Diabetes and COVID-19: Risks, Management, and Learnings From Other National Disasters

It is unclear if PWD are more likely to contract COVID-19. PWD are considered at increased risk of infection generally, which has been extended to COVID-19 (1). Community testing for COVID-19 remains limited; hence, data predominantly come from hospitalized cohorts. Systematic reviews primarily consisting of data from China have estimated rates of 8% (95% CI 6–11%) (2), 7.87% (95% CI 6.57–9.28%) (3), and 9.7% (95% CI 6.9–12.5%) (4) of diabetes in people hospitalized with COVID-19. The percentage hospitalization appears higher in the U.S., where from February 12 to 28 March 2020, PWD accounted for 10.9% of all COVID-19 patients (similar to the proportion of the U.S. population with diabetes), 24% of hospitalized cases (non–intensive care unit [ICU]), and 32% of ICU admissions (5). A recent (preprint) U.K. study found uncomplicated diabetes to be the second most common comorbidity (19%) in patients hospitalized with COVID-19; complicated diabetes is reported in a figure only but prevalence looks to be approximately 6.5% (6).

Though issues with study quality and imprecision make it extremely difficult to quantify, current data suggest that COVID-19 is associated with worse outcomes in PWD. Reports that aggregate data from multiple centers are summarized in Table 1. There is a notable paucity of data on what moderates the relationship between diabetes and COVID-19 severity. Increased age, being male, hypertension, and cardiovascular comorbidities are associated with increased risks for COVID-19 severity (7) and are likely to be closely related to diabetes status. It is plausible that BMI (8), ethnicity (9), type of diabetes, diabetes control, and certain medications (10) all may also play a role. In two cohort studies (France and the U.K., U.K. study preprint) in PWD hospitalized with COVID-19, higher BMI was positively associated with worse COVID-19 outcomes when compared with people with BMI 25–29.9 kg/m² (11,12). Data on diabetes type and COVID-19 outcomes are only beginning to emerge, but preliminary (not yet peer reviewed) data from England (see Table 1) suggest risks may be higher in people with type 1 diabetes compared with type 2 (though risk was increased in all PWD regardless of type) (12). Data on glucose control and COVID-19 outcomes are discussed below.

There are limited data to date on the association between blood glucose control and COVID-19 outcomes. A retrospective study of 451 people with COVID-19 with diabetes and/or hyperglycemia from the U.S. reported that people with uncontrolled hyperglycemia had longer length of stay and higher mortality compared with people without diabetes or uncontrolled hyperglycemia (13). Another retrospective study of people with type 2 diabetes from China reported that well-controlled blood glucose correlated with improved outcomes in infected patients (14). Worse infection may predispose to more difficulty managing blood glucose, so the causal mechanism behind correlations between glucose control and worse COVID-19 outcomes is unclear.

Two recent U.K. studies (both preprint) reported that diabetes was independently associated with a higher risk of death that increased with higher HbA_1c (12,15). Compared with people without diabetes, one study reported that PWD with HbA_1c >7.5% (58 mmol/mol) had a higher chance of in-hospital death than those with HbA_1c <7.5% (<7.5% HR 1.50 [95% CI 1.40–1.60], ≥7.5% HR 2.36 [2.18–2.56]) (15). In a separate analysis, PWD with HbA_1c >10% (86 mmol/mol) had a higher risk of in-hospital death related to COVID-19 than those with an HbA_1c of 6.5–7% (48–53 mmol/mol) (HbA_1c >10% compared with HbA_1c 6.5–7% adjusted HR 2.19 [95% CI 1.46–3.29] for type 1 diabetes, 1.62 [95% CI 1.48–1.79] for type 2; in patients with type 2 diabetes, a significant difference was also detected when comparing HbA_1c values >7.5% [59 mmol/mol] to the reference category) (12). These data suggest that diabetes control preinfection has a role to play in COVID-19 outcomes. In contrast, a French observational study in PWD hospitalized with COVID-19 did not find an association between long-term glucose control and COVID-19 outcomes but had a smaller sample (11).

