Factors associated with self-perceived burden to the primary caregiver in older patients with hematologic malignancies: an exploratory study.

OBJECTIVE: Although cancer patients frequently experience self-perceived burden to others, this perception has not been enough studied. The aim of this study was to investigate the prevalence of self-perceived burden to the primary caregiver (SPB-PC) and associated factors in an older patient population with hematologic malignancies at the time of chemotherapy initiation. METHODS: In total, 166 consecutive patients with hematologic malignancies aged ≥65 years were recruited at the time of chemotherapy initiation. Patients' SPB-PC was assessed using a 100-mm visual analogue scale (VAS). Characteristics potentially associated with SPB-PC, including sociodemographic and medical characteristics, physical functioning status (Karnofsky performance score, activities of daily living (ADL)/instrumental ADL), symptoms (fatigue, pain, nausea, quality of life), psychological distress (Hospital Anxiety and Depression Scale (HADS)), perceived cognitive function (Functional Assessment of Cancer Therapy Cognitive (FACT-Cog) Scale), and patients'/primary caregivers' personal relationship characteristics (family tie, support), were assessed. RESULTS: Thirty-five percent of patients reported moderate to severe SPB-PC (VAS ≥ 50 mm). Patients' SPB-PC was associated with lower Karnofsky performance (ß = -0.135, p = 0.058) and ADL (ß = -0.148, p = 0.037) scores, and higher HADS (ß = 0.283, p < 0.001) and FACT-Cog perceived cognitive impairments subscale (ß = 0.211, p = 0.004) scores. The proportion of explained variance was 23.5%. CONCLUSIONS: Health care professionals should be aware that about one third of older cancer patients experience moderate to severe SPB-PC at the time of chemotherapy initiation. They should adapt their support of patients who report such a feeling. Copyright © 2016 John Wiley & Sons, Ltd.

Improving anxiety regulation in patients with breast cancer at the beginning of the survivorship period: a randomized clinical trial comparing the benefits of single-component and multiple-component group interventions.

OBJECTIVE: To compare in a multicenter randomized controlled trial the benefits in terms of anxiety regulation of a 15-session single-component group intervention (SGI) based on support with those of a 15-session multiple-component structured manualized group intervention (MGI) combining support with cognitive-behavioral and hypnosis components. METHODS: Patients with nonmetastatic breast cancer were randomly assigned at the beginning of the survivorship period to the SGI (n = 83) or MGI (n = 87). Anxiety regulation was assessed, before and after group interventions, through an anxiety regulation task designed to assess their ability to regulate anxiety psychologically (anxiety levels) and physiologically (heart rates). Questionnaires were used to assess psychological distress, everyday anxiety regulation, and fear of recurrence. Group allocation was computer generated and concealed till baseline completion. RESULTS: Compared with patients in the SGI group (n = 77), patients attending the MGI group (n = 82) showed significantly reduced anxiety after a self-relaxation exercise (P = .006) and after exposure to anxiety triggers (P = .013) and reduced heart rates at different time points throughout the task (P = .001 to P = .047). The MGI participants also reported better everyday anxiety regulation (P = .005), greater use of fear of recurrence-related coping strategies (P = .022), and greater reduction in fear of recurrence-related psychological distress (P = .017) compared with the SGI group. CONCLUSIONS: This study shows that an MGI combining support with cognitive-behavioral techniques and hypnosis is more effective than an SGI based only on support in improving anxiety regulation in patients with breast cancer.

Cognitive compensatory processes of older, clinically fit patients with hematologic malignancies undergoing chemotherapy: A longitudinal cohort study.

OBJECTIVE: Despite the well-known negative impacts of cancer and anticancer therapies on cognitive performance, little is known about the cognitive compensatory processes of older patients with cancer. This study was designed to investigate the cognitive compensatory processes of older, clinically fit patients with hematologic malignancies undergoing chemotherapy. METHODS: We assessed 89 consecutive patients (age ≥ 65 y) without severe cognitive impairment and 89 age-, sex-, and education level-matched healthy controls. Cognitive compensatory processes were investigated by (1) comparing cognitive performance of patients and healthy controls in novel (first exposure to cognitive tasks) and non-novel (second exposure to the same cognitive tasks) contexts, and (2) assessing psychological factors that may facilitate or inhibit cognitive performance, such as motivation, psychological distress, and perceived cognitive performance. We assessed cognitive performance with the Trail-Making, Digit Span and FCSR-IR tests, psychological distress with the Hospital Anxiety and Depression Scale, and perceived cognitive performance with the FACT-Cog questionnaire. RESULTS: In novel and non-novel contexts, average cognitive performances of healthy controls were higher than those of patients and were associated with motivation. Cognitive performance of patients was not associated with investigated psychological factors in the novel context but was associated with motivation and psychological distress in the non-novel context. CONCLUSIONS: Older, clinically fit patients with hematologic malignancies undergoing chemotherapy demonstrated lower cognitive compensatory processes compared to healthy controls. Reducing distress and increasing motivation may improve cognitive compensatory processes of patients in non-novel contexts.

