DIABETICMedicine

DOI: 10.1111/j.1464-5491.2012.03605.x

Article: Care Delivery

Insulin adherence behaviours and barriers in the
multinational Global Attitudes of Patients and Physicians
in Insulin Therapy study

M. Peyrot1,2, A. H. Barnett3, L. F. Meneghini4 and P.-M. Schumm-Draeger5

1
Department of Sociology, Loyola University of Maryland, 2Department of Medicine, Johns Hopkins University, Baltimore, MD, USA, 3University of Birmingham
and BioMedical Research Unit, Heart of England NHS Foundation Trust, Birmingham, UK, 4Department of Medicine, University of Miami, Miami, FL, USA and
5
Department of Medicine, Bogenhausen Academic Teaching Hospital, Munich, Germany

Accepted 31 January 2012

Abstract
Aims To examine patient and physician beliefs regarding insulin therapy and the degree to which patients adhere to their
insulin regimens.
Methods Internet survey of 1250 physicians (600 specialists, 650 primary care physicians) who treat patients with diabetes
and telephone survey of 1530 insulin-treated patients (180 with Type 1 diabetes, 1350 with Type 2 diabetes) in China,
France, Japan, Germany, Spain, Turkey, the UK or the USA.
Results One third (33.2%) of patients reported insulin omission ⁄ non-adherence at least 1 day in the last month, with an
average of 3.3 days. Three quarters (72.5%) of physicians report that their typical patient does not take their insulin as
prescribed, with a mean of 4.3 days per month of basal insulin omission ⁄ non-adherence and 5.7 days per month of prandial
insulin omission ⁄ non-adherence. Patients and providers indicated the same five most common reasons for insulin omis-
sion ⁄ non-adherence: too busy; travelling; skipped meals; stress ⁄ emotional problems; public embarrassment. Physicians
reported low patient success at initiating insulin in a timely fashion and adjusting insulin doses. Most physicians report that
many insulin-treated patients do not have adequate glucose control (87.6%) and that they would treat more aggressively if
not for concern about hypoglycaemia (75.5%). Although a majority of patients (and physicians) regard insulin treatment as
restrictive, more patients see insulin treatment as having positive than negative impacts on their lives.
Conclusions Glucose control is inadequate among insulin-treated patients, in part attributable to insulin omission ⁄ non-
adherence and lack of dose adjustment. There is a need for insulin regimens that are less restrictive and burdensome with
lower risk of hypoglycaemia.
Diabet. Med. 29, 682–689 (2012)
Keywords adherence, insulin therapy, psychosocial, questionnaire

longed periods of poor glycaemic control are believed to

Introduction
increase the severity and progression of macrovascular, as well
While patients with Type 1 diabetes must take insulin to as microvascular complications [4]. The American Diabetes
survive, many patients with Type 2 diabetes also require insulin Association and the European Association for the Study of
supplementation in order to control persistent hyperglycaemia. Diabetes recently issued guidelines for the treatment of Type 2
However, patients with Type 2 diabetes often do not receive diabetes that identified insulin as the most effective glucose-
insulin or do not receive insulin in a timely manner [1–3]. lowering agent and insulin replacement therapy as a key
Delays in insulin initiation and consequent exposure to pro- component of effective diabetes management over the course of
the disease [5]. Moreover, among those receiving insulin ther-
Correspondence to: Mark Peyrot, Department of Sociology, Loyola University apy, regimen adherence, persistence and intensity may be poor
Maryland, 4501 North Charles Street, Baltimore, MD 21210-2699, USA. [6–8], resulting in worse glucose control and increased hospital
E-mail: [email protected] admissions for diabetes complications [9,10].
Re-use of this article is permitted in accordance with the Terms and Condi- Effective insulin therapy includes four critical accomplish-
tions set out at http://wileyonlinelibrary.com/onlineopen#OnlineOpen_Terms ments: initiation, adherence, persistence, intensification [11].

ª 2012 The Authors.

