Endo Fellows 2018

March 13-15, 2018; Chicago, IL; Day #3 Highlights – Draft

Executive Highlights

  • Thursday was the third and final day of an outstanding Endo Fellows conference, hosted this year in Chicago, IL. This report details nine day #3 highlights spanning technology, therapy, and big picture topics.
  • Diabetes technology: UW’s Dr. Irl Hirsch emphasized the importance of downloading data, advocating that the download be considered part of the vital signs for all insulin users. He identified a lack of infrastructure in the clinic as one the two main reasons preventing endocrinologists from downloading data, highlighting his hope that eventually all diabetes technology will facilitate passive download to the cloud. Dr. Hirsch also detailed the new consensus on using coefficient of variation (CV) as the metric of choice for glycemic variability, and he elaborated on using “fake carbs” with the 670G. We also sat in on a round table discussion led by Dr. Craig Taplin (Seattle Children’s Hospital) on CGM use in those under six-years old. We loved seeing such enthusiastic participation from the fellows, who shared their perspective on a range of topics, including technology intake, quality of life, and trend arrows for dosing.
  • Diabetes therapy: In back-to-back talks, Drs. Robert Eckel and Janet McGill addressed macrovascular and microvascular disease in type 1 diabetes. Dr. Eckel encouraged the endo fellows to more aggressively treat all CV risk factors in their type 1 patients, and he named A1c as one of these risk factors, dispelling the misconception (from the original DCCT) that tight glucose control has nothing to do with CV morbidity/mortality – in other words, good CV health is highly associated with great diabetes management. Dr. McGill reviewed best practices for treating nephropathy, retinopathy, and neuropathy, suggesting that the latter of these three is perhaps the highest unmet need, with no available drugs today (therapies target the pain, but they don’t address neuropathy itself). This has been the case forever, and it’s because “complications trials” are expensive given the long time it takes to prove that long-term outcomes can be changed. Dr. Roy Beck stressed the importance of reading clinical trial research with a critical eye, and his talk offered key lessons in statistics relevant to diabetes studies.
  • Big picture: Four highlights in this section of our report cover under-utilization of DSME and Medical Nutrition Therapy benefits from Medicare (this was a fascinating breakout session with CDE Ms. Alison Evert), the high prevalence of depression in the type 1 population, 10 pieces of advice for endocrine fellows transitioning to clinical practice from Dr. Luisa Duran (who trained with Dr. Hirsch in Seattle), and consensus on exercise in type 1 (in a breakout session with Seattle Children’s Hospital’s Dr. Craig Taplin).

Hello again from Chicago! Thursday marked the third and last day of an outstanding Endo Fellows conference, and this report details nine highlights spanning technology, therapy, and big picture topics. This gathering took place at the historic Palmer House Hilton (the oldest hotel in Chicago, and widely considered the oldest continually-operating hotel in the US!), where Frank Sinatra, Louis Armstrong, Ella Fitzgerald, and Judy Garland have all performed. For the past few days, the hotel has hosted stars in the diabetes world (equally high-profile, in our humble opinion : >), including Drs. Roy Beck, Robert Eckel, Irl Hirsch, Elizabeth Seaquist, and so many others.

Click here for our eight highlights from days #1-2 of Endo Fellows. You’ll find renowned author Mr. Jim Hirsch’s keynote on living with type 1 diabetes and raising a son with type 1 diabetes, Dr. Hirsch’s brilliant talk on the past and future of insulin therapy, Dr. Guillermo Umpierrez’s positive comments on SGLT-2s for type 1, and Dr. Seaquist’s crash course in hypoglycemia.

Happily, we’re not saying goodbye to Chicago just yet … the 100th (!) annual ENDO meeting starts with preconference sessions tomorrow (Friday), and you can read our preview here.

Diabetes Technology Highlights

1. Dr. Irl Hirsch on Data Downloading: a “Vital Sign”; CV <33% is Achieved By His HCL Users; 670G Download a Step Backwards, “Fake Carbs” An Issue

