Value of SMBG in Type 2 diabetes taking oral antidiabetic agents

1
Value of SMBG inType 2 diabetes taking oral
antidiabetic agents

• LifeScan Reimbursement Team
• September 01, 2006
• Version 1.0

AW 088-723A AW 089-677A
2
Role of SMBG in type 2 diabetes on oral agents

• Safety detect hypoglycemia
• For patients on insulin-sensitizing drugs (e.g.
  sulfonylureas)
• Help improve glycemic control and reduce
  complications

3
Safety Detection of Hypoglycemia
4
Hypoglycemia leads to increased healthcare use
While hypoglycemia is less common in type 2
patients, hospitalization was more common 1

• A German study found the incidence of severe
  hypoglycemia in people with type 1 diabetes was
  3.8 episodes per 100 patients per year, and 0.4
  episodes per 100 patients per year in patients
  with type 2 diabetes.1
• Individuals with type 1 diabetes required
  hospitalization significantly less often, and for
  significantly less time that individuals with
  type 2 diabetes experiencing an episode of
  hypoglycemia

Days hospitalized when admitted for hypoglycemia
Patients with hypoglycemia admitted to the
hospital ()
1. Holstein et al. Diabetes Care 2002 252109
5
Hypoglycemia is less common in type 2 diabetes on
oral medication than in patients on insulin
Risk of severe hypoglycemia requiring emergency
treatment

• In a study of intensified glucose management with
  oral agents or insulin, 15 to 21 of patients on
  sulfonylureas reported an incident of any
  hypoglycemia in a 10 year period compared with 8
  of patients taking metformin.1
• In that same study, 1 of patients on
  sulfonylureas had an episode of severe
  hypoglycemia over a 10 year period compared to
  0.6 of patients taking metformin.1
• A separate study found that 2 patients taking
  long acting sulfonylureas were admitted to the
  hospital for every 1000 patient years of drug
  exposure.2

Severe hypoglycemia episodes per 100 patient
years (Ref 3)
Most hypoglycemia encountered with oral agent use
is found in individuals taking sulfonylurea
medications
1. UKPDS 34, Lancet 1998 352854-865 2. Stahl
M Berger W, 1999 16586- 3. Leese GP et al
Diabetes Care 2003 261176-1180
6
Hypoglycemia is less common in patients on
oral hypoglycemic agents than those taking
insulin
Prevalence of hypoglycemia by type of diabetes
therapy

• The frequency of severe hypoglycemia with type 2
  patients taking insulin is the same as for
  patients with type 1 (taking insulin) and is
  approximately 11 cases per 100 patient years.1
• The frequency of severe hypoglycemia with
  sulphonylureas is 13 fold lower than for insulin
  using type 2 patients.1
• Mild hypoglycemia is common in patients with type
  2 diabetes undergoing aggressive management.2
• One cohort study found that 2.0 of patients
  over 65 and 1.4 of patients under 65 had at
  least one episode per year,3 With results
  confirmed by an additional study published that
  same year.4

1. Leese GP et al Diabetes Care 2003
261176-1180 2. Miller CD et al Arch Intern Med
2001 1653-1658 3. vanStaa T et al J Clin
Epidemiol 1997 50735-741 4. Shorr RI et al Arch
Intern Med 1997 1571681-1686
7
Hypoglycemia leads to increased healthcare use
While hypoglycemia is less common in type 2
patients, hospitalization was more common 1

• A German study found the incidence of severe
  hypoglycemia in people with type 1 diabetes was
  3.8 episodes per 100 patients per year, and 0.4
  episodes per 100 patients per year in patients
  with type 2 diabetes.1
• Individuals with type 1 diabetes required
  hospitalization significantly less often, and for
  significantly less time that individuals with
  type 2 diabetes experiencing an episode of
  hypoglycemia

Days hospitalized when admitted for hypoglycemia
Patients with hypoglycemia admitted to the
hospital ()
1. Holstein et al. Diabetes Care 2002 252109
8
Use of a sulfonylurea with metformin increases
the risk of trauma including motor vehicle
accidents

• In a large retrospective study of motor vehicle
  crashes in elderly patients, the adjusted risk
  ratio for a crash with patients using a
  combination of sulfonylurea and metformin was
  nearly as high as the risk in those taking
  insulin for their diabetes amounting to an
  approximate risk of 32 crashes per 10,000 elderly
  drivers per year.1
• In a separate study, hypoglycemia was
  significantly more common in insulin using type 1
  patients than in insulin or Oral agent users who
  had type 2 diabetes.2

