pain Burden Interview-Youth (mSCPBI-Y) survey. A demographic survey (Supplemental Digital Content Appendix I, assessed baseline menstrual characteristics and SCD-related comorbidities. The validated MSQ was used to define dysmenorrhea (Table 1).[9] The validated SCPBI-Y survey[10] assessed the impact of acute vaso-occlusive pain episodes on QOL during the previous 30 days, excluding pain occurring with menstruation. The mSCPBI-Y survey (Supplemental Digital Content Appendix II, determined the effects of vaso-occlusive pain experienced during a participant's last menstrual cycle on QOL. Chart reviews were completed for all of the participants to determine whether they had an underlying cause of secondary dysmenorrhea, such as endometriosis or fibroids.


  • Dysmenorrhea: Painful menstrual cramps of uterine origin. Primary dysmenorrhea occurs in the absence of pelvic pathology, whereas secondary dysmenorrhea occurs in the presence of pelvic pathology (eg, endometriosis, fibroids).[11] We classified participants as having dysmenorrhea by using select criteria from the MSQ (Table 1) and further stratified dysmenorrhea into primary or secondary based on chart review results.

  • Prolonged menstrual bleeding: Experiencing menses for >7 days on average, as self-reported in the demographic survey.[12]

  • Poor QOL: According to the mSCPBI-Y survey, experiencing "sleep disturbance" or "missed work or school" for "many days" or "every day." Scores on the mSCPBI-Y range from 0 (no pain burden) to 28 (severe pain burden).[10]

Data Analysis and Study Outcomes

All of the survey results were entered into REDCap.[13] Our primary outcomes were rates of prolonged menstrual bleeding, primary and secondary dysmenorrhea, and contraceptive use, as well as QOL both generally and related to menstruation (defined through scores on the mSCPBI-Y survey). Associations of clinical and demographic variables with menstrual characteristics and dysmenorrhea were evaluated by the χ 2 test or the Fisher exact test for small percentages. The proportion of women who met the criteria for dysmenorrhea was determined based on predefined criteria (Table 1). A P value <0.05 was considered significant for all of the analyses.