University of Washington’s Dr. Irl Hirsch cited downloading data as a critical component of modern-day diabetes therapy, remarking that the download should be considered part of the vital signs for all insulin users. We agree! He was pleased to see that almost everyone in the audience indicated that they consistently download data with their patients (a far cry from years past when very few people raised their hands when asked this question), although several attendees made the comment that this of course depends on whether their patients remember to bring their devices (so depressing! We think patients should be reminded that this is the minimum expected, given all the resources spent on care – a productive appointment is very challenging without this). Dr. Hirsch highlighted his hope for all diabetes technology to eventually facilitate passive download to the cloud, removing the barrier of implementing costly infrastructure in the clinic. In fact, Dr. Hirsch argued that one of the two main reasons preventing endocrinologists from not downloading data is a lack of infrastructure in the office leading to poor efficiency and a perception of not having sufficient time or resources. Dr. Hirsch’s other proposed reason was a lack of adequate fellow training, yet he acknowledged this is changing rapidly for the better, as demonstrated by the audience response. Dr. Hirsch detailed the importance of glycemic variability as a key metric in predicting hypoglycemia risk, noting that the new consensus recommends using coefficient of variation (CV) as opposed to standard deviation (SD). He mentioned that Dexcom Clarity introduced CV when it added AGP to Clarity, though we’d point out that SD still remains the focus on Clarity’s central dashboard. Dr. Hirsch cited a paper published in Diabetes Care last year (Monnier et al.) identifying a CV of 36% as a threshold to distinguish between stable and unstable glycemia. Dr. Hirsch personally prefers the slightly more aggressive benchmark of 33% - i.e., standard deviation should be less than one-third of the average, which is also much easier to remember. He highlighted that such limited variability is difficult to achieve in the pediatric population, though his patients on hybrid closed loop systems are typically achieving this goal. As he put it: “It’s been an evolution that’s happened in front of our eyes because of all these new tools.” CGM, which is “becoming the standard of care” has been particularly valuable in assessing glycemic variability, as Dr. Hirsch noted a minimum of 10-15 data points are necessary to make a variability conclusion using SMBG.

  • Much to the agreement of those present in the audience, Dr. Hirsch claimed that the new 670G CareLink download has “taken a step backwards” with the removal of specific insulin dose details. He believes the FreeStyle Libre AGP report allows for “pretty sophisticated” insights, but finds Tidepool’s display even better, as it provides specific insulin dose history (from pumps) and carb data. Dr. Hirsch enjoys using Tidepool to remotely monitor his patients (“this is the whole point of telemedicine”) although he acknowledged that he does not recommend it for everyone. 
  • Dr. Hirsch noted some of the most common patient mistakes including: (i) over- or under-calibrating the CGM; (ii) not looking at the sensor enough – a particular issue with FreeStyle Libre; and (iii) not accounting for insulin lag times, resulting in insulin stacking. He also advised providers to educate patients on the role of “fake carbs” in hybrid closed loop, as he recently found out this happened in the 670G pivotal trial. He admitted to having reservations regarding this practice of falsely adding carbs into the bolus calculator to artificially make the 670G more aggressive, though he recognized that many patients are doing it anyway – occasionally those without this knowledge will come off the 670G due to poor outcomes. Ultimately, as he put it: “We’re playing a game. We’re changing all of this stuff to the make the tools work better.” We hope Medtronic and the FDA get this message so future AID systems can be more aggressive, especially for the early adopters that will find system workarounds. Then again, we obviously do not have the whole story and we are sure there are reasons that this first 670G system is not as aggressive – some have said it may be legal and we’d love to know more.
  • In a valuable beyond A1c discussion, Dr. Hirsch informed the next generation of endocrinologists of the wide mean glucose range encompassed by each A1c reading. He reminded the audience that the older ADAG study, the results of which are used by all CGM programs to estimate A1c, was conducted in a “perfect population” and includes a very wide and unappreciated 95% confidence interval. (The more recent T1D Exchange study found the same thing.) As he has noted at many conferences before, Dr Hirsch has found that in typical diabetes management, 14-25% of A1c measurements are misleading – see diaTribe.org/beyondA1c and our Beyond A1c conference coverage from last summer for more on this. He discussed how glycation gaps can occur due to a variety of relatively common factors, including iron deficiency, and cautioned against comparing A1c between individuals. Dr. Hirsch stressed CGM as critical for understanding the meaning behind A1c values for individuals yet was careful to note that his arguments do not extend to use of A1c in clinical trials, as it remains a valuable metric on a population level.

2. CGM as a “Life-Saver” in <6 Years-Old Population; Remote Monitoring is Key; Quality of Life Concerns for Families?

In a round table led by Dr. Craig Taplin (Seattle Children’s Hospital), speakers discussed CGM use in those less than six-years old, technology uptake, quality of life, and trend arrows for dosing. Participating fellows indicated substantial variability regarding the proportion of their patients kindergarten-age or under on CGM, spanning 20%-80%. Most noted that they mention CGM during the first visit, with Dr. Taplin emphasizing the importance of being proactive – as he pointed out, the majority of parents will arrive at the clinic with questions regarding technology gleaned from the internet. He commented that CGM is more of a “life-saver” than pumps in this very young age group, for both the patients and their parents. In fact, he noted that this generation of parents is the most fearful to date. The potential for remote monitoring can be incredibly liberating, giving parents more peace of mind. As one fellow commented, parents are much more likely to let their children participate in sports or other activities when they have the comfort of being able to track their glucose readings in real time. Indeed, one fellow cited the lack of a share feature as the number one reason parents decide not to use Medtronic for their young children with diabetes. (Medtronic has still not given firm timing on when a Bluetooth-enabled MiniMed 670G is coming. With Guardian Connect just FDA approved, however, we imagine much of groundwork has been laid.) Still, Dr. Taplin cautioned against pushing CGM universally, underscoring that for the “standard situation” CGM could cause more problems when implemented immediately in this age group. He ended the discussion on a somewhat sobering question, asking: “Have we made the lives of families worse with CGM?” While he appreciated the massive benefits associated with CGM, especially for parents previously unable to sleep through the night, he suggested that a patient’s quality of life can “take a dive.” He referenced T1D Exchange data showing that in the <6 years-old age range, it’s “really common” to come off CGM. (We wonder if this is parent-driven or adhesive/on-body real estate driven?) To this end, he highlighted the importance of identifying the family’s goals for CGM and carefully discussing pros and cons. We’d also recommend many of the tips in the introduction and Mindset chapters of Bright Spots & Landmines, which can help reframe CGM data. Presumably automated insulin delivery, once approved in pediatrics, will help a lot with the quality of life issues – driving fewer CGM alarms, especially overnight.

