Women's Health

Comparison of Two Domestic Violence Classification Tools Among Female HMO Enrollees

Bonomi AE, Thompson RS, Anderson M, Sisk C, Rivara FP

Contact:

Amy E. Bonomi, MPH, PhD, Group Health Cooperative-Center for Health Studies, 1730 Minor Avenue, Suite 1300, Seattle, WA 98101. E-mail: bonomi.a@ghc.org.

Background: Coker and colleagues report that the Women's Experience with Battering (WEB) scale is broadly applicable as a classification tool for domestic violence (DV).

Objective: To compare the WEB to 5 questions commonly administered as part of the Behavioral Risk Factor Surveillance Survey (BRFSS) to estimate DV prevalence and the false negative and false positive rates of the WEB in female enrollees of Group Health Cooperative (GHC).

Methods: We conducted telephone interviews with 806 women (mean age, 45) randomly sampled from GHC, who had an intimate partner since age 18, to estimate DV prevalence and to calculate the false negative and positive rates of the WEB. Five BRFSS questions (yes/no) assessed lifetime and current partner exposure to physical, sexual (forced intercourse or sexual contact), and psychological (threats or controlling behavior) DV. The 10-item WEB scale assessed women's perceived loss of power and control in their current or past relationships; scores greater than 20 (single item score range, 1-6) indicated exposure. Considering women in a current relationship, the false negative and false positive rates for the WEB were calculated, using the BRFSS physical/sexual and psychological (threats/controlling behavior) items as gold standards.

Results: Forty-three percent (351/806) of women reported lifetime DV exposure (responded yes to at least 1 of 5 BRFSS items), and 26% (207/806) were positive on the WEB. Fifteen percent (97/650) of women reported DV in their current relationship via the BRFSS items, and 4.5% (29/650) via the WEB scale. The false negative rate for the WEB (current relationship) was 80% (40/50) against the BRFSS physical/sexual items and 75% (60/80) against the BRFSS psychological items. The false positive rate for the WEB was 3% (BRFSS physical/sexual) and 1.5% (BRFSS psychological).

Conclusions: Our results suggest that between 75%-80% of women who experience physical, sexual, or psychological DV in their current relationship would be missed if the WEB were used as a classification tool for characterizing abuse. We will continue to compare the WEB and BRFSS items as additional data are obtained in our study of 7000 randomly selected female enrollees of Group Health Cooperative.

Hormone Replacement Therapy Utilization Pre- and Post- Women's Health Initiative HRT Trial Termination

Wegienka G, Havstad S

Contact:

Ganesa Wegienka, Henry Ford Health System, One Ford Place, 3E, Detroit, MI, 48202. E-mail: gwegien1@hfhs.org.

Background: In July 2002, the Women's Health Initiative Study published their conclusions that led to the early termination of the hormone replacement therapy (HRT) trial of estrogen plus progestin in postmenopausal women with an intact uterus. In JAMA, they reported an increased risk of breast cancer, but also a decreased risk of osteoporosis. While the findings were headline news across the nation, it is not clear whether they impacted clinical practice or the health behavior of women.

Objective: To examine the impact of the WHI results on HRT usage.

Methods: Using claims data from 15 493 women aged 50-79 continuously enrolled in the Health Alliance Plan HMO from January 2001 through November 2003, we examined the counts of HRT prescriptions filled by this fixed cohort in the 18 months before and after the WHI results were published. Using the claims data, we were not able to determine whether each woman had an intact uterus or their exact menopausal status.

Results: Overall, the number of HRT prescriptions filled decreased over time. However, in the months immediately after July 2002, there was a steep drop in the number HRT prescriptions filled by the fixed cohort. In January 2001, 2320 HRT scripts were filled, 1973 in June 2002, 1837 in July 2002, 1617 in August 2002, 1467 in September 2002 and 742 in November 2003. The number of new HRT users increased through the first half of 2002 and dropped continuously after peaking in June 2002. The patterns observed were similar for the most commonly filled prescription, Premarin (estrogen plus progestin/Wyeth-Ayerst), and all other HRT medications.

Conclusions: It appears that the publication of the WHI results and their subsequent discussion in the press affected the use of HRT in our study population. It is not clear whether the women, their clinicians, or both were the driving forces behind this reaction. The symptoms of menopause, such as hot flashes, sleeplessness, and bone loss can be disabling; however, some women have no symptoms. Future research should study the appropriateness of HRT given the different circumstances women may encounter with menopause.

