Achieving target inpatient glycemic management outcomes has been shown to influence important clinical outcomes such as hospital length of stay and readmission rates. However, arguably the most profound, lasting impact of inpatient diabetes management is achieved at the time of discharge—namely reconciling and prescribing the right medications and making referrals for follow-up. Discharge planning offers a unique opportunity to break through therapeutic inertia, offer diabetes self-management education, and institute an individualized treatment plan that prepares the patient for discharge and promotes self-care and engagement. However, the path to a successful discharge plan can be fraught with potential pitfalls for clinicians, including lack of knowledge and experience with newer diabetes medications, costs, concerns over insurance coverage, and lack of time and resources. This article presents an algorithm to assist clinicians in selecting discharge regimens that maximize benefits and reduce barriers to self-care for patients and a framework for creating an interdisciplinary hospital diabetes discharge program.
Diabetes and hyperglycemia are common in hospitalized patients, affecting ∼25% of inpatients (1). People with diabetes may have a higher lifetime likelihood of hospital admissions and readmissions (1–7), nearly twice that of people without diabetes. Hospitals understand that this increased risk may have important financial implications (1,5,8). Interestingly, data on the importance of inpatient glycemic management have been conflicting (5,9,10). Implementation of specialized diabetes teams was found to reduce inpatient dysglycemia rates, infections, and even 30-day readmissions and lengths of stay in some studies (1,10–18).
Arguably the most important component of inpatient diabetes management is not the inpatient glycemic management at all, but rather determining the treatment regimen on discharge. The overwhelming majority of patients’ lives are spent outside of the hospital setting, so, intuitively, the medications that patients take at home will have a much more significant cumulative impact than their medication regimen during hospitalization.
Guidelines for diabetes care in the hospital are published annually by the American Diabetes Association (ADA) and call for a “structured discharge plan tailored to the individual patient with diabetes” (19,20). For example, not all patients are discharged to home; some may require rehabilitation and could be transferred to a skilled nursing facility. Homeless patients may require a referral to a social service agency that provides temporary residence and/or shelter. Comorbidities such as chronic kidney disease, heart disease, or gastrointestinal (GI) disease can influence medication choices and patients’ ability to perform self-care. Additionally, patients on insulin therapy represent a high-risk group for post-discharge complications and may require additional diabetes self-management education or resources (21,22).
Moreover, the moment of discharge presents a unique opportunity to reassess and revise a patient’s home diabetes regimen. Hospitals should deploy algorithms to assist providers with making antidiabetes medication adjustments based on patients’ admission A1C values (23). Additionally, new or worsening medical conditions (e.g., renal failure) or new prescriptions that are not compatible with the prior home regimen may necessitate an adjustment in diabetes outpatient therapy. However, therapeutic inertia is pervasive yet underrecognized (24). In a large study of admissions to U.S. Veteran Affairs hospitals, among patients with diabetes and a high A1C, only 22% received a change in outpatient diabetes therapy upon discharge (24). The reasons for therapeutic inertia are numerous and include anchoring bias, time constraints, unfamiliarity with other medications or treatment options, and provider burnout. Physicians tend to lock in or “anchor” to a patient’s existing medications despite ample clinical evidence of chronically poor glycemic control. Also, clinicians may believe it is inappropriate to change an outpatient regimen managed by another clinician (24).
Systematically and reproducibly instituting a successful diabetes discharge plan for patients is challenging and requires considerable forethought and planning. In this article, we discuss barriers, strategies, and key steps to achieving this goal.
The Three Pillars of Successful Discharge Planning
There are three major components, or pillars, of diabetes care that must be addressed at discharge to help ensure lasting treatment success: medications, glucose monitoring, and outpatient education/resources. These are discussed in more detail below.
I. Medications
Prescribing medication is both an art and a science, particularly in diabetes management. Toward this end, the ADA guidelines offer a flexible treatment algorithm with different options depending on which factor(s) are most pressing in each patient (25). Similarly, we offer here a basic framework for medication selection at the time of discharge. Ultimately, however, the best treatment is the one adapted to the needs and wishes of the individual patient.
It is well known in behavioral economics that the more difficult a task is to accomplish, the less likely a person is to do it consistently over time. Conversely, the simpler or more convenient a task is to perform, the more likely a person will be to continue to do it well (26). It is no surprise that Amazon has become one of the largest, most successful companies in the United States because it has removed much of the “friction” of shopping. Ordering, paying for, and receiving products is simple, fast, and convenient and does not require much effort. This type of thinking should be applied to diabetes management and discharge planning as well.