Improving communication in cancer pain management nursing: a randomized controlled study assessing the efficacy of a communication skills training program.

PURPOSE: Effective communication is needed for optimal cancer pain management. This study assessed the efficacy of a general communication skills training program for oncology nurses on communication about pain management. METHODS: A total of 115 nurses were randomly assigned to a training group (TG) or control group (CG). The assessment included the recording of interviews with a simulated cancer patient at baseline for both groups and after training (TG) or 3 months after baseline (CG). Two psychologists rated the content of interview transcripts to assess cancer pain management communication. Group-by-time effects were measured using a generalized estimating equation. RESULTS: Trained nurses asked the simulated patient more questions about emotions associated with pain (relative rate [RR] = 4.28, p = 0.049) and cognitions associated with pain treatment (RR = 3.23, p < 0.001) and used less paternalistic statements about cancer pain management (RR = 0.40, p = 0.006) compared with untrained nurses. CONCLUSIONS: The general communication skills training program improved only a few of the communication strategies needed for optimal cancer pain management in nursing. General communication skills training programs should be consolidated using specific modules focusing on communication skills related to cancer pain management.

Efficacy of a dyadic intervention to improve communication between patients with cancer and their caregivers: A randomized pilot trial.

OBJECTIVE: Cancer-related communication is critical for patients' and caregivers' adaptation to illness. This randomized pilot study was conducted to test the feasibility, acceptability, and efficacy of a specific dyadic intervention to improve communication. METHODS: A four weekly-session intervention was developed to reinforce cancer-related patient-caregiver communication. Patients receiving treatment for any diagnosed cancer, and their caregivers, were recruited from two oncology clinics in Belgium. Sixty-four patient-caregiver dyads were assigned randomly to intervention and waitlist groups. Cancer-related dyadic communication, dyadic coping and emotional distress were assessed at baseline and post-intervention. RESULTS: The intervention attrition rate was 6 %. Linear mixed models were performed on 60 dyads. Significant two-way group × time interaction indicated improvement in participants' cancer-related dyadic communication frequency (ß = -1.30; SE = 0.31; p = .004), self-efficacy (ß = -10.03; SE = 3.90; p = .011) and dyadic coping (ß = -5.93; SE = 2.73; p = .046) after the intervention. CONCLUSION: These results indicate that the brief dyadic communication intervention is feasible and acceptable, and show preliminary evidence of efficacy. PRACTICE IMPLICATIONS: Encouraging patients and caregivers to discuss personal cancer-related concerns may improve their ability to cope with the illness together.

Insight on variables leading to burnout in cancer physicians.

Although communication skills training programs have been recommended to reduce physicians' burnout, few studies have investigated their efficacy. This study assessed the impact of two training programs on cancer physicians' burnout. Especially, it identified some variables leading to burnout in order to develop effective interventions. Burnout was assessed with the Maslach Burnout Inventory. No statistically significant impact of training programs on burnout was observed. The amount of clinical workload and the overuse of some facilitative communication skills were associated with cancer physicians' burnout. The content of such programs must be redefined to reduce burnout.

Predictors and correlates of burnout in residents working with cancer patients.

There are few studies which have investigated variables associated with the development of burnout among residents working with cancer patients. The aim of this study is to identify variables leading to residents' burnout in order to develop effective interventions. Burnout was assessed with Maslach Burnout Inventory. Person- (i.e., emotional-focused coping) and work-related (i.e., changes in lack of organizational support index) variables explain 28% of the variance in changes in emotional exhaustion. Training programs may be improved by adding specific modules for residents, about problem-focused coping in interviewing patients, and for supervisors, about effective team management.

Factors influencing physicians' detection of cancer patients' and relatives' distress: can a communication skills training program improve physicians' detection?

