Article Text
PDF
Abstract
Objectives This study aimed to investigate patients’ use of electronic Patient-Reported Outcome Measures (ePROMs) and understand the demographic and clinical factors that may be correlated with patient responses to the BREAST-Q at the preoperative stage of breast cancer. The BREAST-Q is a PROM in questionnaire format, developed and validated to assess satisfaction and quality of life for breast surgery patients.The hypothesis tested is that considering disparities in geography, age and education among responders is essential for capturing a diverse patient population in future Patent-Reported Outcome Measures initiatives, examining how these characteristics are associated with Patent-Reported Outcome Measures utilisation and outcomes.
Design Quantitative descriptive study.
Setting Electronic Patient-Reported Outcome Measures were collected between 6 September 2021 and 5 September 2022 from patients recruited from an outpatient clinic at a Plastic- and Breast Surgery Department at a University Hospital in Denmark.
Participants Participants include a total of 629 Danish-speaking women diagnosed with breast cancer and scheduled for breast cancer surgery, with a final participation rate of 468.
Intervention Preoperative ePROMs and demographic data were collected between September 2020 and 2021 through patients’ secure national digital post-box.
Main outcome measures Demographic variables of both responders and non-responders were assessed using t-tests, Mann-Whitney U tests and χ2 tests. Linear regression models were employed to determine the demographic variables associated with BREAST-Q subscale scores.
Results The response rate for ePROMs was 72.5% with a median age of responders at 62 years. Older patients reported lower breast satisfaction (unadjusted coefficient bu=−0.26 (95% CI −0.44; −0.07), p=0.006) but better physical well-being (adjusted coefficient ba=0.23 (0.08; 0.37), p<0.001). Lower educational achievement was correlated with reduced breast satisfaction and psychosocial and sexual well-being; for example, patients with a master’s/doctoral level education scored 14.29 points higher in psychosocial well-being (95% CI 6.50; 22.07, p<0.001) compared with those with lower secondary education. Cohabiting patients reported psychosocial well-being scores approximately four points higher than those living alone (ba=3.91 (0.06; 7.75), p=0.046). Body mass index (BMI) was negatively associated with sexual well-being, with a 0.75-point decline per additional BMI point (ba=−0.75, (-1.12; −0.37), p<0.001).
Conclusions The present study demonstrates a positive attitude towards completing BREAST-Q as ePROMs among women diagnosed with breast cancer in the investigated region in Denmark. However, completion rates for ePROMs varied by demographic factors such as age, marital status and access to healthcare. Younger, more educated, married patients with lower BMI who lived near major cities were more likely to report better pretreatment outcomes.
- Patient Reported Outcome Measures
- Surveys and Questionnaires
- Breast surgery
- Patient Participation
Data availability statement
Data are available upon reasonable request. The data supporting the findings of this study consist of deidentified participant data.The deidentified dataset can be requested from the corresponding author, Julie Hougaard Prüsse, via ORCID 0000-0002-4570-6010. Requests for data will be considered on a case-by-case basis to ensure compliance with ethical guidelines and the General Data Protection Regulation (GDPR). Data sharing is permitted for non-commercial, research purposes only, and requires appropriate data use agreements. Additional supporting materials, including the study protocol are available upon reasonable request from the corresponding author.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
STRENGTHS AND LIMITATIONS OF THIS STUDY
Electronic Patient-Reported Outcomes (ePROMs) proved to be an innovative and feasible approach for collecting data on patients’ health conditions, health-related quality of life and functional status.
The study uses robust methods, including t-tests and linear regression, to identify demographic factors associated with outcomes.
The analysis of demographic and clinical data to identify patterns in ePROM completion can inform targeted interventions and improve patient care.
The methods used in this study do not offer comprehensive explanations for why non-responders do not participate; a study applying qualitative methods could provide deeper insights into this aspect.
This study is conducted as a single-centre study using baseline data from the first year of a longitudinal ePROM intervention, which may limit the generalisability of the findings to other settings or populations.
