Diabetes in the Nursing Home

This article reviews and synthesizes the literature pertaining to the identification and care of the nursing home (NH) resident with diabetes. NH residents with diabetes have heavy illness burden with complex medical management of their comorbid conditions including depression and exacerbations of their other chronic conditions. Future research is necessary to determine models of care in the NH, with nurse practitioners driving improvement in standards of care, and how this may improve outcomes such as cognition and functional ability. Future research is also indicated to determine how residents’ conditions and care change over time, as well as how one can assist in the transition of care to end-stage disease with a terminal decline to better serve these residents. (Annals of Long-Term Care: Clinical Care and Aging 2009;17[8]:42-46)
____________________________________________________________________________________________________________________________________________

Twenty percent of patients over age 65 have diabetes.¹ This number is expected to grow rapidly with the baby boomers’ advancing age. According to the American Diabetes Association (ADA),¹ there are no published long-term studies in individuals over age 65 to demonstrate the benefits of tight glycemic control, blood pressure (BP), and lipid control. High rates of premature death, functional disability, and comorbid conditions are more likely in older patients with diabetes than in those without. These persons are also at greater risk for polypharmacy, depression, cognitive impairment, urinary incontinence, injurious falls, and persistent pain.¹ Life expectancies in this population are highly variable. It is important that clinicians recognize the prevalence of diabetes in nursing homes (NHs) and to provide appropriate management.

Prevalence of Diabetes in Nursing Homes

The prevalence of diabetes in NHs, obtained predominantly via questionnaires, has been estimated from 8.8-26.7%.^2-7 Sinclair et al⁴ used a more scientific approach by administering an oral glucose tolerance test in addition to collecting fasting blood sugar (FBS) levels. Nearly 15% of the residents had diabetes and an additional 30.2% were diagnosed with impaired glucose tolerance.⁴ The prevalence rate using information from the Minimum Data Set (MDS) was 26.4%.⁷

Undiagnosed diabetes is a serious problem because evidence supports a link between diabetes and cognitive impairment.^8-12 Targeted screening of elderly residents with dementia may identify the highest rates of undiagnosed diabetes.¹³ Researchers have estimated that prevalence rates for undiagnosed diabetes in NHs ranged from 13% to 47.2%.^6,13,14 Hauner and colleagues⁶ concluded that undiagnosed diabetes prevalence was 47.2%; this higher prevalence rate may be related to the authors’ stricter definition of diabetes as hemoglobin A1C (HbA1C) value of greater than 6.0%.

Profile of the Older Adult with Diabetes

The composite of available data is from community-dwelling elderly individuals and is not specific to the elderly NH resident with diabetes. Based on this information, elderly persons with diabetes tend to have decreased functional ability^8,9 with increased fall rates as compared to elderly persons without diabetes. Increased hospitalizations and admissions are also seen in the diabetic population. Cognitive functional impairment and depression are strongly correlated with HbA1C values.^8-10 According to the ADA, glycosylated hemoglobin levels of 6% or lower are desirable and reduce target organ damage.¹ Research indicates otherwise for elderly persons with diabetes. In one study, HbA1C levels less than 6% were found to have a deteriorating effect on cognition.⁸ Hagemann et al⁸ concluded that participants with HbA1C levels from 6-8% had higher Mini-Mental State Examination (MMSE) scores than those with levels lower than 6% or higher than 8%. These findings seem to suggest that HbA1C levels of 7.5-8% may be acceptable for elderly individuals. Worrall and colleagues¹⁵ also concluded that slightly higher HbA1C levels resulted in increased performance, which contradicts the ADA’s recommendation for desirable levels.

Several studies reported that cognitive function impairment in persons with diabetes had a direct correlation with HbA1C values from greater than 7.5% to 11.5%.^8-10 Travis et al,⁷ using data from the MDS, found that 37% of persons with diabetes with elevated HbA1C levels had impaired decisional capacity. Maraldi et al¹⁰ found that more participants with HbA1C levels greater than 8.4% were cognitively impaired (15.3%) as compared to those with controlled diabetes (7.6% impaired) and participants without diabetes (8.9% impaired).

Perlmuter and colleagues¹¹ measured serial learning tasks, digit span tests, and reaction times. Cognitive differences in persons with diabetes versus those without were associated with HbA1C levels greater than 7.5%, and severity of peripheral neuropathy resulted in poorer performance.¹¹ Reaven and colleagues¹² reported HbA1C levels of 11.0% in their group with diabetes and FBS of 204.4 mg/dL. Abstract reasoning and verbal learning were related to HbA1C levels.¹² Memory was related to FBS level.¹² Essentially, the higher the HbA1C and the higher the FBS, the poorer the performance.