Adjusted rate ratio of an injurious crash
NS
reporting mild accidents
1. Hemmelgarn B, et al. Canad J Clin Pharmacol
2006 13e112-e120 2. Cox DJ et al. Diab Care
2003 262329-2334
9
Hypoglycemia is associated with a worse quality
of life

• A survey of patients with type 2 diabetes
  participating in the United Kingdom Prospective
  Diabetes Study (UKPDS) found that 27 of patients
  taking insulin and 3 of patients not on insulin
  reported two or more hypoglycemic episodes in the
  previous year.
• Patients reporting at least one or hypoglycemic
  episode within the last year had a significantly
  greater total mood disturbance, and those
  reporting two or more episodes had greater mood
  disturbance and less work satisfaction than those
  who had no episodes of hypoglycemia

UKPDS Study Group. Diabetes Care 1999
221125-1136
10
People with hypoglycemia use more medical
services, have more unproductive days and cost
the health system more

• A large study of 2664 individuals who took
  insulin with or without another agent found one
  in six employees on insulin had a diagnosis of
  hypoglycemia within a 2.5 year follow-up.1
• Risks of hospitalization and emergency room
  visits was increased two fold in the group
  experiencing hypoglycemia.1
• Hypoglycemia was recorded in 19.9 of insulin
  treated patients versus 4.0 of those treated
  with oral agents.1
• Annualized medical costs of hypoglycemia was
  3241 (year 2000 dollars), and the risk of short
  term disability was increased five fold, with
  hypoglycemia patients having 77 more short term
  disability days than those not experiencing a
  hypoglycemic episode.1
• In a separate study of a large population of
  patients in Germany, hospitalization of
  hypoglycemia patients with type 2 diabetes was
  much more common than patients with type 1
  diabetes who had hypoglycemia, and often was
  associated with a need for inpatient monitoring
  when hypoglycemia was secondary to
  sulphonylureas.2

1. Rhoads GG et al J Occupational and
Environmental Med 2005 47447-452 2. Holstein A
et al Diabetes Care 2002 252109
11
Hospital admissions drive costs
Percent of Costs for Diabetes Care in MCOs

• US National Statistics show that inpatient care
  is a major driver of diabetes costs.
• Hypoglycemia significant enough to warrant
  medical attention occurred in 16 of patients on
  insulin with 7.1 episodes per 100 pts/yr
• Costs per episode averaged 1,583
• Costs were driven by hospitalizations occurring
  in 15 of the episodes but representing more than
  60 of the total costs to the Managed Care
  Organization (MCO)

Stephens JM. J Managed Care Pharmacy 2006
12130-142
12
Patients struggle to recognize hypoglycemia and
are unable to predict BG levels accurately

• Individuals with diabetes were asked to estimate
  their blood glucose. Estimates were defined as
  within plus or minus 1 mmol (18 mg/dL) of the
  SMBG meter reading
• 43 underestimated their blood glucose,
• 39 were within the accurate range
• 17 overestimated their blood glucose

Frankum S Ogden J Brit J Gen Pract 2005
55944-948
13
Impaired hypoglycemic awareness can increase the
incidence of hypoglycemia

• In a study of patients with normal awareness of
  hypoglycemic symptoms and patients with
  hypoglycemic unawareness, there was a significant
  difference in the incidence of hypoglycemia
  between the groups with 5 times the number of
  hypoglycemic episodes recorded in patients with
  hypoglycemic unawareness.1
• Missing hypoglycemia can be dangerous
  leading to disorientation, coma and even
  death.2

Mean no. of recordings per patient
(55 mg/dL)
(45 mg/dL)
1. Gold et al Diabetes Care 1994 17 697-703
2. ADA Diabetes Care 2005 28 1245-1249
14
Hypoglycemia with oral agents is more likely with
sulphonylureas

• In a review of patients from 719 practices in the
  UK encompassing 33,243 patients receiving
  sulphonylurea drugs, the incidence of
  hypoglycemia severe enough to report to their
  physician or to result in a visit to their
  physician occurred in 1.4 of patients under 65yo
  and 2.0 of patients over 65yo.
• Glyburide was associated with a rate of
  hypoglycemia nearly twice as high as other
  sulphonylureas.
• A separate study of diabetes patients between 40
  and 65yo, reported hypoglycemic symptoms in 20
  of patients treated with sulfonylureas but none
  of the patients treated with metformin.
• Hypoglycemic symptoms were reported on a monthly
  basis in 6 of the patients on sulphonylureas.
• Glyburide was again associated with a higher
  incidence of hypoglycemia than other
  sulphonylureas.

