What constitutes patient-centred care for women: a theoretical rapid review

We conducted a theoretical review, characterized by a comprehensive search strategy, inclusion of conceptual and empirical primary sources, explicit study selection, no quality appraisal, and content analysis of included items [19]. Theoretical reviews are suitable when the aim is to describe how a given process has been conceptualized. We also adopted a rapid review approach to quickly generate a PCCW framework that could then be elaborated through primary research in a larger planned study. A rapid review is characterized by restriction to a single language (English), a short time frame (last 10 years, 2008+), exclusion of grey literature, quality of included studies is not appraised, and authors of included studies are not contacted [20, 21]. In a typical rapid review, one person performs screening and data abstraction, but we employed independent screening to enhance rigour. As there are no reporting criteria specific to theoretical reviews, we employed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses criteria [22]. Data were publicly available so institutional review board approval was not needed. We did not register a protocol for this review.

We used a PICOT (participants, issue, comparisons, outcomes, type of publication) framework to establish eligibility criteria. Participants included at least 50.0% adult women (18+) participants, and/or clinicians (physicians, nurses) of any specialty in primary, secondary or tertiary care. The problem investigated included any specific condition not included in our previous complementary review, which addressed cardiovascular disease, mental health, and reproductive health topics prioritized by our research team at that time [23]. The current review examined other conditions to assess if PCCW differs for women facing different health care issues. The issue referred to PCC, or a synonymous term such as person-, women-, client-, or family-centred care, or approaches or strategies to promote or support PCC. PCC was viewed as compassionate, respectful care that addresses patient values and preferences, as well as information and supportive care needs, thus requiring patient-level engagement and patient-provider interaction [1‐7]. Patient-centred communication is the cornerstone on which patient-centred care is built. For this reason, we focused our review on the elements of good patient-centred communication (PCC) in order to illuminate the broader application of PCCW. To do this, we adopted McCormack et al.’s conceptualization of PCC of 31 elements organized in six domains: fostering patient-clinician relationship, exchanging information, recognizing and responding to patient emotions, managing uncertainty, making decisions, and enabling patient self-management [7]. Comparisons referred to studies that explored patient or clinician views about PCCW, its barriers and how to achieve it; or studies that evaluated strategies, interventions or tools aimed at supporting PCCW.. Thus, publication type including qualitative (interviews, focus groups, qualitative case studies) and quantitative (questionnaires, randomized controlled trials, time series, before/after studies, prospective or retrospective cohort studies, case control studies) research designs, or mixed methods studies published in English language. Outcomes included but were not limited to awareness, understanding, experiences or impacts of PCCW, or determinants or factors influencing any of these functions, or the impact of strategies implemented to support or improve PCCW. Although systematic reviews were not eligible (to avoid duplication of studies included in reviews and by our search), we screened their references to identify additional eligible primary studies.

On February 9, 2018, ARG (principal investigator) conducted a preliminary search of MEDLINE using [patient-centered care”] AND [wom#n or female] from 2008 to that date. This initial search was carried out to capture relevant studies that did not necessarily employ the term PCC, become familiar with the literature, and develop a more comprehensive search strategy. This search returned nearly 31,000 results of which many were not relevant to PCC, requiring considerable time and effort to screen. For example, many studies arbitrarily used the term PCC referring to clinical care without defining, describing or measuring PCC. Thus, we opted for a more focused search strategy to retrieve articles that specifically employed the term PCC or similar alternatives.

Our search strategy was developed with a medical librarian and complied with the Peer Review of Electronic Search Strategy reporting guidelines (Additional file 1: Table S1) [24]. On February 26, 2018, we searched MEDLINE, EMBASE, CINAHL, and SCOPUS from 2008 to that date. We searched for studies that explicitly used the term “patient-centered”, or an alternative spelling or synonymous option. We supplemented that keyword search with Medical Subject Headings reflecting the concept of PCC to identify studies that employed a synonymous term for PCC that we had not considered, and combined those searches with terms for women. As part of our larger study, we employed the same search strategy, and separately screened for and reviewed studies of PCCW for cardiac rehabilitation, depression, and family planning, topics prioritized by our collaborators (who included health services researchers, clinician investigators, and representatives of professional societies, disease-specific foundations, quality improvement and monitoring agencies, patient advocacy groups, patients and consumers). Thus, our search strategy reflects studies of PCCW for any conditions other than those.