Medication nonadherence is epidemic among diabetes patients, and >30% of patients perform limited self-care behaviors (27). It has been well studied that major factors associated with diabetes medication nonadherence include cost, insulin use, hypoglycemia, perceived treatment efficacy, and treatment complexity (28,29). Indeed, one study found that the greatest predictor of diabetes nonadherence was patients feeling that their “medicines are hard to take” (30). Therefore, a patient’s inpatient basal-bolus insulin regimen should not be heedlessly continued upon discharge; rather, inpatient providers should thoughtfully choose a discharge regimen that removes the friction hindering successful medication adherence. This includes minimizing the number of times per day a patient has to take medications, while simultaneously keeping the treatment regimen simple and affordable.
It is important to note that the following discussion of medications is specific to patients with type 2 diabetes, even though some of the earlier principles may apply to all people with diabetes. Additionally, the following are the authors’ suggestions, but ultimately the number and types of diabetes medications prescribed on discharge should be consistent with existing ADA guidelines (25).
To maximize performance of self-care behaviors, patients with type 2 diabetes should take all of their diabetes-related medications once daily, all at the same time if possible (31). A simple strategy is to consider starting one noninsulin medication for every 1% the patient’s A1C is above target, up to three (or, rarely, four) medications. For example, for an A1c of 10%, consider starting three noninsulin diabetes medications (32). For an A1C >10–12%, consider also adding on insulin, as discussed further below (Table 1).
Percentage Points A1C Is Above Target . | Medications . | |
---|---|---|
Maximize Benefits . | Minimize Costs . | |
1 | Metformin ER | Metformin |
2 | SGLT2 inhibitor | Pioglitazone |
3 | GLP-1 receptor agonist | Sulfonylurea |
≥4 | Basal insulin | Premixed insulin* |
Percentage Points A1C Is Above Target . | Medications . | |
---|---|---|
Maximize Benefits . | Minimize Costs . | |
1 | Metformin ER | Metformin |
2 | SGLT2 inhibitor | Pioglitazone |
3 | GLP-1 receptor agonist | Sulfonylurea |
≥4 | Basal insulin | Premixed insulin* |
For each percentage point A1C is above target, providers should consider adding an additional medication, as listed above, to the treatment regimen (32). SGLT2 inhibitors and GLP-1 receptor agonists can be prescribed at lower A1C levels if other indications exist, as per guidelines (25).
If premixed insulin is initiated, the authors recommend not prescribing a sulfonylurea on discharge.
Metformin
Recommendation: start metformin extended release (ER) 750 mg daily with breakfast. After 1 week, if there are no significant GI side effects, increase dose to 1,500 mg (two tablets) with breakfast.
The ADA’s Standards of Medical Care in Diabetes—2022 (25) recommends metformin as the first-line agent in medically treating patients with type 2 diabetes. Recent reports have suggested that, in addition to promoting some weight loss and not causing hypoglycemia, metformin may have numerous other positive pleiotropic benefits such as anticancer, antiaging, anti-inflammatory, and cardioprotective properties (33–36). Metformin can be prescribed either as an immediate release (IR) formulation that should be taken twice daily or an extended release (ER) formulation that can be taken once daily. Metformin IR tablets comes in doses of 500, 750, 850, and 1,000 mg, whereas metformin ER tablets are available in doses of 500, 750, and 1,000 mg. Studies have shown that doses of metformin ER have comparable A1C effects to the equivalent daily doses of metformin IR (37,38).
All of the metformin IR doses are generic and therefore affordable. However, for metformin ER, it is important to realize that the 500- and 750-mg tablets are also generic and covered by most insurance companies, whereas the 1,000-mg tablet is only available as a brand-name product (e.g., Fortamet or Glumetza). Therefore, prescribing two metformin ER 1,000-mg tablets daily (2,000 mg total) can cost a patient several hundred dollars more than prescribing two 750-mg tablets daily (1,500 mg total). Moreover, the difference in A1C benefit between 1,500 and 2,000 mg daily is only ∼0.3% (37). One may point out that patients could take four generic 500-mg tablets daily, but taking four pills and a protracted uptitration schedule (e.g., 500 mg for the first week, increasing by one tablet each week for the following 3 weeks) represents increased treatment friction and may be more likely to limit self-care than a simpler two-pill regimen of metformin ER.