1. vanStaa T et al J Clin Epidemiol 1997
50735-741 2. Jennings AM et al Diabetes Care
1989 12203-208
15
The relationship between A1C and hypoglycemia is
unclear.

• Hypoglycemic symptoms vary with the type of
  diabetes treatment and have been reported in 12
  of patients treated with diet alone, 16 of
  patients using oral agents and 30 of patients
  taking insulin.2
• Severe hypoglycemia with oral agents is uncommon
  with no cases reported in one study, and in
  another the incidence was 2.24 cases per 1000
  patient years.3
• Risk factors for hypoglycemia with sulphonylureas
  include1-4
• Switching of sulphonylurea agents
• Use of 5 or more concomitant medications
• Lower A1C
• Recent hospital admission
• Missing a meal
• Exercise
• Age

1. Shorr RI et al Arch Internal Med 1997 157
2. Miller CD et al Arch Internal Med 2001
1611653-1659 3. Stahl M and Berger W. Diabet Med
1999 16586-590 4. vanStaa T et al J Clin
Epidemiol 1997 50735-741
16
SMBG can help to identify episodes of
hypoglycemia

• In a study of patients treated with a
  sulphonylurea, the lowest glucose values were
  recorded in the evening compared with morning
  measurements.1
• In a second study, pre-breakfast and pre-lunch
  SMBG captured the largest proportion of
  hypo-glycemic readings(64), and pre-dinner and
  bedtime SMBG captured the largest porportion
  of hyperglycemic readings (66).2
• Another study using timed four point glucose
  profiles also found that the lowest glucose
  values were in the evening (1700hrs) and the
  risk of hypoglycemia was greatest at that
  sampling time.3

Episodes of hypoglycemia
Episodes of hyperglycemia
1. Gillausseau PJ Diabetic Medicine 1997
14798 2. Hoffman RM et al. Diabetes Care 2002
251744 3. Monnier L Europ J Clin Invest 2004
3437
17
Treatment Failure and Glycemic Control

• 1988 to 2000
• According to NHANES data Glycemic control as
  defined by A1Cvalues of lt7 declined from 44.5
  in NHANES (1988-1994) to 35 in NHANES
  1999-2000.1
• Odds of glycemic control was 21 lower in NHANES
  1999-2000 after adjusting for age, ethnicity,
  sex, BMI, medication and duration of diabetes.1
• 2002
• A retrospective study in 2002 involving 1,765
  patients with type 1 and type 2 diabetes from 44
  clinics in the US reported only 34 of patients
  at an A1C goal of lt7. 2
• Fewer then half of the patients with elevated A1C
  values had changes in hypoglycemic therapy during
  clinic visits.2
• The Accumulated Effect?
• In one observational study involving 7,208
  patient completing treatment the average patient
  accumulated nearly 5 A1C years of excess glucose
  burden gt8 and about 10 years of glucose burden
  gt7 before starting insulin. 3
• Glycemic burden varies across treatments, from
  8.6 months to 2.5 years depending on the
  treatment.

3
1 Koro, CE et al, Diabetes Care vol 27, Number 1,
Jan 2004, 17-20 2. Grant, RW et al, Diabetes
Care, vol 28, number 21 Feb 2005, 337-42 3. Brown
JB, et al, Diabetes Care vol 27, number 7, July
2004, 1535-40
18
What We Know From Continuous Monitoring of Well
Controlled Patients

• Subjects remain euglycemic range for 63 of the
  total day.1
• Were hypoglycemic 3 to 8 or time and
  hyperglycemic 29 to 571,2
• Hypoglycemia is more prevalent nocturnally.1
• Hypoglycemia is frequently unreported.2
• Hay, et al reported hyperglycemia (defined as
  gt144mg/dl) was reported after 57 of all meals. 2
• Breakfast 60
• Lunch 58
• Dinner 55
• Bonora E,et al reported that although most
  patient had plasma glucose levels lt 120mg/dl
  before meals most had plasma glucose levels
  gt160mg/dl after meals.
• Breakfast 60.9
• Lunch 61.2
• Breakfast and or lunch 77.1

