American Association of Diabetes Educators – 39th Annual Meeting

August 1-4, 2012; Indianapolis, IN; Report – Devices – Draft

Executive Highlights

Educators throughout the conference expressed an increasingly high level of enthusiasm for the benefits of CGM – higher than last year, in our (admittedly anecdotal) view. Most notably, a study of 26 people with type 2 diabetes on basal insulin presented by Ms. Jean Halford (Rocky Mountain Diabetes and Osteoporosis Center, Idaho Falls, ID) demonstrated that after six months of continuous CGM use, A1c declined by 2.4% (baseline: 9.0%) in patients who saw a diabetes educator and by 1.1% (baseline: 8.7%) in patients who did not. Impressively, three patients reduced their A1cs by 4.1% or more! Although the study was small, did not have a non-CGM control group, and no p-values were reported, the results speak volumes about the immense potential of CGM in people with type 2 diabetes, especially when supported by solid diabetes education. These findings reinforce Dr. Robert Vigersky’s 2010 study (see pg. 26 of our 2011 ADA Report – CGM Pumps and SMBG at http://bit.ly/P3qtWb). Relative to ADA 2012, AADE had much less talk of behavioral barriers to CGM use – the sessions we attended hardly discussed issues such as alarm frustration, unrealistic patient expectations, carrying multiple devices, and information overload. We also heard substantial enthusiasm for professional CGM from a number of speakers – while the value proposition is much different compared to real time CGM, we believe diagnostic CGM has a very important place in the treatment arsenal, especially as HCPs look to individualize therapy in light of the new ADA/EASD position statement.

Similar to ADA 2012, the headlines about glucose monitoring mainly concerned new products, though there was still tremendous focus on how HCPs can use all BGM data more effectively. As a result, software continued to be a major focus of Abbott’s marketing around the FreeStyle InsuLinx, and the benefits of structured testing (Roche’s STeP study) were mentioned in more sessions than we could count. And at a very fundamental level, patients just need to test, a point underscored by Dr. Steve Edelman’s (University of California San Diego, La Jolla, CA) fascinating study comparing patients’ guesses about their blood sugar to their actual values. In his words, “There’s no way you could get any device approved that’s this inaccurate.” And yet, so many are “guessing instead of testing” every day…

The standout presentations on insulin delivery were the launch of Valeritas’ V-Go and an absolutely brilliant session on pump therapy organized by highly regarded educator Gary Scheiner. Over 150 attendees sacrificed sleep to attend Valeritas’ 6:30 am product theater introducing the V-Go, a once- daily, disposable insulin delivery device that supplies basal-bolus therapy without any electronics or tubing. The device is most appropriate for patients with type 2 diabetes who require moderate doses of insulin. We valued hearing a patient’s very positive personal experience with the V-Go, including a 3% A1c reduction! Considering the barriers to insulin therapy adherence, we believe the V-Go can have an important impact on the millions of type 2s who are not optimally controlled, frequently omit insulin, or haven’t found a way to go on insulin at all. We also enjoyed a “corporate symposium” on pump therapy, which likely holds the record for most simultaneous sponsors: Animas, Asante, Insulet, Medtronic, Roche, Tandem, and Unomedical. The jam-packed audience (in the conference’s largest hall) was thoroughly listening to cutting-edge clinical insights from Ms. Jennifer Block (Stanford University, Stanford, CA), Ms. Davida Kruger (Henry Ford Health System, Detroit, MI), Ms. Alison Evert (University of Washington Medical Center, Seattle, WA), and moderator Mr. Gary Scheiner (Integrated Diabetes Services, Wynnewood, PA). It was notable to see so much enthusiasm and interest in pump therapy, a topic that we think has expanded substantially at the past few AADEs.

 

Table of Contents 

 

Devices

Breakout Sessions: Technology

THE SYNERGISTIC EFFECT OF EDUCATION AND CONTINUOUS GLUCOSE MONITORING ON GLYCEMIC CONTROL WITH OLDER T2DM USING INSULIN

Jean Halford, RD, CDE, LD (Rocky Mountain Diabetes and Osteoporosis Center, Idaho Falls, Idaho)

Ms. Halford presented a very encouraging pilot study of CGM use in 26 people with type 2 diabetes on basal insulin (including 10 patients >65 years old). Patients wore an unblinded Dexcom Seven Plus CGM for six months and had the opportunity to meet with a CDE as desired to review and revise their treatment plan at any time. Nine patients had zero CDE visits over the course of the study (the No Education group) and 17 patients had at least one visit to a CDE (the Education group). After six months of CGM use, A1c declined by 2.4% in the Education group (baseline: 9.0%) and 1.1% in the No Education group (baseline: 8.7%). Impressively, three patients reduced their A1cs by 4.1% or more (all in the Education group). Changes in medications differed depending on study group: many patients in the education group added bolus insulin, reduced basal insulin, added a TZD, and stopped sulfonylureas, while those in the No Education group tended to increase their basal insulin and make few other changes. There was no change in the incidence of hypoglycemia (excellent considering the A1c improvement), while BMI improved in the Education group and blood pressure improved in both groups. Unfortunately, the study did not have a control group and p-values were not reported during the presentation. It's also difficult to assess how much of the improvement in the Education group was due to CGM vs. educator visits. Of course, the “synergistic effect” mentioned in the talk’s title might also be a very valid explanation: the greater CGM data may have helped the educators do an even better job managing their patients. Additionally, the 1.1% improvement in the group that did not see CDEs was certainly very promising. We believe CGM has a lot of encouraging potential in the type 2 diabetes population, especially as it gets more accurate and easier to use with the Dexcom G4 and Medtronic Enlite systems. Some of the big remaining roadblocks include collecting enough data to influence payors, getting the costs down, and making the devices even more user-friendly (e.g., reducing/eliminating calibrations, etc.)

  • This study assessed the benefits of CGM and diabetes education in people with type 2 diabetes on basal insulin. It was an open-label, uncontrolled, single-center, investigator- initiated, pilot study. Investigators recruited adults over the age of 18 with type 2 diabetes managed with basal insulin with or without orals. Participants wore an unblinded Dexcom Seven Plus continuously for six months after an initial two-week blinded period. Meetings with a CDE could be scheduled by patients as desired to review and revise the treatment plan at any time (see below for frequency). After beginning the study, medication changes were acceptable, including adding bolus insulin. The primary endpoint was A1c change at six months.
  • The study included 26 individuals with type 2 diabetes with a mean duration of diabetes of 13.4 years, a mean duration of insulin use of 3.6 years, and a mean BMI of 34.6 kg/m2. Ms. Halford emphasized that twenty of the 26 participants were >55 years old and 10 of the 26 were >65 years old – she highlighted that testing CGM in an elderly population was one of study’s main aims. The Education group included 17 individuals that visited with a diabetes educator to receive diabetes self-management training and medical nutritional therapy. As noted above, patients could see the educator as many times as they wanted throughout the study (four patients had one CDE visit, three patients had two visits, three patients had three visits, four patients had four visits, and three patients had five visits). The nine patients in the No Education group did not visit with a diabetes educator during the study. Patients had a “regular physician visit” once every three months.
  • After six months of CGM use, A1c declined by a whopping 2.4% in the Education group (baseline: 9.0%) and 1.1% in the No Education group (baseline: 8.7%). Ms. Halford displayed a histogram of A1c change – half the patients in the study reduced their A1c by 1.1-1.5%. Outliers included three patients that reduced their A1c’s by >4.1% (all in the Education group) and one patient that increased his or her A1c by 2%. As seen in the table below, patients>65 years had a smaller decline in A1c than those <65 years, though the starting baseline A1c was much higher in the younger patients. The slide that presented these statistics did not include p- values.
 

Mean A1c at Baseline (Range)

Mean A1c at 6 Months (Range)

Mean Change in A1c

Education (n=17)

9.0%

(7.0% - 11.6%)

6.6%

(5.9% - 8.8%)

-2.4%

No Education (n=9)

 8.7%

(7.6% - 11.5%)

7.6%

(6.4% - 10.6%)

-1.1%

Education >65 years

8.3%

6.7%

-1.6%

Education <65 years

9.5%

6.6% -2.9%

No Education >65 years

8.3%

7.0%

-1.3%

No Education <65 years

8.9%

7.9% -1.0%
  • The most common medication changes over the course of the study were the addition of bolus insulin, insulin sensitizers, and the discontinuation of sulfonylureas in the Education group and an increase in the basal insulin dose in the No Education Group. We suspect the greater number of education visits allowed educators to finely tune meal-time doses and identify glycemic patterns, while it was easier to treat those not receiving education with an increase in basal insulin alone. Overall there were more medication changes in the Education group than the No Education Group. The number of patients experiencing a change in insulin dosing or basal/bolus split was not reported. No p- values were given for this data either, unfortunately.
 

Mean Daily Basal Insulin (Range)

Mean Daily Bolus Insulin (Range)

Insulin Sensitizers

Sulfonylurea

Education – Baseline

36.8 (10-100)*

0

   

Education – 6 Months

31.9 (8-100)

11.8

8 patients added

6 patients stopped, 2 added Prandin

No Education – Baseline

39.2 (20-92)

0

   

No Education – 6 Months

49.3 (30-100)

1 subject added 30 units per day

No Change

1 patient added, 1 stopped

  • There was no change in the incidence of hypoglycemia, while BMI improved in the Education group and blood pressure improved in both groups. While the baseline incidence of hypoglycemia was low at 0.89%, it was encouraging to see this stay consistent given the large improvements in A1c. Ms. Halford noted, “This technology allowed us to reach that primary endpoint without hypoglycemia.” No p-values were given.
 

Body Mass Index (kg/m2)

Blood Pressure

Education – Baseline

35.8

131/79

Education – Six Months

34.7

129/77

No Education – Baseline

32.4

135/75

No Education – Six Months

32.5

135/81

  • In reviewing the study’s quality of life data, Ms. Halford noted overall improvements in regimen distress and patient concerns about hypoglycemia. Regimen distress (as measured by the Diabetes Distress Scale) decreased slightly over the course of the study in both groups (3.2 to 2.4 in the Education group and 3.3 to 2.7 in the No Education group). On the Insulin Treatment Satisfaction Questionnaire, hypoglycemia was the greatest patient worry at baseline, which also improved slightly by the end of the study (3.3 to 2.9 in the Education group and 3.7 to 3.3 in the No Education group). No p-values were specified.
  • Ms. Halford concluded with interesting data from the Device Use Visual Analog Scale (DUVAS), focusing on the patients >65 years old. The investigators had hoped that they could give patients CGM and have them watch an instructional DVD, and then the patients themselves could initiate use of the device. In the older population, however, “that didn’t work very well.” She explained that older patients wanted “one-on-one, hold my hand, help me with insertion, show me how to use the receiver” type of help. The study team received multiple phone calls from patients in this group. Older patients also did not feel as comfortable on user interface measures such as ease of entering readings or events into the receiver, using menus, and using trend graphs. The lowest score received for any measure was on “ease of downloading data from the receiver into the computer.” This improved over the course of the study, and interestingly, older patients rated this higher than younger patients. Ms. Halford noted that about 50% used thedownload software, skewed towards the older patients. With the recent acquisition of SweetSpot, we hope that Dexcom will be able to make strong improvements in its software.

