Promoting patient participation in healthcare interactions through communication skills training: A systematic review

Abstract

Objective

To present literature on training patients in the use of effective communication skills.

Methods

Systematic searches were conducted in six databases. References were screened for inclusion through several phases. Extracted data included intervention study design, sample characteristics, content and structure of training programs, outcomes assessed, and findings reported.

Results

A total of 32 unique intervention studies were included. Most targeted primary care or cancer patients and used a randomized controlled study design. Interventions used a variety of training formats and modes of delivering educational material. Reported findings suggest that communication training is an effective approach to increase patients’ total level of active participation in healthcare interactions and that some communication behaviors may be more amenable to training (e.g., expressing concerns). Trained patients do not have longer visits and tend to receive more information from their providers. Most studies have found no relationship between communication training and improved health, psychosocial wellbeing, or treatment-related outcomes.

Conclusions

Findings reinforce the importance and potential benefits of patient communication training.

Practice Implications

Additional research is warranted to determine the most efficacious training programs with the strongest potential for dissemination.

1. Introduction

The importance of effective communication in healthcare interactions cannot be overstated. With a movement away from paternalism in the delivery of care over the past several decades, there is ample literature demonstrating the critical role of tailoring communication to the specific values and preferences of patients and their family members [1, 2]. The overwhelming majority of work in healthcare communication has centered on the providers’ side of the clinical encounter, with research aimed at better understanding how providers’ communication impacts patient outcomes [3] and interventions to improve providers’ communication skills [4, 5].

At the same time, training patients to be good communicators remains an essential, yet understudied area. Patient communication training is important for many reasons. Provider-patient encounters are interactive and reciprocal. As noted by Parker and colleagues [6], while the providers’ role is critical, patients are responsible for contributing to the communication process that unfolds. In order to ensure that their perspective is accurately received, patients must be able to effectively communicate their needs, concerns, and preferences. Therefore, to achieve the best possible outcomes, both sides of the healthcare interaction must be engaged and competent communicators [7].

Patient communication training is also important due to the fact that patients encounter certain challenges in interacting with healthcare providers. For example, communication gaps may occur when providers focus on major disease benchmarks at the expense patients’ day-today experience [8, 9] or when providers fail to respond empathically to patients’ expressions of emotion [10, 11]. Research has also shown that ethnic and racial minorities receive suboptimal communication in visits as compared to white patients. For example, African Americans and Latinos rate their visits as less participatory, and physicians may be more likely to adopt a narrowly biomedical communication pattern when interacting with ethnic and racial minority patients [12]. Electronic health records and other technology can also pull providers’ attention away from patients, potentially disrupting the working relationship and impeding effective communication [13, 14].

Three behavioral categories are typically used to operationalize patients’ level of communicative engagement or active participation in healthcare interactions [15]: (1) information seeking and verifying behaviors (e.g., asking questions); (2) assertive statements (e.g., articulating treatment preferences or stating opinions); and (3) expressing emotions or concerns. Empirical findings have linked patients’ level of active participation with a variety of outcomes. For example, breast cancer patients who more frequently state their treatment preferences report higher satisfaction with their providers and experience improved psychological wellbeing [16, 17]. In primary care, active participation is associated with treatment adherence [18]. Patients’ level of active participation also influences provider behavior. Providers offer more information and make more supportive and partnership-building statements to patients who ask questions, express concerns, or otherwise communicate actively in visits [19, 20]. Providers are also more likely to have an accurate perception of patients’ beliefs when interacting with patients who participate more [21].

Despite known benefits of active participation, patients may lack the skills necessary to communicate effectively in healthcare interactions [22]. For example, patients may express emotional concerns using indirect cues that are more difficult for providers to recognize [10, 23]. In addition, asking questions and providing information (e.g., sharing medical history) may come naturally to many patients, while advanced skills such as repeating back or summarizing information that is presented may be more difficult. At the same time, using such information verifying skills can be critical to patients’ understanding of and ability to recall information presented by providers during healthcare interactions.

In light of these points, patient communication training is a potentially effective tool for promoting patient participation and helping patients to maximize their healthcare interactions. Previous reviews of patient communication training literature exist [7, 24–26]. However, these papers are either outdated (e.g., [26]) or topical reviews (e.g., [7]). The purpose of the current systematic review was to present an overview of the status of patient communication training literature. Our goal was to examine the content and structure of existing training programs, the design of intervention studies, and evidence concerning the impact of training on patient behavior, communication process, and other important outcomes.

2. Methods

2.1 Literature Search

We followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines for this review [27]. Systematic literature searches were conducted (September 28, 2015) in six databases for references written in English-only with no specified sex, publication type or date range filters. The databases searched were: MEDLINE (via PubMed), Embase, The Cochrane Library, Web of Science, PsycINFO (via OVID), and ERIC (Education Resource Information Center). For the following databases, both controlled vocabulary and text words were used in the development of the search strategies: PubMed, Embase, The Cochrane Library, and PsycINFO. Web of Science and ERIC do not employ controlled vocabularies so they were searched using keywords only. All search results were combined in a bibliographic management tool (EndNote) and duplicates were eliminated both electronically—using the capabilities in EndNote—and manually, to pick up any duplicates missed by the software. An update search was conducted (September 16, 2016) in the MEDLINE (PubMed) database to capture recent evidence since the original searches took place.