  • The fellows were very interested to hear that Dr. Taplin teaches some parents how to calculate predictive glucose based off CGM trend arrows. It can be tricky, as the arrows are different for every system, but Dr. Taplin emphasized that “we have to start thinking about dosing off arrows.” While Dr. Taplin appreciated that such rigorous monitoring is “way too much” for most parents, it can considerably contribute to the ultimate goal of CGM in this young population: preventing hypoglycemia in the developing brain. See our coverage of the Endocrine Society’s latest recommendations for Dexcom’s G5

Diabetes Therapy Highlights

1. Dr. Eckel on CV Disease in Type 1 Diabetes: A1c-Lowering is Important Based on 30-Year DCCT Follow-Up; Mixed Results on Statins

University of Colorado’s Dr. Robert Eckel encouraged fellows to treat CV risk factors (including A1c) aggressively in their type 1 patients, despite some key differences he highlighted between CV disease in type 1 vs. type 2 diabetes. Although macrovascular events were not significantly reduced by intensive glucose-lowering in the original DCCT, Dr. Eckel described the legacy effect of good glycemic control by showing fewer CV events for participants in the intensive control arm over longer-term, 30-year follow-up. He dispelled the misconception that glycemia has nothing to do with CV morbidity/mortality, writing on his conclusion slide “improved glycemic control almost certainly reduces atherosclerotic CV disease and all-cause mortality.” We were reminded of a presentation by Dr. Lawrence Leiter at ESC 2017 all about why cardiologists, too, should care about glycemia – because it impacts onset and progression of CV disease. We think this is an important point to make and to reiterate to anyone who treats diabetes, since misinterpretations from the original DCCT do persist. Dr. Eckel urged providers in the room to be conscious of the power of glucose-lowering while also managing other CV risk factors – lipid levels, blood pressure, kidney function (eGFR), etc. Data from the Swedish National Diabetes Register (n=24,230) demonstrates the efficacy of statins in type 1s: 40% risk reduction for CV death over 4-6 years follow-up (p<0.001), 44% risk reduction for all-cause death (p<0.001), 44% risk reduction for fatal/non-fatal stroke (p<0.001), 22% risk reduction for fatal/non-fatal MI (p=0.003). Still, Dr. Eckel suggested that the evidence for statins in type 1 is mixed overall. He pointed out that this Swedish study was observational and therefore susceptible to confounders. Indeed, after matching statin-treated patients to non-statin-treated patients 1:1, the CV efficacy effects were muted: Only the data on all-cause death remained statistically significant, with a relative risk reduction of 26% (p=0.001). Nevertheless, Dr. Eckel recommended that people with type 1 start moderate-intensity statin treatment at ≥20 years diabetes duration, if they’re older than 30 with at least one other risk factor for atherosclerotic CV disease, if they have a coronary calcium score above the 75th percentile, or if they’re older than 20 with LDL >130 mg/dl (he clarified that target LDL should be <100 mg/dl). (On a personal note, he shared that he himself has type 1 diabetes and is on a statin – “but that’s an N of 1.”) Dr. Eckel offered a similar review of blood pressure guidelines, goals, and management strategies.

  • In a fascinating discussion of atherosclerosis in type 1 vs. type 2 diabetes, Dr. Eckel explained that plaques tend to be more calcified, and maybe more stable, in type 1. Coronary artery calcification scores are often used by HCPs to determine when an adult should start statin therapy, and there could be important applications of this method in type 1 diabetes care (especially because, as Dr. Eckel outlined, the data on statin efficacy in this patient population as a whole is still somewhat mixed). Moreover, findings from the CACTI and SEARCH studies have revealed interesting lipid patterns in patients with type 1 diabetes. In CACTI, type 1s had a more favorable lipid profile overall compared to their counterparts without diabetes, including lower triglycerides, lower LDL, and higher HDL (p<0.01 for all comparisons); similar trends were observed in SEARCH a few years later. However, Dr. Eckel also alluded to drug studies in which an HDL-raising agent doesn’t actually improve atherosclerosis in type 1s. He implied that HDL may be less functional in people with type 1 diabetes, even when it presents at higher levels (“there’s something preventing HDL from doing what it normally does”). While much more research is needed into these HDL (and overall lipid) patterns, for now, Dr. Eckel suggested that clinicians focus on lifestyle interventions, rather than pharmacotherapy, to improve HDL in their type 1 diabetes practice. Drugs may not be having their intended effect; even if they generate an increase in HDL, they may not be increasing HDL functionality.
  • Bonus reading: Dr. Eckel and three of his family members shared their family’s unique journey with type 1 diabetes during an emotional panel discussion yesterday. This is also, most definitely, worth a read (we provide a summary followed by a full transcript of the conversation).

