O R I G I N A L A R T I C L E
`
`Ability to handle, and patient preference for,
`insulin delivery devices in visually impaired
`patients with type 2 diabetes
`
`Charles Fox*, Carol McKinnon, Anthony Wall, Simon A Lawton
`
`ABSTRACT
`In this comparative study, 86 patients with type 2 diabetes and visual impairment were evaluated on their preference for, and ability
`to operate, three different insulin delivery systems – InnoLet® (NovoNordisk), Humulin® NPH Pen (Eli Lilly) and vial and syringe
`(Becton Dickinson). Patients found the clocklike dose scale on InnoLet® significantly easier to read than that of the other systems
`(92% were able to read four doses, versus 45% and 61% with Humulin® Pen and syringe respectively, both p < 0.001), and showed
`greater ability to set and dispense a 20 unit dose without instruction. After reading the packaging information leaflet and brief ver-
`bal instruction, 99% of patients were able to correctly set and dispense three consecutive insulin doses with InnoLet® compared with
`85% with Humulin® Pen and 64% with syringe (p = 0.001 and p < 0.001, respectively). Ability to set and deliver insulin was not
`however clearly related to visual acuity. On questionnaire, 87% of patients expressed an overall preference for InnoLet®, and 13%
`for Humulin® Pen (p < 0.001); no patients preferred the syringe. In conclusion, insulin delivery systems designed to simplify accu-
`rate, reliable insulin delivery for people with visual impairment can improve the ability of such patients to repeatedly set and deliver
`the correct insulin dose. Copyright © 2002, John Wiley & Sons, Ltd
`
`Practical Diabetes Int, 2002; 19(4): 104–107
`
`KEY WORDS
`type 2 diabetes; insulin delivery system; visual impairment
`
`Introduction
`Type 2 diabetes is a progressive disorder,
`associated with a steady decline in beta-cell
`function, from a mean 51% at diagnosis,
`to 28% after 6 years1. This reduction in
`glucose-lowering capacity necessitates
`introduction of insulin in the majority of
`patients, within 15 years of diagnosis2. By
`lowering HbA1c through the use of inten-
`sive therapy, morbidity associated with
`
`Charles Fox, BM FRCP Consultant
`Physician, Northampton General Hospital,
`Northampton, UK
`
`Carol McKinnon, MBChB, MRCGP,
`DRCOG, General Practitioner, Castlemilk
`Group Practice, Glasgow, UK
`
`Anthony Wall, LRCP MRCS MB.BS DCH,
`General Practitioner and Clinical Trials
`Coordinator, St John’s Health Centre,
`Woking, UK
`
`Simon A Lawton, BHB, MBChB,
`International Medical Advisor, Novo
`Nordisk A/S, 2880 Bagsvaerd, Denmark
`
`*Correspondence to Dr Charles Fox,
`Northampton General Hospital, Cliftonville,
`Northampton NN1 5BD, UK
`
`Received: 28 June 2001
`Accepted: 7 January 2002
`
`microvascular, and probably macrovascu-
`lar, complications can be significantly
`reduced3,4. Any reduction in HbA1c, how-
`ever small, contributes to improving prog-
`nosis3. Nonetheless, many eligible patients
`are denied the benefits of timely insulin
`therapy due to misconceptions about
`appropriate use of insulin in type 2 dia-
`betes, and anxiety concerning the practical
`and organisational skills required to follow
`complex dosing regimens. From the
`patients’ perspective the anticipated pain
`and social stigma of injections compound
`the delay in receiving optimal glucose-low-
`ering therapy5.
`Current insulin delivery systems vary in
`format and ease of use. Insulin is available
`in vials, to be drawn up using a conven-
`tional syringe, but this method provides
`poor dose accuracy6. Patients who experi-
`ence particular difficulty in accurate self-
`dosing include the elderly, as well as
`younger people with poor vision or
`impaired manual dexterity. Visual impair-
`ment is common in diabetes, with acuity
`6/12 or worse in 16% of people over the
`age of 65, rising to nearly 27% by age 757.