Questions and Answers

Ms. Virginia Valentine (University of New Mexico, Albuquerque, NM): Have the patients been seen since the study completion? Did this improvement continue?

A: No. Some patients we’ve followed up with have seen their blood glucose control drop off. In the Medicare population, they’re going back to three tests per day.

Ms. Valentine: So the concept that this would be useful for training purposes doesn’t pan out.

A: I think intermittent use is useful for the provider to help understand glucose control. But I’m not sure wearing it intermittently is the best use.

Q: Medicare will cover blinded CGM studies twice per year.

A: It’s great for us.

Q: If patients self selected how often they saw you, what frequency of contact did it take?

A: Four patients saw us once, three patients saw us twice, three saw us three times, four saw us four times, and three patients came in for five visits. There were 17 patients in the Education group. The No Education group received zero visits.

Q: Would group visits then be possible?

A: Possibly. But we had some patients that needed one-on-one attention.

Q: How often did patients download or look at the graphs?

A: We encouraged them to look at the graphs about once per week. Once we started them in the study, we set no parameters, as if it were their own personal device.

Q: Did they see a physician?

A: We had them come in every three months for a regular physician visit.

MANAGEMENT CHALLENGES SEEN WITH INSULIN PUMP THERAPY – OPPORTUNITIES TO OPTIMIZE CARE

Justine Fierman, FNP-BC, BC-ADM, MSN, CDE (St. Luke’s Center for Diabetes and Endocrinology, Scranton, PA)

Ms. Fierman offered advice on how to successfully treat patients new to pump therapy. Underlying her talk was a message of individualization, an issue that has received increasing attention over the summer and will likely continue given the recent ADA/EASD position statement (see our report at http://bit.ly/JqqW4b). Ms. Fierman emphasized the importance of understanding a patient’s entire story – especially their goals and fears surrounding pump therapy – in order to create a mutually- agreed-upon treatment plan. Interestingly, she noted that her practice does have patients changing from U-100 to U-500 insulin in their pumps, which is off label but an up-and-coming trend – we assume this was for type 2 patients and perhaps the occasional type 1 who is overweight or insulin-resistant. She also echoed some of the important messages we heard earlier at AADE from Ms. Allison Evert (University of Washington Medical Center, Seattle, WA) and Ms. Jennifer Block, (Stanford University, Stanford, CA) regarding the importance of a backup plan for pump malfunctions. Ms. Fierman stressed that providers need to make sure patients know who to call for help when there are problems and know how to recognize problems, especially what symptoms suggest diabetic ketoacidosis. Ms. Fierman also offered her take on managing and interpreting patient data, which was particularly relevant given the emphasis on health information technology and mobile health technology at this conference: “You can manage a patient without a download, but you should not manage a download without the patient.” Pump downloads, she said, are only one piece to the puzzle. Finally, she encouraged educators to “Fight for your patients,” another theme we’ve consistently heard at AADE given some of the reimbursement challenges for diabetes technology. We point out that patients should also fight for their educators and doctors – we continue to be disappointed by the poor reimbursement for CGM and pump analytics analysis.

Questions and Answers

Q: I’ve had supplies that were samples that left the office and were later recalled.

A: There are some facilities that can’t even provide patients with samples. That’s a good point.

Q: On the backup plan, I don’t want that insulin around because I’ll waste it. Does Lantus expire?

A: I try to give patients samples and a backup plan they understand. It is up to the providers to determine how they want their patient to handle it. Typically, I do all of that because I’m the one putting them on pump. I’ll try to give samples and will tell them to keep the samples in the fridge unopened. But if they have to use them, they need to replace the stock in the fridge. I also tell them when they travel to have these supplies with them. I try to prepare patients when they go on vacations.

Q: Whenever you give a manual bolus, it doesn’t appear on the report.

A: The manual boluses do appear as boluses, but depending on the download you won’t see any blood glucose measure or carbs associated with it. On downloads, especially with Medtronic downloads, they tell you how many were manual vs. how many were by the bolus wizard.

Q: I am thinking about the summertime with insulin reservoirs and all my outdoor men with type 1. What about the life of insulin in extreme heat?

A: It’s a hard thing. Often they may need to put less insulin in and change sets more frequently. There’s also the Frio, but the Frio packs are hard to wear. I don’t have experience with patients wearing them. But what I’ve had patients do is put less insulin in and change every day.

Q: Are there any guidelines to go by?

A: It would be patient specific. Watch their blood sugars. Everyone is in a different environment with different types of pumps.

 

Breakout Sessions: Disease Management

EFFECTIVE AND SAFE INSULIN THERAPY WITH A NEW PATIENT-FRIENDLY TOOL

Israel Hodish, MD, PhD (University of Michigan, Ann Arbor, MI)

Dr. Hodish argued that frequent insulin dosage adjustments are necessary to improve patients’ ability to achieve and maintain their target A1c. He presented data indicating that new software being developed by Hygieia, the Diabetes Insulin Guidance System (DIGS), may be a safe and effective manner of automating the dose adjustment process. Dr. Hodish argued that a key reason that two-thirds of US patients taking insulin do not achieve a 7% A1c target is that their insulin dosage is not adjusted frequently enough. Additionally, he provided clinical data from a retrospective study demonstrating that frequent insulin dosage adjustments are required to not only achieve target A1c, but also to maintain target A1c. Furthermore, daily glucose data is enough to inform weekly insulin dosage adjustments, whether by a HCP or a software program, and subsequently lower average A1c. In a study evaluating the DIGS software, which automatically advises patients on weekly insulin dose adjustments based solely on blood glucose readings (Bergenstal et al., Diabetes Technology & Therapeutics 2012) only three out of 1,734 recommended dosage changes made by the software were not approved by the study team, and patients’ A1c declined from 8.4% to 7.9% (n=46, p=0.01). This is impressive early data, and we look forward to the larger phase 2 study in which a physician will not double-check insulin adjustments. He concluded that this tool might empower patients to make their insulin therapy more effective between appointments and relieve much of the burden on HCPs.

  • Dr. Hodish argued that a key reason that two-thirds of US patients taking insulin do not achieve a 7% A1c target is that their insulin dosage is not adjusted frequently enough. The average A1c of insulin-treated patients in the US is 7.9%, with only 35% of these patients at an A1c under 7%. In contrast, the average A1c of patients in clinical studies is 7%, with two-thirds of patients reaching the 7% goal. The key difference between the real world and clinical studies, he argued, is that patients in clinical trials are closely monitored and receive treatment adjustments far more frequently (on the order of every week or few weeks) than patients who are not in clinical trials (who typically see a provider every few months). He concluded that insulin therapy is far more effective when adjusted every few weeks.
  • Frequent insulin dosage adjustments are required to not only achieve target A1c, but also to maintain target A1c. Dr. Hodish’s team performed a retrospective study that reanalyzed 2,380 insulin dosage adjustment episodes from a previously published study of a 16- week trial that targeted <7% A1c. In this retrospective analysis, 26 patients >60 years old with type 2 diabetes that had average baseline A1c of 7.8% were selected for reanalysis (Rosenthal et al., Journal of Diabetes and its Complications 2011). He found that the magnitude and frequency of insulin dosage adjustments during the induction period (the four-week period at the start of the trial when patients brought down their A1c to <7%) were similar to the magnitude and frequency of dosage adjustments required during the 12-week maintenance period. The 23 participants that achieved the 7% A1c target after the induction period averaged two to three weeks between physician contacts during both periods, and averaged 2.4 modifications in insulin dosages per visit during the induction period and 2.0 modifications per visit during the maintenance period. Dr. Hodish concluded that frequent adjustments must continue in order to maintain a 7% A1c target.
  • Glucose data alone is enough to inform weekly insulin dosage adjustments. In another study (Bashan & Hodish, Journal of Diabetes Complications 2012), Dr. Hodish’s team performed a prospective study of 14 patients with type 1 and type 2 diabetes that had an average A1c of 9.8%. Patients were treated with basal/bolus therapy over 12 weeks, and their dosages were adjusted every week after Dr. Hodish reviewed only their logbooks from a remote location. Weekly mean blood glucose declined from 220.3 mg/dl to 151.5 mg/dl (p<0.0001), mean A1c declined from 9.8% to 7.9% (p=0.001), and the frequency of hypoglycemia was low at about 0.5 minor hypoglycemic events per week. Furthermore, when these data were fed to a software program to eliminate human bias, the recommendations were clinically similar 99% of the time to those made by Dr. Hodish. About 65% of the time the recommendations were identical, about 30% of the time they were within 10%, and they were never more than 20% different.
  • Dr. Hodish presented data from a study evaluating the Diabetes Insulin Guidance System (DIGS) software being developed by Hygieia, which automatically advises patients on insulin dose adjustments based on blood glucose readings, concluding that this tool may empower patients to make their insulin therapy more effective between appointments (Bergenstal et al., Diabetes Technology & Therapeutics 2012). In this 16-week study of 46 patients with type 1 and type 2 diabetes, data from patient logbooks were fed to the DIGS software, and the program made recommendations for dosage changes. Contingent upon approval by the study team, the dosage change was relayed to the patient. Only three out of 1,734 recommended dosage changes made by the software were not approved by the study team due to a bug in the software that has since been corrected. Average A1c declined from 8.4% to 7.9% (p=0.01). The next step will be to conduct a study in which the software makes adjustments without doctor supervision. Dr. Hodish noted that the software will be available soon in the UK and expects to report data from its usage there in one year. If proven safe and effective, this device could potentially relieve much of the burden on HCPs in diabetes care. For more detail on this study, please see our Keystone Day #2 highlights at http://www.closeconcerns.com/knowledgebase/r/3d55b14c.

Questions and Answers

Q: Have you correlated a person’s C-peptide levels with outcomes so you can see if some beta-cell function is necessary to benefit from this software?

A: That is a good point, but we did not look at C-peptide. We think it will work best in type 2 patients. In our study, 90% of patients with type 2 diabetes reacted well and reduced their A1c. 10% had a high variability in their glucose readings, with hypoglycemic and hyperglycemic readings too, so it was hard to adjust. Half of patients with type 1 diabetes reacted well because they didn’t have a high frequency of hypoglycemia. This software might enable us to concentrate efforts on patients who need more attention than others by freeing up the time spent on patients who benefit from the software.

Q: What is the minimum number of tests per day that patients need to do to make this software effective?

A: For a basal/bolus patient, four per day. For patients on biphasic insulin therapy, twice a day, so it depends on their regimen. If a patient forgets sometimes its fine, but there is some point at which it is not enough, similar to how we do it in the office.

