Abstract
Importance
Breast reconstruction has the potential to improve a person’s body image and quality of life but has important risks. Variations in who undergoes breast reconstruction have led to questions about the quality of patient decisions.
Objective
To assess the quality of patient decisions about breast reconstruction.
Design, Setting, and Participants
A prospective, cross-sectional survey study was conducted from June 27, 2012, to February 28, 2014, at a single, academic, multidisciplinary oncology clinic among women planning to undergo mastectomy for stage I to III invasive ductal or lobular breast cancer, ductal carcinoma in situ, or prophylaxis.
Exposures
Mastectomy only and mastectomy with reconstruction.
Main Outcome and Measures
Knowledge, as ascertained using the Decision Quality Instrument; preference concordance, based on rating and ranking of key attributes; and decision quality, defined as having knowledge of 50% or more and preference concordance.
Results
During the 20-month period, 214 patients were eligible, 182 were approached, and 32 missed. We enrolled 145 patients (79.7% enrollment rate), and received surveys from 131 patients (72.0% participation rate). Five participants became ineligible. The final study population was 126 patients. Among the 126 women in the study (mean [SD] age, 53.2 [12.1] years), the mean (SD) knowledge score was 58.5% (16.2%) and did not differ by treatment group (mastectomy only, 55.2% [15.0%]; mastectomy with reconstruction, 60.5% [16.5%]). A total of 82 of 123 participants (66.7%) had a calculated treatment preference of mastectomy only; 39 of these women (47.6%) underwent mastectomy only. A total of 41 participants (32.5%) had a calculated treatment preference of mastectomy with reconstruction; 36 of these women (87.8%) underwent mastectomy with reconstruction. Overall, 52 of 120 participants (43.3%) made a high-quality decision. In multivariable analysis, white race/ethnicity (odds ratio [OR], 2.72; 95% CI, 1.00-7.38; P = .05), having private insurance (OR, 1.61; 95% CI, 1.35-1.93; P < .001), having a high school education or less (vs some college) (OR, 4.84; 95% CI, 1.22-19.21; P = .02), having a college degree (vs some college) (OR, 1.95; 95% CI, 1.53-2.49; P < .001), and not having a malignant neoplasm (eg, BRCA carriers) (OR, 3.13; 95% CI, 1.25-7.85; P = .01) were independently associated with making a high-quality decision.
Conclusions and Relevance
A minority of patients undergoing mastectomy in a single academic center made a high-quality decision about reconstruction. Shared decision making is needed to support decisions about breast reconstruction.
This cross-sectional survey study assesses the quality of patient decisions about breast reconstruction.
Key Points
Question
What is the quality of patient decisions about breast reconstruction after mastectomy?
Findings
In this cross-sectional survey study of 126 women, a minority of participants (43.3%) made a high-quality decision, defined as having knowledge of at least half of the important facts and undergoing treatment concordant with one’s personal preferences.
Meaning
Decisions about breast reconstruction after mastectomy could be improved.
Introduction
Breast reconstruction after mastectomy has the potential to improve a person’s body image and quality of life but has important risks. Variations in who undergoes reconstruction, by geography, race/ethnicity, and insurance status, have led to questions about the quality of decision making about the procedure. The decision about whether to undergo breast reconstruction is a preference-sensitive one, in that the optimal choice depends in part on personal goals. For such choices, decision quality has been defined as the extent to which the decision is informed and consistent with patients’ preferences.
It remains unclear, however, how well decisions about breast reconstruction reflect patients’ informed preferences. Most prior research has been retrospective, evaluating decisions months or even years after treatment. In those studies, most patients who underwent mastectomy reported being advised of reconstruction by their clinicians. Their knowledge about the pros and cons of reconstruction, however, was limited. Few studies of reconstruction decisions have elicited patients’ preferences specific to breast reconstruction or explored how well those preferences match treatment choices.
We sought to evaluate the quality of decisions about breast reconstruction after mastectomy. We were specifically interested in how often patients comprehended the major pros and cons of treatment alternatives and received treatment concordant with personal preferences.
Methods
Study Design
This single-center, prospective cross-sectional study was conducted in the breast clinic of the North Carolina Cancer Hospital, a public, National Cancer Institute–designated comprehensive cancer center. The clinic had 4 surgical oncologists and approximately 500 new patients with early-stage breast cancer annually. Study procedures were approved by an advisory board of 2 bioethicists and 2 breast cancer survivors, and the University of North Carolina Institutional Review Board. It was also registered with clinicaltrials.gov (NCT01488357). Participants provided written informed consent.