To pilot test the screening process, KB (research assistant), JUR (graduate student) and ARG independently screened the first 50 titles and abstracts, then compared and discussed discrepancies to achieve a common understanding of how to apply eligibility criteria. KB and JUR independently screened titles and abstracts against eligibility criteria, and ARG resolved queries and discrepancies. Exclusion criteria were generated concurrent with screening. Studies were not eligible if participants were: less than 50% women, solely family members, caregivers, or care partners as their views do not always match those of patients, and allied health care professionals (i.e. pharmacists, dentists) or medical trainees. Studies were also excluded if based in a long term care or residential setting, where what constitutes PCCW might differ from primary, secondary or tertiary care. Studies were excluded if they concluded that PCC was necessary, or arbitrarily used PCC to refer to patient treatment interventions or management models, the illness experience rather than the care experience, or patient-reported outcomes.,. Studies solely referring to one aspect of PCC such as information needs or empathy, rather than examining PCC as a multi-dimensional construct or approach, were also excluded.

JUR and ARG pilot-tested the data extraction process on 3 studies, and compared and discussed discrepancies to achieve a common understanding of what data to extract and how. JUR extracted and tabulated data on study characteristics including author, publication year, country, study objective, research design, participants, term used to refer to PCC, definition or description of PCC, and relevant findings including details of interventions implemented to promote or support PCC, and outcomes of PCC.

We used summary statistics to report the number of studies published per year, and by condition, country, study design, and term used for PCC. We compared definitions or descriptions of PCC that were articulated by participants across studies and conditions, and then mapped those PCC constructs against McCormack’s PCC framework comprised of 31 elements organized in 6 domains. While it reflects the views of cancer patients including both men and women, it was chosen because it was rigorously-developed and more comprehensive than other PCC frameworks [7]. This served to compare expressed views about what constitutes PCCW specific to women against PCC constructs considered ideal by other patients and clinicians, and potentially identify PCC constructs unique to women with conditions other than cancer. Without doing so, we would not have been able to identify gaps in the way PCCW has been studied, and instead would have compiled PCC components, work already done by McCormack and others [6, 7]. To identify gaps in the way PCCW was studied, we summarized the number of domains addressed in each included study. We also compared PCC domains addressed in studies comprised largely of women and mixed-gender studies. Instruments used to measure PCC were specified. We described the impact and determinants of PCC narratively, and the number of studies that evaluated interventions designed to promote or support PCC, and details about those interventions. Research team members, which included health services researchers, physicians of various specialties and experts in women’s health, independently reviewed data and the draft manuscript, and provided feedback that shaped the interpretation of results and conclusions.

We identified a total of 9267 studies, from which 6670 unique studies remained after removal of duplicates. Screening of titles and abstracts eliminated 6513 studies. Screening of 157 full-text articles eliminated another 118 studies that were not eligible because they were not about PCC (43), the number of participating women was not stated to ascertain if at least 50% were women (28), or because the condition (24), publication type (12), participants (6) or setting (5) were not eligible. Ultimately, we included 39 studies in this review (Fig. 1). Additional file 2: Table S2 includes data extracted from included studies.

Studies were published from 2008 to 2018. Among the 39 studies, clinicians were participants in 5 (12.8%). Of the remaining 34 studies involving patients, women were the sole participants in 22 (64.7%), 93.0 and 94.0% of participants in 2 studies [25, 26], and in 10 (29.4%) studies comprised 50.0 to 69.8%. Most studies were conducted in the United States (17, 43.6%). Others were conducted in the United Kingdom, Canada, Belgium, Netherlands, Japan, Australia, Brazil, Chile and South Africa. Most studies involved qualitative interviews with patients or clinicians (18, 46.2%) or surveys (17, 43.6%), and 4 (10.3%) were observational studies. More than half of included studies investigated maternal care (21, 53.0%). Other studies investigated cancer, diabetes, HIV, endometriosis, dementia, distal radius fracture, overactive bladder, and systemic lupus erythematosus. With respect to objectives, most studies (23, 59.0%) explored patient and/or clinician views about PCC and its determinants. The remainder examined whether PCC took place (8, 20.5%), evaluated interventions meant to support PCC (4, 10.3%), solely identified determinants of PCC (3, 7.7%), or developed a scale to measure PCC (1, 2.6%). Of the 39 studies, 17 (43.6%) were qualitative studies, which inductively explored what women perceived as PCC to truly represent what women believed constitutes PCC. Most studies referred to patient-centred care (23, 59.0%). Others referred to woman- or women-centred care (10, 25.6%), person-centred care (3, 7.7%), family-centred care (2, 5.1%) or patient and family-centred care (1, 2.6%).