2. Dr. McGill Tackles Microvascular Complications, Highlighting Heterogeneity in Kidney Disease, Conveying Optimism on Improving Retinopathy Outcomes; Neuropathy Remains Challenging

In a practical rundown of microvascular complications, Dr. Janet McGill especially emphasized thorough, frequent screening for diabetes-related kidney disease. Optimistically, she outlined a new treatment paradigm centered around slowing the progression of DKD. Pathologically, glomerular lesions are responsible for proteinuria, while tubular lesions contribute to the progression of kidney disease, Dr. McGill explained. Despite the strong correlation between increased albuminuria and declining eGFR, she pointed out that many patients fall into stages 3 and 4 CKD without any albuminuria, and some develop to macroalbuminuria while maintaining an eGFR >60 mL/min/1.73 m2. While both of these axes (eGFR and albuminuria) are associated with increased CV risk, there is enormous heterogeneity under the “nephropathy” umbrella. As such, providers shouldn’t rely on the notion that DKD starts with albuminuria, even though that is often the case. Rather, Dr. McGill argued, HCPs need to measure serum creatinine and calculate an eGFR annually; doing so will identify DKD earlier in some patients, and will also allow for more accurate tracking of nephropathy progression. While this can be conceptually confusing, Dr. McGill reiterated that it’s very important: Some patients have a decline in GFR that just seems to be unstoppable, while others can be quite stable at the same level of albuminuria. Unfortunately, it’s hard to tell who will fall into which category, making it all the more important to monitor regularly and over time. In patients with microalbuminuria, Dr. McGill measures serum creatinine every six months. She drove home the point that it isn’t just about identifying nephropathy anymore: It’s about slowing progression. That said, the tools available are still relatively “crude” and include tight glycemic control (she emphasized keeping blood glucose <180 mg/dl) and blood pressure control (<130/80 mmHg). She also focused on avoiding tobacco use, episodes of acute kidney injury (AKI), and nephrotoxins (ibuprofen), which she sees as seriously overlooked. We’d add that GLP-1 agonist liraglutide (Novo Nordisk’s Victoza), in LEADER, gave a significant 22% reduction in renal outcomes (95% CI: 0.67-0.92), and SGLT-2 inhibitors also appear to be renal protective (e.g. 40% reduction on a composite renal endpoint in CANVAS, 95% CI: 0.47-0.77). These CVOTs were of course conducted in type 2 diabetes, but SGLT-2s are already used off-label in many type 1s (Dr. McGill even shared one patient case to this end) and could soon be approved with type 1 indications (AZ has filed Forxiga for type 1 in Europe and will file in the US in 2H18, while Sanofi/Lexicon are on track to submit SGLT-1/2 dual inhibitor sotagliflozin for type 1 to EMA and FDA by end of this month). We’re enormously excited (and curious) about the potential renal benefits of this therapy class in type 1 diabetes, not that we are confident that the right trials will ever be done.

  • Retinopathy leading to vision loss is among the biggest diabetes-related worries for patients, but Dr. McGill conveyed positivity about the impact that anti-VEGF + laser therapy can have when used optimally – and when retinopathy is detected early. To be sure, there’s good reason for concern about retinopathy. Dr. McGill cited the DCCT in explaining that retinopathy is the most sensitive indicator for the impact of glycemic control on any tissue (followed by nephropathy, then neuropathy, then microalbumunuria). She suggested that dropping a teenager’s A1c from, say, 11% to 9% should at least halve his/her risk for retinopathy (the curve really climbs above 9%) and is cause for celebration. She stressed the need to learn how to use an ophthalmoscope properly: Background retinopathy is easy to miss because blood vessels are actually ~one diopter anterior to the retinal surface where bleeds are located. Despite the potentially devastating impact of retinopathy, though, in many ways it’s the most treatable microvascular complication. Dr. McGill was optimistic about the impact of anti-VEGF injections (which actually improve vision) promptly followed by laser treatment (to improve durability), especially for macular edema. She really thinks that anti-VEGF therapies have “changed the face of retinopathy,” but pan-retinal photocoagulation remains the therapy of choice for proliferative retinopathy.
  • On neuropathy, Dr. McGill emphasized that long-term glycemic control is the best bet for prevention, given a lack of disease-modifying therapy for neuropathy. Dr. McGill highlighted a strong relationship between disease duration/age and neuropathy, showing rates north of 50% in some patient populations. Because of the significant variability in symptoms (sensory, location, functional impact), it’s important to perform a thorough history and examination, but, as with all microvascular complications, prevention is the best medicine. Unfortunately, Dr. McGill noted, studies are even conflicting on the impact of lower glucose on peripheral neuropathy risk. She repeatedly underscored the great unmet need for disease-modifying therapy to address this microvascular complication of diabetes. All current treatments for neuropathy are targeted at the comorbid pain and include gabapentin or pregabalin for pain, especially at night when pain causes sleep disturbances. Steroids or IVIg are used for rare cases of inflammatory neuropathy or demyelination. Treating depression can reduce pain or make it more tolerable.  Note that none of these agents are correcting neuropathy itself. Dr. McGill especially recommended, when using gabapentin, to double or triple the dose at night, but keep daytime doses lower.  A major goal of neuropathy treatment is to avoid opiates. But unlike anti-VEGF therapy for retinopathy, no treatment improves the underlying cause of neuropathy; B vitamins and alpha-lipoic acid may help, but more investigation is needed. The only candidate we’re aware of in this space in Daiichi Sankyo’s phase 3 mirogabalin, and we absolutely agree that this remains an area of high unmet need.