Procedures and Complications After Bilateral Prophylactic Mastectomy

West CN, Barton MB, Liu AI, Geiger AM, for the Cancer Research Network PROTECTS Group

Contact:

Carmen N. West, Kaiser Permanente Southern California Research and Evaluation, 100 So. Los Robles Avenue, 2nd Floor, Pasadena, CA 91101. E-mail: carmen.n.west@kp.org.

Background: Little is known about the complications or subsequent procedures needed to rectify complications or cosmetic problems after bilateral prophylactic mastectomy.

Objective: To evaluate complications and procedures occurring after bilateral prophylactic mastectomy.

Methods: We used automated hospitalization and cancer registry records to identify women who underwent bilateral mastectomy without breast cancer at 1 of 6 health maintenance organizations between 1979 and 1999. Confirmation of bilateral mastectomies being done for prophylactic reasons, identification of the timing of initial reconstruction, and ascertaining complications and subsequent procedures were done by structured medical record review.

P

P

P

P

Results: During the study period 270 women underwent bilateral prophylactic mastectomy. Median age of women at surgery was 44 years (range 23 to 74) and the majority (90%) were Caucasian. The majority of women (179, 66%) had simultaneous reconstruction but 36 (13%) had delayed reconstruction and 55 (20%) had none. After bilateral prophylactic mastectomy 466 complications occurred in 171 (63%) women, with a median of 2 per woman (range 1 to 13). More than half (55%) required repair, including excessive scarring and implant leakage or rupture. About a third (167, 36%) were temporary, including hematoma, hemorrhage, and infection. The remaining 42 (9%) complications were permanent or psychological, including lymphedema and depression. A total of 822 subsequent procedures were performed in 167 (62%) women, with a median of 4 per woman (range 1 to 22); the majority (766, 93%) were cosmetic in nature. Complications and subsequent procedures were less common in women with no reconstruction (&#967;², = .067 and = < .001, respectively), but occurred in nearly identical proportions among women with simultaneous or delayed reconstruction (&#967;², = .764 and = .958, respectively).

Conclusions: Women undergoing bilateral prophylactic mastectomy may experience a range of complications and after reconstruction additional procedures may be required. The risks and the potential benefits of bilateral prophylactic mastectomy need to be weighed by women and their physicians.

Risk of Additional Breast Evaluation with Initiation, Discontinuation, and Continuing Use of Hormone Replacement Therapy

Boudreau DM, Buist DM, Rutter CM, Taplin S

Contact:

Denise M. Boudreau, PhD, Group Health Cooperative-Center for Health Studies, 1730 Minor Avenue, Suite 1600, Seattle, WA 98101-1448. E-mail: boudreau.d@ghc.org.

Background: Hormone replacement therapy (HRT) continues to be used for menopausal symptoms and in the prevention of osteoporosis. However, HRT use can adversely affect the accuracy of screening mammography. Estimates of the influence of HRT on the likelihood and costs of additional imaging are important to anticipate the impact of diagnostic evaluations on healthcare costs and to facilitate cost-effectiveness analyses of alternative evaluation strategies.

Objective: To evaluate the effects of hormone therapy (HT) on recall for additional breast evaluation after mammographic screening.

Methods: A population-based cohort of women age 40-80 years with at least 2 mammography screenings in a health maintenance organization was studied. Using automated health plan records on utilization and data collected at each screening, logistic regression was used to estimate the odds of recall among women who initiated, discontinued, and continued use of HT between the mammograms.

Results: The relationship between HT and recall varied by age and type of HT. There was no association between HT and recall among women 40-49 years, regardless of HT type. Among women 50 years and older, current HT users–including initiators and continuers–had an increased risk of recall for additional evaluative services compared to non-users (odds ratio [OR] = 1.2; 95% CI, 1.0-1.4 for 50-59 age group; OR = 1.8; 95% CI, 1.5-2.2 for 60-69 age group; and OR = 1.7; 95% CI, 1.4-2.0 for 70-80 age group). Odds of recall were limited to opposed HT users among women 50-59 years (opposed–OR = 1.3; 95% CI, 1.1-1.5; unopposed–OR = 1.0; 95% CI, 0.8-1.2). No differences in recall were observed between past users and nonusers of HT. Results did not vary by breast density. Increases in recall were for imaging and ultrasound but not fine needle aspiration and biopsy.

