Abstract
Women undergoing surgery for primary breast cancer can choose between breast conserving therapy and mastectomy (with or without breast reconstruction). Patients often turn to outcomes data to help guide the decision-making process. The BREAST-Q is a validated breast surgery-specific patient-reported outcome measure that evaluates satisfaction, quality of life, and patient experience. It was originally developed for paper-and-pencil administration. However, the BREAST-Q has increasingly been administered electronically. Therefore, the aim of this study was to evaluate the psychometric properties of an electronic version of the BREAST-Q in a large online survey. Women with a history of breast cancer surgery recruited from the Love/AVON Army of Women program completed an electronic version of the BREAST-Q in addition to the Impact of Cancer Survey and PTSD Checklist. Traditional psychometric analyses were performed on the collected data. BREAST-Q data were collected from 6748 women (3497 Breast Conserving Therapy module, 1295 Mastectomy module, 1956 Breast Reconstruction module). Acceptability was supported by a high response rate (82%), low frequency of missing data (<5%), and maximum endorsement frequencies (<80%) in all but 17 items. Scale reliability was supported by high Cronbach’s α coefficients (≥0.78) and item-total correlations (range of means, 0.65–0.91). Validity was supported by interscale correlations, convergent and divergent hypotheses as well as clinical hypotheses. The electronically administered BREAST-Q yields highly reliable, clinically meaningful data for use in clinical outcomes research. The BREAST-Q can be used in the clinical setting, whether administered electronically or using paper-and-pencil, at the choice of the patient and surgeon.
Keywords: BREAST-Q, PROMS, Patient reported outcome measures, Patient reported outcomes, Psychometrics, Electronic questionnaires, Breast surgery
INTRODUCTION
Breast cancer is the second most common cancer among women in the United States, with an estimated 246,660 new diagnoses projected in 2016 [1]. Owing to advances in the field of surgical oncology, breast cancer patients and survivors now have many surgical options, which emphasizes the importance of understanding the long-term quality of life (QOL) outcomes [2]. Women undergoing surgery for primary breast cancer are increasingly making their own surgical decisions, and are often choosing between breast conserving therapy and mastectomy with or without breast reconstruction [3]. Although recent studies suggest breast conserving therapy has a survival benefit [4], patients often turn to satisfaction and QOL outcomes data to help guide the decision-making process [5, 6].
Patient-reported outcome measures (PROMs) are increasingly used in clinical research, education, and patient care. The field of PRO measurement is expanding, and the number of condition-specific scales continues to grow. As such, the COnsensus-based Standards for the selection of health Measurement INstruments initiative has suggested guidelines for the validation of PROMs [7]. Hospitals and clinics are incorporating PROMs into patient care and are more frequently capturing them in electronic medical records. It is therefore critical that electronically administered PROMs are demonstrated to have comparable psychometric validity to the paper-and-pencil version [8].
The BREAST-Q is a validated breast surgery—specific PROM that assesses patient satisfaction, QOL, and patient experience within breast surgery—specific domains [9, 10]. The BREAST-Q can provide meaningful data that can be used to support quality metrics and evidence-guided surgical practices in oncologic breast and plastic surgery [11]. It has been widely used by clinicians and researchers to help patients and surgeons better understand outcomes [9–12]. A recent literature review identified 49 publications that used the BREAST-Q as an outcome measure, including 22,457 patients who completed at least one scale of the BREAST-Q [12]. For example, the BREAST-Q was used to compare satisfaction with silicone and saline implant reconstruction after mastectomy; this analysis revealed that satisfaction with the breast was higher among those who received silicone implants [13]. Another study reported psychosocial and sexual well-being after nipple-sparing and skin-sparing mastectomies with reconstruction and found that nipple-sparing mastectomy was associated with higher QOL [14]. Other researchers used the BREAST-Q to evaluate the effect of a decision aid for different breast reconstruction options [15]. An increasing number of cancer hospitals have incorporated the BREAST-Q into clinical care. At these hospitals, physicians use the BREAST-Q in routine clinical practice to follow the progress of individual patients.
The BREAST-Q was originally developed and validated using a paper-and-pencil approach [9]. Given the uptake of the BREAST-Q in research and clinical practice, it is necessary to examine its psychometric performance when data are collected in an electronic format. The aim of this study was to examine the performance and clinical relevance of the BREAST-Q in an online community and to assess the psychometric validity of web-based completion.
METHODS
Study Design
A sample of participants were recruited from the Army of Women (AOW), an online community started in 2008 by the Dr. Susan Love Research Foundation with the goal of connecting breast cancer researchers to women with and without breast cancer [11]. The AOW Scientific Advisory Committee accepted this study, as previously described, with ethics approval obtained from the Duke University Medical Center Institutional Review Board and the Cancer Protocol Committee in June 2012 [16].
A recruitment e-mail (e-blast) was sent to AOW members, with a follow-up e-mail sent to eligible and interested participants. Women were invited to participate if they had undergone surgery for breast cancer and were able to complete the surveys in English. Participants completed a demographic survey and were directed to take either the BREAST-Q Breast Conserving Therapy, BREAST-Q Mastectomy, or BREAST-Q Breast Reconstruction postoperative module on the basis of their responses to demographic questions. Patients who completed the Breast Reconstruction module but had revision reconstruction surgery or had undergone surgery after recurrence were excluded from this validation study, as their experience was complex and could confound results. In addition to the BREAST-Q, participants also completed the Impact of Cancer Version 2 [17, 18] and the PTSD Checklist-Civilian version (PCL-C) [19, 20]. These two breast cancer-specific PROMs were used to establish construct validity via comparisons with the BREAST-Q scales.