The final search strategy had four components, all of which were linked together with the AND operator: (1) professional-patient relations; (2) patient education, training, coaching, teaching; (3) communication, discussion, interaction; and (4) skills, behavior. Keyword and controlled vocabulary terms (where applicable) were included in the search across all six databases in an effort to be as comprehensive as possible. To reduce publication bias, grey literature sources were retrieved from the Embase, Web of Science, and ERIC databases and included in the reviewing phases of the project. For a complete list of MeSH (Medical Subject Headings) and keyword terms used, please refer to the MEDLINE search strategy accompanying this paper (Appendix 1).

2.2 Review and Abstraction Process

References were screened for inclusion through several phases. All titles and abstracts were independently reviewed by two co-authors. If at least one author coded to include a title or abstract, that reference continued to the next round of review. Each potentially eligible reference was then randomly assigned to a pair of co-authors for full-text review. Disagreements regarding inclusion were resolved via discussion between reviewers for this phase, with the first author serving as a third reviewer when needed. All reviewers then met as a group to resolve any remaining discrepancies and to make decisions regarding the final set of included articles.

References were deemed ineligible if they met any of the following exclusion criteria: (1) review or meta-analysis; (2) commentary, letter to the editor, or similar publication; (3) biomedical or psychotherapy study; (4) did not include a patient training program; or (5) purely descriptive study with no outcomes measuring the impact of training. We also excluded intervention studies that focused solely on question asking. This decision was made for two reasons. First, reviews of question prompt lists in healthcare interactions have been published recently (e.g., [28]). Second, while question asking is an important patient communication skill that should not be overlooked, we were interested in more comprehensive training programs.

The final set of included articles was divided equally among co-authors and relevant data was extracted. Data extracted from each article focused on several areas, including intervention study design, sample characteristics (e.g., sample size, demographics, and disease population), content and structure of the communication training program (e.g., skills taught and mode of delivery), outcomes assessed, and findings reported.

We classified each training program by type using three categories previous defined in patient communication training literature [6, 7]: (1) Materials only; (2) Materials plus individual coaching; and (3) Group-based. Materials only trainings involve learners being presented with some form of educational material (e.g., a communication workbook or leaflet) and nothing else. Materials plus individual coaching trainings add a component of individualized instruction (e.g., meeting with a research staff member to review materials and practice skills). Finally, Group-based trainings entail participants receiving an intervention alongside other learners.

To provide an overview of the content of included training programs, we coded whether or not each intervention specifically targeted communication skills within five behavioral categories. The first four behavioral categories were drawn from the PACE System, a widely used patient communication training curriculum that was initially designed for primary care and subsequently applied in a variety of clinical settings [18, 29]. PACE stands for Presenting information, Asking questions, Checking understanding, and Expressing concerns. Presenting information refers to patients sharing details about symptoms, medical history, medications, and any other pertinent facts. The provision of information skills that fall under this behavioral category help to ensure providers’ accurate understanding of each patient, which can facilitate outcomes such as a correct diagnosis and appropriate treatment recommendations. Asking questions includes information seeking skills, which can work toward enhancing patients’ understanding of their medical situation and what they can expect moving forward. Checking understanding includes information verifying skills (e.g., patients’ restating or summarizing information or asking for clarification) that help to ensure patients understand what a provider is telling them and facilitate retaining that information after the interaction. Expressing concerns includes sharing fears, worries, or emotions as well as disclosing potential conflicts (e.g., religious or cultural beliefs) that might impede treatment or undermine the working relationship between patient and provider. We also coded for targeted skills that fell within a fifth behavioral category, Stating preferences. Initially proposed by Bylund and colleagues [30], Stating preferences includes assertive communication skills pertaining to patients sharing opinions (e.g., about treatment options) and desires (e.g., preferred involvement in decision making).

Study quality was assessed using a modified version of the Downs and Black Study Quality Checklist [31] (i.e., 25 of the original study quality indicators most relevant to the types of studies reviewed).¹ Based on quality, an article was considered fit for inclusion if it met at least 33% of the modified Downs and Black study quality indicators. A randomly selected 25% of articles were double coded to ensure reliability.

3. Results

3.1 Study Characteristics

Figure 1 contains the PRISMA flow diagram for our review. The systematic search resulted in 13,218 references. An additional 19 references were identified through bibliography review. A total of 38 articles met final inclusion criteria. After further review it became clear that, in some cases, two or more articles were published from a single intervention. Therefore, our findings are based on a total of 32 unique intervention studies. Each study possessed at least 52% of the relevant study quality indicators [31] and 22 studies (69%) had 75% or more of these study quality indicators.

Figure 1.