`Thus, insulin delivery systems that are easy
`to use by people with poor vision are essen-
`tial for effective diabetes management, and
`to allow patients to maintain their inde-
`pendence. Prefilled insulin cartridges,
`
`inserted into disposable or durable insulin
`delivery devices with short needles, can
`improve dose accuracy and increase com-
`pliance rates compared with more tradi-
`tional methods, and quality of life is often
`enhanced through their use8–11. Since
`patients with type 2 diabetes perform sig-
`nificantly worse than age-matched controls
`on tests of learning and reasoning12, this
`patient group – of whom over 50% will
`require daily insulin therapy at some
`stage13 – may derive particular benefit from
`delivery systems that are easy to handle and
`simple to operate. It is anticipated that
`rational insulin delivery systems that make
`maximum use of existing sensory capacity
`will be preferred by patients, and will help
`those with sensory impairment capitalise
`on their remaining abilities.
`
`Materials and methods
`In this multicentre, open, randomised,
`comparative
`study, 86
`insulin-naive
`patients with type 2 diabetes were drawn
`from three centres in the UK. During a
`handling test, patients tested three insulin
`delivery systems – InnoLet® (Novo
`Nordisk, Denmark), Humulin® NPH Pen
`(Eli Lilly, USA), and vial and 0.5 ml
`(Becton Dickinson, USA).
`syringe
`NovoFine® 30G 8 mm needles (Novo
`Nordisk) were used with the prefilled pen
`
`104 Pract Diab Int May 2002 Vol. 19 No. 4
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
` PFIZER, INC. v. NOVO NORDISK A/S - IPR2020-01250, Ex. 1041, p. 1 of 4
`
`

`

`O R I G I N A L A R T I C L E
`
`Insulin delivery devices in the visually impared
`
`devices; the syringe had a pre-attached
`Micro-Fine + 8 mm needle. The study was
`conducted in accordance with good clini-
`cal practice.
`
`Patients
`Patients were included if they were ≥55
`years with diet- and/or oral-hypogly-
`caemic-agent-treated type 2 diabetes. Since
`one of the study objectives was to assess
`the ease with which patients could learn
`how to operate the respective delivery sys-
`tems, people with a history of insulin self-
`administration were excluded to avoid
`potential bias from previous exposure to
`the systems tested. Visual acuity (corrected
`near vision) in the best eye was between
`0.5 (20/40) and 0.1 (20/200) as assessed
`with a Rosenbaum card.
`
`Assessment methods
`Accuracy when viewing dose
`scale
`Ability to administer insulin safely and
`independently relies on repeated accurate
`dose reading. Patients were therefore asked
`to read four randomly selected whole-
`numeral dose settings on each insulin
`delivery system under standardised direct
`illumination.
`
`Handling test
`The handling test comprised three parts.
`
`Part I: intuitive test
`Patients were asked to set and discharge a
`20 unit insulin dose with a minimum of
`standardised instruction and no specific
`training, using each of the three delivery
`systems. The time allowed to complete the
`task was 2 min. Patients were assessed as
`successful in completing this section if the
`dose was set within 5% of that requested
`and the full dose dispensed. In this, and all
`other parts of the study, doses were deliv-
`ered into a disposable container.
`
`Part II: written instruction
`Patients were provided with those sections
`of the manufacturers’ package information
`leaflets giving details on how to operate the
`device. After a maximum of 3 min reading
`time, participants set and dispensed three
`doses of insulin. Doses were randomly
`selected (dose range 4–50 units, with one
`dose > 30 units) but consistent between
`devices and patients (dose 1 = 23 units,
`dose 2 = 42 units, dose 3 = 17 units), and
`dose setting was deemed accurate if it fell
`within 1 unit for doses < 20 units and
`within 5% for doses > 20 units. Successful
`
`completion of this section required all three
`doses to be correctly set and delivered.
`
`Part III: verbal instruction and
`demonstration
`Patients only entered this section of the
`trial if they were unable to complete the
`previous
`section
`successfully. Verbal
`instruction and a demonstration on how
`to operate the system correctly were pro-
`vided, and a 5 min limit was set. The
`patients were then requested to set and dis-
`pense three further randomly selected
`doses (dose 1 = 40 units, dose 2 = 29 units,
`dose 3 = 19 units). Successful completion
`of this section was assessed on the same cri-
`teria as in part II.