Study Population
We enrolled women 21 years of age or older who were planning to undergo mastectomy for stage I to III invasive ductal or lobular breast cancer, ductal carcinoma in situ, or prophylaxis. We excluded patients with stage IV disease, other cancers, and those with an impaired capacity to make a decision. We excluded patients who could not read or speak English because study measures were in English only. We planned a sample size of 118 participants, to have 77% power to detect a 10-point difference in quality of life, the study’s primary end point.
Enrollment
Enrollment occurred from June 27, 2012, to February 28, 2014. We identified potential participants by screening clinic schedules and confirmed eligibility at a multidisciplinary conference. A research assistant approached eligible patients, described the study, and provided detailed written study information and a consent form. She advised patients that participation was optional and would not affect their care. We attempted to approach all eligible patients and enroll them in the clinic. Patients who were missed in the clinic or who asked to be enrolled later were contacted by mail.
Enrolled participants were given a printed questionnaire to be completed in the clinic or at home. Those who did not return the questionnaire in 2 weeks were called by telephone. Those who did not return the questionnaire in 4 weeks received another copy by mail. Participants who completed the questionnaire before surgery received a $25 gift card. Those who did not complete the questionnaire before surgery were considered ineligible.
Variables
The questionnaire covered demographics, knowledge, preferences, involvement in decision making, quality of life, and body image. It was evaluated with cognitive interviews with 5 healthy individuals and 5 breast cancer survivors and was revised appropriately. The revised questionnaire was pilot-tested with 3 healthy individuals. We collected medical, treatment, and demographic data from the patients’ medical records.
Demographics
We asked questions about educational attainment, marital status, race/ethnicity, employment, and income.
Knowledge
Participants completed all scales of the Decision Quality Instrument. The knowledge scale contains 9 validated, multiple-choice questions about recovery, number of surgical procedures, flaps vs implants, complication risk, radiotherapy effects, surveillance, evidence about satisfaction after reconstruction, and risk of recurrence. A previous study has reported results for the knowledge items.
Preferences
We elicited preferences about 4 attributes of the treatment options: having a breast shape after mastectomy with or without clothes, complication risk, number of surgical procedures, and recovery time. Patients had identified these attributes as important to their decisions in prior qualitative and quantitative work. We considered using the Decision Quality Instrument preferences scale, which includes need for a prosthesis, body image, and sexuality. However, in prior work, body image and sexuality were not key preferences associated with treatment choice. We thought that the prosthesis attribute was directly related to and overlapped with the existing attribute, “having a breast shape with or without clothes.” During measure development, comprehension of ranking scales was higher with 4, rather than 6, attributes.
We used 3 common preference elicitation techniques: rating, ranking, and stated preference. Participants rated the importance of each attribute to their decision on a scale from 0 to 5 (with 0 as least important and 5 as most important). They then ranked the attributes by importance. We used both rating and ranking because they have complementary advantages and disadvantages. Rating is comprehensible, allows for small differences, and does not depend on comprehension of every attribute. However, it permits the same rating for multiple attributes and ceiling effects. Ranking has the advantage of forcing tradeoffs but can result in false precision and can be harder to comprehend.
Calculated Treatment Preference
We used rating and ranking to calculate which treatment was preferred, using approaches that have been described elsewhere. For a given patient, if her rating for having a breast shape was higher than for all other attributes, the calculated treatment preference was mastectomy with reconstruction. If the rating for having a breast shape was lower than for all other attributes, the calculated treatment preference was mastectomy only.
If the rating for having a breast shape was neither higher nor lower than all other ratings, we then used rankings. If the ranking for having a breast shape was first or second, the calculated treatment preference was mastectomy with reconstruction. Otherwise, the calculated treatment preference was mastectomy only. This approach errs on the side of finding that a woman preferred reconstruction. Some might argue that if a patient does not rate having a breast shape higher than all other attributes and does not rank that attribute first, her calculated treatment preference should be mastectomy only. That approach is somewhat more stringent, in that it errs on the side of finding that a woman preferred mastectomy only. We tested that more stringent threshold, defining calculated treatment preference as mastectomy with reconstruction, only if the ranking for having a breast shape was first; otherwise, it was mastectomy only.
Stated Treatment Preference and Treatment Received
A multiple-choice item asked which treatment the participant intended to have. We then examined which treatment was received (mastectomy only or mastectomy with reconstruction). Because clinicians at this site generally delayed reconstruction in patients receiving radiotherapy after mastectomy, we adjusted the definition of treatment received. Specifically, if a patient underwent mastectomy only and received radiotherapy after mastectomy, and her stated treatment preference was mastectomy with reconstruction, we defined her treatment received as mastectomy with reconstruction, assuming she would eventually undergo delayed reconstruction. Otherwise, treatment received was actual treatment within 1 year.
Involvement in Decision Making
The Decision Quality Instrument was used to measure involvement in decision making.
