Abstract
Objective:
To characterize sliding-scale insulin (SSI) use in US nursing homes (NHs) before and after the COVID-19 pandemic.
Design:
Cross-sectional study.
Setting and Participants:
A total of 129,829 US NH residents on SSI (01/2018–06/2022) across 12 NH chains with a common electronic health record system.
Methods:
Among all residents with at least 1 administration of SSI documented in the electronic medication administration record, we described resident demographics, frequency of SSI monotherapy vs combination therapy with another diabetes medication, number of daily capillary blood glucose readings (“fingersticks”), and hypoglycemia (capillary blood glucose <70 mg/dL) and hyperglycemia after first SSI use. We used interrupted time series analysis (ITS) with segmented linear regression models to examine whether the monthly prevalence of SSI use changed at and after the onset of the COVID-19 pandemic (March 2020).
Results:
There were 129,829 unique NH residents with SSI use [51% women, average age 71.3 (SD 11.7) years]. Of these, 36% of residents received SSI monotherapy and 64% received SSI combination therapy. Residents on SSI received an average of 3.96 (SD 1.41) fingersticks per day. Overall, 26% of SSI users experienced a hypoglycemic event within 30 days of the first SSI dose. The ITS analysis identified a step decrease in the rate of SSI use following the onset of the COVID-19 pandemic (43 fewer SSI users per 1000 insulin users) but no change in overall trend over time from before the onset of the pandemic.
Conclusions and Implications:
SSI use and fingerstick burden are high in NH residents. Hypoglycemia occurred commonly among residents on SSI. Future research should compare the safety and effectiveness of SSI monotherapy vs other diabetes medication regimens to guide person-centered prescribing decisions in NHs.
Keywords: Diabetes mellitus, hypoglycaemia, hyperglycemia, Insulin, nursing homes, deprescriptions
More than one-third of older adults in the United States who reside in nursing homes (NHs) have type 2 diabetes mellitus (T2DM).1–3 Management of diabetes in NH residents is complex because of large fluctuations in glucose levels and a high risk of hypoglycemia from polypharmacy and multimorbidity, which are independent risk factors for hypoglycemia and prevalent in NH residents.3–5 Hypoglycemia has extremely detrimental consequences for NH residents, including falls, fractures, frailty, cognitive decline, and reduced quality of life.6–8 Recommended strategies for diabetes management, therefore, focus on decreasing the risk of hypoglycemia while minimizing persistent or severe hyperglycemia.9
Guidelines from the American Diabetes Association (ADA) and the American Medical Directors Association (AMDA) recommend avoiding sliding scale insulin (SSI) monotherapy for long-term diabetes management.9,10 In an SSI regimen, fast-acting insulin is administered several times per day and dosed based on bedside blood glucose (BG) readings taken immediately prior to administration.11 Although SSI may be appropriate in short-term, transitional, or acute care situations, SSI monotherapy is considered a reactive method of glycemic control that is not ideal for long-term treatment because of the potential for hyperglycemic or hypoglycemic episodes.12 Evidence suggests that proactive approaches to manage BG, such as using longer-acting basal insulins, should be preferred.13 Further, SSI use poses an increased provider and caregiver burden as a result of the frequent BG monitoring required compared with other glucose-lowering regimens.14,15
Recent evidence of the use of SSI as monotherapy or as part of a combination diabetes medication regimen among generalizable populations of short- and long-term NH residents is lacking. A 5-year (2016–2020) serial cross-sectional study examining diabetes medication prescribing in US NHs found that the yearly prevalence of SSI prescribing was 41% to 43%.16 Furthermore, a recent national examination of Veterans Affairs (VA) NHs found that more than half of all short-term insulin regimens used were for SSI.11 However, the VA NH resident population (<1% of all US NHs), is not representative of the private sector NH population in characteristics such as race, comorbidities, and sex (eg, >98% men). In addition, little is known about detailed characteristics of SSI use in a generalizable NH resident population, such as the frequency of SSI administration and fingerstick measurements, combinations with other diabetes medications, duration of use, and prevalence of SSI use before vs after the COVID-19 pandemic.11
To address these knowledge gaps and inform diabetes-related quality improvement initiatives, guideline changes, and intervention development, we leveraged national electronic health record (EHR) data from >1600 NHs from 12 US NH chains to conduct a cross-sectional study among a large, nationally representative population of NH residents receiving SSI therapy. Our main objectives were (1) to characterize SSI use, including use with other diabetes medications, in a sample that is representative of the US NH population and (2) to evaluate changes in the rates of SSI use before and after the onset of the COVID-19 pandemic. We hypothesized that the prevalence of SSI use would decrease because of increased infection control measures and staffing burden in the NH setting after the onset of the pandemic.