682 Diabetic Medicine ª 2012 Diabetes UK
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Original article DIABETICMedicine

Each of these requires the involvement of both patient and France, Japan, Germany, Spain, Turkey, the UK and the USA.
healthcare provider: (1) providers must recommend ⁄ prescribe The physician survey was conducted via the Internet and the
insulin and patients must fill their prescriptions and begin taking patient survey was conducted through computer-assisted tele-
the medication; (2) providers must formulate an insulin regimen phone interviewing. Each survey used one questionnaire that
that patients can implement and patients must adhere to that was translated into the primary language of each country. The
regimen; (3) providers must renew prescriptions and patients questionnaires were developed with collaboration of Edelman,
must continue to refill and use their prescriptions; (4) providers StrategyOne, Novo Nordisk and the authors. Before conduct-
must intensify when appropriate (increase dose and frequency of ing the main survey phase, a pretest was conducted among
administration) and patients must accept and implement the primary care physicians and patients to check that the ques-
intensified regimen. Unfortunately, there are failures at each tionnaire was effective and unbiased and that there were no
juncture, some of which can be attributed to patients, some to obvious errors or omissions. The final questionnaires were
providers and all, in part, because of the nature of the insulins and revised based upon the feedback obtained from the pretest.
delivery systems that are available to patients and physicians. The target physician sample size was set at 1250. Quotas
There are a number of provider barriers to initiation of insulin were defined for the number of physicians—a minimum of 50
therapy. Some of these relate to beliefs about the medication primary care physicians (internal medicine, general medicine
itself; some physicians believe that insulin therapy may not be and family practice) and 50 specialists (diabetologists and
effective, may result in weight gain, increase the risk of hypo- endocrinologists) in each country, with higher quotas in the
glycaemia and have other side effects [12]. Physicians also may USA (200 primary care physicians, 150 specialists), China (100
believe that insulin therapy is inconvenient and painful for primary care physicians, 150 specialists) and the UK (100 pri-
patients and will result in patient dissatisfaction [13–15]. Espe- mary care physicians, 50 specialists). Respondents were
cially important is the therapeutic orientation of the provider; recruited via validated healthcare professional panels main-
i.e. whether or not the provider emphasizes normalization of tained by WorldOne Healthcare Research. Physician eligibility
blood glucose and modifies treatment to achieve glucose control criteria were: in practice for more than 1 year since completing
targets [12]. Some barriers may be a function of the provider’s residency, see a minimum number of patients with diabetes per
level of specialization and treatment experience [12,16]. week (primary care physicians 5, specialists 10) and initiate
Patient barriers to some degree parallel those of providers; for insulin treatment for patients with diabetes.
example, concerns about efficacy, safety and weight gain The target patient sample size was set at 1500. Quotas were
[8,15,17]. But patients are also concerned about convenience, defined for the number of patients with Type 2 diabetes—a
interference with daily living and social stigma [8,11,18]. While minimum of 135 in each country, with higher quotas in the
these concerns may be valid, patients also have inaccurate beliefs USA (315), China (180) and the UK (180). Patients with
about insulin therapy, including the idea that insulin causes late- Type 1 diabetes were captured during enrollment of those with
stage diabetes complications and is an indication of imminent Type 2 diabetes in whatever numbers were obtained. Respon-
deterioration and death, or is a result of patients’ failure to take dents were recruited via panels of research consumers main-
good care of themselves [12]. Other barriers may be practical, tained by WorldOne Healthcare Research. Patient eligibility
including medication cost and difficulty with access [19]. criteria were: age 18 years or older; use insulin to control blood
This paper reports the results of the Global Attitudes of Patients sugar; Type 1 or Type 2 diabetes.
and Physicians in Insulin Therapy study, a multinational survey of
patients and providers regarding insulin therapy. This study
Measures
examines patient and physician reports of insulin omission ⁄ non-
adherence and the reasons for these events, physician perceptions The questionnaires assessed patient and physician reports of:
of patient success with insulin treatment tasks, and patient and frequency of insulin omission ⁄ non-adherence and reasons for
physician perceptions and beliefs about insulin therapy. The these events; dissatisfaction with insulin therapy; perceptions of
study allows a comparison of patient and physician responses for patient difficulties; and opinions of insulin therapy. The phy-
a number of beliefs and behaviours, which is important because sician questionnaire also assessed perceptions of patient success
prior research has suggested that there may be differences in with various insulin treatment tasks. The patient questionnaire
beliefs and perceptions and these differences may interfere with also assessed perceptions of the impact of insulin treatment on
optimal diabetes treatment [20,21]. The study also permits a their lives.
comparison of specialist and primary care physicians [12,19]. The key measure was insulin omission ⁄ non-adherence. For
patients, this concept was assessed by a single item that assessed
whether the respondents ever miss an insulin dose or do not take
Participants and methods
insulin exactly as prescribed and, if so, how many days this had
happened in the last month. For patients, the frequency of
Study design
insulin omission ⁄ non-adherence was calculated two ways: (1)
The study consisted of cross-sectional surveys of patients and based on the responses of all patients (treating those who said
physicians in eight developed and developing countries: China, no to the original item as having zero days of insulin omission ⁄

ª 2012 The Authors.