The feared rare complication of metformin toxicity is lactic acidosis, which can lead to hospitalization or, rarely, death. Because metformin is renally cleared, a patient’s renal function influences the decision of whether to initiate this medication, as well as selecting the final goal dose. Studies have suggested that at an estimated glomerular filtration rate (eGFR) >30 mL/min/1.73 m2, a metformin dose of 1,000 mg daily poses minimal risk for developing lactic acidosis, and at an eGFR >45 mL/min/1.73 m2 (chronic kidney disease [CKD] stage 3a or better), metformin 1,500 mg daily is safe (39,40).
With regard to side effects, at doses of 1,000 mg daily, metformin IR has significantly greater rates of nausea and a trend toward significantly more diarrhea. At doses ≥1,500 mg, the rates of GI side effects are similar between the IR and ER formulations (37). Another study found that patients who switched from twice-daily metformin IR to a once-daily metformin ER regimen had fewer GI side effects, equivalent metabolic benefits, and a greater preference for the ER formulation (41).
Given all of these considerations (i.e., cost, side effects, safety, and simplicity), if a patient has CKD stage 3 or better (≥30 mL/min/1.73 m2), we recommend starting metformin ER 750 mg daily with breakfast. If a patient’s eGFR is ≥45 mL/min/1.73 m2, instruct the patient and write on the prescription for the patient to increase the dose after 1 week to 1,500 mg (two tablets) daily with breakfast. However, instruct the patient that if, at 1,500 mg, the patient develops intolerable GI side effects, decrease the dose back to 750 mg daily and discuss further with their outpatient provider.
Sodium–Glucose Cotransporter 2 Inhibitors
Recommendation: start a sodium–glucose cotransporter 2 (SGLT2) inhibitor if it is covered by the patient’s insurance.
SGLT2 inhibitors do not have a profound effect on A1C lowering, with an expected reduction of approximately 0.5% (42,43). However, they all have been shown to reduce congestive heart failure (CHF) exacerbations, they have mild systolic blood pressure–lowering effects, and several agents in this class have renoprotective benefits as well. Additionally, this class of medications rarely causes hypoglycemia and encourages a potential weight loss of 3–5%, which is particularly beneficial in patients with overweight or obesity (44).
The eGFR cut-offs for initiating an SGLT2 inhibitor differ by medication and indication. In general, SGLT2 inhibitors are not dangerous at lower levels of renal function, but their glycosuric (and therefore glucose-lowering) effects become negligible (44). However, the renoprotective and/or cardioprotective benefits of canagliflozin, dapagliflozin, and empagliflozin remain at an eGFR ≤30 mL/min/1.73 m2 (45–48). For this reason, U.S. Food and Drug Administration (FDA)-approved labeling allows for certain medications in this class to continue even if a patient’s renal function worsens to CKD stage 4 or stage 5 (≤30 mL/min/1.73 m2), as long as the patient is not on hemodialysis (49).
Therefore, if a patient needs additional A1C lowering beyond 1% (meaning that metformin alone will be unlikely to get the patient’s A1C to goal), or if the patient has another indication such as CKD, CHF, or atherosclerotic cardiovascular disease, it is recommended to strongly consider also starting an SGLT2 inhibitor on discharge (25).
If cost remains an issue even after insurance coverage, one option is to prescribe the higher dose (e.g., dapagliflozin 10 mg or empagliflozin 25 mg) and instruct the patient to take only half of a pill each time. This half dose should still retain considerable cardioprotective, renoprotective, and weight loss benefits, with only a slight decrease in glucose-lowering potency (48,50,51).
Glucagon-Like Peptide 1 Receptor Agonists
Recommendation: start a glucagon-like peptide 1 (GLP-1) receptor agonist if it is covered by the patient’s insurance.
GLP-1 receptor agonists, like SGLT2 inhibitors, have generated considerable excitement within the diabetes community for their glucose-lowering ability, lack of hypoglycemia, considerable weight loss effects, and potential cardio- and renoprotective benefits (52). However, not all GLP-1 receptor agonists are equivalent. Importantly, only dulaglutide, injectable semaglutide, and liraglutide have been shown to reduce major acute cardiovascular events (MACE). Other drugs in this class (oral semaglutide, exenatide ER, exenatide, and lixisenatide) did not reduce MACE in large clinical trials (53,54). Tirzepatide, a once-weekly GLP-1/gastric inhibitory polypeptide receptor agonist, has had impressive glucose-lowering and weight loss benefits, but whether it confers cardiovascular and/or renoprotective benefits is still unclear (55,56).
