Factors Associated With Hospitalization for Hypoglycemia and Hyperglycemia Among Older People in Long‐Term Care Facilities

ABSTRACT

Background

Individuals with diabetes newly entering long‐term care facilities (LTCFs) encounter changes in care needs and facility and care process‐related factors, which potentially impact diabetes treatment outcomes. This study examined the 12‐month incidence of hospitalizations for hypoglycemia and hyperglycemia in residents with diabetes and factors associated with these hospitalizations following LTCF entry.

Methods

This retrospective cohort study included residents aged ≥ 65 years with diabetes who entered a LTCF between 2015 and 2018 using data from the Registry of Senior Australians. Cumulative incidence of hospitalization for hypoglycemia or hyperglycemia in the 12 months following entry was evaluated. Factors associated with hypoglycemia or hyperglycemia hospitalizations were examined using a Fine–Gray model, accounting for the competing event of mortality. Subdistribution hazard ratios (sHRs) were reported.

Results

Of the 55,734 individuals included (median age 84 years), 1.0% (95% confidence interval [CI]: 0.9–1.1) were hospitalized for hypoglycemia, and 0.5% (95% CI: 0.4–0.6) for hyperglycemia in the 12 months after LTCF entry. Factors associated with a higher rate of hospitalization for hypoglycemia included high (sHR: 2.59, 95% CI: 1.61–4.17) or medium (sHR: 2.61, 95% CI: 1.61–4.24) level of care needs, renal disease (sHR: 1.22, 95% CI: 1.01–1.49), prior hospitalization with hypoglycemia (sHR: 2.18, 95% CI: 1.77–2.67) or hyperglycemia (sHR: 1.61, 95% CI: 1.19–2.18), use of insulin (sHR: 6.15, 95% CI: 4.99–7.59), sulfonylureas (sHR: 1.41, 95% CI: 1.14–1.74), or angiotensin‐converting enzyme inhibitors (sHR: 1.23, 95% CI: 1.02–1.47). Factors associated with a higher rate of hospitalization for hyperglycemia included preferred spoken language other than English (sHR: 1.40, 95% CI: 1.02–1.93), dementia (sHR: 1.39, 95% CI: 1.08–1.80), prior hospitalization with hyperglycemia (sHR: 3.88, 95% CI: 2.72–5.53) or hypoglycemia (sHR: 2.50, 95% CI: 1.83–3.41), use of insulin (sHR: 2.01, 95% CI: 1.51–2.69), or metformin (sHR: 1.42, 95% CI: 1.10–1.84).

Conclusions

The risk of hospitalization for hypoglycemia or hyperglycemia may be reduced through diabetes care planning at LTCF entry informed by the identified risk factors for these complications.

Summary.

• Key points:

  • ○
    Incidence of hospitalization within 12 months of LTCF entry was 1.0% for hypoglycemia, and 0.5% for hyperglycemia.
  • ○
    Risk of hospitalization for hypoglycemia was higher among individuals with greater complex care needs, renal disease, use of insulin, sulfonylurea or angiotensin‐converting enzyme inhibitors, and prior hospitalization with hypoglycemia or hyperglycemia.
  • ○
    Risk of hospitalization for hyperglycemia was higher among individuals whose preferred spoken language was not English, individuals with dementia, those receiving insulin or metformin, and prior hospitalization with hypoglycemia or hyperglycemia.
• Why does this paper matter?

  • ○
    Hypoglycemia or hyperglycemia are important acute diabetes complications that can lead to hospital admission, negatively affecting residents' health and quality of life, and increasing the cost of care.
  • ○
    This study demonstrated that while the incidence of hospitalization for hypoglycemia and hyperglycemia is low, it remains a concern in LTCFs, with risk factors associated with these hospitalizations within their first 12 months of residents' stay identified using national real‐world data.
  • ○
    These complications can be prevented through individualized diabetes treatment plans at the time of LTCF entry that account for identified risk factors.

