Association of Deviation From Glycated Hemoglobin (HbA1c) Target in the Guidelines for the Treatment of Diabetes in Older Patients With Place of Care and Interval Between HbA1c Tests Among Older Patients Receiving Home Medical Care.

Introduction

In Japan, the number of older patients with diabetes is increasing with its super‐aged society [ 1 ]. These patients are more likely to have fewer abilities in terms of activities of daily living (ADL), cognitive impairment, and multiple comorbidities, leading to difficulties in attending regular clinic visits. Today, more older patients choose to receive home medical care. In this research, we define home medical care as regular treatment performed biweekly by a physician at home or at a nursing facility.

The Guidelines for the Treatment of Diabetes in Older Patients in Japan 2016 (hereinafter, the guidelines), proposed by the Japan Geriatric Society and the Japan Diabetes Society in 2016, recommend that the glycated hemoglobin (HbA1c) target for older patients be set in consideration of each individual's background, including age, cognitive function, ADL, comorbidities, risk of serious hypoglycemia, and estimated life expectancy, rather than a uniform standard [ 2 , 3 ]. Many older patients with diabetes who receive home medical care have cognitive impairment, limited ADL, and multiple comorbidities; thus, both their medical and living situations should be considered. The goal of treating these patients should be to avoid deviation from the HbA1c target recommended in the above guidelines [ 4 , 5 ].

The choice of medical treatment for older patients with diabetes depends on the place of care (home or nursing facility) [ 6 ]. Patients treated at home who have cognitive impairment or ADL restrictions have limited self‐management regarding diet and exercise therapy. Inadequate medical adherence is also expected with comorbidities and consequent polypharmacy, even with multi‐occupational support by visiting staff such as nurses, home care workers, and pharmacists [ 7 , 8 ]. However, patients treated in nursing facilities are more likely to receive appropriate support by nursing staff, which leads to stable lifestyle habits and medical management.

This background led us to hypothesize that glycemic control in older patients with diabetes receiving home medical care, deviation from HbA1c targets in the Guidelines for the Treatment of Diabetes in Older Patients is associated with the place of care (home or nursing facility) and the interval between HbA1c tests. Clarifying these associations will help home medical care teams to improve the quality of care for older patients with diabetes.

The aim of this research was to investigate factors associated with deviation of the HbA1c value from the HbA1c target proposed in the Guidelines for the Treatment of Diabetes in Older Patients, focusing on the place of care (home or nursing facility) and interval between HbA1c tests.

Methods

Participants

Study participants comprised older patients with diabetes over age 65 years who received home medical care by physicians from four clinics in Tokyo and Kanagawa Prefecture in Japan, between November 14, 2016 and November 13, 2021 and who met all of the following conditions during the study period: diagnosed with diabetes, receiving home medical care from the same clinic for > 2 months, having ≥ 2 HbA1c tests, and remaining in the same place of care.

Model of Deviation From the Guidelines for the Treatment of Diabetes in the Older Patients

We created a model of deviation from the above Japanese guidelines based on previous research and agreement among three experienced researchers who had treated older patients with diabetes for more than 10 years. This model included age and place of care (home or nursing facility) as well as the presence of terminal‐stage cancer as a background disease, number of oral medications, and interval between HbA1c tests as factors that can be intervened by a visiting physician. Financial dependence on public assistance was also included as a social factor.

Data Collection

We collected individual data on the use of diabetes medication (insulin, insulin secretagogues, other oral antihyperglycemic medications, or no diabetes medication), the number of total oral medications including non‐diabetes medications, the place of care (home or nursing facility), bedridden level (independent, J, A, B, or C), rating of dementia (independent, I, II, III, IV, or M) [ 9 ], Charlson Comorbidity Index score (CCI) [ 10 ], presence of terminal‐stage cancer, and dependence on public assistance. We also calculated the average interval between HbA1c test by dividing the days between the first and last visit by the number of HbA1c tests during that period.

HbA1c Target

We classified participants into three categories of HbA1c target, according to the guidelines. The HbA1c targets are categorized into three grades according to cognitive function, ADL, and comorbidities. In each category, the use of insulin or insulin secretagogues modifies both the upper and lower HbA1c limits. We evaluated the individual HbA1c target category according to comorbidities, use of insulin or insulin secretagogues, and grades for bedridden level and rating of dementia, as recorded by a family doctor in the letter of opinion for the use of Long‐Term Care Insurance at the beginning of the study period. The following category classification was based on the most severe level among the grades for bedridden level, rating of dementia, and the CCI score.