With regard to ease of use, oral semaglutide is the only oral GLP-1 receptor agonist formulation. However, patients must take the medication on an empty stomach while drinking no more than 4 oz (120 mL) of plain water and must not take any other medications, food, or drink for at least 30 minutes (57). Such restrictions may prove challenging to some patients. Injectable semaglutide, dulaglutide, and exenatide ER come as once-weekly formulations, whereas liraglutide and lixisenatide must be taken daily, and exenatide must be taken twice daily. Studies have shown that the once-weekly formulations have improved adherence compared with daily GLP-1 receptor agonists (58,59).
Dulaglutide and exenatide ER pens come as fixed-dose pens that are discarded after a single use. Conversely, semaglutide and liraglutide are multidose pens that must be dialed to the correct dose, similar to an insulin pen. There are pros and cons to each design. Some patients find the design of the dulaglutide pens very easy to use, with minimal steps and an attached pen needle, which can be particularly beneficial for those with poor eyesight, needle phobia, or dexterity issues (60). However, for the multidose pens, if a patient develops significant GI side effects even at the lowest dose, providers may instruct patients to dial a “microdose” by counting clicks (e.g., 8–9 clicks is ∼0.125 mg, or half the typical starting dose of injectable semaglutide) (61). This “microdosing” of GLP-1 receptor agonists is not approved by the FDA but offers the potential for slower titration and reduced risk of GI side effects (62,63).
Some providers are hesitant to start a GLP-1 receptor agonist on discharge for fear of causing significant GI upset. Although these adverse effects may be common among patients, they are often transient. Moreover, severe side effects are rare, and establishing expectations on discharge is vital. Patients must be educated that certain foods such as large and/or high-fat meals can trigger worsening nausea or vomiting. Eating smaller and leaner meals (e.g., poultry or fish) and gradually increasing fiber intake may help to reduce the risk of untoward GI side effects (64).
If a patient’s A1C is >9%, we strongly recommend adding a GLP-1 receptor agonist on discharge, provided that no contraindications exist and the cost after insurance coverage is not prohibitive. Studies have shown that GLP-1 receptor agonists prescribed at discharge have significantly greater A1C-lowering effects, weight loss, and less hypoglycemia than basal insulin (65). Prescribing once-weekly injectable semaglutide or dulaglutide is preferred because patients on this dosing regimen have demonstrated the greatest adherence compared with patients on other GLP-1 receptor agonists. Also, these drugs have demonstrated cardio- and renoprotective benefits in large clinical trials and are safe to use for any degree of CKD (66,67).
Basal Insulin
Recommendation: if needed, start once-daily basal insulin or twice-daily premixed insulin.
A combination of metformin, an SGLT2 inhibitor, and a GLP-1 receptor agonist can likely decrease A1C by ∼3% (37,49,68). Therefore, if a patient’s A1C is >3% higher than target, adding insulin may be necessary, especially if the patient required considerable basal insulin during the hospital stay. Some providers at this point would resort to starting the patient simply on basal-bolus insulin therapy and not bother with noninsulin antihyperglycemic medications. However, this strategy misses the vital nonglycemic benefits of metformin, SGLT2 inhibitors, and GLP-1 receptor agonists, as discussed above.
Moreover, once-daily basal insulin plus a GLP-1 receptor agonist has been shown to be as effective in getting patients to glycemic targets as a basal-bolus insulin regimen (69–72). Also, the simplified strategy of using a once-weekly GLP-1 receptor agonist plus once-daily basal insulin can improve patient acceptance and treatment adherence compared with the four or more daily insulin injections required in a basal-bolus insulin strategy (73).
For simplicity and efficacy, it is preferable to prescribe once-daily basal insulin (e.g., glargine 100 or 300 units/mL, detemir, or degludec) in a pen formulation that is covered by the patient’s insurance. Affordability is a major issue with insulins, with the cheapest out-of-pocket basal insulin pen costing just under $100 for a box of five pens (Semglee [glargine-yfgn]) (74). However, if the provider chooses the wrong insulin (i.e., one that is not covered by insurance or falls into the Medicare drug plan coverage gap or “donut hole”), the cost could be as high as $900 or more (75). If cost is an issue, another viable option is to prescribe Walmart’s ReliOn brand of premixed 70/30 NPH/regular insulin pens twice daily, which at the time of this writing cost $42.88 for a box of five pens (75).