`
`Questionnaire
`On finishing the handling test, partici-
`pants completed a questionnaire on their
`preference for certain features of each
`delivery system. Only the question on
`overall preference required both qualitative
`and quantitative answers.
`
`Data analysis
`Setting sample size to a minimum of 80
`patients ensured that group differences
`would be detected with a power of 85% at
`a 0.05 level of significance (a = 5%). The
`global difference between delivery systems
`was calculated with Fisher’s exact test using
`SAS© proc freq. If global difference was
`significant (p ≤ 0.05), all three pairwise
`comparisons between devices were made
`using Fisher’s exact test. In the case of pair-
`wise comparisons a Bonferroni correction
`to
`the value was applied. Because
`InnoLet® was tested twice, the threshold
`value was lowered from 5% to 2.5%.
`
`Results
`Study population
`characteristics
`The study population comprised 86
`patients (51 male, 35 female), mean age 69
`years and mean disease duration 6 years.
`Visual acuity ranged between 0.29 and 0.5
`in 46 patients (54.1%), 0.2 and 0.29 in 17
`patients (20%) and 0.1 and 0.2 in 22
`(25.9%) patients. Only five (5.8%)
`reported a subjective hand disability.
`
`Withdrawal and non-
`completion of handling test
`No patients withdrew prematurely from
`the trial, and only two subjects (2.3%)
`were unable to complete the handling test.
`In both cases this was due to inability to
`handle the vial and syringe.
`
`Figure 1. More patients were able to read
`four consecutive doses settings with
`InnoLet® than with Humulin® Pen or
`syringe
`
`Visual accuracy when reading
`dose scale
`The majority of patients could read all
`four doses correctly with InnoLet®
`(79 patients, 92%), compared with 39
`(45%) with Humulin® Pen, and 52 (61%)
`with the syringe (Figure 1). The differences
`between InnoLet® and both the other
`devices were highly statistically significant
`(p < 0.001 for both). The number of
`patients in each visual acuity group who
`were able to read all four doses correctly
`decreased with decreasing acuity for all
`devices, but the decline was less apparent
`with InnoLet® than with Humulin® Pen
`or syringe.
`
`Handling test
`Intuitive test
`Eighty-three patients (97%) set the 20 unit
`dose correctly (i.e. not evaluating ability to
`dispense the dose) with InnoLet®, com-
`pared with 41 (48%) and 33 (39%) with
`Humulin® Pen and syringe, respectively.
`Following this, 72 patients (84%) also suc-
`ceeded in dispensing the complete 20 unit
`dose with InnoLet®, compared with 35
`(41%) with Humulin® Pen and 27 (31%)
`with the syringe (p < 0.001 between
`InnoLet® and both Humulin® Pen and
`syringe for setting and dispensing; Table 1).
`Mean time taken to set and dispense the
`correct dose was significantly shorter with
`InnoLet® (26 s) than with Humulin® Pen
`(65 s, p < 0.001 versus InnoLet®) or
`syringe (53 s, p < 0.001 versus InnoLet®;
`Figure 2). There was no clear relationship
`between visual acuity and ability to set and
`dispense the correct dose.
`
`Written instruction
`After receiving written guidance, 83 patients
`(97%) correctly set all three insulin doses
`using InnoLet®, while 69 (80%) set and dis-
`pensed all doses accurately (Table 1). This
`compared with 56 (65%) and 24 (28%)
`
`Pract Diab Int May 2002 Vol. 19 No. 4
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`105
`
` PFIZER, INC. v. NOVO NORDISK A/S - IPR2020-01250, Ex. 1041, p. 2 of 4
`
`

`

`O R I G I N A L A R T I C L E
`
`Insulin delivery devices in the visually impared
`
`Table 1. InnoLet® is the most reliable device for accurate dose-setting and insulin delivery
`
`InnoLet® Humulin® Pen
`
`Syringe
`
`84a,b
`
`80a,c
`
`99a,b
`
`41
`
`61
`
`85
`
`31
`
`27
`
`64
`
`% of patients able to
`intuitively set and
`dispense 20 U insulin dose
`
`% of patients able to set
`and dispense three insulin
`doses after written
`instruction
`
`% of patients able to set
`and dispense three insulin
`doses after both written
`instruction and brief verbal
`instruction/demonstration
`a p < 0.001 versus syringe.