PURPOSE: This study aimed to assess the impact on physicians' detection of patients' and relatives' distress of six 3-h consolidation workshops (CW) following a 2.5-day communication skills basic training (BT) program and to investigate factors associated with detection of distress. METHODS: Physicians, after BT, were randomized to CW or to a waiting list. Physicians' detection of patients' and relatives' distress was measured through differences between physicians' ratings of patients' and relatives' distress (VAS) and patients' and relatives' self-reported distress (HADS). Communication skills were analysed according to the CRCWEM. RESULTS: Mixed-effects modelling of physicians' detection of patients' distress showed a positive group by time effect in favour of physicians in the CW group. Detection of patients' distress was associated negatively with patients' distress, positively with physicians' concurrent use of psychological assessment and supportive skills, and negatively with general assessment skills. Mixed-effects modelling of physicians' detection of relatives' distress showed no significant group by time effect. Detection of relatives' distress was associated negatively with relatives' distress and with general assessment skills. CONCLUSION: CW following a 2.5-day BT are needed to improve physicians' detection of patients' distress in three-person interviews. Results indicate the need to further improve physicians' detection of relatives' distress.

Factors that influence cancer patients' and relatives' anxiety following a three-person medical consultation: impact of a communication skills training program for physicians.

BACKGROUND: No study has yet assessed the impact of physicians' skills acquisition after a communication skills training program on changes in patients' and relatives' anxiety following a three-person medical consultation. This study aimed at comparing, in a randomized study, the impact, on patients' and relatives' anxiety, of a basic communication skills training program and the same program consolidated by consolidation workshops and at investigating physicians' communication variables associated with patients' and relatives' anxiety. METHODS: Consultations with a cancer patient and a relative were recorded and analyzed by the Cancer Research Campaign Workshop Evaluation Manual. Patients' and relatives' anxiety were assessed with the State-Trait Anxiety Inventory-State. RESULTS: No statistically significant change over time and between groups was observed. Mixed-effects modeling of changes in patients' and relatives' anxiety showed that decreases in both patients' and relatives' anxiety were linked with patients' and relatives' self-reported distress (p = 0.031 and 0.005), and that increases in both patients' and relatives' anxiety were linked with physicians' breaking bad news (p = 0.028 and 0.005). CONCLUSION: No impact of the training program was observed. Results indicate the need to further study communication skills which may help reduce patients' and relatives' anxiety especially when breaking bad news.

Development of the LaComm 1.0, A French medical communication analysis software: A study assessing its sensitivity to change.

OBJECTIVE: To test and compare the sensitivity to change of a communication analysis software, the LaComm 1.0, to the CRCWEM's using data from a randomized study assessing the efficacy of a communication skills training program designed for nurses. METHODS: The program assessment included the recording of two-person simulated interviews at baseline and after training or 3 months later. Interview transcripts were analyzed using the CRCWEM and the LaComm 1.0 tools. RESULTS: One hundred and nine oncology nurses (mainly graduated or certified) were included in the study. The CRCWEM detected 5 changes out of 13 expected changes (38%) (e.g., more open directive questions after training) and the LaComm 1.0, 4 changes out of 7 expected changes (57%) (e.g., more empathic statements after training). For open directive question, the effect sizes of the group-by-time changes were slightly different between tools (CRCWEM: Cohen's d=0.97; LaComm 1.0: Cohen's d=0.67). CONCLUSIONS: This study shows that the LaComm 1.0 is sensitive to change. PRACTICE IMPLICATIONS: The LaComm 1.0 is a valid method to assess training effectiveness in French. The use of the Lacomm 1.0 in future French communication skills training programs will allow comparisons of studies.

Predictors of physicians' communication performance in a decision-making encounter with a simulated advanced-stage cancer patient: A longitudinal study.

OBJECTIVES: Physicians' characteristics that influence their communication performance (CP) in decision-making encounters have been rarely studied. In this longitudinal study, predictors of physicians' CP were investigated with a simulated advanced-stage cancer patient. METHODS: Physicians (n=85) performed a decision-making encounter with a simulated patient (SP). Their CP was calculated by analyzing encounter transcripts with validated interaction analysis systems. Potential specific psychological predictors were physicians' empathy towards the SP (Jefferson Scale of Physician Empathy, JSPE) and their decisional conflict about the treatment (Decisional Conflict Scale, DCS). Potential general psychological predictors were physicians' empathy towards cancer patients (JSPE), their decisional conflict about cancer patients' treatments (DCS), and their affective reactions to uncertainty (Physicians' Reactions to Uncertainty, PRU). RESULTS: Physicians' CP was predicted by their decisional conflict about the SP's treatment (DCS) (ß=0.41; p< 0.001) and their affective reactions to uncertainty regarding cancer treatments (PRU) (ß=-0.31; p=0.003). CONCLUSION: During encounters with advanced-stage cancer patients, physicians' awareness of uncertainty about which treatments to consider may facilitate their communication performance, whereas physicians' affective reactions to uncertainty may inhibit their performance. PRACTICE IMPLICATIONS: Physicians' decisional conflict and reactions to uncertainty should be addressed in communication skills training programs.

