Objective Recognising the need for a best-practice and consistent approach in providing care to women with rheumatoid arthritis (RA) in relation to (1) general health, (2) contraception, (3) conception and pregnancy, (4) breast feeding and (5) early parenting, we sought to achieve cross-discipline, clinical consensus on key messages and clinical practice behaviours in these 5 areas.
Design 3-round eDelphi study. In round 1, panellists provided free-text responses to open-ended questions about care for women with RA across the 5 areas. Subsequently, panellists refined and scored the synthesised responses, presented as metathemes, themes and detailed elements. Where ≥5% of panellists did not support a theme in a given round, it was removed.
Setting Panel of practicing Australian rheumatologists (n=22), obstetricians/obstetric medicine physicians (n=9) and pharmacists (n=5).
Results 34 (94.4%) panellists participated in all 3 rounds. The panel supported 18 themes across the 5 areas (support/strongly support: 88.2–100%) underpinned by 5 metathemes. Metathemes focused on coordination in information delivery, the mode and timing of information delivery, evidence underpinning information, engagement of the right health professionals at the right time and a non-judgemental approach to infant feeding. Themes included practices for primary prevention of chronic disease and their sequelae, the importance of contraception and planning pregnancy and breast feeding, close monitoring of medications, supporting mental well-being, managing disease activity and providing practical support for early parenting.
Conclusions A cross-disciplinary clinical panel highly supported key information and clinical practices in the care for women with RA across the continuum of contraception to early parenting within a whole-person, chronic disease management approach.
Statistics from Altmetric.com
Strengths and limitations of this study
A clinical panel of rheumatologists, obstetricians and pharmacists contributed to the development of the recommendations, reflecting ‘real-world’, multidisciplinary clinical perspectives on best-practice care for women with rheumatoid arthritis.
Inductive analysis of responses to open-ended questions ensured that themes were developed empirically from the expert panel.
Very strong support was achieved for the final set of recommendations.
The recommendations relate to clinical best-practice within a high-income economy and therefore may not be readily transferable to middle and low-income economies.
Rheumatoid arthritis (RA) typically occurs in women during their childbearing years,1 with a prevalence that has increased over the last 15 years2 and is expected to increase in coming decades.3 Compounded with the prospect of living with a chronic, painful and sometimes progressive disease associated with substantial physical and mental health impacts,4 young women with RA and their families face additional challenges in navigating a safe and successful pathway to parenthood.5 Specifically, this pathway requires informed, collaborative decision-making and careful planning on the part of the patient and health professional(s) around RA disease activity monitoring and safe pharmacological management, physical and emotional health, and social support as they relate to contraception, conception and pregnancy, birth, breast feeding and early parenting.1 ,5–10
Our recent primary research investigating the information and service needs of Australian women with RA concerning pregnancy, breast feeding and early parenting identified a dearth of contemporary information available to support women to make informed decisions and implement practical coping strategies.6 In particular, our data suggest that women with RA have clear pregnancy-related educational needs and seek specific information from their care providers about medicines safety during pregnancy and breast feeding, physical and emotional support services and practical coping strategies to manage pain and flares during the postnatal period.6 Moreover, in a recent systematic review and editorial, we identified a lack of self-management interventions for women with RA targeted towards these issues,9 ,11 despite European League Against Rheumatism (EULAR) recommendations that specific patient education is critical at important life and disease course stages12 and published Standards of Care for RA that recommend comprehensive and individually tailored education about RA and its management.13 While some consumer texts14 ,15 and arthritis organisation web pages have been developed to address information gaps in this area, they remain limited and inadequately address some of the most important issues raised by women relating to disease management and medicines safety during pregnancy and breast feeding.5 ,6 Consequently, treating health professionals (particularly rheumatologists) remain the primary communicators of such information, which is expected by patients6 ,16 and entirely appropriate for the majority of cases.12 Given that patients’ information needs vary,6 highly prescriptive guidelines are unlikely to be of use in clinical practice and do not align well with a patient-centred model of care.
