Abstract
Objective
This study measures diabetes care perceptions of patients and their providers, and examines perceptions differences of patient-provider pairs.
Methods
Patient and provider perceptions were assessed using the Diabetes Semantic Differential Scales (DSDS) which ask respondents to rate diabetes care concepts using contrasting adjective pairs. The DSDS was scored by two methods: using means and using factor analysis. Persons with diabetes 40-years-old or older were recruited. Using a “snowball” sampling strategy, potential provider participants were identified by their patients; 71 providers agreed. These providers represented 51% of the patient participants and created 138 patient-provider pairs.
Results
For the mean scores, there were significant differences between patients and providers for 5 of the 18 semantic differentials (28%). Similarly, the factor scores indicated significant differences for 14 of 54 factors (26%). The effect sizes indicated practical differences.
Conclusion
Significant differences exist between patient and provider perceptions. Generally, patients have the more positive diabetes perceptions.
Practice Implications
During patient and provider discussions, participants can perceive diabetes concepts differently. The DSDS can determine perception differences. While it is best to use factor analyses to score the DSDS, mean scores are more easily calculated and indicate the broad conceptual areas where patient and provider differ.
Keywords: diabetes perceptions, patient-provider interactions, assessing perceptions
1. Introduction
Diabetes self-management relies on the patient’s ability to perform a complex self-care routine including monitoring blood glucose, taking medication, making healthy dietary choices, and participating in regular physical activity (1). In the context of diabetes care, as in virtually all non-surgical medical care, the provider functions as an educator/consultant to the patient who is managing his own illness. Given this unique health care dynamic, mutual understanding and effective communication between the patient and care provider are critical in achieving positive health outcomes (2–3).
Effective patient-provider communication has been found to be positively associated with a number of patient outcome variables including satisfaction and self-care behaviors (4–11). Sherman et al. (6) found that physicians with better communication skills have more satisfied patients, better patient outcomes, and greater satisfaction with their clinical practice. According to a study of provider-patient dialogues conducted by Carter et al (7), better communication and patient knowledge was positively associated with higher adherence to medical recommendations. Similarly, Rost (8) found that increased patient participation in medical interactions was related to an increase in patient satisfaction and adherence to care recommendations.
Effective patient-provider communication is founded on the shared meaning of the terms and concepts used an encounter or, if meanings are not shared, an understanding of how terms and concepts are perceived by each person. According to Kleinman (12), an individual’s explanatory model of illness influences how illness concepts are understood and perceived (by the patient or provider), and in turn influences patient compliance and satisfaction.
To date, little research has investigated similarities or differences between patients’ and providers’ perceptions of diabetes concepts. Diabetes perceptions are how an individual identifies and understands diabetes and are distinguished from diabetes attitudes, which are an individual’s beliefs and feeling about diabetes. A patient’s perception is connotative, i.e., personal and encompassing. Patients’ experience of an illness is expansive; illness is the patients’ subjective perception of disease and is shaped by many factors (13,14). Conversely, a provider’s focus is to determine the patient’s problem and to provide a solution. The provider’s framework organizes the information and provides context for subsequent decisions. A provider’s perception of diabetes is denotative, i.e., explicit and limited.
Diabetes concepts are assigned attributes that are relevant to each individual. For the provider, diabetes concepts and terms may have a more direct and precise medical meaning. A lack of awareness or understanding of each other’s perception of diabetes increases the potential for misinterpretation between patient and provider. Loewe and Freeman (15) found that diabetes patients were more concerned with the visible signs of complications (e.g., wounds that do not heal) than were providers; providers, on the other hand, focused more on the internal (unseen) processes of the disease. A study by Aufseesser et al (16) regarding eight retinopathy terms found that diabetes patients’ understanding of these terms was varied and inaccurate, and the physicians were unaware of these misunderstandings. Even the concept of “diabetes control” can have different meaning for patients and physicians as demonstrated in a study by Hunt, Arar and Larme (17).
The purpose of the present study is to measure and identify differences in the perceptions of diabetes between patients and their providers using the Diabetes Semantic Differential Scales (DSDS). The study also examines two scoring methods to determine whether their results are comparable and appropriate. The research questions explored in this paper are:
Do patients matched with their providers have measurable differences in their perceptions of diabetes care concepts?
Do two scoring methods, means scores and factor analysis scores, yield similar results?
2. Methods
Patient and provider surveys were used to assess diabetes care perceptions. The instruments used are summarized in Table 1 and described below.
Table 1.