PRISMA flow diagram

Characteristics of the included studies are presented in Table 1. Half of the reviewed studies were conducted in primary care (50%; n = 16). The second most frequently targeted patient population was cancer (28%; n = 9). While the median sample size was 115, three of the included studies had comparatively large samples (ranging from 629 [37] to 2,196 [48]). Most studies were conducted in the United States (88%; n = 28). Two studies were conducted in the United Kingdom [57, 63], one in the Netherlands [33], and one in Indonesia [52].

Table 1.

Summary of intervention study characteristics

Reference	Population	N	Mean age	% Female	Design	Control
[32]	Cancer (mixed; advanced disease)	265	64.4	55	RCT	True
[33]	Cancer (lymphoma)	97	55	60	RCT	True
[34]	HIV	160	45.1	33	Quasi	True
[35]	Primary care (cardiovascular risk; serious mental illness)	17	58.5	53	Quasi	None
[36]	Pediatric surgery (parents)	65	33.1	72.5	RCT	True
[37]	Cancer (mixed; advanced disease)	629	59.5	48.5	RCT	True
[38]	Cancer (breast)	50	50.7	100	RCT	True
[39]	Community (mixed)	38	42.7	84.2	Quasi	None
[40]	Primary care (colorectal cancer screening; low-income and minority)	270	56	63.7	RCT	True
[41]	Primary care (diabetes; African American)	21	66	88	Quasi	None
[42]	Primary care (mixed)	194	50.1	38.5	RCT	True
[43, 44]	Primary care (hypertension)	270	61.3	65.9	RCT	Attention
[30]	Cancer (mixed)	32	53.8	87.5	Quasi	None
[45]	Cancer (mixed)	148	58.2	80.9	RCT	Attention
[46]	Cancer (lung)	151	62	26	RCT	True
[47]	Community (mixed)	40	53.7	77.5	Quasi	None
[48]	Primary care (mixed)	2,196	NR	42.1	RCT	True
[49]	Primary care (mixed; parents)	81	NR	85	Quasi	True
[50]	Community (mixed)	110	53	73	Quasi	None
[51]	Community (cancer)	70	NR	NR	Quasi	None
[52]	Primary care (family planning)	768	NR	100	RCT	Attention
[53]	Cancer (breast)	132	52.5	100	Quasi	Attention
[29]	Primary care (mixed; ≥ 65 yrs old)	33	71.9	57.6	Quasi	True
[18, 54, 55]	Primary care (mixed)	150	45	71.3	RCT	True
[56]	Primary care (mixed)	20	37	70	Quasi	Attention
[57]	Primary care (mixed)	120	42.8	73	RCT	Attention
[58]	Cancer (breast)	60	59.1	100	RCT	Attention
[59]	Pediatric primary care (mixed)	141	8.6	43	RCT	Attention
[60]	Primary care (diabetes)	59	49.7	50	RCT	Attention
[61]	Primary care (mixed)	150	58	0	RCT	Attention
[62]	Gastroenterology (peptic ulcer)	45	55	9	RCT	Attention
[63]	Primary care (mixed)	192	NR	NR	Quasi	True

Notes: NR = not reported; RCT = randomized controlled trial; Quasi = quasi-experimental

The mean age of participants across studies was 52 years. One study specifically targeted pediatric primary care patients [59] and another focused on older adults in primary care [29]. In addition, three studies offered communication training to parents of pediatric patients [36, 49, 59]. Most studies (66%; n = 21) included predominantly or exclusively female participants.

Studies used either a randomized controlled (59%; n = 19) or quasi-experimental design (41%; n = 13). Participants were randomly assigned to either control or experimental conditions in all randomized controlled studies. Quasi-experimental studies either compared control and experimental participants without random assignment to condition or examined changes in participants’ scores pre-training versus post-training. Of the 25 studies that included a control condition, 14 used a true control (i.e., usual care or assessment measures only). The other 11 studies used an attention control. An attention control involved participants being exposed to an intervention that mimicked the amount of time and attention received by the intervention group, but lacked the communication skills training component. For example, Street and colleagues [45] gave control participants a booklet on pain management. A health educator met with control participants to review the booklet, address common misconceptions, and review key aspects of pain-related knowledge. Control participants in this study did not receive communication training content or have an opportunity to role-play communication skills. In analyzing outcomes, the majority of studies (78%; n = 25) conducted comparisons between trained and untrained patient conditions. The remaining studies (22%; n = 7) conducted pre- to post-training comparisons.

3.2 Communication Training Program Characteristics

A summary of communication training program characteristics are presented in Table 2. Materials only trainings accounted for 41% (n = 13) of the interventions and included the use communication workbooks or leaflets [18, 36, 54, 55, 57, 63] or some form of multimedia (e.g., videos [40, 59, 61], web- or computer-based platforms [33, 37, 42, 58], audio CDs [48], or personal digital assistants [38]). Materials plus individual coaching trainings also accounted for 41% (n = 13) of interventions. To give one example, Cegala, Post, and McClure [29] mailed a communication workbook to patients a few days prior to their visit. A research staff member then met with each patient on the day of their visit to answer questions, review the material page by page, and organize an approach to the visit. Group-based trainings accounted for approximately 18% (n = 6) of the interventions and entailed participants attending a community workshop [39, 50, 51] or educational course held across multiple sessions [35, 41, 47].

