These pages are best viewed with Netscape version 3.0 or higher or Internet Explorer version 3.0 or higher. When viewed with other browsers, some characters or attributes may not be rendered correctly.

Behavioral and Psychosocial
Measures for Diabetes Care:
What Is Important to Assess?

Russell E. Glasgow, PhD

So Many Measures, So Little Time
It is increasingly recognized that behavioral and psychosocial issues are critical to good diabetes care.^1-3 Such factors are important to measure for several reasons: because they often determine self-management behaviors; because they help us better understand and work with patients to tailor interventions; and because they are important outcomes in their own right.⁴ For example, Davis and associates⁵ found that behavioral outcomes were stronger predictors of mortality among diabetes patients than a host of physiological and metabolic control measures.

There are hundreds of psychosocial measures available, many of which assess highly related but slightly different factors. It is difficult for busy practitioners to decide which of these are best to use, especially given the time and fiscal constraints in most health-care settings today. This article will provide a framework for determining the important psychosocial and behavioral issues and will offer recommendations from among current procedures. More importantly, though, it will provide guidelines for how to select measures. This is important because available measures, their costs, and the data supporting them are constantly changing.

This section discusses issues to consider in measure selection. Following this, a framework is presented for thinking about key factors that influence diabetes management and adjustment. First, however, an important note on terminology. The term measure is used as shorthand for assessment procedure. In many cases, the most relevant assessment action that one can take is to conduct an interview, have patients demonstrate a skill, or participate in an assessment task rather than complete a questionnaire, which is what we most commonly think of when hearing the word measure.

Four Questions to Ask
As illustrated in Table 1, there are four key questions to ask about any psychosocial assessment. The most basic question concerns the utility of a measure for your specific purpose. Most research articles devote a great deal of attention to the reliability of a measure, which is one element of utility. Reliability refers to consistency of measurement results: for example, will a measure produce a similar result if administered a second time (test-retest reliability), will two raters produce similar results on a rating scale (interrater reliability), or do different questions on a scale assess the same quality (internal consistency reliability)?

Reliability is one consideration, but it is often less important than the validity of a measure. That is, does the procedure measure what it is supposed to measure? Like reliability, there are several different types of validity,^6-7 but the most common uses in diabetes care are predictive validity and sensitivity to change. Predictive validity asks how well the results of this assessment predict what will happen regarding X at some point in the future? The longitudinal, prospective time perspective is important and often not well addressed in many assessment studies because it is difficult and expensive to conduct longitudinal research that follows participants over time.

When selecting a measure to help evaluate a treatment or intervention, the issue of sensitivity to change becomes paramount. Some measures are predictive and very reliable but do not reflect changes resulting from interventions, so it is important to ascertain if an instrument will work for your specific purpose. Similarly, a procedure may be reliable and valid for one population, such as older adults with type II diabetes, but not so for another, such as adolescent patients with type I diabetes. This issue, as well as whether instruments have been demonstrated to be culturally sensitive (the majority have not) and the reading level of the measure,⁸ should be considered when selecting procedures for the population and setting in which you work.

Practicality is a dimension on which a large number of psychosocial measures fail and is a large part of the reason that they are not used more systematically in clinical practice. As shown in Table 1, three issues combine to determine the practicality of a measure for a given setting. The first is response burden on the part of participants and includes time, attention span, and amount of effort. If informed of the rationale for the measures ahead of time, patients are often willing to complete questionnaires before a clinic visit. They are often less consistent about completing complex monitoring records such as food diaries.

The second issue related to practicality concerns the costs of assessment. Costs include the purchase price of the instrument, as well as the amount of time or money required to score and provide feedback on assessment results. The final practicality issue concerns how well an assessment procedure fits with your clinic philosophy and patient flow. For example, a measure of patient empowerment⁹ makes the most sense if one of the goals of your clinic is to increase patient involvement and participation.