Methods
Data Source
We leveraged data from 12 NH chains with a common EHR system (PointClickCare). These data contain person-level information from >1600 NH facilities across almost every state in the United States, including information on a daily census (person-level file that contains each resident’s disposition on a given day, including transfers, discharges, and deaths), resident demographics (eg, age, sex, race/ethnicity), immunization records, vital signs (eg, blood pressure, temperatures, bedside “fingerstick” glucose measurements), diagnosis codes, medication orders (medication initiation and discontinuation), nonmedication orders (eg, procedures, diagnostic testing, advance directives), and the electronic medication administration record (eMAR). In addition, the EHR contains Minimum Data Set assessments, which are federally mandated clinical evaluations that must be conducted at admission for all NH residents and at least quarterly thereafter during their stay in the NH.17 A map showing the geographical distribution of facilities that provided data is available in Supplementary Material. Brown University’s Institutional Review Board approved this study and waived the requirement for informed consent.
Population and Characterizing SSI Use
We identified all residents at first administration of SSI in the NH recorded in the eMAR (study figure illustrating cohort identification available in Figure 1) between January 2, 2018 and June 30, 2022. We identified SSI use by administrations of rapid-acting, short-acting, and intermediate-acting insulins that also contained order directions with the term “scale” in the order direction field, identified using the FIND function in SAS.11,18 Preliminary analyses using random samples of 100 selected orders from each NH chain found that the sensitivity of this method for identifying SSI use was greater than 97%. Rapid-acting insulins of interest included lispro, aspart, and glulisine, but we also included other types of insulin that may be unconventionally used as an SSI: short-acting (regular insulin), intermediate-acting (insulin NPH), and combination insulins. We described resident demographics, clinical characteristics, and laboratory values using Minimum Data Set assessments and NH facility clients data sets measured up to 90 days before or within 30 days after the first SSI administration. We excluded residents missing age or sex data in both data sources.
Fig. 1.
Study diagram. GLM, glucose-lowering medication; MDS, Minimum Data Set.
We stratified the study population based on whether they were receiving SSI as monotherapy or in combination with other glucose-lowering medications (henceforth “SSI combination therapy”). Concurrent use of other diabetes medications (eg, metformin, sodium-glucose cotransporter 2 inhibitors, dipeptidyl peptidase 4 inhibitors) was determined by examining the eMAR-recorded administration of eligible diabetes medications on the same day as the first day of administration of SSI, except for weekly doses of glucagon-like peptide-1 (GLP-1) agonists (dulaglutide, exenatide, and semaglutide), which were assessed within 7 days before or after first SSI administration. We measured the average and maximum number of doses of SSI and fingerstick glucose measurements per day in the first 7 and 30 days of SSI therapy. We also estimated the length of SSI therapy from first administration of SSI to either death, discontinuation of SSI defined as 7 days without use, and discharge from the NH. We compared SSI use and resident characteristics between residents with SSI monotherapy vs SSI combination therapy using absolute standardized mean differences (aSMDs; with aSMD ≥0.10 indicating a meaningful difference).19 Finally, we measured the prevalence of abnormal glycemic events as measured by bedside “fingerstick” glucose measures, including hypoglycemia (BG < 70 mg/dL), severe hypoglycemia (BG < 50 mg/dL), and hyperglycemia (BG > 300 mg/dL) within 7 days and 30 days of first SSI use.
Changes in SSI Use During the COVID-19 Pandemic
To examine whether SSI use changed at or after the onset of the COVID-19 pandemic, we calculated the monthly proportion and rate of residents with SSI among residents with any insulin use from January 2019 to May 2021. Select NH chains did not contribute data outside of this time because of different start dates for data availability and divestitures. Studying a period wherein the same chains provided data, therefore, would reduce the likelihood that the addition or removal of facilities with different SSI prescribing practices would not falsely induce a time trend. However, a stability analysis included all data from January 2018 to June 2022 to study SSI prescribing periods over a longer time frame. We examined whether the monthly rate of use of SSI changed after the onset of the COVID-19 pandemic via an interrupted time series (ITS) analysis using ordinary least squares regression with Newey-West standard errors to account for autocorrelation.20 The ITS estimated the linear trend in the rate of SSI users over time, the immediate effect of the onset of the pandemic (defined as March 2020) on the rate of users, and the effect of the onset of the pandemic on the linear trend in use over time. After preliminary visualization of rates per month over time, we incorporated a 3-month lag to account for a potential lagged effect of the onset of the COVID-19 pandemic on SSI use; however, we also conducted a stability analysis that did not use a lag for comparison. All analyses were conducted using SAS, version 9.4 (SAS Institute), and Stata, version 17 (StataCorp).