1. Introduction

Diabetes disproportionately affects the older population, with prevalence reaching up to 37% for individuals in long‐term care facilities (LTCFs) [1, 2]. Diabetes complications are a leading but potentially preventable cause of hospitalization among older people with diabetes, contributing to increased morbidity, mortality, and health care costs [3, 4, 5]. In the LTCF population, hospitalizations can increase the risk of further complications such as infection and cognitive impairment, and compromise quality of life [6]. Hypoglycemia and hyperglycemia are acute, life‐threatening diabetes complications which could have deleterious effects for frail older people in LTCFs and may contribute to hospitalizations [7, 8].

LTCF entry is associated with changes in individuals' care needs, care providers, food intake, and a higher risk of infection, potentially affecting diabetes treatment outcomes [9, 10]. Previous small studies conducted in LTCFs that report hypoglycemia and hyperglycemia incidence largely focus on blood glucose test results [11, 12, 13], with few examining severe complications resulting in hospitalization [14, 15]. Although studies conducted among community‐dwelling older adults have identified various individual and medicine‐related factors associated with the incidence of hypoglycemia or hyperglycemia [16, 17, 18], there is limited evidence from LTCFs [19], where individuals are likely to encounter unique facility and care process‐related factors at admission. Hence, identifying residents at increased risk of hospitalization for hypoglycemia or hyperglycemia following LTCF entry can enable proactive care planning and tailored glycemic targets, and potentially lead to hospitalization avoidance. This study examined the 12‐month incidence of hospitalization for hypoglycemia and hyperglycemia and factors associated with these hospitalizations following LTCF entry.

2. Methods

2.1. Study Design, Setting, and Data Source

This retrospective cohort study was conducted using the National Historical Cohort of the Registry of Senior Australians (ROSA). ROSA links aged care, health care, and social welfare data, providing de‐identified details of sociodemographic characteristics, health conditions, care needs, use of outpatient health care services, use of medicines, hospitalizations, emergency department presentations, and mortality for older individuals who received government‐subsidized aged care services from 2002 onwards [20]. This study used the aged care assessments and services, mortality, pharmaceutical, and hospital admission datasets (Text S1).

2.2. Study Cohort

Individuals aged ≥ 65 years with diabetes from four Australian states (New South Wales, Victoria, Queensland, and South Australia), who did not identify as Aboriginal or Torres Strait Islander, and first entered a LTCF between 01/01/2015 and 31/12/2018 were included (Figure S1). Diabetes status was ascertained using aged and/or health care data (pharmaceutical and hospital claims) as previously published [1]. The cohort was followed up for up to 12 months from LTCF entry until LTCF exit, death (i.e., competing risk) or 31/12/2019, whichever occurred first.

2.3. Outcomes of Interest

The main outcomes of interest were time to first unplanned hospitalization or emergency department presentation with a primary diagnosis of hypoglycemia or hyperglycemia at a public hospital in the 12 months following LTCF entry, identified using International Statistical Classification of Disease and Related Problems, 10th Revision, Australian Modification (ICD‐10‐AM) code (Table S1) [21].

2.4. Covariates

Individual, LTCF, health service, and medicine utilization‐related covariates were used to characterize the study cohort and examine factors associated with hospitalization for hypoglycemia or hyperglycemia (Text S2). Individual and LTCF characteristics included age at LTCF entry, sex, place of birth, preferred spoken language, Rx‐Risk‐V pharmaceutical‐based comorbidity score [22], specific health conditions, level of complex care needs, LTCF ownership, state, and remoteness. Health service and medicine utilization covariates included hospitalization with hypoglycemia or hyperglycemia and use of glucose‐lowering medicine and medicines potentially affecting blood glucose levels before LTCF entry.