The third question to consider is whether your needs will be best served by a diabetes-specific or a general measure. When general versus diabetes-specific measures have been compared, diabetes-specific assessments usually appear to be stronger measures.^10-12 Clare Bradley¹¹ has published a book of diabetes-specific instruments that are available without charge and have validational data to support them. The one drawback to assessing only diabetes-specific issues is that you may not get a picture of the whole patient. For example, many older patients also have multiple other chronic diseases in addition to diabetes, and it may be important to know about the impact of these factors in planning interventions.

The final question to consider is what modality of presentation will be optimal. There are an increasing number of options, including computer-administered scales, phone surveys, interviews, role-play assessments, skill demonstrations, and questionnaires. The more interactive formats (e.g., computer and interview) are often more interesting for patients and contain built-in procedures to check for errors, but the initial costs are also often more.

Conceptualizing Psychosocial Influences: Multi-Level Thinking
It is helpful to think about the various levels of psychosocial factors that can influence diabetes management and adjustment, as summarized in Table 2. As can be seen, there are six levels of psychosocial influences ranging from highly personal to social group and, finally, to societal and cultural influences. All of these factors combine to form the psychosocial context within which people respond to their diabetes. At each of these six levels, there are both supports that can be used to facilitate coping and barriers or challenges that must be overcome.¹³ A comprehensive assessment will assess influences at each of these levels.

Of course, it is often not practical to conduct such a comprehensive assessment. The question then becomes, which level(s) of influences are most important given a patient’s goals, challenges, and living situation? It is an error to always assess factors at the personal level to the exclusion of other levels, but this is what often happens.

The remainder of this section discusses key issues and questions at each of the six levels described in Table 2 and briefly comments on currently recommended measures at each level. Please consider that these recommendations will likely need to be updated by the time this article is published. Rather than relying on this or any other source to determine which measures are best for your situation, you should ask the series of questions discussed in the previous section to determine whether a given measure is appropriate for your population and purpose.

At the personal level, hundreds of scales have been developed to measure an amazing array of patient beliefs, attitudes, knowledge, characteristics, and behaviors. Given the literature on what most strongly predicts behavior change, however,^14-15 I feel it is most important to measure four things. These are the specific behavior that patients desire to change; their self-efficacy that they will be able to successfully accomplish this behavior; their coping or problem-solving skills; and their beliefs about diabetes that are relevant to this behavior (e.g., if dietary change is the goal, what are the patients’ beliefs about the relationship between diet and desirable outcomes, and how effective do they believe that dietary change will be?).

In addition, one of the most frequently overlooked personal factors is a patient’s level of depression or distress. Depression is prevalent among people with diabetes and other chronic diseases¹⁶ and is related to both levels of self-management and clinical outcomes. It can be measured by either standard depression scales or by diabetes-specific distress measures such as the Problem Areas in Diabetes survey.¹⁷

Examples of the above measures that meet at least most of the criteria outlined in the previous section are listed in the third column of Table 2. Notations after a measure indicate whether it is applicable for type II diabetes patients (II), type I diabetes patients (I) or both (B); whether the instrument is a self-report questionnaire (Q), an interview/phone measure (IP), or some other type of assessment (O); and if it is diabetes-specific (D) or general (G). In cases where there were several measures from which to choose, I listed instruments that, other things being equal, were available without cost, were brief, and were relatively widely used so there are normative data with which to compare results.

At the family or significant other level, it is important to measure both things that significant others do to provide support and things they do that interfere with coping with one’s diabetes. Key issues concern the balance of responsibilities between patients and family members and what is appropriate for children of different developmental ages.^18-19

At the health-care system level, the key issues to consider are patients’ satisfaction with their health care; patient-provider interaction with various team members; what collaboratively set goals patients and the health-care team are working to achieve; and how many recommended preventive services patients have received. Health-care system measures are one of the most active, frequently changing, and confusing assessment areas. It is hoped that in the near future there will be commonly accepted guidelines for preventive services and standard measures of patient satisfaction. The best current measures of patient-provider interaction require audio or videotaping and scoring by trained raters and thus are not practical for most clinical settings.

Until such practical measures are available, one may want to ask interview questions about whether patients understand recommendations from their providers and whether they are receiving consistent messages from all team members. Finally, it is strongly recommended that patients’ current behavioral goals be documented in their medical records and that patients be provided a copy of those goals to have at home.

