Background
Breast cancer (BC) is the most common cancer in women worldwide. According to WHO, the burden of BC is increasing in less developed countries where mortality rates are higher due to late diagnosis and lack of treatment facilities [
1]. In Argentina, approximately 30–40% of BCs are diagnosed at advanced stages of the disease (stages III and IV) [
2], and the survival rate of BC is 68.2%, below the 85% considered an international benchmark [
3]. In addition, inequalities in diagnosis are observed within the different health subsystems whereby, according to RITA hospital database, private centres detect most cases at clinical stages 0, I and II, whilst public hospitals do so at stages III and IV [
4] (p.36).
Treatment developments in the last decades, in particular hormonal therapy (HT) for women with oestrogen positive receptors, which accounts for ~ 70% of all BC cases, has proven effective in reducing the risk of recurrence and in extending survival from the disease: in postmenopausal women, 5 years of the anti-oestrogen drug Tamoxifen (TAM) reduces the risk of recurrence by about a half and mortality by about 30%, and Aromatase Inhibitors (AI) reduce recurrence by about two-thirds and mortality rate by around 40%, during 10 years after initiation of treatment [
5,
6].
In Argentina, BC is the cancer with the highest incidence considering both sexes, followed by colorectum and prostate cancer. According to GLOBOCAN (2018) [
7] the (ASR) incidence is 73, and (ASR) mortality is 18 cases × 100.000. Although mortality rates for BC in Argentina have followed a decreasing trend since 1997 [
8] (p.31), it still occupies the second highest in South America after Uruguay.
In the province of Córdoba, the second most populated province with 3.308.876 inhabitants in the last census (2010) [
9], BC has an (ASR) incidence of 65.8 and mortality of 21.4 × 100.000, whereas in the capital city both rates are higher 77.1 and 23.7 respectively [
10]. According to epidemiological regional data collected by the National Institute of Cancer (2016), Córdoba is located in the Centre Region - alongside Buenos Aires, Entre Ríos, Santa Fe and the city of Buenos Aires (Federal capital) - a region that has concentrated a higher BC mortality rate (18.1) than the national average (17.4) [
11]. Socio-demographic data for the province of Córdoba indicates that the population group most affected by BC (65+) is 23.4% of the population in 2016, and female life expectancy, according to the last census (2010), is 79.2 years, in both cases showing higher values than that for Argentina [
12]. Also, an 8.7% of the population of Córdoba has unmet basic needs (lower than national level at 12,5%), and 1.3% of women are classified as illiterate (lower than national average) [
12]. However, a recent ecology study on sociodemographic determinants associated to the spatial distribution of BC in the province, concluded that urbanisation was inversely associated to BC incidence, whereas deprivation (measured as index of unmet basic needs per households) showed a direct relationship [
13]. Other indicators associated to BC such as a good supply of health services in the capital (see Study setting below) that has traditionally served the demand of nearby provinces, two University teaching hospitals with specialisation in oncology, together with an ageing female population and a higher than average BC mortality rate make Córdoba a relevant case to explore. Scholars have long emphasised the fragmented nature of the Argentinian health system, which comprises three main subsystems (public, private, and social security) with scarce synergy amongst them. Deficiencies in cancer control have been observed not only in terms of providers and resource management [
14‐
16], but also in terms of the lack of a National Cancer Programme [
17,
18], and inequalities in access to diagnosis and treatments [
19‐
21]. The health system in Córdoba is broadly organised as follows: The majority of the working-age population has the social security health insurance system (
Obra Social), each insurance plan being organised according to the occupation of the beneficiary and are administered by different workers unions. Within this subsystem are also the provincial insurance plans for civil servants, and a Comprehensive Medical Assistance Program for retired people (
Programa de Asistencia Médica Integral, PAMI). A second group of high-income earners have private health insurance (
Medicina pre-paga) offered by employers or contracted on an individual basis. Finally, the public sector subsystem, for uninsured people, is offered free of charge and financed with resources from the provincial budget and national funds for specific health programmes. According to data from 2010, 69% of women were insured by the social security (49.73%), private health (17.44%), or state plan (1.83%); whilst a 31% were uninsured [
22].