Survey Instruments and Measures
| Quantitative Measures | Number of Items | Sample Item |
|---|---|---|
| Diabetes Semantic Differential Scales(DSDS) | 18 | Please rate “Diabetes” |
| Constrained __ : __ : __ : __ : __ : __ : __ Free | ||
| Weak __ : __ : __ : __ : __ : __ : __ Strong | ||
| Dominant __ : __ : __ : __ : __ : __ : __ Submissive | ||
| Worthless __ : __ : __ : __ : __ : __ : __ Valuable | ||
| Difficult __ : __ : __ : __ : __ : __ : __ Easy | ||
| Unsafe __ : __ : __ : __ : __ : __ : __ Safe | ||
| Tense __ : __ : __ : __ : __ : __ : __ Relaxed | ||
| Routine __ : __ : __ : __ : __ : __ : __ Varied | ||
| Independent __ : __ : __ : __ : __ : __ : __ Dependent | ||
|
| ||
| Diabetes Care Profile (DCP) | Sample Item | |
| Positive Attitude | 5 | “I feel satisfied with my life.” |
| Negative Attitude | 6 | “I am afraid of my diabetes.” |
2.1. Measures
2.1.1. The Diabetes Semantic Differential Scales (DSDS)
Semantic differential scales were used to assess patients’ and providers’ perceptions of diabetes concepts. This technique, developed by Osgood, Suci and Tannenbaum (18), asks respondents to rate a specified object or concept using contrasting adjective pairs. The method evaluates the semantic meaning of an object or concept to the individual. Semantic differentials have three components: 1) the concept to be measured, 2) the contrasting adjective pairs to rate the concept, and 3) the undefined rating positions between the adjective pairs. It is suggested that the number of rating positions should be between five and nine with seven positions being the optimal number (19) (See Table 1). The rating positions for each scale are converted to numeric values for analyses. As learned during the pilot testing of the DSDS (20), the relevance of the adjective pairs can only be determined empirically. Adjective pairs appearing to be unrelated to the concept under investigation may prove to be valuable, while pairs appearing to be appropriate may prove to be unremarkable.
The DSDS contains 18 diabetes care concepts. Patients and providers evaluated each of these concepts using nine scales anchored by adjective pairs developed by Wikblad, Wibell and Montin (21). In order to present an unambiguous orientation, there is a patient version and a provider version of the DSDS. The versions differ in only 2 of the 18 diabetes care concepts. In the patient version there is the concept “caring for diabetes”, while in the provider version it is presented as “patient’s self-care for diabetes.” Similarly, the patient version concept of “your emotions about diabetes” was revised to “patient emotions about diabetes” for the provider version.
For each of the nine adjective pairs, the undefined rating positions were converted to numbers (leftmost space = 1, rightmost space = 7) and treated as continuous. Scoring of the DSDS was performed using two methods. The first method was simple and straightforward. Using the nine adjective pairs of each diabetes care concept, a mean response score was calculated for each of the 18 concepts. Low scores were considered to represent more negative perceptions and high scores more positive. The scoring was reversed for 2 of the adjective pairs (the dominant-submissive and the independent-dependent pairs) to make their negative-positive endpoints and scoring consistent with the other adjective pairs. This method provided summary measures that were labeled DSDS “mean scores.”
The second method was the more traditional method of scoring semantic differentials using principal components analysis with varimax rotation of the first three factors. Each construct was expected to contain three orthogonal meaningful dimensions, similar to those found by Osgood, Suci and Tannenbaum (18), i.e., the dimensions of evaluation, potency and activity (19, 22). There were no a priori assumptions about the meaning or strength of the dimensions (how each adjective pair would correlate with any of them), nor was it assumed that the factor structure would be the same for all constructs. Three factor scores were computed for each construct for each patient on the construct’s rotated factors, and then three factor scores were similarly computed for each construct for each provider using the factor structures from the patients’ data. The decision matrix for labeling factors is provided in Table 2. Patient data were used in these analyses with listwise deletion for missing data. Unlike the mean scores, no adjective pair score were reversed for the factor analyses.
Table 2.
Decision Matrix for Factor Labeling
| Low | High | Challenge1 | Power2 | Lifestyle Autonomy3 | Certainty4 | Adaptability5 |
|---|---|---|---|---|---|---|
| Constrained — Free | + | |||||
| Weak — Strong | + | + | ||||
| Dominant — Submissive | − | + | + | |||
| Worthless — Valuable | + | |||||
| Difficult — Easy | + | |||||
| Unsafe — Safe | + | |||||
| Tense — Relaxed | + | |||||
| Routine — Varied | + | + | + | |||
| Independent — Dependent | + | |||||
Required three of the four to be categorized as “Challenge.”
Required two of the three to be categorized as “Power.”
Required two of the three to be categorized as “Lifestyle Autonomy.”
Required both to be categorized as “Lifestyle Autonomy.”
Required both to be categorized as “Adaptability.”
The three factors derived from these analyses were labeled by the investigators with terms descriptive and appropriate for diabetes: challenge, power, lifestyle autonomy, certainty and adaptability. These sub-measures are considered DSDS “factor scores.”