Measures
BREAST-Q
The BREAST-Q was designed to assess the unique outcomes of breast surgery patients. The originally described core domains, which were identified through patient interviews, were physical well-being, psychosocial well-being, sexual well-being, satisfaction with breast, satisfaction with outcome, and satisfaction with care. Modules specific to different types of breast surgery were developed, with scales measuring different domain constructs. For example, the Breast Reconstruction module contains a different satisfaction with breast scale than the Breast Conserving Therapy module. Items for each scale are evaluated with multiple response options. Raw scores are added and converted to a score from 0 (worse) to 100 (best). In the original BREAST-Q development study and subsequent psychometric validation study, BREAST-Q data were collected via postal surveys sent to breast surgery patients [9, 21]. The BREAST-Q was developed using Rasch measurement theory analysis. Paper-based assessments have demonstrated high internal consistency, with a Cronbach’s α of 0.81–0.97, and test-retest reproducibility, as measured by intraclass correlation coefficients of 0.73–0.96, for each scale within the modules [9].
Impact of Cancer Version 2
The Impact of Cancer Version 2 is a 47-item questionnaire intended to assess the effect that cancer has had on the health-related QOL of long-term survivors [17, 18]. The questionnaire consists of two higher-order scales, the positive impact summary scale and the negative impact summary scale, each with four scales. The negative impact summary has the following scales: appearance concerns, body change concerns, life interference, and worry. For each item, respondents indicate his or her agreement with five response options that range from “strongly disagree” to “strongly agree.” The scale scores represent the mean of the item responses within the scale. High scores represent a higher impact; high negative impact scores are worse than low negative impact scores. The negative impact and positive impact summary score is a mean of the scale scores. The Impact of Cancer Version 2 has demonstrated high internal consistency (Cronbach’s α statistics, 0.76–0.89) when distributed by mail to survivors of early-stage breast cancer [18].
PCL-C
The PCL-C is a self-administered checklist of key symptoms of posttraumatic stress disorder (PTSD) in the civilian population and consists of 17 items that correspond to the Diagnostic and Statistical Manual of Mental Disorders 4th ed. [19]. The items address four distinct characteristics of PTSD: re-experiencing, avoidance, numbing, and hyperarousal. Respondents indicate how bothered by a symptom they have been during the past month using a 5-point scale, ranging from “not at all” (1 point) to “extremely” (5 points). Items are summed to produce a total PTSD severity score. The instrument has been used previously to assess symptoms of PTSD in breast cancer survivors [22, 23]. The PCL-C has shown high internal consistency among domestic violence victims [24], college students who have experienced trauma [20], and bone marrow transplant patients [25] when completed in person, by mail, and by telephone interview. The Cronbach’s a for the total score in these populations ranged from 0.89 to 0.94 [20, 24, 25].
Statistical Analysis
Scores for the electronically administered BREAST-Q were computed for each scale of the three procedure-specific modules on the basis of the BREAST-Q scoring tables. Scores were excluded from the analysis if any item included in the scale was missing. With two exceptions, all scales were completed by the entire sample (women who had abdominal surgery were the only patients to complete the physical well-being—abdomen and satisfaction with abdomen scales). In the evaluation of the performance of the electronically administered BREAST-Q, traditional psychometric methods were used in accordance with international instrument guidelines for data quality, reliability, and validity [10, 26].
For reliability, internal consistency was tested by computing the Cronbach’s α coefficient for each scale supported by values >0.70 [26–28]. Internal consistency was also tested by measuring the correlation of single items to the total score supported by values >0.30 [9, 28]. Acceptability was determined by examining maximum endorsement frequencies (MEFs) and missing data. For each item response, the proportion of respondents was determined, and a MEF >80% was considered a violation, in accordance with the traditional psychometric approach. For each item, the proportion of missing responses >5% was considered a violation [9, 28]. Construct, convergent, and clinical validity were tested as described below.
Construct Validity
Construct validity was evaluated by measuring interscale correlations to determine the agreement between separate but related constructs. Validity was supported by Spearman’s coefficients of 0.30–0.70 [28, 29].
Convergent Validity
The convergent validity of each scale was tested by measuring correlations between scale scores and previously validated scales (PCL-C and negative impact summary scale with appearance concerns, body change concerns, life interference, and worry scales). We measured only correlations with the IOCv2 that had items related to the BREAST-Q scales. Hypotheses were made on the basis of the constructs within each external scale as they related to the BREAST-Q constructs. For example, we hypothesized that the satisfaction with breast scale and the appearance concerns scale would be correlated, as they both measure similar surgical outcomes. Moderate correlations were supported by absolute Spearman’s coefficients of 0.30–0.70 [28]. Discriminant validity was tested by measuring correlations of scale scores between clinical groups without known differences (age, education), with low correlations supported by Spearman’s coefficients <0.30 [28].