First, we classified patients who met all of the following criteria into Category I: independent or J grade for bedridden level, independent or I grade for the rating of dementia, and CCI score < 3.

Secondly, we classified the patients into Category II if they had CCI scores < 3 with either an A grade for bedridden level or a II grade for rating of dementia. Category II was characterized by mild cognitive impairment, impaired instrumental ADL, and stable basic ADL.

Thirdly, we classified the patients that met either of the conditions below into Category III: B or C grade for bedridden level, III, IV, or M grade for rating of dementia, or CCI score ≥ 3. Category III included patients with moderate to severe dementia, impaired basic ADL, or multiple comorbidities and functional disability.

Statistical Analysis

For the primary outcome, among participants whose first HbA1c value met the HbA1c target according to the guidelines, we defined those whose last HbA1c value missed the HbA1c target as the “deviating” group. Patients whose last HbA1c value met the HbA1c target were included in the “non‐deviating” group. In cases where the HbA1c target had changed according to cognitive or ADL decline, the outcome was evaluated at the time of the change. Throughout the observation period, we excluded patients with any diagnosis that could influence the HbA1c value, such as gastrointestinal bleeding, leukemia, multiple myeloma, or aplastic anemia.

Sex, age, place of care (home or nursing facility), dependence on public assistance, presence of terminal‐stage cancer, number of oral medications, and interval between HbA1c were compared between the deviating group and non‐deviating group, as defined using the management targets in the guidelines. Comparisons between groups were performed using the chi‐square test or Fisher's exact test and the Mann–Whitney U test, as appropriate.

In older patients with diabetes receiving home medical care, deviation from the HbA1c target in the above guidelines (coded as 1) was set as the outcome variable. We performed multivariable logistic regression analysis with forced entry of the following covariates in our model: age, sex, place of care (home or facility), interval between HbA1c tests (days), number of oral medications, dependence on public assistance, and presence of terminal‐stage cancer. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated.

Analysis of Longitudinal Glycemic Indicators (Sensitivity Analysis)

As a sensitivity analysis, we calculated the time‐in‐target (TiT) for HbA1c, defined as the proportion of HbA1c measurements that met each patient's individualized target (i.e., number of times HbA1c was within the individualized target range divided by the total number of HbA1c tests during the observation period). We also calculated the coefficient of variation (%CV), defined as the standard deviation of HbA1c divided by its mean, which indicates glycemic variability. Furthermore, we performed linear regression analysis using the %CV as the dependent variable, with the same covariates as those used in the main analysis including the place of care and interval between HbA1c tests, using forced entry.

All tests were two‐sided, with a significance level set at p < 0.05. Statistical analyses were performed using IBM SPSS Statistics, version 29.0.1 (IBM Corp., Armonk, NY, USA).

Results

Of the 1120 eligible patients, 838 met the HbA1c target value at the beginning of the observation period. Among these, we excluded four patients diagnosed with gastrointestinal bleeding during the observation period, leaving 834 patients for analysis. No patients were diagnosed with leukemia, multiple myeloma, or aplastic anemia.

The median participant age was 87.0 years (interquartile range [IQR] 82.0–90.0), with 292 (35.0%) men and 542 (65.0%) women. The median observation period was 400.0 days (IQR 203.8–743.5).

For the primary study outcome, deviation from the HbA1c target in the guidelines, 71 patients (8.5%) were classified as belonging to the deviating group and 763 patients (91.5%) to the non‐deviating group (Figure 1 ).

Participants' Characteristics

Table 1 shows that the proportion of patients whose place of care was at home was significantly higher in the deviating group than in the non‐deviating group. No significant difference was observed between the groups in terms of age, sex, number of oral medications, dependence on public assistance, presence of terminal‐stage cancer, observation period (days), oral feeding ability, CCI score, ADL, and cognitive function.

Treatment Status for Diabetes and Other Lifestyle‐Related Diseases

Among patients in both groups, most were classified as Category III according to the guidelines, with no significant difference between groups (Table 2 ). The proportion of patients treated with diabetes medications was significantly higher in the deviating group than that in the non‐deviating group. No significant differences were observed between groups in the use of antihypertensive medications, statins, or steroid medications. The interval between HbA1c tests was significantly longer in the non‐deviating group than in the deviating group.

HbA1c Targets in the Guidelines for the Treatment of Diabetes in the Older Patients

As shown in Table 3 , the proportion of the patients with the grades for bedridden level and rating of dementia that match Category I according to the guidelines (which included patients who were nearly independent both physically and mentally) was significantly higher in the deviating group than that in the non‐deviating group.