2.1.2. Positive Attitudes and Negative Attitudes Towards Diabetes
The Diabetes Care Profile (DCP) (23) is a diabetes specific instrument that measures social and psychological factors related to diabetes and its treatment. Patients completed two of the fourteen DCP subscales: Positive Attitude Towards Diabetes and Negative Attitude Towards Diabetes. The scores of these subscales range from 1 to 5. Higher scores for the Positive Attitude subscale and lower scores for the Negative Attitude subscale are preferred.
These established scales were used to provide a measure of validation to the two DSDS scoring methods. Diabetes perceptions and diabetes attitudes are not independent from each other. It is expected that an association will exist between the DSDS and the DCP subscales. If the DSDS mean scores are valid measures of perceptions, positive correlations with the Positive Attitude subscale and negative correlations with the Negative Attitude subscale can be expected. The DSDS factor scores should also correlate with these two subscales, but the direction of these correlations will vary, i.e., for the factor scores it is the existence and magnitude of the correlation that is important not whether it is positive or negative.
2.2. Study Participants
2.2.1. Selection and Data Collection
Patient and provider recruitment were interconnected. Patients were recruited from southeastern Michigan, an area that includes the metropolitan Detroit area and the cities of Ann Arbor, Ypsilanti and Jackson. Patient recruitment was limited to individuals 40-years-old or older, not cognitively impaired, under the care of a health care professional, and who had been diagnosed with diabetes mellitus for at least one year. Patients were recruited from Human Subjects Core of the U of M Older Americans Independence Center, and through newspaper ads and flyers. Potential participants were instructed to call a toll-free telephone number in which the project personnel evaluated eligibility. If found eligible, participants were scheduled for an interview (the appointment was conducted in the subject’s home or at a place suggested by either the subject or a research assistant). Before the scheduled interview, a confirmation letter, a consent form (approved by the Institutional Review Board of the University of Michigan Medical School), and the patient survey were mailed to the participant. Patients were paid $35 for their participation. Recruitment continued until the target sample size was attained. Two hundred and seventy-three people agreed to participate and completed the survey and the interview.
A targeted “snowball” sampling strategy was used to recruit health care providers. Snowball sampling is a technique that uses enrolled individuals to recommend or suggest potential participants. In this study, providers were identified by their patients and asked to participate. During their interviews, patients’ were asked to identify and provide the name and location of the single individual that the patient indicates as the primary provider of their diabetes care (whether a physician or a nurse) and permission to contact him/her. These providers were mailed an approved health care provider consent form and survey, a copy of the consent form signed by their patient, and $15 for their participation. One hundred eighty-three unique providers were asked to participate; seventy-one providers (39%) agreed and returned their provider survey. These providers represented 138 patients (51%).
2.2.2. Demographics
Demographics for entire patient sample (n=273) were calculated. Patients’ average age was 62 (±12) years old and 61% were women. The majority of the patient participants were Caucasian (63%), although the proportion of African Americans was substantial (33%). Most had type 2 diabetes (99.6%). The average time since diagnosis was 10 (±9) years.
The demographics of patients whose providers participated in the survey (n=138) were compared to patients whose providers did not participate (because names weren’t obtained or the provider declined/did not respond) (n=135). There were no statistically significant differences between the two patient groups for age, gender, ethnicity, marital status, education, type of diabetes, or time since diagnosis. There were differences for age at onset of diabetes (50±13 years old for patients with provider participation vs. 53±12 years old for patients without provider participation, p=.03) and whether the patient previously had diabetes education (78% for patients with provider participation vs. 66% for patients without provider participation, p=.03).
Providers’ average age was 47 (±10) years old, 44% were women, and 89% were Caucasian. The majority of the provider participants were physicians (96%) and had been in practice for an average of 18 (±11) years. Over half the providers (57%) had less than 25% of their practice devoted to diabetes care.
2.3. Statistical Analyses
The internal reliability of the 18 DSDS mean scores was estimated using Cronbach’s coefficient alpha. All 273 patients were used to calculate these measures, however, the numbers vary due to missing responses.
The 138 paired patient-provider DSDS mean score differences and DSDS factor score differences were determined using paired t-tests. Bonferroni adjustments for multiple statistical tests were used to determine significance levels for both scoring methods. For the DSDS mean scores, a p ≤.003 was used to indicate a significant difference between patient and provider scores. For the DSDS factor scores, a p ≤.001 was used to indicate a significant difference between patient and provider scores.
Effect sizes were used to determine significant practical differences between the scores of patients and providers. Effect sizes of .2, .5 and .8 correspond to small, medium and large effect differences (24).