Health care services, and in some cases access to medication and supplies, have been disrupted by COVID-19. Evidence from other national emergencies shows that such disruptions can lead to worse diabetes outcomes during and after these events (16–18). Diet and physical activity are mainstays of diabetes self-management and can reduce risk of worse outcomes in PWD and those with cardiometabolic multimorbidities (19). Though yet to be addressed by the COVID-19 literature, the pandemic presents significant disruption to both: a U.S. survey of PWD found more than a third of respondents reported their diet is now less healthy, and half report exercising less (20). The current pandemic and social isolation are likely to increase rates of anxiety and depression, which may also lead to poor adherence to medications and worsening of risk factor control (21,22).

There is little evidence on how PWD can reduce their risk during the COVID-19 pandemic beyond following general infection control guidance within each country. More frequent blood glucose monitoring (in people who self-monitor) has been suggested, though it is unclear what evidence was used to make these recommendations (1).

Figure 1 summarizes specific considerations for diabetes management during national emergencies.

There is little information on the effectiveness of self-management/self-education during the pandemic. A variety of online services have been implemented but have yet to be assessed (see EXPERIENCES FROM COUNTRIES WITH HIGH RATES OF COVID-19). The wider literature on care of long-term conditions during national emergencies suggests a role for educational materials (23). Evidence for interventions aiming to optimize self-management in PWD that are potentially feasible in disaster contexts include mobile phone apps (24), web- or computer-based interventions (25), text messaging (26,27), and self-monitoring of blood glucose (28,29). The latter two show the most promise based on the available literature. However, the interventions tested may not be widely available or may require health care resources to be set up. In addition, choice of format should be tailored to patient preferences, which will vary by age and sociodemographic group (30).

Guidelines for routine diabetes care during the COVID-19 pandemic vary by country. Themes are summarized in Table 2. Studies of remote consultations have generally found positive results, though their generalizability to the current pandemic may be limited (31). Within diabetes, there is little evidence to support or oppose remote support (32,33).

There are overarching concerns about the impact of the COVID-19 pandemic on mental health and well-being (34,35). PWD are more prone to mental health issues than the general population (36). Adherence to treatment can worsen when people are distressed or have depression (21), as seen both during and after disasters (37,38). We found no evidence directly pertaining to management of diabetes-related anxiety or distress during this pandemic. In the general literature, there is mixed evidence on interventions to reduce diabetes-related distress, and the vast majority of interventions tested involve face-to-face contact—an unsuitable format in the current context (39).

Management of PWD with COVID-19 generally follows standard sick-day rules. Considerations that may be specific to COVID-19 are summarized below.

Most COVID-19–related consensus statements recommend stopping metformin and sodium–glucose cotransporter 2 inhibitors (SGLT2i) during acute illness and following the sick-day rules. Dipeptidyl peptidase 4 inhibitors (DPP-4i), glucagon-like peptide 1 receptor agonists (GLP-1RA), and insulin are the preferred options in particular for hospitalized patients (see Fig. 2) (40,41). There has been some discussion regarding use of ACE inhibitors and angiotensin receptor blockers (ARBs) being associated with worse outcomes in COVID-19, particularly in PWD. However, in view of data showing potential benefits, the current recommendation is to continue with these therapies (40).

A number of studies are currently testing chloroquine/hydroxychloroquine for prevention or management of COVID-19. Careful glucose monitoring will be required in PWD due to these drugs’ antidiabetic properties, with the potential risk of hypoglycemia that is associated with increased risk of cardiac arrhythmia, cardiovascular events, and mortality (42).

Guidance emphasizes the importance of referring to local specialist diabetes teams and of managing hyperglycemia (43). A narrative review suggests insulin is the preferred agent for control of hyperglycemia in this context (1). In the U.S., a major early focus for hospital management was to move inpatient care for PWD to “virtual” formats where possible, to reduce the need for personal protective equipment. This included use of electronic health records to interrogate data, telephone communication between diabetes care providers and inpatients and hospital staff, expanded “diabetes self-management protocols” allowing selected inpatients to monitor their own glucose and self-administer insulin with oversight and advice from the virtual care team, and, in some institutions, initiating inpatient continuous glucose monitoring and/or flash systems. To minimize the need for ICU beds, several institutions launched subcutaneous insulin protocols for the treatment of diabetic ketoacidosis on floors with adequate nursing staffing. Virtual diabetes care teams focused on supporting transitions to lower levels of care or outpatient settings.