Dealing with these issues is complex for health professionals, particularly in the light of limited data related to drug safety in pregnancy and breast feeding,1 ,17 highlighting the importance of shared decision-making. Recent publications, however, provide updated evidence on medicines safety during pregnancy and breast feeding.18–20 While these medications safety data and guidelines are fundamental to optimising clinical practice, they reflect only one component of care for women with RA at this important life stage. A broader counselling approach from health professionals is therefore important1 ,10 ,21 and there is currently limited guidance in this area.20
Considering that a multidisciplinary approach to care is most likely to confer optimal fetal and maternal outcomes,1 cross-discipline communication is important, as is consistency in the interdisciplinary messages that are provided to women with RA.1 These issues are particularly relevant in settings with limited specific healthcare resources.22 The Australian healthcare system supports interdisciplinary care for people with chronic health conditions, such as RA. For most patients, care is coordinated by family physicians, and in the context of RA and pregnancy, is delivered by rheumatologists and obstetricians, and supported by allied health practitioners and community and hospital pharmacists. Outside public tertiary hospital clinics, these health professionals are rarely colocated, so coordination of care and communication between health professionals can be fragmented.23 Notably, proceedings from the American College of Rheumatology (ACR) Reproductive Health Summit suggest that interprofessional and patient–professional communication in issues relating to fertility, pregnancy and lactation need to be improved.1 Further, women with RA have expressed great concern and frustration at the contradictory messages they receive from their healthcare providers relating to RA management during conception, pregnancy, breast feeding and early parenting.6 Having already undertaken primary research to identify the health information and service needs of women with RA concerning pregnancy, postnatal care and early parenting6 and evidence for the effectiveness of interventions to improve women’s knowledge and self-management skills related to contraception, pregnancy and breast feeding;11 our programme of translational research sought to develop guidance for clinicians in interprofessional care. Specifically, this study aimed to establish cross-discipline consensus on the important educational messages (ie, ‘what to say’) and practice behaviours (ie, ‘what to do’) in the management of women with RA relating to contraception, pregnancy, breast feeding and early parenting.
Materials and methods
A multidisciplinary national clinical panel was recruited, consisting of Australian practising rheumatologists, obstetricians/obstetric medicine physicians and pharmacists. Given our earlier primary research with patients6 and that the focus of this study was to develop guidance for clinicians, we did not include patients on the Delphi panel. Panellists were sampled via a range of strategies, including self-nomination, purposive sampling via personal invitation from the researchers and snowballing; consistent with standard practice and recent Delphi studies and protocols in rheumatology research.13 ,20 ,26–31 The specific sampling strategies used for each discipline are summarised below:
Rheumatologists: a study flyer was provided to all attendees of the 2015 Australian Rheumatology Association Annual Scientific Meeting, inviting rheumatologists who met the inclusion criteria to contact one of the investigators (SVD) to express their interest in participating. The same investigator also sent email invitations to rheumatologists in each Australian state and territory who were known to the research team to practice clinically and/or who had an interest in pregnancy-related care.
Pharmacists: pharmacists known to have a special interest in medicines safety with pregnancy or breast feeding were invited via email from one of the investigators (SVD) to participate. In addition, pharmacy departments at two tertiary women's hospitals (Victoria and South Australia) and one university (Western Australia) were contacted to nominate suitably qualified pharmacists to facilitate snowballing.
Obstetricians/Obstetric medicine physicians: clinicians known to have a special interest in inflammatory arthritis, chronic disease or high-risk pregnancies were invited to participate via email from one of the investigators (SVD). Websites of obstetric medicine practices were searched to identify clinicians who advertised a special interest in high-risk pregnancies.
Across each of the three disciplines, clinicians were invited to nominate other colleagues who met the study inclusion criteria, to facilitate snowballing.
Inclusion criteria for the panellists included: current registration to practice in Australia; a minimum of 4 years consultant experience (rheumatologists, obstetricians/obstetric medicine physicians) or clinical practice experience (pharmacists); and currently clinically active for at least 8 hours (or two clinical sessions) per week or had undertaken this level of clinical practice within the last 5 years. All potential panellists were asked to complete an online screening questionnaire that was used to confirm eligibility and collect demographic data.
eDelphi protocol and analysis
All eDelphi rounds were administered using Qualtrics software (Provo, Utah, USA). Each round started by sending panellists an email with a unique hyperlink to access the portal.
In round 1, panellists were asked to provide free-text responses to 10 open-ended questions about important information and resources women with RA and their families needed to receive across five domains, including: (1) general health, (2) contraception, (3) conception and pregnancy, (4) breast feeding and (5) early parenting. Free-text data were analysed inductively for each question using a content analysis approach to develop a framework of key themes and detailed elements underpinning each theme within each of the five domains of interest.32 Overarching metathemes (guiding principles) were also developed, representing concepts that spanned across the five domains. Data were analysed independently by two analysts with clinical and patient communication expertise (AMB and JEJ, respectively). Where discrepancies were identified, these were resolved by consensus and review of the raw data.