Oncological treatments and cancer drugs approved by national protocols are covered free of charge in Argentina through the compulsory medical programme (
Programa Médico Obligatorio - PMO) [
23], which applies to all health insurers including the public sector that provides drugs (approved by a protocol) through provincial health ministries. Despite these provisions to guarantee free access to oncological drugs, a study based on a survey of oncologists prescribing adjuvant BC treatment in 2008, concluded that there were considerable disparities between what oncologists thought was an ideal treatment and what they actually could prescribe to patients due to different restrictions. Revealingly, only 40% were satisfied with the hormonal treatment given [
24]. This raises questions about accessibility to treatment in a context of, in principle, universal drug coverage.
Inequalities related to BC treatments outcomes have been extensively documented in the US and European countries, where modifiable social factors such as ethnicity, literacy, doctor-patient’s communication, socio-economic status (SES), drug accessibility, and health system provider amongst others, have been variously identified as drivers for different outcomes between affluent and disadvantaged groups [
25]. Understanding how these different factors interact in complex ways is relevant to ensure an equitable access to HT treatment, especially if we consider that the benefits associated with HT imply a long-term process, as recent guidelines recommend a 10-year course of therapy. The significance of the latter has led authors, such as Beryl and others [
26], to differentiate patients’ decision-making process into acute treatments (surgery, radiotherapy and chemotherapy) and chronic care (HT), because of the irreversible/reversable nature of the decision, and the passive/active role of treatment administration respectively. Much of the quantitative and qualitative research on HT treatments have focused on patients’ perceptions and experiences in relation to adherence to treatments. Yet there is a dearth of analysis focusing on what factors influence oncologists’ prescribing practices, the challenges they face and how they overcome them. In addition, oncologists’ gender is a variable that has rarely been explored in terms of interactions with patients and decision-making patterns, although it is known from studies on physicians more broadly, that female doctors tend to deliver a more patient-centred style of communication [
27].
Hence, to help fill this gap in the literature, the aim of this study was to explore the factors influencing oncologists’ prescribing practices of HT for women with early and advanced BC in the city of Córdoba. The research questions the study explored were: 1) What biological and non-biological factors influence oncologists prescribing HT?; 2) How does the oncologist’s gender affect HT prescribing practices?; 3) How does the health system provider affect oncologists' prescribing practices?
Methods
Study setting
The study was undertaken in the city of Córdoba, capital of the homonymous province, where cancer treatment is provided by different specialised services available through the three health subsystems. These comprise the following: 1) the institute of oncology (public sector) concentrates services on clinical oncology and radiotherapy for the province, and provides oncological drugs to patients there assisted and by referral from other provincial/university hospitals. 2) oncology services provided by 4 private hospitals and approximately 14 clinics with various levels of complexity that have contracts with the different social security and private health insurance plans. The difference between these providers is that the latter tend to offer the most prestigious, state-of-the-art hospitals and clinics, many of which are owned by the providers themselves. It is also worth emphasising that the patients assisted in each of these subsystems often move across services making the boundaries between them not so demarcated. On the one hand, the public institution grants access to any person free of charge, so a patient in the private/social insurance sector can seek a second opinion for their treatment in the public sector or become a service user when they have lost their jobs and their insurance plans. On the other, a private clinic that predominantly receives users from the social security system can also offer services to a handful of private health insurers. What is clearer, is that uninsured patients only have access to the public health system. For this study, the main public oncology institute, three private clinics (mostly social security), and three private hospitals (mostly private health) were purposively selected to encompass the three health subsystems based on the list of oncology services available at the provincial social security Web site [
28] and through communication with hospitals directors and heads of oncology services (See Table
1).
Table 1
Study Participants (n = 16)
Female | 7 |
Male | 9 |
Profession | Participants (n) |
Clinical Oncologist | 13 |
Gynaecologic Oncologist | 2 |
Radiation Oncologist | 1 |
No. Years Oncology practice | Participants (n) |
0˗5 | 2 |
6˗10 | 3 |
11˗20 | 4 |
21˗30 | 4 |
More than 30 | 3 |
Health System Provider | Participants (n) |
Public health | 6 |
Social Security (Obra social) | 5 |
Private health insurance (Pre-paga) | 5 |
Study design
A qualitative study design was developed to obtain insight from oncologists who prescibe HT in the city of Córdoba. The study employed individual semi-structured in-depth interviews to explore biological and non-biological factors, structural factors (health system) and personal characteristics (gender) associated with HT decision-making and prescribing practices as perceived by oncologists themselves. Adopting a thematic analysis approach (Framework method) allowed for the identification of these pre-selected themes as well as emergent themes generated from the data. Moreover, a qualitative approach can offer depth and detail on the experiences of oncologists’ prescribing practices that could elicit the development of complementary quantitative studies.