Table 2.

Summary of communication training program characteristics

Reference	Mode of presentation	Time of delivery	Duration	Behaviors encouraged	Brief description of intervention
Materials only (n = 13)
[33]	Multimedia	Few days pre-visit	Self-paced	Present info; ask questions; check understanding; express concerns; state preferences	Included 58 exemplary videos and a question prompt list; Patients were also able to upload and replay an audio-recorded visit
[36]	Written	Few days pre-visit	Self-paced	Present info; ask questions; check understanding; express concerns	Included sections on providing detailed information about symptoms and treatment, common questions, information verifying skills, and ways of identifying and expressing concerns
[37]	Multimedia	Sometime pre-visit	15 mins	Present info; ask questions; check understanding; express concerns	Utilized written patient narratives that focused on how to prepare for visits, how to talk to doctors, what to ask doctors, and important lifestyle needs to discuss; Patients were able to print questions
[38]	Multimedia	Multiple points	Self-paced	Present info; ask questions; check understanding; express concerns	PDA-based program recorded symptoms, displayed summaries for doctors, and delivered race-concordant videos on how to effectively communicate about pain, depression, and fatigue
[40]	Multimedia	Immediately pre-visit	15 mins	Present info; ask questions; check understanding; express concerns	Provided information about colorectal cancer screening, effective communication skills, and patient testimonials; Two educational leaflets were used to supplement the video
[42]*	Multimedia	N/A	Self-paced	Present info; ask questions; check understanding; express concerns; state preferences	Included 228 exemplary videos; Self-assessment used to identify communication issues and to tailor educational material
[48]*	Multimedia	Immediately pre-visit	20 mins	Ask questions; check understanding; express concerns	Focused on planning and organizing questions/concerns; Encouraged discussion of treatment options, negotiating plan, repeating/checking understanding, and asking questions
[18, 54, 55]	Written	Few days pre-visit	Self-paced	Present info; ask questions; check understanding; express concerns	Included sections on information provision, seeking, and verifying skills; Prompted patients to list topics they wanted to discuss, relevant personal/family medical history, and questions/concerns
[57]	Written	Immediately pre-visit	Self-paced	Ask questions; express concerns	Included sections asking patients to list ideas about their illness and encouraging them to express these ideas, ask questions, and be certain that they understand and agree with the doctor
[58]	Multimedia	Immediately pre-visit	30–45 mins	Ask questions; express concerns; state preferences	Provided information about breast cancer, treatment options, and patient testimonials; Encouraged to use specific communication skills
[59]*	Multimedia	Immediately pre-visit	10 mins	Ask questions; express concerns; state preferences	Focused on importance of considering visit goals; Recommended long-term goal of increasing child’s development as an active participant in his/her care; Provided evidence on the importance of a child-inclusive model of communication; Modeled effective communication skills
[61]	Multimedia	Immediately pre-visit	14 mins	Ask questions; check understanding; express concerns	Displayed an educator presenting information about hypertension; Portions of the video were designed to be complex or unclear (i.e., “confusion points”); Depending on study condition, displayed patient responding to “confusion points” in different ways
[63]	Written	Immediately pre-visit	Self-paced	Ask questions; check understanding	Focused on importance of asking questions and checking understanding; Provided instructions on how to use these skills
Materials + individual coaching (n = 13)
[32]*	Mixed	Multiple points	3.25 hrs total	Ask questions; check understanding; express concerns; state preferences	Pre-visit: Workbook included a list of questions and important issues to share in visits; Guided in formulating questions, requesting clarification, and expressing a desire to participate in care planning; Follow-up: Telephone check-ins to reinforce skills and address concerns; Summary and thank you letter sent
[34]*	Face to face	Immediately pre-visit	< 10–20 mins	Ask questions; express concerns	Explored barriers to medication adherence; Main questions/concerns written on reminder card
[43, 44]*	Face to face	Multiple points	Approximately 1 hr total	Ask questions; express concerns; state preferences	Pre-visit: Asked about desired changes to interactions; Given opportunity to role-play; Provided with a diary to record appointments, medications, and questions; and Guided in identifying sources of support and strategies to overcome barriers; Post-visit: Debriefed immediately after visit; Follow-up: Telephone check-ins, bi-monthly photonovels, and monthly newsletters reinforced skills
[30]	Mixed	Immediately pre- or post-visit	20–30 mins	Present info; ask questions; check understanding; express concerns; state preferences	Focused on presenting detailed information about symptoms, asking questions, checking understanding of information given, expressing emotions and concerns, and stating preferences about treatment and communication; Included a didactic lecture, exemplary videos, and discussion
[45]	Face to face	Immediately pre-visit	20–40 mins	Ask questions; express concerns; state preferences	Assessed needs, goals, and values to tailor messages and exercises; Included a lecture on pain control and communication skills and role-plays
[46]	Mixed	Multiple points	Approximately 20 mins total	Present info	At time of study consent: Video provided information on pain and approaches to self-monitoring and reporting changes in pain; At follow-ups: Received tailored reinforcement coaching in person or via telephone
[49]*	Face to face	Immediately pre-visit	15 mins	Present info; ask questions; check understanding; express concerns	Reviewed and discussed material; Included sections on providing detailed information about how child is feeling, asking questions about illness and treatment, checking understanding about information given, and approaches to expressing concerns about recommended treatment
[52]	Mixed	Immediately pre-visit	19 mins	Ask questions; check understanding; express concerns	Reviewed reason for visit; Discussed any anxiety about talking with provider; Encouraged to ask questions during visit and informed that provider wanted to hear from him/her; Guided through leaflet that included structured exercises and examples of communication skills; Role-played using skills
[53]	Mixed	Immediately pre-visit	20 mins	Ask questions; express concerns; state preferences	Discussed upcoming visit; Recorded questions/concerns, with a flowchart produced that organized questions/concerns based on relevance and priority; Offered an opportunity to role-play
[29]	Face to face	Immediately pre-visit	30 mins	Present info; ask questions; check understanding; express concerns	Reviewed and discussed workbook; Guided in organizing approach to the visit; Included sections on common questions and information verifying skills; Provided space to list questions/concerns
[56]	Face to face	Immediately pre-visit	NR	Present info; ask questions; check understanding; express concerns	Focused on typical communication goals, defined roles of information seeker and provider, and explained how roles shift in visits; Discussed and role-played effective communication skills
[60]	Mixed	Immediately pre-visit	20 mins	Ask questions; state preferences	Reviewed medical chart and treatment plan; Encouraged asking questions and recognizing/negotiating relevant decisions
[62]	Mixed	Immediately pre-visit	20 mins	Ask questions; state preferences	Reviewed medical chart and treatment plan; Encouraged asking questions and recognizing/negotiating relevant decisions
Group-based (n = 6)
[35]*	Mixed	N/A	1.5 hrs × 9 sessions	Present info; ask questions; express concerns; state preferences	Organized into 6 sections: basics of heart health, personal assessment, setting lifestyle goals, making most of visits, communicating effectively, and getting help from family; Included lectures, modeling, exemplary videos, experiential role-plays, and discussion
[39]	Mixed	N/A	1 hr	Present info; ask questions; express concerns	Covered benefits and barriers to CAM-related communication; Included a lecture, exemplary videos, role-play, and discussion; Focused on preparing for visits in advance, initiating CAM- related discussions, disclosing CAM use, and asking questions
[41]	Mixed	N/A	1.5 hrs × 10 sessions	Present info; ask questions; check understanding; state preferences	Divided into 2 content areas: diabetes education/self-management and shared decision-making/communication skills; Included: lectures, audiovisual aids, role-plays, testimonials, and, discussions
[47]	Mixed	N/A	1–3 hrs	Present info; ask questions; express concerns; state preferences	Divided into 6 sections: structure of visits, communication styles, preparing for visits, setting an agenda/telling a biopsychosocial story, asking questions to effectively negotiate, and post-visit problem solving; Included a didactic lecture, exemplary videos, role-plays, and discussion
[50]	Mixed	N/A	2 hrs	Ask questions; express concerns	Divided into 4 sections: barriers to effective communication, examples of communication styles, communication skills, and practice exercises; Included a didactic lecture, exemplary videos, role-plays, provocative questions, and discussion
[51]	Mixed	N/A	7 hrs	Present info; Ask questions; check understanding; express concerns; state preferences	Divided into 3 sections: getting through the diagnosis/prognosis phase, exploring options, and asking difficult questions; Included a didactic lecture and small group discussions