What If Cost Is the Greatest Barrier to Self-Care?
An estimated 30 million people in the United States do not have medical insurance (76), and even for those who do, high deductibles and copays may still put many of the medications discussed above out of reach (77). For these patients, following the same general strategy above—one noninsulin medication for each 1% A1C is above target—can still be followed by using a combination of generic drugs and prescription discount cards to minimize costs.
First, begin with metformin, as discussed above. Both metformin IR and metformin ER are covered by most insurances with minimal out-of-pocket costs. For patients without medical coverage, a 1-month supply of metformin IR 2,000 mg (four 500-mg tablets) daily commonly ranges from $5 to $24, whereas a monthly supply of metformin ER 1,500 mg (two 750-mg tablets) daily is slightly more expensive, typically ranging from $11 to $30 depending on the pharmacy (78,79). For this reason, some providers prefer starting patients on metformin IR 500 mg daily for 1 week and then increasing the dose by one tablet per week as tolerated to a maximum dose of 1,000 mg twice daily.
Second, providers should consider adding the thiazolidinedione pioglitazone. Pioglitazone is dosed once daily, has negligible hypoglycemia risk, and has purported significant cardiovascular, renal, and β-cell protective effects (80,81). However, side effects (namely, water retention, weight gain, and bone loss) must be considered when deciding whether to prescribe this drug (82,83). In particular, this medication is contraindicated in patients with New York Heart Association class III or IV CHF, and clinicians should consider avoiding its use in any patient with CHF. Finally, there have been conflicting studies regarding the possibility of a slightly increased risk of bladder cancer in individuals taking pioglitazone (84–86).
After metformin ER and pioglitazone, providers can also consider adding a sulfonylurea such as glimepiride, glipizide, or glipizide ER. Sulfonylureas unfortunately carry a significant risk of hypoglycemia, with 2–3% of patients experiencing severe hypoglycemia requiring medical assistance. Therefore, they should be prescribed judiciously in the elderly or in individuals with advanced CKD (87,88).
If a patient’s A1C is >10–12%, insulin is most likely necessary to mitigate glucotoxicity and bring the patient’s glucose closer to goal (32). If the patient is uninsured or insurance copays for basal insulins are prohibitively expensive, the authors recommend prescribing ReliOn 70/30 NPH/regular insulin pens twice daily with meals, as discussed previously. Although a vial of NPH or premixed insulin can be purchased as cheaply as $25 (75), a box of pens provides 1,500 units of insulin (300 units per pen), whereas a vial only provides 1,000 units. Therefore, on a per-unit cost basis, the savings difference (35 units/dollar with pens vs. 40 units/dollar with vials) becomes negligible. Moreover, both pens and vials, once opened, must be discarded after 28 days. Because a box contains five pens, a single prescription could, in theory, last for several months depending on the daily dosage.
If premixed insulin is being initiated, metformin ER with or without pioglitazone may still be considered for their additional pleiotropic and insulin-sparing benefits. However, the authors recommend not including a sulfonylurea in the discharge regimen because of their long duration of action, which results in a high risk of hypoglycemia, especially in older adults. In some cases, patients report having significant post-lunch glycemic excursions on premixed insulin. To combat this issue, providers can consider also prescribing a short-acting insulin secretagogue (e.g., a meglitinide) to be taken just once daily with lunch (89).
II. Glucose Monitoring
Recommendation: consider prescribing a continuous glucose monitoring (CGM) system to patients with high A1C values or obesity or who are taking medications known to cause hypoglycemia.
Similar to medications, it is vital to consider simplifying glucose monitoring on discharge to reduce barriers and empower patients. Capillary blood glucose monitoring (BGM) with a traditional glucose meter is associated with significant diabetes distress, anxiety, and burnout (90). Many patients either fear needles or hate the pain that comes from repeated BGM, which limits self-care (91,92). Limited frequency of BGM, in turn, is associated with increased hospitalization rates and diabetes complications (93).