Diabetic Medicine ª 2012 Diabetes UK 683
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DIABETICMedicine Insulin adherence in the GAPP Study • M. Peyrot et al.

non-adherence) and (2) the number of days reported by those Table 1 Characteristics of study populations
who answered yes to the original item. For physicians, this
concept was assessed by two questions; the first an item that Characteristic Response
assessed whether any of the respondents’ typical patients fail to
Patient sample n = 1530
take their insulin as prescribed. Physicians who responded yes
Country
were asked how many days in the last month a typical patient China 13.1%
would miss an insulin dose or not take insulin exactly as pre- France 10.5%
scribed for (1) basal insulin and (2) meal-related insulin. For Germany 9.9%
physicians, the frequency of insulin omission ⁄ non-adherence Japan 9.8%
Spain 10.3%
was calculated two ways: (1) based on the responses of all
Turkey 10.1%
physicians (treating those who said no to the original item as UK 13.4%
reporting zero days of insulin omission ⁄ non-adherence) and (2) USA 22.9%
the number of days reported by those who answered yes to the Female 50.4%
original item. Age (years, mean  sd) 60.1  13.7
Race ⁄ ethnicity
White 44.2%
Statistical analysis Asian 23.9%
Middle Eastern 9.4%
Significance of differences between patient and physicians, Hispanic 5.7%
between patients with Type 1 and Type 2 diabetes, and Black 4.5%
None of the above 12.4%
between specialists and primary care providers are analysed
Type 2 diabetes 88.2%
using the v2 statistic. Other than data regarding characteristics Duration of diabetes (years, mean  sd) 14.7  10.2
of the study populations (Table 1), all data are weighted so that Duration of insulin treatment 8.6  8.3
every country is equally represented. For the physician data, (years, mean  sd)
specialists and primary care physicians are weighted so that Insulin delivery system
Pen only 74.2%
they are equally represented. For patients, data are weighted so
Syringe only 20.6%
that the percentage of Type 1 and Type 2 diabetes is the same Pen and syringe 3.7%
for each country as for the overall study population. Weights Other 1.5%
maintain original sample sizes. Thus, the results are not Physician sample n = 1250
influenced by disproportionate country sample sizes. Country
China 20.0%
France 8.0%
Ethical approval Germany 8.0%
Japan 8.0%
The study, which received ethical approval from the Human Spain 8.0%
Subjects Committee at Loyola University Maryland, complies Turkey 8.0%
UK 12.0%
with the recommendations of the 1964 Declaration of Helsinki
USA 28.0%
and all relevant ethical standards. Male 70.0%
Specialty
Diabetology 17.2%
Results Endocrinology 30.8%
Family practice 16.1%
Study populations General practice 20.4%
Internal medicine 15.5%
Patient recruitment targets were met or exceeded in each coun- Primary clinical setting
try, resulting in a total sample of 1530 respondents (Table 1). Private practice office 42.4%
Respondents were almost equally divided by gender, with a Hospital outpatient 31.7%
Hospital inpatient 10.4%
mean age of approximately 60 years (Type 1 diabetes
Community health centre 9.1%
47 years, Type 2 diabetes 62 years). White people were the Public health service 5.4%
largest racial ⁄ ethnic group, but less than half the total sample. Other 1.0%
Most respondents had Type 2 diabetes and the mean duration of Duration of clinical practice 17.0  8.3
diabetes was almost 15 years (Type 1 diabetes 18 years, (years, mean  sd)
Type 1 patients (no. weekly, 10.3  13.5
Type 2 diabetes 14 years). Respondents had been using insulin
mean  sd)
for an average of almost 9 years (Type 1 diabetes 15 years, Type 2 patients (no. weekly, 56.7  54.5
Type 2 8 years) and the majority used an insulin pen all or part mean  sd)
of the time (Type 1 diabetes 76%, Type 2 diabetes 78%). Type 2 patients using insulin 24.0  28.2
Physician recruitment targets were met in each country, (no. weekly, mean  sd)
resulting in a total sample of 1250 respondents (Table 1).

ª 2012 The Authors.