Introduction
A body of evidence supports the benefits of implementing Patient-Reported Outcome Measures (PROMs) in outpatient cancer clinical practice,1 2 and with the recognition of the potential of PROMs to improve healthcare, their use in clinical practice has increased rapidly.3–5 PROMs are instruments used to report the status of a patient’s health condition, health-related quality of life (HRQOL), or functional status directly from the patient, without interpretation by a clinician or anyone else.6 Often, PROMs are collected through paper questionnaires; however, the usage of electronic PROM (ePROM) systems is increasing.7 8
The assessment of PROMs is highly relevant in the cross-field of breast cancer and plastic surgery, where the goal is to support and ultimately satisfy the patient regarding physical functioning, psychosocial well-being and their perception of the aesthetic result.9 As the most prevalent cancer worldwide, breast cancer leaves many women living with physical, emotional and practical changes as survivors.10–13 These changes include affected body image, sexuality, psychosocial well-being and HRQOL.14–18 These circumstances support that patient satisfaction and HRQOL are essential outcomes for evaluating breast cancer surgery.9
One PROM designed for use in cosmetic and reconstructive breast surgery and clinical practice is BREAST-Q.19 It was developed to assess the impact and effectiveness of breast surgery by addressing patient satisfaction and aspects of HRQOL.9 However, knowledge about the clinical application of BREAST-Q is limited,20 including the feasibility of BREAST-Q as ePROMs in various healthcare settings and geographical areas.2 21
In a prospective observational cohort study, Ramírez et al (2019) found that most publications on BREAST-Q conduct postoperative evaluations and overlook the importance of the preoperative BREAST-Q module. This inhibits the availability of valuable information critical for both routine clinical practice and understanding changes in PROMs over time,22 for instance, this is particularly relevant when assessing a patient who had already experienced breast pain before undergoing surgery. Furthermore, it was noted as a limitation that social or environmental factors were not taken into account despite their direct effect on the variable.22 A study by Kebede et al (2023) investigating demographic differences in psychosocial well-being before breast reconstruction further highlighted the importance of addressing the demographic and psychosocial variation among patients diagnosed with breast cancer. The study showed an association between survey responses and various socioeconomic factors and patient demographics. The authors argued that integrating the preoperative version of BREAST-Q into active discussions with patients prior to reconstructive surgery offers a valuable baseline for understanding patients’ current body perceptions and for discussing potential postsurgical changes.23 Therefore, understanding the baseline characteristics of the population to which PROMs are applied is crucial for informing preoperative discussions and interpreting postoperative PROMs.
In 2020, the outpatient clinic for plastic and breast surgery at a large university hospital in Denmark launched a multimethod feasibility study to investigate whether BREAST-Q as ePROMs in treatment and care of women diagnosed with breast cancer is practical and sustainable in clinical practice.21 Approximately 700–800 women are diagnosed with breast cancer at the department yearly.24 Previous research has found that multiple factors influence PROM implementation and patients’ engagement in completing PROMs, such as complexities within the healthcare system, the patient’s capability, relevance to treatment and technology.25 Further research is required to understand which patients complete PROMs.25 Research on engagement with ePROMs and the determinants influencing the response rates indicates that an individual’s socioeconomic status serves not only as a predictive factor for the probability of completing ePROMs, but also for clinical outcomes.25–28 These studies have found that variables such as social deprivation and advancing age are associated with a higher likelihood of the non-completion of PROMs. This suggests that certain demographic groups may be under-represented in PROM research, potentially introducing bias that favours individuals who are more likely to report better outcomes.26–29
A recent study revealed that the region where the present university hospital is located has a large proportion of citizens facing socioeconomic challenges regarding educational level, labour market attachment and income.30 Additionally, the report indicates that these citizens have greater health problems related to mental well-being, obesity, long-term illness and multiple diseases compared with the national population on average. This appears to be particularly prominent in municipalities located at greater distances from the university hospital.30 31 Therefore, the introduction of ePROMs warrants further investigation, particularly in terms of inclusion and non-participation of patients in the region.
The hypothesis in this study was that demographic factors such as geography, age and educational level are associated with the completion rates and outcomes reported in preoperative BREAST-Q questionnaires among women scheduled for breast cancer surgery.
To enable adjustments to our ePROM intervention and consider how to capture more diverse and under-represented patient populations in future PROM initiatives, we sought to explore whether patient characteristics are associated with the utilisation of ePROMs and the outcomes reported by patients. The aim is twofold: (1) to investigate the demographic and clinical characteristics associated with the completion of preoperative BREAST-Q as ePROMs in women scheduled for breast cancer surgery at a Danish university hospital and (2) to investigate and describe demographic and clinical patterns concerning the outcomes reported in the BREAST-Q questionnaire at baseline.