CGM may carry many advantages over traditional BGM. CGM sensors are easy to apply and much less painful than the fingersticks needed for BGM (94). Whereas fingerstick BGM only shows the glucose value at that moment in time, CGM demonstrates the direction in which the glucose level is trending, as well as the historical data. Additionally, the newer CGM systems (e.g., the FreeStyle Libre 2, FreeStyle Libre 3, and Dexcom G6 systems) allow for alarms to be set at particular glucose levels, which empowers patients to proactively prevent hypoglycemia and severe hyperglycemia (95). CGM systems have associated smartphone applications (apps), which enable patients to discreetly check glucose levels with minimal additional supplies and even share the results in real time with family members. This last feature can be particularly attractive for individuals with a history of hypoglycemia or the elderly, because other individuals can be notified of a hypoglycemic event and intervene on the patient’s behalf if needed.
Studies have shown that CGM reduces hypoglycemic events, fear of hypoglycemia, health care visits, and A1C, while simultaneously increasing time spent in the target glycemic range, patient engagement, medication adherence, treatment satisfaction, and quality of life (92,96–98). Moreover, by having continuous real-time data, patients better understand how other lifestyle choices such as food and exercise affect glycemic outcomes, which in turn may motivate patients to make healthier choices (95,99).
Consequently, many patients with suboptimal diabetes management or obesity may derive benefit from being prescribed a CGM system. As with medications on discharge, it is imperative to determine whether the patient’s insurance covers the CGM system. As of this writing, when using a prescription discount card, the cheapest out-of-pocket CGM system is the FreeStyle Libre 2 at approximately $130 per month ($65–70 per sensor) (100). Although expensive, this cost may be affordable for some patients who are willing and able to spend the extra money to forgo the anxiety and pain of BGM. Given that a box of 100 glucose meter test strips can cost the same or more (101), it is not surprising that several studies have found potential cost savings for patients in switching from BGM to CGM (102,103). Of note, the cheapest Walmart ReliOn brand glucose meter currently costs $9, and 100 ReliOn strips cost $18 (104).
If prescribing a CGM system on discharge, it is highly recommended that the patient be taught how to place and use a CGM sensor, as well as download the correct smartphone app before discharge if a smartphone is to be used in place of a reader or receiver.
III. Outpatient Education/Resources
It is unlikely for patients to master diabetes self-management after a single discussion with a provider or diabetes educator. Establishing continued care and education after discharge is an often-overlooked but essential component to encouraging patient engagement and treatment success (19). Key referrals are described below.
Outpatient Endocrinology Referral
Endocrinology specialty referral could be considered for any patient with diabetes, but specifically should be considered for those with elevated A1C, frequent hypoglycemia, or a new diagnosis of diabetes, as well as those newly starting insulin and those who have developed complications from diabetes (105,106). Unfortunately, in many areas of the United States, access to specialty diabetes care is limited, and wait times for appointments can span months (107). However, the increased adoption of telemedicine may help to bridge this care gap (108).
Outpatient Diabetes Self-Management Education
Diabetes self-management education and support (DSMES) is a service for people living with diabetes who want to learn how to cope with and manage diabetes (109). According to the ADA’s guidelines, “All people with diabetes should participate in diabetes self-management education and receive the support needed to facilitate the knowledge, decision-making, and skills mastery for diabetes self-care” (110,111).
With the increased adoption of virtual communication, as well as the availability of online materials, patients no longer are restricted to local in-person diabetes outpatient education programs (or limited by the lack thereof). For example, the ADA offers a virtual Living With Type 2 Diabetes program at no charge, and the Association of Diabetes Care & Education Specialists (ADCES) has a website that offers assistance in finding in-person and online diabetes education programs (112,113).
Diabetes Support Groups
In addition to education, patients can benefit from support groups. Diabetes can be isolating and is associated with significant depression, anxiety, and burnout. Finding safe spaces in which to share experiences, learn from others who have had similar frustrations, and obtain social support has been shown to improve glycemic outcomes and increase physical activity, knowledge, optimism, confidence, and psychosocial functioning (114,115). Both the ADA and the ADCES have diabetes support directories and other resources online (116,117).
Medical Nutrition Therapy
Food plays a large role in determining a patient’s glycemic outcomes, and nutrition therapy and education empowers patients to make healthier food choices and cooking decisions (118). Providers should refer patients with diabetes for individualized medical nutrition therapy provided by a registered dietitian who is knowledgeable in providing diabetes-specific nutrition therapy and education (110,119). Outpatient dietitian appointments can be expensive, so patients may need to contact their insurance carrier to see what services are covered by their plan.