In round 2, panellists were presented with a synthesis of the themes and elements for each domain and asked to assign their level of agreement with the information presented. Panellists were asked to rate their agreement at the key theme level, while being cognizant of the detailed elements underpinning each theme, using a three-point Likert scale: ‘completely agree’; ‘partly agree (modifications required)’; ‘disagree’. Where panellists selected ‘partly agree’ or ‘disagree’ for a particular theme, they were asked to provide a free-text response to explain their selection. Where a theme was scored as ≥5% ‘disagree’ by the panel, it was removed. Free-text data were analysed as described above and used to further refine the themes and elements. Elements were further categorised into clinical practices (‘what to do’) for health professionals and key messages to be disseminated (‘what to say’).
In round 3, panellists were presented with a final synthesis of the domain-specific themes and elements relating to clinical practices and key messages. Panellists assigned their level of support at the key theme level, while being cognizant of the detailed elements underpinning each theme, using a five-point Likert scale: ‘strongly support’; ‘support’; ‘undecided’; ‘oppose’; ‘strongly oppose’. Panellists were given the opportunity to provide additional comments using free-text fields for each of the five domains. Where a key theme was scored as ≥5% ‘oppose’ or ‘strongly oppose’ by the panel, it was removed. Free-text data were analysed as described above and used to further refine the themes and elements.
Descriptive statistics were used to characterise the panel. All members of the research team reviewed the synthesised themes and elements at each round for clarity and clinical meaningfulness. Descriptive statistics were used to analyse frequency data from rounds 2 and 3.
Thirteen (35.1%) individuals self-nominated, 16 (43.2%) were purposively sampled and a further 8 (21.6%) were identified from snowballing. Of these, 36 (97.3%) were deemed eligible to participate in round 1 (1 was ineligible due to having <4 years consultant experience). Demographic characteristics of the 36 panellists are summarised in table 1.
Thirty-five panellists (97.2%) participated in round 2 and 34 (94.4%) in round 3.
Analysis of free-text data from round 1 revealed 21 themes, supported by detailed elements across the five domains. At the second round, two themes exceeded the inclusion threshold of ≥5% disagreement and were removed. At the third round, one theme exceeded the inclusion threshold of ≥5% oppose or strongly oppose, leaving a final set of 18 themes supported by five metathemes (guiding principles) with support/strongly support for key themes ranging from 88.2% to100%.
Metathemes (guiding principles) focused on coordination in information delivery across health professionals, the mode and timing of information delivery, evidence underpinning information, engagement of the right health professionals at the right time and a non-judgemental approach to infant feeding approaches (box 1).
Metathemes (guiding principles) derived from the data that spanned all domains of interest
Consistent information should be conveyed to women with rheumatoid arthritis (RA) by a range of health professionals across the care continuum in a coordinated manner, based on the expertise of the health professional and relevance to the patient (ie, it is not suggested that one health professional carries the responsibility to convey all this information)
The mode and format in which information is delivered; timing of information; and level of details provided by a health professional(s) should be tailored to the needs of the patient, stage of their decision-making and with consideration of their health literacy, emotional well-being and overall situation at the time
Information provided by health professionals should be based on the best level of evidence currently available (or best practice in the absence of clear evidence) and relevant to their disease status
The recommended actions are those that should be undertaken across the health professional team, in a coordinated approach if and when appropriate. It is important to recognise that not all team members will be needed at all times: some team members will have different roles at different times (eg, general practitioners and gynaecologists may take a greater role in counselling about contraception, while rheumatologists would likely take a greater role in reviewing medication profiles)
Information and actions are intended to support women with RA who plan to, or choose to, breast feed in a non-judgemental manner. It is acknowledged that a range of infant feeding options are available and that breast feeding may not always be possible
In the final presentation of themes and elements data for each of the five domains (tables 2⇓–4), the elements for ‘saying’ and ‘doing’ are not intended to be linked and interpreted together within a specific table row. Rather, they should be interpreted independently and as a non-hierarchical list.
For general health approaches, panellists recommended primary prevention practices for chronic disease management, discussion around RA management and guidance on identifying trustworthy information (table 2). Across the four themes, final round scores for strongly support or support ranged from 94.1% to 100.0%.
For contraception, discussing the importance of contraception and its various options were identified as important themes (table 3). Across the two themes, final round scores for strongly support or support ranged from 97.1% to 100.0%.