Calculations of Knowledge Score, Decision Quality, and Involvement Score
Knowledge Score
We computed a knowledge score for each participant by dividing the number of correct knowledge questions by 9. Missing responses were considered incorrect. Details on the knowledge findings have been reported separately.
Decision Quality
For a decision to be considered high quality, a patient needed to comprehend her treatment alternatives and receive treatment most concordant with her calculated preference (based on rating and ranking). Thus, we calculated how many participants had a knowledge score of 50% or greater and received their calculated treatment preference. We based this approach on the assessment by Hersch et al of “informed choice.” We conducted a sensitivity analysis for the knowledge cutoff.
Involvement Score
We computed an individual involvement score by assigning 1 point for each of 4 involvement questions (“yes” reconstruction was described as an option, “a lot/some” discussion of advantages, “a lot/some” discussion of disadvantages, and “yes” the clinician asked her preference).
Statistical Analyses
We calculated descriptive summary statistics for demographic variables and compared patients with a stated treatment preference of mastectomy only with patients who had a stated treatment preference of mastectomy with reconstruction, using 2-tailed t tests and the Fisher exact test. Seven participants had missing data, so we excluded them from analyses, rather than imputing values. Analyses were performed with SAS, version 9.4 (SAS Inc). P < .05 was considered significant.
Preferences
For rating data, we calculated mean (SD) values for each attribute. We compared responses by stated treatment preference using the Jonckheere-Terpstra test. For ranking data, we calculated the percentage of participants who ranked each attribute as the most important. We compared rankings by stated treatment preference, using the Fisher exact test. We performed separate analyses of patients with or without a malignant neoplasm (eg, BRCA1/2 [OMIM 113705/600185] carriers). Those results were similar, so they are not reported separately.
Preference Concordance
We estimated the concordance between calculated treatment preference and treatment received using a κ statistic.
Factors Associated With a High-Quality Decision
We used generalized estimating equation modeling with a logistic link to identify potential factors (ie, demographic, clinical, and involvement) associated with a high-quality decision. The generalized estimating equation models perform logistic regression while accounting for variability or clustering by surgeon. All variables were candidates for inclusion in a multivariable model. Employment status was strongly associated with income and insurance, so we excluded it from the final model, resulting in a better fit.
Results
Study Population
During the 20-month period, 214 patients were eligible, 182 were approached, and 32 missed. We enrolled 145 patients (79.7% enrollment rate), and received surveys from 131 patients (72.0% participation rate). Nearly all patients chose to complete surveys at home rather than in the clinic. Five participants became ineligible (had surgery elsewhere or completed the surveys after surgery). The final study population was 126 patients (Table 1). Participants were relatively educated (104 of 125 [83.2%] with some college or more), had higher incomes (62 of 121 [51.2%] with income ≥$60 000), and were insured (70 [55.6%] with private insurance). A total of 51 patients (40.5%) underwent immediate reconstruction, 40 patients (31.7%) had adjuvant radiotherapy, and 32 patients (25.4%) had adjuvant chemotherapy. Among the 37 eligible patients who declined enrollment, the mean age was 53 years, 17 were nonwhite (46%), and 8 (22%) had immediate reconstruction.
Table 1. Demographic and Treatment Characteristics of Study Sample.
| Characteristic | Stated Treatment Preferenceb,c | P Value | ||
|---|---|---|---|---|
| Total (N = 126)a |
Mastectomy Only (n = 41) |
Mastectomy With Reconstruction (n = 78) |
||
