|
Trust in Physician Scale (TPS)
Please note that this section is an archive (last updated in June 2006). [disclaimer]
Sections:
Overview |
Instrument Reviews |
Construct Overviews |
Book Compendium Reviews |
Internet Site Reviews
|
|
|
|
Practical Information
|
Instrument Name:
|
Trust in Physician Scale (TPS)
|
Instrument Description:
|
The Trust in Physician Scale assesses a patient’s interpersonal trust in one’s physician with 11 items. The developers defined interpersonal trust as “a person’s belief that the physician’s words and actions are credible and can be relied upon.” (Ref: 1)
|
Price:
|
Free. (Ref: 1) Users are requested to send the results of their studies to the author.
|
Administration Time:
|
Approximately 5-8 minutes.
|
Publication Year:
|
1990
|
Item Readability:
|
Flesch-Kincaid Grade Level of 7.0 for items and 9.4 for the instructions. Most items have less than 15 words each.
|
Scale Format:
|
5-point Likert scale.
|
Administration Technique:
|
Interviewer-administered (Ref: 1); Self- and mail-administered (Ref: 2)
|
Scoring and Interpretation:
|
A raw total score after reversing four items is transformed to a 0-to-100 scale score. (Ref: 2) A higher score indicates a higher level of trust in one’s physician.
|
Forms:
|
Thom and others modified wording on one item and labels of the scale. The modified version uses “totally agree” and “totally disagree” instead of “strongly agree” and “strongly disagree” in the scale. (Ref: 2) The modified version by Thom and others had a mean score of 74.7 (on a 0- to-100 scale) with only 5% of participants scoring at the maximum, while the original version had a mean score of 78.9 and 18% of participants scoring at the maximum. (Ref: 2)
|
|
|
Research Contacts
|
Instrument Developers:
|
Lynda A. Anderson and Robert F. Dedrick
|
Instrument Development Location:
|
Department of Health Behavior and Health Education
School of Public Health, The University of Michigan
1420 Washington Heights
Ann Arbor, Michigan 48109-2029
|
Instrument Developer Email:
|
LAnderson4@cdc.gov
|
Instrument Developer Website:
|
No information found.
|
|
|
Annotated Bibliography
|
1. Anderson LA, Dedrick RF. Development of the Trust in Physician scale: a measure to assess interpersonal trust inpatient-physician relationships. Psychol Rep 1990 Dec;67(3 Pt2):1091-100. [PMID:2084735]
|
Purpose:
To develop an instrument that measures a patient’s interpersonal trust in his/her physician, and to collect evidence of reliability and validity.
|
|
|
Sample:
Study 1) N = 160 from an outpatient clinic in the Fayetteville Veterans Administration Medical Center (VAMC), North Carolina. Criteria for inclusion were diagnosis of noninsulin-dependent diabetes mellitus, at least one prior encounter between the physician and patient, and the patient being free of severe cognitive or communicative deficits. Sample patients had a mean age of 55.2 years (SD= 10.5) and an average of 11.8 year (SD = 2.9) of formal education. 78% of participants were married and 56.3% were white.
Study 2) An independent sample of106 patients from the General Medical Clinic of the Durham VAMC, North Carolina. Inclusion criteria were the same as for study 1. Study patients had a mean age of 60.9 years (SD = 9.7), and an average of 10.3 years (SD = 3.5) of formal education. 84% of the participants were married, and 62% were white; 28% were employed, and 56% were retired.
|
|
|
Methods:
Study 1) 25 items were generated from a review of previously developed instruments and interviews with patients and providers. Three dimensions of patient’s trust in physician were targeted: perceived dependability of the physician, confidence in physician’s knowledge and skills, and confidentiality and reliability of shared information. Both negatively and positively worded items were included. The developmental pool of items were administered to 160 participants. The Multidimensional Health Locus of Control scales and a 9-item version of the Marlowe-Crowne Social Desirability Scale were also administered. Items were administered verbally. Only items that evidenced relatively high variance and an item-to-total score correlation of 0.40 or greater were retained. Investigators dropped 2 of the remaining items because of concern that their content focused on outcomes and satisfaction rather than trust in physician.