Far less attention has been paid to assessment of factors in the lower half of Table 2, and thus it is more difficult to recommend specific measures. Most people under the age of 65 years spend a significant portion of their time either at work or at school, yet there has been little attention to key features of these settings that facilitate or impede diabetes care.

Little diabetes-specific work has been conducted on either the identification or modification of key worksite or school influences. Issues that seem relevant from other areas of worksite health promotion research include policies, resources to support healthy lifestyles, supervisor (or teacher) support, co-worker (or peer) support, and as Padgett and associates²⁰ have recently pointed out, flexibility of schedules. At present, the most pragmatic assessment procedure may be to begin with some open-ended questions about how work or school affects a patient’s diabetes and then to follow up with relevant subscales from an instrument such as the Work Environment Scale²¹ that seem most relevant.

Given the emerging data on the consistently strong influence of socio-economic factors on health and health behaviors,²² it is ironic that procedures have not been developed to assess aspects of one’s neighborhood and community setting that support or interfere with diabetes management and control. Key factors that seem especially related to the lifestyle elements of diabetes self-management include availability of safe, free, or low-cost areas to exercise and availability, affordability, and labeling of low-fat food choices in grocery stores, fast-food outlets, and restaurants. Assessment of social support in the form of self-help groups for people with chronic disease, a strong local American Diabetes Association (ADA) chapter with many patient-education opportunities, and community programs such as recreation and senior citizen centers that offer activities and events for chronic-disease patients is also relevant. At present, the most relevant instruments are either questionnaire measures of the physical environment²³ or observational measures²⁴ of community resources.

Finally, it is important to keep in mind the shared influences that determine the overall health patterns of people in a given geographic region, state, or nation. Psychological assessment has tended to focus on differences among individuals in beliefs, behaviors, and environments. In terms of opportunities for making changes that result in the greatest good for the greatest number of people with diabetes, however, we also need to keep in mind societal and cultural policies and practices that have substantial, often unrecognized influences.²⁵ Examples include whether health insurance programs such as Medicare and Medicaid reimburse expenses for diabetes education and self-management supplies; regulatory issues, such as whether tax structures encourage or discourage tobacco and alcohol consumption; and the usual level of physical activity within a given cultural group. Also relevant is the way local and national media portray diabetes and lifestyle issues, which has a tremendous influence on people’s self-perceptions and behavior in this information-based society.

Putting It All Together
The previous section discussed assessment of the various levels of factors that influence a person’s adjustment to diabetes. It is also helpful to understand how all these factors combine to influence an individual’s quality of life and overall adjustment. A strong case can be made that, historically, health-care providers have not considered adequately the impact of medical interventions on patients’ quality of life.²⁶ Fortunately, this situation is changing, as evidenced by the goal of the ADA "to make an everyday difference in the quality of life for all people with diabetes."²⁷

Both diabetes-specific and general quality-of-life scales are available. It has been said that the SF-36 scale from the Medical Outcomes Study²⁸ is now the most widely used instrument in the world. Use of this scale will allow one to compare results with a huge number of other people both with diabetes and with other chronic diseases. A case can also be made for understanding the specific impact that diabetes has on a patient’s life. For this purpose, the most widely used and best validated scale currently available is the Diabetes Quality of Life Scale developed for the Diabetes Control and Complications Trial.²⁹

A brief measure that assesses overall adjustment to diabetes or the extent to which diabetes has become integrated into a person’s life is the ATT19.¹² While not providing the various scales of the lengthier quality-of-life scales, the ATT19 is shorter (19 items) and may be useful as an overall indicator of adjustment. Finally, the Diabetes Care Profile30 is one of the most carefully constructed and validated diabetes-specific scales. It comprehensively assesses factors at several of the levels shown in Table 2. For settings in which it is not possible to administer the entire DCP, use of the subscale(s) most relevant for your purpose may be possible.