`b p ≤ 0.001 versus Humulin® Pen.
`c p < 0.01 versus Humulin® Pen.
`
`between InnoLet® and syringe). Again,
`there was no clear relationship between
`visual acuity and ability to use the devices
`correctly.
`
`Questionnaire
`InnoLet® was the preferred delivery system
`of 75 (87%) patients, compared with 11
`patients (13%) who chose Humulin® Pen
`(p < 0.001); no patients preferred the
`syringe. All the specific device features
`assessed were rated as significantly superior
`for InnoLet® (p < 0.001 where applicable;
`Figure 3). All five patients with subjective
`hand impairment found that the button
`on InnoLet® was the easiest to depress
`during injection. Four patients found
`InnoLet® the easiest to hold while setting
`a dose, while one preferred Humulin® Pen
`in this respect. Two found InnoLet® the
`easiest to hold while injecting, whereas two
`preferred Humulin® Pen. Formal statisti-
`cal testing of these differences was omitted
`because of the small sample size.
`
`Discussion
`The need to self-inject can be a barrier to
`starting insulin therapy at any age. Many
`people with type 2 diabetes are introduced
`to insulin self-injection during their more
`advanced years, when confidence in learning
`new skills is declining and both physical and
`psychological limitations add to the burden
`of performing complex tasks. If patients
`believe that assistance from caregivers will be
`necessary, reluctance to be dependent may
`be an additional barrier to starting therapy5.
`Visual impairment – most commonly
`due to cataract – affects 16% of people
`with type 2 diabetes over the age of 65, and
`
`Figure 2. Patients were intuitively able to
`set and dispense insulin significantly more
`quickly using InnoLet® than Humulin®
`Pen or syringe
`
`who set the dose correctly using Humulin®
`Pen and syringe respectively, and 52 (61%)
`and 23 (27%) who dispensed insulin accu-
`rately. Significantly more patients were suc-
`cessful operating InnoLet® compared with
`the other two delivery systems (dose-setting,
`p < 0.001 versus both other systems; dose-
`setting plus delivery, p = 0.007 versus
`Humulin® Pen, p < 0.001 versus syringe).
`Only when using the syringe did decreasing
`visual acuity impair patients’ ability to set
`and fully dispense the correct dose.
`
`Verbal instruction and
`demonstration
`Patients who entered part III of the trial
`were those unable to successfully complete
`part II. InnoLet® accounted for 17
`patients, Humulin® Pen for 34 and syringe
`for 62. After verbal instruction and
`demonstration, 17 (100%) correctly set all
`doses with InnoLet®, compared with 23
`(68%) and 31 (50%) with Humulin® Pen
`and syringe, respectively (p = 0.009 for the
`
`106 Pract Diab Int May 2002 Vol. 19 No. 4
`
`difference with Humulin® Pen; p < 0.001
`for syringe). Following this, 94% (n = 16)
`of patients correctly dispensed all three
`insulin doses with InnoLet®, compared
`with 62% (n = 21) and 50% (n = 31) with
`Humulin® Pen and syringe, respectively (p
`= 0.019 between InnoLet® and Humulin®
`Pen, and p < 0.001 between InnoLet® and
`syringe; Table 1). The proportion of sub-
`jects who set and fully dispensed all three
`doses correctly showed no clear relation-
`ship with visual acuity.