Study 2) In an interview 1 week prior to a scheduled clinic visit (generally, by telephone), participants completed the Trust in Physician scale, the Multidimensional Health Locus of Control Scale, and the Multidimensional Desire for Control Scale. A post-visit interview within one week of the clinic visit, participants completed the Medical Interview Satisfaction Scale (summary score). Coefficient alpha for the Trust in Physician was .85. Demographic and treatment variables were not significantly associated with score on the Trust in Physician scale.
|
|
|
Implications:
In both studies, coefficient alpha values were high supporting the reliability of the scale. Correlations with other measures were as predicted supporting the construct validity of the scale.
|
2. Thom DH, Ribisl KM, Stewart AL, Luke DA. Further validation and reliability testing of the Trust in Physician Scale. The Stanford Trust Study Physicians. Med Care 1999 May;37(5):510-7. [PMID:10335753]
|
|
Purpose:
To assess reliability and validity of the Trust in Physician Scale.
/td>
|
|
|
Sample:
N = 414 patients and 20 physicians from primary care clinics in northern California. Physicians were recruited first, then patients of these physicians were recruited. These 20 physicians were younger and more likely to be in a group practice compared to physicians in practice in the region. The sample patient characteristic were 47.3 years of age (SD = 16.2), female (62%), high school graduate (93%), and white (67%). The median length of relationship with physicians was 2.2 years.
|
|
|
Methods:
The modified 11-item Trust in Physician Scale was self-administered along with six other measures before the patient saw the physician. A post-visit questionnaire was completed by the patient immediately after he/she saw the physician. The Trust in Physician Scale was mailed to patients one month after the visit, and other measures were mailed to patients six months after the visit. Cronbach’s alpha, item to total correlations, and test/retest correlations (ICC) were calculated. Positive relationships were hypothesized between score on the Trust in Physician scale and the following variables: patients’ satisfaction, physician humanistic behaviors, patients’ active choice of physician, duration of the relationship, preference for physician playing more active role, and expectation for proportion of care from physician. Additionally, the association between scores on the Trust in Physician scale and scores on a general measure of trust (Trust in People Scale) were calculated.
|
|
|
Implications:
The modified version of the Trust in Physician scale exhibited less of a ceiling effect than the original version. Reliability and validity of the measure were supported by the results.
|
3. Freburger JK, Callahan LF, Currey SS, Anderson LA. Use of the Trust in Physician Scale in patients with rheumatic disease; psychometric properties and correlates of trust in the rheumatologist. Arthritis Rheum 2003 Feb 15;49(1):51-8. [PMID:12579593]
|
|
Purpose:
To examine the psychometric properties of the Trust in Physician scale and to identify variables related to rheumatologic patients’ trust in physicians.
|
|
|
Sample:
N = 713 rheumatologic patients who were being treated in North Carolina during the fall of 1999. 741 patients out of 1759 potential participants responded to the mail-delivered questionnaire. Respondents were younger and had more years of education compared to non-respondents. 28 respondents were excluded due to incomplete responses. The participants were predominantly female (77%), and non-Hispanic white (86%). Their mean age, years of education, and duration of disease were 59.58 (SD = 12.68), 13.51 (SD = 2.73), and 10.09 (SD = 7.99) years, respectively.
|
|
|
Methods:
This study was conducted as a part of a study measuring a variety of outcomes in patients with rheumatic disease at the University of North Carolina Hospitals and other private clinics in 1999. The questionnaire was mailed to these participants, and 741 (42%) responded. This questionnaire included Trust in Physician Scale, self-reported health status, Modified Health Assessment Questionnaire (MHAQ), and demographic and disease information. Factor analysis of the Trust in Physician scale items and medical skepticism items was conducted.
|
|
|
Implications:
The reliability, construct, and discriminate validity of the Trust in Physician Scale were supported in this large population of rheumatology patients.
|
|
|
top
|
Factors and Norms
|
Factor Analysis Work:
|
A principal components factor analysis was conducted on the data from 713 rheumatologic patients in North Carolina, using the 11 items from Trust in Physician Scale and the 4 items from the Medical Skepticism scale. The results suggested one factor for each scale, which was expected. (Ref: 3)
|
Normative Information Availability:
|
No information found.
|
|
|
Reliability Evidence
|
Test-retest:
|
Test-retest reliability (ICC) over a one-month period was reported as 0.77. However, the two testings were conducted using different modes of administration (Self-administered vs. Mail-administered). (Ref: 2)
|
Inter-rater:
|
No information found.