Clinical Validity
Finally, the ability of the scales to differentiate between known clinical subgroups was tested using the Wilcoxon signed rank sum test. As adjuvant chemotherapy has been shown to negatively affect the functional and aesthetic outcomes of breast conserving therapy, we hypothesized that, among patients who underwent breast conserving therapy, scores for participants who received adjuvant chemotherapy would be lower than those for patients who received no chemotherapy [30]. We also compared scores between patients who had free and pedicle transverse rectus abdominis muscle (TRAM) flap procedures and patients who had implant-based reconstruction. Previous evidence suggests that patients are more satisfied with TRAM flap reconstruction than with implants [31]. In addition, we hypothesized that scores for patients who self-reported complications would be lower than those for patients who experienced no complications, which has been shown in previous work [6]. Complications included bleeding, open wounds, infections requiring antibiotics, scarring, seroma, infection of implants requiring removal, fat necrosis requiring removal, failed reconstruction requiring different reconstruction, partial loss of flap, and complete loss of flap. Respondents selected all complications that they experienced. We dichotomized respondents into those who reported at least one complication and those who reported no complications.
RESULTS
As previously reported, 7619 women consented to and completed the AOW survey (response rate, 82%) [16]. In total, 6748 women were included in this evaluation. Demographic data and available clinical data for the procedure-specific groups are listed in Table 1: breast conserving therapy (52%; n=3497), mastectomy (19%; n=1295), and breast reconstruction (29%; n=1956). Most women were postmenopausal (mean age, 58 years; SD, 9.4; range, 18–70), white (94%), married (71%), and had some higher education (95%). Most were treated for stage 1 or 2 breast cancer (77%) and did not have a recurrence (94%). Twenty-nine percent experienced a complication (Table 1).
Table 1.
Patient demographic and clinical characteristics
| Characteristic | All (n=6748) | Procedure-specific groups
|
||
|---|---|---|---|---|
| Breast-conserving therapy (n=3497) | Mastectomy (n=1295) | Breast reconstruction (n=1956) | ||
| Age, years, mean (SD) | 58 (9.4) | 59 (8.9) | 61 (9.2) | 55 (9.3) |
| Stage | ||||
| 1 | 3155 (47) | 1886 (54) | 385 (30) | 884 (45) |
| 2 | 2025 (30) | 986 (28) | 428 (33) | 611 (31) |
| 3 | 638 (9) | 180 (5) | 281 (22) | 177 (9) |
| 4 or metastatic disease | 90 (1) | 26 (1) | 41 (3) | 23 (1) |
| Unknown | 840 (12) | 409 (12) | 160 (12) | 261 (13) |
| Type of breast cancer | ||||
| Ductal carcinoma in situ | 2035 (30) | 1087 (31) | 291 (22) | 656 (34) |
| Invasive ductal carcinoma | 2649 (39) | 1393 (40) | 507 (39) | 748 (38) |
| Invasive lobular carcinoma | 401 (6) | 174 (5) | 85 (7) | 142 (7) |
| Other/Unknown | 1666 (25) | 843 (24) | 412 (32) | 410 (21) |
| Recurrence | ||||
| Yes | 369 (5) | 62 (2) | 132 (10) | 175 (9) |
| No | 6334 (94) | 3405 (97) | 1161 (90) | 1768 (90) |
| Complications | ||||
| Yes | 1938 (29) | 789 (23) | 397 (31) | 752 (39) |
| No | 4803 (71) | 2706 (77) | 897 (69) | 1200 (61) |
| Marital status | ||||
| Married/Significant other | 5114 (76) | 2617 (75) | 935 (72) | 1562 (80) |
| Widowed/Separated/Divorced | 1143 (17) | 600 (17) | 265 (20) | 278 (14) |
| Single, never married | 474 (7) | 272 (8) | 93 (7) | 109 (6) |
| Education | ||||
| High school | 336 (5) | 173 (5) | 85 (7) | 78 (4) |
| College, trade, university | 3639 (54) | 1878 (54) | 675 (52) | 1086(56) |
| Master/Doctoral degree | 2754 (41) | 1436 (41) | 531 (41) | 787 (40) |
| Employment | ||||
| Full-time | 2551 (38) | 1313 (38) | 386 (30) | 852 (44) |
| Part-time | 984 (15) | 504 (14) | 183 (14) | 297 (15) |
| Retired | 1629 (24) | 920 (26) | 411 (32) | 298 (15) |
| Other | 1567(23) | 747 (21) | 313 (24) | 507 (26) |
| Race | ||||
| White, Non-Hispanic | 6163 (91) | 3216 (92) | 1178 (91) | 1769 (90) |
| White, Hispanic | 87 (1) | 43 (1) | 16 (1) | 28 (1) |
| Black or African American | 16 (2) | 51 (1) | 27 (2) | 28 (1) |
| Asian or Pacific Islander | 65 (1) | 34 (1) | 16 (1) | 15 (1) |
| Other | 317 (5) | 153 (4) | 58 (4) | 106 (5) |
| Years from surgery | ||||
| <5 | 3061 (45) | 1594 (46) | 507 (39) | 960 (49) |
| 5–10 | 1663 (25) | 951 (27) | 279 (22) | 433 (22) |
| 11–15 | 969 (14) | 501 (14) | 218 (17) | 250 (13) |
| 16–20 | 480 (7) | 214 (6) | 126 (10) | 140 (7) |
| >20 | 302 (4) | 99 (3) | 126 (10) | 77 (4) |
The internal consistency of each scale score for the BREAST-Q was supported by high Cronbach’s α coefficients (0.78–0.96) and high item-total correlations (range of means, 0.65–0.91). Missing data was low for each item within the scales, with no items having >5% missing data (Table 2). Seventeen of 205 items had a MEF >80%; otherwise, the items met the criteria for acceptability.