We summarize here experiences from five countries that have had significant COVID-19 outbreaks.

During the outbreak in China, many PWD were unable to access health care providers (HCPs) or medical supplies due to stay-at-home orders or quarantine. Hospitals reduced the number of appointments drastically. To mitigate the impact of those changes on diabetes management, several organizations issued guidance to PWD on how to cope with the situation (44,45). The guidance developed by the Chinese Geriatric Endocrine Society mainly focused on prevention and early discovery of hyperglycemic crises and management of medications and provided detailed instructions on how to get access to certified internet-based medical services through smartphones (45). If PWD urgently needed to see an HCP in hospital, detailed guidance on how to prepare for consultations before leaving home and minimize exposure to the virus were given en route and during the hospital visit. Guidance was promoted in the form of reading material and lectures given by medical professionals through internet-based public media such as Baidu Health (an equivalent of Google in China) and the WeChat mobile app. An expert recommendation on insulin treatment of hyperglycemia in patients affected with COVID-19 was developed (46).

On 12 March, President Macron ordered most people to stay at home, especially “at-risk” groups, including PWD. Many PWD were struck by the announcement and were not expecting to be publicly identified as such, without answers to basic questions. On 18 March, a group of HCPs and researchers from the Federation of Diabetology launched a web app, CoviDIAB, to provide PWD with diabetes-specific, scientifically based information and to provide support. This free service includes access to a frequently updated media library and to live webinars with nurses, physicians, and experts. Twice a week, registrants also receive short questionnaires about COVID-19 symptoms and tests for anxiety/depression. If indicated, registrants automatically receive a notice suggesting medical contact, and phone calls may follow. At the time of writing, tens of thousands of patients were registered. Empathy was a driving principle in design, and comments suggest that the service helps people to understand and self-manage their individual risk and to limit disruption in lifestyle and care. An evaluation will indicate to what extent it reduces the burden of disease. Recommendations have been formulated by the French-speaking Diabetes Society (Société Francophone du Diabète) to promote the intensification of glycemic control by remote support by professionals and frequent self-monitoring. However, extremely limited evidence was available to guide both patients and HCPs. In response, a group of physicians and researchers from 50 French hospitals designed and launched a large register to collect data on PWD hospitalized with COVID-19 and their prognosis (COVID-19 and Diabetes Outcomes [CORONADO], NCT04324736, ClinicalTrials.gov). Recruitment has exceeded expectation; early results are now available (11).

The first cases COVID-19 in Italy were recorded early in February. Tragically, Italy has suffered among the highest numbers of deaths in the world (47). At the time of this writing, the death rate in Italy was estimated at 122.52 per 1,000 infections, i.e., a 10-fold higher rate than in Germany (14.14 per 1,000 infections) (48). This could be due to demographic features of the Italian population, which include a large proportion (23%) of people aged ≥65 years (49). From 9 March, social confinement rules were issued including restricted access to outpatient clinics. Access to diabetes clinics has been limited to urgent cases, and remote contacts via telephone or video teleconsultation have been officially approved and reimbursed by the National Health System. Prescriptions of glucose-lowering agents requiring specialist approval (DPP-4i, SGLT2i, GLP-1RA, and new basal insulin analogs) were automatically renewed until the end of May. Special joint recommendations have been promptly issued by the Italian Society of Diabetes (SID), the Association of Italian Diabetologists (AMD), the Italian Society of Endocrinology (SIE), and the Italian Society of Pediatric Endocrinology (SIEDP) (50). The same organizations have opened a Facebook page entitled “One hour with AMD, SID, and SIEDP” allowing PWD and their relatives to contact specialists (51). A government web page providing simple and pragmatic recommendations has been created focusing on disruptions to physical activity and diet as a result of social confinement (52). At the time of writing, discussions were ongoing regarding restarting clinical activities and how to organize post-COVID diabetes clinics.