Notes:

^*= included provider training; CAM = complementary and alternative medicine; N/A = not applicable; NR = not reported; PDA = personal digital assistant

Most interventions were given immediately pre-visit (53%; n = 17) and half were one hour or less in duration (50%; n = 16). Approximately 13% (n = 4) of intervention studies delivered content to learners through written materials and 19% (n = 6) used a face-to-face mode of delivery. About a quarter of programs (n = 9) utilized some form of multimedia (e.g., web- or computer-based [33, 37, 42, 58]). Finally, 41% (n = 13) used a mixed mode of presentation. Nearly 38% (n = 12) of intervention studies offered participants an opportunity to practice the communication skills being taught (e.g., using experiential role-play).

The overwhelming majority of interventions (97%; n = 31) targeted skills within multiple behavioral categories. About half of the intervention studies targeted skills in Presenting information (53%; n = 17) and Checking understanding (56%; n = 18). All but one intervention study included Asking questions skills (97%; n = 31) and most targeted Expressing concerns skills (85%; n = 27). Finally, 47% (n = 15) of intervention studies targeted Stating preferences skills.

Twenty-five percent (n = 8) of reviewed studies included a training component for providers. For example, in their training to improve medication adherence among HIV patients, Beach et al. [34] offered providers a one hour didactic lecture on motivational interviewing techniques. Other studies that included a provider intervention focused more specifically on training providers in the use of effective communication skills (e.g., [32, 42, 49]).