1 Bode, BW, et al. Diabetes Care, vol 28, number
10, Oct 2005, 2361-66 2. Hay, L.C. et al. Diab
Tech and Therap., Vol 5, Number 1, 2003, 19-26 3.
Bonora, E et, al, Diabetes Care, vol 24, number
12, Dec 2001 2023-29
19
A1C Values and Complications

• In patients with type 2 diabetes the risk of
  diabetic complications was strongly associated
  with previous hyperglycemia.
• Any reduction in A1C is likely to reduce the risk
  of complications.1
• The Relative risk for coronary heart disease
  (CHD) in persons with type 1 diabetes was 1.15
  (95 CI, 0.92-1.43) for each 1 increase in A1C.2
• The Relative risk for peripheral vascular disease
  (PVD) in person with type 1 diabetes was 1.32
  (CI, 1.19 to 1.45) for each 1 increase in A1C2
• Improved A1C control has been shown to lower
  low-density lipoprotein cholesterol levels which
  should theoretically lower risk of developing CHD
  in patient with diabetes.3

1 Stratton I.M., et al, BM vol 321, 12 Aug 2000,
405-412 2. Selvin, E et al, Ann of Inter Med, Vol
141, Number 6 Sept 2004, 421-431 3. Khaw, K, et
al, Ann of Inter Med, Vol 141, Number 6 Sept
2004, 4413-419
20
Incidences of Hypoglycemia

• Factors effecting predictive risk of developing
  hypoglycemia1,2.
• Duration of therapy
• Concomitant use of insulin (clear dose response)
  13 fold higher incidence vs orals alone
• Sulfonylurea-potentiating or antagonizing and
  concomitant use of beta blockers
• Patients 65 years or older have a 2 annual risk
  of developing hypoglycemia vs. 1.4 for people
  less than 65 years.1
• Females are more likely to have reported
  hypoglycemia compared to males.
• It is likely that less severe episodes of
  hypoglycemia do not lead to physician visits or
  reporting to the physician.1
• Hypoglycemia requiring emergency assistance from
  health service personnel is common in patients
  with type 2 diabetes treated with insulin as it
  is with type 1 patients.2
• In one study the rate of hypoglycemia was similar
  to that observed with intensively treated type 1
  patients in the DCCT even though the average A1C
  was 8.5 vs 7 in the DCCT.2

1. Va Staa T, J Clin Epidemiol vol 50, No 6, 1997 pp
   735-741
2. Leese, G.P., Diabetes Care, Vol 26, No 4, April
   2003 pp 1176-1180

21
SMBG Pre vs Post-prandial measurements alone and
Measurement with Counseling
Study Patients Therapy adjusted Therapy adjusted Results
Gerstein H.C. et al Diabet. Med. 21 2004 Drug naive Type 2 On Metformin Started on repaglinide Based on Pre-meal SMBG Based on Post Meal SMBG Similar A1C results (slightly better results if repaglinide is adjustedbased on post-meal SMBG when patient is taking metformin)
Kolendorf, K et al, Diab Research and Clin Practice 2004 Drug naive Type 2 Started on Repaglinide Based on Pre meal SMBG (FBG) Based on Post Meal SMBG (PPBG) 57 of FBG met target 86 of PPBG met target Slightly greater A1C drop in FBG group.
Results Results Results Results Results
Schwedes U, et al Diabetes Care vol 25 No 11 Nov 2002 Non-insulin treated Type 2 Standardized counseling on SMBG, diet and lifestyle Significant improvement in glycemic control during follow-up. 87 continued SMBG Metabolic status remained stable. Statistically significant reduction in A1C Significant improvement in glycemic control during follow-up. 87 continued SMBG Metabolic status remained stable. Statistically significant reduction in A1C
Non-standardized counseling on diet and lifestyle
22
Does SMBG Alone Improved Control?

• The Kaiser Permanente Model
• Evaluated the relationship between self
  monitoring frequency via strip utilization and
  A1C results.
• 24,312 patients
• Patient with type 1 diabetes monitoring gt3 times
  per day had a 1 lower A1C than those testing
  less frequently.
• Pharmacologically treated type 2 patients had a
  0.6 lower A1C than those testing less frequently
• Non-pharmacologic treated patients practicing
  SMBG at any frequency had a 0.4 lower A1C than
  those not testing.