`Following written or verbal instruction
`and demonstration, insulin was accurately
`set and dispensed by 85 (99%) patients
`using InnoLet®, compared with 73 (85%)
`and 54 (64%) with Humulin® Pen and
`syringe, respectively (p = 0.001 between
`InnoLet® and Humulin® Pen, p < 0.001
`
`Figure 3. Patients preferred InnoLet® to Humulin® Pen or syringe
`% of patients
`10 20 30 40 50 60 70 80 90 100
`
`0
`
`Easiest number scale to read
`
`Easiest to set insulin
`
`Simplest to operate
`
`Easiest to learn how to use
`
`Easiest button for injection
`
`Easiest to hold when setting dose
`
`Easiest to hold while injecting
`
`Feel most confident about
`
`Overall preference
`
`InnoLet®
`Humulin®
`Pen
`Syringe
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
` PFIZER, INC. v. NOVO NORDISK A/S - IPR2020-01250, Ex. 1041, p. 3 of 4
`
`

`

`O R I G I N A L A R T I C L E
`
`Insulin delivery devices in the visually impared
`
`Key points
`● Visual impairment is common in people with type 2 diabetes.
`● Accurate, reliable insulin delivery is facilitated by devices that are tailored to
`physical and psychological ability.
`● With only minimal instruction – of a duration reasonable within a busy clinic
`– the majority of patients, including those with severe visual impairment,
`were able to demonstrate the ability to repeatedly set and dispense insulin
`accurately using InnoLet®.
`● In comparison with Humulin® Pen and syringe, patients could accurately set
`and dispense insulin more reliably with InnoLet®.
`● Patients became competent in using InnoLet® more rapidly, preferred the
`design features of InnoLet® and felt more confident about using it.
`
`this figure rises to 27% at age 757.
`Repeated setting and delivery of the correct
`insulin
`dose
`poses
`a
`challenge.
`Furthermore, over 50% of people with type
`2 diabetes have limited joint mobility in
`the hands14, and 25% have symptomatic
`peripheral neuropathy15. Arthritis and
`tremor increase with advancing age16,
`adding to the difficulty in manually operat-
`ing handheld devices. In order for patients
`to be willing to self-inject insulin regularly,
`both regimen and device should be tailored
`to physical and psychological ability.
`In this study of 86 patients with type 2
`diabetes and visual impairment, patients
`were able to read the InnoLet® dosing
`scale more reliably than that of the other
`systems. Although accuracy fell with
`reduced visual acuity, this was least pro-
`nounced with InnoLet®, and over 80% of
`patients with severe visual impairment
`(<0.2 to ≥0.1) were still able to read the
`dose scale, compared with less than 20%
`with Humulin® Pen and 41% with
`syringe. At every level of instruction, a
`greater number of patients were able to
`accurately set and dispense insulin using
`InnoLet®. In the absence of specific train-
`ing, 84% set and administered the dose
`reliably, compared with 41% and 32%
`with Humulin® Pen and syringe, respec-
`tively. With only minimal training, in the
`form of written instruction or a 5 min
`explanation, 99% became proficient.
`Overall, 87% of patients preferred
`InnoLet® to Humulin® Pen or the syringe.
`
`Ratings for specific design features were
`significantly higher for InnoLet® across all
`eight categories assessed.
`Since people with type 2 diabetes are sig-
`nificantly more likely to have difficulty per-
`forming tasks that require recall of learned
`skills12,17,18, one could speculate that the
`patient preference for InnoLet® demon-
`strated in this study rests on design features
`that address the specific patient needs asso-
`ciated with type 2 diabetes. The large clock-
`like dial, which is set like a familiar kitchen-
`timer, and large-scale numbers assist accu-
`rate dose selection; audible clicks accom-
`pany delivery of each insulin unit dialled,
`reassuring patients about dose selection.
`This may account for the greater number of
`patients in the handling study being able to
`repeatedly set and dispense insulin with
`InnoLet® (99%) than were able to read the
`dose scale without error (92%). Although
`the preference of the five patients with sub-
`jective hand impairment in this study
`tended towards InnoLet®, there were too
`few patients to provide conclusive results.
`However, it seems logical that the comfort-
`able handling of the device and large dose-
`delivery buttons would simplify insulin
`delivery for patients with weak hands.
`In conclusion, the rational design of
`InnoLet® renders it a device suitable for
`accurate, reliable delivery of insulin, even
`in patients with visual impairment. Given
`a choice, the majority of patients favour
`InnoLet® over currently available alterna-
`tives.