Results
Population and Type of SSI Therapy
We identified a total of 129,829 residents on SSI [51% female, mean age 71.3 (SD 11.7) years; Table 1]. In total, 46,816 (36.1%) residents were on SSI monotherapy and 83,013 (63.9%) were on combination therapy with another diabetes medication. The SSI monotherapy and combination therapy groups generally had similar baseline characteristics, except for age [72.4 years (SD 11.7) in the monotherapy group vs 70.7 years (SD 11.6) in the combination therapy group; aSMD 0.14] and the time to start of SSI from NH admission [3.2 days (SD 24.8) in the monotherapy group vs 7.4 days (SD 38.5) in the combination therapy group; aSMD 0.13].
Table 1.
Characteristics of Nursing Home Residents on Sliding Scale Insulin Monotherapy vs Combination* Glucose-Lowering Therapy
| Total (N = 129,829) | SSI Monotherapy (n = 46,816) | Combination Therapy* (n = 83,013) | aSMD | |
|---|---|---|---|---|
|
| ||||
| Age, y, mean ± SD | 71.33 ± 11.68 | 72.38 ± 11.71 | 70.74 ± 11.62 | 0.14 † |
| Female sex | 66,179 (51.0) | 24,087 (51.5) | 42,092 (50.7) | 0.02 |
| Poor prognosis (life expectancy <6 mo) | 1730 (1.3) | 678 (1.5) | 1052 (1.3) | 0.02 |
| Any falls | 19,036 (14.7) | 6763 (14.5) | 12,273 (14.8) | <0.01 |
| Major falls | 626 (0.5) | 221 (0.5) | 405 (0.5) | <0.01 |
| Race/ethnicity | ||||
| White | 89,627 (69.03) | 31,029 (66.3) | 58,598 (70.6) | 0.09 |
| Black | 26,729 (20.6) | 10,851 (23.2) | 15,878 (19.1) | 0.01 |
| Hispanic/Latino‡ | 5435 (4.2) | 1986 (4.2) | 3449 (4.2) | <0.01 |
| Asian or Pacific Islander | 2981 (2.3) | 1140 (2.4) | 1841 (2.2) | 0.01 |
| Indigenous‡ | 550 (0.4) | 178 (0.4) | 372 (0.4) | 0.01 |
| Missing | 4507 (3.5) | 1632 (3.5) | 2875 (3.5) | <0.01 |
| Concurrent diagnoses, conditions, and care states | ||||
| Comatose | 177 (0.1) | 86 (0.2) | 91 (0.1) | 0.02 |
| Hospice | 1605 (1.2) | 647 (1.4) | 958 (1.2) | 0.02 |
| Cancer | 11,688 (9.0) | 4664 (10.0) | 7024 (8.6) | 0.05 |
| Alzheimer disease and related dementias§ | 27,669 (21.3) | 10,957 (23.4) | 16,712 (20.1) | 0.08 |
| Coronary artery disease | 34,042 (26.2) | 11,776 (25.2) | 22,266 (26.8) | 0.04 |
| Heart failure | 36,626 (28.2) | 13,470 (28.8) | 23,156 (27.9) | 0.02 |
| Hypertension | 96,672 (74.5) | 34,266 (73.2) | 62,406 (75.2) | 0.05 |
| Peripheral vascular disease | 17,300 (13.3) | 5952 (12.7) | 11,348 (13.7) | 0.03 |
| Renal insufficiency | 44,436 (34.2) | 16,636 (35.5) | 27,800 (33.5) | 0.04 |
| Hyperlipidemia | 70,857 (54.6) | 24,070 (51.4) | 46,787 (56.4) | 0.09 |
| Foot infection | 3917 (3.0) | 1103 (2.4) | 2814 (3.4) | 0.06 |
| Diabetes foot ulcer | 6357 (4.9) | 1807 (3.9) | 4500 (5.5) | 0.08 |
| Steroid use | 11,405 (8.8) | 3651 (7.8) | 7754 (9.2) | 0.06 |
| SSI characteristics, mean ± SD | ||||