Guerci B, et al, Diab Metab. 2003, 29, 587-94
23
SMBG and Glucose Control
Study n Results Associated Observations for Glucose Control
Franciosi M, Et al, Diabetes Care 241870-1877 2001 2855 Higher frequency of SMBG was associated with better metabolic control among insulin adjusting subjects. No relationship found for other treatment groups Patients with poorer glucose control tended to test more frequently. There is an association between testing at least once a week and increased depression and worry about diabetes. The knowledge and ability to adjust therapy based on SMBG results is required to facilitate improved glucose control.
Karter, A. J et al Diabetes Care vol 291757-1763, 2006 16,091 Longitudinal study Greater frequency in SMBG was associated in a graded decrease in A1C relative to nonusers. Changes in SMBG frequency among prevalent users were associated with an inverse graded change in A1C ONLY among pharmacological treated patients. Poor glycemic control may likely motivate healthcare providers and patients to initiate or intensify SMBG. Patient education or instruction was not known.
Fremantle Diabetes Study Diabetes Care vol 29No 8 Aug 2006 1286 Data from two studies support use of SMBG in insulin-treated type 2 diabetes. No significant difference in A1C between SMBG users and non users. Glycemic improvement was limited to patients who were able to adjust insulin doses. More frequent testing (gt7 times per week) may be required to facilitate improvement Reasons for lack of monitoring include No education on how to do SMBG (45) No motivation to start or continue SMBG (31) Fear of finger sticks (9)
Harris, M et al, Diabetes Care, Vol 24 No 6 Jun 2001 979-982 National survey SMBG is considered to be a tool to guide patient and physician with respect to changes in diet, physical activity and medication. The true impact of SMBG could be assessed in a randomized clinical trial with pre-established guidelines for how glucose results would be used to facilitate achievement of glucose targets.
24
Value of Immediate A1C Feedback

• Immediate feedback of A1C results at time of
  visit resulted in significant improvement of
  glycemic control at 6 months and persisted at 12
  months but did not change in the control
  group.1,2
• Resource usage was the same for both groups1
• Improvement in glycemic control was NOT
  accompanied by an increase in severe hypoglycemic
  episodes, or emergency room visits.1
• Patients in the immediate assay group changed
  insulin regimens LESS frequently than patients in
  control group.1
• Under conventional testing (results NOT obtained
  at time of patient visit) doctors did not appear
  to change patient management following the
  receipt of information.2
• Compared to usual insulin titration (no contact
  between visits), active titration and achieved
  greater A1C reduction.3
• Usual titration and laboratory A1C testing
• Usual titration and POC A1C testing
• Active titration and laboratory A1C testing
• Active titration and POC A1C testing
• Providers intensified therapy in 51 of patient
  in POC testing vs. 32 on patients in
  conventional.4
• Among Active titration groups, a greater
  proportion of patient in POC group achieved
  reductions in SMBG and achieved A1C goals.2

Type 2 Patients any therapy

• Providers are more likely to intensify therapy
  when A1C results are immediately available and
  A1Cgt83

POC Point of Care
1. Cagliero, E, et al, Diabetes Care, Vol 22 No
11, Nov 1999, 1785-1789 2. Grieve, R, Health
Technology Assessment 1999, Vol 3, No 15 3.
Kennedy, L, et al, Diabetes Care, vol 29, No 1,
Jan 2006, 1-8 4. Miller, C.D., Diabetes Care, Vol
26, No 4, Apr 2003 1158-1163
25
Quality of Life and Intensive Therapy

• Cross-sectional studies showed that therapies
  were neutral in effect, with neither improvement
  nor deterioration in QOL scores for mood,
  cognitive mistakes, symptoms, work satisfaction
  or general health.
• Complications of the disease do affect QOL
  whereas therapeutic and glucose control policies
  shown to reduce the risk of complications have no
  effect.
• Only 3 of patients not on insulin compared to
  27 of patient on insulin reported two or more
  hypoglycemic episodes in the previous year.
• UKPDS QOL data does not confirm as reported in
  some studies that insulin therapy has an adverse
  impact on QOL than other therapies.