Over 90% of PWD in the U.K. are managed in primary care, by enhanced diabetes‐skilled primary care physicians, nurses, and health care assistants. These models have shown to be effective in reducing hospitalizations, outpatient attendance, and admissions for diabetes‐related complications (53). In March, the government issued guidance on social distancing and self-isolation (54). In England, people considered especially clinically vulnerable (including some but not all PWD) were contacted by the government with advice on shielding. All people at high risk, including PWD, were advised to only leave the house for limited periods for essential shopping and one form of exercise (55). The Primary Care Diabetes Society and the Association of British Clinical Diabetologists have issued guidance on managing PWD in primary care (56). The national charity Diabetes UK has been active in giving advice to PWD through their website and social media. Both primary and secondary care are providing emergency and routine follow-up using telephone or video consultations including support for mental well-being, though there are some regional variations. Social media channels have been set up where HCPs are sharing experience in managing people in the community and hospital and exchanging new guidance. A number of self-management education programs have been made freely available. There are also discussions regarding longer-term plans to phase in face-to-face consultations for routine chronic disease management.

The lack of universal health coverage poses additional challenges to PWD and their care providers during the COVID-19 pandemic in the U.S. There has been an almost wholesale switch to virtual care for outpatient appointments, but there are many failings in this approach. Phone visits provide a much lower rate of reimbursement than video visits, but many of the most vulnerable patients have inadequate equipment or connectivity to support video visits. Contacts with certified diabetes educators by phone or video are not reimbursed. Additional efforts have been made to communicate with PWD at home to ensure that they are safe. Laboratory and physical exam monitoring of complications has virtually ceased for most patients as long as they symptomatically remain well. Many patients have lost insurance coverage. Prescription fills for diabetes medications are down 10%; a survey of PWD found one in six respondents needing insulin experienced a problem. A similar proportion reported issues with obtaining test strips, and a quarter of respondents reported issues with obtaining pump or continuous glucose monitoring supplies (20). Many pharmaceutical companies have stepped in to increase access to otherwise unaffordable medications. In data available through the end of March, relatively early in the U.S. COVID-19 course, approximately 80% of clinicians reported serious strain and nearly two-thirds were uncertain whether they would be able to keep their practices open due to insufficient financial resources and low volume of reimbursable work. Overall visits (virtual and face-to-face) for chronic asymptomatic care were down ∼50% (57). As the future course of the epidemic in the U.S. is uncertain and seems likely to persist for many weeks, the threat to PWD is grave. The hope is that new government initiatives and innovation on the part of stakeholders will fill the emerging cracks in an already fragmented system.

Evidence on the management of long-term conditions during national emergencies suggests various ways to mitigate the risks presented by these events, which predominantly fall under two phases: planning and response (23). These strategies are outlined in Table 3.

There is much uncertainty as to how the COVID-19 pandemic will end and what will be left in its wake. Disruptions that arise due to national emergencies can lead to increased HbA_1c in those affected up to 16 months later, with some evidence that this is particularly the case for people of lower socioeconomic status and those treated with insulin (18,23). A lack of access to routine health care is a leading cause of morbidity and mortality after disasters; stroke, acute myocardial infarctions, and diabetes complications are all shown to increase after the immediate threat has dissipated (26,58). Services such as diabetes clinics may also rethink their organization to minimize risk of ongoing transmission.

The need for decisive action creates an important tension when evidence is limited. An example here is the classification of PWD as being at increased risk from COVID-19 and therefore subject to increased preventive measures. Though risk is clearly increased, quantification is scant. There is little to no evidence on potentially moderating factors, despite the fact that these data are routinely collected in data sets used for existing analyses; the results presented are often unadjusted and use single disease categories, ignoring potential differences between type 1 and type 2 diabetes and multimorbidities, which are associated with worse outcomes (59).

In the face of a limited evidence base relating directly to COVID-19, decisions can be informed by international experiences to date and, to some extent, from the literature as it relates to other national emergencies. This latter source of evidence shows the toll of disruptions to diabetes care is often most pronounced after the acute phase of the emergency or disaster has passed. In some cases, the excess morbidity and mortality in the aftermath of national emergencies is higher than the toll during the emergency itself. History issues a stark warning here when considering the balance between diverting resources toward the acute COVID-19 crisis and maintaining routine care for people living with long-term conditions.