Methods
Study design
The study is part of a multimethod study evaluating the feasibility of BREAST-Q as ePROMs to enable systematic follow-up in treatment and care of women diagnosed with breast cancer in a plastic and breast surgery outpatient setting.21 This substudy applied a quantitative study design. The study includes data from the first year after PROMs were introduced into the outpatient setting.21
The BREAST-Q questionnaire is recommended by the International Consortium for Health Outcomes Measurement and was designed specifically to evaluate outcomes among women undergoing breast surgery, including breast cancer surgery.32 33 BREAST-Q has modules for mastectomy, breast-conserving therapy, breast reconstruction, breast reduction and breast augmentation. All modules contain three subdomains regarding HRQOL (physical, psychosocial, and sexual well-being) and three subdomains on patient satisfaction (satisfaction with breasts, satisfaction with outcome and satisfaction with care). The modules are divided into preoperative and postoperative scales to address the subdomains and can be used independently, allowing questionnaires to be tailored to suit the specific needs of clinicians and researchers.32 Each scale is accompanied by a conversion table and is individually transformed into scores in which the overall score of the full module is not attainable. The scores are calculated through Rasch score conversion by adding the response items and converting the raw sum scale into a score from 0 to 100.33 34 For all BREAST-Q scales, a higher score indicates greater satisfaction or better quality of life, depending on the scale. If missing data are less than 50% of the scale items, the mean of the completed items is inserted.32
Patient and public involvement
A patient representative from the Danish Cancer Society was involved in the research from November 2020. The representative participated in all meetings as an equal member of the study steering group and was involved in the development of the study design, decisions on the conduct of the study, choice of methods, outcome measures and recruitment to the study. The patient representative also assisted in formulating and designing patient information and the questions regarding patient characteristics in the questionnaire. Furthermore, he contributed to disseminating knowledge about the project through the Danish Cancer Society.
Participants and setting
Patients eligible for inclusion in the multimethod study were recruited from a large outpatient plastic and breast surgery clinic at a Danish university hospital from 6 September 2021 to 5 September 2022.21 Patients newly diagnosed with breast cancer were screened for possible inclusion according to inclusion and exclusion criteria (figure 1). The patients were invited to participate in the multimethod study after their treatment trajectory was planned.21 An invitation, including a link to an information video, was sent to the patients’ secure national digital post-box (e-Boks), which is linked to Danish citizens’ personal registration numbers.35 The patient invitation included a section in which the patient could provide written consent or decline to participate in the study. Patients included in the study could withdraw from participating in the study at any time without justification and without affecting their present or future treatment.
Overview of excluded and included patients based on the criteria and the respective sample sizes (n).
The identification and invitation procedures were performed daily to include new patients over a period of 1 year. The recruitment procedure was conducted by a research assistant (JHP) in collaboration with a surgeon who specialised in plastic surgery (LBH); both were employed at the department.
Sample size
The sample size for this study was based on the number of participants who responded during the first year of data collection in the planned 3-year multimethod study. As this study presents initial data from the first year and focuses on evaluating the feasibility of using BREAST-Q as ePROMs, no formal power calculation was conducted for this single-year analysis. The sample size reflects the natural cohort of eligible participants who met the inclusion criteria during the first year.
Data collection
The multimethod study on BREAST-Q in clinical practice21 involved three different timepoints (T) for using ePROMs among women diagnosed with breast cancer (figure 2). This substudy was based on data from T1 (baseline—see figure 2).
ePROM Intervention and study flow chart. Adapted from the publication by Thestrup et al. 21 Dark boxes illustrate the intervention features. Timepoints (T) T1, T2 and T3 refer to the timely and specific questionnaires sent to the patients.
The ePROMs were prepared using the electronic data capture system REDCap, with ePROMs as output.34 Patients received a version of the BREAST-Q questionnaire that contained the following scales: satisfaction with breasts, physical well-being, psychosocial well-being and sexual well-being.32
The ePROMs were completed and stored in REDCap, which allowed surgeons and nurses, research assistants and heads of research to access their responses. If patients declined the invitation to participate in the multimethod study, the questionnaire was closed, and rejection was displayed in REDCap. An encrypted database was used to document (1) the patients’ initial therapy (neoadjuvant or surgical), (2) which patients were excluded (hence the exclusion criteria), and (3) those who had received written reminders and who had been contacted by phone (as described within the study plan for reminders and offering digital assistance).21
Patient characteristics were collected as part of the baseline questionnaire (T1). The patient characteristics included age, height, weight, body mass index (BMI), marital status, educational level and place of residence defined by municipality. The covariate of patient characteristics was purposefully selected to investigate the potential impact of patient characteristics on completion of ePROMs or on patients’ perceptions of their HRQOL and to determine whether additional interventions were warranted for specific patients.