UKPDS Prospective Diabetes Study Group, Diabetes
Care Vol 22, N 7, July 1999, pp1125-1136
26
SMBG in Type 2 Diabetes Study Analysis

• Welschen, L.M. et al1
• Systematic review of 6 studies meeting the
  criteria for assessing the impact of SMBG in
  patients with type 2 diabetes not taking insulin
  indicated that
• The overall effect of SMBG was a statistically
  significant decrease of 0.39 in A1C compared to
  control groups.
• Based on UKPDS results this reduction is expected
  to reduce the risk of complications by 14.1
• Jansen, J.P2
• Meta-analysis of direct and indirect comparison
  of studies from 1966 -2005 indicated that
• Interventions with SMBG seem to be more effective
  the interventions without SMBG
• An additional A1C reduction of 0.4 was observed
• Interventions with SMUG (urine testing) were no
  more effective than interventions without
  self-monitoring
• Sarol, J.N. et al3
• Meta- analysis of 8 studies involving SMBG as
  part of multi-component therapy in type 2
  non-insulin users indicated that
• Type 2 non insulin using patients on intervention
  with SMBG when integrated with educational advice
  achieved significantly greater A1C reductions the
  on interventions without SMBG
• Mean testing frequency of 5 to 7 time per week
  was associated with better control
• Coster, S. et al 4
• Meta-analysis of 8 studies on monitoring in type
  2 diabetes indicated that no study had sufficient
  power to detect differences in A1C of less the
  0.5
• No evidence for clinical effectiveness

1. Welschen, L.M. et al, Diabetes Care, Vol 28, No
   6, Jun 2005
2. Jansen J. P. Curr Med Research and Opinions, Vol
   22, No 4, 2006 671-681
3. Sarol, J. N., Curr Med Research and Opinions, Vol
   21, No 2, 2005 1730-183
4. Coster, S., Diabetic Medicine, 17, 755-761

27
A1C and Average Plasma Glucose
1
A1C 4 5 6 7 8 9 10 11 12
Mean plasma glucose mmol/l 3.5 5.5 7.5 9.5 11.5 13.5 15.5 17.5 19.5
mg/dl 65 100 135 170 205 240 275 310 345
Relative Contribution of Plasma Glucose Toward
Measured A1C Value
1
2
50
25
15
10
120
0
90
30
60
Days Since Test
Blood glucose levels that were above a certain
threshold at different times of the day made
good glucose control very unlikely2
1. Rohlfing, C.L et al, Diabetes Care, vol 25, No
2, Feb 2002, 275-278 2. Monnier L et al. Diabetes
Care. 200326881-885.
28
The Costs of Complications in US

• The annual cost of care in 1993 dollars for a and
  average male age 64 years1
• No cardiovascular complications 2033
• With initiation of cardiovascular
• complications 3120
• After major CV event 9385
• Patient with diabetes are 3- 10 time more
  expensive in terms of healthcare costs.1
• Treatment for CHD disease and CVD events are the
  most costly portion of care.
• In adults with diabetes CHD, hypertension, and
  depression are more strongly predictive of future
  costs than A1C level.2
• A1C is insignificantly predictive of cost when
  lt7.5
• Per capita health care cost in 20023
• With diabetes 13,243
• Without diabetes 2,560

• Brown, J. B. et al, Arch Internal Med, Vol 159,
  Sep 13 1999
• Gilmer, T. P. et al, Diabetes Care, Vol 28, No 1,
  Jan 2005, 59-64
• Economic Cost of Diabetes in the US in 2002,
  Diabetes Care Vol 26, No 3 Mar 2003

29
Costs of type 2 Diabetes in Europe
The Cost of treating complications is far greater
than the cost of therapy for glucose control.

• Average annual cost per patient was estimated to
  be 2834 EUR (1999 values)
• Cost distribution revealed the greatest
  percentage of expenditures for patients with
  diabetes were
• Hospitalizations 55
• Cardiovascular and
  lipid lowering agents 42
• Glucose lowering medications represent a much
  lower percentage of total health care
  expenditures
• Antidiabetic medications represent only 11.9 of
  average total costs per patient
• Insulin represents 9.6 of average total costs.
• Cardiovascular agents represent 29of average
  total costs