 

DIFFERENCES BETWEEN PERCEIVED VERSUS MEASURED BLOOD GLUCOSE RESULTS IN PEOPLE WITH TYPE 2 DIABETES

Steven Edelman, MD (UCSD, San Diego, CA)

With the tagline “Test, Don’t Guess,” Dr. Edelman presented an interesting study assessing 294 type 2 patients’ abilities to guess their own blood sugar. Participants were asked to fill out a pre-test survey, guess their blood sugar, have it tested using the Bayer Contour meter, and fill out a post-test survey. The accuracy of their guesses was assessed according to current ISO standards (±15 mg/dl or ±20% for meter values <75 mg/dl and ≥75 mg/dl, respectively) and proposed ISO standards (±15 mg/dl or ±15% for meter values <100 mg/dl and ≥100 mg/dl, respectively). Despite prior beliefs to the contrary, a high percentage of patients were unable to accurately estimate their blood glucose value: 46% and 58% of participants estimated blood glucose values that were outside current and proposed ISO accuracy guidelines, respectively. Dr. Edelman noted that laminated teaching tools will be developed for in-clinic use, which will highlight the importance of testing and illustrate how most people cannot guess accurately. We’re very glad to hear this, and we would note it’s similar to Roche’s educational/marketing strategy with the Accu-Chek 360 tool (STeP study). We also see this study as potentially useful for CGM manufacturers, as it truly reinforces how “feeling” one’s blood sugar can lead to large inaccuracies, and consequently, bad decision making. We think a similar study done for type 1 diabetes is also warranted, as we suspect the results would likely be even more inaccurate.

  • This study was conducted at two Taking Control of Your Diabetes Conferences. A PowerPoint slide shown during the conference advertised the study in general terms: “Purpose: To learn more about how people feel about testing their blood glucose levels. What happens? You will be asked a questionnaire and have your blood glucose tested.” Participants were asked questions about their blood glucose testing behaviors and then asked, “What do you think your blood sugar level is now?” Trained phlebotomists using the Bayer Contour meter then tested blood sugars in a private testing area. Proper hand washing was undertaken and tests occurred throughout the day (in the morning, before eating, and after eating). Throughout the presentation, Dr. Edelman emphasized the scientific rigor of the study. A post-test questionnaire was also given and those who participated received a $25 gift card. An analysis of covariance (ANCOVA) was used to compare blood glucose values estimated by subjects to the meter result, adjusting for various factors that could potentially affect the subject’s estimate (e.g., frequency of SMBG, time since last SMBG, time since last meal, and self-reported A1c level).
  • Accuracy results were evaluated based on current and proposed ISO standards. As a reminder, current standards mandate that compared to reference, 95% of meter values must be within ±15 mg/dl (<75 mg/dl) or ±20% (>75 mg/dl). Proposed ISO standards mentioned by Dr. Edelman were ±15 mg/dl (<100 mg/dl) or ±15% (>100 mg/dl). In other words, new standards (as proposed here) would expand the size of the low range and tighten the accuracy bar in the high range. Dr. Edelman noted, “We have a group of four or five companies that are going to meet the new standard.” We would guess he is referring to the Big 4 meter companies and perhaps Sanofi/AgaMatrix.
  • The study enrolled 294 individuals with type 2 diabetes and a mean age of 59 years. Mean diabetes duration was nine years, 79% were on oral agents, 28% were on insulin, and 14% were on diet/exercise only. Most were white/Caucasian (71%), female (60%), and had an associate degree or higher (87%; “a fairly educated group”).
  • In the pre-fingerstick questionnaire, 77% of patients strongly agreed, agreed, or were neutral with the statement: “My body tells me without testing if my blood glucose is low or high.” A similarly high 71% of participants strongly agreed, agreed, or were neutral that “I make decisions about my diabetes, such as my food intake or my insulin dose even when I do not test my blood sugar.” Only including those who strongly agreed or agreed reduced the numbers slightly to ~65% and ~67%, respectively.
  • “There’s no way you could get any device approved that’s this inaccurate.” Compared to measured blood glucose values, participants’ estimates did not come close to meeting current or proposed ISO standards. See table below for detailed results. A Parkes Error Grid Analysis comparing guesses and actual blood glucose results revealed: 56% of guesses in the A-Zone, 42% in the B-Zone, and 3% in the C-Zone. Dr. Edelman also noted that guesses widely varied, with some participants overestimating their blood sugar by as much 100% and others underestimating by as much as 70%.

Percentage of Inaccurate Blood Glucose Estimates

 

Based on Current ISO Standards

Based on Proposed ISO Standards

Overall (n=286)

46% inaccurate

58% inaccurate

Low glucose range*

Current ISO: <75 mg/dl (n=7) Proposed ISO: <100 mg/dl (n=43)

43% inaccurate

65% inaccurate

High glucose range

(>200 mg/dl; n=59)

63% inaccurate

78% inaccurate

  • The post-fingerstick questionnaire revealed nearly 100% of participants that recognized the value of obtaining blood glucose information. Participants were asked if they agreed with the following statements: “Knowing my blood sugar by checking…” – could help me make different diabetes decisions (99% strongly agreed, agreed, or were neutral); could give me more confidence in my ability to manage my diabetes (98% strongly agreed, agreed, or were neutral); could help me prevent low blood sugar (98% strongly agreed, agreed, or were neutral); give me more confidence in adjusting my daily insulin dose (98% strongly agreed, agreed, or were neutral).

Questions and Answers

Q: I was pleased to see the Parkes Grid. But I think the Consensus Error Grid is more valuable. Have you looked at plotting on that one?

A: We’ll probably do that once we write the manuscript.

Q: You referenced possible hypoglycemia unawareness. Did you assess that?

A: We didn’t. No one passed out during test. [Laughter] If they had hypoglycemia unawareness, that would have made a difference in the low range. But there were only seven patients in that range.

 

SELF-MONITORING OF BLOOD GLUCOSE FOR TYPE 2 DIABETES

Lisa McAndrew, PhD (Department of Veterans Affairs, New Jersey Health Care System, Newark, NJ)

Dr. McAndrew provided a valuable research overview of self-monitoring of blood glucose for type 2 diabetes. She began by delving into why adherence is so poor in type 2 diabetes and explained the common sense model of self-regulation. Using this feedback loop, she translated SMBG into four “required steps” that must be performed for it to be effective – 1) a patient needs to know how to use a blood glucose meter; 2) how to interpret a blood glucose reading; 3) how to respond to the reading; 4) and how to appraise the response’s efficacy. While many recent studies have included the first three steps, only one of 11 recent RCTs has incorporated the fourth step. Dr. McAndrew concluded that SMBG is effective at helping manage type 2 diabetes and lowering A1c, but it must be part of a comprehensive behavioral program. A key part is a provider’s use of the SMBG information, emphasized on her final slide: “Patients are frustrated with the medical community for asking them to self-monitor their blood glucose and then not using the information of the SMBG to guide medical care.” We couldn’t agree more and hope better tools emerge to make data interpretation easier for patients and HCPs. In the great Q&A following this presentation, we enjoyed hearing more focus on individualizing SMBG and using structured testing. We are happy to see educators embracing these paradigm shifts in diabetes care, particularly because other HCPs are really going to need their help in our view to assess how patients are doing from quarter to quarter.

  • To show how SMBG is useful for type 2 diabetes, Dr. McAndrew reviewed the common sense model of self-regulation. Patients link their symptoms and how they feel (step one) to a representation of their illness (step two), which leads to self-management (e.g., diet, physical activity; step three). However, since diabetes “is invisible and imperceptible,” the model breaks down (i.e., no symptoms are recognized and the corresponding self-management does not occur). Dr. McAndrew asserted that self-monitoring could replace the first step in the model, encouraging patients to eventually perform behaviors such as diet and exercise. To support the benefits of self-monitoring for this purpose, Dr. McAndrew noted that if you want your patients to lose weight, you have them write down what they’re eating – studies have shown that this is the best predictor of weight loss. Additionally, patients are more adherent to self- monitoring recommendations than other self-care behaviors.
  • There are four steps for SMBG to be effective: 1) a patient needs to know how to use a blood glucose meter; 2) how to interpret a blood glucose reading; 3) how to respond to the reading; 4) and how to appraise the response’s efficacy. In other words, “what does a reading mean,” “what should I do with it,” and “is what I did effective.”
  • Since 2007, four systematic reviews have concluded that SMBG in type 2 diabetes leads to small improvements in A1c (0.22-0.31%) – St John, 2010; Poolsup, 2009; Allemann; 2009; Malanda, 2012. However, Dr. McAndrew explained that none of these studies used the common sense model of self-regulation to guide their systematic review.
  • Dr. McAndrew’s review of studies examining SMBG in type 2 diabetes (2007-2011) included an analysis of the required four SMBG steps from the common sense model of self-regulation. Half of the 11 RCTs included information about how to monitor and how to interpret at least some SMBG readings (steps one and two). Most included information on how to respond to readings in general (step three). However, only one study taught participants how to evaluate the efficacy of those responses and that was only in relation to hypoglycemia (step four). Some of the 11 RCTs studied were: Polonsky, 2011 (the STeP study); Kleefstra, 2010; Barnett, 2008; Mohan, 2010; Lim, 2011; Yoo, 2008; Scherbaum, 2008; Bonomo, 2010; Farmer 2007; O’Kane 2008.

Questions and Answers

Q: Is there any difference in terms of frequency of testing for type 2 diabetes?

A: It’s a question we don’t know the answer to. There were two RCTs that examined that question and found that increased frequency a couple times per day was associated with better care. I think it’s going to be very personalized. We’ve been moving away from RCTs to interventions that are tailored for a specific individual.

Q: I work with the elderly and they tell me the requirements keep changing on how many strips they get. If you are not on insulin, you get one strip per day. If you are on insulin, you get three strips per day. What do you recommend?

A: At the VA where I work, patients with type 2 diabetes who are not on insulin get no strips. Because the evidence isn’t out there, policy decisions are made on the available evidence. I don't’ have a short answer

for how to change that or improve it. The onus comes on all of us to push for SMBG. If you have one strip per day, it’s important to do monitoring not just at set times, but before and after meals at different times.

Comment: I also work with the elderly. I have patients that say, “I only have one strip per day.” I will tell them to monitor morning sugars for two weeks. Then, monitor once a day before lunch, etc. That has worked for me, and patients like that idea.

Dr. McAndrew: Those are great suggestions. It goes back to the idea of personalizing it.

Comment: I work in primary care inner-city clinics. We run against one strip per day too. I say, “Your insurance will cover 30 strips per month. You tell me what will work for you.” If it’s checking once per day – Monday, Wednesday, and Friday, check when you wake up. Test Sunday at bedtime and Tuesday, Thursday, and Saturday at a meal. Or, they can take three days and check before and after every meal and then take a vacation from testing. It’s their choice. I also give a list – if it’s high, why is it high? I tell them, “Would you drive in a car with no speedometer? Are you going to drive your diabetes without any idea of your control?” That’s what working for us. [Applause]

Comment: We do something termed intentional monitoring. A patient wants to know something like can they get by with two slices of thin crust versus one slice of thick crust? Should they eat mashed or baked potatoes? As long as they want to know something, we can guide them. I tell them that they’re Albert Einstein and the monitor is their laboratory to tell them if the experiment worked or not.