2.5. Statistical Analysis

The cumulative incidence function estimated the probability of hypoglycemia or hyperglycemia in the 12 months following LTCF entry, accounting for the competing risk of mortality. Fine–Gray subdistribution hazard regression models were used to estimate subdistribution hazard ratios (sHRs) with 95% confidence intervals (CIs) and to analyze the relative effect of covariates on the cumulative incidence of the outcomes of interest. The proportionality assumption was checked using Schoenfeld residuals and examined using log (−log) plots. Statistical significance was set at α = 0.05. All estimations used complete case analyses as only ~2.7% of cases (n = 1530 for hypoglycemia and n = 1527 for hyperglycemia) were excluded due to missing data.

2.6. Sensitivity Analysis

The incidence of hospitalization and associated factors for hypoglycemia or hyperglycemia using either primary and/or secondary diagnoses ICD‐10 AM codes, the primary reason for hospitalization among those with hypoglycemia or hyperglycemia in their secondary diagnoses, and cause of death in the 12 months following LTCF entry were examined.

Data were analyzed using SAS (version 9.4, SAS Institute Inc., Cary, NC, USA), Stata Statistical Software (Release 18, TM: StataCorp LLC), and R V 4.3.3.

3. Results

Of the 55,734 individuals included, 53.9% (n = 30,066) were female, 46.5% (n = 25,905) were living with dementia, and the median age was 84 years (interquartile range [IQR]: 78–88) (Table 1). In the 12 months preceding LTCF entry, 8.5% (n = 4734) were hospitalized with hypoglycemia, and 2.8% (n = 1580) were hospitalized with hyperglycemia. Over the 12‐month follow‐up period, 33.1% (n = 18,435) of individuals died, with cardiovascular disease accounting for 28.4% (n = 5231) of these deaths (Table S2).

TABLE 1.

Characteristics of the study cohort (n = 55,734).

Characteristics of residents	n (%) or median (IQR)
Sociodemographic
Age at LTCF entry, median (IQR)	84 (78–88)
Female, n (%)	30,066 (53.9)
Born in Australia, n (%) a	34,556 (62.1)
Preferred spoken language: English b	47,376 (85.4)
Australian state of residence, n (%)
New South Wales	21,493 (38.6)
Victoria	17,479 (31.4)
Queensland	10,886 (19.5)
South Australia	5876 (10.5)
Remoteness of residence, n (%) c
Major city	38,709 (69.7)
Outside major city	16,877 (30.4)
LTCF ownership, n (%)
Not‐for‐profit	29,197 (52.4)
For‐profit	23,787 (42.7)
Government	2750 (4.9)
Health conditions
Rx‐Risk‐V comorbidity score, median (IQR) d	6 (4–7)
0–4, n (%)	18,779 (33.7)
5–6, n (%)	16,967 (30.4)
≥ 7, n (%)	19,988 (35.9)
Specific health conditions, n (%)
Dementia	25,905 (46.5)
Heart failure	17,558 (31.5)
Ischemic heart disease	17,371 (31.2)
Fall	17,185 (30.8)
Cerebrovascular disease	14,458 (25.9)
Renal disease	12,096 (21.7)
Health care needs
Activities of daily living needs, n (%) e
Nil or low	8433 (15.5)
Medium	17,183 (31.5)
High	28,957 (53.1)
Behavioral daily living needs, n (%) e
Nil or low	13,170 (24.1)
Medium	14,908 (27.3)
High	26,495 (48.5)
Complex health care needs, n (%) e
Nil or low	9289 (17.0)
Medium	15,407 (28.2)
High	29,877 (54.7)
Health services utilization
History of hospitalization with diabetes complications in the 12 months before LTCF entry, n (%) f
Renal complication	15,461 (27.7)
Neurological complication	5910 (10.6)
Hypoglycemia	4734 (8.5)
Peripheral vascular complication	4641 (8.3)
Ophthalmic complication	3513 (6.3)
Hyperglycemia	1580 (2.8)
Number of emergency department presentations in the 12 months before LTCF entry, n (%)
None	13,364 (24.0)
1	14,640 (26.3)
2–4	20,988 (37.7)
≥ 5	6742 (12.1)
Number of unplanned hospitalizations in the 12 months before LTCF entry, n (%)
None	14,419 (25.9)
1	18,118 (32.5)
2–4	19,638 (35.2)
≥ 5	3559 (6.4)
Medicines use
Number of unique medicines dispensed in the 6 months before LTCF entry, median (IQR)	10 (7–14)
0–4, n (%)	5643 (10.1)
5–10, n (%)	22,595 (40.5)
≥ 11, n (%)	27,496 (49.3)
Glucose‐lowering medicines dispensed in the 3 months before LTCF entry, n (%)
Any glucose‐lowering medicine	31,681 (56.8)
Metformin	20,595 (37.0)
Sulfonylurea	11,222 (20.1)
Insulin g	10,073 (18.1)
Dipeptidyl peptidase 4 inhibitors	7324 (13.1)
Sodium‐glucose co‐transporter 2 inhibitors	738 (1.3)
Thiazolidinediones	357 (0.6)
Glucagon‐like peptide 1 receptor agonists	220 (0.4)
Acarbose	164 (0.3)
Medicines potentially affecting blood glucose levels that were dispensed in the 6 months before LTCF entry, n (%)
Beta blockers	20,461 (36.7)
Angiotensin‐converting enzyme inhibitors	17,414 (31.2)
Systemic corticosteroids	8181 (14.7)
Antipsychotics	7343 (13.2)
Thiazide diuretics	1308 (2.3)
Fibrates	1610 (2.9)