For conception and pregnancy, medication reviews, discussion about maternal and fetal health and preconception care, maintaining optimism, health and disease monitoring and practical considerations related to being pregnant with RA were deemed important (table 4). Across the six themes, final round scores for strongly support or support ranged from 94.1% to 100.0%.
Similarly, medication reviews and discussion concerning maternal and fetal health were considered important for breast feeding (table 5). For both of the two themes, final round scores for strongly support or support were 100.0%.
For early parenting, discussing impacts of RA, medication reviews and whole-person management were considered important (table 6). Across the four themes, final round scores for strongly support or support ranged from 88.2% to 100.0%.
While clinical practice guidelines, drug surveillance data from registries, databanks and electronic medical records are critical to informing clinicians’ decisions about safe and appropriate RA management decisions,33 these resources do not necessarily bridge the ‘know-do’ gap for clinicians or empower patients to engage in shared decision-making.34 Supporting clinicians with clinician-centred, practical guidance (eg, guidance on ‘what to say’ and ‘what to do’) on how to practically implement evidence into routine clinical practice is therefore important,34 particularly in diseases like RA where evidence–practice gaps exist.35 ,36 To the best of our knowledge, this is the first study that has addressed this issue as it relates to cross-discipline management of RA in the context of contraception, pregnancy, breast feeding and early parenting. While recent publications provide excellent contemporary evidence on concerning the safety and effectiveness of pharmacological management in the perinatal and postnatal periods,18–20 they do not consider a broader approach to management, particularly as it relates to implementing a person-centred model of care at the level of the clinical encounter. Consensus-based recommendations, supported by metathemes, have been developed to guide information delivery and clinical practices in a model of person-centred care for RA. Importantly, this study complements the eumusc.net standards of care for RA13 and EULAR overarching principles for antirheumatic drug use before, during and after pregnancy20 by operationalising these standards and principles in the context of contraception, pregnancy, breast feeding and early parenting.
The themes and elements identified across the five domains reasonably balance clinical practice behaviours and recommendations for the delivery of health information across health professionals with support for self-management, shared decision-making and recognition of non-physical impacts of pregnancy and early parenting on a background of RA. We suggest these components and this balance are reflective of best practice in a person-centred model of care for a chronic disease, such as RA.37 The consensus clinical recommendations also align with the findings of a recent review around managing pregnancy in women with rheumatological disease,38 and extend beyond the physical implications of RA. Although a previous small study reported no mental health impact of pregnancy in a group of women with RA compared with controls, as assessed by the SF-36,39 panellists in our study recognised the potential for psychological health to be impacted. The metathemes acknowledge that these recommendations should be considered in the context of a person's health literacy and disease status, which are critical for informed decision-making and capacity to engage in cocare.40 Importantly, many of the themes and elements were oriented towards empowering women, through tailored education, to make decisions (eg, about breast feeding) in collaboration with their health professionals and maintaining a positive outlook to pregnancy and early parenting. In this context, an important theme that emerged was the risk of relying on web-based materials for information about medicines safety related to pregnancy and breast feeding and disease control. Panellists were unequivocal in recommending that women seek professional advice from their healthcare team and consult appropriately qualified organisations, such as arthritis consumer organisations and breast feeding organisations, for more information. Notably, while internet searching is important to consumers in this context,6 ,41 data from recent studies examining website quality as it relates to chronic pain suggest it is highly variable.42 ,43
Panellists provided the most detailed responses for Theme 3—Conception and Pregnancy. In addition to supporting women in conception and pregnancy within a chronic disease model of care through education, practical skills and situation planning, the most significant clinical issues raised related to the critical need for pregnancy planning due to the possible teratogenic effects of disease-modifying agents and achieving disease control to ensure maternal and fetal health and safety. These issues are consistent with those raised in recent reviews on this topic, confirming face validity of the recommendations20 ,22 ,38 and reflect the priorities of women with RA.6 We suggest, therefore, that our findings, while intentionally clinician-centred, align well with the primary research we conducted with women with RA that formed the basis for the current study.6
Panellists considered that assessment of indications and contraindications for various contraceptive options related predominantly to care provided by general practitioners or gynaecologists and this theme was, therefore, not included in the final set. Notably, a recent systematic review44 and guidelines45 have been published on the issue of contraception in RA, which provides some guidance to clinicians.