Is it possible to improve communication around radiotherapy delivery: A randomized study to assess the efficacy of team training?

BACKGROUND AND PURPOSE: The aim of this study was to assess the efficacy of a 38-h communication skills training program designed for multidisciplinary radiotherapy teams. MATERIALS AND METHODS: Four radiotherapy teams were randomly assigned to a training program or to a waiting list. Assessments were scheduled at baseline (T1) and then after the training was completed or four months later (T2), respectively. Communication around radiotherapy delivery was assessed based on audio recordings of the first and last radiotherapy sessions in order to assess team members' communication skills and the expression of concerns by breast cancer patients (analyzed with content analysis software LaComm). RESULTS: 198 radiotherapy sessions were recorded. During the first radiotherapy sessions, members of the trained teams exhibited more assessment skills (p=0.048), provided more setting information (p<0.001), and used more social words (p=0.019) compared to the members of the untrained teams. During the last radiotherapy session, members of the trained teams used more assessment skills (p=0.004) and patients interacting with members of the trained teams expressed more sadness words (p=0.023). CONCLUSION: Training of multidisciplinary teams has the potential to transfer skills that affect the short exchanges that take place around radiotherapy delivery.

How to optimize physicians' communication skills in cancer care: results of a randomized study assessing the usefulness of posttraining consolidation workshops.

PURPOSE: Although there is wide recognition of the usefulness of improving physicians' communication skills, no studies have yet assessed the efficacy of post-training consolidation workshops. This study aims to assess the efficacy of six 3-hour consolidation workshops conducted after a 2.5-day basic training program. METHODS: Physicians, after attending the basic training program, were randomly assigned to consolidation workshops or to a waiting list. Training efficacy was assessed through simulated and actual patient interviews that were audiotaped at baseline and after consolidation workshops for the consolidation-workshop group, and approximately 5 months after the end of basic training for the waiting-list group. Communication skills were assessed according to the Cancer Research Campaign Workshop Evaluation Manual. Patients' perceptions of communication skills improvement were assessed using a 14-item questionnaire. RESULTS: Sixty-three physicians completed the training program. Communication skills improved significantly more in the consolidation-workshop group compared with the waiting-list group. In simulated interviews, group-by-time repeated measures analysis of variance showed a significant increase in open and open directive questions (P =.014) and utterances alerting patients to reality (P =.049), as well as a significant decrease in premature reassurance (P =.042). In actual patient interviews, results revealed a significant increase in acknowledgements (P =.022) and empathic statements (P =.009), in educated guesses (P =.041), and in negotiations (P =.008). Patients interacting with physicians who benefited from consolidation workshops reported higher scores concerning their physicians' understanding of their disease (P =.004). CONCLUSION: Consolidation workshops further improve a communication skills training program's efficacy and facilitate the transfer of acquired skills to clinical practice.

Transfer of communication skills to the workplace: impact of a 38-hour communication skills training program designed for radiotherapy teams.

PURPOSE: This study assessed the efficacy of a 38-hour communication skills training program designed to train a multidisciplinary radiotherapy team. METHODS: Four radiotherapy teams were randomly assigned to a training program or a waiting list. Assessments were scheduled at baseline and after training for the training group and at baseline and 4 months later for the waiting list group. Assessments included an audio recording of a radiotherapy planning session to assess team members' communication skills and expression of concerns of patients with breast cancer (analyzed with content analysis software) and an adapted European Organisation for Research and Treatment of Cancer satisfaction with care questionnaire completed by patients at the end of radiotherapy. RESULTS: Two hundred thirty-seven radiotherapy planning sessions were recorded. Compared with members of the untrained teams, members of the trained teams acquired, over time, more assessment skills (P = .003) and more supportive skills (P = .050) and provided more setting information (P = .010). Over time, patients interacting with members of the trained teams asked more open questions (P = .022), expressed more emotional words (P = .025), and exhibited a higher satisfaction level regarding nurses' interventions (P = .028). CONCLUSION: The 38-hour training program facilitated transfer of team member learned communication skills to the clinical practice and improved patients' satisfaction with care.