Comment: We’re a tertiary care center with really out of control patients. We’ve sent letters in supporting more than one strip per day. With the A1c value, you have a greater power to say, “This person is out of control and needs to be monitoring three to four strips per day.”

A: If anyone has data, I’m happy to help you get it out there. That’s how we’re going to change these policies. It’s not that SMBG is not important – it’s that the studies haven’t shown that it’s that effective.

Q: How can we put pressure on insurances companies or Medicare?

A: The strongest thing we can do is to keep showing evidence. It will be people in the clinic capturing data on SMBG to show that it’s effective. The more that data is out there, the more people will make those changes in policy.

 

A NEW APPROACH TO USING INSULIN PUMPS IN THE HOSPITAL

Nancy Morwessel, CNP, MSN, CDE (Cincinnati Children’s Hospital, Cincinnati, OH); Jane Stock, RN, BSN, CDE (Cincinnati Children’s Hospital, Cincinnati, OH); and Tammy DiMuzio, RN, MSN, CDE (Cincinnati Children’s Hospital, Cincinnati, OH)

Ms. Morwessel began by describing the benefits of hospital insulin pump therapy, including the ability to precisely deliver small doses of basal and bolus insulin using fewer injections. However, there is a lack of evidence-based protocols for the best way to manage patents’ insulin pumps in the hospital. Interestingly, the majority of audience members believed that pumps are safe in the hospital setting, though Ms. Morwessel said that pumps raise many additional questions and complications – who provides the supplies, who takes responsibility for the device’s operation, which models can be used, how targets are adjusted, when patients should be switched to injections, and more. Ms. Stock followed with a detailed description of the educational program Cincinnati Children’s Hospital developed to train nurses to safely operate insulin pumps. The hospital owned one brand and model of pump and one type of angled infusion set to avoid the issue of training every nurse to use every pump. Ms. DiMuzio concluded the session by presenting the results of encouraging patient and staff satisfaction surveys. Across 13 months, 72 patients (66 with type 1 diabetes, six with cystic fibrosis-related diabetes) experienced two pump-related safety events (one hypoglycemic episode and one pump site not being changed). Of the 33% of participants who responded, all were either completely or somewhat satisfied with the hospital insulin pump, and a majority found diabetes/glucose control to be the same or better with the hospital insulin pump. However, in addition to the low response rate and small sample size, patients and participants were only given the choice between “Yes, completely,” “Yes, somewhat,” and “No, not at all,” suggesting that the results may be positively skewed. Additionally, 92% of staff respondents (n was not reported) believed the pump to be beneficial in the hospital compared to injections. Finally, Ms. DiMuzio urged insulin pump companies to develop a pump specific to hospitals, a plea received very warmly by the audience, who applauded her firm stance – we are not certain about the practicality of this.

Questions and Answers

Q: First, did you develop your own online training? Second, how did you prepare patients for discharge and switching back to their own pump?

Ms. DiMuzio: We developed our own training but used a lot of the material from the companies.

Ms. Stock: An educator always meets with someone, especially if they were hospitalized for DKA and talks about their skills. They can see their actual history in the pump and we have demo pumps for people so we can verify that we are comfortable with their skills on their own pump. We also have them transition in the hospital.

Q: How do you justify with a patient, or parent, that you are capable of taking over care versus a parent who has been doing well with management for a long time?

Ms. Stock: Things in the hospital are different – your child is on IV fluids with dextrose, the food and activities are different. It’s getting parents to understand that the whole situation is different, and what may work beautifully for you at home may not work in the hospital setting.

Comment: I’m totally impressed; I’ve been working on this since 2009. The solution you created won’t work for me, but the important thing is making patients safe in that particular environment. I really want you to write an article, because I need to prove it to people, and they won’t just listen to a CDE.

 

THE IMPORTANCE OF GLUCOSE METER ACCURACY IN CLINICAL DECISION MAKING*

Steve Edelman, MD (University of California San Diego, San Diego, CA)

Dr. Edelman gave a very persuasive talk on the importance of blood glucose meter accuracy. Above all, he emphasized that self-monitoring of blood glucose (SMBG) is important for type 1s, type 2s (on orals and on insulin), and even in people with prediabetes (he believes once or twice a week is beneficial) to make changes in their lifestyle and medications. Dr. Edelman highlighted a study that asked patients to estimate their blood glucose based on feeling alone – as might be expected, they radically overestimated their ability to do so accurately (more detail on the study will be presented on Saturday). He also emphasized the vast number of variables that affect blood glucose, which should not be exacerbated by meters that require coding, are hard to use, or are inaccurate. Finally, Dr. Edelman summarized the times when accurate blood glucose readings are most critical: when dosing insulin (especially in highly insulin sensitive patients), using an insulin pump, and calibrating a CGM (“As a CGM user myself, accuracy is the number one thing” and “one of most important things for CGM is calibrating with an accurate meter”).

  • “A1c is only one measure. It does not provide immediate feedback about glucose levels. Patients need immediate feedback.” With this in mind, Dr. Edelman believes that people with prediabetes could benefit from occasional blood glucose testing a few times after meals and when fasting. He also illustrated how A1c does not (and cannot) tell the whole story of a patient’s glucose control – Dr. Edelman displayed a wildly erratic CGM trace (ranging from 40- 400 mg/dl) despite the patient’s A1c of 6.7%.
  • In people with type 2 diabetes, glucose monitoring is necessary due to the vast differences between patients’ perceived and measured blood glucose results. Dr. Edelman described a study undertaken at a TCOYD conference, where individuals were asked a number of questions, including a guess of their current blood glucose. Then, their actual blood glucose was measured and compared to their estimate. Prior to the fingerstick, 77% of patients strongly agreed, agreed, or were neutral that they could feel what their blood glucose was. However, 46% of patients were inaccurate according to current ISO standards, and 58% were inaccurate according to proposed ISO standards. In short, despite a belief that they can accurately predict their blood sugar based on feeling alone, people with type 2 diabetes are not very good at doing so. To Dr. Edelman, this supports the need for blood glucose monitoring in type 2 diabetes.
  • In another study undertaken at a TCOYD conference, attendees’ meter settings were randomly checked to see if the code in the meter matched that on the strip bottle. Among type 1s, 11.2% had their meter improperly coded, a number that rose to 16.2% in type 2s (11% for those on insulin and 19% for those on orals). Dr. Edelman emphasized that these mis- codings can result in very large meter inaccuracies, which are “fairly significant for a type 1 with a correction factor of 1:50.” On a somewhat related note, we wonder how often mis-calibrations occur on CGMs (either due to patient error or inaccurate meters), how drastically these affect CGM accuracy and/or clinical outcomes, and whether these errors might increase patients’ frustration with the devices (or even encourage discontinuation). From a patient perspective, we look forward to future CGMs that hopefully won’t require fingerstick calibration at all.
  • Dr. Edelman presented a case study to illustrate the importance of glucose meter accuracy in an insulin-sensitive patient. The example patient had a target glucose of 100 mg/dl and a correction factor of 1:50. Assuming her actual blood glucose was 174 mg/d (lab draw) and a 20% meter error gives a range of 139 mg/dl to 205 mg/dl. Dr. Edelman explained that at 139 mg/dl, she would take one less unit of insulin, which would lead to hyperglycemia all afternoon. At 205 mg/dl, she would take one extra unit of insulin, which would lead to hypoglycemia. He illustrated the same example for a blood glucose of 74 mg/dl (a 20%-error range of 59 mg/dl to 89 mg/dl). She would consume carbs at 59 mg/dl, leading to hyperglycemia, while a reading of 89 mg/dl would give her a false sense of security and put her at risk for hypoglycemia. The caveat to this example, said Dr. Edelman, is “If errors are larger than 20%, the situation could be worse.”

 

Product Theater: Introducing V-Go – The Simple Delivery Of Basal-Bolus Insulin Therapy For Adult Patients With Type 2 Diabetes (Sponsored by Valeritas)

INTRODUCING V-GO: THE SIMPLE DELIVERY OF BASAL-BOLUS INSULIN THERAPY FOR ADULT PATIENTS WITH TYPE 2 DIABETES

Matt Nguyen, PharmD (Valeritas, Bridgewater, NJ); Beverly Everitt, APRN, BC-ADM (Bayview Physicians Group, Virginia Beach, VA); Patricia Munz, MSN, APNC, CDE (Endocrinology Associates of Central New Jersey, Freehold, NJ), Peter Jackson (patient)

The focus of this popular 6:30 AM session (there were about 150 attendees) was the V-Go, a disposable insulin delivery device that supplies basal-bolus therapy with a once-daily application without any electronics or tubing. V-Go delivers insulin at a preset basal rate of 20, 30, or 40 units per day, and its reservoir contains an additional 36 units of insulin that can be released for on-demand bolusing (in two-unit increments). Valeritas has identified adults with type 2 diabetes currently on insulin but in suboptimal control – especially those leading busy lives or for whom insulin injections are not practicalas ideal candidates for the V-Go, along with those not yet on mealtime insulin who need prandial coverage. Speakers touted the many benefits of the V-Go: discretion; once-daily applications; more physiologically representative than once-daily basal insulin, as it also allows for mealtime boluses; and easier than a pen since it is always with them. Learning how to fill and apply the V-Go was quite easy; it took us about five minutes to learn all of the steps and apply it correctly. Skin irritation and adhesive residue left behind after V-Go removal were cited as the main disadvantages of using the device, though we thought the strength of the adhesive was certainly a plus. We are looking forward to following V- Go’s progress, as there’s no question more patients should be on insulin and there is a need for more insulin delivery alternatives.