| Time to start of SSI from NH admission, d | 5.86 ± 34.22 | 3.15 ± 24.76 | 7.39 ± 38.45 | 0.13 † |
| Duration of SSI use, d | 30.57 ± 72.4 | 26.34 ± 59.42 | 32.95 ± 78.69 | 0.10 † |
| Number of fingersticks per day, mean ± SD | ||||
| 7 d | 4.01 ± 1.45 | 3.70 ± 1.26 | 4.19 ± 1.52 | 0.35 † |
| 30 d | 3.96 ± 1.41 | 3.68 ± 1.25 | 4.13 ± 1.46 | 0.32 † |
| Maximum administrations per day, mean ± SD | ||||
| 7 d | 5 ± 1.98 | 5 ± 1.78 | 5 ± 2.06 | 0.35 † |
| 30 d | 5 ± 2.12 | 5 ± 1.91 | 6 ± 2.20 | 0.35 † |
| Insulins used for SSI | ||||
| Insulin lispro only | 75,215 (57.9) | 27,222 (58.2) | 47,993 (57.8) | 0.01 |
| Insulin aspart only | 27,556 (21.2) | 9051 (19.3) | 18,505 (22.3) | 0.01 |
| Insulin glulisine | 71 (0.05) | 14 (0.03) | 57 (0.07) | 0.02 |
| Insulin aspart and lispro | 13,834 (10.7) | 4409 (9.4) | 9425 (11.4) | 0.06 |
| Other insulins only∥ | 10,119 (7.8) | 5007 (10.7) | 5112 (6.2) | 0.16 † |
| Oral antidiabetes medications or GLP-1 | ||||
| Metformin | 23,230 (17.9) | — | 23,230 (28.0) | — |
| SU | 9966 (7.7) | — | 9966 (12.0) | — |
| DPP4I | 5436 (4.2) | — | 5436 (6.6) | — |
| GLP-1A | 4136 (3.2) | — | 4136 (5.0) | — |
| SGLT2I | 1518 (1.2) | — | 1518 (1.8) | — |
| Concurrent non-SSI insulin use | ||||
| Basal-bolus | 14,562 (11.2) | — | 14,562 (17.5) | — |
| Basal only | 42,775 (33.0) | — | 42,775 (51.5) | — |
| Bolus only | 5887 (4.5) | 5887 (7.1) | ||
| Number of antidiabetes medications concurrent with SSI | ||||
| 1 | 61,894 (47.7) | — | 61,894 (74.6) | — |
| 2 | 16,674 (12.8) | — | 16,674 (20.1) | — |
| 3+ | 4443 (3.4) | — | 4443 (5.3) | — |
| Glycemic events | ||||
| Prevalence of hypoglycemia (blood glucose <70 mg/dL) | ||||
| 7 d | 17,320 (13.4) | 4548 (9.8) | 12,772 (15.4) | 0.17 † |
| 30 d | 33,122 (25.6) | 9366 (20.1) | 23,756 (28.7) | 0.20 † |
| Prevalence of severe hypoglycemia (blood glucose <50 mg/dL) | ||||
| 7 d | 4727 (3.7) | 1458 (3.1) | 3269 (4.0) | 0.04 |
| 30 d | 10,775 (7.3) | 3342 (7.2) | 7433 (9.0) | 0.07 |
| Prevalence of hyperglycemia (blood glucose >300 mg/dL) | ||||
| 7 d | 65,921 (50.9) | 18,110 (38.8) | 47,811 (57.7) | 0.38 † |
| 30 d | 79,778 (61.6) | 23,054 (49.2) | 56,724 (68.5) | 0.39 † |
DPP4I, dipeptidyl peptidase 4 inhibitor; GLP-1A, glucagon-like peptide agonist; SGLT2I, sodium-glucose cotransporter-2 inhibitor; SU, sulfonylurea.
Unless otherwise noted, values are n (%).
SSI in addition to other glucose-lowering medications.
aSMD ≥0.10 indicates a meaningful difference between groups.
Self-identified label from Minimum Data Set assessment.
Measured using the Minimum Data Set indicator for dementia, which likely underestimates prevalence.
“Other insulins”: insulins other than insulin aspart, insulin lispro, or insulin glulisine.