[Treatment of cognitive impairments in oncology: a review of longitudinal controlled studies]. / La prise en charge des altérations cognitives en oncologie : une revue des études longitudinales contrôlées.

Various studies highlight cognitive impairments in cancer patients. This paper proposes a review of longitudinal controlled studies evaluating the efficacy of interventions aiming to reduce these cognitive impairments. Longitudinal controlled studies evaluating the efficacy of interventions aiming to reduce cognitive impairments in adult cancer patients and published between 1993 and 2013 were identified, with the exception of studies that implied patients suffering from CNS tumor or metastasis. Pharmacological interventions (n = 11) suggested the positive impact of modafinil on memory and executive functions. Non-pharmacological interventions (n = 10) suggested the positive impact of cognitive revalidation and stimulation programs, psycho-education and meditation on several memory, attentional and executive objective as well as subjective functions. Non-pharmacological interventions show more significant cognitive benefits than pharmacological interventions. Some longitudinal controlled studies support the usefulness of interventions aiming to reduce cognitive impairments in cancer patients. Further studies should evaluate the effectiveness of programs combining technics aiming to reduce cognitive impairments and psychotherapeutic technics aiming to support patients' coping with illness.

The effect of communication skills training on residents' physiological arousal in a breaking bad news simulated task.

OBJECTIVE: Breaking bad news (BBN) is a complex task which involves dealing cognitively with different relevant dimensions and a challenging task which involves dealing with intense emotional contents. No study however has yet assessed in a randomized controlled trial design the effect of a communication skills training on residents' physiological arousal during a BBN task. METHODS: Residents' physiological arousal was measured, in a randomized controlled trial design, by heart rate and salivary cortisol before, during and after a BBN simulated task. RESULTS: Ninety-eight residents were included. MANOVA showed significant group-by-time effects. Trained residents' mean heart rate levels remained elevated after training and cortisol areas under the curve increased after training compared to untrained residents. CONCLUSION: Communication skills training has an effect on residents' physiological arousal. Residents' self-efficacy and communication skills improvements in a BBN simulated task are associated with an elevated physiological arousal, which becomes proportional to the complexity of the task and reflects a better engagement and performance. PRACTICE IMPLICATIONS: Residents should be informed that, to perform a task, they need to engage in the task with a physiological arousal proportional to the complexity of this task. Communication skills training should be adapted.

Is it possible to improve radiotherapy team members' communication skills? A randomized study assessing the efficacy of a 38-h communication skills training program.

BACKGROUND AND PURPOSE: Optimizing communication between radiotherapy team members and patients and between colleagues requires training. This study applies a randomized controlled design to assess the efficacy of a 38-h communication skills training program. MATERIAL AND METHODS: Four radiotherapy teams were randomly assigned either to a training program or to a waiting list. Team members' communication skills and their self-efficacy to communicate in the context of an encounter with a simulated patient were the primary endpoints. These encounters were scheduled at the baseline and after training for the training group, and at the baseline and four months later for the waiting list group. Encounters were audiotaped and transcribed. Transcripts were analyzed with content analysis software (LaComm) and by an independent rater. RESULTS: Eighty team members were included in the study. Compared to untrained team members, trained team members used more turns of speech with content oriented toward available resources in the team (relative rate [RR]=1.38; p=0.023), more assessment utterances (RR=1.69; p<0.001), more empathy (RR=4.05; p=0.037), more negotiation (RR=2.34; p=0.021) and more emotional words (RR=1.32; p=0.030), and their self-efficacy to communicate increased (p=0.024 and p=0.008, respectively). CONCLUSIONS: The training program was effective in improving team members' communication skills and their self-efficacy to communicate in the context of an encounter with a simulated patient. Future study should assess the effect of this training program on communication with actual patients and their satisfaction. Moreover a cost-benefit analysis is needed, before implementing such an intensive training program on a broader scale.

How much is residents' distress detection performance during a clinical round related to their characteristics?