  • V-Go is a once-daily, disposable insulin delivery device that supplies basal-bolus therapy without any electronics or tubing. V-Go just launched in April 2012. The device delivers insulin at a preset basal rate of 20, 30, or 40 units per day. The reservoir holds an additional 36 units of insulin that can be released on demand (in two-unit increments) when a bolus dose is needed. Therefore, it is most appropriate for patients with type 2 diabetes that require moderate doses of insulin (between 20 and 76 total units per day, which Ms. Munz estimated to be about 80% of the type 2 diabetes population – this assumes patients cut their total daily dose by 12-20%, which is typical when going on a pump), but the speakers indicated that patients with type 1 diabetes have used the device successfully. It is cleared for use with Humalog and NovoLog. Because the needle is fully retractable, it can be disposed of in the regular household trash.
  • Valeritas has identified suboptimally controlled adults with insulin-treated type 2 diabetes – especially those leading busy lives or for whom insulin injections are just not practical – as ideal candidates for V-Go use. The presenters focused on V-Go’s ease of use and physiological delivery compared to alternative insulin delivery methods. To illustrate this, a four-quadrant graph was displayed with axes for complexity (ranging from simple to complex) and physiological imitation (from less to more physiologic). The current progression of diabetes therapies (orals, basal, basal + one bolus, MDI, insulin pump) runs the gamut from simple and less physiologic (orals) to complex and more physiologic (insulin pump). Valeritas maintained that V-Go resides in the simple and more physiologic quadrant where there are currently no other options.
    • Complexity often increases barriers to adherence – in one study of 692 type 2s on three insulin injections per day, 72% of them never gave an injection away from home. Factors such as hassle, fear of pain, embarrassment of use in public, a mental or physical handicap, or simply forgetting to bring insulin often impeded optimal insulin usage. (Editor’s note – this is quite staggering from our view.)
  • At this product theater, the company’s AADE exhibit, and at ADA, we noticed that Valeritas has tried to carefully distinguish the V-Go from insulin pumps. The company referred to the three available sizes as different “dosing options,” and Mr. Nguyen indicated during Q&A that the device is covered under Medicare Part D (prescription drug coverage) rather than Medicare Part B, where durable medical equipment is traditionally covered. We think this is a very smart move for the company – from a free association perspective, we believe many associate insulin pumps with a high degree of complexity or having type 1 diabetes. The pump has clear points of differentiation from traditional insulin pumps, and we have no doubt this will be an integral part of the marketing message going forward. Targeting patients on MDI also offers the company the largest initial target market.
  • We were shown a video detailing the step-by-step process for using the V-Go and then permitted to fill and apply one (sans-needle) ourselves. The device is filled with insulin using the EZ Fill (an ergonomic handheld piece of plastic about the size of a small remote control).
    • Filling: To fill the V-Go, an insulin vial is inserted into the appropriate chamber of the EZ Fill, the V-Go is slid into the filling chamber, a plunger is primed (taking about 5 seconds) and then slowly depressed (over about 25 seconds). One can watch the device fill through a filling window. The EZ Fill is stored in the refrigerator and can be re-used daily for one month (this is due to the testing criteria used with FDA as we understand it).
    • Application: The filled V-Go can then be applied to the abdomen or arm via a sturdy adhesive. A protective film must be removed and then the device can be pressed onto the body. Pushing one button inserts the needle and begins basal delivery.
    • Bolusing: When it is time to bolus, a push of the “Bolus-Ready” button activates the “Bolus-Delivery” button, and pushing the “Bolus-Delivery” button delivers two units of insulin. The patient would then repeat the “two-click” bolus delivery process as many times as needed to reach the necessary dose for that meal. Although it is not possible to visually check how much insulin has already been administered, based on human factors data provided to FDA, the company (and presumably the FDA) did not believe that a bolus counter was necessary. In our view, keeping count of “clicks” during this process is easy to understand upon testing the product.
    • Removal: After 24 hours, the device is removed by pressing the needle retraction button and peeling off the adhesive. After the needle is retracted, it cannot be engaged again, so the device must be discarded at this point. We note that there is an outside chance that the needle retraction button could be pressed by accident – it is located opposite the “Bolus-Ready” button. However, the needle retraction button takes considerably greater force to press than the “Bolus-Ready” button, so should prevent accidentally rendering the device unusable too early. Plus, we would assume that FDA’s recent hardcore focus on human factors would have examined this issue thoroughly.
  • In Valeritas’ patient survey of 31 V-Go users who wore the device for an average of 202 days (6.5 months), V-Go’s ease of use, discreetness, comfort, and helpfulness of patient education materials (including information on diet and exercise) were all rated around nine on a scale of one to 10. Compared to their previous therapy, 86% ofrespondents indicated that the V-Go provided better glycemic control. Additionally, in a retrospective analysis for A1c data for those who gave consent (an n was not provided), patients experienced ~1% reduction in A1c from a baseline of ~8.5%; A1c worsened again after going off of the V-Go after the study period ended.

Questions and Answers

Q: Could you comment on the V-Go’s 24-hour capacity vs. other pumps that go for 72 hours? Is there going to be a change toward longer wear time?

A: V-Go is designed to be once a day – there is no anticipated change in design to go beyond 24 hours because we’ve experienced that patients think once a day is appropriate.

Q: You mentioned that reimbursement is better than patients tend to expect – could you expand on what type of reimbursement there is, and are you thinking of Medicare/Medicaid?

A: Reimbursement for V-Go will vary according to patients’ individual insurance. Some patients are covered by their pharmacy and some patients go through medical benefits. An easy way to test this is to send patients to their pharmacy – if they are denied, then there is information in V-Go starter kits on who to call to find out what their benefits are. With regards to Medicare Part D (drugs), V-Go is covered. It is not covered under Medicare Part B (traditional electronic devices). I have no information regarding Medicaid at this point, but they are traditionally slower at adopting products.

Q: Peter, do you remember what your total daily dose was when you started the V-Go?

A: I maxed out – I was using the full 40 and was clicking 18 times a day (36 units) until I got my diet together and could go down from there.

 

Product Theater: Introducing FreeStyle InsuLinx – The Latest Advancement for Insulin Users (Sponsored by Abbott Diabetes Care)

UNDERSTANDING THE BARRIERS OF PATIENT LOGGING AND PATTERN MANAGEMENT

Ralph DeFronzo, MD (University of Texas Health Science Center, San Antonio, TX)

Dr. DeFronzo opened the early morning corporate symposium by noting, “The Freestyle InsuLinx is a major, major advance in terms of helping patients understand their diabetes and take control of it.” He framed the session with a review of the suboptimal state of glycemic control, the low rates of SMBG and logging, and the key patient barriers (it is burdensome, it is perceived as not worth the effort, and it is never seen or discussed with the physician). He concluded with a short review of the beneficial features of the FreeStyle InsuLinx, which he believes can help overcome these obstacles.

  • “Despite this enormous amount of knowledge, glycemic control remains suboptimal.” Dr. DeFronzo reminded the audience that 50% of patients have an A1c above 7%, including 35% of people with an A1c above 8% (he did not specify a source). Additionally, he highlighted that the incidence of complications – as observed in the Pittsburgh Epidemiological Study – is much higher than in the DCCT.
  • Dr. DeFronzo highlighted the “bad” news about SMBG: a significant portion of both type 1s and type 2s test far too infrequently. In one survey of type 1s, 24% performed SMBG less than once weekly (this was very disturbing to hear), while only 39% performed SMBGonce daily. The data is more “dismal” in type 2 insulin users, where 29% performed SMBG less than once monthly or never, and only 39% performed SMBG at least once daily. No specifics on the study design or population were collected, though we’d be interested to hear more given the surprising numbers.
  • Insulin users struggle to log insulin and blood glucose values. In a 2011 self-reported survey of 500 people with type 1 diabetes and 504 people with type 2 diabetes, 46% used paper logbooks, only 18% used electronic logbooks, and 36% did not log glucose values at all. The data was similar for insulin dosing, where just 32% used paper logbooks, 17% used electronic logbooks, and 51% did not log insulin doses at all.
  • Dr. DeFronzo summarized the key obstacles to tracking blood glucose and logging insulin: it is burdensome, it is perceived as not worth the effort, and it is never seen or discussed with the physician. He emphasized that the FreeStyle InsuLinx is a very simple system to use that can help address these burdens. Additionally, Dr. DeFronzo highlighted the results from Dr. Bill Polonsky’s STeP study (Diabetes Care 2011), which found that “if patients log the data and discuss with their doctor…it could drop A1c by as much as 0.5%” (the result from the study’s per- protocol analysis in the structured testing group (-1.3%) vs. the active control group (-0.8%)).
  • “The FreeStyle InsuLinx has all the qualities that will make it easy for patients, CDEs, and HCPs to really improve glycemic control.” The features he highlighted included: providing easy-to-interpret data for patients and physicians, built-in reminders, and meal and event marking to match blood sugars with swings in glucose.

 

AN ADVANCEMENT FOR INSULIN USERS

Anne Peters, MD (University of Southern California, Los Angeles, CA)

Dr. Peters detailed the features and benefits of the FreeStyle InsuLinx system in what was the session’s main presentation. She has been using the meter for the past couple months and really likes it – key for her is the easy downloading and report transmission, the large touchscreen (making it easy for patients to input information), the on-meter software (especially because it’s Mac-compatible), and the combined insulin dosing/glucose data. Through use of an on-stage iPad, Dr. Peters demonstrated the user interface in lightning quick fashion by virtually inserting a strip, testing, meal marking, and logging an insulin dose (“and then you’re done!”). Although she was initially negative on the black-and-white screen, she conceded that it makes the battery life last much longer (“I’ll forgive them on that because it’s better for the environment”) – we would agree that the screen is significantly harder to read than LifeScan’s OneTouch Verio IQ or Sanofi’s iBGStar on an iPhone, though the battery tradeoff is likely a very important one for many patients. Much like Dr. Bode did in a similar symposium at ADA 2012 (see page 70 of our report at http://bit.ly/OE7i2S), Dr. Peters reviewed the reports on the FreeStyle Auto- Assist software. Her favorite was the Modal Day report (“The idiot’s guides to pattern recognition in diabetes”) and the 90-day average blood glucose trend graph on the Snapshot report. The latter is a particularly unique display that we cannot recall seeing on other glucose meter software before. Dr. Peters emphasized that the trend graph helps provide encouragement when patients can see improvement over time (i.e., a downward sloping average glucose trend line).

 

PATIENT CASES AND CLINICAL APPLICATION

Donna Tomsky, MSN, RN, CDE, C-NP, FAADE (ABQ Health Partners in the Department of Endocrinology and Diabetes, Albuquerque, NM)

In the session’s final presentation, Ms. Tomsky reviewed “where the rubber meets the road” using two case studies of the FreeStyle InsuLinx: 1) an 80 year-old type 2 on insulin (confirmation for her that the touchscreen device can be effectively used in the elderly) and 2) a 41-year old type 1 with nocturnal hypoglycemia. The FreeStyle Auto-Assist software reports were the focus of both cases, and like Dr. Peters, Ms. Tomsky was most positive on the Modal Day report. In the 80-year-old patient, the combined insulin and glucose data allowed her to quickly see that mealtime insulin was being taken far too long after eating, leading to postprandial hyperglycemia. The takeaway was somewhat similar in the younger patient, who had been snacking between meals at work and forgetting to dose insulin. Ms. Tomsky was quite positive on the software overall, emphasizing that she liked the reports “because they’re visual and numeric.”

 

PANEL DISCUSSION

Ralph DeFronzo, MD (University of Texas Health Science Center, San Antonio, TX); Anne Peters, MD (University of Southern California, Los Angeles, CA); and Donna Tomsky, MSN, RN, CDE, C-NP, FAADE (ABQ Health Partners in the Department of Endocrinology and Diabetes, Albuquerque, NM)

Dr. DeFronzo: What are your thoughts on why people do such a poor job of recording glucose values?

Dr. Peters: For many it’s a burden – they don’t want to disrupt their lives by checking. Whether you are rich or poor, you can be non-adherent for different reasons: time, money, not wanting to do it, etc. I find that it’s often a psychological reaction to a number – “I don’t want to test when I’m high or I’m worried that Dr. Peters will say I did a bad job.” It’s really burdensome to have to log information. Patients also feel like physicians are not doing anything with the data. So there are a lot of different barriers to doing it. The tools that make it easier to do that will make diabetes easier.