After the section about patient characteristics, the baseline questionnaire (T1) comprised preoperative versions of the BREAST-Q scales, including satisfaction with breasts, physical well-being, psychosocial well-being and sexual well-being.32 Evaluations of T2 and T3 will be reported elsewhere.
Data analysis
Data were extracted from REDCap and the encrypted database and subsequently anonymized. The data analysis was conducted by a statistician (AM). Descriptive statistics were calculated for demographic variables for both responders and non-responders to the BREAST-Q questionnaire based on data from T1 (baseline). Depending on the normality of the numerical variables, assessed using the Shapiro-Wilk test, means (SD) or medians (IQRs) were calculated, whereas categorical variables were expressed as proportions. Differences in demographic variables were analysed using t-tests, Mann-Whitney U tests and χ2 tests.36 Furthermore, linear regression models were used to identify which of the demographic variables—age, marital status, educational level, initial therapy, BMI and municipality were associated with the subscale scores from the BREAST-Q questionnaire.
In the analysis, the 17 municipalities in which the respondents resided were grouped into four areas as defined by Statistics Denmark: (1) provincial village municipalities, (2) upland municipalities, (3) rural municipalities and (4) capital municipalities.37 Statistics Denmark also defines metropolitan municipalities; however, the region where this study was conducted does not include patients from these municipalities. Educational level was categorised according to the International Standard Classification of Education,38 and marital status was divided into three groups: (1) single (including widowed, divorced and separated), (2) cohabiting/married and (3) other/unknown.
All variables were entered into univariate and multivariable regression models to identify the demographic variables that were independently associated with the questionnaire scores. The multivariable regression model included all the above variables. RobustSEs were used to account for slight violations of model assumptions.36 Data were analysed using the Stata software package (StataCorp LLC, USA).39 The significance level was set at p<0.05, and all tests were two-tailed.
Theoretical framework
The theoretical foundation of this study is based on the concept of access to care.40 The framework by Levesque et al supports the interpretation and discussion of the results by highlighting how factors such as geographical location and digitalisation may be associated with patients’ access to ePROMs and, consequently, their access to follow-up treatment and care in a clinical setting. Levesque et al define access to care as the individual’s opportunity to identify healthcare needs, to seek healthcare services, to reach, obtain or use healthcare services, and to have the need for services fulfilled in practice. The framework suggests five dimensions of accessibility (approachability, acceptability, availability and accommodation, affordability, appropriateness) and five corresponding abilities of populations (ability to perceive, ability to seek, ability to reach, ability to pay and ability to engage). According to Levesque et al, these are determinants that are related to access to healthcare40 such as PROM initiatives.
Results
Description of the sample
A total of 629 patients were enrolled in this study (as shown in figure 1). Among these patients, 468 provided their consent, and 12 individuals later withdrew from participating in the study. This resulted in a final participation count of 456 participants, representing a response rate of 72.5%. Reasons for withdrawing from participation included not wishing to participate, lacking the energy to participate and finding the questions unsuitable.
Excluded patients
In this study, it is crucial to report on the patients who were excluded from participation to ensure a clear understanding of the sample population and the potential impact on the study’s results. Exclusions were made based on predefined criteria to maintain the quality and reliability of the data. Based on the exclusion criteria, 183 patients were excluded from the study (figure 2). The mean age of excluded patients was 76 years. The main reason for exclusion was that the patients were initially candidates for surgery but chose non-surgical treatment with letrozole aromatase inhibitor hormone therapy as primary treatment (45%). Moreover, of the excluded patients, 35% did not have a mail account with e-Boks, the national digital post-box. Other reasons for being excluded from the study were as follows: palliative treatment had been planned (nonsurgical; 12%); did not speak/understand Danish (4%); was not planned for surgical treatment for other reasons (2%); had a disability that made ePROM follow-up impossible (1%); and those who wished to receive no further information about research.