Factors Associated With Deviation From the HbA1c Target in Patients Receiving Home Medical Care

Multivariable logistic regression analysis showed that receiving care at home was significantly associated with deviation from the HbA1c targets in the guidelines (adjusted OR [aOR] = 2.06, 95% CI: 1.18–3.61), as shown in Table 4 . Additionally, each 1‐day increase in the interval between HbA1c tests was associated with an approximately 2% decrease in the aORs of HbA1c deviation (aOR = 0.98, 95% CI: 0.98–0.99). Age, sex, number of oral medications, dependence on public assistance, and presence of terminal‐stage cancer were not significantly associated with deviation from the HbA1c target.

The chi‐square test in the model was significant ( p < 0.001), and the Hosmer–Lemeshow test ( p = 0.308) indicated good model fit. The maximum correlation coefficient among independent variables was 0.23, indicating no strong correlations. All variance inflation factor values were ≤ 1.2, indicating no evidence of multicollinearity.

Analysis of Longitudinal Glycemic Indicators (Sensitivity Analysis)

For the TiT, 83% of participants showed a value of 100%, suggesting stable glycemic control within individualized HbA1c targets. Lower %CV values, indicating more stable glycemic control, were associated with care at a nursing facility ( p = 0.01) and extended interval between HbA1c tests ( p < 0.001). These results match the main results of our research and also suggest an association between stable glycemic control and the clinical decision of the physician regarding the place of care and intervals between HbA1c tests.

Discussion

In outpatient clinics, the treatment goal for patients with diabetes is usually set as achieving a good prognosis as well as maintaining good glycemic control to prevent diabetes complications. However, the goals for older patients with diabetes are preventing a decline in ADL and severe complications such as hypoglycemia, diabetic ketoacidosis, and a hyperosmolar hyperglycemic state. For these goals, physicians should avoid maintaining tight glycemic control to lower the risk of hypoglycemia through a strict diet and exercise therapy that could impair patient quality of life. By the end of the median 400‐day observation period in this study, 8.5% of older patients with diabetes receiving home medical care had HbA1c levels that deviated from the targets in the Guidelines for the Treatment of Diabetes in the Older Patients. Receiving care at home was significantly associated with deviation from the HbA1c targets in the above guidelines, as compared with receiving care at a nursing facility. This difference is possibly owing to environmental features of nursing facilities that are suitable for diet and exercise therapy as well as well‐controlled medical adherence, including for antihyperglycemic medication. Few studies have compared glycemic control among older patients with diabetes according to the place of care. Our findings suggest that it is important to consider the place of care in older patients with diabetes who have suboptimal glycemic control.

A longer interval between HbA1c tests was significantly associated with stable control within the HbA1c targets in the Guidelines for the Treatment of Diabetes in Older Patients. This result was in contrast to previous reports of a positive association between good glycemic control and a shorter interval between HbA1c tests [ 11 , 12 , 13 , 14 ]. This may be attributed to considerably limited access to blood testing in the home medical care setting and better understanding of the treatment environment in patients receiving home medical care than in those being treated at an outpatient clinic where intervals between HbA1c tests are modified according to physicians' expectations regarding the stability of individual glycemic control. These results suggest that modifications decided by physicians engaged in home medical care are reasonable, to a certain extent. However, the intervals between HbA1c tests may be intentionally shortened in cases where the physician expects unstable glycemic control.

In the setting of home medical care as well as outpatient care, shorter intervals between HbA1c tests will not necessarily result in better glycemic control. Ukai et al. reported no significant difference in glycemic control between monthly and bimonthly clinic visits among patients with stable diabetes [ 15 ]. Because the interval between HbA1c tests remains controversial, further research in the home medical care setting is warranted.

Factors that are considered more frequently in the home medical care setting than in an outpatient clinic include HbA1c fluctuation; lifestyle habits; variation in personal factors, such as the wishes of patients and their family; the care environment including caregiver status, history of hypoglycemia or hyperglycemia, cognitive function, ADL, and nursing care services received in individual cases. Few reports have documented clinical intervals between HbA1c tests and patient outcomes in the home medical care setting, which render our research findings uniquely novel. As applied to clinical situations, our findings suggest an exploratory association between the extended interval between HbA1c tests and the achievement of individualized HbA1c targets in patients receiving home medical care. Although these results should not be interpreted as causal, the findings suggest that physicians engaged in home medical care tend to extend the intervals between HbA1c tests in patients with stable glycemic control. Further prospective studies should evaluate the optimal intervals of HbA1c tests in home‐care settings.