Dr. DeFronzo: Do you find that patients are more adherent in your upper scale clinic vs. the barrio?

Dr. Peters: The problems exist for both groups due to different psychological issues. In my East LA clinic, the patients are quite poor and not numeric. They often don’t understand the meaning of the numbers. Their lives are tumultuous and they have stressors – maybe not having food, their home was foreclosed upon, things that make you not want to think about blood glucose levels. Anyone is capable of not adhering. It’s hard.

Ms. Tomsky: A lot of people don’t know what to do with the information. That’s the common question I ask.

Dr. Peters: I had this guy – he was the meanest guy I ever met. His A1c was 12% and he said he was giving insulin multiple times a day and blamed everyone else for his poor control. He started using this meter and it was amazing to see him face that he was only giving insulin once a day. He faced his own reality. Now, he is so proud that he downloads his report and sends it to me. He may get better and I think this helped him.

Dr. DeFronzo: We work at the VA and these are difficult patients. My CDE picked out four of the most difficult patients to use this meter. They had incredibly dramatic results when they could see what the issues were and they had to face what they were doing.

Dr. DeFronzo: (Referring to the Snap Shot report) I don’t find standard deviation to be helpful. I personally like the modal day report.

Ms. Tomsky: I think the average is helpful to correlate to A1c.

Dr. Peters: This report gives you the fast gestalt. Standard deviation is less useful. I find this report useful for the number of tests per day that patients are doing. People like averages, although we as clinicians use A1c.

Q: The threshold on the reports was less than 60 mg/dl. Can you change the thresholds?

Dr. DeFronzo: Yes, that’s very simple to do. You can also decide where threshold and alarms come up. Ms. Tomsky: Because the software resides in the meter, you can do customization with it. You can put in

remarks or whatever.

Q: How does the information incorporate into a patient’s medical record?

Ms. Tomsky: It depends on how you record data in your particular practice. I like the numeric. In your dictation, you can note the percentage above range, the average, and the number of tests per day. In terms of interfacing with medical records, you’d have to go to the booth and speak with the Abbott folks.

Dr. Peters: You can print the reports out in color or black-and-white PDF and keep them on your computer. It’s harder to do this integration, but we’re getting closer to that point.

Q: Is there a bolus calculator in the device?

Dr. Peters: It would be nice and it has that capacity for the future. That’s something we all want. Currently, it’s a logbook of sorts. Hopefully it will become more of a dose calculator. This is the first step for now.

Q: Can you put the insulin dose in retrospectively?

Dr. DeFronzo: Yes. You can also set reminders to put the doses in. The little reminder system is one of the advantages of it. There are also reminders to check your glucose.

Dr. Peters: It’s better if they can remember to put the dose in at the exact time. It’s so quick, you saw me do it in the demo.

Dr. DeFronzo: Also, that means they thought about how to dose the insulin based on the glucose reading.

Q: I don’t test my blood glucose before every insulin dose (believe it or not). Can we log an insulin dose without first doing a test?

Dr. Peters: Yes, you can.

Ms. Tomsky: What’s we’re trying to do is change patients’ behavior. One patient I showed was forgetting to take insulin. Now that she is logging it, she saw. This really facilitates behavior change; believe me.

Q: What do you recommend for patients on insulin doses greater than 50 units at a time?

Dr. Peters: Once I get above 50 to 60 units of basal insulin, I give mealtime insulin. I use a fair amount of U-500 in patients on high doses. In my East LA practice, there are lots of insulin-resistant patients. It works out well and you just multiply by five for this dose logging on the meter.

Dr. DeFronzo: Once you start above 60 units of long acting, I recommend splitting the dose or combining it with rapid acting insulin.

Q: How often do you get complete data on insulin users – including both insulin doses and blood glucoses?

Dr. DeFronzo: Our experiences are similar – about one third gives you reasonable data, and the other two thirds you’re missing data, particularly the insulin doses. People are less enthusiastic about recording insulin values.

Dr. Peters: For patients that are more reluctant to put in data, I have them do it intensively one week per month. I use it to really give me a sense of what’s going on. If I get one really good week, that’s very helpful. Then, they might see it’s useful and become more motivated to do it. Get patients to send you the download and see the numbers. The goal is helping the individual patient in real time and getting them in communication with me. It’s help plus data, and the plus data part is where there is utility for me.

Dr. DeFronzo: It’s easy to record what a patient has done on a particular day. Exercise can be put in and related to glucose values.

Q: What has been your general experience in terms of how this has been able to help your patients? How many times do you find people are doing something really wrong?

Dr. Peters: I cannot give a percentage. There is always this phenomenon of newness. People like the way data is entered on the meter. There is one patient who lives about an hour from me – I gave him one of these meters. He was using it and then something happened and he told me he needed to get a new one. I told him, “You’ve got other meters.” But he said, “I really want to keep using this one.” It’s about the patient feeling this is useful. The data is valuable when the insulin dosing values are in there. Also, I focus on two weeks – I want to know everything you did in the last two weeks. Studies show that two weeks is the optimal amount to give you a sense of what’s coming in the immediate future. A week is too short and a month is too long.

Ms. Tomsky: I also like the 90-day trend graph.

Dr. Peters: The worse thing on that is someone is really trying hard and then they get worse over time. My only issue with trend data is that you want to make sure it’s never discouraging. I do a lot of cheerleading. However negative the data is, I always make it into a positive. It’s so hard to live with this and I don’t want people to feel like terrible failures.

Q: How could the FreeStyle InsuLinx affect patient outcomes?

Dr. DeFronzo: I would refer to the study from Bill Polonsky [Diabetes Care 2011]. It was in a large number of non-insulin requiring type 2s. They trained patients to do home glucose monitoring and to record their results. One group was required to meet with a physician who had to review the data every three months. The other group received standard care. At the end of one year, the group that was required to review their data with a physician had a 0.3% greater drop in A1c compared to the control group. But within the group that was supposed to review, one-third actually did it. So if you simply looked at those who actually reviewed data, the reduction in A1c was 0.5%. That’s quite clinically significant in terms of microvascular complications. The Freestyle InsuLinx will make it even easier to review the data with physicians. It will provide more reinforcement and allow physicians to see and more likely to discuss the data.

Ms. Tomsky: Don’t just say physicians. Also diabetes educators – we do a lot of the insulin dose adjustments.

Dr. DeFronzo: Sure, we all know that you do the work.

Dr. DeFronzo: Just some concluding remarks. I think you can see from the demo Anne showed that the meter is simple to use. The benefits are the touchscreen, the ability to quickly enter data, and to quickly get back information in a visual way. We like the modal distribution to get a sense of the average and a good picture of the swings. The reminders are very good and there are a number of things you can set. There is also good logging of events so physicians can line up and see what’s happening. Then there is the ability to download information. There are a lot of advantages of the system. In all of our experiences, this is very easy to use and patients like it.

 

BAYER’S NORTH AMERICAN COMPARATOR TRIAL (NACT)

Solveig Halldorsdottir, PhD (Director Scientific Affairs, Bayer)

Dr. Halldorsdottir began her presentation with a trip down memory lane. We’ve come a long way since Bayer introduced the first portable BGM system in 1969, and Dr. Halldorsdottir suggested that the Contour Next products would usher in the next era of BGM. The contour Next EZ (launching August 20), Contour Next Link (launching September 4), Contour Next USB (FDA cleared with launch expected in early 2013), and Contour Next (pending 510(k) clearance) all use Bayer’s new Contour Next test strip. Dr. Halldorsdottir highlighted the notable features of the Contour Next strip including: 1) proprietary mediator which can detect glucose at very low levels (and is used in addition to the GDH-FAD chemistry utilized by Contour’s legacy strips); 2) multiple pulsing – with seven pulses throughout the test time; and 3) an advanced algorithm. The strips have demonstrated impressive accuracy in the hands of trained operators, but more importantly in the hands of patients – 100% of the time results were within 10 mg/dl or 10% at glucose concentrations <75 mg/dl (n=7) and 96.3% of the time within 10 mg/dl or 10% at glucose concentrations > 75 mg/dl (n=108). The highlight of the talk was Dr. Halldorsdottir’s review of Bayer’s North American Comparator Trial (NACT) comparing the Contour Next EZ, Accu- Chek Aviva Plus (Roche), FreeStyle Freedom Lite (Abbott), OneTouch Ultra 2 (LifeScan), and Truetrack (Nipro Diagnostics) meters. The Contour Next EZ demonstrated superior accuracy (assessed by mean absolute relative difference from laboratory reference values), with the Accu-Chek Aviva Plus in second place. We are glad to see what looked like a well-conducted comparative study from Bayer. We hope to see an independent study of the latest meters at some point in the future (e.g., LifeScan’s OneTouch Verio IQ, Sanofi’s iBGStar, etc.).

  • Bayer’s North American Comparator Trial (NACT) compared the Contour Next EZ, Accu-Chek Aviva Plus, FreeStyle Freedom Lite, OneTouch Ultra 2, and Truetrack meters. One to three blood samples were taken per participant (n=146) for a total of 388 blood samples per meter. All lancing and testing was performed by trained operators. The NACT primary endpoint was to evaluate accuracy assessed by mean absolute relative difference from laboratory reference values, with the secondary endpoint to evaluate differences in accuracy. Bayer is in the process of publishing this study.
  • Contour Next EZ was first in accuracy across overall the blood glucose range (23- 386 mg/dl) and low blood glucose range (<70 mg/dl); Accu-Chek Aviva Plus meter was second in both categories. We would be interested to see how the newest meters (like the OneTouch Verio IQ) fit into this ranking.
 

Overall blood glucose range (23-386 mg/dl)

Low blood glucose range (<70 mg/dl)

Blood glucose meter

MARD (% error)

Ranking*

MARD (% error)

Ranking*

Contour Next EZ

4.73%

1

0.65%

1

Accu-Chek Aviva Plus

6.32%

2

2.53%

2

FreeStyle Freedom Lite

18.29%

3

18.27%

3

OneTouch Ultra 2

23.39%

4

22.39%

4

Truetrack

26.17%

4

33.23%

5

* Systems with different rankings are significantly different from each other (p <0.05), while those with the same ranking do not differ significantly from each other (p >0.05).