Table 2.
Summary of classical test theory psychometrics
| Module/scale | Items, no. | Cronbach’s α coefficient | Item-total correlations, mean (range) | Inter-scale Spearman’s coefficient, range | Missing data %, range of maximums | MEF %, range of maximums |
|---|---|---|---|---|---|---|
| BREAST-Q Breast Conserving Therapy module | ||||||
| Satisfaction with Breast | 11 | 0.96 | 0.83 (0.75–0.89) | 0.22–0.57 | 0.1–1.0 | 33–68 |
| Adverse effects of radiation | 7 | 0.80 | 0.67 (0.58–0.73) | 0.25–0.39 | 0.5–1.2 | 72–83 |
| Psychosocial well-being | 10 | 0.95 | 0.85 (0.79–0.89) | 0.32–0.64 | 0.3–0.7 | 46–76 |
| Sexual well-being | 8 | 0.93 | 0.83 (0.59–0.89) | 0.30–0.64 | 0.5–1.7 | 20–38 |
| Physical well-being | 14 | 0.89 | 0.65 (0.41–0.74) | 0.21–0.36 | 0.2–1.5 | 44–81 |
| Satisfaction with information | 12 | 0.93 | 0.76 (0.70–0.83) | 0.21–0.49 | 0.2–1.9 | 41–76 |
| BREAST-Q Mastectomy module | ||||||
| Satisfaction with breast | 4 | 0.82 | 0.81 (0.73–0.86) | 0.44–0.67 | 0.2–4.9 | 31–41 |
| Psychosocial well-being | 10 | 0.95 | 0.85 (0.81–0.87) | 0.44–0.69 | 0.1–0.8 | 32–62 |
| Sexual well-being | 6 | 0.94 | 0.72 (0.52–0.82) | 0.34–0.69 | 0.3–1.5 | 26–49 |
| Physical well-being | 16 | 0.93 | 0.87 (0.78–0.93) | 0.34–0.44 | 0.2–1.5 | 43–83 |
| BREAST-Q Breast Reconstruction module | ||||||
| Satisfaction with breast | 16 | 0.96 | 0.78 (0.69–0.84) | 0.33–0.68 | 0.4–2.3 | 33–59 |
| Satisfaction with outcome | 7 | 0.89 | 0.79 (0.75–0.84) | 0.32–0.68 | 1.2–1.5 | 43–82 |
| Psychosocial well-being | 10 | 0.96 | 0.85 (0.78–0.89) | 0.28–0.73 | 0.2–0.7 | 35–64 |
| Sexual well-being | 6 | 0.94 | 0.87 (0.75–0.93) | 0.31–0.73 | 0.5–1.6 | 29–43 |
| Physical well-being | 16 | 0.92 | 0.69 (0.49–0.79) | 0.22–0.53 | 0.3–1.5 | 44–83 |
| Physical well-being (abdomen) | 8 | 0.88 | 0.74 (0.61–0.84) | 0.28–0.53 | 1.8–2.8 | 48–75 |
| Satisfaction with abdomen | 3 | 0.78 | 0.83 (0.82–0.85) | 0.28–0.47 | 1.9–2.5 | 35–50 |
| Satisfaction with information | 15 | 0.95 | 0.76 (0.68–0.81) | 0.22–0.33 | 0.2–1.7 | 33–77 |
| Common scales | ||||||
| Satisfaction with surgeon | 12 | 0.97 | 0.87 (0.76–0.92) | — | 0.2–0.7 | 72–88 |
| Satisfaction with medical team | 7 | 0.96 | 0.91 (0.87–0.93) | — | 1.4–2.0 | 77–85 |
| Satisfaction with office staff | 7 | 0.96 | 0.91 (0.87–0.93) | — | 1.3–1.8 | 77–84 |
MEF, maximum endorsement frequency.
Construct validity was supported by interscale correlations with moderate to high Spearman’s coefficients (breast conserving therapy, 0.21–0.64; mastectomy, 0.34–0.69; breast reconstruction, 0.22–0.73) (Table 2).
Hypotheses of relationships between the breast conserving therapy scales and external instrument scales were supported by moderate correlations (0.31–0.62). One correlation was lower than expected: body change concerns and satisfaction with breast (0.29). Hypotheses of correlations between the mastectomy scales and external instrument scales were supported by moderate correlations (0.46–0.70). Hypotheses of correlations between the breast reconstruction scales and other instrument scales were supported by moderate correlations (0.31–0.71). Three correlations were lower than expected: worry and satisfaction with outcome, worry and satisfaction with information, and negative impact summery scale and satisfaction with info (0.14 and 0.18). Discriminant validity was supported with low coefficients in each module with two variables (0.00–0.22) (Table 3).