The most common insulins used for SSI were insulin lispro (57.9%), insulin aspart (21.2%), or a combination of the two (10.7%). Insulin product use was similar between the monotherapy and combination therapy groups. In the combination therapy group, most residents (74.6%) received only 1 other diabetes medication in addition to SSI; the most common were metformin (28.0%) followed by sulfonylureas (12.0%). Among residents who were taking a non-SSI insulin in addition to SSI, the most common regimen was a basal-only regimen (51.5%), followed by basal-bolus (17.5%), followed by bolus only regimen (7.1%).
Characteristics of SSI Use
The mean duration of SSI use was 30.6 days (SD 72.4) but slightly differed between the monotherapy and combination groups [26.3 (SD 59.4) vs 33.0 (SD 78.7) days, respectively; aSMD 0.10]. The primary reasons for the end of follow-up were discharge from the home (76%) or death (1.6%); overall, 23% of SSI users discontinued use while still in the NH. The mean number of fingersticks per day was lower in the monotherapy group [7 days: 3.70 (SD 1.45) vs 4.19 (SD 1.52), aSMD 0.35; 30 days: 3.7 (SD 1.3) vs 4.1 (SD 1.5), aSMD 0.32]. The average maximum administrations per day was also lower in the monotherapy group [7 days: 4.6 (SD 1.78) vs 5.2 (SD 2.06), aSMD 0.35; 30 days: 4.9 (SD 1.91) vs 5.5 (SD 2.20), aSMD 0.35].
Hypoglycemia During SSI Use
Overall, 25.6% of residents with SSI use experienced at least 1 instance of hypoglycemia (BG level <70 mg/dL) in the first 30 days after SSI initiation, and the prevalence within 7 days of initiation was 13.4%. Hypoglycemia was more common in the combination therapy group at both the 7-day and 30-day follow-up points (7 days: 15.4% vs 9.8%, aSMD 0.17; 30 days: 28.7% vs 20.1%, aSMD 0.20). Severe hypoglycemia (BG < 50 mg/dL) occurred in 7.3% of the total cohort within 30 days, with similar rates between the 2 groups. Hyperglycemia (BG > 300 mg/dL) was prevalent in 50.9% of all residents by 7 days and 61.6% of all residents by 30 days. Hyperglycemic events occurred more commonly in the combination therapy group by 7 days (57.7% vs 38.8%, aSMD 0.38) and 30 days (68.5% vs 49.2%, aSMD 0.39).
Impact of the COVID-19 Pandemic on SSI Use
The average monthly proportion of SSI use among all residents with any insulin use was 63.1% (SD 0.03), January 2019 to May 2021. Rates ranged from 644 to 670 per 1000 insulin users prior to and ranged from 589 to 642 per 1000 insulin users following the onset of the pandemic. The ITS analysis identified a step decrease in SSI use at the onset of the pandemic, with 42.5 fewer SSI users per 1000 insulin users (95% CI 65–19 fewer SSI users per 1000 insulin users; P = .004). However, the overall trend over time for SSI use after the pandemic (ie, the “slope” change) was not significant (coefficient of intervention × time = 1.7 more SSI users per 1000 insulin users, 95% CI −0.95 to 4.37; P = .197) (Figure 2). In the stability analysis that included extended data from January 2018 to June 2022, the extended calendar time showed a more cyclical pattern of SSI use with a postpandemic peak in April 2021. Results were similar to the primary analysis when removing the 3-month lag in the step effect of the onset of the COVID-19 pandemic. Results from the primary and stability ITS models are available in Supplementary Material.
Fig. 2.
Monthly rate of sliding scale insulin use per 1000 residents (January 2019 to May 2021).
Discussion
In this study of 129,829 US NH residents, we observed that approximately two-thirds of NH residents with insulin use were on SSI therapy, and one-third of residents with SSI therapy were on monotherapy. Use of SSI therapy appeared to decrease at the onset of the COVID-19 pandemic, with similar trends in SSI use over time thereafter as before the pandemic. Finally, abnormal glycemic events were common among SSI users regardless of combination treatment or monotherapy, with about 25% experiencing hypoglycemia and 60% experiencing hyperglycemia.
