Diabetes Spectrum
Volume 13 Number 4, 2000, Page 234
Feature Article/Quality Improvement

A Quality Improvement Project in Diabetes Patient Education During Hospitalization

Ellen D. Davis, MS, RN, CDE


One institutional quality improvement model, the FADE process, offers a useful approach to assessing and improving diabetes inpatient education. FADE stands for focus, analyze, develop, and execute. First, we focused on the 17% of all inpatients on one day who had diabetes. Second, we analyzed the current level of diabetes education provided, using a quality improvement (QI) survey tool. The most interesting data revealed that we offered education less frequently to people over age 65, but that older patients who received education had higher scores than younger patients who received education. Third, we developed a diabetes teaching protocol for use by staff nurses with inpatients. Finally, 6 months after protocol implementation, we re-executed our QI survey tool. We found that test scores were slightly better than pre-protocol-implementation survey scores, but that fewer patients had received self-management education. The poor implementation of the protocol was investigated, and steps are being taken to improve diabetes care and education.

An institution's quality improvement (QI) model offers a useful approach to assessing and improving diabetes patient education. Diabetes patient education fits easily into QI processes; inclusion in such processes has many advantages, not the least of which is that education can then be viewed as an integral part of care for a person with diabetes as a primary or secondary diagnosis. In inpatient settings, diabetes patient education is a part of total care of the patient and is by its nature shared by all health care providers. These providers will feel a more committed investment in patient education if that process shares similarities with other QI programs.

Diabetes patient education may be appropriately done by generalist staff nurses or by designated specialized educators in the inpatient arena. While many authors have argued that staff nurses do not have sufficient knowledge to educate patients,19 the reality is that with the current shortened lengths of hospital stays and other cost-saving measures, patients may of necessity receive all of their self-care information from staff nurses. The outcome of patient interventions—whoever provides the care—is the important piece of information to be shared with surveying bodies. More informed decisions about who should provide care and education can be made using QI data.

In our large university medical center, the FADE process is used in QI.10 This is one of many processes described in the literature to facilitate QI. The steps of this process are focus, analyze, develop, and execute. To focus, a team articulates an identified problem. In the second step, analyze, the team collects baseline data, the pre-test part of the project. Develop means developing a plan to address the problem. Finally, to execute, the team puts the plan into action. Quality action teams are at work in many of our institutional areas to identify and solve problems. The diabetes patient education project described in this article used the FADE methodology of QI.

The goals of diabetes management are to reduce the personal tragedy and public health cost of diabetes and its complications and to enhance the quality of life for people with diabetes. People with diabetes need sufficient self-care knowledge to manage their diabetes effectively. They may lack this essential knowledge unless they receive education.

One place where people with diabetes receive care is in hospitals. They are admitted to hospitals more frequently than people without diabetes and for longer stays.11 Indeed it is estimated that 20–25% of all hospitalized patients have diabetes. However, diabetes is an underreported disease in hospitalization documentation. It is estimated that 40% of all people with diabetes who are hospitalized do not have diabetes listed on their hospitalization discharge record. A 1990 study found that 9% of all hospitalizations listed diabetes as a primary or secondary diagnosis.11 In our institution, in 1996 and 1997, we found that 17% of patients had diabetes.

While outpatient facilities remain the most appropriate setting for diabetes education, more that 50% of people with diabetes receive limited or no diabetes self-management education.12 Hospitalization offers an important opportunity to provide education on survival skills. During a stay at our tertiary care medical center, patients are often physiologically stressed and must begin a more complex diabetes self-care plan. For example, a patient who has been taking oral diabetes medication may require insulin injections for home management following coronary artery bypass graft surgery. The patient or family will have to be taught to do the more sophisticated home care.6 The patient must be taught in the hospital to perform self-care starting immediately after discharge. Additional outpatient education will be recommended and set up, whenever possible. But from a tertiary care setting, it is sometimes impossible to set up long distance outpatient education, and there may be no outpatient education program in the patient's home setting.

