| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
ORIGINAL ARTICLES |
From the Departments of Family Medicine (R.M.E., C.G.S., K.F., S.M., P.R.D.), Psychiatry (R.M.E., C.G.S., P.C.D.), and Community and Preventive Medicine (K.F.), University of Rochester Medical Center, Rochester, NY; the Rochester Center to Improve Communication in Health Care (R.M.E., C.G.S., K.F., S.M., P.R.D.), Rochester, NY; and the Department of Community and Family Medicine (P.A.C.), Dartmouth Medical School, Hanover, NH.
Address correspondence and reprint requests to Ronald M. Epstein, MD, 1381 South Avenue, Rochester, NY 14620. E-mail: Ronald_Epstein{at}urmc.rochester.edu
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
Methods: After providing informed consent, 100 primary care physicians in greater Rochester, New York, were visited by two unannounced covert standardized patients (actors, or SPs) portraying two chest pain roles: classic symptoms of gastroesophageal reflux disease (GERD) with nausea and insomnia (the GERD role) and poorly characterized chest pain with fatigue and dizziness (the MUS role). The visits were surreptitiously audiorecorded and analyzed using the Measure of Patient-Centered Communication (MPCC), which scores physicians on their exploration of the patients’ experience of illness (component 1) and psychosocial context (component 2), and their attempts to find common ground on diagnosis and treatment (component 3).
Results: In multivariate analyses, MUS visits yielded significantly lower scores on MPCC component 1 (p = .01). Subanalysis of component 1 scores showed that patients’ symptoms were not explored as fully and that validation was less likely to be used in response to patient concerns in the MUS than in the GERD visits. Component 2 and component 3 were unchanged.
Conclusion: Physicians’ inquiry into and validation of symptoms in patients with MUS was less common compared with more medically straightforward patient presentations. Further research should study the relationship between communication variables and poor clinical outcomes, misunderstandings, mutual distrust, and inappropriate healthcare utilization in this population, and test interventions to address this problem.
Key Words: somatization • physician–patient relations • patient-centered care • symptoms • communication • standardized patients
Abbreviations: GERD = gastrointestinal reflux disease; MPCC = Measure of Patient-Centered Communication; MUS = medically unexplained symptoms; PCC = patient-centered communication.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
Although psychopathology, patient personality, history of trauma or abuse, and sociocultural factors have been associated with somatoform disorders (8–10) and MUS (3,5,11–24), the physician may also unwittingly reinforce the expression of somatic symptoms (25). Qualitative studies suggest several ways that ineffective communication between physicians and patients may underlie some of these patients’ negative experiences. Patients with MUS often provide clues to their psychological concerns, which are rarely acknowledged by the physician (26,27). Physicians tend to attempt to reassure these patients by normalizing their symptoms without providing an adequate explanation for their impressions (28). Cognitive and affective factors may underlie difficulties between physicians and patients with MUS. Dealing with ambiguity, for example, increases the cognitive complexity of the encounter and physician anxiety (29). These, in turn, may conspire to promote premature closure. In an effort to manage their own anxiety, physicians are reported to either reject the patient’s symptoms (or ideas about causation) as not legitimate or collude with the patient’s proposed explanations and requests in an attempt to please the patient. Physicians may interpret patients’ distress as requests for healthcare services of marginal benefit (30,31); physicians may test ostensibly to reassure patients, but may also be motivated by their own anxiety about missing a diagnosis or feelings of inadequacy in the face of unexplained symptoms (32,33). The physician’s collusion may limit consideration of a wider range of diagnostic alternatives (1), whereas premature reassurance may paradoxically raise patients’ anxieties (34).
Prior reports from this study suggest that physicians use one of two communication styles when confronted with ambiguity: usual care, in which ambiguity is denied and closure sought, and a "partnering" approach in which the patient’s experience is understood, ambiguity is acknowledged, and patient input is sought. Partnering is one aspect of a patient-centered communication (PCC) style that emphasizes the need to explore and validate the patient’s illness experience, develop an understanding of the patient as a person, and respect each patient’s unique needs, wishes, and contexts (16). Other PCC skills include coming to agreement on a name for the illness and a plan for follow-up visits, diagnostic testing, and treatment, recognizing that ambiguity about the nature of some symptoms may persist for months or years.