The high identified rate of SSI use, with almost 6 in 10 residents on insulin therapy receiving SSI, aligns with previous studies in other NH populations. In one US NH chain, 54% of new orders for insulin among newly admitted residents were for SSI.21 Another study of NH residents from 117 different US NHs found that >50% of all residents on basal insulin were also receiving SSI.22 Further, a recent study describing SSI use in the VA NHs found that among residents on short-acting insulin, the prevalence of SSI regimen use was high (57%) compared with corrective dose (20%) or fixed-dose regimens (23%).11 The present study extends these important foundational findings on the prevalence of SSI use to the private NH sector, which comprises >99% of NHs in the United States; our study includes about 11% of all US NHs. Finally, we extended previous work by examining the prevalence of SSI monotherapy vs combination therapy, finding that most residents were on other diabetes medications in addition to SSI and that hypoglycemia was more common among those with combination treatment.
Most importantly, there was a high prevalence of hypoglycemia among all residents on SSI, with approximately 1 in 4 residents experiencing hypoglycemia and 1 in 10 having a severe hypoglycemic event. Hypoglycemia was more common in the SSI combination therapy group compared to monotherapy. This aligns with previous evidence that some glucose-lowering therapies may enhance the hypoglycemic effects of insulin or vice versa.23,24 We also observed an increase in the proportion of residents experiencing a glycemic event from day 7 to day 30, suggesting that there may be an increased risk over time. Further, about 2 in 3 residents also had hyperglycemic events, suggesting that the tradeoff for glucose control with SSI regimens may be limited. However, future studies need to be conducted to evaluate the causal relationship between SSI use and abnormal glycemic events.
Next, there was a high prevalence of SSI use even after the onset of the COVID-19 pandemic, though overall use did decrease immediately after the pandemic onset. During the pandemic, 21% of NHs reported staffing shortages,25 and SSI use in NHs has been associated with an increase in fingerstick burden.14 The shortage of nursing aides, who commonly perform fingerstick glucose readings in the NH, has been the highest among NH staff, and increasing since the onset of the pandemic.25,26 We found that, on average, NH residents on SSI received approximately 4 fingerstick readings per day, exceeding the recommended number of 2 BG readings per day for older adults on insulin and presumably placing a substantial burden on both staff and residents throughout the day.10 Because SSI is not recommended per guidelines from both the AMDA and ADA, the fingerstick burden due to SSI may not be worth the extra staffing hours and resources.10,27,28 Furthermore, fingerstick readings for SSI cause significant discomfort to residents and decrease their quality of life.29
Recent guidance from the ADA Standards of Care in Diabetes recommends avoiding hypoglycemia and overtreatment of diabetes, deintensification of treatment goals, and simplification of complex treatment plans to reduce hypoglycemia and polypharmacy for older adults with diabetes.27 Simplification strategies focusing on transitioning prandial insulin (including SSI) to basal insulin or noninsulin agents can reduce the risk of hypoglycemia while maintaining glucose control.30 With higher rates of abnormal glycemic events, long duration of therapy, and treatment burden associated with SSI, alternative diabetes medication regimens to SSI should be evaluated in comparison to SSI. A randomized control trial comparing SSI to basal-bolus insulin in the NH setting found that participants who volunteered to be on a basal-bolus regimen had a significantly lower fasting BG posttrial vs pretrial compared with the SSI participants who had no significant change in fasting BG. This study also found no significant differences in rates of hypo- and hyperglycemia.31 These findings suggest that switching from SSI to a basal-bolus regimen is feasible, safe, and effective in the NH setting and may result in better FBG management with minimal changes in rates of abnormal glycemic events. Research examining the safety and effectiveness of switching from SSI to basal-only insulin, oral, and injectable noninsulin diabetes medications needs to be conducted to broaden our understanding of optimal pharmacologic glucose management for older adults in NHs.
There were limitations to this study. First, we could not assess the dose of the sliding scale administered. Second, we were unable to assess longer-term glucose control measured by hemoglobin A1c, because these measures were not consistently available for residents within the EHR data. Third, because this is a descriptive study, we could not make any inferences on the effect of SSI use on hypoglycemia and hyperglycemia. In addition, we were unable to measure exposure to SSI before NH admission. Future studies should use causal inference methods to examine the comparative safety of SSI monotherapy on rates of hypoglycemia and the benefits and harms of deprescribing SSI.
Conclusions and Implications
In conclusion, we found that SSI use remains a common practice in the NH setting. Hypo- and hyperglycemia were common for residents on SSI, and that these events occurred more frequently in residents on combination therapy. This research calls for policy change and clinical interventions aligned with current guidance to simplify diabetes regimens and reduce the risk of hypoglycemia in NH residents.