Responder versus non-responder characteristics
Table 1 presents a summary of the characteristics of the enrolled patients. The median age of the responders was 62 (range 29–94) years; 19% of the responders were scheduled for neoadjuvant therapy as initial therapy, and 81% were planned for surgical treatment upfront. Most of the responders (36%) lived in a provincial village municipality. 29% of the responders were from rural municipalities, 26% were from upland municipalities and 9% were from capital municipalities. Furthermore, 72% of the responders were cohabiting or married. In terms of education, the highest percentage of responders (33%) held a bachelor’s or equivalent degree.
Characteristics of enrolled patients
During the study period, 161 patients were characterised as non-responders, which was defined as those who declined to participate and those who never replied. The median age of non-responders was 68 (range 28–91) years. Most patients were planned for surgical treatment upfront (80%), and 20% were planned for neoadjuvant therapy as initial therapy. Of the 161 non-responders, 38% lived in provincial village municipalities, 20% in upland municipalities, 33% in rural municipalities and 9% in capital municipalities. It was not possible to obtain data on marital status and educational level from the non-responders.
BREAST-Q scores
The results of the univariate and multivariable analyses are reported in tables 2 and 3.
Results of the analysis of the four subscales in BREAST-Q
Results of the multivariable analysis of the four subscales in BREAST-Q
For the subscale ‘satisfaction with breasts’, higher age was significantly associated with a lower score in the unadjusted analysis (unadjusted coefficient bu=−0.26 (95% CI −0.44; −0.07), p=0.006). Having a short-cycle tertiary education compared with a lower secondary education (bu=7.16 (95% CI 0.6; 13.71), p=0.033) or being planned for neoadjuvant therapy before surgical therapy compared with surgery therapy upfront (bu=6.63 (95% CI 1.59; 11.66) p=0.010) were both significantly associated with a higher score, whereas coming from rural municipalities compared with provincial village municipalities was significantly associated with a lower score (bu=−5.50 (95% CI −10.75; −0.24), p=0.040). However, none of these factors were significantly associated with the subscales when combined in the multivariable model (table 3).
Concerning the subscale ‘physical well-being’, age, initial therapy and municipality category were significantly associated with the outcome in both the unadjusted and adjusted analyses. Specifically, in the adjusted analysis, individuals scored significantly higher with higher age (adjusted coefficient ba=0.23 (0.08; 0.37), p<0.001) and considerably lower if their planned treatment was neoadjuvant therapy compared with upfront surgical therapy (ba=−7.40 (-11.42; −3.37), p<0.001). Their score was also approximately four points lower if they came from an upland municipality compared with a provincial village municipality.
Educational level and marital status were significantly associated with the ‘psychosocial well-being’ subscale in both the unadjusted and the adjusted analyses. Patients who were cohabiting, compared with those living alone, had a higher average score by about four points (ba=3.91 (0.06; 7.75), p=0.046), whereas all levels of education were associated with significantly higher scores compared with the lower secondary level, ranging from 9.21 points for short-cycle tertiary education (95% CI 2.20; 16.23, p=0.010) to 14.29 for a master’s/doctoral level education (95% CI 6.50; 22.07, p<0.001) in the adjusted analysis.
Finally, educational level and BMI were associated with the subscale ‘sexual well-being’. All levels of education were associated with significantly higher scores compared with the lower secondary level, ranging between 18.19 points for an upper secondary education (95% CI 4.46, 31.92, p=0.010) and 25.40 for a master’s/doctoral level education (95% CI 10.97; 39.84, p=0.001) in the adjusted analysis. BMI was negatively associated with the subscale, with a 0.75-point decline in the score for every additional BMI point (ba=−0.75 (−1.12; −0.37), p<0.001).
Missing data
Among those who agreed to participate, the rate of missing data on the three subscales ‘satisfaction with breasts’, ‘physical well-being’ and ‘psycho-social well-being’ was 2% or less, thus no correction for missing data was conducted.
The subscale ‘sexual well-being’ was not mandatory to complete and had a substantial dropout of 43%. Participants who answered this subscale were generally younger, with a mean age of 56.5 years, cohabiting and highly educated.
Discussion
The present study investigated demographic and clinical patterns in the baseline BREAST-Q questionnaire from the first year’s data of the prospective study ‘BREAST-Q as ePROMs for systematic follow-up in women diagnosed with breast cancer’.21 Demographic and clinical characteristics were analysed using response rates and outcomes reported in the preoperative BREAST-Q module.