| Age, mean (SD), y | 53.2 (12.1) | 58.7 (12.7) | 49.6 (10.3) | <.001 |
| Educational level | ||||
| ≤High school | 21/125 (16.8) | 10/40 (25.0) | 10 (12.8) | .26 |
| Some college | 44/125 (35.2) | 12/40 (30.0) | 28 (35.9) | |
| College graduate or more | 60/125 (48.0) | 18/40 (45.0) | 40 (51.3) | |
| Marital status | ||||
| Married or committed | 82/125 (65.6) | 22/40 (55.0) | 56 (71.8) | .10 |
| Single, divorced, separated, or widowed | 43/125 (34.4) | 18/40 (45.0) | 22 (28.2) | |
| Race | ||||
| White | 95/124 (76.6) | 28/39 (71.8) | 62 (79.5) | .08 |
| Black | 23/124 (18.5) | 11/39 (28.2) | 11 (14.1) | |
| Other or multiracial | 6/124 (4.8) | 0 | 5 (6.4) | |
| Ethnicity | ||||
| Hispanic | 6/122 (4.9) | 1/37 (2.7) | 5 (6.4) | .66 |
| Not Hispanic | 116/122 (95.1) | 36/37 (97.3) | 73 (93.6) | |
| Employment status | ||||
| Working full time | 50/124 (40.3) | 13/39 (33.3) | 34 (43.6) | .55 |
| Working part time or temporary leave | 19/124 (15.3) | 6/39 (15.4) | 11 (14.1) | |
| Not working | 55/124 (44.4) | 20/39 (51.3) | 33 (42.3) | |
| Annual household income, $ | ||||
| <30 000 | 37/121 (30.6) | 16/39 (41.0) | 19/75 (25.3) | .41 |
| 30 000-59 999 | 22/121 (18.2) | 7/39 (17.9) | 15/75 (20.0) | |
| 60 000-100 000 | 24/121 (19.8) | 6/39 (15.4) | 15/75 (20.0) | |
| >100 000 | 38/121 (31.4) | 10/39 (25.6) | 26/75 (34.7) | |
| Primary insurance, No. (%) | ||||
| No insurance | 8 (6.3) | 1 (2.4) | 7 (9.0) | .04 |
| Medicaid only | 9 (7.1) | 4 (9.8) | 3 (3.8) | |
| Medicare or Tricare | 39 (31.0) | 18 (43.9) | 19 (24.4) | |
| Private insurance | 70 (55.6) | 18 (43.9) | 49 (62.8) | |
| Diagnosis | ||||
| No malignant neoplasm | 17 (13.5) | 0 (0.0) | 17 (21.8) | .01 |
| Ductal carcinoma in situ | 21 (16.7) | 6 (14.6) | 14 (17.9) | |
| Invasive ductal carcinoma | 88 (69.8) | 35 (85.4) | 47 (60.3) | |
| Surgical treatment | ||||
| Unilateral mastectomy | 70 (55.6) | 28 (68.3) | 37 (47.4) | .03 |
| Bilateral mastectomy | 56 (44.4) | 13 (31.7) | 41 (52.6) | |
| Had adjuvant radiotherapy | 40 (31.7) | 17 (41.5) | 20 (25.6) | .10 |
| Had adjuvant chemotherapy | 32 (25.4) | 10 (24.4) | 17 (21.8) | .82 |
| Had immediate breast reconstruction | 51 (40.5) | 2 (4.9) | 48 (61.5) | <.001 |
| Knowledge score, mean (SD), % | 58.5 (16.2) | 55.2 (15.0) | 60.5 (16.5) | .10 |
| Calculated treatment preference | ||||
| Mastectomy only | 82/123 (66.7) | 38/40 (95.0) | 39 (50.6) | <.001 |
| Mastectomy with reconstruction | 41/123 (33.3) | 2/40 (5.0) | 38 (49.4) | |
Not all patients had data on all demographics.
Data are presented as number/total number (percentage) of patients unless otherwise indicated.
Only 119 patients gave stated treatment preference.
Knowledge
The detailed findings on knowledge have been reported previously. The mean (SD) knowledge score was 58.5% (16.2%), which did not differ by treatment (mastectomy only, 55.2% [15.0%]; mastectomy with reconstruction, 60.5% [16.5%]). A total of 88 participants (69.8%) had a knowledge score of 50% or greater and were therefore considered to be informed. For the question on risk of major complications in the first 2 years, 18 participants (14.3%) answered correctly that the risk was 16% to 40%. All but 1 person who responded incorrectly underestimated the risk.
Preferences
Rating
Participants’ ratings (Table 2) showed that they had high concern about risk of complications, with a mean (SD) importance rating of 4.9 (0.3) of 5, which did not differ by stated treatment preference (mastectomy only, 5.0 [0.2]; mastectomy with reconstruction, 4.9 [0.4]). Women whose stated treatment preference was mastectomy with reconstruction placed greater value on having a breast shape with or without clothes (mean [SD] rating, 4.4 [1.0]) than did women whose stated treatment preference was mastectomy only (mean [SD] rating, 2.8 [1.8]) (P < .001). Women whose stated treatment preference was mastectomy with reconstruction were also less concerned than women whose stated treatment preference was mastectomy only about recovery time (mean [SD] rating, 3.9 [1.1] vs 4.5 [0.9]) and number of surgical procedures (mean [SD] rating, 3.9 [1.1] vs 4.5 [0.9]).