Barriers to Excellent Discharge Care and Strategies to Overcome Them
Therapeutic inertia encompasses educational deficiencies for both clinical providers and patients, as well as health care system barriers (e.g., limited insurance coverage, the need for prior authorizations, and high costs of treatment). The ADA has spearheaded the effort to overcome therapeutic inertia and has developed a three-component model to achieve this goal: education, collaboration, and research (110).
One component necessary to overcoming therapeutic inertia is education and awareness for clinicians and the larger health care team (120). Because of their lack of experience with outpatient diabetes management, inpatient clinicians require ongoing education about the standards of care for diabetes, as well as treatment modalities—specifically pharmaceutical and technology updates. Examples of education that uses both educational and collaborative elements include diabetes-focused grand rounds; quarterly diabetes-related presentations for internal medicine teams, care management teams, patient access teams, and nursing staff; individual case reviews with interdisciplinary teams; and the development of a diabetes champion program (a 1-year diabetes care–focused education program for clinical staff) (4,121–124). Additionally, research demonstrates that empowering nonprescriber providers (i.e., pharmacists, nurses, and diabetes care and education specialists [DCESs]) results in reduced therapeutic inertia and greater achievement of glycemic goals (125).
Education for patients during hospitalization is also a key component of discharge planning. Diabetes can be confusing, complicated, and overwhelming for patients and their families. Barriers to care for people with diabetes include inadequate knowledge about treatment modalities, health literacy and numeracy deficits, social determinants of health, diabetes distress, and depression (126,127). Assessing and intervening on these issues can be achieved by different members of the care team, including DCESs, care managers, nursing staff, social workers, and prescribers (both hospitalists and endocrinologists). Topics to discuss can include the use of technology, medication administration, hypoglycemia prevention and treatment, community health resources, and continuation of education after discharge (e.g., referral for outpatient DSMES) (110). Assessing a patient’s understanding of the treatment plan and ability to implement a new plan, using the teach-back method, is essential. A discussion regarding the burden of treatment is the first step in discharge plan development and should be completed on day 1. Additionally, screening for diabetes distress and depression is important (19). It is known that treatment-related distress often results in suboptimal self-care behaviors and that reducing the burden of treatment is an effective strategy to reduce diabetes distress (128).
Collaboration is another component of overcoming therapeutic inertia and tackling existing health care system barriers to optimal diabetes care. Once the discharge prescriptions and outpatient referrals have been determined, the assignment of tasks, with a clear delineation of responsibilities, is essential. Although the specifics must be decided at each institution, dedicated unit (or discharge) pharmacists and/or care management teams can assist with prior authorizations and determining the costs of proposed treatments.
One very important strategy for discharge is to start planning for discharge on day 1 of the hospital admission. Early initiation will allow for the determination of insurance coverage and costs, completion of prior authorizations, identification of available cost-saving processes, and determination of whether an alternative treatment regimen may be necessary. Table 2 outlines the tasks, responsible team members, and actions required for the development of safe and successful discharge plans.
Task . | Team Member Responsible . | Action . |
---|---|---|
Assess burden of treatment | Bedside RN, care manager, provider | Screen for social determinants of health and diabetes distress |
Assess financial burden | Bedside RN, care manager, provider | Review home medication list and inquire about affordability |
Change discharge medications, as appropriate | Provider | Electronically send prescriptions to pharmacy |
Investigate insurance coverage, cost, and financial assistance programs | Care manager, inpatient pharmacist, provider | Confirm cost/copay; based on financial assessment above, consider whether further medication changes are needed, using low-tier formulary, providing coupons, and providing patient assistance applications |
Assess self-care knowledge | CDCES, bedside RN, provider | Provide inpatient education/DSMES and refer to continuing outpatient education/DSMES |
Assess lifestyle and regimen match | CDCES, bedside RN, provider | Review typical day, meal times, and work schedule |
Task . | Team Member Responsible . | Action . |
---|---|---|
Assess burden of treatment | Bedside RN, care manager, provider | Screen for social determinants of health and diabetes distress |
Assess financial burden | Bedside RN, care manager, provider | Review home medication list and inquire about affordability |
Change discharge medications, as appropriate | Provider | Electronically send prescriptions to pharmacy |
Investigate insurance coverage, cost, and financial assistance programs | Care manager, inpatient pharmacist, provider | Confirm cost/copay; based on financial assessment above, consider whether further medication changes are needed, using low-tier formulary, providing coupons, and providing patient assistance applications |
Assess self-care knowledge | CDCES, bedside RN, provider | Provide inpatient education/DSMES and refer to continuing outpatient education/DSMES |
Assess lifestyle and regimen match | CDCES, bedside RN, provider | Review typical day, meal times, and work schedule |
CDCES, certified diabetes care and education specialist; RN, registered nurse.