Abbreviations: IQR, interquartile range; LTCF, long‐term care facility.

^a Missing (n = 133).

^b Missing (n = 234).

^c Missing (n = 138).

^d Excluding diabetes.

^e Missing (n = 1161).

^f Principal and secondary diagnosis codes evaluated in the year before long‐term care facility entry.

^g Supplies were assessed 6 months before long‐term care facility entry.

The cumulative incidence of hospitalization for hypoglycemia over 12 months was 1.0% (95% CI: 0.9–1.1, n = 530 individuals) (Figure 1). The 12‐month cumulative incidence of hospitalization for hyperglycemia was 0.5% (95% CI: 0.4–0.6, n = 276 individuals).

FIGURE 1.

Cumulative incidence of hospitalizations for (a) hypoglycemia or (b) hyperglycemia (primary diagnosis) and competing risk of mortality in the 12 months after long‐term care facility entry.

Factors associated with hospitalization for hypoglycemia included high (sHR: 2.59, 95% CI: 1.61–4.17) or medium (sHR: 2.61, 95% CI: 1.61–4.24) complex care needs compared to low or none, prior hospitalization with hypoglycemia (sHR: 2.18, 95% CI: 1.77–2.67) or hyperglycemia (sHR: 1.61, 95% CI: 1.19–2.18), use of insulin (sHR: 6.15, 95% CI: 4.99–7.59), sulfonylureas (sHR: 1.41, 95% CI: 1.14–1.74), or angiotensin‐converting enzyme inhibitors (sHR: 1.23, 95% CI: 1.02–1.47) (Table 2).

TABLE 2.

Factors associated with hospitalization for hypoglycemia or hyperglycemia (primary diagnosis) in the 12 months after long‐term care facility entry (n = 54,204).