3.3 Assessment of Outcomes

Approximately two-thirds of studies (n = 20) examined the impact of training on patient communication behavior. Sixty percent of studies (n = 19) measured communication process outcomes such as visit length or amount of information exchanged. Finally, 75% of studies (n = 24) included a measure of additional outcomes (e.g., psychosocial wellbeing, visit satisfaction, and disease-related knowledge).

Over half of the studies (n = 18) collected and analyzed recordings of provider-patient interactions, with the remaining studies measuring outcomes through self-report questionnaires alone. Of the 18 studies that recorded visits, 14 used a previously validated coding method (e.g., Roter Interaction Analysis System; RIAS; [64]) to analyze this data and four studies used investigator-developed approaches [46, 56, 57, 63].

3.3.1 Communication behavior outcomes

Evidence was strongest for the relationship between communication training and patients’ overall level of participation within interactions. Total active participation was measured by summing patients’ total use of different communication skills or by asking patients to rate their overall level of participation within a visit. Of the ten studies (31%) that included a measure of total active participation, eight found a significant difference between intervention and control groups or in pre-post intervention change.

Looking specifically at the five behavioral categories described above, support was strongest for Expressing concerns skills. Of the five studies (16%) that assessed skills in this behavioral category, four observed a significant difference in favor of training. Findings were mixed for the other behavioral categories. Seven studies (22%) examined Presenting information skills, with four finding no significant differences. Similarly, of the 13 studies (41%) that assessed Asking questions, seven found no significant differences. Eight studies (25%) looked at Checking understanding skills, with half finding a significant difference in favor of training and half finding no differences. Five studies (16%) examined Stating preferences skills and three of those studies found no significant differences.

While low cell counts left us unable to conduct statistical comparisons, brief (i.e., ≤ 20 minutes) and self-paced interventions demonstrated meaningful impacts on patients’ total active participation and the use of specific communication skills. For example, Kim and colleagues [52] offered women a “Smart Patient” coaching session for family planning and contraceptive use discussions. The face-to-face intervention took an average of 19 minutes to complete and was shown to increase the number of questions and concerns articulated by patients. Cegala, Chisolm, and Nwomeh [36] delivered a communication workbook to parents of pediatric surgery patients a few days before their appointment and found that intervention group parents participated significantly more in visits as compared to control group parents. Intervention parents asked more questions and engaged in more information verifying and expressing of concerns. Significant improvements in total active participation were also consistently found within studies that offered communication training to providers [32, 35, 42, 59]. Moreover, many of the provider interventions were brief (e.g., a 15 minute video [59] and a two hour in-person training [32]). Surprisingly, studies that did not offer participants an opportunity to practice skills appeared to be just as likely, if not more than likely, to find a significant increase in total active patient participation as compared to studies that did offer participants an opportunity to practice.

3.3.2 Communication process outcomes

There was notable overlap in several of the process outcomes assessed across studies. Seven studies (22%) examined the impact of training on visit length and six found no statistically significant differences between intervention and control conditions. One study found that trained patients had significantly longer visits as compared to untrained patients [57]. Ten studies (31%) evaluated the amount of information exchanged between providers and patients, with seven observing that trained patients received significantly more information within visits. Such findings were more frequently observed in studies that used a randomized controlled design [18, 42–44, 54, 55, 60, 62] and, surprisingly, those that didn’t offer patients an opportunity to practice skills [18, 29, 34, 42, 54, 55, 60, 62].

Several studies measured outcomes that are consistent with patient-centered care. Four studies (13%) looked at providers’ use of facilitative behaviors such as inviting questions and encouraging patients to share concerns. Of those studies, the two that included a communication training component for providers observed that providers’ facilitative behavior increased when interacting with trained patients [34, 43, 44]. Four studies (13%) assessed shared decision making, either directly [42, 44] or as implied through patient involvement in care [41, 58]. Results were mixed. Half of those studies found that shared decision making was significantly greater among trained patients and half found no difference between trained and untrained patients. Next, five studies (16%) looked at linguistic control within interactions as measured by the ratio of patient to provider statements. Three of these studies found that trained patients were more likely to be verbally dominant as compared to untrained patients. Finally, Epstein and colleagues [32] observed a significant difference in favor of training for a composite measure of patient-centered communication. No significant differences were noted for the individual components that included responding to patients’ emotions, informing patients about prognosis and treatment choices, and balanced framing decisions.

3.3.3 Additional outcomes

Studies also assessed several additional outcomes. Five studies (16%) examined the impact of training on patient communication self-efficacy, with three of those studies observing no significant differences. Ten studies (31%) explored the relationship between patient communication training and some form of patient satisfaction (e.g., communication satisfaction or general satisfaction with care). The majority of those studies (n = 6) found no differences in satisfaction ratings between trained and untrained patients. Six studies (19%) included a measure of at least one health- or treatment-related outcome (e.g., pain severity or hospice utilization). Five of these studies observed no significant differences or mixed findings. An early study conducted by Greenfield et al. [60] found intervention participants demonstrated significant improvement in blood sugar control. This finding was attributed to greater involvement in care and closer adherence to treatment regimens [60]. Six studies (19%) examined psychosocial outcomes (e.g., anxiety, depression, or role limitations). Five of these studies found no differences or mixed results. Greenfield, Kaplan, and Ware [62] delivered a face-to-face individual coaching intervention to gastroenterology patients. Results showed that six to eight weeks after training, experimental group participants reported fewer limitations in physical and role-related activities. Finally, seven studies (22%) evaluated the link between communication training and patients’ level of information recall or disease knowledge. Most of these studies (n = 4) found that trained patients demonstrated significantly greater recall of information presented during interactions or knowledge of their disease.