684 Diabetic Medicine ª 2012 Diabetes UK
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Original article DIABETICMedicine

Specialists (diabetologists and endocrinologists) made up cannot be compared with percentage of patients who report
almost half of the sample and primary care physicians were these reasons, because physicians report whether this is a
relatively equally divided among internal medicine, general reason for the behaviour of their ‘typical patient’ rather than
medicine and family practice. Private offices and hospitals were the percentage of their patients for whom this is a reason, but
the predominant practice sites. Respondents had been practic- the rank order of reasons by patients and physicians can be
ing for an average of 17 years and saw an average of over 30 compared. The Spearman rank order correlation was 0.86 and
insulin-treated patients during a week. the top five reasons were the same for patients and physicians.
The one major discrepancy was forgetting, which was more
highly ranked by patients than physicians.
Survey responses
Patients and physicians reported a number of negative
Physicians rated patient success with insulin treatment tasks as perceptions about insulin treatment (Table 3). The two most
low (Fig. 1). Only a minority of patients were rated as ‘very commonly reported difficulties patients have with insulin
successful’ with any of the tasks, with very low rates for treatment (as reported by both patients and physicians) were
starting insulin when needed and adjusting doses. Less than one
third of physicians rated patients as very successful in being
able to take their basal insulin everyday (28.9%) or take their Table 2 Patient (n = 530) and physician (n = 964) reported reasons* for
insulin omission ⁄ non-adherence 
bolus ⁄ premixed insulin as prescribed (11.2%).
Patients and physicians reported a high level of insulin
Patients % Physicians %
omission ⁄ non-adherence. One third of patients (33.2%)
Reason and rank and rank
reported insulin omission ⁄ non-adherence, with a mean of
3.3 days in the last month; the calculated rate of insulin Too busy 18.9% 1 41.9% 3
omission ⁄ non-adherence for the entire sample was 1.1 days. Travelling 16.2% 2 43.6% 2
The majority of physicians (72.5%) reported that some of their Skipped meal 15.0% 3 44.8% 1
Stress or emotional problems 11.7% 4 32.2% 5
typical patients do not take their insulin as prescribed (this
Embarrassing to inject in public 9.7% 5 36.8% 4
percentage cannot be compared with the percentage of patients Challenging to take it at the 9.4% 6 29.1% 6
reporting insulin omission ⁄ non-adherence because physicians same time everyday
did not report the number or percentage of their patients who Forgot 7.4% 7 2.0% 11
did so). For physicians who said yes, the reported days per Too many injections 6.0% 8 26.4% 7
Avoid weight gain 4.0% 9 13.4% 9
month of insulin omission ⁄ non-adherence was 4.3 for basal
Regimen is too complicated 3.8% 10 16.8% 8
insulin and 5.7 for meal-related insulin. The calculated rate of Injections are painful 2.6% 11 7.8% 10
insulin omission ⁄ non-adherence for the entire sample was
3.1 days for basal insulin and 4.1 days for meal-related insulin. *Respondents were asked to select top three reasons (order of
These data support the physicians’ report that there is less reasons randomized, ‘Forgot’ responses volunteered as Other);
data are % of respondents choosing a reason as one of the
success with administration of bolus ⁄ premixed insulin than
three.
basal insulin (Fig. 1).  Absolute percentages reported for physicians cannot be com-
Patients and physicians were asked to choose the top three of pared with percentage of patients who report these reasons
10 possible reasons for insulin omission ⁄ non-adherence because physicians report whether this is a reason for the
(Table 2); one reason (whether omission ⁄ non-adherence was a behaviour of their ‘typical patient’ rather than the percentage of
their patients for whom this is a reason; rank order of reasons
result of forgetting) was volunteered by a substantial number of
by patients and physicians can be compared.
respondents. The absolute percentages reported for physicians

Starng insulin treatment when it is needed 9.7%

Taking basal insulin every day 28.9%

Taking basal insulin at the same me every day 18.6%

Taking bolus/premixed insulin as prescribed 11.2%

Adjusng insulin doses 6.2%

0 10 20 30 40
Percentage (%)

FIGURE 1 Physician report: patient success with insulin treatment tasks (n = 1250)* . *Percentage of physicians reporting that their patients are very
successful (vs. somewhat successful, not very successful, not at all successful, ‘don’t know’).  Specialists and primary care physicians not significantly
different for any measure

ª 2012 The Authors.

Diabetic Medicine ª 2012 Diabetes UK 685
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DIABETICMedicine Insulin adherence in the GAPP Study • M. Peyrot et al.