Covariates	Hypoglycemia		Hyperglycemia
	sHR (95% CI)	p	sHR (95% CI)	p
Sociodemographic
Age (per 1 year increment)	1.01 (0.99–1.02)	0.288	0.98 (0.97–1.00)	0.074
Female vs. male	1.10 (0.92–1.31)	0.313	0.91 (0.72–1.17)	0.470
Preferred spoken language (other vs. English)	1.19 (0.93–1.52)	0.164	1.40 (1.02–1.92)	0.038
Australian state of residence
Queensland vs. New South Wales	1.24 (0.99–1.56)	0.066	1.54 (1.11–2.14)	0.010
South Australia vs. New South Wales	1.00 (0.74–1.35)	0.989	1.36 (0.90–2.05)	0.146
Victoria vs. New South Wales	0.76 (0.60–0.95)	0.017	1.11 (0.82–1.51)	0.499
Remoteness of residence
Outside major city vs. Major city	1.07 (0.87–1.32)	0.498	1.02 (0.76–1.38)	0.874
LTCF ownership
Not for profit vs. Government	1.52 (0.89–2.59)	0.129	1.32 (0.63–2.76)	0.464
For profit vs. Government	1.43 (0.83–2.47)	0.198	1.72 (0.82–3.62)	0.154
Health conditions
Dementia (yes vs. no)	1.18 (0.99–1.42)	0.064	1.39 (1.08–1.80)	0.011
Heart failure (yes vs. no)	0.70 (0.57–0.86)	0.001	0.55 (0.40–0.75)	< 0.001
Ischemic heart disease (yes vs. no)	0.93 (0.76–1.13)	0.449	0.91 (0.69–1.21)	0.516
Cerebrovascular disease (yes vs. no)	0.89 (0.73–1.09)	0.265	1.07 (0.82–1.40)	0.638
Renal disease (yes vs. no)	1.22 (1.01–1.49)	0.042	0.91 (0.67–1.24)	0.563
Complex health care needs
Medium vs. nil or low	2.61 (1.61–4.24)	< 0.001	1.05 (0.68–1.63)	0.811
High vs. nil or low	2.59 (1.61–4.17)	< 0.001	1.26 (0.84–1.90)	0.266
Health services utilization
History of hospitalization with diabetes complications in the 12 months before LTCF entry
Hypoglycemia (yes vs. no)	2.18 (1.77–2.67)	< 0.001	2.50 (1.83–3.41)	< 0.001
Hyperglycemia (yes vs. no)	1.61 (1.19–2.18)	0.002	3.88 (2.72–5.53)	< 0.001
Ophthalmic complication (yes vs. no)	1.24 (0.96–1.61)	0.100	1.04 (0.68–1.58)	0.857
Peripheral vascular complication (yes vs. no)	1.06 (0.82–1.37)	0.672	0.91 (0.63–1.33)	0.627
Neurological complication (yes vs. no)	1.15 (0.91–1.46)	0.232	1.04 (0.69–1.56)	0.845
Medicines use
Number of unique medicines dispensed in the 6 months before LTCF entry
5–11 vs. 0–4	0.92 (0.62–1.38)	0.700	1.09 (0.68–1.75)	0.706
≥ 11 vs. 0–4	0.86 (0.57–1.31)	0.493	0.94 (0.57–1.55)	0.814
Glucose‐lowering medicines dispensed in the 3 months before LTCF entry
Metformin	0.76 (0.62–0.93)	0.006	1.42 (1.10–1.84)	0.008
Sulfonylurea	1.41 (1.14–1.74)	0.001	1.24 (0.93–1.65)	0.139
Insulin a	6.15 (4.99–7.59)	< 0.001	2.01 (1.51–2.69)	< 0.001
Dipeptidyl peptidase 4 inhibitors	1.04 (0.82–1.33)	0.725	1.28 (0.93–1.75)	0.129
Sodium‐glucose co‐transporter 2 inhibitors	1.13 (0.63–2.03)	0.670	0.79 (0.32–1.92)	0.600
Medicines potentially affecting blood glucose levels that were dispensed in the 6 months before LTCF entry
Systemic corticosteroids	0.98 (0.76–1.26)	0.879	0.63 (0.41–0.97)	0.037
Beta blockers	1.17 (0.97–1.42)	0.094	Not included	—
Angiotensin‐converting enzyme inhibitors	1.23 (1.02–1.47)	0.031	Not included	—
Antipsychotics	Not included	—	1.27 (0.92–1.76)	0.144

Note: The bolded values indicate statistically significant associations.

Abbreviations: LTCF, long‐term care facility; sHR, subdistribution hazard ratio.

^a Supplies were assessed 6 months before long‐term care facility entry.