Pearson correlation coefficients were used to determine relationships between each of the two types of DSDS scoring (mean and factor), and the DCP Positive Attitude Towards Diabetes subscale and the DCP Negative Attitude Towards Diabetes subscale. All 273 patients were used to calculate these correlations. Separate Bonferroni adjustments were used to determine the appropriate significance levels for the DSDS mean score correlations and for DSDS factor score correlations. For the DSDS mean score correlations, a Bonferroni adjustment determined the appropriate significance level to be p ≤ .003. For the DSDS factor score correlations, the appropriate significance level was p ≤ .001. Furthermore, only correlations of absolute value equal to or greater than .25 were considered to be of sufficient strength to be of practical significance.
3. Results
3.1. DSDS Mean Scores
Internal reliability measures of the DSDS mean scores are provided in Table 3. The alphas had a mean and standard deviation of 0.75±0.02 and ranged from 0.70 to 0.80.
Table 3.
Internal Reliability of the DSDS Mean Scores
| DSDS Mean Score | n | Cronbach’s Coefficient Alpha |
|---|---|---|
| Diabetes | 239 | 0.73 |
| Heart Disease and Diabetes | 253 | 0.75 |
| Caring for Diabetes | 253 | 0.73 |
| Low Blood Sugar | 245 | 0.77 |
| Having Diabetes | 260 | 0.76 |
| Blood Sugar Testing | 260 | 0.73 |
| Your Emotions about Diabetes | 257 | 0.77 |
| Controlling your Blood Sugar | 258 | 0.73 |
| High Blood Sugar | 254 | 0.76 |
| Diabetes Complications | 257 | 0.74 |
| Diabetes Diet | 257 | 0.73 |
| Exercise and Diabetes | 262 | 0.74 |
| Using Insulin | 206 | 0.75 |
| Taking Diabetes Pills | 244 | 0.70 |
| Lifelong Disease | 255 | 0.78 |
| Help with Diabetes from Family | 260 | 0.77 |
| Help with Diabetes from Friends | 249 | 0.80 |
| Paying for Diabetes | 254 | 0.75 |
Table 4 provides the mean scores for the patient and provider semantic differentials using the mean response of the nine adjective pairs method of scoring. For five of the eighteen semantic differentials (28%), a difference in patient and provider perceptions was indicated (Low Blood Sugar, Your Emotions about Diabetes, Diabetes Complications, Help with Diabetes from Family, and Paying for Diabetes). The effect sizes for these concepts were moderate indicating that the differences have practical significance. The patients had the more positive (higher) score for each concept.
Table 4.
Diabetes Semantic Differential Scales Mean Scores of Patient and Provider Matches
| Patient and Provider Matches | Patients with Diabetes | Primary Diabetes Care Providers | Effect Size | |
|---|---|---|---|---|
| n | Mean±SD | Mean±SD | ||
| Diabetes | 131 | 4.00 ± 1.03 | 4.14 ± 0.84 | 0.14 |
| Heart Disease and Diabetes | 130 | 3.90 ± 1.01 | 3.85 ± 0.98 | 0.05 |
| Caring for Diabetes | 135 | 4.23 ± 1.03 | 4.08 ± 0.83 | 0.16 |
| Low Blood Sugar | 127 | 3.99 ± 1.08 | 3.56 ± 0.77 | 0.44* |
| Having Diabetes | 134 | 3.67 ± 1.09 | 3.56 ± 0.76 | 0.12 |
| Blood Sugar Testing | 134 | 4.54 ± 1.02 | 4.43 ± 1.06 | 0.10 |
| Your Emotions about Diabetes | 136 | 4.10 ± 1.13 | 3.52 ± 0.92 | 0.54* |
| Controlling your Blood Sugar | 134 | 4.18 ± 1.05 | 4.29 ± 0.78 | 0.12 |
| High Blood Sugar | 132 | 3.65 ± 1.11 | 3.43 ± 0.73 | 0.23 |
| Diabetes Complications | 132 | 3.67 ± 1.12 | 3.24 ± 0.85 | 0.42* |
| Diabetes Diet | 135 | 4.09 ± 1.02 | 4.23 ± 1.00 | 0.14 |
| Exercise and Diabetes | 132 | 4.35 ± 0.97 | 4.60 ± 0.98 | 0.25 |
| Using Insulin | 107 | 3.94 ± 1.10 | 4.29 ± 0.91 | 0.35 |
| Taking Diabetes Pills | 131 | 4.70 ± 1.00 | 4.74 ± 1.01 | 0.04 |
| Lifelong Disease | 132 | 3.60 ± 1.15 | 3.72 ± 1.02 | 0.11 |
| Help with Diabetes from Family | 134 | 4.51 ± 1.16 | 3.86 ± 0.89 | 0.60* |
| Help with Diabetes from Friends | 128 | 4.28 ± 1.11 | 3.91 ± 0.80 | 0.37 |
| Paying for Diabetes | 130 | 4.09 ± 1.06 | 3.46 ± 0.72 | 0.66* |
Paired t-test significant difference at the p level of ≤.003.