Conclusion: There is a strong relationship between HT use and recall rates that is greatest among older women and independent of measured breast density. However, recall rates return to baseline upon discontinuation of HT.

Validation of an Automated Pregnancy Episode Grouper

Hornbrook MC, Whitlock EP, Bachman D, et al.

Contact:

Mark C. Hornbrook, PhD, Kaiser Permanente Center for Health Research, 3800 North Interstate Avenue, Portland, OR 97227-1110. E-mail: mark.c.hornbrook@kpchr.org.

Objective: To validate a computerized algorithm for identifying pregnancy occurrence, pregnancy start and end dates, and pregnancy outcomes using automated patient records containing dates of laboratory tests, diagnoses, procedures, medications, delivery/termination of pregnancy, and type of pregnancy termination. The algorithm will be used to establish the denominator population of episodes for determining the incidence and prevalence of partum and postpartum morbidities.

Current Procedural Terminology, 4th Edition and ICD-9-CM

a priori

Methods: All pregnancy episodes occurring to KPNW members (1998-2001) were identified through a computerized algorithm run against the electronic medical record and other utilization files. The algorithm assigns pregnancy outcome data identified from International Classification of Diseases, 9th Edition–Clinical Modification (ICD-9-CM) diagnoses, text diagnoses from the emergency department, procedures, and diagnostic related groups. If the data suggest similar but different outcome dates, an hierarchy is used to select the outcome date. A controlled selection sample, by type of pregnancy outcome, of 511 women was drawn from a universe of nearly 25 000 pregnant women. All pregnancy episodes for the sample within the study period (n = 650) were abstracted from medical charts by technicians who were blinded as to the results of the computer algorithm. Analyses measured the extent of agreement between the 2 sources on various dimensions of the episodes.

Results: Very high rates of agreement were found for the following episode elements: diagnosis of pregnancies, liveborn outcomes, ectopic pregnancies, and molar pregnancies. The algorithm had some difficulty confirming spontaneous and therapeutic abortions and discriminating between them. The algorithm produces a number of indefinite outcome classes (eg, early loss—outcome unknown), and a significant portion of these were classified to a definite outcome by the abstractors (who had access to more information than the algorithm).

Conclusions: Our computerized pregnancy episode algorithm has strong validity for establishing the presence of a pregnancy episode, and performs well in identifying most episode outcomes and beginning and ending dates. Episodes with minimal healthcare contacts or poorly coded diagnoses and procedures significantly reduce algorithm performance.

Women and Hormone Use in the Light of New Evidence

Rolnick S, Kopher R, Kelley M.

Contact:

Sharon J. Rolnick, PhD, MPH, HealthPartners Research Foundation, 8100 34th Avenue South, 11th Floor, Bloomington MN 55425. E-mail: cheri.j.rolnick@healthpartners.com.

Background: Prior to the findings of the Women's Health Initiative (WHI), it was believed that hormone therapy (HT) provided numerous health benefits. With WHI, however, considerable doubt was raised regarding prior assumptions. We conducted a survey to better understand what women who were using HT are doing and thinking in the light of recent findings.

Objective: To assess HT usage (maintenance, discontinuation, or change), health behaviors, symptom control and overall concerns among women at a large Midwestern health plan.

Subjects: Women who had a prescription for HT (PremPro or FemHRT) between January 1, 2002 and June 30, 2002.

Methods: Participants were identified through pharmacy records and stratified by duration of hormone use (recent users [< 1 year], women with 1-5 years of use, and long-term users [>5 years]). A random sample of 10% from each stratum was selected for a total of 1200 subjects. Surveys were mailed with questions on hormone use, awareness of study findings, changes in medication use or health behaviors, what influenced changes, symptoms, use of alternative medications, and overall concerns.

Results: The response rate was 70%. Of these, 69% claimed they discontinued HT. The main reasons stated for starting HT were symptom relief (70%), and bone health (46%). Many women (35%) did not consider how long they would take hormones. Of those that did, however, only 21% thought it would be short term. Most assumed they would continue on HT "as long as they felt okay." When asked how they learned of the new study findings, twice as many cited television and magazines than their health plan. Only 19% claimed to make no changes, others discontinued, stopped then restarted, or changed their HT. Minimal change was reported in use of alternative medications or in life style.

Conclusions: Women appear to be heeding the warnings and taking initiative for changes regarding HT, rather than being encouraged at the initiation of their providers. They are concerned about current symptom control and also if past usage puts them at risk for future health problems.