3. Dr. Roy Beck on Tips for Evaluating Scientific Literature: Confidence Intervals vs. P-Values, Identifying Confounding Factors

In a highly informative session to wrap up Endo Fellows 2018, Jaeb Center’s Dr. Roy Beck provided useful tips on critically evaluating scientific literature. For much of this final talk, he focused on the added value of confidence intervals as compared to p-values. The p-value, which defines the chance of there being no actual difference between groups (most clinical trialists look for this to be <0.05), fails to account for issues related to study design and conduct. Confidence intervals shed light on the variance, reflecting precision as well as sample size. To illustrate this point, Dr. Beck presented a theoretical example where the p-value indicated non-significance, potentially leading investigators to conclude that the therapy being studied was not effective, but the confidence interval spanned a wide range, meaning that a conclusion can’t really be made either way. In another case, the point estimate indicated superiority, but the confidence interval extended into inferiority, weakening the strength of a possible conclusion. Often, a wider confidence interval is indicative of a small sample size, Dr. Beck explained. He also detailed the difference between confounding and bias. While both can influence outcomes, confounding can be controlled for, whereas bias cannot. Biases lead to uncertainty and are based on the study design or conduct. Dr. Beck acknowledged that diabetes studies in particular are very difficult, as so many factors other than the treatment itself can influence glucose levels. Whenever possible, Dr. Beck recommended conducting blinded studies to minimize bias. To identify potential confounds as a reader of clinical research, Dr. Beck advised attendees to look at the baseline characteristics table for any meaningful differences between groups that might be associated with the study outcome; if differences exist, the next step is to determine whether investigators have taken the confounding factor into account. To adjust for confounding, investigators can perform separate analyses. As an example, Dr. Beck described a study which appeared to conclude that those with type 2 diabetes exhibited significantly greater levels of Vitamin D deficiency as compared to those with type 1. However, when the analyses were performed separately by race, it was clear that the difference was due to greater numbers of African Americans in the type 2 study population. When designing a study, confounding can be minimized by randomization, stratifying for the potential confounder, and restriction in eligibility. We’ll certainly be keeping these points in mind as we continue to digest and analyze clinical trial research (for diabetes drugs and devices alike), and we think it’s so valuable for endocrine fellows to hear this lesson as well.

Big Picture Highlights

1. Registered Dietitian/CDE Ms. Alison Evert Decries Low Rates of DSME and Nutrition Counseling: Only 5% of Newly-Diagnosed Patients on Medicare Use DSME Benefits (Which Do Exist!)