Table 2. Rating and Ranking of Attributes, by Stated Treatment Preferencea.
| Measure | Total (N = 126) |
Stated Treatment Preference | P Valueb | |
|---|---|---|---|---|
| Mastectomy Only (n = 41) |
Mastectomy With Reconstruction (n = 78) |
|||
| Ratings, mean (SD) | ||||
| Appearance | 3.8 (1.5) | 2.8 (1.8) | 4.4 (1.0) | <.001 |
| No. of surgical procedures | 4.0 (1.2) | 4.3 (1.3) | 3.9 (1.1) | .03 |
| Recovery time | 4.1 (1.0) | 4.5 (0.9) | 3.9 (1.1) | .01 |
| Risk of complications | 4.9 (0.3) | 5.0 (0.2) | 4.9 (0.4) | .07 |
| Rankings, No. (%)c | ||||
| Appearance | 23/116 (19.8) | 0 | 23/76 (30.3) | <.001 |
| No. of surgical procedures | 8/116 (6.9) | 4/40 (10.0) | 4/76 (5.3) | |
| Recovery time | 9/116 (7.8) | 7/40 (17.5) | 2/76 (2.6) | |
| Risk of complications | 76/116 (65.5) | 29/40 (72.5) | 47/76 (61.8) | |
A total of 7 of the 126 patients did not state a treatment preference, and 3 had missing rating or ranking data.
The Jonckheere-Terpstra test was used for the ratings; the Fisher exact test was used for the rankings.
Patients who ranked the attribute most important.
Although mean ratings were generally consistent with stated treatment preference, some individual ratings were not. For example, among 41 women whose stated treatment preference was mastectomy only, 15 rated having a breast shape highly (4 or 5 of 5). Among 78 patients whose stated treatment preference was mastectomy with reconstruction, 13 rated having a breast shape in the 1 to 3 range.
Ranking
Most participants (76 of 116 [65.5%]) ranked complication risk as the most important attribute (Table 2). Among 40 participants whose stated treatment preference was mastectomy only, no one ranked having a breast shape as most important, 7 (17.5%) ranked recovery time as most important, and 29 (72.5%) ranked complication risk as most important. In contrast, among 76 participants whose stated treatment preference was mastectomy with reconstruction, 23 (30.3%) ranked having a breast shape as most important, 4 (5.3%) ranked number of surgical procedures, 2 (2.6%) ranked recovery time, and 47 (61.8%) ranked complication risk as most important.
Calculated Treatment Preference
Based on rating and ranking, 82 of 123 participants (66.7%) had a calculated treatment preference of mastectomy only. Using the more stringent threshold for ranking, 104 of 123 participants (84.6%) had a calculated treatment preference of mastectomy only.
Preference Concordance
Among the 82 participants whose ratings and rankings favored mastectomy only, 39 (47.6%) underwent mastectomy only and therefore had preference-concordant care (Table 3). Among the 41 participants whose ratings and rankings favored mastectomy with reconstruction, 36 (87.8%) underwent mastectomy with reconstruction and therefore had preference-concordant care. Overall concordance was 61.0% (κ = 0.29; 95% CI, 0.15-0.42).
Table 3. Preference Concordance: Calculated Treatment Preference Based on Rating and Ranking vs Treatment Receiveda.
| Calculated Treatment Preference | Treatment Receivedb | Total | |
|---|---|---|---|
| Mastectomy Only | Mastectomy With Reconstruction | ||
| Mastectomy only | 39c | 43 | 82 |
| Mastectomy with reconstruction | 5 | 36c | 41 |
| Totald | 44 | 79c | 123 |
Concordance: (39 + 36)/123 = 61.0%; κ = 0.29 (95% CI, 0.15-0.42).
Patients who had radiotherapy and stated treatment preference of reconstruction assumed to have delayed reconstruction.
Indicates concordant treatment.
Does not include 3 patients who were missing rating or ranking data.
Decision Quality
A minority of participants made a high-quality decision (52 of 120 [43.3%]; 6 had missing knowledge or preference data). In multivariable analysis, white race/ethnicity (odds ratio [OR], 2.72; 95% CI, 1.00-7.38; P = .05), having private insurance (OR, 1.61; 95% CI, 1.35-1.93; P < .001), having a high school education or less (vs some college) (OR, 4.84; 95% CI, 1.22-19.21; P = .02), having a college degree (vs some college) (OR, 1.95; 95% CI, 1.53-2.49; P < .001), and not having a malignant neoplasm (OR, 3.13; 95% CI, 1.25-7.85; P = .01) were independently associated with making a high-quality decision (Table 4). Sensitivity analysis was performed for the knowledge criterion. With the 60% criterion, 34 of 120 participants (28.3%) made a high-quality decision. With the 70% criterion, 16 of 120 (13.3%) made a high-quality decision.