The presence of diabetes in older adults places them at risk for falls.^7,16 A prospective cohort study completed on long-term care (LTC) residents reported that diabetes is an independent risk factor for falls.¹⁶ While the MDS reported a 25% fall rate 30 days prior to admission,⁷ Maurer et al¹⁶ reported fall rates of 78% for persons with diabetes as compared to only 30% for those without. Fall rates after 1 year of follow-up were essentially unchanged at 73% for those with diabetes as compared to 31% for those without.¹⁶ The difference in fall reports may be due to self-reported data from the MDS versus incident reports and physical therapy evaluations collected. One flaw in the study was that there was no measurement of diabetic control to correlate the fall rates.¹⁶

In addition to fall risk, elderly persons with diabetes have higher hospitalization rates. Benbow and colleagues² reported that 43% of the patients with diabetes studied were admitted to the hospital as compared to 25% of the control group. Increased hospitalization rates were also supported by the Third National Health And Nutrition Examination Survey (NHANES III): 36.3% of age 65-74–year-old individuals were hospitalized as compared to 34.0% of the age 75-and-over age group.¹⁷ No indication of the reason for admission or length of stay was provided in either study.

One-third of all NH residents are classified with depression upon admission.⁷ No depression studies completed specifically on NH residents with diabetes were found in the literature. For community-dwelling elderly persons, however, depression rates for elderly persons with diabetes ranged from 18.7% to 66%.^9,10,18 Reasons for these variations may be related to study design, multiple depression screening tools, and varied classifications of depression.

Ciechanowski et al⁹ examined relationships between depression and diabetes in older adults. Approximately 34% of the participants were classified in the highly depressed symptom severity category and 33% in the medium depressed symptom severity category. Grouping both medium and high symptom severity together resulted in an approximately 66% prevalence rate. HbA1C levels greater than 7.9% were correlated with increased depression in the highly depressed symptom severity group. Maraldi et al¹⁰ also found that elderly persons with diabetes with a mean HbA1C of 8.4% were at higher risk for depression. Recurrence of depression was higher in those with uncontrolled diabetes with HbA1C levels greater than 8.1%.¹⁰

Management of Diabetes

Standards of Medical Care in Diabetes is published annually. The 2009 recommendations for diabetes care recommend assessment of glycemic control through self blood glucose monitoring and HbA1C values. The prevention of diabetic complications recommends the following: BP control with BP checks at each routine visit, systolic BP less than 130 mm Hg, and diastolic BP less than 80 mm Hg; lipid screening at least annually; antiplatelet therapy; cardiovascular screening to evaluate 10-year cardiovascular risk, then assess risk factors annually; nephrology screening with annual microalbumin and serum creatinine; retinopathy screening annually or every 2-3 years if there has been one or more normal exams; and neuropathy screening annually.¹

The ADA recommends that elderly persons who are functional, cognitively intact, and with significant life expectancy follow the same goals as those for younger adults.¹ The glycemic goals for elderly individuals not meeting criteria should be individualized but caution against hyperglycemia symptoms and complications. When reviewing cardiovascular risk and glycemic control in persons with type 2 diabetes, the potential benefit of improved glycemic control has not been demonstrated in a randomized clinical trial to date.¹

Treatment of hypertension is indicated. Lipid and aspirin therapy may benefit elderly persons. Screenings for diabetic complications should be individualized for elderly individuals, but attention should be placed on complications that would lead to functional impairment.

Appropriate diabetes management in the LTC setting is influenced by the ambivalence about the benefits and risks of glycemic control, and the relative lack of scientific studies of patients with diabetes in LTC settings.¹⁹ Diabetes management may also be determined by other factors such as cognition and functional status. Researchers in one study examined the variability of physician practice based on cognitive and functional status of the residents.¹⁹ Three profiles were presented: (1) a functionally impaired, but cognitively intact elderly person; (2) a functionally intact, but cognitively impaired elderly person; and (3) a functionally and cognitively impaired elderly person. Ordering was more likely to occur in the profile 1 individuals and least likely to occur in the profile 3 individuals. Limitations to this study were that the rationales on survey responses were not studied, and members who designed the study also participated. Likewise, NH residents are less likely to receive appropriate diabetes management when compared to non−NH persons.