 

FACILITATING BEHAVIOR CHANGE THROUGH SMARTER GLUCOSE MONITORING

Jane Seley, DNP, MPH, MSN, BC-ADM, CDE (New York Presbyterian-Weill Cornell Medical Center, New York, NY)

Dr. Seley shared her insights on some of the challenges to diabetes self-care and behavior change and how providers can help patients overcome these challenges. She reminded the audience that diabetes is a demanding disease with many barriers to self-care – depression, denial, poor understanding of the disease process, lack of social support, lack of privacy in checking blood sugar, limited reimbursement, pain of blood glucose monitoring, overly complex regimens, limited access to care, influence of other health problems, and lack of physical and cognitive ability to perform self-care behaviors safely and correctly – just to name a few! Part of the role of being a diabetes educator is to help patients solve these problems, she said; her secret to success was building relationships with patients by reviewing blood glucose data as an equal partner. Dr. Seley emphasized that providers need to assess whether patients know how to monitor their blood glucose correctly and check that their meter is accurate (she also noted it is important that providers match the patient with the meter – an older patient likely does not have the same meter preferences as a young, more active patient). She highlighted that providers need to engage with their patients to make sure patients know when to check blood glucose and what their targets are. Finally, she encouraged providers to find ways to motivate patients to monitor their blood glucose, as the real key to diabetes care, she said, is using CGM information to make “smart” changes in the treatment plan.

 

PANEL DISCUSSION

Steve Edelman, MD (University of California San Diego, San Diego, CA); Solveig Halldorsdottir, PhD (Director Scientific Affairs, Bayer); Jane Seley, PhD, DNP, MPH, MSN, BC-ADM, CDE (New York Presbyterian-Weill Cornell Medical Center, New York, NY)

Q: I’ve had experiences teaching classes where blood glucose was 160 mg/dl in one finger and 200 mg/dl in another. You’re saying with this new monitor, I shouldn’t be seeing that? Or is that natural physiologic variability?

Dr. Seley: Did the patient wash his or her hands? That is the first thing I would worry about.

Q: In one classroom there is a sink. In the other, there isn’t one. Even when that’s not a factor, sometimes I see even greater variability than 160 to 200. This happens even when you check against the venous draw. If you’re 180 mg/dl, is that 180 mg/dl throughout your entire blood stream?

Dr. Halldorsdottir: You can have differences between venous and capillary blood. But the differences you are referring to are quite large. What you are referring to is precision. That was not a subject of my talk, but as you can infer from the plots I showed, precision is quite improved on the new platform.

Dr. Edelman: The blood glucose is pretty much the same in large vessels. Depending on where you’re testing – forearm, finger, earlobe – it can be different. It takes time for the blood to go from the main arteries to peripheries. People test two times in a row to calibrate CGM. They see differences like you said. It could be that one hand handled a glucose tab hours ago, or it could be meter inaccuracy. I usually have them test four times and take the two closest.

Q: I do blood glucose screening in my area. I’m a festival queen. I was testing blood glucoses and asked a patient, “Do you have a history of diabetes?” No. She had a big Pepsi in her hand. Her blood glucose was 396 mg/dl. I then asked her, “Did you spill your Pepsi?” She said yes, so she washed her hands and we tested again. Turns out she had a blood glucose of 175-180 mg/dl – prediabetes. This is to say that it’s important to make sure your hands are clean. I would have said she had diabetes just because she had spilled a Pepsi on her hand.

Dr. Halldorsdottir: That’s an important point – our system cannot figure out if your hands are clean or not.

Dr. Edelman: In the studies that were showed, hands were super cleansed. This is the best data without any other interfering factors.

Q: I was disturbed by the OneTouch Ultra accuracy that you showed. When we first starting using CGMs, we were giving all our patients OneTouch meters. How good were those results?

Dr. Halldorsdottir: We were also surprised by the results. We were careful to go back and see that there was nothing suspicious about the lots, recalls, etc. It’s just what we found in this particular study. (Editor’s Note: As we understand it, OneTouch Verio IQ strips were originally slated to be tested in the study, but were removed from the protocol due to the manufacturer recall earlier this year).

Q: This speaks to sensor users who say they’ve seen this too. I can say we won’t be using OneTouch Ultra for CGMs any more.

Dr. Halldorsdottir: Well that’s your decision, but I agree with you, it’s not what we would have expected.

 

Corporate Symposium: Get Pumped – Best Practices for Success With Insulin Pump Therapy (Sponsored by Animas, Asante, Insulet, Medtronic, Roche, Tandem, and Unomedical)

INTRODUCTION

Gary Scheiner, MS, CDE (Integrated Diabetes Services, Wynnewood, PA)

Audience members arriving to the largest hall at AADE were greeted with blasting, topic-appropriate music: “Pump the Jam” by Technotronic and “Pumped Up Kicks” by Foster the People. Mr. Gary Scheiner introduced the insulin pump-focused session to a completely packed, attentive room. He started with examples of “Great Feuds through the Ages” – Hatfields and McCoys, Red Sox and Yankees, Captain Kirk and the Klingons (Star Trek), and, historically for diabetes, the various insulin pump companies. However, Mr. Scheiner gave credit to the sponsors of the session (Animas, Asante, Insulet, Medtronic, Roche, Tandem, and Unomedical), who “put aside their differences to support the program.” We congratulate AADE on making such a unique, joint educational symposium happen. It’s very encouraging to see fiercely competing companies working together to help support and improve educators’ pump skills – everyone benefits from this type of commitment. Mr. Scheiner concluded with a summary of how being better pump trainers and educators can help industry (increased use of pumps, optimal use of supplies, less dependence on customer support, improved word-of-mouth referrals, fewer returns), payers (ensure pump use by qualified individuals, less complications and ER visits, less drop out), patients (improved control, enhanced self-efficacy, minimizing complications), and HCPs (more effective use of time, marketable skills, job/financial security, improved outcomes, and industry collaboration).

 

TAKING IT TO THE NEXT LEVEL

Jennifer Block, MSN, RN, CDE (Stanford University, Stanford, CA)

As a type 1 patient and CDE herself, Ms. Block spoke with a mix of passion and humility that enamored the audience: “I recognize for many of us, we don’t have a choice on whether we take insulin, but we do have a choice as to how we take insulin. Any choice we can give our patients when they are living with something so challenging is really a gift.” Her presentation addressed critical aspects and possibilities of insulin pump therapy, as well as some of its limitations. Underlying her discussion was a clear emphasis on individualization.

  • Ms. Block gave her recipe for successful pump use: 1) test frequently, take insulin, and adjust insulin doses based on patterns and repeat as needed; 2) “when in doubt, take it out!” (referring to infusion sets) and 3) providers and patients should work together to assess blood glucose data. As to the latter, Ms. Block emphasized the importance of looking for patterns on a regular basis and adjusting therapy in response to trends. In doing this, she encouraged providers to explain what the patterns are and why they warrant certain responses – if you can teach your patient how do this, she said, your patient will be more likely to do it.
  • Insulin doses need to be fine-tuned on an individualized basis – “One basal rate does not fit all,” emphasized Ms. Block. Even for a single individual, one basal rate isn’t necessarily maximizing the benefits of pump therapy. Patients should consider alternate basal rates for menstrual cycles, extended illness, work vs. weekend days, and extended travel as well as temporary basal rates for illness, stress, or change in activity. The timing of insulin boluses should also be individualized – by looking at data and responding to it, insulin dose timing can be optimized. Additionally, Ms. Block reminded the audience that an important part of fine-tuning insulin therapy is appropriately considering the impact of carbohydrates and determining the best correction factor for a patient. She also highlighted the fact that not all foods with the same carbohydrate content have the same impact on blood glucose
  • Every patient is different and has different infusion set preferences. Ms. Block described discontinuing her first insulin pump because she had only one option for an infusion set at the time (it didn’t work well for her). Insulin doesn’t work, she said, if the insulin doesn’t go in, so finding an infusion set that is an appropriate fit for each individual is important. She described three types of sets and the possible candidates. First, the steel 90° set might be a good option for those who have a history of kinked cannulas. Second, the plastic 90° set offers a variety of insertion devices and may be preferred by those who don’t want to see the needle or people who have dexterity issues. Third, the plastic 30-45° sets might be better for people who are prone to infection, who are lean and muscular, or who are pregnant.
  • Injections are still our friends. Ms. Block reminded the audience that sometimes, insulin pumps don’t work – there could be malabsorption, insulin spoiling, or kinks in the infusion site, just to name a few. Consequently, it is important for patients to test their blood glucose frequently. She also encouraged blood ketone testing anytime someone on insulin pump therapy feels DKA symptoms, is ill, or when blood glucose is high and is not responsive to correction doses. When a patient is concerned something is awry with their pump, he or she should change everything, she said. And, echoing Ms. Evert, she reminded us that even insulin pump users will need to use injections every once in a while when their pump fails.
  • Pumps have great new features, although there are few studies to guide the use of these features. Ms. Block highlighted extended/square wave boluses (ideal for low carb meals high in fat and protein) and the dual wave bolus feature (ideal for a carb containing meal that is also high in fat and protein). These capabilities increase patients’ ability to fine-tune their insulin dosing, although more data is needed to optimize their use.
  • Ms. Block spoke to new advances in pump therapy, the marriage of CGM and insulin pump therapy, and moving towards a completely closed loop. Without going into too much detail on each product, she briefly mentioned new advances in durable pump therapy (Tandem’s t-slim, Asante’s Pearl, and Roche’s Accu-Chek Combo), patch pumps (“for whom traditional tethered pumps is not their cup of tea” – Debiotech’s Jewel, Cellnovo, Valeritas’ V-Go, Insulet’s OmniPod, and Roche’s Solo), automated pump suspension (Medtronic MiniMed 530G), and remote monitoring (Medtronic’s mySentry). She encouraged the audience to visit the exhibit hall for more information on each product.

 

INSULIN PUMP THERAPY – IT’S A CONTINUUM OF CARE

Alison Evert, MS, RD, CDE (University of Washington Medical Center, Seattle, WA)

Ms. Evert began her presentation by considering the available evidence to guide practice for insulin pump therapy. While significant evidence supports the use of CSII in people with type 1 diabetes, within the pediatric population, there is not as much evidence to guide practice. As for people with type 2 diabetes, she noted that clinical evidence was mixed at this point. Next, Ms. Evert asked why patients should consider pump therapy. She reflected on the burden that providers place on patients to calculate their insulin dose while on MDI – a multi-step process that includes counting carbs, dividing carbs by the insulin to carb (ICR) ratio, checking blood sugar, subtracting blood sugar from glycemic target, and delivering the appropriate insulin dose – and suggested that one of the primary benefits of insulin pumps was that the bolus calculator simplified this process. For patients switching to CSII, CDE’s play an incredibly important role. First, she explained, CDE’s need to engage with patients to set realistic goals for therapy. Next, she encouraged providers to invest the necessary time to help patients build their foundation of knowledge – CSII only works as well as the operator. She also emphasized that we need to help patients “think like a pancreas” (a nod to moderator Gary Scheiner’s well-known book), meaning understanding insulin action (onset, peak, duration, and insulin-on-board) and the effect of nutrition on insulin demands. In addition to developing patient knowledge, she implored providers to give ongoing support – a critical component to successful diabetes care. Ms. Evert also presented data from a continuous quality improvement (CQI) project at the Diabetes Care Center at the University of Washington Medical Center, which gave a fascinating glimpse into just how much time CDEs spend on the phone and what areas require the most time investment.