Table 3 Patient and physician perceptions of insulin treatment tion of ability to control blood glucose where they were similar.
However, blood glucose control was ranked second in patient
Categories and Patients Physicians dissatisfaction, but sixth in physician dissatisfaction. Finally, a
items (n = 1530) (n = 1250) majority of both patients and physicians felt that diabetes is
restrictive and controlled patients’ lives and about half felt that
Patient difficulties*
Taking insulin at prescribed time 27.6%*1 54.5% it is hard to live a normal life while managing diabetes. Patient
or with meals every day and physician agreement was strongest for the wishes that
Number of daily injections 23.1% 58.5%*2 insulin should be flexible to fit patients’ lives and that good
Following healthcare professional 16.9% 45.4% control with insulin should not require injections every day.
instructions
Table 4 examines country differences for two key measures
Preparing injections 10.3% 35.0%
Adjusting insulin doses 16.8% NA reported by both patients and physicians—assessment of insu-
Changing timing of insulin to meet NA 57.7% lin omission ⁄ non-adherence and overall dissatisfaction with
daily needs insulin treatment. Patient, specialist physician and primary care
Dissatisfaction  physician responses differed significantly (P < 0.05) across
Choose frequency of injections 17.6% 43.3% 2
countries for all measures. Specialists reported significantly
Choose time of injections 15.2% 32.2% 1
Blood glucose control 15.8% 15.9% (P < 0.05) more insulin omission ⁄ non-adherence than primary
Simplicity of regimen 12.9% 27.8% care physicians, both overall and within most countries, but did
Safety regarding low blood sugar 11.4% 32.0% not differ in dissatisfaction with insulin treatment (physician
Insulin treatment overall 10.0% 18.2% responses are not comparable with patient responses). Patients
Opinionsà
were significantly (P < 0.05) less satisfied with insulin treat-
Wish for good control with insulin 92.5% 91.2%
not injected every day ment than either specialist physicians or primary care physi-
Wish insulin regimen would fit 81.4% 85.8%à2 cians, both overall and within most countries. Country
daily life changes rankings of dissatisfaction by patients, specialists and primary
Insulin-treated diabetes controls 66.7% 66.1%à1 care physicians were not significantly correlated (Spearman’s
life
rho < 0.5); country rankings of omission ⁄ non-adherence were
Insulin regimen can be restrictive 59.8% 68.2%
Hard to live normal life while 54.4% 49.6% stronger (Spearman’s rho > 0.5), although only specialist and
managing diabetes primary care physicians rankings were significantly correlated,
indicating that there may be a country effect for insulin omis-
*Very difficult or somewhat difficult (vs. very easy, somewhat sion ⁄ non-adherence.
easy, not applicable, ‘don’t know’). Absolute percentages
In spite of the negative attitudes and perceptions of insulin
reported for physicians cannot be compared with percentage
of patients for whom this is difficult because physicians treatment, Fig. 2 shows that more patients reported positive
report whether this is difficult for their ‘typical patient’ rather than negative impact on life for all domains except finances
than the percentage of their patients for whom this is (P < 0.05), although the trend was stronger for patients with
difficult; rank order of reasons by patients and physicians Type 1 diabetes than for those with Type 2 diabetes.
can be compared.
Figure 3 presents physician beliefs about insulin treatment.
*1Average of response for two items (insulin at prescribed
times, insulin with each meal). While there are statistically significant differences between
*2Physician item is ‘taking insulin frequently’. specialist and primary care physicians, physicians generally
 Very dissatisfied or somewhat dissatisfied (vs. very satisfied, agree that many patients on insulin are not adequately con-
somewhat satisfied, not applicable). trolled. Physicians report that the possibility of hypoglycaemia
 1Physician item is ‘insulin regimens that better fit patients’
limits treatment aggressiveness, and that it is difficult to manage
dynamic lives’.
 2Physician item is ‘total number of injections per week’. efficacy (hyperglycaemia) and safety (hypoglycaemia) simulta-
àStrongly agree or somewhat agree (vs. strongly disagree, neously. Finally, physicians wish there was an insulin treatment
somewhat disagree, neither, ‘don’t know’). that would have sustained efficacy if patients miss a dose or, to
à1Physician item is if his ⁄ her patients feel diabetes controls extrapolate, delay a dose.
their lives.
à2Physician item is ‘which insulin treatments could be more
flexible’. Discussion
NA, not asked.
Physicians perceived patients as relatively unsuccessful in terms
the number of injections taken and taking insulin at prescribed of starting insulin when needed, adjusting insulin doses and
times; these two aspects of insulin therapy were also among adhering to prescribed regimens. Insulin omission ⁄ non-adher-
those receiving the highest level of dissatisfaction (as reported ence was a common problem in all countries, although the rate
by both patients and physicians). Patients reported difficulty in was twice as high in some countries as in others. Physicians
adjusting insulin doses, in agreement with physician views of were aware of this problem, especially specialists, and reported
their success in this area. In general, physicians were more that it was more common for mealtime and premixed insulin
dissatisfied with insulin therapy than patients, with the excep- than for basal insulin. Physicians and patients generally agreed

ª 2012 The Authors.