A descriptive study of elderly persons with diabetes examined factors associated with providers’ decisions to initiate treatment after diabetes was diagnosed in the hospital or to continue prior diabetic management post-discharge. Treatment rates for NH residents were 7.9% as compared to 18% for those not in a NH.²⁰ Treatment odds in multivariate analyses for the NH resident were 54% lower than those not residing in a NH. Relative odds of treatment after discharge were 67-70% lower in participants who resided in the NH prior to admission.²⁰

There is continued supporting evidence that management in NHs is suboptimal. Benbow et al² reported that 64% of the NH residents studied had no record of anyone being medically responsible for their diabetes care and management from the previous year. Taylor and Hendra³ reported that regular care visits by the general practitioner or practice nurse addressed diabetes care “always” in only seven NHs, “sometimes” in 12, and “never” in 12.

Glycemic Control
The American Medical Directors Association (AMDA)²¹ noted a lack of evidence regarding the optimal frequency and timing of self-monitoring blood glucose (SMBG) in the LTC setting. Comprehensive monitoring is inappropriate for every individual in LTC, but some recommendations based on control and insulin use exist (Table).

A descriptive study attempted to outline the frequency of capillary blood glucose (CBG) testing and the feasibility of allowing nursing staff to adjust testing frequency based on CBG values.²² Sixty-nine percent of the residents were ordered the CBG protocol testing four times per day.²² Twenty-two residents met their CBG goals and advanced to lower testing frequency. Fifteen residents did not meet CBG goals and continued testing four times per day. Fifteen residents advanced to lower testing frequencies but not according to protocol parameters. Seventeen residents were not advanced to lower testing frequency but should have been according to the protocol. Approximately 70% of the providers used the CBG protocol, and correctly in 54% of the cases.²²

Blood glucose testing rates vary. No monitoring for diabetic control was reported in 27% of the residents, with one-third in NHs in England.² Nicholls²³ reported a 97% rate for daily testing but included urine glucoses in the statistic. One study reviewed the resident’s ability to perform SMBG; findings showed that 11% of the NHs provided this opportunity, 39% provided self-management sometimes, and 50% never allowed it.³ We must question including a functional resident in SMBG and how this would impact control. Should we really take away all management upon admission to the NH?

HbA1C testing rates varied from 19-54%.^2,23 HbA1C measurement was performed in 54% of the NH residents as compared to 85% of the community-dwelling persons studied by Fahey et al.²⁴ The most significant difference in the rates may be related to nurse chart reviews versus self-reported testing. NHANES III demonstrated that HbA1C control could be achieved in older persons with diabetes.¹⁶ Of the 1482 participants, 30% were in the age 65-74 age group and 24% were in the age 75-and-older age group. The mean HbA1C was 7.71% and 7.27%, respectively. HbA1C less than 7.0% was reported in 38.3% and 55.2% of all age groups that showed control within the ADA guidelines. HbA1C levels above 8.0% in the two groups were 37.2% and 26.9% of the cases that demonstrated poor control.¹⁶ This study done on community-dwelling persons may impact how these results would compare to age-, sex-, and race-matched counterparts who are institutionalized.

Blood Pressure
There are no studies to date that reviewed control in NH residents. Benbow et al² reported that only 29% of the residents had their BP measured.

Lipids
There are no studies that reviewed control in NH residents.

Antiplatelet agents
There are no studies that reviewed antiplatelet compliance in NH residents with diabetes.

Nephropathy
Benbow et al2 reported that only 28% of NH residents had renal function assessed.

Retinopathy
Benbow et al² reported that 72% of the residents did not have dilated exams.

Neuropathy
Benbow et al² reported that only 18% of residents had foot exams.

Practice Implications

With the increase in the diagnosis of diabetes and the baby boomers getting older, we must improve care. We need to increase diabetes screening, as prevalence rates were widely variable and reflected undiagnosed diabetes. With a direct correlation between diabetes and cognition, one has to wonder if earlier diagnosis would assist elderly individuals in remaining functional in the community. Depression is also a problem. The higher the HbA1C level, the greater the risk for depression.

The clinical community needs to examine its current practices with NH residents. Failure to comply with recommended standards of care, which includes reviewing control through SMBG and laboratory testing, may contribute to long-term complications, falls, cognitive decline, and depression.

Decisions about medical treatment or care should not be determined on the basis of the cognitive status, functional status, or place of residence, as demonstrated by McNabney et al¹⁹ and Glynn et al.²⁰ Benbow and colleagues² and Taylor and Hendra3 demonstrated that there are residents who are not receiving help with managing diabetic care. Questions remain as to the optimal SMBG and HbA1C values for elderly NH residents. Clinicians may want to contribute to the development of new standards for these individuals with diabetes given the complexity of their care.