If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The Effect of Communication Skills Training for Generalist Palliative Care Providers on Patient-Reported Outcomes and Clinician Behaviors: A Systematic Review and Meta-analysis
Address correspondence to: Dr. Lucy Ellen Selman, Population Health Sciences, Bristol Medical School, University of Bristol, 39 Whatley Road, Bristol BS8 2PS, United Kingdom.
King's College London, Cicely Saunders Institute, London, United KingdomPopulation Health Sciences, Bristol Medical School, University of Bristol, Bristol, United Kingdom
As most end-of-life care is provided by health care providers who are generalists rather than specialists in palliative care, effective communication skills training for generalists is essential.
Objectives
To determine the effect of communication training interventions for generalist palliative care providers on patient-reported outcomes and trainee behaviors.
Methods
Systematic review from searches of 10 databases to December 2015 (MEDLINE, EMBASE, PsycINFO, ERIC, CINAHL, CENTRAL, Web of Science, ICTRP, CORDIS, and OpenGrey) plus hand searching. Randomized controlled trials of training interventions intended to enhance generalists' communication skills in end-of-life care were included. Two authors independently assessed eligibility after screening, extracted data, and graded quality. Data were pooled for meta-analysis using a random-effects model. PRISMA guidelines were followed.
Results
Nineteen of 11,441 articles were eligible, representing 14 trials. Eleven were included in meta-analyses (patients n = 3144, trainees n = 791). Meta-analysis showed no effect on patient outcomes (standardized mean difference [SMD] = 0.10, 95% CI −0.05 to 0.24) and high levels of heterogeneity (chi-square = 21.32, degrees of freedom [df] = 7, P = 0.003; I2 = 67%). The effect on trainee behaviors in simulated interactions (SMD = 0.50, 95% CI 0.19–0.81) was greater than in real patient interactions (SMD = 0.21, 95% CI −0.01 to 0.43) with moderate heterogeneity (chi-square = 8.90, df = 5, P = 0.11; I2 = 44%; chi-square = 5.96, df = 3, P = 0.11; I2 = 50%, respectively). Two interventions with medium effects on showing empathy in real patient interactions included personalized feedback on recorded interactions.
Conclusions
The effect of communication skills training for generalists on patient-reported outcomes remains unclear. Training can improve clinicians' ability to show empathy and discuss emotions, at least in simulated consultations. Personalized feedback on recorded patient interactions may be beneficial.
Given population aging and shortages of palliative care specialists, training generalist palliative care providers is essential to ensure those who require palliative care are able to access it and that EoLC is of high quality.
Parliamentary and Health Service Ombudsman Dying without dignity: Investigations by the Parliamentary and Health Service Ombudsman into complaints about end of life care.
Parliamentary and Health Service Ombudsman,
London2015
for the American College of Physicians High Value Care Task Force Communication about serious illness care goals: a review and synthesis of best practices.
However, evidence regarding the effectiveness of communication skills training interventions is fragmented, and studies have predominantly been small scale, often assessing the effect of training on clinicians' self-reported confidence or attitude rather than patient outcomes or staff behaviors.
The Cochrane Consumers and Communication group in 2016 identified communication in EoLC as the top priority for research in health communication and participation.
Stakeholder priorities for research in health communication and participation: Findings from the Cochrane Consumers and Communication priority setting project.
This reflects the importance of the topic across stakeholder groups, and the need for a comprehensive review integrating effectiveness data. Pulsford et al.
conducted a review of training in EoLC, but the interventions and evaluation methods identified were highly heterogeneous and not specific to communication skills. The review did not differentiate between specialist and generalist training and provided no comparable estimates of the effectiveness of training programs. A systematic review of studies of end-of-life communication interventions published up to March 2014 included interventions not specifically for staff and did not integrate data across studies.
limited their review to training on communication in end-of-life decision making, excluding other important EoLC interventions (e.g., breaking bad news, EoLC preferences and the dying process).
Oncology-specific communication skills training has been reviewed to determine efficacy of training,
The exclusion of noncancer clinical specialties from these reviews also limits their applicability in generalist EoLC: it is estimated that >30% of deaths that need palliative care are from nonmalignant conditions.
We aimed to determine the effectiveness of EoLC communication skills training interventions for generalist palliative care providers, in relation to patient-reported outcomes and real and simulated clinician interactions with patients.
