Decision aids related to hormone replacement therapy (HRT), whether delivered in written form,1,2 along with audiotapes,3 or as part of discussion groups,4 have outperformed generic brochures in increasing knowledge and accuracy of risk assessments. However, these decision aids have provided women with population-based estimates of average risk, not individual risk levels that may have bearing on their decisions about HRT. Decision aids individually customized or “tailored” to include only the most relevant information could make it easier for women to consider HRT's risks and benefits.5,6 Tailored interventions have yet to be evaluated for HRT decisions. We describe the effect of a tailored decision aid on women's accuracy of perceived risk for breast cancer, confidence to decide about HRT, and satisfaction with the decision.
METHODS
Study Design
Between October 1998 and February 1999, interviewers called households from a purchased list to identify women aged 45 to 54 years who were willing to receive written materials about HRT and who did not have a history of breast cancer. Eligible women stratified by baseline HRT use were randomized to either a delayed or an active intervention arm. Women in the active arm received materials 2 weeks after the baseline survey; those in the delayed arm received materials after completing the study. Telephone surveys were conducted at 1 and 9 months. Study protocols were approved by the institutional review board.
Intervention
The trifold decision aid7 included (1) “Step 1 The Facts” (19 pages), which was tailored to baseline perceived menopausal status, hysterectomy status (no or yes), prior HRT use, and accuracy of perceived risk for breast cancer8; (2) “Step 2 What's Important to You,” a worksheet to record preferences; “Step 3 Next Steps” (13 pages), which included vignettes of women at decision points similar to those of women receiving the intervention and a checklist of questions for the health care providers of women receiving the intervention.
Outcome Measures
Accuracy was the agreement between women's perceived and objective 10-year risk for breast cancer as measured by the Gail score.8 Perceived risk was assessed on a 0 (certain not to happen) to 100 (certain to happen) scale. Breast cancer risk factors were used to calculate a Gail score.8 Accuracy was computed as the difference between the woman's perceived and objective risk score. The woman's perception was accurate unless the absolute value of the difference score exceeded 10%.9
Level of confidence in ability to understand the risks and benefits of HRT, make a decision about HRT, and discuss HRT with a health care provider was rated (0 = low to 10 = high; Cronbach α = .78). Items were summed to yield an average level of confidence.
Women's satisfaction was assessed by agreement (1 = strongly disagree to 5 = strongly agree) with 6 statements related to being informed about HRT, whether the decision (for those who made a decision) was consistent with their personal values, and overall satisfaction with the decision among those who had made a decision (Cronbach α = .78).10
Statistical Analysis
Logistic regression models were tested to predict dichotomized confidence (based on the median baseline value), accuracy of perceived breast cancer risk, and satisfaction outcomes at each follow-up. Covariates were intervention arm, baseline value, race/ethnicity, education, marital status, working for pay, perceived menopausal status, ever use of HRT, hysterectomy status, decision status, and numeracy (for the accuracy outcome).
RESULTS
Recruitment and Follow-Up
Of the 2388 telephone numbers called, 158 (7%) numbers were not working, 844 (35%) people were ineligible, 444 (19%) calls were never answered, and 361 (15%) people refused to participate. Of the 581 women who were randomized, 557 (96%) and 541 (93%) completed the 1- and 9-month surveys, respectively. Complete data are available for 520 (90%) of the women.
Study Outcomes
Women in the active intervention arm were significantly less likely than those in the delayed arm to be working for pay (P = .01) and to have had a hysterectomy (P = .02) (Table 1 ▶).
TABLE 1.