OBJECTIVE: The aim of this study was to investigate residents' characteristics associated with their performance in detecting patients' distress (detection performance). METHODS: Residents' detection performance was assessed in a clinical round. A mean detection performance score was calculated for each resident by comparing residents' rating of patients' distress (VAS) with patients' reported distress (HADS). Residents' characteristics include general (socio-demographic, professional and psychological), detection (self-efficacy, attitudes and outcome expectancies) and performance characteristics (communication skills (LaComm), psychological arousal (STAI) and physiological arousal (heart rate and blood pressure) in a highly emotional and complex simulated interview task). RESULTS: Ninety-four residents and 442 inpatients were included. 30% of the variance in residents' detection performance was related to residents' performance characteristics: anxiety level (p=.040) and mean arterial blood pressure (p=.019) before the task; empathy (p=.027) and mean heart rate (p=.043) during the task; mean arterial blood pressure changes (p=.012) during the assessment procedure. CONCLUSION: Residents' detection performance is partly related to their performance characteristics. Psychological and physiological arousals are key characteristics--beside empathic skills--that need to be considered in models designed to determine detection performance. PRACTICE IMPLICATIONS: Future interventions designed to improve residents' detection performance should focus notably on their performance characteristics.

Transfer of communication skills to the workplace during clinical rounds: impact of a program for residents.

BACKGROUND AND PURPOSE: Communication with patients is a core clinical skill in medicine that can be acquired through communication skills training. Meanwhile, the importance of transfer of communication skills to the workplace has not been sufficiently studied. This study aims to assess the efficacy of a 40-hour training program designed to improve patients' satisfaction and residents' communication skills during their daily clinical rounds. METHODS: Residents were randomly assigned to the training program or to a waiting list. Patients' satisfaction was assessed with a visual analog scale after each visit. Transfer of residents' communication skills was assessed in audiotaped actual inpatient visits during a half-day clinical round. Transcripted audiotapes were analyzed using content analysis software (LaComm). Training effects were tested with Mann-Whitney tests and generalized linear Poisson regression models. RESULTS: Eighty-eight residents were included. First, patients interacting with trained residents reported a higher satisfaction with residents' communication (Median=92) compared to patients interacting with untrained residents (Median=88) (p=.046). Second, trained residents used more assessment utterances (Relative Risk (RR)=1.17; 95% Confidence intervals (95%CI)=1.02-1.34; p=.023). Third, transfer was also observed when residents' training attendance was considered: residents' use of assessment utterances (RR=1.01; 95%CI=1.01-1.02; p=.018) and supportive utterances (RR=0.99; 95%CI=0.98-1.00; p=.042) (respectively 1.15 (RR), 1.08-1.23 (95%CI), p<.001 for empathy and 0.95 (RR), 0.92-0.99 (95%CI), p=.012 for reassurance) was proportional to the number of hours of training attendance. CONCLUSION: The training program improved patients' satisfaction and allowed the transfer of residents' communication skills learning to the workplace. Transfer was directly related to training attendance but remained limited. Future studies should therefore focus on the improvement of the efficacy of communication skills training in order to ensure a more important training effect size on transfer.

Physicians are different when they learn communication skills: influence of the locus of control.

PURPOSE: Although it is widely recognised that educational interventions may be more effective for people with an 'internal' Locus of Control (who believe that life outcomes are controlled by their own characteristics or actions) compared to people with an 'external' Locus of Control (who believe that life outcomes are controlled by external forces such as luck, fate or others), no study has yet assessed the influence of physicians' Locus of Control (LOC) on communication skills learning. This study aims to test the hypothesis that, in a communication skills training program, physicians with an 'internal' LOC would demonstrate communication skills acquisition to a greater degree than those with an 'external' LOC. METHODS: A non-randomised longitudinal intervention study was conducted between January 1999 and April 2001. Sixty-seven volunteer physicians from private and institutional practice in Belgium participated in a learner-centred, skills-focused, practice-oriented communication skills training program. Communication skills changes were assessed in 2 standardised simulated interviews before and after training (one two-person and one three-person interview). Communication skills were assessed using the Cancer Research Campaign Workshop Evaluation Manual. Physicians' LOC was assessed using the Rotter I-E scale. Communication skills changes of the upper and lower third of physicians in respect of their scores on this scale were compared using group by time repeated measures of variance. RESULTS: In the two-person and three-person interviews, changes in the use of open directive questions were more important among physicians with an "internal" LOC compared with changes observed among physicians with an 'external' LOC (P=0.066 and P=0.004, respectively). In the three-person interview, changes in the use of directive questions, assessing functions and moderate feelings stated explicitly were more important among physicians with an 'internal' LOC compared with changes observed among physicians with an 'external' LOC (P= 0.001; P=0.002 and P=0.011 respectively). CONCLUSION: This study shows that physicians' LOC is a psychological characteristic that could influence the efficacy of a communication skills training program. This evidence supports the idea that a psychological characteristic such as 'internal' LOC may facilitate communication skills acquisition through physicians' belief that communication with patients may be controlled by physicians themselves.