We previously reported that that MUS visits were longer than visits for unambiguous straightforward symptoms. Normally, longer visits are associated with higher scores on measures of PCC (30). However, qualitative findings that physicians tend to discount and misunderstand patients with MUS suggest that visits with patients with MUS would exhibit less PCC. We hypothesized that MUS visits would score lower on measures of PCC even when controlling for important confounders, especially visit time.
We used unannounced covert standardized patients (SPs) to provide physicians with nearly identical stimuli and to avoid three challenges. First are Hawthorne effects—physicians may change their interactional style if they know that they are being observed. Second is case-mix; it is difficult to compare physician behavior with patients who present different symptom clusters. Third is mutual accommodation—patients influence physicians’ communication style. Fourth is self-selection; real patients are likely to select and remain with physicians who share their views (38) and whom they trust (39). Thus, by using unannounced SPs, we could study the physician’s contribution to communication, in particular, self disclosure.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
This analysis is part of a large study that used three different sources of data to arrive at descriptions of physicians’ overall clinical style. We report only on data derived from surreptitious audiorecordings of clinical encounters between unannounced standardized patients and primary care physicians. The methods, physician sample, SP training protocols, and characteristics of real patients who completed surveys (not reported here) have been reported in detail previously (30,36).
Physician Sample
We used personal physician-to-physician recruitment methods, which have demonstrated higher recruitment yields than postal, e-mail, or telephone recruitment (37). In late 1999, 12 physician-recruiters identified primary care physicians in active clinical practice within 45 minutes’ drive of Rochester belonging to a large managed care organization (MCO) serving the eight-county Rochester, New York, region (population 1.1 million). To achieve stable measures of costs for another aspect of the study, only physicians who had greater than 100 patients in the MCO were enrolled; thus, enrolled physicians, compared with nonenrolled physicians, had larger practices. We also intentionally oversampled family physicians to allow for future comparisons between family physicians and internists. A maximum of two physicians per practice were recruited to avoid clustering effects and minimize physician detection of SPs. Thus, 297 of the 594 local physicians were eligible for recruitment. They recruited in random order until a total of 100 physicians were recruited. Physicians gave informed consent to participate in a study of "patient care and outcomes." They agreed to have two unannounced covert surreptitiously audiorecorded SP visits without advance notice at any point in the subsequent 12 months during 2000 to 2001. Physicians were reimbursed $100 for each SP visit (slightly less than usual charges for a new patient acute appointment) and $100 for participation in the study; $100 was also provided to the office staff at each site. The study received Institutional Review Board approval. Physician demographic data are summarized in Table 1.
|
Standardized Patient Role Development
Detailed clinical biographies were developed for the two clinical presentations (gastrointestinal reflux disease [GERD] and MUS). Role outlines were prepared by the coinvestigators and were revised iteratively until they were judged by a consensus of investigators and advisors to be clinically credible and manageable within the context of a 15- to 20-minute new-to-doctor acute visit. We then piloted the roles with family physicians and asked them for their most likely diagnosis, seeking 90% agreement on the GERD diagnosis and less than 30% agreement on any single diagnosis for the MUS role. Standardized patients were middle-aged, white, nonobese men and women, some with professional acting experience. Training focused on depicting the historical and emotional features of the roles, including response to physical examination maneuvers and physician recommendations. Each SP was assigned one of the 2 roles for the entire study and was required to portray role details with 95% accuracy.