3.2. DSDS Factor Analysis Scores
Table 5 provides the three factor scores for the patient and provider semantic differentials. The 3 factors are characterized as “Challenge” (always the first component for each semantic differential), and variously as “Power”, “Lifestyle Autonomy”, “Certainty” or “Adaptability”.
Table 5.
Diabetes Semantic Differential Scales Factor Scores of Patient and Provider Matches
| Factor | Patient and Provider Matches
n |
Patients with Diabetes
Mean±SD |
Primary Diabetes Care Providers
Mean±SD |
Effect Size | |
|---|---|---|---|---|---|
| Diabetes | Challenge | 116 | 0.135 ± 1.047 | 0.109 ± 0.747 | 0.03 |
| Power | 116 | 0.062 ± 0.950 | 0.359 ± 0.914 | 0.31 | |
| Lifestyle Autonomy | 116 | −0.036 ± 1.057 | −0.025 ± 0.779 | 0.01 | |
| Heart Disease and Diabetes | Challenge | 121 | 0.059 ± 1.052 | −0.168 ± 0.910 | 0.23 |
| Power | 121 | 0.041 ± 1.040 | 0.055 ± 0.899 | 0.01 | |
| Lifestyle Autonomy | 121 | 0.009 ± 1.046 | 0.420 ± 1.095 | 0.38 | |
| Caring for Diabetes | Challenge | 123 | 0.066 ± 1.022 | −0.276 ± 0.843 | 0.36 |
| Lifestyle Autonomy | 123 | −0.019 ± 1.064 | −0.209 ± 0.865 | 0.20 | |
| Power | 123 | −0.017 ± 0.981 | 0.214 ± 1.022 | 0.23 | |
| Low Blood Sugar | Challenge | 121 | 0.047 ± 1.040 | −0.327 ± 0.712 | 0.41* |
| Lifestyle Autonomy | 121 | −0.067 ± 1.049 | −0.119 ± 0.888 | 0.05 | |
| Power | 121 | 0.013 ± 0.951 | 0.094 ± 0.928 | 0.09 | |
| Having Diabetes | Challenge | 127 | 0.115 ± 1.046 | −0.066 ± 0.728 | 0.20 |
| Power | 127 | 0.003 ± 1.039 | −0.002 ± 0.773 | 0.01 | |
| Lifestyle Autonomy | 127 | −0.027 ± 1.012 | 0.201 ± 0.778 | 0.25 | |
| Blood Sugar Testing | Challenge | 130 | 0.042 ± 0.963 | −0.133 ± 0.849 | 0.20 |
| Lifestyle Autonomy | 130 | 0.039 ± 1.029 | −0.324 ± 0.949 | 0.36 | |
| Certainty | 130 | 0.124 ± 1.048 | −0.399 ± 0.923 | 0.51* | |
| Your Emotions about Diabetes | Challenge | 126 | 0.053 ± 1.015 | −0.369 ± 0.549 | 0.52* |
| Lifestyle Autonomy | 126 | 0.019 ± 1.048 | −0.615 ± 0.710 | 0.67* | |
| Adaptability | 126 | 0.043 ± 1.026 | 0.144 ± 0.846 | 0.11 | |
| Controlling Your Blood Sugar | Challenge | 127 | 0.070 ± 1.017 | −0.329 ± 0.891 | 0.41* |
| Lifestyle Autonomy | 127 | −0.041 ± 1.008 | 0.128 ± 0.835 | 0.18 | |
| Power | 127 | −0.057 ± 0.944 | 0.383 ± 0.967 | 0.45* | |
| High Blood Sugar | Challenge | 122 | 0.037 ± 1.008 | −0.137 ± 0.641 | 0.20 |
| Power | 122 | 0.007 ± 0.945 | −0.077 ± 0.696 | 0.10 | |
| Lifestyle Autonomy | 122 | −0.027 ± 0.976 | 0.035 ± 0.830 | 0.01 | |
| Diabetes Complications | Challenge | 126 | −0.085 ± 1.028 | 0.325 ± 0.690 | 0.46* |
| Lifestyle Autonomy | 126 | −0.068 ± 1.072 | 0.363 ± 0.964 | 0.41* | |
| Power | 126 | 0.057 ± 1.007 | 0.070 ± 0.989 | 0.01 | |
| Diabetes Diet | Challenge | 126 | 0.015 ± 1.059 | 0.104 ± 0.857 | 0.09 |
| Lifestyle Autonomy | 126 | −0.104 ± 1.048 | 0.137 ± 0.992 | 0.24 | |
| Power | 126 | 0.010 ± 0.971 | −0.233 ± 0.836 | 0.27 | |
| Exercise and Diabetes | Challenge | 110 | 0.183 ± 0.939 | −0.119 ± 0.976 | 0.31 |
| Lifestyle Autonomy | 110 | −0.008 ± 0.881 | 0.484 ± 0.961 | 0.52* | |
| Power | 110 | 0.003 ± 0.968 | −0.182 ± 0.767 | 0.21 | |
| Using Insulin | Challenge | 98 | −0.067 ± 1.002 | −0.243 ± 0.904 | 0.18 |
| Lifestyle Autonomy | 98 | −0.010 ± 0.934 | 0.009 ± 0.861 | 0.02 | |
| Power | 98 | −0.191 ± 0.908 | −0.087 ± 0.739 | 0.13 | |
| Taking Diabetes Pills | Challenge | 119 | 0.000 ± 1.033 | 0.167 ± 0.781 | 0.18 |
| Lifestyle Autonomy | 119 | −0.119 ± 0.991 | −0.295 ± 0.843 | 0.19 | |
| Certainty | 119 | 0.016 ± 1.041 | −0.381 ± 0.860 | 0.41 | |
| Lifelong Disease | Challenge | 123 | 0.021 ± 1.033 | 0.181 ± 0.850 | 0.17 |
| Power | 123 | −0.048 ± 1.025 | 0.250 ± 1.003 | 0.29 | |