`
`Conflict of interest: This research project was funded by NovoNordisk and Charles Fox
`received an honorarium for taking part in the launch of the InnoLet® injection device.
`However, he has no ongoing commercial interest in this product.
`
`References
`1. UK Prospective Diabetes Study Group. UK
`Prospective Diabetes Study 16. Diabetes 1995; 44:
`1249–1258.
`2. Clauson P, Linnarsson R, Gottsater A et al.
`Relationships between diabetes duration, metabolic
`control and beta-cell function in a representative pop-
`ulation of type 2 diabetic patients in Sweden. Diabet
`Med 1994; 11: 794–801.
`3. UK Prospective Diabetes Study (UKPDS) Group.
`Intensive blood glucose control with sulphonylureas
`or insulin compared with conventional treatment and
`the risk of complications in patients with type 2 dia-
`betes (UKPDS33). Lancet 1998; 352: 837–853.
`4. Ohkubo Y et al. Intensive insulin therapy prevents the
`progression of diabetic microvascular complications
`in Japanese patients with NIDDM. Diabetes Res Clin
`Pract 1995; 28: 103–117.
`5. Bashoff E, Beaser R. Insulin therapy and the reluctant
`patient. Postgrad Med 1995; 97: 86–90.
`6. Puxty J, Hunter D, Burr W. Accuracy of insulin injec-
`tion in elderly patients. BMJ 1983; 287: 1726.
`7. Klein R, Klein B, Moss S. Visual impairment in dia-
`betes. Ophthalmology 1984; 91; 1–9.
`8. Hörnquist J, Wikby A, Andersson P, Dufva A.
`Insulin-pen treatment, quality of life and metabolic
`control:
`retrospective
`intra-group evaluations.
`Diabetes Res Clin Pract 1990; 10: 221–230.
`9. Paterson K, Sandler D. Pen injectors – the way for-
`ward? Diabet Med 1992; 9: 684.
`10. Coscelli C, Lostia S, Lunetta M et al. Safety, efficacy,
`acceptability of a pre-filled insulin pen in diabetic
`patients over 60 years old. Diabetes Res Clin Pract
`1995; 28: 173–177.
`11. Corsi A et al. Pre-filled insulin pen in newly insulin-
`treated diabetic patients over 60 years old. Diab Nutr
`Metab 1997; 10: 78–81.
`12. Reaven G, Thompson L, Nahum D, Haskins E.
`Relationship between hyperglycaemia and cognitive
`function in older NIDDM patients. Diabetes Care
`1990; 13: 16–21.
`13. Barnett A. Tablet and insulin therapy in type 2 dia-
`betes in the elderly. J Royal Soc Med 1994; 87:
`612–614.
`14. Starkman H, Gleason R, Rand L et al. Limited joint
`mobility (LJM) of the hand in patients with diabetes
`mellitus: relation to chronic complications. Ann
`Rheum Dis 1986; 45: 130–135.
`15. Ziegler D, Gries F, Spuler M et al. The epidemiology
`of diabetic neuropathy. Diabet Med 1993; 10:
`82S–86S.
`16. Frey M, Barrett-Connor E, Sledge P et al. The effect
`of non insulin dependent diabetes mellitus on the
`prevalence of clinical osteoarthritis. A population-
`based study. J Rheumatol 1996; 23: 716–722.
`17. Strachan M, Diary I, Ewing F, Frier B. Is type II dia-
`betes associated with an increased risk of cognitive
`dysfunction? A critical review of published studies.
`Diabetes Care 1997; 20: 438–445.
`18. Cosway R, Strachan MWJ, Dougal A et al. Cognitive
`function and information processing in type 2 dia-
`betes. Diabet Med 2001; 18: 803–810.
`
`Pract Diab Int May 2002 Vol. 19 No. 4
`
`Copyright © 2002 John Wiley & Sons, Ltd.
`
`107
`
` PFIZER, INC. v. NOVO NORDISK A/S - IPR2020-01250, Ex. 1041, p. 4 of 4
`
`