3.3.4 Assessment of interaction effects

Seven studies (22%) considered the impact of patient characteristics on the effectiveness of communication training. Epstein et al. [32] conducted exploratory analysis and compared intervention effects across pre-specified subgroups (e.g., race and gender). No significant modifier effects were found. Aboumatar and colleagues [43] stratified experimental and control conditions by health literacy level and found that when both physician and patient received communication training, low literate patients asked significantly fewer questions. Results also showed that low literate patients reported significantly lower participatory decision making in conditions where their physician received no communication training. Meropol et al. [37] conducted post-hoc analyses and found that education and physical functioning moderated the effects of communication training on satisfaction. More specifically, among patients with greater than high school education and those with poorer physical functioning at baseline, communication training was associated with significantly greater ratings of communication satisfaction. Findings also suggested a benefit from communication training in reducing decisional conflict among patients with lower physical functioning at baseline.

Kim and colleagues [52] found that older, more educated, higher income, and less assertive women benefited most from receiving communication training. Post et al. [55] explored the impact of race on the effectiveness of communication training. Findings revealed that while communication training had a strong and significant effect on white patients, small to no effects were seen for African American patients. McCann and Weinmann [57] conducted subgroup analyses on age, gender, socioeconomic status, and communication self-efficacy. Findings suggested that increases in question asking were related to lower SES, and that increases in consultation length were associated with younger age, male gender, and lower SES. Finally, Anderson et al. [61] found that patient assertiveness did not moderate the relationship between communication training and patient communication behavior.

4. Discussion and Conclusion

4.1 Discussion

We conducted this systematic review to present an overview of the status of patient communication training literature. Our goal was to examine the content and structure of existing training programs, the design of intervention studies, and evidence concerning the impact of training on patient behavior, communication process, and other important outcomes.

Our findings suggest that patient communication training is an effective approach to increase patients’ total level of active participation in healthcare interactions and that some communication behaviors may be more amenable to training. The majority of studies that targeted and assessed patients’ expression of emotions and concerns observed a significant difference in favor of training. Research has shown that patients report several barriers to disclosing such issues in healthcare interactions, including fears about burdening their provider or disrupting the therapeutic alliance and a belief that these topics fall outside the scope of their medical care [65, 66]. By specifically targeting these active participation behaviors, training programs may not only give patients essential skills, but also help patients recognize that sharing emotions and concerns is desired and expected.

It is unclear why the evidence base across other behavioral categories (i.e., Presenting information, Asking questions, Checking understanding, and Stating preferences) is not as strong. Patients may already be proficient in using skills within some of these behavioral categories and some may be consistent with the established role expectations for medical patients. For example, patients are expected to respond to the questions of providers and to present information about symptoms and medical history during their healthcare interaction. Alternatively, it is possible that even after communication training some patients may view stating preferences as being too assertive. Better understanding of patients’ perspectives and possible barriers to stating their preferences is warranted. Considering the diversity of training programs, it is promising that approximately half of all intervention studies observed significant differences in favor of training within each of these behavioral categories.

An important finding of our review is that patient communication training is not associated with longer visits. Trained patients may communicate more clearly and efficiently in healthcare interactions, which would ultimately benefit healthcare providers who are largely pressed for time and express frustrations with patients’ lack of effective communication [22]. Consistent with this point, results also suggest that trained patients tend to receive more information from their providers.

The relationship between patient communication training and patient-centered care remains unclear. For example, one fundamental component of patient-centered communication is providers’ encouraging patients to ask questions and share concerns, or otherwise working to organize an interaction around the patient’s agenda [67]. Findings from our review were mixed in terms of studies that included providers’ use of facilitative behavior as an outcome. In looking closer at these studies, we see that those with significant findings frequently incorporated a training component for providers and those with non-significant findings often did not. The same is true for studies that measured shared decision making. Although additional work is needed to better understand such results, these studies suggest that patient communication training alone may not be enough to enhance patient-centered care. Patient-centered care is a mutually constructed process and, therefore, intervening with communication training on both sides of the interaction may be necessary to impact outcomes related to patient-centered care [7].

Evidence regarding an association between communication training and patients’ health, psychosocial, or treatment-related outcomes was weak. The pathways between effective communication and distal outcomes such as physical functioning, treatment utilization, and emotional distress are complex, and the former may impact the latter through more indirect routes [3]. Additional work is needed to improve the effectiveness of training programs in achieving mutual agreement, trust, satisfaction, understanding, and other intermediate outcomes that are associated with improved health and wellbeing. For example, better communication may enhance patients’ knowledge of their disease, which may ultimately improve decision making, treatment adherence, and health- and treatment-related outcomes (e.g., pain control) in turn. At the same time, pre-training satisfaction levels were extremely high in many of the intervention studies. A limited range in the distribution of satisfaction scores leaves little room for growth and carries implications for concomitant impacts on long-term outcomes that might otherwise result from improvements in satisfaction. Researchers investigating the benefits of patient communication training must consider such ceiling effects and recruit samples with a wider range of pre-training scores in satisfaction and other constructs that may be associated with health, treatment, and psychosocial outcomes.