Methods
Protocol and Registration
The systematic review protocol was prospectively registered with PROSPERO.
End of life care communication skills training for generalist palliative care providers: a systematic review.
PROSPERO.2014; (CRD42014014777 2014. Available from http://www.crd.york.ac.uk/PROSPERO_REBRANDING/display_record.asp?ID=CRD42014014777. Accessed August 24, 2017)
Here we report data from randomized controlled trials (RCTs) testing the effect of training interventions on patient-reported outcomes (primary outcome) and clinician behaviors in real and simulated interactions with patients (secondary outcomes). We describe the development, content, and evaluation of training programs across study designs in a separate publication.
A systematic review of end-of-life care communication skills training for generalist palliative care providers: research quality and reporting guidance.
We included studies of training interventions for individuals who work or expect to work with patients with advanced, progressive, incurable illness but do not have and are not training for specialist palliative care qualifications (defined as “generalist” palliative care providers). Examples include family physicians, oncologists, social workers, nurses in hospital and community, care home staff, and volunteers in these settings. Studies including both generalist and specialist palliative care providers were included if generalists accounted for ≥80% participants (to ensure data would be applicable to those who were not specialist palliative care providers) or if their results could be separated from specialists'. Interventions directed at specialist palliative care staff (e.g., hospice nurses, palliative care physicians) or those training to be such specialists were excluded. Studies in which >20% of participants had (or were undertaking) specialist palliative care qualifications and generalist course participant results could not be separated were excluded.
Interventions
We included studies of training interventions intended to enhance generalist palliative care providers' communication skills in relation to EoLC topics. EoLC topics were defined broadly to include issues related to incurable progressive disease and the final stages of advanced disease (e.g., discussing poor prognosis, advance directives, EoLC preferences, and the dying process). We included studies of communication skills training interventions with an EoLC component and EoLC training interventions with a communication skills component. If the content of the training intervention was chosen by course participants, inclusion was assessed by authors' reports of frequently chosen topics. No exclusions were made based on comparison group. Studies of training interventions that did not include any EoLC communication skills, or which were specific to pediatric populations or to communication with individuals other than the patient, were excluded. After protocol registration, we added the following exclusion criteria to ensure relevance: training interventions not delivered to staff/volunteers (e.g., for patients or family members), “train the trainer” programs, or training interventions that occurred alongside extensive system intervention (e.g., change in clinic structure or record keeping), meaning the effect of training alone was unclear.
Study Design and Outcomes
Studies were included if they tested the effectiveness of a training intervention in an RCT and assessed our primary and/or secondary outcomes. Our primary outcome was patient-reported outcomes; secondary outcomes were clinician behaviors in simulated interactions with patients and in real patient interactions. Unpublished studies were included, if sufficient information to satisfy the inclusion criteria could be obtained from the authors. Studies were excluded if they did not test the effectiveness of a training intervention using a randomized controlled design or did not assess effect on either our primary or secondary outcomes. Review articles were excluded. There was no exclusion on the basis of year of publication or language.
Information Sources
The following databases were searched for all available years until December 1, 2015: MEDLINE, EMBASE, and PsycINFO via Ovid; ERIC and CINAHL via EBSCOhost; CENTRAL via Wiley, Web of Science; the WHO International Clinical Trials registry; CORDIS; and OpenGrey.
This was supplemented with hand searching of six relevant reviews,
and the five journals found to be most relevant during scoping: Journal of Palliative Medicine, American Journal of Hospice and Palliative Medicine, Palliative Medicine, Journal of Cancer Education, and Palliative and Supportive Care, from January 1, 2004, to December 1, 2015.
Where searches found published abstracts but no full report, authors were contacted to obtain full study results. If no further results, or insufficient results to determine eligibility, were available, the study was excluded.
Search
Free text terms for searching titles, abstracts, and key words were combined with database-specific subject headings following the structure of [end of life care] AND [communication skills] AND [training]. See Supplementary Figure S1 for an example search strategy for MEDLINE.
Study Selection
Titles and abstracts were screened for inclusion in the review by one author (either L.J.B. or A.H.). Full articles were obtained for studies that could not be excluded based on the information in the title and abstract. Each article was then independently assessed for eligibility by two authors (L.J.B., A.H., C.M., and S.O.), with disagreements resolved through discussion with a third author (L.E.S./J.K.).