Table 3.
Convergent and discriminant construct validity of the BREAST-Q
| Module/scale | PCL-C [19] | Negative impact scale | Negative impact of cancer subscales [17]
|
Age | Education | |||
|---|---|---|---|---|---|---|---|---|
| Appearance concerns | Body change concerns | Life interference | Worry | |||||
| BREAST-Q Breast Conserving Therapy module | ||||||||
| Satisfaction with breast | .29 | .34† | .57† | .29‡ | .26 | .18 | .03^ | .13^ |
| Adverse effects of radiation | .36† | .36† | .31† | .35† | .3 | .24 | .17^ | .08^ |
| Psychosocial well-being | .55† | .57† | .62† | .48† | .50† | .39† | .21^ | .09^ |
| Physical well-being | .37 | .35† | .27 | .36† | .32† | .25 | .18^ | .09^ |
| Sexual well-being | .50† | .54† | .61† | .47† | .47 | .35 | .14^ | .11^ |
| Satisfaction with information | .31 | .33† | .41 | .26 | .28 | .22† | .05^ | .03^ |
| BREAST-Q Mastectomy module | ||||||||
| Satisfaction with breast | .47 | .53† | .64† | .46† | .46 | .37 | .15^ | .13^ |
| Psychosocial well-being | .62† | .67† | .70† | .54† | .60† | .51† | .22^ | .10^ |
| Physical well-being | .53 | .51† | .37 | .53† | .46† | .40 | .17^ | .13^ |
| Sexual well-being | .54† | .62† | .66† | .53† | .56 | .48 | .10^ | .13^ |
| BREAST-Q Breast Reconstruction module | ||||||||
| Satisfaction with breast | .34 | .39† | .58† | .35† | .32 | .23 | .00^ | .05^ |
| Satisfaction with outcome | .31† | .34† | .47† | .32 | .30† | .18‡ | .03^ | .03A |
| Psychosocial well-being | .59† | .63† | .71† | .54† | .56† | .42† | .20^ | .06A |
| Physical well-being | .50 | .45† | .36 | .46† | .42† | .33 | .15^ | .11^ |
| Sexual well-being | .50† | .54† | .66† | .46† | .48 | .35 | .05^ | .07^ |
| Physical well-being (abdomen) | .36 | .39† | .31 | .43† | .34† | .27 | .11^ | .12^ |
| Satisfaction with abdomen | .33 | .37† | .46† | .38† | .27 | .25 | .08^ | .09^ |
| Satisfaction with information | .23 | .21‡ | .26 | .18 | .18 | .14‡ | .01^ | .08^ |
Shown are absolute values by Spearman correlation. All correlations were included for completeness. Hypotheses tested are indicated with the symbols described below. PCL-C, PTSD Checklist-Civilian Version.
Correlations were consistent with predictions.
Correlations were lower than predicted.
Discriminant correlations were consistent with predictions.
Clinical hypotheses are summarized in Table 4. Differences between scores for women who reported any complications were significant for all scales in every procedure-specific module (p<0.01). Among the women who received breast conserving therapy, those who received adjuvant therapy had lower scores for all scales (p<0.01) except adverse effects of radiation (p=0.48). Women who received either a free or pedicle TRAM flap had higher scores in the breast reconstruction scales than women who received implants (p<0.01), with the exception of the satisfaction with information scale (p=0.11).
Table 4.
Clinical validity hypothesis testing
| Module/scale | Any complication Med, (IQR) | No complication Med, (IQR) | p | Adjuvant therapy Med, (IQR) | No adjuvant therapy Med, (IQR) | p | Implant reconstructi on Med, (IQR) | TRAM reconstructi on Med, (IQR) | p |
|---|---|---|---|---|---|---|---|---|---|
| BREAST-Q Breast Conserving Therapy module | n= 789 | n= 2706 | n= 1662 | n= 1651 | |||||
| Satisfaction with breast | 55 (41–71) | 66 (50–85) | <0.01 | 62 (44–77) | 66 (50–80) | <0.01 | — | — | — |
| Adverse effects of radiation | 89 (73–100) | 100 (80–100) | <0.01 | 89 (80–100) | 89 (80–100) | 0.48 | — | — | — |
| Psychosocial well-being | 76 (61–100) | 87 (65–100) | <0.01 | 78 (63–100) | 82 (68–100) | <0.01 | — | — | — |
| Physical well-being | 72 (62–86) | 78 (67–92) | <0.01 | 75 (64–86) | 77 (67–92) | <0.01 | — | — | — |
| Sexual well-being | 52 (40–63) | 58 (46–69) | <0.01 | 55 (42–66) | 60 (48–69) | <0.01 | — | — | — |
| Satisfaction with information | 64 (52–79) | 75 (58–100) | <0.01 | 70 (56–84) | 74 (58–100) | <0.01 | |||
| BREAST-Q Mastectomy module | n= 397 | n= 897 | |||||||
| Satisfaction with breast | 49 (38–61) | 52 (44–71) | <0.01 | — | — | — | — | — | — |
| Psychosocial well-being | 63 (52–78) | 70 (57–87) | <0.01 | — | — | — | — | — | |
| Physical well-being | 72 (57–86) | 78 (69–92) | <0.01 | — | — | — | — | — | — |
| Sexual well-being | 41 (26–54) | 45 (30–60) | 0.014 | — | — | — | — | — | — |
| BREAST-Q Breast Reconstruction module | n= 752 | n= 1200 | n= 1547 | n= 669 | |||||
| Satisfaction with breast | 60 (46–74) | 63 (51–76) | <0.01 | — | — | — | 59 (47–71) | 70 (57–86) | <0.01 |
| Psychosocial well-being | 70 (65–87) | 76 (59–92) | <0.01 | — | — | — | 72 (55–92) | 79 (63–100) | <0.01 |
| Physical well-being | 75 (63–86) | 78 (69–86) | <0.01 | — | — | — | 76 (66–86) | 79 (69–92) | <0.01 |
| Sexual well-being | 48 (35–64) | 52 (37–64) | <0.01 | — | — | — | 49 (35–64) | 55 (41–68) | <0.01 |
| Physical well-being (abdomen) | 80 (60–90) | 90 (71–100) | <0.01 | ||||||
| Satisfaction with abdomen | 61 (44–71) | 71 (52–84) | <0.01 | — | — | — | — | — | — |
| Satisfaction with information | 59 (48–74) | 64 (48–77) | <0.01 | — | — | — | 63 (48–74) | 65 (48–77) | 0.11 |
Shown are the results of comparison of scores between clinical groups. p values were derived using the Wilcoxon signed rank sum test.