Standardized Patient Roles
To compare ambiguous with straightforward patient presentations, we created two contrasting roles; both portrayed a mildly ill patient with chest pain of 2 weeks’ duration who worried about what the symptoms might represent. We attempted to present patients with the same level of illness severity. The "GERD" role portrayed a 48-year-old patient with symptoms of heartburn: nocturnal chest pain exacerbated by food and partially relieved by antacids, typical of GERD. There were secondary concerns about nausea, insomnia (resulting from pain) and fatigue (resulting from insomnia), which clinicians would normally attribute to the underlying GERD. The "MUS" role portrayed a 48-year-old patient with poorly characterized chest pain. There were secondary concerns of dizziness, fatigue, and "not feeling quite right," which clinicians might normally not connect into a single unifying diagnosis; assuming that the symptoms were chronic, this role would meet criteria for multisomatoform disorder (40). Consistent with observations by Salmon et al. (31) that patients with MUS express worry that then leads to physician test-ordering, both roles incorporated an identical prompt from the patient after several minutes of the history-taking portion of the visit that expressed concern that the symptom represented "something serious" followed by more specific concerns about heart disease and cancer. There were no specific requests for testing or treatment in the role scripts. If the physician suggested that stress or anxiety might be related to the symptoms, the SP would endorse such an explanation but not volunteer psychological causation spontaneously, consistent with Kirmayer’s description of "facultative somatization" (41). Recognizing that some patients with GERD share psychological features of patients with MUS (42,43), we constructed the GERD role to exhibit few of these features (e.g., depression, anxiety, history of trauma or abuse, ambiguous or dramatic symptom portrayal, doctor shopping, poor response to treatment). One male and one female were trained to the ambiguous symptoms role; one male and two females were trained to the GERD role when it became apparent that one of the females could not do all of the visits. In pilot testing, the symptom presentations were calibrated so as to communicate patient discomfort and concern but not to prompt referral to an emergency department or administration of medications in the physician’s office.
Standardized Patient Visits
Each physician saw two SPs: one male and one female, and one of each illness condition. The first SP visit was randomly assigned stratified by SP illness condition and gender; the second visit was by an SP of the other gender who portrayed the other illness condition. In that way, each physician could act as his or her own control. Visits occurred at least 2 months from securing consent from participating physicians and were at least 2 months apart.
To ensure realism, SPs were provided factitious insurance cards obtained from local insurance companies; false identities (including pseudonym, local home and work address, and "mobile phone number" corresponding to the cellular phone number of the study coordinator); and cash to make any applicable copayments.
Project staff enlisted practice managers at local clinical sites to help the SPs make medical appointments. Clinic personnel were told that the patient wanted to establish as a "new patient" with the doctor but also had an acute issue ("chest pain for a couple of weeks") that required attention within 1 to 2 weeks.
SPs were monitored throughout training and data collection. Experienced trainers at each site reviewed audiotapes and reporting forms corresponding to each SP’s first six visits plus the first two visits after any sustained break in activity (>1 month). Trainers completed a checklist of behaviors and gave feedback to each SP after each visit.
Two days after the visit, a fax was sent to the physician to determine whether, when prompted, the physician could identify the patient as a SP, not a real patient. The fax notified the physician that a SP had visited in the past few days and prompted the physician to describe any patient that they may now suspect was an SP. If the physician identified the SP after prompting, they were asked to indicate how realistic the SP portrayal was and whether detection occurred before, during, or after the visit. These visits were subject to additional scrutiny. Four physicians from outside of Rochester, blinded to the study hypotheses, were recruited to review randomly selected visits in which the physician identified the SP correctly; we provided paired audiotapes of identified and unidentified SP encounters and asked them to guess which were identified (detected) SPs.
Each SP visit was recorded using a digital audio-disk recorder with a high-quality microphone; equipment was completely hidden within a coat pocket, backpack, or handbag. Visit length was calculated (in minutes) excluding waiting time in the examining room before the visit and any period of more than 1 minute during which the physician left the room.
Measures of Patient-Centered Communication
We analyzed the audiorecordings using the Measure of Patient-Centered Communication (MPCC) (44). The MPCC measures physician responsiveness to patient concerns; component 1 has been positively correlated with patient trust (36) and patient perceptions of patient-centeredness (45).