In order to plan for improved diabetes patient education at our institution, we first needed to examine the current level of education provided. To do this, we developed a QI tool (Figure 1). This QI tool was administered by staff nurses during their normal duties to all available inpatients with diabetes on one day. The tool was designed to ascertain survival-level self-management knowledge and also asked if patients had received education during their current hospitalization.

pg229fig1.jpg (80927 bytes)
Figure 1. A quality improvement tool to assess survival-level diabetes self-management knowledge

The QI survey tool was designed by the diabetes clinical nurse specialist (CNS) with input from other diabetes experts and staff nurses. It was tested for inter-rater reliability and validity.

This one-day hospital-wide QI initiative required a great deal of labor. When it became apparent that the project would require more time than was available from the CNS or any other regular health care provider, we sought funding for another person. We wrote a proposal for a grant for materials and salary for a staff assistant to the CNS, and a pharmaceutical company provided the funding. After gaining commitment from a school of nursing graduate student and former nurse manager for temporary work, we wrote a contract using a format sanctioned by the institution. The staff assistant performed specific tasks designated by the CNS.

pg230fig2.gif (19674 bytes)

Figure 2. Pre-protocol-implementation survey: patient scores by group

Analysis of the data from the QI tool demonstrated that out of the entire inpatient population on the survey day, 111 patients had bedside blood glucose monitoring ordered, and 54 of these patients were interviewed, for an inclusion rate of 49%. The average score on the knowledge test was 60.7%. Twelve patients (22.2%) had received education during the current hospitalization. The average score for patients who had received education was 68.3% (Figure 2).

Interestingly, the average score on the knowledge test for patients under age 65 was 63.6%, and the average score for patients over age 65 was 56.2%. Among patients under age 65, 24.2% received education, while among those over age 65, only 19.0% received education. The average score for patients under age 65 who received education was 60.0%, and the average score for patients over age 65 who received education was 85.0%. To summarize, we offered education less frequently to people over age 65, but older patients who received education had higher scores than younger patients who received education (Figure 2). Therefore, we see that older patients benefit greatly from education and should not be slighted because of "ageist" assumptions.

pg230fig3.gif (34100 bytes)
Figure 3. Pre-protocol-implementation survey: patient scores of educational questions.

More than 85% of the respondents gave an adequate definition of diabetes, and more than 90% relayed that food is a primary element of control. (See Figure 3.) Only about 50% could state the short-term rationale for good blood glucose control, while more than 80% could state the long-term rationale for control. When asked what things people can do to take care of diabetes, only 46% answered "exercise," but almost 80% said "take medications." Medical care and foot care received low scores as elements of care.

Perhaps the most interesting finding was that 57.7% of the patients interviewed performed self-monitoring of blood glucose at home. There are no current reliable data on the national average of people with diabetes who self-monitor. While the percentage we found seems high, pharmaceutical industry spokespeople suggest that it accurately reflects the percentage of people who self-monitor. Certainly, this figure supports the important assumption that patients will engage in positive self-management practices if given professional guidance.

To meet the educational needs of patients with diabetes, we developed a diabetes teaching protocol for staff nurses. A diabetes patient education interest group of staff nurses and the CNS had met earlier on a quarterly basis. This group was reconvened and asked to prepare the teaching protocol. The protocol workgroup met six times and also conducted business through internal mailings. After the group developed a draft, expert reviewers examined it and made suggestions for revisions. The protocol was then sent to the Clinical Practice Council where it went through the process necessary to become an approved protocol.

We needed the involvement of more than 2,000 staff nurses (all of the full-time and part-time nurses employed by the hospital) to administer the pre-test (pre-teaching protocol implementation) QI survey and then to use the new protocol to improve diabetes patient education in our institution. Buy-in by key groups was thus essential. Therefore, the next steps involved developing marketing strategies for the protocol and its attendant QI survey. Key groups selected, in addition to the Clinical Practice Council, included the unit QI representatives and nurse managers. Posters were put up on each unit, and nurse managers advertised the initiative through staff meetings and unit communication books.

Because diabetes is almost always a secondary diagnosis in acute and tertiary care settings, there is not a constituent base of staff nurses dedicated to diabetes care, as was once the case. Previously at our institution, patients were admitted to an endocrine unit for prolonged stays for diabetes control and education. This unit no longer exists, and diabetes patients are found everywhere in the hospital. However, at times, one can find a cluster of patients with diabetes on the general medical units (primary diagnosis, in some cases) or on the cardiology unit, for example.