In his book The Checklist Manifesto, Atul Gawande explains how many processes, and particularly those in the health care sector, have become so complex in modern society that it is impossible to consistently remember to do them all (129). The results are frequent unintended errors and poor outcomes. However, by instituting a simple tool—the checklist—in the workflow, even very complicated tasks can be accomplished successfully, reproducibly, and with minimal mistakes. In the hospital setting, a discharge checklist is an essential foundation of the discharge plan for patients with diabetes (Table 3).
Check as Completed . | Task . |
---|---|
☐ | Confirm with patient/family/outpatient provider the home medication regimen. |
☐ | If diabetes is not well managed,* assess whether the patient would be willing to adjust the outpatient treatment regimen. |
☐ ☐ | Electronically prescribe: Medications Supplies (e.g., CGM system, glucose meter, glucose test strips, lancets, and syringes/pen needles) |
☐ | Confirm with pharmacist/care manager the cost of medications and supplies, prior authorization needed, and other medication and insurance details. |
☐ ☐ ☐ ☐ ☐ ☐ | Educate patient about: Medications (e.g., side effects and titration) Hypoglycemia prevention and treatment Insulin/GLP-1 receptor agonist injection technique BGM CGM Diet |
☐ ☐ ☐ ☐ | Make outpatient referrals as needed: Endocrinologist Diabetes education/DSMES Diabetes support group Registered dietitian |
☐ | Write discharge instructions for the patient. |
Check as Completed . | Task . |
---|---|
☐ | Confirm with patient/family/outpatient provider the home medication regimen. |
☐ | If diabetes is not well managed,* assess whether the patient would be willing to adjust the outpatient treatment regimen. |
☐ ☐ | Electronically prescribe: Medications Supplies (e.g., CGM system, glucose meter, glucose test strips, lancets, and syringes/pen needles) |
☐ | Confirm with pharmacist/care manager the cost of medications and supplies, prior authorization needed, and other medication and insurance details. |
☐ ☐ ☐ ☐ ☐ ☐ | Educate patient about: Medications (e.g., side effects and titration) Hypoglycemia prevention and treatment Insulin/GLP-1 receptor agonist injection technique BGM CGM Diet |
☐ ☐ ☐ ☐ | Make outpatient referrals as needed: Endocrinologist Diabetes education/DSMES Diabetes support group Registered dietitian |
☐ | Write discharge instructions for the patient. |
Diabetes can be considered not well managed if the patient’s A1C is significantly above goal, the patient presents with a severe exacerbation of diabetes (e.g., hypoglycemia or hyperglycemic emergency), or if the patient reports having frequent hypoglycemia at home.
A recent trend in acute care is the addition of endocrine hospitalist and inpatient diabetes management services to hospital systems. This model is solely inpatient-based and can consist of an endocrinologist plus an advanced practice provider, pharmacist, dietitian, nurses, and DCESs. Evidence shows that endocrine hospitalists and inpatient diabetes management teams can reduce readmissions and hospital lengths of stay and improve glycemic outcomes (1,17). Among the responsibilities of the inpatient diabetes management service is providing education for clinicians and nursing staff on discharge planning. The incorporation of this team into inpatient care can reduce therapeutic inertia and improve the long-term health of patients with diabetes.
Conclusion
Discharging patients with diabetes can be complicated, involves many decisions, and is fraught with numerous barriers and hurdles. With regard to prescriptions, the goal is to maximize benefits and medication adherence and minimize harm. Toward this end, providers should minimize the number of BGM checks and insulin doses required and ensure that the regimen is affordable and easy to implement for the patient. On a grander scale, executing a reproducibly successful discharge plan involves considerable forethought, planning, and education on the system level. Strategies to achieve a successful workflow include developing checklists, order sets, and template language; fostering collaboration; and clearly defining the roles of team members and specific necessary steps within the workflow.
Article Information
Duality of Interest
A.P.D. receives funding from DexCom for a study investigating glycemic effects in patients with type 1 diabetes. No other potential conflicts of interest relevant to this article were reported.
Author Contributions
All of the authors wrote, reviewed, and edited the manuscript. A.P.D. is the guarantor of this work and, as such, takes responsibility for the integrity and the accuracy of the article.