Interested readers may want to explore a number of issues in greater depth than is possible here. A recommended reading list is provided at the conclusion of this article for those wishing to explore measurement evaluation issues or acquire more information about particular scales.

A Look into the Future
As a result of both internal and external forces, behavioral measurement procedures are undergoing rapid change. It is not possible to predict what psychosocial assessment for diabetes will look like in 25 years, but below are discussed some likely trends to look for over the next decade. The information revolution will likely result in at least two promising developments.

First, with the availability of resources such as Bradley’s¹¹ compendium of psychological diabetes measures and the Internet, we will hopefully see more standardization of measures and a greater percentage of researchers and practitioners using standard, validated measures. This will allow easier comparison across settings, programs, and populations, and is greatly preferable to the pattern that we have seen in the past of each person making up his or her own measure.

Second, computer and other technologies make it much more feasible to administer, score, and provide immediate feedback on psychosocial measures than was previously possible. In the near future, patients may be able to complete questionnaires over the Internet, fax in scannable surveys, or participate in interactive touchscreen computer assessments while in the waiting room and have the results available for all members of the health-care team at that visit.

With the increasing recognition of the importance of behavioral and quality-of-life issues in diabetes,^1,3 it is hoped that more behavioral endpoints will be added to HEDIS (Health Plan Employer Data and Information Set)³¹ and other report-card and guideline systems. Incorporation of patient-focused behavioral measures in such systems would do more than any other single activity to improve the psychosocial quality of care. This is something for which diabetes health professionals should advocate.

We are also starting to witness more practical measures becoming available as collaboration increases between researchers and practitioners, as well as a greater focus on population-based application of knowledge.³² Previously, measure developers focused primarily on lengthy multidimensional scales that were impractical to administer or score for many settings. There will likely be a trend toward more practical, user-friendly instruments for specific purposes, especially the measurement of change.

The Bottom Line
This article has covered a number of areas and measurement issues. Lest the reader be left thinking "this is interesting, but how can I possibly incorporate all these issues into my already over-booked patient encounters?" I conclude with the suggestions in Table 3, which provides recommendations for settings that have extremely limited assessment time or resources. It is probably most important to measure what both the patient and the health-care system are doing about diabetes. It may be most practical for a given period of time to collaboratively identify goals in one or two specific areas and to assess only these behaviors, rather than all self-management behaviors. With the increasing use of electronic patient records and preventive action guidelines, it is hoped that health-care system practices related to diabetes care will be automatically recorded and prompted by record-keeping systems.

To develop individualized plans with patients, it is also necessary to understand both the patients’ perspective and the social environment in which they are managing their diabetes. Brief measures of patients’ self-efficacy and health beliefs are available and can be restricted to the specific behavioral issues being addressed at that time. At present, the social environment may be best assessed by talking with patients about aspects of their home, work, and social life that may make it harder and easier to achieve their goals. Finally, to understand how a patient is doing overall and the impact a patient’s care plan is having, a measure of patient satisfaction and/or quality of life is recommended.

Remember that it is not necessary to measure everything at every visit. At a given visit, issues most relevant to goals on which the patient is focusing should be assessed. Over time, the record of measures collected can be compared to Table 2 to see if some key areas have been missed. To paraphrase Gordon Paul’s classic quote,³³ at each visit ask yourself "What measures, for this patient, for this particular goal, are most important at this time to inform our treatment planning and evaluation?"

Recommended Reading
and Resources

Anderson BJ, Rubin RR: Practical Psychology for Diabetes Clinicians: How to Deal with the Key Behavioral Issues Faced by Patients and Health Care Teams. Alexandria, VA, American Diabetes Association, Inc., 1996.

Bradley C, Ed.: Handbook of Psychology and Diabetes: A Guide to Psychological Measurement in Diabetes Research and Practice. Berkshire, England, Harwood Academic Publishers, 1994.

Glasgow RE, Osteen VL: Evaluating diabetes education: are we measuring the most important outcomes? Diabetes Care 15:1423-32, 1992.

Grady KE, Wallston BS: Research in health care settings. Applied Social Research Methods Series, Vol. 14. London, Sage Publications, 1988.