Data Collection Process and Data Items
A digital data extraction form was created, piloted, and refined. Data were extracted by one author and independently checked by a second (L.J.B, A.H., C.M., and S.O.) and included general study information (e.g., country, year of publication), how the training interventions were developed, descriptions of the interventions, how they were tested for effectiveness, and the results (see protocol for full list
End of life care communication skills training for generalist palliative care providers: a systematic review.
PROSPERO.2014; (CRD42014014777 2014. Available from http://www.crd.york.ac.uk/PROSPERO_REBRANDING/display_record.asp?ID=CRD42014014777. Accessed August 24, 2017)
The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions.
Statistical power was scored zero or one, where one point is awarded for the presence of power calculation and zero for no evidence of power calculations.
Where sufficient data were available, continuous results were converted to standardized mean differences and dichotomous outcomes into odds ratios (with 95% CIs). To allow comparison across continuous and dichotomous outcomes, odds ratios were converted into standardized mean differences
Scores were reversed where necessary, so that positive effect sizes reflected improvements. Where outcomes were measured multiple times post-training, the latest time point was used. Effect sizes were interpreted as follows: 0.2 = small, 0.5 = medium, and 0.8 = large.
To select the outcome, we used hierarchies, agreed post hoc based on the frequency with which each outcome was assessed in the included RCTs, plus evidence of the importance of outcomes to patients and their families.
A systematic review of prognostic/end-of-life communication with adults in the advanced stages of a life-limiting illness: patient/caregiver preferences for the content, style, and timing of information.
To assess effect on patient outcomes, the hierarchy was anxiety > depression > perceived empathy > satisfaction with communication skills. For real and simulated clinician behavior, the hierarchy was showing empathy > discussing emotions. Standardized mean differences and SEs were pooled using a random-effects model. Assessments for heterogeneity used chi-square test and I2 statistics.
If the data allowed, meta-regression analyses at study level were planned to determine which training and evaluation characteristics explained variations in effectiveness and between-study heterogeneity.
As meta-regression was not possible because of the small number of studies, we tabulated study variables (presence/absence of different teaching methods and length of follow-up) with effect sizes for visual comparison.
funnel plots were planned to assess potential publication bias.
Results
Study Selection and Characteristics
Of 11,441 unique records identified and screened, 845 full-text articles were examined and 19 found eligible, representing 14 RCTs (Fig. 1). Most interventions were palliative/EoLC courses for oncology staff with communication skills components
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Factors affecting patient and clinician satisfaction with the clinical consultation: can communication skills training for clinicians improve satisfaction?.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
PC/EoLC communication skills Nurses and junior doctors
“Codetalk”: 8 × 4-h taught sessions on PC/EoLC communication skills (e.g., talking about advance directives, talking about dying). Participants were given a brief didactic overview followed by role-play demonstrated by faculty, skills practice using simulated patients/families, and reflective discussion. Topics were presented in the context of two unfolding cases, following a patient from diagnosis to death.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
Intervention vs. 'usual education' control (not specified)
194 Staff, 1082 patients, 565 relatives
-
Trainee, patient, and relative measures of quality of communication and end-of-life care
-
Patient and relative depression
-
Patient functional status (within 10 mo)
De La Cruz, USA
PC/EoLC including communication skills Medical students
“Dying wish” film shown to trainees. Tells the story of a retired surgeon with end-stage cancer and raises issues about discussing nutrition at the end of life.
Observed communication skills (simulated interview) (immediate)
Delvaux/Razavi, Belgium
Specialism specific (cancer), including PC/EoLC communication skills Nurses
105 h across 3 wk to teach skills for working in oncology, including PC/EoLC communication. This comprised 20 h theoretical information and 75 h of role-playing exercises (each participant completed four each).
Intervention vs. no training control (waiting list)
115 Staff, 110 patients
-
Observed communication skills (real and simulated interview)
-
Observed responses (real and simulated patients) (immediate, plus delayed 3 mo. Data shown: delayed 3 mo)
Fallowfield/Jenkins/Shilling, UK
Specialism specific (cancer), including PC/EoLC communication skills Doctors
3-Day retreat using role-plays and discussion to teach PC/EoLC communication skills for oncology. At the start of the course, participants reviewed and received feedback on baseline videos of their own, real consultation, plus patient satisfaction scores, and comments on these interactions. They identified communication problems most pertinent to them and worked on these in simulated interactions, video reviews, and group discussion. Randomized to A) course plus additional written feedback, B) course only, C) written feedback only, and D) control.