Med, Median; IQR, Interquartile Range; TRAM, transverse rectus abdominis muscle.
Comparisons were made for Adjuvant Therapy vs No Adjuvant Therapy for the Breast Conserving Therapy module only.
Comparisons were made for Implant Reconstruction vs TRAM Reconstruction for the Breast Reconstruction module only.
DISCUSSION
In this study, the Breast Conserving Therapy, Mastectomy, and Breast Reconstruction modules of the BREAST-Q were shown to have reliability, validity, and clinical relevance when administered electronically. Building on the expansive use of the BREAST-Q in clinical research, quality measurement, and clinical practice, these analyses support the validity and reliability of the electronically administered version. The e-version of the BREAST-Q scales were shown to be valid in a breast cancer population and to fulfill the criteria for published psychometric guidelines.
Our study highlights the benefits of using an electronic survey. Sending surveys via email allows patients to complete them at the time and place of their choosing, and results in high response rates. Patients can answer questions about satisfaction with breast and sexual well-being, for example, in the privacy of their own home. A large sample of patients can be included and results are delivered instantly and accurately. Surveys can be tailored to individual patient experiences or surgery type like we have demonstrated in this study. Breast surgeons can incorporate the BREAST-Q with a variety of scales and procedures into their practice, with the click of a button.
Further study of the electronic BREAST-Q will be directed toward evaluating test-retest reliability, as well as responsiveness. We also anticipate considering the utility of computer adaptive testing for the BREAST-Q. The computer adaptive testing version will allow patients to answer fewer but more-specific and-informative questions, giving clinicians and researchers information in a more efficient manner.
This study has several strengths. Performing recruitment through the AOW resulted in a large sample size and a diverse population not limited to a single institution or region. This methodology allowed for the evaluation of subgroups based on type of surgery, treatment, and time from surgery. Another strength was the study’s inclusion of only cancer patients, which eliminated confounders associated with cancer such as anxiety and depression. This study used scales that had previously been validated for the paper form of the survey, and, therefore, we were able to compare psychometric properties between the paper and electronic formats.
A limitation of this study was that participation was voluntary and only interested participants of the AOW were offered the survey. Moreover, the AOW participants were mostly white, educated, upper-middle-class women, which is reflected in our population. Another limitation relates to the length of time that had passed since breast cancer treatment for some participants. For those who were longer out from their surgery, some recall bias may be present for the scales asking about satisfaction with information and surgeon. Furthermore, clinical data were provided by the participants and could not be confirmed through medical records. As we did not evaluate test-retest reliability in this study, we cannot comment on consistency over time for electronic administration. In addition, the sample was cross-sectional, and we were therefore unable to demonstrate the responsiveness of the BREAST-Q to clinical change [13, 14].
Conclusions
We have shown that the electronically administered BREAST-Q is psychometrically valid. For each scale, evidence demonstrated validity, reliability, and acceptability. These scales can be used in clinical practice for patient evaluation and education or as primary outcomes in quality and outcomes research. The BREAST-Q continues to be a valuable tool to inform and support clinical care, comparative effectiveness research, and quality improvement for patients with breast cancer.
HIGHLIGHTS.
The BREAST-Q is a validated breast surgery-specific patient-reported outcome measure that has been validated in electronic format.
Psychometric evaluation of the BREAST-Q provided evidence that the BREAST-Q is acceptable, valid and reliable in an electronic distribution.
The BREAST-Q can be administered electronically in clinical outcomes research and yields clinically meaningful data.
The BREAST-Q can be used in the clinical setting, whether administered electronically or using paper-and-pencil, at the choice of breast cancer patients and their surgeon.