Data collection
For data collection, a stratified purposeful sample was used to obtain representatives of male/female oncologists working at the different health subsystems. Because this study wanted to explore the perspectives of oncologists in prescribing HT, social aspects such as gender, and health service provider – here used as a proxy to class – were considered as relevant variables within the sample. According to Patton [
29] (p. 240), ‘the purpose of a stratified purposeful sample is to capture major variations rather than to identify a common core, although the latter may also emerge in the analysis. Each of the strata would constitute a fairly homogeneous sample.’ A total of 16 individual semi-structured in-depth interviews were conducted with oncologists who regularly prescribed HT to women in adjuvant and metastatic stages of BC disease.
Participants were identified and recruited through initial contact with oncologists located at the 3 health subsystems (1 director of hospital, 2 heads of services). All oncologists interviewed were asked to further identify other potential participants with personal contacts, to whom the researcher contacted via telephone and email. The sample size was initially planned for around 20 participants and it followed the concept of ‘information power’ [
30] whereby the narrow aim of the study, the specificity of the participants included and the theoretical background (health system and gender) would offer sufficient focus for the interviews. 18 oncologists were approached and only 16 participated (1 interested but did not provide interview dates; 1 non-respondent). All interviews were conducted in Spanish by the author, who is a native speaker, and is familiar with health studies research on BC endocrine treatment and its use amongst eligible patients. Interviews were digitally recorded at participants’ consulting rooms and hospital offices during July 2016, and lasted between 37 and 101 min.
An interview question guide was developed prior to the recruitment process, which included a set of questions for three different themes related to HT: prescribing, adherence, and novel hormonal therapies. In this article, only results on prescribing are presented, and the question guide is available in Additional file
1.
Study sample
Participants for this study were 9 male and 7 female doctors, with specialisation in clinical oncology, radiation oncology, and gynaecological oncology. All with experience in prescribing HT, and with a wide breath of years in BC practice: For women, the median number of practising years was 16.4 (range: 5–40); and for men 21.3 (range: 5–54). This reflects the late feminisation of the oncology profession in the last 40 years.
Regarding the health service provider, participants worked at different settings (clinics, hospitals, and institutes) corresponding to the 3 health subsystems. For this study, the identification given to each of the 3 subsystems has followed, in the case of private/social security, the main type of population that the service assisted. Finally, it is also common for doctors in Argentina to work at different institutions, and within this sample, a few oncologists worked simultaneously in two different subsystems (one worked in the private and public sector, and three did so at the social security and the public sector). In this sense, the questions were focused on their perceptions and working experience of the specific setting where the interview took place.
After preliminary analysis of the data it was considered that sufficient information power regarding relevant patterns of prescribing practices was obtained before completing the 16 interviews. Characteristics of the sample are provided in Table
1.
Data analysis.
All interviews were recorded with participants’ consent, and transcribed verbatim by a research assistant with experience in qualitative data management. The author subsequently double checked the transcripts with the recordings and translated it into English. Framework analysis was used to analyse the data where a combined approached was adopted, first, through a deductive process based on the literature that informed the research questions and secondly, through and inductive process based on participants’ accounts [
31]. Data were analysed following the five methodological steps of the framework analysis: familiarisation with the data, identification of a thematic framework, indexing, charting, and mapping and interpretation of themes [
32]. Data was entered into a case chart for each respondent were notes and extracts of relevant passages were included for all identified themes (see Additional file
2 for an example of the case chart used). This allowed further identification of patterns and associations during the mapping and interpretation process of the similarities and differences in relation to gender and the health system provider.
Data interpretation is reported here by using relevant verbatim quotes to illustrate.
Quality assurance
In order to increase the internal validity of the data collected, member checking was systematically used during the interview and through the presentation of a summary of the information collected to each participant at the end. This allowed the investigator to paraphrase answers provided by each respondent, and to ensure understanding and accuracy in the presentation of ideas.