Progress has been made since the publication of previous systematic reviews [24, 26]. The number of patient communication training programs continues to grow. In addition to the intervention studies reviewed here, a number of trials are currently underway [68, 69] and other training programs have been developed, but not empirically tested [70–73]. Evidence also suggests that training programs have become more comprehensive. As recommended by Anderson and Sharpe [24], recent training programs have moved beyond a central focus on question asking and now target a broader range of communication skills. An increasing number of studies have also included a training component for providers. While the number of studies is still too few to draw conclusions, this approach enables investigators to consider the added benefit of training both sides of the healthcare interaction.

A number of limitations are also apparent in patient communication training literature. To begin, over three quarters of the reviewed studies were conducted in either primary care or oncology settings. Moving forward, more research is needed to evaluate training programs specifically designed for more diverse disease populations and that can be applied in a wider variety of clinical settings. Given that only four studies were conducted outside of the United States, communication training is clearly needed for patients in others parts of the world. Additional work is also need to develop a better understanding of individual, cultural, and contextual factors that contribute to patients’ response to training. Of the reviewed studies that examined the impact of patient characteristics on the effectiveness of training, findings suggest that age, gender, race, ethnicity, socioeconomic status, education, and health literacy are important and should be considered in future research. We also recommend that future studies explore the influence of training intensity and timing of intervention delivery. Unfortunately, low cell counts left us unable to conduct statistical comparisons that would otherwise help answer these questions. Our preliminary findings do suggest that brief (≤ 20 minutes) and self-paced interventions demonstrated meaningful impacts on patients’ total active participation and the use of specific communication skills.

Trainings may also demonstrate stronger effects by implementing more realistic and intensive experiential exercises. For example, patients may benefit from practicing skills in simulated encounters with standardized providers (i.e., an actor trained to play a physician or other healthcare provider). This approach is used widely in communication training and evaluation for healthcare professionals [74].

Nearly all studies also suffered from a lack of grounding in theory or conceptual framework. One notable exception is the work of Epstein et al. [32], which was based on an ecological model of provider-patient communication [75]. In failing to present the theory or conceptual framework used to develop a training program or to guide study rationale and design, it becomes harder to build a case for how patient communication training influences outcomes. Furthermore, this issue makes it more difficult to select appropriate outcome variables and to bridge findings across studies. Suggestions regarding unifying frameworks have been proposed elsewhere [6, 7] and will not be discussed here. However, this point should be considered to improve research moving forward.

4.2 Conclusions

The current systematic review reinforces the importance and potential benefits of training patients in the use of effective communication skills. Reported findings indicate that communication training is a useful approach to increase patients’ total level of active participation in healthcare interactions and that some communication behaviors may be more amenable to training (e.g., expressing concerns). It appears that trained patients do not have longer visits and tend to receive more information from their providers. Our findings also suggest that there is little evidence to support a link between patient communication training and improved health, psychosocial wellbeing, or treatment-related outcomes.

4.3 Practice implications

We cannot conclude that a superior type of patient communication training program exists or that one method for delivering educational materials is better than another. If anything, the findings of our review indicate that there are many different ways of improving outcomes through patient communication skills training and that more work is needed to determine the most efficacious training programs with the strongest potential for dissemination.

Highlights.

• Reviewed 38 studies describing 32 patient communication intervention studies

• Training increases patients’ total level of active participation

• Trained patients do not have longer visits and receive more information in visits

• Little evidence for improved health, psychosocial wellbeing, or treatment outcomes

• Additional research is warranted to determine most efficacious training programs

Appendix 1. MEDLINE Search Strategy

1. (Patient OR patients OR “Patients”[Mesh] OR “Physician-Patient Relations”[Mesh] OR “Nurse-Patient Relations”[Mesh] OR “Professional-Patient Relations”[Mesh])

2. (Educate* OR educating OR education OR educator OR teach OR teaching OR teacher OR train OR training OR trainer OR intervention* OR workshop* OR trial OR coaching OR “Education”[Mesh] OR “Models, Educational”[Mesh] OR “Patient Education as Topic”[Mesh] OR “Teaching”[Mesh] OR “Intervention Studies”[Mesh])

3. (Communicate* OR communicator OR communicating OR communication OR Conversation* OR Discuss OR discussing OR discussion* OR interaction* OR “Communication”[Mesh] OR “Teach-Back Communication”[Mesh])

4. (“communication skills” OR “conversation skills” OR “discussion skills” OR “communication behavior” OR “communication behaviour” OR “linguistic behavior” OR “linguistic behaviour”)

5. #1 AND #2 AND #3 AND #4