Table 4. Factors Associated With Making a High-Quality Decisiona.
| Independent variable | Univariable P Value | Multivariable model | |
|---|---|---|---|
| OR (95% CI) | P Value | ||
| Age, 10-y increment | .57 | 1.00 (0.93-1.08) | .96 |
| Race/ethnicity | |||
| White vs nonwhiteb | .01 | 2.72 (1.00-7.38) | .05 |
| Educational level | |||
| ≤High school vs some collegeb | .05 | 4.84 (1.22-19.21) | .02 |
| College degree vs some collegeb | <.001 | 1.95 (1.53-2.49) | <.001 |
| Employment statusc | |||
| Full-time vs other | .05 | NA | NA |
| Income, $ | |||
| ≥60 000 vs <60 000 | .001 | 1.95 (0.93-4.10) | .08 |
| Primary insurance | |||
| Private vs otherb | <.001 | 1.61 (1.35-1.93) | <.001 |
| Cancer stage | |||
| No malignant neoplasm vs stage 3b | <.001 | 3.13 (1.25-7.85) | .01 |
| DCIS vs stage 3 | .60 | 1.29 (0.57-2.94) | .54 |
| Stage 1 vs stage 3 | .10 | 1.82 (0.56-5.90) | .32 |
| Stage 2 vs stage 3 | .31 | 1.57 (0.65-3.80) | .32 |
| Involvement score (1 point) | .85 | 0.89 (0.76-1.04) | .15 |
Abbreviations: DCIS, ductal carcinoma in situ; NA, not applicable; OR, odds ratio.
The analysis was performed on 120 patients because 6 patients had missing data.
Significant factors.
Employment status was excluded from the multivariable model because of strong correlation with income and insurance.
Discussion
In this sample of patients undergoing mastectomy, a minority of patients made a high-quality decision about breast reconstruction. Specifically, 43.3% of patients had adequate knowledge and underwent treatment concordant with their preferences. Most of the suboptimal decision quality resulted from preference discordance; a smaller proportion had inadequate knowledge. Much of the preference discordance was due to patients undergoing breast reconstruction despite having preferences more consistent with having mastectomy only.
Our findings confirm those of prior retrospective work describing moderately informed decision making about breast cancer treatments. In a study of patients who had undergone mastectomy approximately 2 years before, patients generally received the treatment (reconstruction or not) they preferred but had major knowledge deficits. In a population-based cohort of patients 1 year after treatment, patient knowledge was moderate, at about 50%. Many patients have reported less decision involvement than they preferred, and minority women have reported receiving inadequate information.
A substantial number of patients in our study appeared to undergo more surgery than they preferred, based on their values for key attributes. This pattern of possible overuse may be associated with patients’ strong concern about complications and low knowledge about the actual risk. It may also be associated with more discussion of the advantages than the disadvantages of breast reconstruction. Some patients received less surgery than they apparently preferred. This outcome was relatively uncommon but suggests issues with access to reconstruction.
Our finding of limited preference concordance is novel because, to our knowledge, few studies have prospectively elicited preferences about reconstruction or concordance with treatment. Most studies have examined preferences about involvement, and some have examined treatment motivations, which are associated with, but not equivalent to, preferences. In one retrospective study, the goals of using one’s own tissue and waking up after mastectomy with reconstruction under way were associated with reconstruction. The goal of avoiding foreign material was associated with mastectomy only.
Strengths and Limitations
Our method for eliciting preferences has strengths and limitations. We based the preference measures on prior qualitative and quantitative work about concerns that are most important to reconstruction decisions. The measures used rating and ranking to address each technique’s limitations and were highly comprehensible. They were not validated, however, and 1 attribute (risk of complications) did not significantly discriminate between treatments (Table 2). The remaining attributes, however, did discriminate between treatments.
We used a somewhat novel approach to evaluating preference concordance. Alternative approaches include comparing stated preference with treatment received or calculating preferred treatment using aggregate multivariable models. Our approach avoids some inherent biases of stated preference approaches. However, for calculated treatment preference, we did assign a somewhat arbitrary cutoff (using the first- or second-ranked attribute), so we erred on the side of preference concordance. Other attributes besides the 4 that we used may be relevant to this decision.
Making a high-quality decision was more common among women who were white, were privately insured, had a high school education or less, or had a college degree. These findings suggest socioeconomic influences on decisions about breast reconstruction, consistent with findings of prior studies. We are uncertain, however, why high school education was associated with decision quality. Participants undergoing mastectomy for prophylactic reasons made higher-quality decisions than did women who had a malignant neoplasm. This finding may be due to the longer time frame for decisions about prophylactic mastectomy and reconstruction.
We defined high-quality decisions as those made with knowledge of at least half of the key facts and concordant with preferences. Some may ask why the definition includes knowledge, contending that if a patient values having a breast shape above all other attributes, then regardless of her knowledge, reconstruction is her preferred treatment. However, knowledge is often necessary for people to settle on preferences. For example, a patient who does not understand the experience of reconstruction may rate having a breast shape highly, but on fully understanding the procedure, she may reduce its importance. The minimal amount of comprehension that one needs to be informed is subject to debate. We chose half of the key facts as our criterion for knowledge because it is intuitive, has been used before for this purpose, and errs on the side of categorizing patients as knowledgeable. Ultimately, the ideal approach to assessing decision quality remains unclear and would benefit from further research and discussion.