|
Internal Consistency:
|
Item-to-total correlations ranged from 0.41 to 0.70 in the developmental study, and Coefficient alpha values of 0.90 and 0.85 were obtained. (Ref: 1) Thom and colleagues modified the scale to decrease ceiling effects. In their study, a Cronbach’s alpha value of 0.89 and item-to-total correlation of 0.39 to 0.72 were obtained. (Ref: 2) Item-to-total correlations of Trust in Physician scale ranged from 0.40 to 0.70, and Cronbach’s alpha of 0.87 was reported. (Ref: 3)
|
Alternate Forms:
|
No information found.
|
|
|
Validity Evidence
|
Construct/ Convergent/ Discriminant:
|
The Trust in Physician Scale correlated with the Multidimensional Health Locus of Control scales as follows: the powerful-other subscale (r = 0.28, p < 0.01; r = 0.38, p < 0.01), internal locus of control (r = -0.07, p > 0.05; r = 0.17, p < 0.05), and chance locus of controls (r = 0.16, p < 0.05; r = 0.06, p > 0.05). Further, correlations with the Multidimensional Desire for Control scales were: Desire for personal control, r = -0.34, p < 0.01; and Desire for clinician, r = 0.48, p < 0.001. Correlation between the Trust in Physician Scale and the Medical Interview Satisfaction Scale was 0.62 (p < 0.001). (Ref: 1) Another study reported that the Trust in Physician Scale was strongly correlated with a subset of the Consumer Satisfaction Survey (r = 0.73, p < 0.001), and with perceived humaneness of physician behavior during the visit (r = 0.68, p < 0.001), while weakly correlated with the Trust in People Scale (r = 0.08, p = 0.22). (Ref: 2)
|
Criterion-related/ Concurrent/ Predictive:
|
Thom and others assessed predictive validity by examining whether baseline Trust in Physician scores predicted continuity with the physician, self-reported adherence to medication, and satisfaction at six months following the physician-patient visit. The researchers found these relationships were statistically significant after adjusting other variables such as age and education; therefore, predictive validity was concluded to be supportive. (Ref: 2)
|
Content:
|
25 items were generated from literature review and interviews with patients and providers. (Ref: 1)
|
Responsiveness Evidence:
|
No information found.
|
Scale Application in VA Populations:
|
Yes. (Ref: 1)
|
Scale Application in non-VA Populations:
|
Yes. (Ref: 2-3)
|
|
|
Comments
|
The TPS is “the first trust measurement instrument specific to the physician-patient relationship,” (Pearson & Raeke , JGIM 2000 (15), p. 510). The TPS is supported by evidence of validity, although primarily with respect to middle-age, white patients with chronic illness. The evidence for the instruments reliability and validity is strong. Hypotheses regarding expected associations with other health measures have been supported suggesting the construct validity of the TPS. The minimal association between a measure of general trust in people and TPS scores supports the discriminate validity of the instrument.
Usefulness for a Certain Population: Since its inception in 1990, the TPS has produced positive results in health outcomes studies involving a variety of patient and physician populations. Most of the evidence we found indicates the usefulness with patients who have chronic illness (e.g., diabetes, rheumatic disease), but the study involving primary care physicians and their mostly well educated, white female patients (Ref: 2) also produced strong evidence of usefulness. The TPS has been used successfully with veteran patients (Ref: 1).
Advantages: The TPS has been used and cited often, is relatively easy to administer and score, and is written to an appropriate reading level for most patient populations.
Disadvantages: The complexity and multidimensional nature of trust cautions against assuming the instrument provides exhaustive or definitive information on patients’ trust in their physician.
Recommendation: The title of the 2000 JGIM article by Pearson and Raeke, “Patients’ trust in physicians: Many theories, few measures, and little data,” still holds true. Because there are few measures of patient trust that have been cross-validated with patients outside the setting in which they were created, those that have, such as the TPS, have grown in popularity.
We recommend the TPS [specifically, the modified version used by Thom (Ref: 2)], especially for use in middle-age patients with chronic disease. Additional testing should evaluate its properties in other populations.
|
Updates
No information found.
|
|
|
|