The general acceptance of ePROMs by patients is a key factor in successful implementation.29 41 Completion rates from other studies using BREAST-Q have been reported as low as 27%.42 The response rate in this study was 72.5%, indicating a positive attitude towards BREAST-Q as ePROMs among patients with breast cancer within the geographical region where we conducted this study. However, 64 patients (mean age: 76 years), representing 10% of eligible participants, were unable to participate in the study due to the exclusive use of e-Boks for disseminating the BREAST-Q questionnaire. Data from the Danish Agency for Digital Government indicate that among the five Danish regions the inhabitants of this particular region represent the second highest percentage of the population without e-Boks (7.6%). Notably, the report highlights a higher prevalence of individuals without access to e-Boks among older individuals, women and individuals living in rural municipalities.43
To understand and interpret the differences between responders and non-responders, as well as the clinical implications of participation in ePROMs, it is relevant to consider how patients’ individual characteristics are linked to access to healthcare services. Access to treatment and care through ePROMs may have direct health implications for patients. The introduction of ePROMs to follow-up on women planned for surgical treatment is a treatment and care initiative, where ePROMs are actively used by surgeons and nurses to offer patients person-centred follow-up based on their outcomes and the active discussion of these.21 Levesque et al (2013) describe access to healthcare as a complex notion, and analysing access to and the utilisation of healthcare is therefore not an easy task.40
The conceptual framework of access to healthcare by Levesque et al (2013) provides an evidence-based synthesis of determinants that may influence patients’ access to healthcare. Based on the framework, one can posit that the introduction of ePROMs in a clinical outpatient clinic has the potential to enhance patient care by offering a more person-centred approach. However, this improvement relies on patients having the necessary abilities to engage with the ePROM set-up provided by the department.
In this study, the arguments for the introduction of ePROMs were supported by digital access, which was an enabler for most patients, and the department’s tailored interventions, such as personalised outreach telephone calls on follow-up and digital guidance support. In terms of abilities, the introduction of ePROMs seemed well suited to patients with a high educational level and younger age, as these were characteristics of the participants. However, through the lens of the access to healthcare framework, our data showed that the introduction of ePROMs seemed to be problematic for older patients, defined as age >65, and patients with a low educational level. The framework by Levesque et al and the determinants relating to abilities, including lack of digital access, inability to engage because of high age, and the ability to perceive healthcare services may explain this (figure 3). The ability to perceive healthcare services is associated with patients’ health literacy and their capacity to perceive the service, in our case the ePROM intervention. This finding is supported by previous research which points out that low health literacy is associated with poorer health outcomes and poorer use of healthcare services.44 Studies have also identified digital health literacy as an essential factor for engagement in digital health interventions45 46—a factor which was not included in the framework by Levesque et al in 2013, but which seems highly relevant today. Digital health literacy is defined as the ability to seek, find, understand and appraise health information from electronic sources and apply the knowledge gained to address or solve a health problem.46 Examples of digital health literacy include being able to search and assess health information online, use telehealth services such as ePROMs, and communicate with healthcare providers electronically. Therefore, more consideration should be given to patients’ health literacy and digital health literacy before ePROMs are introduced.
Illustration of how patients’ completion of ePROMs is influenced by multiple determinants on access to healthcare. The figure is inspired by the access to healthcare framework by Levesque et al. ePROM, electronic Patient-Reported Outcome Measures.
The present study highlighted that geography, marital status and digital access are factors associated with the completion of ePROMs. Levesque et al’s framework suggests that other determinants such as culture, social support and income also impact access to healthcare.40 Although these factors were not addressed by our study, the framework confirms that there is no ‘one-size-fits-all’ approach in ePROM initiatives, as some patients may benefit from ePROMs as an easy means to access healthcare services, whereas others will be prevented (for various reasons) from responding to the ePROM invitation or may not be invited due to lack of digital access. The framework of access to healthcare may be useful in further studies on how to approach patients and provide follow-up for patients unable to engage in the ePROM-based follow-up.