While recommendations have been developed concerning the management of RA at this important life stage,18–20 ,38 ,44 ,45 there is a lack of RA-specific guidelines for clinicians, largely attributed to a lack of definitive clinical trials data concerning drug toxicity and safety during pregnancy and breast feeding, particularly as they relate to the newer disease-modifying agents.10 A recent review highlights this issue to be particularly relevant in the Middle East.22 Variability in practice is, therefore, unsurprising,46 resulting in uncertainty and dissatisfaction for patients.6 ,47 Our data offer some practical strategies for health professionals as a means to facilitate consistency in an approach to management and a prompt for cross-discipline care, which has previously been identified as inadequate,47 despite being recommended.38 For women with RA and their families, we support the use of the RA Standards of Care Checklist as a means to ensure RA care, as it relates to contraception, pregnancy, breast feeding and early parenting, is optimised.13
Our study has a number of important strengths: we recruited a large expert sample relative to other comparable studies in arthritis care,12 ,13 ,20 ,27 ,30 ,31 our sample was cross-disciplinary with a spread of practice locations across the country, our response rate remained very high across all three rounds, and we achieved a high level of support for the final set of themes. Our sample was purposely weighted towards rheumatologists to reflect the discipline most usually involved across the continuum of care from contraception through to early parenting. Importantly, however, the findings do not relate solely to rheumatologist-delivered care, but rather all clinicians involved in a woman's care and reflect a best-practice approach. It would be unreasonable to expect a single clinician or clinical discipline to adopt all the recommendations. The transferability of our recommendations may be limited in some aspects since we sampled only Australian clinicians and therefore, the cross-cultural relevance and meaningful transferability to other health systems (eg, those in low and middle-income economies or rural settings) should be explored in future work.22 ,48 Australians living in urban settings enjoy a health system that supports access to coordinated care from multiple health professionals. In rural Australia and in other nations, this is rarely accessible and therefore the recommendation to establish multidisciplinary care teams may not be feasible in a real-world setting in these contexts, particularly at the primary care level. Multidisciplinary team care may, in some settings, be more appropriately established at the tertiary hospital level where access to medical specialists and upskilled allied health providers is more achievable. Although we did not include patients in our sample, the foci for the Delphi were directly informed by previous empirical, consumer-based research, consistent with the approach adopted by Hawker et al.30 The results from this study and our recent systematic review11 point to the opportunity to develop and evaluate targeted educational interventions for this group of consumers.
Coordinated, multidisciplinary care for women with RA concerning contraception, pregnancy, breast feeding and early parenting is important. Although clinical guidelines provide necessary information on ‘what care’ should be provided, guidance on ‘how’ to provide the care is rarely described. Cross-discipline recommendations for care delivery include: counselling and practices relating to primary prevention of chronic disease and their sequelae, supporting women to actively use contraception and to plan pregnancy and breast feeding, close monitoring of medications, supporting mental well-being, managing disease activity and providing practical support for early parenting. A chronic disease model of care should underpin these practices.
The following expert panellists are acknowledged for their valuable time and expertise in contributing to this project: Dr Claire Barrett, Dr Helen L. Barrett, Dr Lisa Begg, Dr Andrea Bendrups, Associate Professor Paul Bird, Ms Melinda Boss, Dr David Careless, Dr Sabina Ciciriello, Dr Stephen Cole, Dr Nicola Cook, Professor Julien de Jager, Dr Marie Feletar, Dr Andrew Gibson, Dr Ilana Ginges, Dr Patrick Hanrahan, Dr Alberta Hoi, Associate Professor Helen Keen, Dr Anita Lee, Associate Professor Karin Lust, Dr Mona Marabani, Dr Geraldine Moses, Associate Professor Peter Nash, Dr Peter Neil, Dr Gene-Siew Ngian, Dr Mark Reed, Dr Joylene Rentsch, Dr Bethan Richards, Dr Helen Robinson, Dr Janet Roddy, Dr Renuka Sekar, Ms Tricia Taylor, Dr Christine Tippett, Mr Rodney Whyte and Dr Laurel Young.
Twitter Follow Ilana Ackerman at @arthritis
Contributors All coauthors contributed to the study design and planning. AMB analysed and interpreted the data and wrote the manuscript. JEJ collected, analysed and interpreted the data and reviewed and edited the manuscript. INA reviewed and edited the manuscript and assisted with data interpretation. SVD procured funding, led recruitment, reviewed and edited the manuscript and assisted with data interpretation.
Funding Funding for this investigator-initiated study was provided by unrestricted research grants from AbbVie Australia, UCB Australia Pty Ltd and Janssen Australia.
Competing interests None declared.
Ethics approval Human Research Ethics Committees of the University of Melbourne and Curtin University.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No additional data are available.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.