The MPCC includes three components of patient-centered communication. For component 1 ("exploring both the disease and the illness experience"), the coder notes patient statements that fall into one of six mutually exclusive subcomponents: reason for visit (symptoms) (e.g., "I’ve been having these headaches."), feelings (e.g., "I’m really worried about this."), ideas (e.g., "Could it be because I’m having allergies?"), effect of the symptoms on functioning (e.g., "The headaches wake me up at night"), expectations (e.g., "I just wanted to see if some medication might help."), and prompts (any concern that was repeated to prompt the physician to respond). For each stated patient concern, the rater determines whether it is "cutoff", whether there has been "preliminary exploration" of the concern, "further exploration" (more than one physician question about the concern), or "validation" (physician expression of understanding or empathy). For example, if a patient says, "I have a headache" and the physician completely ignores the concern or immediately changes the topic, it would be coded as a cutoff. If the physician asked for more information, it would be coded as "preliminary exploration," and if the physician asked more than one question exploring the symptom, it would be coded as "further exploration." Saying, "I think I understand" or "This sounds difficult" after having explored the concern would receive the highest number of points—for having "validated" the patient’s concern through expression of empathy, legitimation, understanding, or support. Given that each subcomponent may include several concerns, the subcomponent score is the mean physician response to all concerns within the subcomponent. The mean score for each component is the mean of the subcomponent scores. Component 2 ("understanding the whole person") uses a similar method to measure the degree to which the physician explores the patient’s family, social network, job, and interests. Component 3 ("finding common ground") uses an analogous method to measure the degree to which the physician explains and involves the patient in discussions about the nature of the problem and the management plan. Developers of the scale report interrater reliabilities of 0.80 to 0.83 for the total MPCC score (45); component score reliabilities and correlations among the components are not published.
Two coders were trained to score the audiorecordings using the MPCC; each coded half of the recordings. The coders and the coding supervisor (CGS) were trained in a 20-hour course offered by the developers of the scale. The first 10% were dual-coded for reliability and then an additional 10% at random. In the present study, intraclass correlation coefficients for the total and three components were 0.79, 0.67, 0.89, and 0.43, respectively. There were weekly coding meetings in which all visits were discussed and coding issues were addressed; the developers of the coding system were available for ongoing consultation. Our mean (0.50) and standard deviation (SD) (0.17) of the total MPCC score were identical to those reported by the developers of the scale (45).
Analyses
At least one SP visit recording was available from each of the 100 physicians included in the investigation. The analysis was based on measures taken during 189 SP visits. The 11 missing visits included seven GERD and four MUS portrayals. Reasons for missing visit data included physician withdrawal as a result of practice changes (n = 7) and equipment failure (n = 4). Raw MPCC and component scores were transformed to a standard normal distribution (mean 0; SD 1) to facilitate comparison of effects across measures.
To aid interpretation of effects, outcome measures of MPCC and its components were standardized to have a mean of zero and a SD of one for all analyses. Simple associations between measures of MPCC and role (MUS versus GERD) were explored using paired t-tests and Pearson correlation coefficients using only the 178 measures obtained from the 89 physicians that saw SPs portraying both roles.
Random-effects multiple regression models were used to adjust the estimates for physician age and gender, whether the physician correctly identified the patient as an SP, and visit length. Both physician age and gender modify some measures of communication (35,46,47). Age was dropped from our analyses, because it did not appear to confound the relationship between SP condition and MPCC scores. Gender did modify those relationships and was kept in the models, although it did not achieve statistical significance. Random-effects models allowed us to adjust for the clustering of measures within physicians and to make use of cases in which only one observation was obtained; random-effects models included 189 observations from all 100 physicians. We conducted separate analyses for each of the subcomponents of PCC separately consistent with prior reports by our group and the authors of the scale. All analyses were conducted using SAS version 8.2 (SAS Corp., Cary, NC).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
There were 189 MPCC measures from 100 physicians available for multivariate analyses, including 93 GERD and 96 MUS interviews. In multivariate analyses controlling for physician gender, visit length, and SP detection, MUS visits yielded significantly lower scores on MPCC component 1 (p = .005). There was no significant effect of the patient’s presentation (MUS or GERD) on total MPCC scores or the other components (Table 2).
|
Table 3 shows the frequencies of specific physician response codes in component 1 and the scores adjusted for the number of concerns addressed in each category. These analyses demonstrate that, compared with GERD visits, there were more cutoffs and preliminary and further exploration responses in the MUS visits and an equal number of validation responses. However, because more concerns were coded in the MUS visits for any given concern, the MUS role scored about the same for preliminary explorations and cutoffs but lower for further exploration and validation.
|
Detailed data on retrospective identification of the SP have been presented previously (30). Forty percent of physicians identified the SP when prompted 2 days later. The most common reasons were that the physician had a closed practice (n = 19; 63%) making retrospective identification easier, the physician was notified by the office staff (n = 10, 33%), and poor acting by the SP (n = 1, 3%). Of the detected visits, the average rating of realism was 8.1 on a 10-point scale. Audit of recordings by blinded judges could not distinguish between visits in which the SP was or was not identified. Adjustment for SP identification and realism ratings did not affect the results shown, although for component 1, the association of SP identification with higher MPCC scores approached statistical significance.