This study took place at a single academic institution among a small cohort of relatively educated patients. Thus, its generalizability is limited. Assessment of decision quality in a larger, diverse sample, including community practices, would be valuable.
The finding of limited decision quality for breast reconstruction has implications for clinical care and policy. Surgeons should discuss the advantages and disadvantages of reconstruction. They should ensure understanding of the risk of complications because knowledge about this risk appears to be particularly low. They should assess how strongly a patient feels about having a normal breast shape in and out of clothes, weighing that against the downsides of additional surgery. Decision aids and decision coaching have proven efficacy, and several breast reconstruction decision aids now exist. Adoption of shared decision making in practice has been limited, however, and needs to become a higher priority for surgeons for the decision about breast reconstruction.
Conclusions
A minority of patients undergoing mastectomy in our study made a high-quality decision about breast reconstruction. Much of the deficit in decision quality was due to patients undergoing reconstruction despite having preferences more consistent with undergoing mastectomy only. Shared decision making, including the use of decision aids, is needed for breast reconstruction.
References
- 1.Lee C, Sunu C, Pignone M. Patient-reported outcomes of breast reconstruction after mastectomy: a systematic review. J Am Coll Surg. 2009;209(1):123-133. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Morrow M, Li Y, Alderman AK, et al. . Access to breast reconstruction after mastectomy and patient perspectives on reconstruction decision making. JAMA Surg. 2014;149(10):1015-1021. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Christian CK, Niland J, Edge SB, et al. . A multi-institutional analysis of the socioeconomic determinants of breast reconstruction: a study of the National Comprehensive Cancer Network. Ann Surg. 2006;243(2):241-249. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Tseng JF, Kronowitz SJ, Sun CC, et al. . The effect of ethnicity on immediate reconstruction rates after mastectomy for breast cancer. Cancer. 2004;101(7):1514-1523. [DOI] [PubMed] [Google Scholar]
- 5.Joslyn SA. Patterns of care for immediate and early delayed breast reconstruction following mastectomy. Plast Reconstr Surg. 2005;115(5):1289-1296. [DOI] [PubMed] [Google Scholar]
- 6.Alderman AK, Hawley ST, Janz NK, et al. . Racial and ethnic disparities in the use of postmastectomy breast reconstruction: results from a population-based study. J Clin Oncol. 2009;27(32):5325-5330. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.O’Connor AM, Llewellyn-Thomas HA, Flood AB. Modifying unwarranted variations in health care: shared decision making using patient decision aids. Health Aff (Millwood). 2004;(suppl variation):VAR63-VAR72. [DOI] [PubMed] [Google Scholar]
- 8.Elwyn G, O’Connor A, Stacey D, et al. ; International Patient Decision Aids Standards (IPDAS) Collaboration . Developing a quality criteria framework for patient decision aids: online international Delphi consensus process. BMJ. 2006;333(7565):417. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Sepucha KR, Fowler FJ Jr, Mulley AG Jr. Policy support for patient-centered care: the need for measurable improvements in decision quality. Health Aff (Millwood). 2004;(suppl variation):VAR54-VAR62. [DOI] [PubMed] [Google Scholar]
- 10.Frisell A, Lagergren J, de Boniface J. National study of the impact of patient information and involvement in decision-making on immediate breast reconstruction rates. Br J Surg. 2016;103(12):1640-1648. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Morrow M, Mujahid M, Lantz PM, et al. . Correlates of breast reconstruction: results from a population-based study. Cancer. 2005;104(11):2340-2346. [DOI] [PubMed] [Google Scholar]
- 12.Morrow M, White J, Moughan J, et al. . Factors predicting the use of breast-conserving therapy in stage I and II breast carcinoma. J Clin Oncol. 2001;19(8):2254-2262. [DOI] [PubMed] [Google Scholar]
- 13.Chen JY, Malin J, Ganz PA, et al. . Variation in physician-patient discussion of breast reconstruction. J Gen Intern Med. 2009;24(1):99-104. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 14.Sue GR, Lannin DR, Au AF, Narayan D, Chagpar AB. Factors associated with decision to pursue mastectomy and breast reconstruction for treatment of ductal carcinoma in situ of the breast. Am J Surg. 2013;206(5):682-685. [DOI] [PubMed] [Google Scholar]
- 15.Lee CN, Belkora J, Chang Y, Moy B, Partridge A, Sepucha K. Are patients making high-quality decisions about breast reconstruction after mastectomy? Plast Reconstr Surg. 2011;127(1):18-26. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Brennan ME, Spillane AJ. Uptake and predictors of post-mastectomy reconstruction in women with breast malignancy–systematic review. Eur J Surg Oncol. 2013;39(6):527-541. [DOI] [PubMed] [Google Scholar]