To ensure reliability and confirmability of data collected and analysed by a single researcher, the process of indexing (using textual codes) and charting was conducted in two different stages. First, indexing and charting was developed by using the Spanish transcript. After translation into English, a second round of indexing and charting took place fourth months after the first one. The recoding (crosschecking) of the two versions enhanced the process of refining themes and subthemes, and ensured the elimination of ambiguity of terms and lack of clarity, as well as the researcher’s subjectivity and bias.
Discussion
This qualitative study explored factors associated with oncologists’ prescribing practices of HT for women with oestrogen positive BC. To my knowledge, is the first qualitative study on this topic conducted in Córdoba, Argentina. HT is highly standardised in Western medicine through consensus guidelines elaborated by professional associations in the US and Europe, based on evidence-based clinical studies on populations. Whilst adherence to these guidelines, in the sample studied, seems to offer a consistent approach to treatment decisions, patient-specific factors such as tumour biology, clinical morbidity, drug toxicity, and tumour resistance allowed oncologist to develop an individual-case approach. The role of experience in the older generation also offered oncologists a way to exert choice and regain clinical judgment in relation to the standardisation imposed by guidelines. In addition, two other factors regarding guidelines’ applicability explored by the literature have been expressed by the respondents: firstly, the local resource implications (drug availability), which requires adaptability and consensus by the team of specialists, as observed by a study on cytotoxic drugs prescribing in the city of Rosario (Argentina) [
34]. Secondly, the constant following of updates to guidelines to keep peace with emerging data [
35].
More importantly, a range of non-biological factors appears to add complexity to the way the prescribing of treatment is formulated. Factors associated with patients’ SES was an influencing component in decision-making observed in this study, comprising three interrelated elements, socio-economic position, discussing treatment options, and health literacy. Although patients’ SES has not been specifically analysed in Argentina, a report exploring inequalities in access to drugs in the public sector pointed to the existence of ‘a cultural profile of the public sector patient and the health professionals that assist them, [as characterised by] a tendency to accept the disease and its associated problems with resignation, including the limitations of the care services’ [
36] (n/p).
Overall, this study has shown that SES factors played a role in prescribing HT treatment, and that there were different approaches observed according to oncologists’ gender. Even when most respondents acknowledged the difficulties that women living in very deprived conditions can pose to chemotherapy, male tended to consider HT as a more acceptable treatment ‘for all social classes’ (14, male public sector). This approach is, however, problematic, as it associates HT with ‘taking a pill’, without considering women’s self-management skills and understanding of the treatment. Health literacy, i.e. having the skills, knowledge and confidence to take decisions on medical instructions, is largely associated with socio-economic circumstances, whereby the most deprived groups are more likely to have low health literacy [
37]. However, as shown in this study, highly educated women could also have limited health literacy − which is understandable considering the complexity associated with treatments regimes in HT − and therefore, they were perceived by male oncologists as often demanding and asking too many questions. There are only a handful of studies that have measured health literacy in specific contexts in Argentina [
38,
39], but there are none for cancer. This is an area that will require more research to find out the implications it may have for health inequalities: a study has demonstrated that BC oncologists spend more time in consultations with highly educated patients than with low-income, less educated ones [
40], whilst other studies, similarly to what have been identified here, have shown that most deprived patients have greatest information needs and support in understanding their disease and treatment [
41,
42].
Previous research has noted that discussion of BC treatment choices between oncologists and patients has been associated with better health outcomes [
43], and more shared decision-making in adjuvant therapy was associated with greater treatment satisfaction [
44]. Whilst the process of shared decision-making is not incorporated into Argentina’s health policy, where a rather ‘paternalistic model’ prevails with a right to informed consent [
45], by considering oncologists’ gender as a variable, this qualitative study has shed light into how gendered power dynamics may affect the process of prescribing in different ways: On the one hand, female oncologists were more considered of the different side-effects that drugs can have in women’s quality of life and so they were more prone to consider women’s life circumstances and allow more educated women to share their views and discuss treatment options. On the other, the data collected suggest that prescribing practices in male oncologists seem not to be influenced by these considerations. Furthermore, male perceptions of both low and upper-class women can inadvertently normalise issues regarding women’s needs and understanding of treatment options, foster a sense of adherence to treatment initiation which may have implications for the amount of health information provided. Moreover, this can also have further implications for effective self-management, where oncologists and coordinated support from the health team is needed for patients to make treatment decisions and manage chronic conditions such as BC [
26]. Hence, further research is needed on health information interventions such as patient decision aids, as evidenced by a Cochrane review [
46], designed to meet the information needs of the different group of patients, according to their level of health literacy.