Factors affecting patient and clinician satisfaction with the clinical consultation: can communication skills training for clinicians improve satisfaction?.
Course vs. no course (inclusion of those with feedback not specified; offered training after)
160 Staff, 861 patients
-
Trainee and patient satisfaction (delayed, 3 mo)
Fujimori, Japan
Specialism specific (cancer), including PC/EoLC communication skills Doctors
2-Day course to teach PC/EoLC communication skills in oncology, utilizing the SHARE communication model (setting, how to deliver bad news, additional information, reassurance, and addressing emotion with empathy). Comprises a 1 h didactic lecture, 8 h role-play with simulated patients (typically tailored to participants' specialties), plus discussions and ice breaking.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Intervention vs. no training control (offered training after)
30 Staff, 601 patients
-
Trainee confidence/self-efficacy
-
Patient satisfaction, trust in oncologist, and distress
-
Observed communication skills (simulated interview) (within 1 wk)
Fukui, Japan
Specialism specific (cancer), including PC/EoLC communication skills Nurses
Two 6-h courses over 3 mo to teach PC/EoLC communication skills in oncology. Used the SPIKES model (setting, assessing patient perception, obtaining patient invitation to disclose information, giving knowledge and information, addressing emotions with empathy, and summarize). Included 2 h didactic lecture, plus role-plays with each other, and facilitated discussions.
Intervention vs. no training control (waiting list)
Eight Staff, 86 patients
-
Patient coping and distress (delayed, post-intervention 1 wk, 1 mo, and 3 mo post- patient cancer diagnosis. Data shown: 3 mo post-diagnosis)
Goelz, Germany
Specialism specific (cancer), including PC/EoLC communication skills Doctors
2-Day course teaching communication skills regarding transition to palliative care in oncology. The workshop comprised primarily skills practice with actors in small groups, focusing on individual learning goals defined by video analysis of baseline interactions with simulated patients. All participants had a 30-min individual coaching session discussing transferring learning goals into practice.
Specialism specific (cancer), including PC/EoLC communication skills Nurses
Six 3-h sessions to teach skills for working with cancer patients, including EOLC communication skills. Comprised theoretical education and role-playing with feedback. Lessons ended with practical homework assignments to complete in practice or at home.
Intervention vs. no training control (waiting list)
46 Staff
-
Observed communication skills (simulated interview) (delayed, 1 mo)
Liénard, Belgium
Communication skills including PC/EoLC Junior doctors
30 h across four sessions over 8 mo to teach communication skills, including PC/EoLC discussions. Of this, 1 h was theoretical didactics and the remaining time was spent participating in role-plays (pre-defined and self-led topics) with immediate feedback. Facilitators adjusted feedback to individuals' skill level and encouraged transfer of skills to clinical practice.
Observed communication skills (real interview) (immediate)
Murray, Canada
PC/EoLC communication skills Nurses
6-Wk course to teach skills for end-of-life care (including communication). Began with an online self-directed module with didactics, case studies, quizzes, and feedback. Next, participants attended a 3-h workshop providing personalized feedback on baseline interactions with simulated patients, appraisal of example scenarios, and skills practice using role-play. A facilitator then called participants 2–3 wk after the workshop to reinforce and support learned behaviors.
Specialism specific (primary care), PC/EoLC including communication skills Doctors
Approximately 96 h of learning time over 3 mo using an online platform (Moodle) to teach PC/EoLC skills (including communication). Each of the four modules included objectives, content directed to clinical practice, PC bibliography and websites, presentations, and self-guided questions.
Knowledge assessment (delayed, 18 mo for trainees, within 18 mo for patients/caregivers)
Szmuilowicz, USA
PC/EoLC communication skills Junior doctors
One 7-h session teaching PC/EoLC communication skills. This included small group discussions, observing an example interaction, didactics, and skills practice. All participants had the opportunity to interview a standardized patient and received feedback from trained faculty.