| Lifestyle Autonomy | 123 | 0.025 ± 1.034 | −0.298 ± 0.862 | 0.34 | |
| Help with Diabetes from Family | Challenge | 130 | 0.020 ± 1.046 | 0.447 ± 0.605 | 0.49* |
| Lifestyle Autonomy | 130 | 0.013 ± 1.087 | −0.486 ± 0.775 | 0.51* | |
| Certainty | 130 | 0.120 ± 1.014 | 0.067 ± 0.730 | 0.06 | |
| Help with Diabetes from Friends | Challenge | 122 | −0.005 ± 1.056 | 0.273 ± 0.579 | 0.32 |
| Certainty | 122 | −0.021 ± 1.075 | −0.073 ± 0.618 | 0.06 | |
| Lifestyle Autonomy | 122 | −0.006 ± 1.011 | 0.518 ± 0.921 | 0.52* | |
| Paying for Diabetes | Challenge | 122 | 0.093 ± 0.992 | 0.589 ± 0.559 | 0.59* |
| Lifestyle Autonomy | 122 | 0.031 ± 1.074 | 0.096 ± 1.259 | 0.06 | |
| Power | 122 | 0.002 ± 0.945 | 0.768 ± 1.116 | 0.70* |
Paired t-test significant difference at the p level of ≤.001.
Nine of the eighteen semantic differentials had at least one factor where a significant difference between the patient and provider scores was indicated (14 of all 54 factors or 26%). Again, the effect sizes for these concepts were moderate indicating that the differences have practical significance.
3.3. DSDS Scores and Patient Attitudes
3.3.1. DSDS Mean Score Correlations
The DSDS mean score were highly correlated with both DCP attitude scales. Sixteen of 18 correlations (89%) with Positive Attitude were significant (ranging from .29 to .46), while 18 of 18 correlations (100%) with Negative Attitude were significant (ranging from −.26 to −.57).
3.3.2. DSDS Factor Score Correlations
Seventeen of the 54 factor score (31%) had correlations of absolute value equal to or greater than .25 with Positive Attitude ranging from an absolute r of |.25| to |.50|. Eighteen factors score (33%) had a significant correlation with Negative Attitude ranging from an absolute r of |.27| to |.57|.
4. Discussion and Conclusion
4.1. Discussion
The purpose of this study was to measure diabetes care perceptions of patient and providers, to identify differences between a patient and their providers, and to examine the two scoring methods. Given that the patients and providers in this study were self-selected, results should be interpreted with caution, as they may not be generalizable to other populations.
The first research question sought to identify measurable differences between patients and providers in their perceptions of diabetes care concepts. The results of the paired t-tests indicate that perception differences do exist. Significant differences between the patient and provider perceptions were found for approximately 25% of the diabetes concepts. These differences were consistent across both scoring methods. The mean score differences suggest the patients are more positive than their providers for the concepts of Low Blood Sugar, Your Emotions about Diabetes, Diabetes Complications, Help with Diabetes from Family, and Paying for Diabetes.
Factor score differences generally support the mean score differences; Low Blood Sugar’s Challenge factor, Your Emotions about Diabetes’ Challenge and Lifestyle Autonomy factors, Help with Diabetes from Family’s Challenge and Lifestyle Autonomy factors, and Paying for Diabetes’ Challenge and Power factors all indicate that patients have the more positive perceptions. The exception to this agreement was Diabetes Complications; for this concept patients had the more negative perceptions for Diabetes Complications’ Challenge and Lifestyle Autonomy factors. The factor scores also indicated significant differences for several additional concepts (Blood Sugar Testing, Controlling Your Blood Sugar, Exercise and Diabetes, and Help with Diabetes from Friends).