Analyses of interaction effects demonstrated no tendency of the scores to converge with longer visits, indicating that differences between the GERD and MUS roles persisted regardless of visit length.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
Previous studies have demonstrated differences in communication related to patient demographic factors (race (48), gender (46,49,50), severity of illness (51), and patient activation (49,52). Although we consider our findings preliminary pending further study, our data support qualitative observations that suggest that physicians’ interactional style is less patient-centered with patients with MUS and that physicians alter their communication style in response to the nature and expression of symptoms. Given that patient-centered communication is associated with greater trust (36), more appropriate use of health services (53), more socioemotional talk (53), and lower likelihood of malpractice litigation (54), our findings suggest that future research should examine whether lack of PCC mediates the distrust, conflict, and inappropriate utilization of healthcare services noted in encounters between patients with MUS and their physicians.
We can only speculate about the reasons for the changes in communication patterns, or lack thereof, in PCC with patients who presented with ambiguous symptoms, but it is likely that cognitive, time management, and attitudinal issues play a role. First, patients presenting the same number of symptoms may require different levels of cognitive effort on the part of the physician depending on whether the symptoms seem coherent from the physician’s perspective. Educational psychologists suggest that complex heuristics or "scripts" are activated by these recognizable and coherent patterns of illness, resulting in efficient diagnostic reasoning (55). However, when the same number of symptoms does not tie together into a familiar pattern of disease, those same "scripts" will either not be helpful or lead the clinician astray; greater cognitive effort is required to resolve the problem, and the physician may experience greater anxiety at not being able to resolve the problem efficiently. Both of these factors conspire toward premature closure.
Second, a greater level of communication skill may be required to explore symptoms that compete for physicians’ time and attention during relatively brief office visits (56). Although visit time is partially under physician control, it is also subject to considerations of the health system, other personnel schedules, and lack of additional reimbursement for longer visits in some settings. Although the MUS visits were longer, they may not have been lengthened enough to accommodate the greater complexity of the consultation. Thus, physicians might benefit from training in agenda setting and organizing the visit in addition to training in self-awareness to recognize when they are tending toward premature closure as a result of cognitive overload or anxiety (57,58).
Third, attitudinal factors such as stigmatizing or discounting patients’ concerns may have resulted in labeling the patients as "hypochondriacs"; at that point, physicians may have used communication moves to circumscribe, shorten, or terminate the visit. Future research using real (not standardized) patients, data on physician personality and response to ambiguity, and interpersonal process recall to understand physicians’ cognitive processes could test the relative contributions of patients, physicians, the health system, and physician reimbursement to these miscommunications.
Prior results from this study showed positive correlations between increased patient-centered communication behaviors (MPCC component 1) observed in SPs and survey measures of patient trust conducted with real patients of the same physician (36). The current analysis suggest further study of whether physician responses, in addition to patient presentation (personality, mental disorders, communication style), contribute to decreased trust between patients with MUS and their physicians (1,27,31,59–61). Diminished trust may play a role in failure of physicians to reach common ground, mutual frustration, and doctor-shopping. Anxious patients who lack trust in their physicians may further amplify their symptoms in an attempt to regain the physician’s attention; anxiety can also worsen the quality and severity of symptoms (22,62).
Study Limitations
Challenges in coding patient-centered communication, limitations of SP methodology, and the cross-sectional observational nature of this study all suggest caution in interpreting these results, which should be treated as preliminary, pending further verification.
Patient-centered communication is difficult to measure; our subsequent analyses suggest that it is not a single construct; even components of it may not be adequately operationalized as a single scale (35). At the time of the study, the MPCC was the only available observational instrument based on a theory of PCC. In this and other analyses from this study (36), only component 1 scores were correlated with meaningful outcomes. Component 3 of the scale had lower interrater reliability and problems with face validity in that it does not explicitly correspond to theories of decision-making developed since the study was initiated (63). The null findings for component 3 of the MPCC may be ascribed to its poor psychometric properties.