Supplementary Material
Acknowledgments
This work was supported by the National Institute of Aging (NIA) of the National Institutes of Health under award number U54AG063546, which funds NIA Imbedded Pragmatic Alzheimer’s Disease and AD-Related Dementias Clinical Trials Collaboratory (NIA IMPACT Collaboratory). Supplemental funding was provided under grant numbers U54AG063546-S07 and U54AG063546-S08. Drs Hayes’s and Zullo’s time was also supported, in part, by grant RF1AG061221 from the NIA. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health or the Department of Veterans Affairs. This research was also supported by the Irwin E Ginsberg Aging Research Fund at Brown University School of Public Health.
Footnotes
Disclosure
K.N.H. has received grant funding paid directly to Brown University for collaborative research from Insight Therapeutics, Sanofi, and Genentech for research on complex insulin regimens and influenza outbreak control in nursing homes. K.N.H. has also served as a consultant for the Canadian Agency for Drugs and Technologies in Health. A.R.Z. has received grant funding paid directly to Brown University by Sanofi for collaborative research on the epidemiology of infections and vaccinations among nursing home residents and infants.
Supplementary Data
Supplementary data related to this article can be found online at https://doi.org/10.1016/j.jamda.2024.01.004.
References
- 1.Newton CA, Adeel S, Sadeghi-Yarandi S, et al. Prevalence, quality of care, and complications in long term care residents with diabetes: a multicenter observational study. J Am Med Dir Assoc. 2013;14:842–846. [DOI] [PubMed] [Google Scholar]
- 2.Zarowitz B, Allen C, O’Shea T, Dalal MR, Haumschild M, DiGenio A. Type 2 diabetes mellitus treatment patterns in U.S. nursing home residents. Postgrad Med. 2015;127:429–437. [DOI] [PubMed] [Google Scholar]
- 3.Nelis SM, Wu YT, Matthews FE, et al. The impact of co-morbidity on the quality of life of people with dementia: findings from the IDEAL study. Age Ageing. 2019;48:361–367. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Growdon ME, Gan S, Yaffe K, Steinman MA. Polypharmacy among older adults with dementia compared with those without dementia in the United States. J Am Geriatr Soc. 2021;69:2464–2475. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Migdal A, Yarandi SS, Smiley D, Umpierrez GE. Update on diabetes in the elderly and in nursing home residents. J Am Med Dir Assoc. 2011;12:627–632. e2. [DOI] [PubMed] [Google Scholar]
- 6.Abdelhafiz AH, Rodríguez-Mañas L, Morley JE, Sinclair AJ. Hypoglycemia in older people - a less Well Recognized risk factor for frailty. Aging Dis. 2015;6:156–167. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Signorovitch JE, Macaulay D, Diener M, et al. Hypoglycaemia and accident risk in people with type 2 diabetes mellitus treated with non-insulin antidiabetes drugs. Diabetes Obes Metab. 2013;15:335–341. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Johnston SS, Conner C, Aagren M, Ruiz K, Bouchard J. Association between hypoglycaemic events and fall-related fractures in Medicare-covered patients with type 2 diabetes. Diabetes Obes Metab. 2012;14:634–643. [DOI] [PubMed] [Google Scholar]
- 9.Munshi MN, Florez H, Huang ES, et al. Management of diabetes in long-term care and Skilled nursing facilities: a Position Statement of the American diabetes association. Diabetes Care. 2016;39:308–318. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.American Medical Directors Association. Diabetes Management in the Post-Acute and Long-Term Care Setting. Columbia, MD: AMDA; 2015. [DOI] [PubMed] [Google Scholar]
- 11.Lam K, Gan S, Nguyen B, Jing B, Lee SJ. Sliding scale insulin use in a national cohort study of nursing home residents with type 2 diabetes. J Am Geriatr Soc. 2022;70:2008–2018. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Golightly LK, Jones MA, Hamamura DH, Stolpman NM, McDermott MT. Management of diabetes mellitus in hospitalized patients: efficiency and effectiveness of sliding-scale insulin therapy. Pharmacotherapy. 2006;26:1421–1432. [DOI] [PubMed] [Google Scholar]
- 13.Schoeffler JM, Rice DAK, Gresham DG. 70/30 insulin algorithm versus sliding scale insulin. Ann Pharmacother. 2005;39:1606–1610. [DOI] [PubMed] [Google Scholar]