Previous studies have largely used qualitative methods to examine patient-provider differences in diabetes experiences and perspectives (e.g., 15 and 17). While qualitative approaches such as in-depth interviews with patients and providers offer rich descriptions of diabetes perceptions, these approaches do not provide a standardized method for interpreting the magnitude of similarity or difference. A study by Aufseesser et al (16) examined patient understanding of diabetes terms, but limited the topic area to retinopathy rather than the wide range of concepts addressed by the DSDS. Wikblad, Wibell and Montin’s work (21) and our previous study (20) surveyed independent groups of patients and health professionals for their diabetes perception. In contrast, this study used a quantitative survey method to examined patient-provider pairs and measures actual perceptions differences in patient-provider relationships.
The results support the DSDS as an instrument that provides a valid method to reliably assess perceptions of key diabetes concepts of both patients and providers. Further, the DSDS not only allows the researcher and the clinician to determine the perception differences between patient and provider, but also to estimate the magnitude of these differences.
The second research question concerned the comparability of two scoring methods. In other words, what method should be used to measure diabetes perceptions? The technical aspect of this question is being addressed in another study (25), however, it is clear that the factor scores have a greater sensitivity, i.e., appear to be more precise, to the nuances of patient-provider perceptions differences. Mean scores indicate differences, but provide only broad areas were differences exist. Another difference between the scoring systems is ease of use. Computing the mean of a specific DSDS is a relatively simple computation even with 9 adjective pairs. Factor analysis on the other had requires a certain degree of statistical sophistication in both execution and interpretation.
4.2. Conclusion
This study suggests that the patient’s perception of diabetes differs from the provider’s perception. For the patient, diabetes is experienced and given meaning within a personal sociocultural world. Diabetes concepts and terms are assigned attributes that are relevant to each individual. For the provider, diabetes concepts and terms most likely have a more direct and precise medical meaning. While the emergence of perception differences is not surprising, the number of differences was greater than anticipated, particularly since the differences are between actual patient-provider pairs. As in the pilot study (20), patients and providers differed on concepts that were not anticipated and did not differ on concepts where differences were expected (e.g., Exercise and Diabetes).
When possible the preferred method of scoring the DSDS is the factor method. This is also the method that should be used in research studies where statistical expertise is more likely to be available. However, in a clinical environment this refinement may not be necessary or possible. The administration of the DSDS and the mean scoring may also be of use for assessment of patient and provider attitudes in practices. The mean scores would be able to indicate general areas where perception difference exist and should be considered in interactions with the patient.
The consistent results found by both the mean and factor scoring methods underscore the strength of these findings (as well as has robust the DSDS is). Further, the DCP attitude subscales correlated with both the mean and the factor scores providing support for the validity (construct) of the instrument. Future research in the area of patient-provider perception differences should explore the relationships among perception differences, patient satisfaction, and self-care behavior. It would also be beneficial if these factors were examined for their impact on patient complication expectations and blood sugar levels.
4.3. Practice Implications
The findings also have practice applications for providers and diabetes educators. During patient and provider discussions, participants perceive diabetes concepts differently. The relationships among patient-provider perception differences are complex and do not appear to be intuitive. Providers and educators should avoid making assumptions about a patient’s understanding of diabetes or that they correctly understand instructions and advice. The DSDS can assist in determining the areas where differences exist.
Acknowledgments
This study was supported by Grant R01 DK061694 from the National Institute of Diabetes, Digestive and Kidney Diseases, National Institutes of Health.
Footnotes
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.
References
- 1.Mensing C, Boucher J, Cypress M, Weinger K, Mulcahy K, Barta P, Hosey G, Kopher W, Lasichak A, Lamb B, Mangan M, Norman J, Tanja J, Yauk L, Wisdom K, Adams C. National standards for diabetes self-management education. Diabetes Care. 2000;23:682–689. doi: 10.2337/diacare.23.5.682. [DOI] [PubMed] [Google Scholar]