—Baseline Characteristics, by Intervention Arm
| Total (N = 581) | Delayed Intervention (n = 292) | Active Intervention (n = 289) | P | |
| Demographics | ||||
| Age, % | .98 | |||
| 45–49 | 55 | 55 | 55 | |
| 50–54 | 45 | 45 | 45 | |
| Some college or more, % | 75 | 75 | 76 | .83 |
| Married or living as married, % | 78 | 77 | 78 | .82 |
| African American, % | 24 | 24 | 23 | .73 |
| Work for pay, % | 85 | 89 | 81 | .01 |
| Money for special things, % | 67 | 67 | 67 | .95 |
| Health services | ||||
| Had health insurance, % | 96 | 96 | 96 | .70 |
| Had regular doctor, % | 94 | 94 | 94 | .85 |
| Menopause related | ||||
| Perceive menopause beginning or begun, % | 71 | 71 | 70 | .49 |
| Clinically postmenopausal, % | 50 | 51 | 49 | .65 |
| Symptoms present | 90 | 91 | 89 | .47 |
| Hormone replacement therapy related | ||||
| Ever used hormone replacement therapy, % | 41 | 40 | 41 | .89 |
| Decided about hormone replacement therapy, % | 47 | 46 | 48 | .68 |
| Decided to use hormone replacement therapya | 80 | 79 | 80 | .81 |
| Satisfaction with decision, mean (SD)a | 4.2 (0.7) | 4.1 (0.7) | 4.2 (0.7) | .17 |
| Currently using hormone replacement therapy, % | 29 | 30 | 28 | .59 |
| Confidence to decide about hormone replacement therapy, mean (SD) | 6.6 (2.2) | 6.6 (2.2) | 6.7 (2.2) | .47 |
| Health-related variables | ||||
| Health good or excellent, % | 90 | 90 | 90 | .97 |
| Currently smoking, % | 12 | 10 | 13 | .23 |
| Perceived likely to get breast cancer in lifetime, % | 49 | 50 | 47 | .42 |
| Perceived breast cancer risk accurately, % | 32 | 34 | 29 | .19 |
| Had hysterectomy, % | 24 | 28 | 20 | .02 |
| Has family history of breast cancer, % | 11 | 13 | 10 | .31 |
aAssessed among the 269 women who had decided about hormone replacement therapy.
Confidence in ability to decide about HRT.
Women in the active arm were more likely than those in the delayed arm to be confident about making a decision at both follow-ups (1 month: odds ratio [OR] = 2.5; 95% confidence interval [CI] = 1.6, 3.9; 9 months: OR = 2.8; 95% CI = 1.8, 4.5, respectively) (Table 2). Women in the active arm who were confident at 1 month were more likely to remain confident in their decision at 9 months than were comparable women in the delayed arm (OR = 2.5; 95% CI = 1.6, 4.0).
Accuracy of perceived risk for breast cancer.
At both follow-ups, women in the active arm were more likely to accurately perceive their level of risk for breast cancer than were those in the delayed arm (1 month: OR = 1.9; 95% CI = 1.3, 2.9; 9 months: OR = 1.9; 95% CI = 1.2, 2.8, respectively). Among those with accurate risk perceptions at 1 month, women in the active arm were more likely than those in the delayed arm to retain those perceptions at 9 months (OR = 2.2; 95% CI = 1.3, 3.7).
Satisfaction with decision.
At 1 month only, women in the active arm were more likely than those in the delayed arm to report that they were very satisfied with their HRT decision (OR = 2.5; 95% CI = 1.5, 4.3). However, among those who reported being satisfied at 1 month, women in the active arm were more likely than women in the delayed arm to remain satisfied with their decision at 9 months (OR = 2.8; 95% CI = 1.5, 5.3).
DISCUSSION
The decision aid improved the accuracy of women's perceptions of breast cancer risk, confidence to make decisions about HRT, and satisfaction with decisions. These intervention effects were sustained between 1- and 9-month follow-ups. HRT decisions might benefit by additional customization or brief telephone counseling calls, which have been effective for other health-related outcomes.11
More than 40% of the undecided women in the active arm made a decision by the 9-month follow-up; in addition, among those decided at baseline and 1 month, women in the active arm reported greater satisfaction at 9-month follow-up than did those in the delayed arm. Providing such decision aids to women before clinic appointments could enable them to make better use of limited visit time.
Although most of the participants were more educated and health oriented than the general population, our use of purchased lists resulted in recruitment of a broader crosssection of women than have been included in prior research on this topic; 24% of our participants (vs 10% of those in previous studies2,3) were African Americans. Involving community groups should be considered as a means of further expanding intervention reach.
As ever-increasing numbers of women enter menopause, rapidly changing knowledge about HRT requires innovative and flexible communication strategies to meet their information needs.
TABLE 2.
—Odds Ratios (ORs) and 95% Confidence Intervals (CIs) for Study Outcomes at 1 Month, 9 Months, and Sustained Between 1 and 9 Months
| n | Active vs Delayed Intervention, OR (95% CI) | P | |
| Confidencea | |||
| 1 mo | 531 | 2.5 (1.6, 3.9) | < .0001 |
| 9 mo | 514 | 2.8 (1.8, 4.5) | < .0001 |
| 1 and 9 mo | 507 | 2.5 (1.6, 4.0) | < .0001 |
| Accuracy of perceived risk for breast cancerb | |||
| 1 mo | 500 | 1.9 (1.3, 2.9) | .002 |
| 9 mo | 471 | 1.9 (1.2, 2.8) | .004 |
| 1 and 9 mo | 464 | 2.2 (1.3, 3.7) | .004 |
| Satisfied with decision about hormone replacement therapyc | |||
| 1 mo | 256 | 2.5 (1.5, 4.3) | .0005 |
| 9 mo | 310 | 1.5 (0.9, 2.4) | .09 |
| 1 and 9 mo | 212 | 2.8 (1.5, 5.3) | .001 |
aDichotomous outcome, proportion above the cutpoint (baseline sample median) at each follow-up. The confidence models were adjusted by race/ethnicity, education, age, marital status, work for pay, perceived menopausal status, ever use of hormone replacement therapy, hysterectomy status, baseline confidence, and hormone replacement therapy decision status.