686 Diabetic Medicine ª 2012 Diabetes UK
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Original article DIABETICMedicine

Table 4 Comparison of patient and physician belief and behaviour by country

Non-adherence** Dissatisfaction
Physician* Physician*
All All
patients*  Specialist* Primary care patients*  Specialist* Primary care
Country % and rank % and rank % and rank % and rank % and rank % and rank

China 33.3% 5 80.0% 4 68.0% 4 7.3%§– 5 28.0% 2 18.0% 4

France 19.4% 8 70.0% 6 52.0%à 7 3.1%§– 7 16.0% 6 18.0% 5
Germany 39.8% 4 96.0% 1 86.0%à 2 2.1%§– 8 8.0% 7 26.0%à 2
Japan 43.8% 1 74.0% 5 58.0%à 6 21.5%– 2 32.0% 1 36.0% 1
Spain 22.5% 7 56.0% 7 40.0%à 8 9.9% 3 12.0% 3 14.0% 6
Turkey 24.1% 6 58.0% 8 64.0% 5 22.5%§– 1 6.0% 8 10.0% 8
UK 41.4% 3 96.0% 1 78.0%à 3 4.7%§– 6 20.0% 5 22.0% 3
USA 41.9% 2 90.0% 3 94.0% 1 8.4% 4 12.0% 4 12.0% 7
Total 33.2% 77.5% 67.5%à 10.0%§– 17.0% 19.5%

*Responses in this study populations differ significantly by country (P < 0.05).

 No significant differences between patients with Type 1 diabetes and those with Type 2 diabetes.
àResponses for specialist and primary care physicians differ significantly (P < 0.05).
§Responses for patients and specialist physicians differ significantly (P < 0.05).
–Responses for patients and primary care physicians differ significantly (P < 0.05).
**For non-adherence, absolute percentages reported for physicians cannot be compared with percentage of patients who report insulin non-
adherence because physicians report whether ‘any of your typical patients fail to take their insulin as prescribed’ rather than the percentage of
their patients who have missed their dose or fail to take their insulin exactly as prescribed; rank order of countries for patients and physicians
can be compared.

FIGURE 2 Patient reports: impact of insulin treatment on life domains in patients with Type 1 (non-hatched) and Type 2 (hatched) diabetes mellitus
(n = 1530). *Positive and negative responses differ significantly (P < 0.05);  reports from patients with Type 1 and Type 2 diabetes differ significantly
(P < 0.05).

on the ranking of reasons for insulin omission ⁄ non-adherence; Patient and physician dissatisfaction with insulin treatment
in addition to situationally appropriate reasons (e.g. after varied substantially across countries, with no strong evidence of
skipping a meal), respondents identified logistical problems a consistent ranking of patient and physician levels of dissat-
(too busy, travelling) and psychosocial problems (stress ⁄ emo- isfaction across countries. Patient dissatisfaction with various
tions and embarrassment) as key factors. Patients also volun- aspects of insulin treatment was low (10–20%). Physician
teered that forgetting was a common reason, a possibility that dissatisfaction with insulin treatment overall and with glucose
was not as salient for physicians. control was in the same range as patients, but more physicians
Several negative perceptions of insulin treatment were (25–45%) were dissatisfied with several other aspects of insulin
reported. Taking insulin at the prescribed time and frequency treatment, including regimen simplicity, hypoglycaemia and
were the difficulties identified as most common by both patients injection timing and frequency.
and physicians. A majority of patients and physicians agreed Respondents also expressed opinions about how insulin
that insulin regimens were restrictive and can control one’s life. treatment could be improved. Patients and physicians reported

ª 2012 The Authors.

Diabetic Medicine ª 2012 Diabetes UK 687
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
DIABETICMedicine Insulin adherence in the GAPP Study • M. Peyrot et al.