- 2.Clark NM, Nothwehr R, Gong M, Evans D, Maiman LA, Hurwitz ME, Roloff D, Mellins RB. Physician-patient partnership in managing chronic illness. Academic Medicine. 1995;70:957–90. doi: 10.1097/00001888-199511000-00008. [DOI] [PubMed] [Google Scholar]
- 3.Von Korff M, Gruman J, Schaefer J, Curry SJ, Wagner EH. Collaborative management of chronic illness. Annals of Internal Medicine. 1997;127:1097–102. doi: 10.7326/0003-4819-127-12-199712150-00008. [DOI] [PubMed] [Google Scholar]
- 4.Ong LML, De Haes JC, Hoos AM, Lammes FB. Doctor-patient communication: A review of the literature. Social Science & Medicine. 1995;40:903–18. doi: 10.1016/0277-9536(94)00155-m. [DOI] [PubMed] [Google Scholar]
- 5.Roter DL, Hall JA, Katz NR. Relations between physicians’ behavior, and analogue patients’ satisfaction, recall and impressions. Medical Care. 1987;25:437–51. doi: 10.1097/00005650-198705000-00007. [DOI] [PubMed] [Google Scholar]
- 6.Sherman SE, Putnam SM, Lipkin M, Jr, Lazare A, Stoeckle J, Kekker V, Carroll JG. Annotated Bibliography of Doctor-Patient Communication. West haven, CT: Miles Institute For Health Communication, Inc; 1992. [Google Scholar]
- 7.Carter WB, Inui TS, Kukull W, Haigh V. Outcome-based doctor-patient interaction analysis: II. Identifying effective provider and patient behavior. Medical Care. 1982;20:550–66. doi: 10.1097/00005650-198206000-00002. [DOI] [PubMed] [Google Scholar]
- 8.Rost C. Something in common. Geriatric Nursing. 1989;10:198. doi: 10.1016/s0197-4572(89)80202-2. [DOI] [PubMed] [Google Scholar]
- 9.DeMatteo MR, Reiter RC, Gambone JC. Enhancing medication adherence through communication and informed collaborative choice. Health Communication. 1994;6:253–65. [Google Scholar]
- 10.Speedling EJ, Rose DN. Building an effective doctor-patient relationship from patient satisfaction to patient participation. Social Science & Medicine. 1985;21:115–20. doi: 10.1016/0277-9536(85)90079-6. [DOI] [PubMed] [Google Scholar]
- 11.Fook L, Morgan R, Sharma P, Adekoke A, Turnbull CJ. The impact of hearing on communication. Postgraduate Medicine Journal. 2000;76:92–95. doi: 10.1136/pmj.76.892.92. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kleinman A. Patients and Healers in the Context of Culture. Berkeley, CA: University of California Press; 1980. [Google Scholar]
- 13.Kleinman A, Eisenberg L, Good B. Culture, illness, and care: clinical lessons from anthropologic and cross-cultural research. Annals of Internal Medicine. 1978;88:251–8. doi: 10.7326/0003-4819-88-2-251. [DOI] [PubMed] [Google Scholar]
- 14.Lau RR. Cognitive representations of health and illness. In: Gochman DS, editor. Handbook of Health Behavior Research I: Personal and Social Determinants. New York: Plenum Press; 1997. pp. 51–69. [Google Scholar]
- 15.Loewe R, Freeman J. Interpreting diabetes mellitus: Differences between patient and provider models of disease and their implications for clinical practice. Culture, Medicine and Psychiatry. 2000;24:379–401. doi: 10.1023/a:1005611207687. [DOI] [PubMed] [Google Scholar]
- 16.Aufseesser M, Lacroix A, Binyet S, Assal JP. Diabetic retinopathy. Interpretation of medical terms by patients Journal Francais d’. Ophtalmologie. 1995;18:27–32. [PubMed] [Google Scholar]
- 17.Hunt LM, Arar NH, Larme AC. Contrasting patient and practitioner perspectives in type 2 diabetes management. Western Journal of Nursing Research. 1998;20:656–82. doi: 10.1177/019394599802000602. [DOI] [PubMed] [Google Scholar]
- 18.Osgood CE, Suci GJ, Tannenbaum PH. The Measurement of Meaning. Urbana, IL: University of Illinois Press; 1957. [Google Scholar]
- 19.Isaac S, Michael WB. Handbook in Research and Evaluation. 2. San Diego, CA: Edits Publishers; 1997. [Google Scholar]
- 20.Fitzgerald JT, Gruppen LD, Wray LA, Shay BR, Anderson RM. Measuring perceptions of diabetes-related concepts: a preliminary study. The Diabetes Educator. 2003;29:480–87. doi: 10.1177/014572170302900312. [DOI] [PubMed] [Google Scholar]
- 21.Wikblad KF, Wibell LB, Montin KR. The patient’s experience of diabetes and its treatment: Construction of an attitude scale by a semantic differential technique. Journal of Advanced Nursing. 1990;15:1083–91. doi: 10.1111/j.1365-2648.1990.tb01990.x. [DOI] [PubMed] [Google Scholar]
- 22.Miller DC. Handbook of Research Design and Social Measurement. Newbury Park, CA: Sage Publications; 1991. [Google Scholar]
- 23.Fitzgerald JT, Davis WK, Connell CM, Hess GE, Funnell MM, Hiss RG. Development and validation of the Diabetes Care Profile. Evaluation & the Health Professions. 1996;19:208–30. doi: 10.1177/016327879601900205. [DOI] [PubMed] [Google Scholar]
- 24.Cohen J. Statistical Power Analysis for the Behavioral Sciences. 2. Hillsdale, NJ: Erlbaum; 1988. [Google Scholar]
- 25.Stansfield RB, Fitzgerald JT. Semantic Differentials and Perception of Diabetes. Evaluation and the Health Professions. submitted. [Google Scholar]