The fragmented nature of the Argentinian health system has often been identified as a cause of main health inequalities, between and within provinces, including cancer diagnosis and treatment [
47] as well as access to oncological drugs [
36]. The perspective of oncologists indicates that HT is strongly contingent on the health system providers, which limit the line of treatments (drugs) available from approximately 2 in the public sector to 7 in the private health insurance. Although oncologists tended to view these differences in the adjuvant setting as non-fundamental in terms of a reported 2–3% variation in drug effectiveness, in cases of metastatic BC the differences between lines of treatments and health outcomes can be more marked in overall survival as shown in a recent review [
48]. New drugs improving the action of anti-oestrogens as well as different generations of anti-oestrogens (SERD) and AIs open different sequence and combinations of treatments (Everolimus with Exemestane or TAM; Palbociclib with Letrozole or Fulvestrant). Precisely some of these targeted drugs are the ones that appear as more challenging to access in the social security and the public sector.
The recurrent mentioning of bureaucracy and paperwork involved in accessing drugs that are not covered within the Provincial/National protocol menu (public sector), or that are covered (social security) but its access is made very difficult by insurance plans can also make the prescription of treatment dependent on both determined oncologists and patients. Participants’ responses are in line with findings from Argentinian reports whereby oncological patients in the public sector referred to delays in access to medication due to wrong prescriptions and complicated paperwork from social services, as well as from national or provincial drug suppliers [
18,
36]. For insured patients, lawsuits have increasingly become an alternative way of accessing oncological medicines [
49]. In the private health subsystem respondents saw the request of drugs as less challenging. This could be due to a more manageable workload for professionals, a better coverage from the insurance plan or a simpler administrative procedure from the insurer. Finally, in requesting drugs outside the protocol, oncologists needed to consider, as one of them expressed, ‘The possibility to sustain treatment in the future’ (07, male private health insurance), that is, to balance the risk of discontinuing treatment after an initial medication approval.
Conclusions
The findings of the study identified a number of themes expressed by oncologists in their prescribing practices of HT for BC women in Córdoba. Overall, the empirical data collected suggests that HT is largely perceived by oncologists as a ‘different type’ of treatment, which is set ‘in comparison’ to the acute ones (surgery, chemotherapy and radiation). Considering HT as a form of chronic care is fundamental to address the specificities involved in oncologists’ prescribing practices and in the understanding of HT use among breast cancer patients. The intersections of biological (tumour type, age and co-morbidities) and non-biological factors (HT guidelines, patients’ SES, women’s ability to understand HT, and access to drugs), alongside the physician’s gender and the health system provider can variously and simultaneously influence oncologists’ decision-making. Given the different lines of treatments available within HT (according to tumour type, age, co-morbidities, and prognostic factors) effective communication of medication options can pose a challenge to oncologists and may disadvantage women affected by the disease, by inhibiting their comprehension of treatments options, benefits and risks. The use of tailored interventions such as patient decision aids delivered to BC women pre or during consultation, and designed according to the perceived health literacy needs (e.g. for low-literacy groups, a web-based interactive audio-visual intervention provided at the hospital with the assistance of a nurse; for high-literacy groups, a booklet or DVD using plain language), can facilitate women’s informed decisions and make it easier for oncologists to discuss treatment options.
In this study, oncologists’ gender can be a contributory factor on treatment decision-making. Female oncologists in the private and public sector, tended to be more responsive to women’s needs, more prone to discuss treatment options and drug toxicity. More qualitative research, including patients, will contribute to advance knowledge on oncologists’ gender as a specific factor, before policy implications can be drawn.
Of significance, the fragmented health system can lead to a differential access to drugs or lines of HT in the metastatic BC setting between the public and private sector; and also to a restricted access to drugs that are covered through the insurance plans. Health system coordination amongst the subsectors, a role assumed by the Ministry of Health of the Province, should be improved in terms of drugs accessibility if equity in access is to be achieved. Initiatives should be taken to make the provision of oncological drugs more streamlined and accessible to all service providers.