Factors associated with hyperglycemia hospitalizations included preferred spoken language other than English (sHR: 1.40, 95% CI: 1.02–1.92), presence of dementia 1.39 (95% CI: 1.08–1.80), prior hospitalization with hyperglycemia (sHR: 3.88, 95% CI: 2.72–5.53) or hypoglycemia (sHR: 2.50, 95% CI: 1.83–3.41), use of insulin (sHR: 2.01, 95% CI: 1.51–2.69), or metformin (sHR: 1.42, 95% CI: 1.10–1.84) (Table 2).

Using the primary and secondary diagnostic codes in our sensitivity analysis, the cumulative incidence of hospitalization for hypoglycemia was 3.1% (95% CI: 3.0–3.3, n = 1700 individuals) (Figure S2a), and hyperglycemia was 2.1% (95% CI: 2.2–2.2, n = 1156 individuals) (Figure S2b). Factors associated with hospitalization for hypoglycemia or hyperglycemia remained largely unchanged from the primary analysis (Table S3). The most common primary diagnosis among individuals hospitalized with hypoglycemia or hyperglycemia was diseases of the respiratory system followed by circulatory system disease (Table S4).

4. Discussion

This population‐based study of older people with diabetes following entry to LTCFs identified a 1.0% cumulative incidence of hospitalization for hypoglycemia and 0.5% for hyperglycemia over 12 months. Ten factors were associated with hospitalizations for hypoglycemia, including complexity of health care needs, prior hospitalization with hypoglycemia or hyperglycemia, and insulin or sulfonylurea use. Nine factors were associated with hospitalization for hyperglycemia, including preferred spoken language, dementia, prior hospitalization with hypoglycemia or hyperglycemia, and insulin use. The findings highlight that hypoglycemia and hyperglycemia remain a concern in LTCFs. These events can potentially be prevented through diabetes management interventions targeting the identified modifiable associated risk factors.

The incidence of hospitalization for hypoglycemia and hyperglycemia in our study is lower than reported in the previous studies from LTCFs [11, 12, 13, 14]. These studies largely defined hypoglycemia and hyperglycemia based on blood glucose test results as documented in case notes. Our study included only the most severe cases that resulted in hospitalization as identified from principal diagnosis codes. However, the inclusion of primary and secondary hospitalization diagnoses identified a higher incidence, highlighting the contribution of other health conditions to hypoglycemia and hyperglycemia incidence.

Several of the factors associated with hypoglycemia or hypoglycemia identified were concordant with previous studies in community‐dwelling older people [16, 17, 18]. In the present study, insulin was associated with a higher risk of both hypoglycemia and hyperglycemia. Previous studies have shown an increase in insulin use around the time of LTCF entry, along with a high overall prevalence of use [10, 23]. Our study is also consistent with reports indicating an increased risk of hypoglycemia among sulfonylurea users [19]. Further, the higher risk of hypoglycemia or hyperglycemia among people with a prior episode(s) of both complications suggests potential glycemic variability, hence a need for closer glucose monitoring in these individuals. These findings support current guideline recommendations to limit insulin and sulfonylurea use among frail older adults living in LTCFs [24].

In the current study, individuals with greater complex health care needs were at higher risk of hospitalization for hypoglycemia. People with a higher level of health care needs are more dependent on care providers for their diabetes management, and variability in practices could increase the risk of hypoglycemia [25]. A higher risk of hospitalization for hyperglycemia among individuals with dementia in our study disagrees with previous studies that have largely reported a higher risk of hypoglycemia [16, 26]. While this may be due to care providers adopting a cautious approach in terms of diabetes management, it highlights the potential for undertreatment. Further, individuals whose preferred spoken language was not English had a higher risk of hospitalization for hyperglycemia, which may potentially be associated with communication difficulties. Signs and symptoms associated with hyperglycemia are often attributed to other diseases in older people, and these, particularly in people with communication barriers, may result in failure of early recognition and treatment initiation [27].