Six months after protocol implementation, we again used the QI survey tool (Figure 1) to measure patient knowledge of diabetes self-management and level of patient education during the current hospitalization. Fifty patients were interviewed on one day for an inclusion rate of 50%. We decided that staff nurses were too stressed to administer the tool. Therefore, the contract-hired staff assistant interviewed patients.

Results of the post-protocol-implementation QI survey showed that the average knowledge score was 62.5%, and only 2 patients (4%) had received education during this hospitalization. While the self-care knowledge test scores were slightly better than the pre-protocol-implementation survey scores, fewer patients had received self-management education. Thus, we can say that the protocol was not truly implemented.

Many reasons can be given for the poor use of the protocol. The diabetes teaching protocol was the first teaching protocol approved and disseminated by the Clinical Practice Council, and support mechanisms were not fully in place. Further, during this time staff went through a major change in documentation systems and were operating in "overload mode." Staff turnover also played a part. Another reason for poor protocol use may be that since diabetes occurs everywhere, usually as a secondary diagnosis, there is no defined team to champion this initiative.

One of the benefits of this project was illumination of a part of our health care system that needs attention. This came at a time when the costs of caring for people with diabetes were also receiving new attention. As new initiatives were being designed, leaders had data in front of them about the current status of diabetes patient education in the hospital. New initiatives were designed starting at that point.

The major contribution of the project was to realistically define the current state of diabetes patient education. Steps are now being taken to improve diabetes care and patient education at our institution. An insulin therapy management protocol has been approved by the Clinical Practice Council. Two insulin order sets were placed into the system, one for subcutaneous insulin and the other for intravenous insulin. Also, a new blood glucose/insulin flowsheet was initiated in all areas where computerized documentation was not done. Finally, a new patient and family education record to accompany the diabetes teaching protocol has been implemented.

This study was labor-intensive, but it was very revealing. The institution has benefited from its results.

1Scheiderich S, Freibaum C, Peterson L: Registered nurses' knowledge about diabetes mellitus. Diabetes Care 6:57–61, 1983

2Etzwiler D: Who's teaching the diabetic? Diabetes 16:111–117, 1967

3Leichter SB, Ferguson SK, Collins P, Rhodes A, Garrity T, Hernandez C: Survey of knowledge among primary health care workers in diabetes. South Med J 73:1243–1246, 1980

4Essig MJ, Thielen PL: A study of diabetes educators in Ohio hospitals. Diabetes Educ 8:33–35, 1982

5Leichter SB: Diabetes patient education in hospital settings. Diabetes Educ 12:277–280, 1986

6Davis ED, Midgett L, Gourley CS: Teach less, teach better at every opportunity. Diabetes Educ 20:236–240, 1994

7Davis ED: Role of the diabetes nurse educator in improving patient education. Diabetes Educ 16:36–38, 1990

8Pretto Z, Koproski J, Poretsky L: Hospital management of diabetes in the post-DCCT era. Pract Diabetol 14:27, 1995

9Wollenberg P: A redesigned inpatient diabetes education program. Pract Diabetol 18:3339, 1999

10Joint Commission on Accreditation of Healthcare Organizations: An Introduction to Quality Improvement in Health Care. Oakbrook, Ill., Joint Commission on Accreditation of Healthcare Organizations, 1991, p. 56–58

11National Diabetes Data Group: Diabetes in America. 2nd ed. (NIH publication 95-1468.) Bethesda, Md., National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases, 1995

12Clement S: Diabetes self-management education. Diabetes Care 18:1204–1214, 1995

Portions of this work were funded by a grant from Lifescan, Inc., a Johnson & Johnson Company.

Ellen D. Davis, MS, RN, CDE, is a diabetes clinical nurse specialist at Duke University Health System in Durham, N.C.

Address correspondence and reprint requests to: Ellen D. Davis, MS, RN, CDE, Box 3677, Duke University Health Systems, Durham, NC 27710.

Return to Issue Contents

Copyright 2000 American Diabetes Association

Last updated: 12/00
For Technical Issues contact