In a breakout session on nutrition, dietitian/CDE Ms. Alison Evert emphasized how woefully underutilized Diabetes Self-Management Education and Support (DSMES) and Medical Nutrition Therapy (MNT) benefits are, despite generally good reimbursement across private and public payers. If it seems like patients haven’t received diabetes education, she said, it’s because they haven’t: Medicare data show that only 5% of newly-diagnosed diabetes patients use their DSMES benefit. Only 1.7% of Medicare beneficiaries with diabetes had a claim for DSMES in 2012, though this rate was higher among those on insulin (~14%). In Ms. Evert’s population health work, her database of ~15,000 patients with diabetes across 14 PCP clinics revealed similar rates: 6% had seen a dietician in the last three years, while only 3% had received DSME. Unbelievable! Indeed, fellows in the room seemed to be generally unfamiliar with the reimbursement available on this front; we think spreading this sort of knowledge is so, so important, as most patients likely assume that this type of care isn’t covered. According to Ms. Evert, Medicare will reimburse 10 hours of DSMES in the first year of diabetes (she noted that this can actually be any year – it’s really just the first year of utilization), plus two hours in each subsequent year. Most private payers mimic this setup. For MNT, three hours are covered by Medicare in the first year, with one-two hours covered in each year thereafter. Medicaid also offers similar benefits, although Medicare and Medicaid benefit coverage for DSMES must be provided by a program accredited by either the ADA or AADE, Ms. Evert clarified. Patients can benefit in so many different ways. DSME is quite broad in scope, encompassing healthy eating, activity, monitoring, taking medications, and coping skills. In Ms. Evert’s view, too many providers forget that lifestyle management is meant to be a constant component of care at every stage of diabetes, per all the major treatment algorithms. Several times throughout Ms. Evert’s talk, she brought up the point that dieticians and CDEs have more time to spend with patients than do endocrinologists. A typical first visit with a nutritionist is an hour, and follow-up is 30 minutes; this allows the nutritionist to take a significant chunk of diabetes management off the endo’s plate, but more importantly, it affords time for a deeper dive into diet/lifestyle information with patients. This was a theme not only of this session, but of the entire meeting. During a panel yesterday evening, Dr. Robert Eckel reinforced the importance of CDEs, saying, “I insist on having a CDE come with me. If you [the endocrinologist] only have 20 minutes, the CDE may have 25 or 30.” Ms. Evert also reinforced the extremely positive impact of diabetes MNT, which she believes should come from a dietician who also has CDE certification (“more bang for your buck”). She cited RCTs, though small in comparison to pharmaceutical studies, that have associated MNT with 0.3%-1.0% reductions in A1c for type 1 diabetes and 0.5%-2.0% reductions for type 2 diabetes, which she correctly asserted is similar to the A1c drop seen with medications. As her bottom line, Ms. Evert implored the audience: “Let us help you.”

  • Ms. Evert also illustrated how patient education on nutrition is an unmet need in diabetes care. The #1 question the ADA consumer call center receives is “what can I eat?”, which makes all the more sense when you learn that <50% of people with diabetes report receiving any type of nutrition counseling or advice. Ms. Evert cited common misconceptions – white foods contain more carbs, brown rice and quinoa are low-carb, a dose of mealtime insulin should be based on meal size rather than carb content – and offered practical advice. Her tips include asking open-ended questions, counting to seven and waiting for patients to answer, and asking, “what are sources of carbohydrates?” or “do you know when your insulin peaks in action?”. The latter is one of many questions important in building an understanding of the connection between insulin and food: Many patients don’t realize that rapid-acting insulin works in the body for 4-5 hours, and Ms. Evert demonstrated how to draw this out for patients – she asserted that helping patients understand this concept reduces bolus stacking. She also likes to have people keep short-term food journals, but also – and we thought this was SO smart – asks people to take pictures of everything they eat if they don’t want to write it down, which helps her understand portion size. Notably, Ms. Evert characterized the foods one eats as one of the most variable and personal aspects among people with diabetes, making individualization of blood glucose management strategies all the more important.

2. Diabetes Psychologist Dr. Jill Weissberg-Benchell Calls for Greater Emphasis on Psychosocial Care; Points to Bright Spots in ADA’s Recent Position Statement, Person-First Language Movement, JDRF Funding for Psych Fellows

Northwestern’s Dr. Jill Weissberg-Benchell described (i) the high prevalence of depression in people with type 1 diabetes and (ii) what needs to be done about it. She set the stage for this presentation by emphasizing the sheer emotional burden that comes with diabetes: This is a condition that imposes unrelenting daily demands (i.e. multiple insulin doses, frequent blood sugar monitoring) and that requires constant decision-making (Do I need more insulin? Is it safe to exercise?). It’s easy to get overwhelmed and demoralized – especially when, in the words of Mr. Jim Hirsch, “all of your failures are public and all your successes are private.” Dr. Weissberg-Benchell went on to explain that adults with type 1 diabetes are 24%-52% more likely to develop a diagnosable depressive disorder than the general population. To this end, 12%-18% of the type 1 diabetes population has a depressive illness vs. 3%-6% of the background population. Among youth with type 1, 20% show depressive symptoms. Dr. Weissberg-Benchell further noted that elevated risk for depression can extend to family members: 22%-33% of mothers with a child with type 1 diabetes experience depressive symptoms. Type 2 diabetes is no exception, with a 60% higher risk of depression and 9%-18% prevalence of a diagnosable depressive illness (again, vs. 3%-6% in the background population). Dr. Weissberg-Benchell underscored that depression in diabetes is a major cause for concern. Beyond taking a significant toll on mental well-being and quality of life, the impact of depression and distress in people with diabetes is worsened glycemic control, decreased adherence to medication, and greater severity of complications – leading to increased medical costs, greater functional disability, and even elevated rates of premature all-cause mortality. Last year’s Endo Fellows meeting was the first to include a session specifically dedicated to the psychology of diabetes (delivered by key opinion leader Dr. Bill Polonsky), and we’re hopeful that this will become a regular part of the agenda from here forward. Indeed, this is essential training for the next generation of endocrinologists, and we hope for a future where diabetes care is more closely interfaced with psychosocial support.