bAccurate if the difference between the 10-year perceived risk and the 10-year objective risk is less than or equal to 10% (based on the Gail score8). The accuracy-of-perceived-risk models were adjusted by race/ethnicity, education, age, marital status, work for pay, perceived menopausal status, ever use of hormone replacement therapy, baseline accuracy of perceived risk, and numeracy.
cDichotomous outcome, proportion above the cutpoint (baseline sample median) at each follow-up. The satisfaction models were adjusted by race/ethnicity, education, age, marital status, work for pay, perceived menopausal status, ever use of hormone replacement therapy, and hysterectomy status.
Acknowledgments
This work was supported by a grant from the National Cancer Institute (PO1-CA-72099-05).
The authors would like to acknowledge Dr Celette Sugg Skinner for helpful comments on earlier drafts. The authors also thank Maragatha Kuchibhatla and Pauline Lyna, who assisted in data analyses, and Pamela Harris for her assistance in preparing the manuscript.
All authors contributed to the conception, analysis, interpretation, and writing of the brief. C. M. McBride, L. A. Bastian, L. Fish, I. M. Lipkus, H. B. Bosworth, B. K. Rimer, and I. C. Siegler were key to the development, implementation, and evaluation of the intervention. C. M. McBride, L. A. Bastian, S. Halabi, and I. C. Siegler oversaw data collection, analysis, and interpretation.
Peer Reviewed
References
- 1.O'Connor AM, Fiset V, DeGrasse C, et al. Decision aids for patients considering options affecting cancer outcomes: evidence of efficacy and policy implications. J Natl Cancer Inst Monogr. 1999;25:67–80. [DOI] [PubMed] [Google Scholar]
- 2.Rothert ML, Holmes-Rovner M, Rovner D, et al. An educational intervention as decision support for menopausal women. Res Nurs Health. 1997;20:377–387. [DOI] [PubMed] [Google Scholar]
- 3.O'Connor AM, Tugwell P, Wells GA, et al. Randomized trial of a portable, self-administered decision aid for postmenopausal women considering long-term preventive hormone therapy. Med Decis Making. 1998;18:295–303. [DOI] [PubMed] [Google Scholar]
- 4.Hampson SE, Hibbard JH. Cross-talk about the menopause: enhancing provider-patient interactions about the menopause and hormone therapy. Patient Educ Couns. 1996;27:177–184. [DOI] [PubMed] [Google Scholar]
- 5.Rimer BK, Conaway M, Lyna P, et al. The impact of tailored interventions on a community health center population. Patient Educ Counseling. 1999;37:125–140. [DOI] [PubMed] [Google Scholar]
- 6.Kreuter MW. Dealing with competing and conflicting risks in cancer communication. J Natl Cancer Inst. 1999;25:27–35. [DOI] [PubMed] [Google Scholar]
- 7.Bastian LA, McBride CM, Fish L, Lipkus IM, Rimer BK, Siegler I. Evaluating participants' use of an HRT decision-making intervention. Patient Educ Couns. In press. [DOI] [PubMed]
- 8.Gail MH, Brinton LA, Byar DP, et al. Projecting individualized probabilities of developing breast cancer for white females who are being examined annually. J Natl Cancer Inst. 1989;81:1879–1886. [DOI] [PubMed] [Google Scholar]
- 9.Lerman C, Lustbader E, Rimer B, et al. Effect of individualized breast cancer risk counseling: a randomized trial. J Natl Cancer Inst. 1995;87:286–292. [DOI] [PubMed] [Google Scholar]
- 10.Holmes-Rovner M, Kroll J, Rovner DR, et al. Patient decision support intervention: increased consistency with decision analytic models. Med Care. 1999;37:270–284. [DOI] [PubMed] [Google Scholar]
- 11.McBride CM, Rimer BK. Using the telephone to improve health behavior and health service delivery. Patient Educ Couns. 1999;37:3–18. [DOI] [PubMed] [Google Scholar]