Observed communication skills (simulated interview) (delayed, within 28 wk: average 14 wk)
Szmuilowicz/Wayne, USA
PC/EoLC communication skills Junior doctors
Two 2-h sessions (original class and booster session) plus online self-study to teach PC/EoLC communication skills, in addition to usual clinical rotations. The original training comprised didactics, observing an example interaction, discussion, skills practice using role-play, provision of self-study materials (including online modules), and completing of a reflective diary regarding two real interactions. The booster sessions included discussion of themes from the reflective diaries, didactics, observation of the example interaction, and group discussion.
Observed communication skills (simulated interview) (delayed, 1 yr)
Tulsky, USA
Specialism specific (cancer), including PC/EoLC communication skills Doctors
5 × 15 min CD-ROM modules to teach communication skills for oncology (including PC/EoLC discussions). Each module included introduction of a skill, video clips demonstrating it, a summary of teaching points, and an opportunity to hear clips from their own baseline audio recording of a real clinic interaction (related to this skill), with tailored feedback.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
using the Hospital Anxiety and Depression Survey), and satisfaction (none used validated measures). Effect sizes could be calculated for all eight studies. Meta-analyses showed no effect of training on anxiety,
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
Factors affecting patient and clinician satisfaction with the clinical consultation: can communication skills training for clinicians improve satisfaction?.
(standardized mean difference [SMD] = 0.10, 95% CI −0.05 to 0.24; Z = 1.33, P = 0.18), although this must be interpreted with caution due to heterogeneity (chi-square = 21.32, degrees of freedom [df] = 7, P = 0.003; I2 = 67%) (Fig. 2). Sensitivity analysis including only high-quality studies showed similar results (SMD = 0.14; 95% CI −0.05 to 0.32, Z = 1.42, P = 0.002). Tabulation of effect sizes with training and evaluation characteristics showed no distinct patterns in relation to use of role-play, personalized feedback on a recorded interaction, duration, or outcome measurement timing (Supplementary Table S2).
Fig. 2Meta-analysis of the effect of training on patient-reported outcomes and experiences.
Effect of Communication Skills Training on Simulated Patient Interactions
Nine RCTs assessed trainees' behaviors during simulated interactions. Some used established dictionaries and coding systems to measure behaviors: Fallowfield et al.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
in simulated interactions (SMD = 0.50, 95% CI 0.19–0.81; Z = 3.19, P = 0.001), with moderate heterogeneity (chi-square = 8.90, df = 5, P = 0.11; I2 = 44%) (Fig. 3). Sensitivity analysis including only high-quality studies showed similar results (SMD = 0.53; 95% CI 0.25 to 0.91, Z = 3.45, P = 0.0006). Tabulation of effect sizes with training and evaluation characteristics showed no distinct patterns (Supplementary Table S3).
Fig. 3Meta-analysis of the effect of training on trainee behaviors in simulated interactions.
Effect of Communication Skills Training on Interactions With Real Patients
Four RCTs measured the impact of training on clinician behaviors during real patient interactions. Most used established dictionaries and coding systems (Fallowfield et al.
(Fig. 4). Meta-analysis indicated a smaller effect of training on behaviors in real patient interactions than was found in simulated interactions, with CIs crossing the line of no effect (SMD = 0.21, 95% CI −0.01 to 0.43, Z = 1.87, P = 0.06). Heterogeneity was moderate (chi-square = 5.96, df = 3, P = 0.11; I2 = 50%). Sensitivity analysis including only high-quality studies showed similar results (SMD = 0.13; 95% CI −0.17 to 0.43, Z = 3.45, P = 0.10). Tabulation of effect sizes with training and evaluation characteristics showed that the two interventions with medium effects on showing empathy in real patient interactions included personalized feedback on a recorded interaction (Supplementary Table S4).
Fig. 4Meta-analysis of the effect of training on trainee behaviors in real interactions.
The online supplementary material contains effect size plots for all patient-reported outcomes and trainee behaviors assessed in the included RCTs (Fig. 2, Fig. 3, Fig. 4).
Risk of Bias (Across Studies)
Because of the small number of studies in the meta-analyses, publication bias could not be assessed.