The incorporation of diverse and under-represented patient populations is a well-known challenge in implementing PROMs in routine clinical practice.28 A systematic review of socioeconomic inequality in cancer in the Nordic countries showed a pattern across the Nordic countries in terms of socioeconomic inequality in cancer stage at diagnosis, treatment and survival.47 Pappot and Taarnhøj (2020) have argued that the introduction of tools such as ePROMs may introduce inequality based on age or social status.48 It is possible that the patients in the present study who were not assigned digital mail in an e-Boks account would have responded if BREAST-Q had been paper based: some studies have shown a >90% completion rate using PROMs.49 Conversely, other studies, including a meta-analysis, have revealed no major differences or limitations between ePROMs and paper-based PROMs.50–52 However, our results showed that the responders were younger than the non-responders, and many were well educated. In our study, the mean age of patients who could not be invited to participate because they were not assigned digital mail via e-Boks was 14 years higher than the mean age of the participants, and many of them were from rural municipalities. On this basis, we acknowledge that our study contributes to inequality based on age and social status, and we might have enhanced the response rate and achieved greater population variation by offering both the ePROM and a paper version. However, using paper versions of PROMs conflicts with the Danish digitalisation strategy of strengthening common welfare through digitalisation and increasing digital skills among the population.53 One aim of the strategy is to ensure that the population can use and benefit from digital services, and Danish regions are responsible for ensuring that disparities in healthcare are not exacerbated in interactions with the health services.53 54 In conjunction with the fact that the group of patients who did not respond or could not be invited is unrepresented in the ePROM intervention, there are implications for proactive application in clinical practice. This should lead to reflections on how to improve the quality of care for patients who do not fit into the intervention, how to reduce the digital divide and how to ensure surgical or other psychosocial follow-up for these patients. Thus, the project team will reconsider how patients are invited to complete BREAST-Q, for example, by providing digital guidance to patients at the hospital and making ePROMs available for those without e-Boks. This could be done by physically helping the patients complete the ePROM questionnaire on a tablet.
Strengths and limitations
A key strength of this study is the detailed description of the demographic and clinical characteristics that correlate with the completion of preoperative BREAST-Q as ePROMs and the reported outcomes. In line with the results from Kebede et al (2023), we identified variations in body satisfaction and perception, depending on diverse demographic factors.23 However, to our knowledge, the current study and the study by Kebede et al are the only studies describing demographic characteristics from the baseline BREAST-Q questionnaire prior to treatment. Thus, we cannot conclude whether the characteristics or patterns seen are scientifically, socially or clinically generalisable. Nevertheless, our results highlight the importance of tailored preoperative counselling that takes each patient’s demographic characteristics, perceptions and expectations into account. The framework on access to healthcare further provided nuanced perspectives on how healthcare services are perceived and operationalised on an individual level.
The benefit of using a digital platform was that ePROMs were easily distributed to participants for completion before starting neoadjuvant or surgical treatments, often within a week. A weakness of this study is that some patients were excluded from participation because PROMs were only offered electronically. Another limitation of this study is that we included only baseline data from the first year. It is possible that the completion rate will change over time. This will be further investigated in future research.
Given the lack of participation of certain patient populations in the PROM intervention,28 it is possible that the patient-reported outcomes might differ if more non-responders had replied. The inclusion criterion requiring Danish-speaking participants, regardless of ethnicity, may have excluded and limited the representation of patients from diverse ethnic backgrounds. However, this study does not include data on ethnicity. A further limitation of our study is that it was not possible to obtain information on the educational level, BMI and marital status of the non-responders.
This study is observational and exploratory in nature. While associations between independent variables and the outcomes have been reported, these results should not be interpreted as evidence of causal relationships. The models presented, both univariate and multivariable, include variables that were available, but the selection was not based on a predefined hypothesis or causal framework. As such, residual confounding or reverse causality cannot be ruled out.
Further research, with a hypothesis-driven approach and appropriate study design, is necessary to establish causal pathways.
The large sample size enhances the statistical power of the study, allowing for more reliable and generalisable findings. The high response rate further supports the robustness of the data, indicating a strong engagement from the target population. This level of participation is particularly noteworthy given the challenges often associated with obtaining high response rates in ePROM studies.