- 17.Lee CN, Dominik R, Levin CA, et al. . Development of instruments to measure the quality of breast cancer treatment decisions. Health Expect. 2010;13(3):258-272. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Lee CN, Ubel PA, Deal AM, et al. . How informed is the decision about breast reconstruction after mastectomy? a prospective, cross-sectional study. Ann Surg. 2016;264(6):1103-1109. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Fagerlin A, Pignone M, Abhyankar P, et al. . Clarifying values: an updated review. BMC Med Inform Decis Mak. 2013;13(suppl 2):S8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Pignone MPHK, Howard K, Brenner AT, et al. . Comparing 3 techniques for eliciting patient values for decision making about prostate-specific antigen screening: a randomized controlled trial. JAMA Intern Med. 2013;173(5):362-368. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Hersch J, Barratt A, Jansen J, et al. . Use of a decision aid including information on overdetection to support informed choice about breast cancer screening: a randomised controlled trial. [published correction appears in Lancet. 2015;385(9978):1622]. Lancet. 2015;385(9978):1642-1652. [DOI] [PubMed] [Google Scholar]
- 22.Hawley ST, Fagerlin A, Janz NK, Katz SJ. Racial/ethnic disparities in knowledge about risks and benefits of breast cancer treatment: does it matter where you go? Health Serv Res. 2008;43(4):1366-1387. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lantz PM, Janz NK, Fagerlin A, et al. . Satisfaction with surgery outcomes and the decision process in a population-based sample of women with breast cancer. Health Serv Res. 2005;40(3):745-767. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 24.Katz SJ, Lantz PM, Janz NK, et al. . Patient involvement in surgery treatment decisions for breast cancer. J Clin Oncol. 2005;23(24):5526-5533. [DOI] [PubMed] [Google Scholar]
- 25.Hawley ST, Griggs JJ, Hamilton AS, et al. . Decision involvement and receipt of mastectomy among racially and ethnically diverse breast cancer patients. J Natl Cancer Inst. 2009;101(19):1337-1347. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Lee CN, Chang Y, Adimorah N, et al. . Decision making about surgery for early-stage breast cancer. J Am Coll Surg. 2012;214(1):1-10. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.Manne SL, Topham N, Kirstein L, et al. . Attitudes and decisional conflict regarding breast reconstruction among breast cancer patients. Cancer Nurs. 2016;36(6):427-436. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Sepucha K, Ozanne EM. How to define and measure concordance between patients’ preferences and medical treatments: a systematic review of approaches and recommendations for standardization. Patient Educ Couns. 2010;78(1):12-23. [DOI] [PubMed] [Google Scholar]
- 29.Elwyn G, Miron-Shatz T. Deliberation before determination: the definition and evaluation of good decision making. Health Expect. 2010;13(2):139-147. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Slovic P. The construction of preference. Am Psychol. 1995;50(5):364-371. doi: 10.1037/0003-066X.50.5.364 [DOI] [Google Scholar]
- 31.Sepucha K, Ozanne E, Silvia K, Partridge A, Mulley AG Jr. An approach to measuring the quality of breast cancer decisions. Patient Educ Couns. 2007;65(2):261-269. [DOI] [PubMed] [Google Scholar]
- 32.Ubel PA. Beyond Comprehension: Figuring Out Whether Decision Aids Improve People’s Decisions. Princeton, NJ: Princeton University Press; 2013. [Google Scholar]
- 33.Stacey D, Légaré F, Col NF, et al. . Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2014;1(1):CD001431. [DOI] [PubMed] [Google Scholar]
- 34.Hawley ST, Newman L, Griggs JJ, Kosir MA, Katz SJ. Evaluating a decision aid for improving decision making in patients with early-stage breast cancer. Patient. 2016;9(2):161-169. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 35.Sherman KA, Shaw LK, Winch CJ, et al. ; BRECONDA Collaborative Research Group . Reducing decisional conflict and enhancing satisfaction with information among women considering breast reconstruction following mastectomy: results from the BRECONDA randomized controlled trial. Plast Reconstr Surg. 2016;138(4):592e-602e. [DOI] [PubMed] [Google Scholar]
- 36.Causarano N, Platt J, Baxter NN, et al. . Pre-consultation educational group intervention to improve shared decision-making for postmastectomy breast reconstruction: a pilot randomized controlled trial. Support Care Cancer. 2015;23(5):1365-1375. doi: 10.1007/s00520-014-2479-6 [DOI] [PubMed] [Google Scholar]
- 37.The Option Grid Collaborative. Option Grid decision aid: breast reconstruction after surgery for cancer: options. Dartmouth Institute for Health Policy and Clinical Practice website. http://optiongrid.org/option-grids/grid-landing/52. Accessed January 1, 2017.