FIGURE 3 Physician beliefs about insulin treatment (n = 1250). *Strongly agree or somewhat agree (vs. strongly disagree, somewhat disagree, neither,
‘don’t know’);  strongly agree or somewhat agree (vs. strongly disagree, somewhat disagree)

that they would like insulin treatment to be more flexible so its occurrence. Additional attention should also be given to the
that it could be adapted to situational variation in daily activ- problems of initiating insulin therapy and getting patients to
ities. Both groups indicated that it would be better if insulin did self-manage their insulin doses. Research is needed to better
not have to be injected everyday, and physicians indicated that understand physician concern about hypoglycaemia, especially
it would be an improvement if insulin would maintain its as it affects their choices of diabetes management goals and
efficacy when patients miss a dose. Physicians also indicated strategies. Research in these areas could help to identify mea-
that they would be more aggressive in treating diabetes if there sures to improve patient and physician diabetes management
was no concern about hypoglycaemia, suggesting that insulins and outcomes.
with less risk of hypoglycaemia could be used more aggres-
sively, potentially leading to improvements in blood glucose
Clinical implications
control and reductions in complications that result from sub-
optimal glucose control. Although patients using insulin are not entirely satisfied with
In spite of the drawbacks of current insulin regimens, their treatment, insulin is well received. The primary problem is
patients reported that the net impact of insulin on their lives that it is seen as restrictive, making it difficult to take all doses
was positive. In six of seven domains examined (the exception as prescribed, especially given patients’ difficulty in adjusting
was financial impact) more patients reported the impact of insulin doses to respond to daily changes. Physicians should
insulin to be positive than to be negative, with a substantial consider prescribing more flexible insulin regimens and reduc-
number saying it had no impact. The advantage was most ing the burden of the treatment regimen. An ideal regimen
pronounced in terms of physical well-being, but also was would minimize the number of injections required [21], the risk
present for emotional well-being, as well as social relationships of hypoglycaemia and the consequences of a delayed or missed
and work. The advantages were more pronounced in patients insulin dose.
with Type 2 diabetes, especially in areas other than physical
well-being. However, the other data presented here identified a
Competing interests
number of unmet needs that could be addressed by adjusting
regimens for currently available insulins and ⁄ or developing MP has received research grant support from: Amylin, Animas,
improved insulins. Genentech, MannKind, Medtronic MiniMed and Novo
There were substantial differences among the countries in Nordisk. He has received consulting fees from: Amylin, Animas,
respondents’ insulin-related beliefs and behaviours. However, Eli Lilly, Genentech, MannKind, Medtronic MiniMed and
the pattern of country differences was not consistent across Novo Nordisk. He has received speaking honoraria from Novo
measures or subgroups of respondents. Thus, the explanation Nordisk and has participated in scientific advisory committees
for the country differences must be more complex than a set of for Eli Lilly, Novo Nordisk and Roche. He has ben reimbursed
country-specific clusters of insulin-related beliefs and behav- by Amylin, Animas, Eli Lilly, Genentech, MannKind, Med-
iours that are shared by all members of a culture. This suggests tronic MiniMed and Novo Nordisk for attending conferences.
that country differences reflect the interplay of cultural beliefs, AHB has received honoraria for lectures and advisory work
healthcare provider training and health system characteristics. as well as research funding from BMS ⁄ Astrazeneca, Boehrin-
ger-Ingelheim, Eli Lilly, Novo Nordisk, MSD, Roche Diag-
nostics, Sanofi-Aventis and Takeda.
Research implications
LFM has received research grant support from Biodel,
Insulin omission ⁄ non-adherence is common and further MannKind, Medtronic, Novo Nordisk, Pfizer and Sanofi-
research is needed to determine the risk factors associated with Aventis. He has received consulting fees from Biodel, Nipro

ª 2012 The Authors.

688 Diabetic Medicine ª 2012 Diabetes UK
 14645491, 2012, 5, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/j.1464-5491.2012.03605.x by Cochrane Colombia, Wiley Online Library on [27/03/2023]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License
Original article DIABETICMedicine