The high response rate may be associated with the fact that patients were invited to take part in a research project and that ePROMs are not part of standard clinical protocol incorporated into the electronic health record. Through the data highlighted in this work, we hypothesise a potential increase in the response rate on the integration of ePROMs into clinical practice, whereby these data are incorporated into the patient’s medical record alongside blood tests and X-rays. Conversely, an opposing perspective could posit a potential decrease in the response rate, postulating that certain patients may exhibit a heightened likelihood of completing BREAST-Q when they are engaged in research and are motivated by altruistic intentions to contribute to the broader patient community or to help researchers.55 However, the current study does not provide an explanation for the high response rate. The high level of engagement supports the representativeness of the responder cohort. The cohort is considered representative of the wider population of interest, as the demographic characteristics, including median age and regional distribution, align with those of the broader population of women diagnosed with breast cancer in the region. The inclusion of various socioeconomic factors, such as educational level and marital status, ensures that the findings are applicable to a diverse patient population. While some patients were excluded based on predefined criteria, these exclusions represented a small proportion of the overall population and are unlikely to significantly impact the generalisability of the study findings. Overall, the sample size and high response rate provide a solid foundation for the study’s conclusions, ensuring that the findings are both statistically significant and applicable to the wider population of women diagnosed with breast cancer in the region. Our study included data reported by patients prior to treatment for breast cancer but after diagnosis. When the patients completed the BREAST-Q questionnaire, they were in a life phase in which they were confronted with a breast cancer diagnosis that may have influenced the reported baseline outcomes compared with healthy women. To distinguish the cancer-related data from normative BREAST-Q data, further studies including the BREAST-Q baseline outcomes from general female populations in the region or nationally would be useful.
Conclusion
This study highlights a positive attitude towards completing BREAST-Q as ePROMs among women diagnosed with breast cancer in Denmark, with a high response rate of 72.5%. However, demographic factors such as age, marital status, educational level and BMI were significantly associated with the utilisation of ePROM completion and outcomes. Older patients were more likely to report lower breast satisfaction but better physical well-being, while lower educational achievement was correlated with lower breast satisfaction, psychosocial and sexual well-being. Marital status was linked to variations in psychosocial well-being, and higher BMI was associated with lower reported sexual well-being. The findings suggest associations between demographic characteristics and pretreatment outcome reports, with younger, more educated, married patients with lower BMI reporting better outcomes.
The study also shows specific demographic factors that are correlated with ePROM completion rates, helping to identify barriers faced by certain patient groups. This information can guide targeted interventions to improve participation and ensure equitable access to ePROMs.
The findings can help develop strategies that address healthcare access inequalities and understanding these demographic and clinical factors allows healthcare providers to tailor their approaches, leading to more personalised and effective care.
The study provides a foundation for future research on the long-term acceptability and impact of ePROMs in similar populations.
Data availability statement
Data are available upon reasonable request. The data supporting the findings of this study consist of deidentified participant data.The deidentified dataset can be requested from the corresponding author, Julie Hougaard Prüsse, via ORCID 0000-0002-4570-6010. Requests for data will be considered on a case-by-case basis to ensure compliance with ethical guidelines and the General Data Protection Regulation (GDPR). Data sharing is permitted for non-commercial, research purposes only, and requires appropriate data use agreements. Additional supporting materials, including the study protocol are available upon reasonable request from the corresponding author.
Ethics statements
Patient consent for publication
Ethics approval
The Danish Data Protection Agency (REG-154-2020) and the Scientific Ethics Review Committee of Region Zealand of Denmark (SJ-914, EMN-2021-01530) granted ethical approval. Participants gave informed consent to participate in the study before taking part.
Acknowledgments
The authors thank: patient representative from the Danish Cancer Society, KH Karlsen for representing patients and the public in the study design. Patients who have participated in the study. Clinical, research, and leading staff participating in the execution of BREAST-Q as ePROMs in clinical practice at the Department of Plastic and Breast Surgery, Zealand University Hospital, Denmark.
References
Footnotes
Contributors JHP and STH were primarily responsible for the study conception and design, data collection, visualisation, drafting and revisions of the manuscript. Guarantor is JHP. STH and VJS were responsible for funding acquisition. AM and JHP were primarily responsible for analysis and interpretation of data. LBH and VJS were responsible for the study conception and design and revisions of the manuscript. LGØ and KP were responsible for editing the manuscript. All authors have approved the final version of this manuscript for publishing and agree to be held accountable for all aspects of the work. Microsoft co-pilot was used for grammar checking on few add-ons after language editing service.
Funding This work was supported by Region Zealand, National Cancer Plan IV [grant number 20-082, 2020]. The Region Zealand has not been part of designing the study, nor was the region part of the analysis or interpretation of data.
Competing interests None declared.
Patient and public involvement Patients and/or the public were involved in the design, or conduct, or reporting, or dissemination plans of this research. Refer to the Methods section for further details.
Provenance and peer review Not commissioned; externally peer reviewed.