Lorig K, Stewart A, Ritter P, Gonzalez V, Laurent D, Lynch J: Outcome Measures for Health Education and Other Health Care Interventions. Thousand Oaks, Calif., Sage Publications, 1996.

Padgett DL, Nord WR, Heins JM, Arfken CL: Managing diabetes in the workplace: critical factors. Diabetes Spectrum 9:13-20, 1996.

Acknowledgments
This project was supported by grant #3DK-R01-35524 from the National Institute of Diabetes and Digestive and Kidney Diseases

References

¹Fisher EB, Jr., Arfken CL, Heins J, Houston C, Jeffe D, Sykes R: Acceptance of diabetes in adults. In Handbook of Health Behavior Research. Gochman DS, Ed. New York, Plenum Publishing Corp, 1996. In press

²Glasgow RE, Osteen VL: Evaluating diabetes education: are we measuring the most important outcomes? Diabetes Care 15:1423-32, 1992.

³Lorenz RA, Bubb J, Davis D, Jacobson A, Jannasch K, Kramer J, Lipps J, Schlundt D: Changing behavior: practical lessons from the Diabetes Control and Complications Trial. Diabetes Care 19:648-52, 1996.

⁴Kaplan RM: Behavior as the central outcome in health care. Am Psychol 45:1211-20, 1990.

⁵Davis WK, Hess GE, Hiss RG: Psychosocial correlates of survival in diabetes. Diabetes Care 11:538-45, 1988.

⁶Cook TD, Campbell DT: Quasi-Experimental Design and Analysis Issues for Field Settings. Chicago, Rand McNally, 1979.

⁷Kazdin AE: Research Design in Clinical Psychology. Boston, Allyn & Bacon, 1992.

⁸Wysocki T, Tarnowski KJ, Allen DM: Readability of diabetes self-report measures. Diabetes Care 12:752-53, 1989.

⁹Anderson RM, Funnell MM, Butler PM, Arnold MS, Fitzgerald JT, Feste CC: Patient empowerment: results of a randomized controlled trial. Diabetes Care 18:943-49, 1995.

¹⁰Glasgow RE, Toobert DJ: Social environment and regimen adherence among type II diabetic patients. Diabetes Care 11:377-86, 1988.

¹¹Bradley C, Ed.: Handbook of Psychology and Diabetes: A Guide to Psychological Measurement in Diabetes Research and Practice. Berkshire, England, Harwood Academic Publishers, 1994.

¹²Welch G, Dunn SM, Benney LJ: The ATT39: a measure of psychological adjustment to diabetes. In Handbook of Psychology and Diabetes. Bradley C, Ed. Berkshire, England, Harwood Academic Publishers, 1994, p. 223-45.

¹³Glasgow RE, Eakin EG: Issues in diabetes self-management. In The Handbook of Health Behavior Change. Schron E, Ockene J, Shumaker S, Eds. New York, Springer Publishing Company, 1997. In press

¹⁴Bandura A: Social Foundations of Thought and Action: A Social Cognitive Theory. Englewood Cliffs, NJ, Prentice Hall, 1986.

¹⁵Strecher VJ, DeVellis BM, Becker MH, Rosenstock IM: The role of self-efficacy in achieving health behavior change. Health Educ Q 13:73-91, 1986.

¹⁶Lustman PJ, Gavard JA: Psychosocial aspects of diabetes in adult populations. In Diabetes in America. Harris MI, Ed. Washington, DC, US Department of Health and Human Services, National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 1996, p. 507-17.

¹⁷Polonsky WH, Anderson BJ, Lohrer PA, Welch G, Jacobson AM, Aponte JE, Schwartz CE: Assessment of diabetes-related distress. Diabetes Care 18:754-60, 1995.

¹⁸Anderson BJ, Coyne JC: Miscarried helping in the families of children and adolescents with chronic diseases. In Advances in Child Health Psychology. Johnson JH, Johnson SB, Eds. Gainesville, FL, University of Florida Press, 1993, p. 167-77.