- 14.Pandya N, Wei W, Meyers JL, Kilpatrick BS, Davis KL. Burden of sliding scale insulin use in elderly long-term care residents with type 2 diabetes mellitus. J Am Geriatr Soc. 2013;61:2103–2110. [DOI] [PubMed] [Google Scholar]
- 15.Jeon SY, Shi Y, Lee AK, et al. Fingerstick glucose monitoring in Veterans Affairs nursing home residents with diabetes mellitus. J Am Geriatr Soc. 2021;69:424–431. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Pandya N, Jung M, Norfolk A, Goldblatt C, Trenery A, Sieradzan R. Medication prescribing for type 2 diabetes in the US long-term care setting: observational study. J Am Med Dir Assoc. 2023;24:790–797.e4. [DOI] [PubMed] [Google Scholar]
- 17.Minimum data Set 3.0 public reports | CMS. Accessed June 27, 2022. https://www.cms.gov/Research-Statistics-Data-and-Systems/Computer-Data-and-Systems/Minimum-Data-Set-3-0-Public-Reports [Google Scholar]
- 18.SAS Help center: FIND function. Accessed June 2, 2023. https://documentation.sas.com/doc/en/pgmsascdc/9.4_3.5/lefunctionsref/p00ab6ey29t2i8n1ihel88tqtga9.htm
- 19.Austin PC. Balance diagnostics for comparing the distribution of baseline covariates between treatment groups in propensity-score matched samples. Stat Med. 2009;28:3083–3107. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Turner SL, Forbes AB, Karahalios A, Taljaard M, McKenzie JE. Evaluation of statistical methods used in the analysis of interrupted time series studies: a simulation study. BMC Med Res Methodol. 2021;21:181. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Pandya N, Thompson S, Sambamoorthi U. The prevalence and persistence of sliding scale insulin use among newly admitted elderly nursing home residents with diabetes mellitus. J Am Med Dir Assoc. 2008;9:663–669. [DOI] [PubMed] [Google Scholar]
- 22.Davis KL, Wei W, Meyers JL, Kilpatrick BS, Pandya N. Use of basal insulin and the associated clinical outcomes among elderly nursing home residents with type 2 diabetes mellitus: a retrospective chart review study. Clin Interv Aging. 2014;9:1815–1822. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Mogensen UM, Andersson C, Fosbøl EL, et al. Sulfonylurea in combination with insulin is associated with increased mortality compared with a combination of insulin and metformin in a retrospective Danish nationwide study. Diabetologia. 2015;58:50–58. [DOI] [PubMed] [Google Scholar]
- 24.Fu H, Xie W, Curtis B, Schuster D. Identifying factors associated with hypoglycemia-related hospitalizations among elderly patients with T2DM in the US: a novel approach using influential variable analysis. Curr Med Res Opin. 2014;30:1787–1793. [DOI] [PubMed] [Google Scholar]
- 25.McGarry BE, Grabowski DC, Barnett ML. Severe staffing and personal protective equipment shortages faced by nursing homes during the COVID-19 pandemic. Health Aff Proj Hope. 2020;39:1812–1821. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Shen K, McGarry BE, Grabowski DC, Gruber J, Gandhi AD. Staffing patterns in US nursing homes during COVID-19 outbreaks. JAMA Health Forum. 2022;3:e222151. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 27.ElSayed NA, Aleppo G, Aroda VR, et al. 13. Older adults: standards of care in diabetes—2023. Diabetes Care. 2022;46(Supplement_1):S216–S229. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Haas LB. Diabetes Management in Long-Term Care Settings. Alexandria, VA: American Diabetes Association; 2014. [Google Scholar]
- 29.Barnhart C, McClymont K, Smith AK, Au-Yeung A, Lee SJ. “Everyone else gets ice cream here more often than I do—it burns me up” - Perspectives on diabetes care from nursing home residents and their doctors. BMC Geriatr. 2016;16:28. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 30.Munshi MN, Slyne C, Segal AR, Saul N, Lyons C, Weinger K. Simplification of insulin regimen in older adults and risk of hypoglycemia. JAMA Intern Med. 2016;176:1023–1025. [DOI] [PubMed] [Google Scholar]
- 31.Dharmarajan TS, Mahajan D, Zambrano A, et al. Sliding scale insulin vs basal-bolus insulin therapy in long-term care: a 21-day randomized controlled trial comparing efficacy, safety and feasibility. J Am Med Dir Assoc. 2016;17:206–213. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.