Abbreviations
- AOW
Army of Women
- MEF
maximum endorsement frequency
- PCL-C
PTSD Checklist-Civilian version
- PROM
patient-reported outcome measure
- PTSD
posttraumatic stress disorder
- QOL
quality of life
- TRAM
transverse rectus abdominis muscle
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
Disclosures
Drs. Cano, Klassen, and Pusic are co-developers of the BREAST-Q and receive royalties when it is used in industry-sponsored, for-profit clinical trials.
Ethical Approval
The Army of Women Scientific Advisory Committee accepted this study with ethics approval obtained from the Duke University Medical Center Institutional Review Board and the Cancer Protocol Committee in June 2012.
References
- 1.U.S. Breast Cancer Statistics. 2016 doi: 10.3322/caac.21332. Available from: http://www.breastcancer.org/symptoms/understand_bc/statistics. [DOI] [PubMed]
- 2.Effects of radiotherapy and surgery in early breast cancer. An overview of the randomized trials. Early Breast Cancer Trialists’ Collaborative Group. The New England journal of medicine. 1995;333:1444–55. doi: 10.1056/NEJM199511303332202. [DOI] [PubMed] [Google Scholar]
- 3.Morrow M, Jagsi R, Alderman AK, Griggs JJ, Hawley ST, Hamilton AS, et al. Surgeon recommendations and receipt of mastectomy for treatment of breast cancer. Jama. 2009;302:1551–6. doi: 10.1001/jama.2009.1450. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Hwang ES, Lichtensztajn DY, Gomez SL, Fowble B, Clarke CA. Survival after lumpectomy and mastectomy for early stage invasive breast cancer: the effect of age and hormone receptor status. Cancer. 2013;119:1402–11. doi: 10.1002/cncr.27795. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Pusic AL, Chen CM, Cano S, Klassen A, McCarthy C, Collins ED, et al. Measuring quality of life in cosmetic and reconstructive breast surgery: a systematic review of patient-reported outcomes instruments. Plastic and reconstructive surgery. 2007;120:823–37. doi: 10.1097/01.prs.0000278162.82906.81. discussion 38–9. [DOI] [PubMed] [Google Scholar]
- 6.Cano S, Klassen AF, Scott A, Thoma A, Feeny D, Pusic A. Health outcome and economic measurement in breast cancer surgery: Challenges and opportunities. Expert Review of Pharmacoeconomics and Outcomes Research. 2010;10:583–94. doi: 10.1586/erp.10.61. [DOI] [PubMed] [Google Scholar]
- 7.Mokkink LB, Terwee CB, Patrick DL, Alonso J, Stratford PW, Knol DL, et al. The COSMIN checklist for assessing the methodological quality of studies on measurement properties of health status measurement instruments: an international Delphi study. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation. 2010;19:539–49. doi: 10.1007/s11136-010-9606-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Schulenberg SE, Yutrzenka BA. The equivalence of computerized and paper-and-pencil psychological instruments: implications for measures of negative affect. Behavior research methods, instruments, & computers: a journal of the Psychonomic Society, Inc. 1999;31:315–21. doi: 10.3758/bf03207726. [DOI] [PubMed] [Google Scholar]
- 9.Pusic AL, Klassen AF, Scott AM, Klok JA, Cordeiro PG, Cano SJ. Development of a new patient-reported outcome measure for breast surgery: the BREAST-Q. Plastic and reconstructive surgery. 2009;124:345–53. doi: 10.1097/PRS.0b013e3181aee807. [DOI] [PubMed] [Google Scholar]
- 10.Cano SJ, Klassen AF, Scott AM, Cordeiro PG, Pusic AL. The BREAST-Q: further validation in independent clinical samples. Plastic and reconstructive surgery. 2012;129:293–302. doi: 10.1097/PRS.0b013e31823aec6b. [DOI] [PubMed] [Google Scholar]
- 11.Atisha DM, Rushing CN, Samsa GP, Locklear TD, Cox CE, Shelley Hwang E, et al. A national snapshot of satisfaction with breast cancer procedures. Annals of surgical oncology. 2015;22:361–9. doi: 10.1245/s10434-014-4246-9. [DOI] [PubMed] [Google Scholar]
- 12.Cohen WA, Mundy LR, Ballard TN, Klassen A, Cano SJ, Browne J, et al. The BREAST-Q in surgical research: A review of the literature 2009–2015. Journal of plastic, reconstructive & aesthetic surgery: JPRAS. 2016;69:149–62. doi: 10.1016/j.bjps.2015.11.013. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.McCarthy CM, Klassen AF, Cano SJ, Scott A, Vanlaeken N, Lennox PA, et al. Patient satisfaction with postmastectomy breast reconstruction: a comparison of saline and silicone implants. Cancer. 2010;116:5584–91. doi: 10.1002/cncr.25552. [DOI] [PubMed] [Google Scholar]
- 14.Wei CH, Scott AM, Price AN, Miller HC, Klassen AF, Jhanwar SM, et al. Psychosocial and Sexual Well-Being Following Nipple-Sparing Mastectomy and Reconstruction. The breast journal. 2016;22:10–7. doi: 10.1111/tbj.12542. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Luan A, Hui KJ, Remington AC, Liu X, Lee GK. Effects of A Novel Decision Aid for Breast Reconstruction: A Randomized Prospective Trial. Annals of plastic surgery. 2016;76(Suppl 3):S249–54. doi: 10.1097/SAP.0000000000000722. [DOI] [PubMed] [Google Scholar]
- 16.Atisha DM, Locklear TD, Rogers UA, Rushing CN, Samsa GP, Abernethy AP. Partnering with engaged patients accelerates research. Journal of surgical oncology. 2014;109:504–5. doi: 10.1002/jso.23515. [DOI] [PubMed] [Google Scholar]
- 17.Zebrack BJ, Ganz PA, Bernaards CA, Petersen L, Abraham L. Assessing the impact of cancer: development of a new instrument for long-term survivors. Psycho-oncology. 2006;15:407–21. doi: 10.1002/pon.963. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Crespi CM, Ganz PA, Petersen L, Castillo A, Caan B. Refinement and psychometric evaluation of the impact of cancer scale. Journal of the National Cancer Institute. 2008;100:1530–41. doi: 10.1093/jnci/djn340. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Weathers FW, Litz BT, Keane TM, Palmieri PA, Marx BP, Schnurr PP. The PTSD Checklist for DSM-5 (PCL-5) Scale available from the National Center for PTSD at. 2013 www.ptsd.va.gov.