Finally, in reviewing what has been written on the topic of diabetes and COVID-19, we have been struck by two noticeable absences. The first is the absence of literature on wider contextual factors. PWD are likely to be impacted by COVID-19 just as much outside the health care setting as within it, with particular concerns relating to disruptions to diet and physical activity, increased stress, and burdens on mental health and well-being, yet the literature to date focuses almost exclusively on clinical management. The other unspoken issue in the literature we reviewed is that of inequality. COVID-19 is not an equal-opportunity disease. The burden will disproportionally be borne by people from less-advantaged groups (60). Emerging data also suggest that COVID-19 may pose more of a risk to nonwhite ethnic groups (9). Diabetes discriminates in similar ways, and the intersection of diabetes and COVID-19 creates a maelstrom in which existing health disparities risk exacerbation with profound and long-lasting consequences. COVID-19 holds a mirror to our health care systems and care of PWD; may we do all we can now to make that reflection favorable in hindsight.

Acknowledgments. Parts of this article are based on Rapid Reviews conducted for the Centre for Evidence-Based Medicine’s COVID-19 Evidence Service (23,61,62).

Funding. This work was not funded. E.M. and C.G. are funded by Wellcome Trust fellowships. J.B.B. is supported by grants from the National Institutes of Health (UL1TR002489 and P30DK124723). K.K. acknowledges support from the National Institute for Health Research (NIHR) Applied Research Collaboration East Midland (NIHR ARC-EM) and the NIHR Leicester Biomedical Research Centre.

The views expressed are those of the authors and not those of the universities, the NIHR, or the U.K. Department of Health and Social Care.

Duality of Interest. J.B.B. reports nonfinancial support and other from Adocia, AstraZeneca, Dance Biopharm, Dexcom, Eli Lilly, Fractyl, GI Dynamics, Intarcia Therapeutics, Lexicon, MannKind, Metavention, NovaTarg, Novo Nordisk, Orexigen, PhaseBio, Sanofi, Senseonics, vTv Therapeutics, and Zafgen; grants and nonfinancial support from AstraZeneca, Eli Lilly, Intarcia Therapeutics, Johnson & Johnson, Lexicon, Medtronic, NovaTarg, Novo Nordisk, Sanofi, Theracos, Tolerion, and vTv Therapeutics; personal fees from Cirius Therapeutics Inc. and CSL Behring; and personal fees and other from Mellitus Health, Pendulum Therapeutics, PhaseBio, and Stability Health, outside the submitted work. R.R. reports grants, personal fees, and nonfinancial support from Sanofi, personal fees and nonfinancial support from Merck Sharp & Dohme, personal fees from AstraZeneca, grants and personal fees from Novo Nordisk, personal fees from Janssen, personal fees from Eli Lilly, personal fees from Abbott, personal fees from Medtronic, grants and personal fees from Diabnext, and personal fees from Mundipharma, outside the submitted work. K.K. has received grants from Bohringer Ingelheim, AstraZeneca, Novartis, Novo Nordisk, Sanofi, Eli Lilly, Merck Sharp & Dohme, and Servier, outside the submitted work. K.K. has served as a consultant, speaker, or an advisory board member for AstraZeneca, Bayer, Napp Pharmaceuticals, Eli Lilly, Merck Sharp & Dohme, Novartis, Novo Nordisk, Roche, Berlin-Chemie AG/Menarini Group, Sanofi, Servier, and Boehringer Ingelheim. No other potential conflicts of interest relevant to this article were reported.

Author Contributions. J.H.-B., E.M., C.G., D.N., K.M., and K.K. were involved in concept and design. J.H.-B., E.M., C.G., J.K., and J.P. reviewed studies for inclusion and extracted relevant data. J.B.B., S.D.P., L.J., R.R., and K.K. contributed experiences from individual countries. J.H.-B. drafted the remainder of the article. All authors edited and approved the final manuscript. J.H.-B. had full access to all data and final responsibility for the decision to submit for publication.