The exploratory association of interval between HbA1c tests with reduced deviation from HbA1c targets likely reflects physicians' tendency to extend intervals between HbA1c tests in cases where glycemic control is expected to remain stable. Thus, the interval between HbA1c tests reflects the clinical judgment of a physician, which results in deviation from the HbA1c targets. However, it remains unclear whether more frequent HbA1c testing would prevent deviation from HbA1c targets.

Regarding the number of oral medications, despite our expectation that polypharmacy would be positively associated with deviation from guideline HbA1c targets [ 16 ], no significant association was found. This may be owing to either little influence of adherence to oral medication among originally stable patients treated mainly with diet and exercise therapy, or owing to intensive support to ensure medication adherence among unstable patients treated with antihyperglycemic medication or insulin.

This research has several limitations. First, we could follow cognitive and ADL decline that would influence the HbA1c target but did not follow the added prescription of insulin or insulin secretagogues, thereby potentially underestimating the deviation from the lower limit.

Second, we set the HbA1c target during the period in which cognitive function and ADL were recorded; consequent changes in the place of care or intervals between HbA1c tests after HbA1c deviation were not examined. We did not include such changes in this study and may have overestimated the deviation using a fixed HbA1c target, even in cases where the HbA1c target should have been relaxed.

Third, there is a certain possibility that not all individual physicians are familiar with the HbA1c target in the guidelines, which may have led to differences in setting actual targets in the clinic. Although we did not consider these in our research, the intervals between HbA1c tests may also be affected by the clinical skill of the physician in charge.

Although we evaluated the overall deviations from HbA1c targets in this research, distinguishing between deviations upward and downward, and examining the backgrounds of those patients who are more likely to deviate in either direction, are of clinical importance. Further investigation is needed in older patients with diabetes in the home medical care setting.

It is very important for older patients with diabetes receiving home medical care to avoid deviating from the HbA1c targets in the guidelines. Further investigation of factors associated with meeting the HbA1c target is needed in consideration of the place of care, appropriate intervals between HbA1c tests, and other clinical factors.

In conclusion, receiving care in a nursing facility and extended interval between HbA1c tests according to a physician's clinical decision were associated with stable HbA1c control in older patients with diabetes receiving home medical care. In future, evaluating the place of care and the appropriate interval between HbA1c tests will improve care quality among older patients with diabetes receiving home medical care.

Author Contributions

M.F. and T.K. conceived the research questions. M.F. contributed to the design of the research protocol. M.F. and T.K. conducted the statistical analysis. M.F. and T.K. drafted the initial manuscript in collaboration with Y.S., S.F., S.M., K.O., M.Y., and M.H. All authors revised and approved the final manuscript.

Funding

This research was supported by a Grant‐in‐Aid for Scientific Research (C), a category of competitive research funding, from the Japan Society for the Promotion of Science (JSPS) (Grant Number: 20K18845).

Ethics Statement

This study was approved by the Ethics Review Committee of the Institute of Science Tokyo (approval number: M2021‐182).

Consent

Informed consent was obtained through an opt‐out method, with the study outline, objectives, protection of personal information, and contact information posted on the clinics' website.

Conflicts of Interest

Mari Fukuhara and Takuma Kimura: Funding: CUC Inc. The other authors declare no conflicts of interest.

Acknowledgments

We would like to express our sincere gratitude to Miu Kitayama and Mao Kawamura for their valuable contributions to this study. We thank Analisa Avila, MPH, ELS, of Edanz ( https://jp.edanz.com/ac ) for editing a draft of this manuscript.

Data Availability Statement

Research data are not shared.

References

1 Ministry of Health, Labour and Welfare of Japan , Patient Survey ( Ministry of Health, Labour and Welfare of Japan , 2023 ), https://www.mhlw.go.jp/toukei/saikin/hw/kanja/23/index.html .

2 Japan Geriatrics Society & Japan Diabetes Society , Guidelines for the Treatment of Diabetes in the Elderly ( Nankodo , 2023 ).

3 T. Sugimoto , T. Sakurai , K. Uchida , et al., “ Impact of Type 2 Diabetes and Glycated Hemoglobin Levels Within the Recommended Target Range on Mortality in Older Adults With Cognitive Impairment Receiving Care at a Memory Clinic: NCGG‐STORIES ,” Diabetes Care 47 , no. 5 ( 2024 ): 864 – 872 . 38470970 10.2337/dc23-2324

4 S. Hirahara , “ Diabetes Treatment in Home Medical Care ,” Journal of the Japan Diabetes Society 60 , no. 5 ( 2017 ): 350 – 352 .