  • The CQI project tracked non-reimbursable CDE and medical assistant time for four months at the Diabetes Care Center at the University of Washington Medical Center. The patient population consisted of 2,224 individuals: 50% with type 1 diabetes, 82% on insulin, and 26% on insulin pumps.
  • Data showed that the mean time per call was 13.4 minutes, which translated to 2.9 hours per day of non-reimbursable time for a provider. While 64% of the time CDEs and MAs could resolve the problem, 37% of the time MDs or ARNPs needed to assist. Pump malfunction calls required a minimum of 20 minutes per call.
  • Most patients calling for pump malfunction problems lacked any back up plan for insulin delivery. Ms. Evert explained that many did not know what to do or have the means to do it – lacking long acting insulin, current prescriptions, syringes, pen needles, or basal or bolus settings recorded outside of their pump. Notably, many had not seen a CDE in years. In terms of pump malfunction and diabetic ketoacidosis (DKA) prevention, many could not recall DKA prevention procedures nor did they possess ketone test strips. These findings inspired Ms. Evert and the Diabetes Care Center to implement several important changes in their practice: 1) educators provide a wallet card or pump setting report at every visit; 2) the center created a DKA prevention handout specific to pump DKA; 3) long-term pumpers are encouraged to visit annually with their CDE (and at a bare minimum patients must return for an education session to receive a pump upgrade); and 4) patients not using personal CGM are scheduled annually for retrospective CGM.

 

BEST OF THE BEST PRACTICES

Davida Kruger, MSN, APN-BC, BC-ADM (Henry Ford Health System, Detroit, MI)

Ms. Kruger gave a comprehensive overview of the insulin pump program at the Henry Ford Health System. She addressed essential elements in individual patient success, pre-pump classes, selecting a pump, training, warranties, and many other topics. The major emphasis throughout was comprehensive education, setting patient expectations properly, and staying up-to-date on the current pumps on the market. Ms. Kruger emphasized that pump programs are not all or nothing – she encouraged the audience to pick what works, to partner with other clinics, other educators, and even pump companies.

  • Ms. Kruger reviewed some of the essential elements in individual patient success with pumps: discussion and understanding of the importance of SMBG; realistic expectations (she still gets people saying, “I was sent here to get an insulin pump. When can we do the surgery?”); a pump should be the patient’s choice; close work with a dietitian; carb counting; a partnership with the health care team; and a technical understanding of the pump and pump therapy.
  • The pre-pump class at the Henry Ford Health System is taught by a nurse practitioner and dietitian. It’s a small class with three to four patients/family members, with an agenda designed to assist them in making the best choices.
  • To help patients decide on a pump, Ms. Kruger emphasized that educators should help guide the patient and it’s “very important” to know every pump on the market. Speaking to the latter, she encouraged all educators to take advantage of the exhibit hall. Ms. Kruger noted that her clinic helps patients complete the requested paper work and C-peptide requirements for the Henry Ford HMO (Medicare has a similar requirement). She also emphasized that “cost is a consideration,” so her team makes sure to inform patients of the upfront and ongoing costs of pump therapy.
  • New pump patients get lots of upfront education and follow-up at the Henry Ford Health System. Patients call to schedule a pump training, which typically takes place within a week. Prior to training, they are instructed to look at the manual, DVD, web training, and tools. Ms. Kruger was adamant that patients should not start using the pump without receiving formal, in-clinic training. Company trainers usually spend about four hours training new pump patients, which is then followed by another two hours with the clinic’s nurse practitioner. Patients return in two days, one week, one month, two months, and then every three months – we thought this was really impressive follow-up and a great way to manage patients’ expectations early on. Upon starting the pump, patients are instructed to fax/email blood glucose levels one to two times weekly for one to two months. They are also encouraged to come back to the nurse practitioner at least once per year.
  • “Warranties for pumps are usually four years. Insurance companies usually reimburse within five years.” At her clinic, they typically wait until there’s a problem with the pump until a decision is made to upgrade. One of the main reasons is that waiting to upgrade as long as possible ensures that patients will have the latest technology. She noted that a lot of times pumps last seven to eight years. Interestingly, her patients are put on a “no-call” list, which means pump companies cannot call them when their warranties are up.

 

PANEL DISCUSSION

Gary Scheiner, MS, CDE (Integrated Diabetes Services, Wynnewood, PA); Davida Kruger, MSN, APN-BC, BC-ADM (Henry Ford Health System, Detroit, MI); Jennifer Block, MSN, RN, CDE (Stanford University, Stanford, CA); and Alison Evert, MS, RD, CDE (University of Washington Medical Center, Seattle, WA)

Q: Will the slides be available online?

Mr. Scheiner: AADE is putting a video presentation of this program together. It will be available online in the coming weeks. [Applause; Editor’s note: at time of publication a video presentation of this program was unavailable online.]

Q: What are the minimum requirements before you start a patient on a pump?

Ms. Kruger: Blood glucose monitoring is a good place to start. Four injections a day. An understanding that it is an external device that will be your new best friend. The importance of follow-up care and the risks of DKA.

Ms. Block: Ultimately, the expectations with this therapy are huge. We need to be very realistic about it. It’s important to have an expectation of frequent glucose monitoring. Also, when people are not in stellar glycemic control and they come in and really want a pump – and they believe it will make a difference – it’s important to consider them too. I think we need to play off their motivation.

Ms. Evert: The main time patients touch their pump throughout the day is for bolusing for food, so some method of carb counting – either small, medium, and large, or grams of carb.

Q: What’s the number one reason for DKA in pump use? How do you prevent it?

Ms. Kruger: The person wearing the pump doesn't realize their infusion set is dislodged, and it’s been a long time since they checked their blood sugar.

Ms. Block: Pump therapy is fairly reliable. So people are used to this therapy working and they’re not used to it not working. The frequency of events where it doesn't work is far apart. So when I see an emergency or problem, typically a failure plan hasn’t been developed.

Ms. Evert: It’s not the $7,000 piece of equipment. It’s the least expensive component – the adhesive on the tape.

Ms. Kruger: This speaks to the importance of education. If you have an elevated blood glucose, do not ignore it. Change the site. I have been impressed – having used pumps for 30 years – how much they’ve improved and our knowledge has improved. We have less DKA and less site infections.

Q: How do you get paid for pre- and post-pump education?

Ms. Kruger: We have four NPs in our group, so we can bill.

Ms. Evert: We use facility building in our institution. In previous places, we billed it as diabetes self- management education. Many people haven’t been through the ten-hour course, so that’s how we would bill it.

Q: Do you consider carb counting and glycemic index important topics to cover?

Ms. Evert: Yes, it’s job security for dietitians.

Ms. Kruger: It’s the cornerstone. Every person with diabetes deserves to see a dietitian as part of his or her therapy.

Q: How do you convince physicians that an insulin pump isn’t appropriate for every person?

Ms. Kruger: Some things you can’t change.

Ms. Block: It’s all about the way you communicate. Ms. Kruger: And having a big enough baseball bat.

Ms. Block: As educators we really advocate for patients and there are individuals for whom we especially need to advocate for to providers.

Ms. Kruger: As pumps become more user-friendly, people who weren’t previously pump candidates now might be.

Ms. Evert: It’s about building good relationships with physicians and nurse practitioners you work with so that they trust you. This is especially important when you go back to them and say you have grave reservations about a certain patient on a pump. We might discover during our time with patients something that the physician hasn’t been able to discover.

Q: Do you do group visits for pre and post insulin pump starts?

Ms. Kruger: Some of our education with the pump trainers is in groups and then we go to the individual.

Ms. Evert: Our manager wants us to do education as a group because people learn so much from each other, but we’ve had too many issues with different learning styles to do group education for pump starts.

Ms. Block: It all comes down to the individual. Some do well in groups; some do not. If you have people who will do well as a group, it has a tremendous amount to offer. They go on to help and learn from each other outside of the group setting.

Q: How can hospitals safely manage patients on pumps?

Ms. Block: It’s challenging to educate hospitals as a whole on pump therapy. In our hospital what it comes down to is that the patient or family member needs to be qualified to use a therapy. Management falls on individuals rather than training the whole staff.

Ms. Kruger: If a patient is on a pump, that’s an automatic consult to an endo. We have a new protocol that I’m excited about for surgery that looks at the amount of time for the surgery and decides whether the pump stays on or comes off. And it instructs surgeons and other practitioners in the room how to manage the pump during surgery.

Ms. Evert: We are working with our glycemic management team – while all the team members were great clinicians, they were not as savvy with pumps.

Q: Is it safe to insert pumps into breast tissue?

Ms. Block: I have no experience in doing it. Ms. Kruger: I wouldn’t want that experience.

Ms. Block: I’m not sure why that would be the option when we do have a fair number of injection states. Ms. Kruger: Perhaps if you were placing the pump in your bra it would make sense.

Q: What are your thoughts on pump use for gestational diabetes?

Ms. Block: I feel very strongly that it is a great option for people at any point in pregnancy. It does require a fair amount of teaching, but so does gestational diabetes. Intensive glucose control has been shown to make a tremendous difference. For everyone one person you are seeing, you are treating two.

Ms. Kruger: For gestational diabetes, we often see it so late that by the time we have the equipment and provide the adequate training it would be a problem. I agree people with type 1 and type 2 diabetes should use pumps during pregnancy, but not so for gestational diabetes. And it is very easy to control gestational diabetes with injections.

Q: How long can a pump be safely disconnected before insulin needs to be given?

Ms. Block: This is challenging for pump users and we need to talk about situations with patients where other forms of insulin in addition to pump therapy need to be used.

Ms. Evert: Two hours is our frame of reference.

Ms. Kruger: Patients can come off of pumps, but if its going to be a day or two they should have back up long acting insulin. People want breaks, but more than two hours make me nervous.

Q: Are there challenges to using U-500 in pumps?

Ms. Kruger: It can be done off label, but you have to be very focused to do it.

Ms. Evert: There are a lot of insulins in the pipeline and coming to market. I want to make sure providers have a good understanding of the insulins and that the patient knows to check, check, check…

Ms. Kruger: It can be done and it can be done successfully, it just has to be done very carefully.

Q: Under what conditions would you rip a pump off a patient?

Ms. Kruger: Patients come in and have not changed infusions sites for seven or eight days. Are you kidding me? There are things for safety, like not doing any blood glucoses or not changing infusions sets.

Ms. Block: Like anything else, we need to weigh the risks and benefits.

Ms. Evert: When we download the pump and 100% of the delivery is basal and they’re not bolusing. We’ve had that scenario.

Q: Can people with low literacy be successful pumpers?

Ms. Evert: Absolutely. Pumps are great for literacy. Putting those algorithms in for prandial and corrections – it’s a lifesaver for reducing highs and lows.

Q: Which company makes the best and most user-friendly pump?

Mr. Scheiner: I’m sorry, but our time is up. [Laughter] I encourage you all to visit the pump companies in the exhibit hall.

 

-- by Adam Brown, Cati Crawford, Hannah Deming, Jessica Dong, Benjamin Kozak, Kira Maker, Nina Ran, Tanayott Thaweethai, Katrina Verbrugge, Vincent Wu, David Zhang, and Kelly Close