Discussion
This systematic review is the first to comprehensively examine the effectiveness of EoLC communication skills training interventions for generalist palliative care providers. Evidence regarding the impact of EoLC communication skills training on patient-reported outcomes was inconclusive. Training interventions do appear to be effective at improving physicians' ability to show empathy and discuss emotions. However, the effects of training on clinicians' behaviors during simulated interactions are not reflected in their behaviors when interacting with real patients, with no effect found in the latter. Although in most cases this finding was across different studies using different measures, this pattern is present in the work by Liénard et al.
in which the same measures were administered to the same participants in both simulated and real patient interactions. Measurement during simulated interactions might overestimate clinicians' skills, or skill levels during interactions with real patients might be more difficult to change or measure.
Although eight RCTs measured patient-reported outcomes, there was inconsistency in the constructs assessed and tools used. This may reflect a lack of consensus regarding the primary purpose and theoretical model informing EoLC communication skills training, including its core active components and mechanisms of action. Current RCT evidence suggests there is a potential for positive outcomes and experiences at the patient level. However, overall we found no effect on patient outcomes, with high levels of heterogeneity and a small number of studies. The varied results across studies are likely attributable to not only the selection of different outcomes and use of different measures (only some of which show demonstrable validity and reliability), but also the different timings of measurement.
The importance of EoLC communication skills training for health care staff is demonstrated by the recent flurry of systematic reviews in this area.
Stakeholder priorities for research in health communication and participation: Findings from the Cochrane Consumers and Communication priority setting project.
who found that although communication skills training in oncology can improve professionals' behaviors, including demonstration of empathy, there were greater effects on behaviors in simulated interactions than in real interactions. We identified more RCTs, showing a positive impact on patient outcomes than previously (e.g., Uitterhoeve et al.
Stakeholder priorities for research in health communication and participation: Findings from the Cochrane Consumers and Communication priority setting project.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
However, as in other reviews, the heterogeneity of outcomes and measures used means it is impossible to determine whether our effect estimate is a true representation of training effect or confounded by inconsistent measurement.
Our review has limitations and strengths. Our review benefitted from no restrictions on the basis of language, year of publication, or publication status, and being conducted following PRISMA requirements. Our search strategy was comprehensive and included both oncology and non-oncology training. We also made significant efforts to collect unpublished data by e-mailing at least two co-authors from conference abstracts. Because of the comprehensiveness of our search strategy and the large number of unique records identified, initial screening was completed by only one author. However, the inclusion criteria were applied conservatively at this stage, as seen by the assessment of >800 full texts by two independent authors. The protocol was developed a priori and published online, but refinements to the process were made post hoc, a practice recognized by PRISMA
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.
: we enhanced the exclusion criteria to ensure relevance, included only RCTs, and selected primary and secondary outcomes at this stage. Our original protocol did not specify study design or outcomes as a more inclusive approach was necessary to meet additional review objectives reported elsewhere (e.g., describing methods used to evaluate training).
A systematic review of end-of-life care communication skills training for generalist palliative care providers: research quality and reporting guidance.
A hierarchical approach to meta-analyses was used for two reasons. First, the diversity of outcomes encountered in our searches limited the potential usefulness of a nonhierarchical method. Second, we believed that assessing the impact of studies across patient-reported outcomes and clinician behaviors was valuable given that impact in these areas is vital to ensure care is patient-centered, yet these outcomes are commonly neglected in evaluating communication training interventions. The hierarchies used were formulated based on frequency of measurement, which demonstrates the theoretical relevance of outcomes in this field, and the clinical relevance of outcomes in this population. The outcomes included in the meta-analyses did not always reflect the primary outcome of the particular study. Although additional meta-analyses using solely primary outcomes would have been valuable, only half the studies clearly stated this information. For these reasons and because of the high level of heterogeneity in studies assessing impact on patient outcomes, the findings of the meta-analyses should be interpreted cautiously. The small number of studies in the meta-analyses meant we were unable to formally assess publication bias. However, in our view, selective reporting is likely to be a greater concern than publication bias: not all RCTs stated primary and secondary outcomes, and few cited prospective trial registration details.
Our findings have implications for clinical commissioning and future research. We recommend only commissioning training with proven effectiveness, which means more investment in funding rigorous evaluations of training is needed. In routine practice, evaluation of EoLC training programs should be embedded and ongoing. It is crucial that future research assesses the effectiveness of training interventions at the level of patients and families, the people who suffer most when communication is done poorly. To progress the field, all trials should use validated outcome and experience measures and interactional coding procedures. Standardized ways of describing interventions should be used across trials.