5 A. Araki , “ Individualized Treatment of Diabetes Mellitus in Older Adults ,” Geriatrics & Gerontology International 24 , no. 12 ( 2024 ): 1257 – 1268 . 39375857 10.1111/ggi.14979 PMC11628902

6 K. Kostev , T. Rockel , and L. Jacob , “ Prescription Patterns and Disease Control in Type 2 Diabetes Mellitus Patients in Nursing Home and Home Care Settings: A Retrospective Analysis in Germany ,” Journal of Diabetes Science and Technology 12 , no. 1 ( 2018 ): 136 – 139 . 28539088 10.1177/1932296817710477 PMC5761966

7 K. Shinohara , “ Continuous Pharmacological Management and Support for Elderly Diabetic Patients in Home Care Settings ,” Japanese Journal of Community Medicine and Pharmaceutical Sciences 8 , no. 1 ( 2021 ): 10 – 16 .

8 H. Koyano , “ Challenges in Supporting Home‐Based Insulin Self‐Injection and Self‐Monitoring of Blood Glucose Among Elderly Diabetic Patients ,” Japanese Journal of Geriatrics 64 , no. 5 ( 2019 ): 427 – 433 .

9 Ministry of Health, Labour and Welfare of Japan , Survey of Institutions and Establishments for Long‐Term Care ( Ministry of Health, Labour and Welfare of Japan , 2010 ).

10 Y. W. Hu , C. M. Yeh , C. J. Liu , T. J. Chen , N. Huang , and Y. J. Chou , “ Adapted Diabetes Complications Severity Index and Charlson Comorbidity Index in Predicting All‐Cause and Cause‐Specific Mortality Among Patients With Type 2 Diabetes ,” BMJ Open Diabetes Research & Care 11 , no. 2 ( 2023 ): e003262 . 10.1136/bmjdrc-2022-003262 PMC10069524 36977521

11 C. Imai , L. Li , R. A. Hardie , and A. Georgiou , “ Adherence to Guideline‐Recommended HbA1c Testing Frequency and Better Outcomes in Patients With Type 2 Diabetes: A 5‐Year Retrospective Cohort Study in Australian General Practice ,” BMJ Quality and Safety 30 , no. 9 ( 2021 ): 706 – 714 . 10.1136/bmjqs-2020-012026 PMC8380884 33542066

12 C. J. Duff , I. Solis‐Trapala , O. J. Driskell , et al., “ The Frequency of Testing for Glycated Haemoglobin, HbA1c, Is Linked to the Probability of Achieving Target Levels in Patients With Suboptimally Controlled Diabetes Mellitus ,” Clinical Chemistry and Laboratory Medicine 57 , no. 2 ( 2018 ): 296 – 304 . 30281512 10.1515/cclm-2018-0503

13 O. J. Driskell , D. Holland , J. L. Waldron , et al., “ Reduced Testing Frequency for Glycated Hemoglobin, HbA1c, Is Associated With Deteriorating Diabetes Control ,” Diabetes Care 37 , no. 10 ( 2014 ): 2731 – 2737 . 25249670 10.2337/dc14-0297

14 C. Fu , L. Ji , W. Wang , et al., “ Frequency of Glycated Hemoglobin Monitoring Was Inversely Associated With Glycemic Control of Patients With Type 2 Diabetes Mellitus ,” Journal of Endocrinological Investigation 35 , no. 3 ( 2012 ): 269 – 273 . 21606668 10.3275/7743

15 T. Ukai , S. Ichikawa , M. Sekimoto , S. Shikata , and Y. Takemura , “ Effectiveness of Monthly and Bimonthly Follow‐Up of Patients With Well‐Controlled Type 2 Diabetes: A Propensity Score Matched Cohort Study ,” BMC Endocrine Disorders 19 , no. 1 ( 2019 ): 43 . 31046742 10.1186/s12902-019-0372-5 PMC6498692

16 R. Koto , A. Nakajima , T. Miwa , and K. Sugimoto , “ Multimorbidity, Polypharmacy, Severe Hypoglycemia, and Glycemic Control in Patients Using Glucose‐Lowering Drugs for Type 2 Diabetes: A Retrospective Cohort Study Using Health Insurance Claims in Japan ,” Diabetes Therapy 14 , no. 7 ( 2023 ): 1175 – 1192 . 37195511 10.1007/s13300-023-01421-5 PMC10241751