Of the 39 included studies, 5 (12.8%) [27, 28, 29, 30, 31] did not define or describe PCC; instead, they explored whether patients experienced PCC. Of the 34 (87.2%) studies that defined or described PCC, none addressed all 6 PCC domains (Table 1). One study involving interviews with mothers of premature babies to explore their views on family-centred neonatal care addressed 5 of 6 domains, and 6 (17.6%) studies addressed 4 of 6 domains [43]. More commonly addressed domains were exchanging information (25, 73.5%) and fostering the patient-clinician relationship (22, 64.7%). Fewer studies addressed making decisions (16, 47.1%), enabling patient self-management (15, 44.1%), or responding to emotions (12, 35.3%). Managing uncertainty was addressed in only 1 study (1, 2.9%) [43]. There were too few studies to compare PCC across different medical conditions, or involving clinicians to compare their PCC priorities with those of women. Among the 34 studies that defined or described PCC, the frequency of addressed domains differed between studies involving largely women (22 studies all women, 2 studies at least 93.0% women) and mixed-gender studies (10 studies of 50.0 to 69.8% women), suggesting that women’s PCC preferences may differ from those of men. In studies involving largely women, 78.0% addressed exchanging information (40.0% mixed-gender), 54.2% addressed fostering the relationship (70.0% mixed-gender), 33.3% addressed each of making decisions and enabling self-management (50.0 and 60.0%, respectively, mixed-gender), 29.9% addressed addressing emotions (50.0% mixed-gender), and 4.2% addressed managing uncertainty (0.0% mixed-gender). No studies identified PCC components not already captured by the McCormack framework.

Eleven (28.2%) studies employed validated instruments to assess PCC. They included: Endometriosis Care Questionnaire [61, 62], Perceptions of Family Centred Care – Parents and Perceptions of Family Centred Care – Staff [27], Patient Expectations and Satisfaction with Prenatal Care Instrument [41], Patient Perception of Patient Centred Care [64], Women Centred Care Pregnancy Questionnaire [44, 40], Patient Assessment of Chronic Illness Care [54], Communication Assessment Tool [50], Wake Forest Trust Scale [50], Global Rating Scale [53], Japanese Measure of Processes of Care for Service Providers, and Japanese Measure of Beliefs about Participation in Family Centred Service [48]. One study validated a new PCC instrument, the Person Centred Maternity Care Tool [34]. However, as noted, instruments varied in the domains they assessed, and none addressed all 6 McCormack PCC domains [7], suggesting that current instruments may not be fully or accurately assessing PCC. No instruments included PCC components not already captured by the McCormack framework.

Of the 39 included studies, 11 (28.2%) reported outcomes associated with PCC. These were captured via validated instruments, qualitative interviews, or non-validated surveys created specifically for the study, and were captured either following a PCC intervention or regular consultation. Patient ratings of PCC were found to be positively correlated with health-related quality of life [61], care satisfaction [44, 47, 50], disease understanding [54], sense of control during birth [44, 47], and clinical outcomes including pain reduction [64], functional recovery [64], and physiological child birth [33]. Patient ratings of PCC were negatively correlated with perceived consequences of illness [54]. PCC behaviours displayed by clinicians were positively correlated with patient knowledge [58], patient well-being [58], treatment satisfaction [31], emotional health [31], and use of prevention of mother-to-child transmission services [54]. Additionally, PCC ratings influenced care preferences regarding induction of labour (direction unspecified) [45].

Fifteen (38.5%) studies identified challenges or barriers to PCC. Clinician traits such as being present during procedures [37, 46], less than 5 years of experience [40], racism [29, 57], training [58], overconfidence [28], and low self-efficacy [48] impeded the use or effectiveness of PCC. Patient barriers to PCC included language (i.e. whether clinician spoke Spanish) [29], social determinants of health [56], HIV stigma [56, 58], lack of preparation for meetings [36], and discomfort with asking questions [36]. Power struggles within the patient-provider relationship [43] and lack of tools to support patient-provider interaction [25, 46] also challenged PCC.

Four studies examined programs or interventions to support PCCW. Two of these were implementation of PCC models during maternity care: the Patient and Family Centred Care Model [27], which involved an organizational shift towards family being accepted as the primary source of support and strength for newborns, including concepts such as unrestricted access to newborns, service flexibility, cooperation, respect, and increased autonomy for the families; and the Model of Integrated and Humanised Midwifery Health Services, which involved specific activities such as strengthening the patient-clinician relationship, continuous emotional support, encouraging variation in birthing position and pain relief methods, and promotion of mother and child bonding [27]. Other programs were the Nurse Patient Navigation Program for HIV care, aiming to retain women of colour in HIV care by supporting autonomy via orientation, care coordination and regular phone calls from a registered nurse for up to 8 months [57]; and specialized diabetes education sessions with a registered nurse and a dietician that included individualized information about self-care, lifestyle habits, treatment priorities and action plans [30]. Three of these studies reported positive results including improvement in patient perceptions of PCC [27, 30, 57], happiness and connection with clinician [57], and increased self-efficacy for self-management [30]. In contrast, 1 study found that women were unhappy with their care, and there were multiple discrepancies between women’s PCC needs and clinician practices despite the implementation of this program [36].