As simulation techniques may be used as a standardized and pragmatic alternative to evaluation using real patient interactions, research is needed to understand why this review demonstrated differences in the measured effects of training interventions on simulated as opposed to real interactions. This finding might be because of measurement factors. Trainees may be “test-ready” for simulated interviews in line with their scoring system, but this might not translate into improvements in real consultations. Lower effect sizes in encounters with real patients may reflect poorer adherence to scoring protocols, but better, more person-centered consultations that are adapted to the individual. Measuring skill levels during encounters with real patients might also be more complex and perhaps less sensitive than in discussions with actors. A fundamental debate here concerns the content validity of objective assessments of communication skills using simulated encounters and prescribed criteria that are often culturally and linguistically specific.
These issues could be explored by analyzing how training changes language and interactional behaviors during simulated and real consultations and how this corresponds to the experience of the patient and their families.
Finally, we found that providing personalized feedback on a recorded interaction with a patient was associated with improvements in clinicians showing empathy in encounters with real patients. A previous review suggested that multifocal interventions (e.g., training plus patient education and altering clinic processes) may be more effective in removing barriers to EoLC communication than training alone.
These types of interventions and models should be subjected to further testing and potentially considered by commissioners to improve communication in EoLC.
Conclusions
Current evidence regarding the effect of EoLC communication skills training for generalist palliative care providers shows no overall effect on patient-reported outcomes. Training generalist staff in EoLC communication skills does appear to improve clinicians' ability to show empathy and discuss emotions; the use of recorded patient interactions may be of particular benefit in this regard. The effects of training on clinicians' behaviors during simulated interactions are not reflected as strongly in their behaviors when interacting with real patients. More patient-centered research and consistency in the use of validated measures is urgently needed to establish best practice.
Disclosures and Acknowledgments
The authors thank the service users in our project advisory group for informing the study and colleagues at the Cicely Saunders Institute for their comments on a previous draft of this article. This systematic review and meta-analysis was funded by Health Education South London as part of the Transforming End of Life Care project. The funders played no role in study design, the conduct of the review, analysis and interpretation of data, the writing of the article, or the decision to submit it for publication. The authors have no competing interests.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Stakeholder priorities for research in health communication and participation: Findings from the Cochrane Consumers and Communication priority setting project.
End of life care communication skills training for generalist palliative care providers: a systematic review.
PROSPERO.2014; (CRD42014014777 2014. Available from http://www.crd.york.ac.uk/PROSPERO_REBRANDING/display_record.asp?ID=CRD42014014777. Accessed August 24, 2017)
A systematic review of end-of-life care communication skills training for generalist palliative care providers: research quality and reporting guidance.
The feasibility of creating a checklist for the assessment of the methodological quality both of randomised and non-randomised studies of health care interventions.
Higgins J.P.T. Green S. Cochrane handbook for systematic reviews of interventions, version 5.1.0. The Cochrane Collaboration,
2011 ([updated March 2011]. Available from www.cochrane-handbook.org. Accessed August 24, 2017)
A systematic review of prognostic/end-of-life communication with adults in the advanced stages of a life-limiting illness: patient/caregiver preferences for the content, style, and timing of information.
Effect of communication skills training program for oncologists based on patient preferences for communication when receiving bad news: a randomized controlled trial.
Factors affecting patient and clinician satisfaction with the clinical consultation: can communication skills training for clinicians improve satisfaction?.
Effect of communication skills training for residents and nurse practitioners on quality of communication with patients with serious illness: a randomized trial.
The PRISMA statement for reporting systematic reviews and meta-analyses of studies that evaluate healthcare interventions: explanation and elaboration.
30311-1&id=gr1.jpg){kind=link}
30311-1&id=gr2.jpg){kind=link}
30311-1&id=gr3.jpg){kind=link}
30311-1&id=gr4.jpg){kind=link}
30311-1&id=fx1.jpg){kind=link}
30311-1&id=fx2.jpg){kind=link}
30311-1&id=fx3.jpg){kind=link}
30311-1&id=fx4.jpg){kind=link}