The strengths of SP methodology are also its weaknesses. We created one MUS role from a heterogeneous spectrum of possible patient presentations. The choice of the SP roles may have induced or selected for certain specific types of physician behavior. Use of more SP visits would likely have produced higher reliability but would have made physician recruitment and retention very difficult. SPs are new patients; physicians’ behavior with a new patient may not predict their subsequent behavior as the relationship develops over time (64). However, to have used audiotaped real patient visits could have introduced Hawthorne effects and suffered from physician–patient self-selection and accommodation to each other’s interactional styles.
SP detection did not affect the results we report. In a post hoc analysis limited to only unsuspected visits, associations between MUS and component 1 were significant. Nonetheless, the relationship between MPCC scores and prompted suspicion approached statistical significance. This may reflect physicians who altered their behavior when the suspected that they were being observed, that more interpersonally sensitive physicians were more likely to pick up on elements of the visit that might distinguish an SP from a real patient, or that physicians who are likely to detect SPs tend to have closed practices as a result of their popularity with patients. Also, there is an inverse relationship between timing of the inquiry and the reported detection rate (65). We chose to inform physicians 2 days after the visit as a courtesy. However, the proximity of the prompt to the SP visit allowed the physician to choose from a small pool of recent patients that they might recall (availability bias). Data from the same physician pool from a more recent study in which the fax was sent 10 working days after the SP visit resulted in a meaningful detection rate of 12.8% (66). Other studies that report detection rates as low as 2% simply asked the physicians to contact study personnel if they suspected they had seen an SP and were never prompted (67).
We also could not adequately determine whether it was the number of concerns raised by the patient or the nature of symptoms that prompted differences in physicians’ actions; however, the observation that the curves describing the relationship between MPCC and time were parallel for the MUS and GERD conditions provides some comfort that more time would not change underlying physician communication patterns.
The results apply only to those physicians selected into the study sample. Although the patients of enrolled and nonenrolled physicians appear to be similar, the physicians themselves, given their agreement to participate in a relatively intrusive study, are likely to exhibit some important differences from others in the community. Generalization to patients and primary care physicians outside the Rochester area cannot be guaranteed.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
Peter Franks, MD, was instrumental in the design, execution, and analysis of data from other parts of this study. We thank Kit Miller, project coordinator, the standardized patients—Janet Cashin, Mike Kochersberger, Judi Lardner, Alison Venuti, Tom Weber—and the physicians who were willing to expose their clinical interactional styles to scrutiny.
|
NOTES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
Support for this study provided by Patient Centered Care and Health Care Costs, AHRQ R01-HS1610-01A1, R. M. Epstein, Principal Investigator.
DOI:10.1097/01.psy.0000204652.27246.5b
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONCONCLUSIONSNOTESREFERENCES |
|---|
This article has been cited by other articles:
 |
|
 |
|
  L. J. Nogueira Borden, C. L. Adams, and L. D. Ladner The Use of Standardized Clients in Research in the Veterinary Clinical Setting J Vet Med Educ, September 1, 2008; 35(3): 420 - 430. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. olde Hartman, L. Hassink-Franke, C Dowrick, S Fortes, C Lam, H. van der Horst, P. Lucassen, and E. van Weel-Baumgarten Medically unexplained symptoms in family medicine: defining a research agenda. Proceedings from WONCA 2007 Fam. Pract., August 1, 2008; 25(4): 266 - 271. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. M. Nezu, C. M. Nezu, D. Jain, M. S. Xanthopoulos, T. A. Cos, J. Friedman, and M. Lee Social Problem Solving and Noncardiac Chest Pain Psychosom Med, November 1, 2007; 69(9): 944 - 951. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 S. H. McDaniel, H. B. Beckman, D. S. Morse, J. Silberman, D. B. Seaburn, and R. M. Epstein Physician Self-disclosure in Primary Care Visits: Enough About You, What About Me? Arch Intern Med, June 25, 2007; 167(12): 1321 - 1326. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Salmon, G. M. Humphris, A. Ring, J. C. Davies, and C. F. Dowrick Why Do Primary Care Physicians Propose Medical Care to Patients With Medically Unexplained Symptoms? A New Method of Sequence Analysis to Test Theories of Patient Pressure Psychosom Med, July 1, 2006; 68(4): 570 - 577. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