¹⁹Wysocki T, Taylor A, Hough BS, Linscheid TR, Yeates KO, Naglieri JA: Deviation from developmentally appropriate self-care autonomy: association with diabetes outcomes. Diabetes Care 19:119-25, 1996.

²⁰Padgett DL, Nord WR, Heins JM, Arfken CL: Managing diabetes in the workplace: critical factors. Diabetes Spectrum 9:13-20, 1996.

²¹Moos RH, Moos BS: Adaptation and the quality of life in work and family settings. J Comm Psychol 11:158-70, 1983.

²²Marmot MG, Bobak M, Smith GD: Explanations for social inequities in health. In Society and Health. Amick BC, Levine S, Tarlov AR, Walsh DC, Eds. New York, Oxford University Press, 1995.

²³Sallis JF, Owen N: Ecological models. In Health Behavior and Health Education, 2nd edition. Glanz K, Lewis FM, Rimer BK, Eds. San Francisco, Jossey-Bess. In press

²⁴Cheadle A, Wagner E, Koepsell T, Kristal A: Environmental indicators: a tool for evaluating community-based health promotion programs. Am J Prev Med 8:345-50, 1992.

²⁵Biglan A: Changing Cultural Practices: A Contextualist Framework for Intervention Research. Reno, NV, Context Press, 1995.

²⁶Robbins J: Reclaiming Our Health. Tiburon, Calif., H. J. Kramer, 1996.

²⁷American Diabetes Association: Mission statement. 1995.

²⁸Stewart AL, Hays RD, Ware JE: The MOS short-form general health survey: reliability and validity in patient population. Med Care 26:724-35, 1988.

²⁹Jacobson AM, de Groot M, Samson JA: The evaluation of two measures of quality of life in patients with type I and type II diabetes. Diabetes Care 17:267-74, 1994.

³⁰Fitzgerald JT, Davis WK, Connell CM, Hess GE, Funnell MM, Hiss RG: Development and validation of the diabetes care profile. Eval & Health Prof 19:208-30, 1996.

³¹Health Plan Employer Data and Information Set (HEDIS). Washington, DC, National Committee for Quality Assurance, 1993.

³²Wagner EH: Population-based management of diabetes care. Patient Educ Counsel 26:225-30, 1995.

³³Paul GL: Behavior modification research: design and tactics. In Behavior Therapy: Appraisal and Status. Franks CM, Ed. New York, McGraw-Hill, 1969, p. 29-62.

³⁴Johnson SB, Silverstein J, Rosenbloom A, Carter R, Cunningham W: Assessing daily management in childhood diabetes. Health Psychol 5:545-64, 1986.

³⁵Toobert DJ, Glasgow RE: Assessing diabetes self-management: the summary of diabetes self-care activities questionnaire. In Handbook of Psychology and Diabetes. Bradley C, Ed. Berkshire, England, Harwood Academic, 1994, p. 351-75.

³⁶Grossman HY, Brink SH: Self-efficacy in adolescent girls and boys with insulin-dependent diabetes mellitus. Diabetes Care 10:324-29, 1987.

³⁷Schafer LC, McCaul KD, Glasgow RE: Supportive and nonsupportive family behaviors: relationships to adherence and metabolic control in persons with type I diabetes. Diabetes Care 9:179-88, 1986.

³⁸Davies AR, Ware JE: GHAA’s Consumer Satisfaction Survey and User’s Manual. Washington, DC, Group Health Association of America Inc., Department of Research and Analysis, 1991.

³⁹Wilson AE, Home PD for the Diabetes Audit Working Group of the Research Unit of the Royal College of Physicians and the British Diabetic Association: A dataset to allow exchange of information for monitoring continuing diabetes care. Diabetic Med 10:378-90, 1993.

⁴⁰Glasgow RE: Social-environmental factors in diabetes: barriers to diabetes self-care. In Handbook of Psychology and Diabetes. Bradley C, Ed. Berkshire, England, Hardwood Academic, 1994, p. 335-49.

Russell E. Glasgow, PhD, is a research scientist and member of the Chronic Illness Research Group at Oregon Research Institute, in Eugene, Oreg.