- 20.Ruggiero KJ, Del Ben K, Scotti JR, Rabalais AE. Psychometric properties of the PTSD Checklist-Civilian Version. Journal of traumatic stress. 2003;16:495–502. doi: 10.1023/A:1025714729117. [DOI] [PubMed] [Google Scholar]
- 21.Cano SJ, Klassen AF, Scott AM, Cordeiro PG, Pusic AL. The BREAST-Q: further validation in independent clinical samples. Plastic & Reconstructive Surgery. 2012;129:293–302. doi: 10.1097/PRS.0b013e31823aec6b. [DOI] [PubMed] [Google Scholar]
- 22.Andrykowski MA, Cordova MJ, Studts JL, Miller TW. Posttraumatic stress disorder after treatment for breast cancer: prevalence of diagnosis and use of the PTSD Checklist-Civilian Version (PCL-C) as a screening instrument. Journal of consulting and clinical psychology. 1998;66:586–90. doi: 10.1037//0022-006x.66.3.586. [DOI] [PubMed] [Google Scholar]
- 23.Cordova MJ, Andrykowski MA, Kenady DE, McGrath PC, Sloan DA, Redd WH. Frequency and correlates of posttraumatic-stress-disorder-like symptoms after treatment for breast cancer. Journal of consulting and clinical psychology. 1995;63:981–6. doi: 10.1037//0022-006x.63.6.981. [DOI] [PubMed] [Google Scholar]
- 24.Khadra C, Wehbe N, Lachance Fiola J, Skaff W, Nehme M. Symptoms of post-traumatic stress disorder among battered women in Lebanon: an exploratory study. Journal of interperson violence. 2015;30:295–313. doi: 10.1177/0886260514534774. [DOI] [PubMed] [Google Scholar]
- 25.Smith MY, Redd W, DuHamel K, Vickberg SJ, Ricketts P. Validation of the PTSD Checklist-Civilian Version in survivors of bone marrow transplantation. Journal of traumatic stress. 1999;12:485–99. doi: 10.1023/A:1024719104351. [DOI] [PubMed] [Google Scholar]
- 26.Aaronson N, Alonso J, Burnam A, Lohr KN, Patrick DL, Perrin E, et al. Assessing health status and quality-of-life instruments: attributes and review criteria. Quality of life research: an international journal of quality of life aspects of treatment, care and rehabilitation. 2002;11:193–205. doi: 10.1023/a:1015291021312. [DOI] [PubMed] [Google Scholar]
- 27.Cronbach LJ. Coefficient alpha and the internal structure of tests. 1951;16:297–334. [Google Scholar]
- 28.Cano SJ, Browne JP, Lamping DL, Roberts AH, McGrouther DA, Black NA. The Patient Outcomes of Surgery-Hand/Arm (POS-Hand/Arm): a new patient-based outcome measure. Journal of hand surgery (Edinburgh, Scotland) 2004;29:477–85. doi: 10.1016/j.jhsb.2004.06.002. [DOI] [PubMed] [Google Scholar]
- 29.McDowell I. Measuring Health; A Guide to Rating Scales and Questionnaires. New York: Oxford University Press Inc.; 2006. [Google Scholar]
- 30.Ozmen T, Polat AV, Polat AK, Bonaventura M, Johnson R, Soran A. Factors affecting cosmesis after breast conserving surgery without oncoplastic techniques in an experienced comprehensive breast center. The surgeon: journal of the Royal Colleges of Surgeons of Edinburgh and Ireland. 2015;13:139–44. doi: 10.1016/j.surge.2013.12.005. [DOI] [PubMed] [Google Scholar]
- 31.Alderman AK, Kuhn LE, Lowery JC, Wilkins EG. Does patient satisfaction with breast reconstruction change over time? Two-year results of the Michigan Breast Reconstruction Outcomes Study. Journal of the American College of Surgeons. 2007;204:7–12. doi: 10.1016/j.jamcollsurg.2006.09.022. [DOI] [PubMed] [Google Scholar]