and Novo Nordisk. He has received speaking honoraria from treatment, glycemic control, and ketoacidosis in insulin-dependent
Amylin, Eli Lilly, Merk, Novo Nordisk and Sanofi-Aventis. diabetes mellitus. Lancet 1997; 350: 1505–1510.
11 Peyrot M, Rubin RR. Perceived medication benefits and their
PMSD has received funds for research support, consulting,
association with interest in using inhaled insulin in type 2 diabetes:
speaking, organizing education and staff support from a model of patients’ cognitive framework. Patient Prefer Adherence
Astrazenica, Berlin Chemie, BMS, Boehringer-Ingelheim, Eli 2011; 5: 255–265.
Lilly, GSK, MSD, Novartis, Novo Nordisk and Sanofi-Aventis. 12 Peyrot M, Rubin RR, Lauritzen T, Skovlund SE, Snoek FJ,
Matthews DR et al. on behalf of the International DAWN Advisory
Board. Resistance to insulin therapy among patients and providers:
Acknowledgements results of the cross-national Diabetes Attitudes, Wishes, and Needs
study (DAWN). Diabetes Care 2005; 28: 2673–2679.
Schinichi Harashima, Doron Schneider and Ji Linong provided 13 Rubin RR, Peyrot M. Factors associated with physician perceptions
consultation regarding project design. of and willingness to recommend inhaled insulin. Curr Med Res
Opin 2011; 27: 285–294.
14 Nakar S, Yitzhaki G, Rosenberg R, Vinker S. Transition to insulin
References in type 2 diabetes: family physicians’ misconception of patients’
fears contributes to existing barriers. J Diabetes Complications
1 Dailey GE. Early insulin: an important therapeutic strategy. 2007; 21: 220–226.
Diabetes Care 2005; 28: 220–221. 15 Peyrot M. Psychological insulin resistance. Overcoming barriers to
2 Davidson MB. Early insulin therapy for type 2 diabetes. Diabetes insulin therapy. Practical Diabetology 2004; 23: 6–12.
Care 2005; 28: 222–224. 16 Grant RW, Wexler DJ, Watson AJ, Lester WT, Cagliero E,
3 Home PD, Boulton AJM, Jimenez J, Landgraf R, Osterbrink B, Campbell EG et al. How doctors choose medications to treat type 2
Christiansen JS. Issues relating to the early or earlier use of insulin diabetes: a national survey of specalists and academic generalists.
in type 2 diabetes. Practical Diabetes Int 2003; 20: 63–71. Diabetes Care 2007; 30: 1448–1453.
4 Del Prato S, Penno G, Miccoli R. Changing the treatment paradigm 17 Peyrot M, Skovlund S, Landgraf R. Epidemiology and correlates of
for type 2 diabetes. Diabetes Care 2009; 32: S217–S222. weight worry in the multi-national Diabetes Attitudes, Wishes and
5 Nathan DM, Buse JB, Davidson MB, Ferrannini E, Holman RR, Needs (DAWN) Study. Curr Med Res Opin 2009; 25: 1985–1993.
Sherwin R et al. Medical management of hyperglycemia in type 2 18 Brod M, Kongso JH, Lessard S, Christensen TL. Psychological
diabetes: a consensus algorithm for the initiation and adjustment of insulin resistance: patient beliefs and implications for diabetes
therapy: a consensus statement from the American Diabetes Asso- management. Qual Life Res 2009; 18: 23–32.
ciation and the European Association for the Study of Diabetes. 19 Wagner TH, Heisler M, Piette JD. Prescription drug co-payments
Diabetes Care 2009; 32: 193–203. and cost-related medication underuse. Health Econ Policy Law
6 Cramer J. A systematic review of adherence with medications for 2008; 3: 51–67.
diabetes. Diabetes Care 2004; 27: 1218–1224. 20 Peyrot M, Rubin RR, Lauritzen T, Snoek FJ, Matthews DR,
7 Rubin RR. Adherence to pharmacologic therapy in patients with Skovlund SE, on behalf of the International DAWN Advisory Panel.
type 2 diabetes mellitus. Am J Med 2005; 118: 27S–34S. Psychosocial problems and barriers to improved diabetes manage-
8 Peyrot M, Rubin RR, Kruger D, Travis L. Correlates of insulin ment: results of the cross-national Diabetes Attitudes, Wishes and
injection omission. Diabetes Care 2010; 33: 243–248. Needs (DAWN) study. Diabet Med 2005; 22: 1379–1385.
9 Cramer JA, Pugh MJ. The influence of insulin use on glycemic 21 Rubin RR, Peyrot M, Kruger DF, Travis LB. Barriers to insulin
control: how well do adults follow prescriptions for insulin? injection therapy: patient and health care provider perspectives.
Diabetes Care 2005; 28: 78–83. Diabetes Educ 2009; 35: 1014–1022.
10 Morris AD, Boyle DIR, McMahon D, MacDonald TM, Newton
RW, for the DRTS ⁄ MEMO Collaboration. Adherence to insulin

ª 2012 The Authors.

Diabetic Medicine ª 2012 Diabetes UK 689