  • On an optimistic note, Dr. Weissberg-Benchell highlighted two recent signs of progress in bringing psychological management to the forefront of diabetes care: (i) ADA’s new Position Statement on Psychosocial Care and (ii) a high-profile paper from ADA/AADE detailing the importance of person-first language in diabetes. The ADA Position Statement, written in partnership with the American Psychological Association (APA), marks a crucial step in defining best practices, general recommendations, and future aspirations for how psychological care should be delivered within the context of diabetes management. At a high level, it recommends regular screening for diabetes distress and depression (along with anxiety, disordered eating, and cognitive capacity), close monitoring of diabetes self-management behaviors as a window into psychological factors affecting diabetes management, careful consideration of a patient’s life circumstances, and referral to behavioral health providers – essentially integrating psychosocial care as a fundamental part of diabetes management on par with retinal, renal, and CV care. Very closely related to this is the ongoing movement for person-first, non-stigmatizing language in diabetes, which gained a mandate with the publication of ADA/AADE’s paper and its own section in ADA’s 2018 Standards of Care. The work was spearheaded by diabetes psychologist Dr. Jane Dickinson and the paper itself, published simultaneously in Diabetes Care and The Diabetes Educator, outlines a set of specific recommendations about language use in diabetes, advocating for the adoption of empowering, non-judgmental, non-stigmatizing vocabulary to motivate, educate, and foster patient/provider collaboration. For example, “adherence” and “compliance” should be replaced with “engagement,” “participation,” or “involvement.” A practical guide is available here
  • Though increased discussion on psychosocial care in diabetes is certainly welcome, Dr. Weissberg-Benchell underscored that we are still waiting on a corresponding change in real-world clinical settings. To illustrate this, she shared that even in the major metropolitan area of Chicago, she is the only psychologist with ADA-certified diabetes training within a 250-mile radius. That said, she was quick to clarify that for people dealing with depressive symptoms or diabetes-related distress, any psychologist is better than none, especially if the psychologist is open to learning more about the specific demands of diabetes management (and if the patient is willing to provide this education). Dr. Weissberg-Benchell also mentioned that JDRF is now funding post-doctoral fellowship positions to train psychologists in the unique psychological demands of type 1 diabetes management – another sign of progress, and a bridge to real-world clinical care. She touted the benefits of cognitive behavioral therapy (CBT), which (in combination with an exercise regimen) recently demonstrated a compelling ability to both resolve the symptoms of depression and lower A1c in people with diabetes in the Program ACTIVE II study. First presented at ADA 2017 and subsequently elaborated upon at AADE 2017, the study’s combined CBT/exercise intervention reduced A1c by a mean 1.3% in individuals with a starting A1c >7% (p=0.02); neither intervention alone had a significant A1c-lowering effect. Notably, Program ACTIVE II enrolled adults from diverse income levels and educational backgrounds (each of which independently affects psychosocial health) in both urban and rural areas, so the results underscore that this kind of intervention is generalizable to a wide range of patient populations in the US. That said, it’s unclear how well-reimbursed CBT is, and whether there are enough therapists to address the growing population of people with diabetes and comorbid depression.

3. Ten Pieces of Advice from a Former Endo Fellow, From How to Deal with Health System Challenges to Endos as Valuable Members of Healthcare Community

Endocrinologist Dr. Luisa Duran provided an honest and insightful reflection on her recent transition from fellowship to clinical practice. After training with none other than Dr. Irl Hirsch at the University of Washington, Dr. Duran joined John Muir Health, a community medical group in the San Francisco Bay Area, where she takes care of many people living with diabetes. She assured the audience of fellows, many of whom expressed uncertainty about the future, with 10 pieces of advice she wishes she had known about the move to clinical practice:

  • (1) Productivity matters. Dr. Duran underscored that clinical practice involves much higher patient volume than fellowship. She typically sees up to 20 patients every day and is allotted only 30-40 minutes for new patients and 15-20 minutes for follow-up appointments (regardless of whether it is a simple thyroid or osteoporosis case or a complex type 1 diabetes case). While not surprising, this was still upsetting to hear – as we continue to push for higher-quality patient/provider interactions, the health system continues to push back with constraints. Yesterday, Dr. Irl Hirsch called for a study that investigates the impact of time spent with a patient on clinical outcomes in diabetes; we would love to see this data collected, and perhaps it could help move the needle on issues like physician burnout, rushed patient appointments, etc. Dr. Duran admitted that she is “still figuring it out” when it comes to managing her high case load, and she advised fellows to use their support staff wisely and to negotiate a “ramp up” period when beginning practice in order to transition steadily to this higher volume.
  • (2) Clinical resources matter. Not all clinics have the resources for comprehensive diabetes care, and Dr. Duran advised the fellows to take a leadership role in making sure their clinic is organized in the best way possible – with appropriate patient resources for SMBG and insulin pump training, and ideally a team of CDEs, nutritionists, and pharmacists on board.
  • (3) Time management matters. Dr. Duran explained that high patient volume makes it easy to fall behind in clinic. To avoid this, she encourages her type 1 patients to pre-download their CGM and insulin pump data before arriving in the office. She suggested to fellows that they spend extra time learning the EMR system, pre-programming templates, and dot phrases ahead of time.
  • (4) Teamwork matters. Medical assistants and office assistants are a vital resource, and Dr. Duran emphasized the importance of training them intensively on various diabetes medications and technologies so they can provide support with meter downloads, injection teaching, and other tasks. This frees up more of the endocrinologist’s time for direct patient care.
  • (5) Staying up-to-date matters. As we know all too well, the diabetes therapy and technology landscape is constantly changing. Dr. Duran shared how patients will ask about new drugs and devices they see advertised on TV or hear about from friends, so it is imperative to be knowledgeable on these updates.
  • (6) Insurance companies matter. Treatment decisions are often influenced by the patient’s insurance coverage (this is a theme we hear about regularly on the conference circuit, how cost determines care). At times, Dr. Duran outlined, this can be circumvented by prescribing a similar in-formulary agent (i.e. switching between Tresiba and Toujeo or between Lantus and Levemir), but if the patient doesn’t react well to the alternative medication, or if the claim is consistently denied, the endocrinologist must call the insurance company or write a formal appeal. Dr. Duran acknowledged that going to bat for patients in this way can be frustrating, but on the flip side, can make a life-saving difference (especially, she noted, when it comes to getting coverage for CGM).
  • (7) Money matters. This is relevant from both the endocrinologist’s perspective (clinic billings can affect compensation) and the patient’s perspective (insurance coverage and formulary status have a major impact on out-of-pocket expenses). It’s critical to be mindful of small details of treatment and medical practice that could have a significant financial impact, Dr. Duran said.
  • (8) Relationships matter. “The joy of endocrinology is that you can build a long-term relationship with your patients.” We loved this sentiment. Dr. Duran emphasized the importance of building trust with patients and approaching clinical issues from an empathetic perspective that takes into consideration the myriad “life factors” beyond medication and technology that play into diabetes management (lifestyle, job situation, motivational status, family support, social stressors, personal beliefs, and the list goes on…). This evoked the person-first language movement, and hit on the critical theme of HCPs understanding the psychology of diabetes, the challenges of living with this disease every day.
  • (9) Staying connected matters. Dr. Duran advised keeping close with mentors and co-fellows for advice and support, especially early in the transition to clinical practice.
  • (10) You matter! Dr. Duran concluded her list of tips with an emphasis on endocrinologists as a vital part of the healthcare community. “PCPs depend on you as an expert, people living with type 1 diabetes need you, and their family members appreciate you.”

4. Dr. Craig Taplin on Exercise in Type 1 Diabetes – All Exercise is Not the Same

Dr. Craig Taplin (Seattle Children’s Hospital) underscored two main points for HCPs to master prior to helping type 1 patients exercise: (i) all exercise is not the same, and (ii) a sedentary patient just initiating an exercise lifestyle responds differently than a trained athlete. Pure aerobic activity (low intensity/steady state) leads to falling glucose, whereas pure anaerobic activity (high intensity) results in rising glucose. Mixed activity like team sports, which Dr. Taplin noted are most common in the pediatric population, have less predictable results. For aerobic activity, patients should be advised to plan ahead with carbs. More carbs are needed as intensity increases up to a point – data indicate that at the very highest limits of aerobic activity, less glucose is taken up, therefore requiring less carb intake prior to exercise. Although substantial variation exists between patients, Dr. Taplin noted that patients themselves tend to be fairly consistent, at least for aerobic activity. Still, as patients become more active, the rate at which they burn through glycogen stores declines, meaning that fitter individuals will actually store more glycogen in favor of burning fat, ultimately requiring changes in their management approach. Dr. Taplin recommends his patients eat a high-quality meal three-four hours before exercise, allowing a build-up of glycogen stores and adequate time for insulin to clear the system. Exercising on a peak bolus is “really challenging,” he explained. However, he cautioned against fueling before high intensity interval training (HIIT), as it tends to dramatically spike glucose. Transient hyperglycemia followed by late-onset hypoglycemia can also occur as a result of exercise, leading Dr. Taplin to advise a 25%-50% reduction in the post-exercise meal bolus followed by a 20% basal reduction during the bedtime to 3 AM time-range. We found the below guide very useful in setting basic blood glucose goals prior to exercise. Dr. Taplin also referenced a new consensus exercise decision tree published in The Lancet, from Dr. Michael Riddell and colleagues.


-- by Ann Carracher, Abigail Dove, Payal Marathe, Maeve Serino, Adam Brown, and Kelly Close