4.1. Strengths and Limitations

This study used comprehensive, national‐level data covering ~87% of the total LTCF population in Australia to examine factors associated with both hypoglycemia and hyperglycemia in the 12 months following LTCF entry. Our findings highlight the importance of identifying individuals at higher risk for these complications by health and aged care providers to inform care planning at the time of LTCF entry.

This study has limitations. Diabetes laboratory test results, including glycated hemoglobin (HbA1c) are not captured in ROSA. We were unable to examine by type of diabetes, but a previous study identified that > 99% of residents with diabetes in LTCFs have type 2 diabetes [28]. This study included public hospital claims data only, as private hospital data was not available for all states included. This may underestimate the true incidence of hypoglycemia and hyperglycemia, but its overall impact is likely minimal, as 92% of all unplanned hospitalizations in Australia are managed in public hospitals [29]. In Australia, complex cases are typically managed in public hospitals, highlighting the potential for variations among individuals treated in public versus private hospitals [30]. Although we were unable to examine factors associated with the studied outcomes for a subset of the cohort with private hospital data due to the small number of events, the identified factors are expected to be largely similar. This study included hypoglycemia and hyperglycemia hospitalizations only, and while this provides good specificity, it underestimates the true incidence of hypoglycemia and hyperglycemia within LTCFs [31]. Finally, the glucose‐lowering medicine used at LTCF entry may not reflect use over the 12 months following LTCF entry, and future studies should consider the potential impact of diabetes treatment changes during the follow‐up period.

5. Conclusion

While hospitalization incidence for hypoglycemia and hyperglycemia was low in the 12 months following LTCF entry, an opportunity for early identification and risk mitigation for individuals at increased risk of poor diabetes‐related outcomes exists. The risk of hospitalization for hypoglycemia was higher with greater health care needs, insulin or sulfonylurea use, whereas individuals whose preferred spoken language was not English, those with dementia, and dispensed insulin were at higher risk of hospitalization for hyperglycemia. Implementation of individualized diabetes treatment plans at the time of LTCF entry that account for identified risk factors may help to reduce hospitalizations for hypoglycemia and hyperglycemia.

Author Contributions

Y.A.W., G.E.C., M.C.I., and J.K.S. conceived the study. Y.A.W. drafted the study protocol, analyzed the data, wrote the manuscript, and is the guarantor of this work and, as such, had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. All authors participated in study design, contributed to the interpretation of results, critically reviewed the manuscript, and read and approved the final manuscript.

Conflicts of Interest

J.K.S. is a nonexecutive director of Southern Cross Care SA, NT & VIC (aged care provider organization). The other authors declare no conflicts of interest.

Supporting information

Text S1. Data source.

Text S2. Covariates.

Figure S1. Cohort selection flowchart.

Figure S2. Cumulative incidence of hospitalization with a primary and/or secondary diagnosis of (a) hypoglycemia or (b) hyperglycemia and competing risk of mortality in the 12 months after LTCF entry.

Table S1. International Statistical Classification of Disease and Related Problems, 10th Revision, Australian Modification (ICD‐10‐AM) codes used to identify hospitalizations for hypoglycemia or hyperglycemia.

Table S2. Cause of death among individuals with diabetes in the 12 months after long‐term care facility entry.

Table S3. Factors associated with hospitalization for hypoglycemia or hyperglycemia (primary diagnosis and/or secondary diagnosis) in the 12 months after long‐term care facility entry (n = 54,204).

Table S4. Primary reason for all‐cause hospitalization, and hospitalization with a secondary diagnosis of hypoglycemia or hyperglycemia among individuals with diabetes in the 12 months after long‐term care facility entry, n (%).

Wondimkun Y. A., Caughey G. E., Inacio M. C., et al., “Factors Associated With Hospitalization for Hypoglycemia and Hyperglycemia Among Older People in Long‐Term Care Facilities,” Journal of the American Geriatrics Society 73, no. 9 (2025): 2859–2867, 10.1111/jgs.19553.