Jonsson, B. Diabetologia (2002) 45 S5-S12
30
SMBG Improves Metabolic Control

• Auto-Surveillance Intervention Active (ASIA)
  Study
• SMBG Training Provided with requirement of at
  least 6 tests per week
• A1C at baseline
  Fasting BG at baseline
• Control 8.9 1.3 7.5
  4.8 mmol
• SMBG group 9.0 1.3 7.2
  5.1 mmol
• A1C change 3 mo.
  Fasting BG at 3mo. Endpoint
• Control -0.45 1.27 6.91 4.56 mmol
• SMBG Group - 0.70 1.28 6.66
  4.83 mmol
• Results
• No significant difference in Fasting BG between
  the two groups at 3 or 6 mo. Endpoints
• A greater of patients in the SMBG group showed
  improvement in A1C.
• Improvement was greater with higher Baseline A1C
  values
• The percentage of patients following dietary
  instructing remained stable in the SMBG group.
  Whereas in the conventional group it declined.
• Patients were trained to monitor their blood
  glucose levels
• Which would give them the necessary feedback to
  support changes in lifestyle

Guerci, B et al, Diab Metab. 2003, 29, 587-594
31
Glycemic Control and Costs

• Getting to goal saves money ( A1C lt7)1
• In a retrospective, longitudinal study database
  analysis after controlling for confounding
  factors revealed
• Diabetes related costs for Above-target group 32
  higher than the at target group.
• 1540 vs. 1171 per patient per year
• Diabetes
  Medical Costs Diabetes
  Pharmacy Costs
• At Target 534 663
• Above Target 682 924
• Patients at target used fewer health care
  resources than patient above target

• Improved glycemic control of type 2 DM is
  associated with substantial short term
  symptomatic QOL and health economic benefits2.
• Improving control save an estimated 11 per
  patient per month
• Patients given glipizide compared to placebo
  achieved better glycemic control, better QOL and
  fewer days off work
• In seven studies from multiple countries the
  benefit-cost gained from self-management training
  ranged was for every 1 spent a net savings of
  0.44 to 8.76 was gained.3

1. Sharaschandra, S, et al, JMCP Vol 11, No 7,
559-564 2. Testa, M.A., JAMA, Nov 1998, Vol 280,
No 17, 1490-1496 3. Klonoff, D. C., Schwartz,
D.M., Diabetes Care, Vol 23, No 3, Mar 2000,
390-404
32
Cost - effectiveness of
Intensive Glycemic Control

• Based on UKPDS and other data the Incremental
  cost-effectiveness for intensive glycemic control
  is estimated at 41384 per Quality Adjusted Life
  Year (QALY) This ratio increased with age from
  9614 per QALY for patients aged 25-34 years to
  2.1million for patients aged 85 to 94 years.1
• Patients would be expected to live longer thus
  increasing cost but the cost for complications
  decreased by about 12
• Intensive glucose control increased treatment
  costs by 695 per patient but reduced the cost of
  complications by 957 compared to conventional
  treatment.2
• When trial visits and tests were replaced with
  those likely in clinical practice the cost per
  patient for conventional vs. intensive management
  was not significant.2
• The intensive policy itself does not seem to
  affect QOL although more hypoglycemia and weight
  gain may occur.3

1. CDC Diabetes Cost-effectiveness Group, JAMA, Vol
   287, No 19, May 15,2002, 2542-2551
2. Gray, A, et al, BMJ, Vol 320, 20 May 2000,
   1373-1378
3. UKPDS Study Group on Quality of Life, Diabetes
   Cae, 1999,Vol 22, 1125-1136

33
Managed Care Benefits of
Improved Glycemic Control
Poor Glycemic Control Increases Short Term Health
Care Costs

• In a large US managed care plan, over a period of
  3 years mean adjusted inpatient charges for acute
  short term complications for 2,394 patient with
  diabetes were1
  w/ no long-term w/ long-term
• complications complications
• Good glycemic control (A1C,8) 970
  2,610
• Fair glycemic control (A1C 8-10) 1,380
  3,810
• Poor glycemic control (A1C gt10) 3,040
  8,320
• Improving Glycemic Control was Cost Effective in
  the Long Term in USA, Canada and UK
• The pharmacy component accounts for only 20-30
  of overall costs for MCO patients with DM. The
  balance is spent on complications and
  co-morbidities such as hypertension and
  hyperlipidemia.2
• Total costs increase for the first 2 months after
  initiating insulin then decrease by 40 over the
  next 2 to 8 months
• Patients gt60 years using insulin experience a
  reduction of 42 in total costs and 32 in
  ambulatory costs relative to patient on oral
  antibiotic agents.
• Programs associated with the greatest
  improvements in A1C levels include
• Pharmacist evaluation and counseling
• Medication adjustment
• Physician / patient interactions

Defined as infections, hyperglycemia,
hypoglycemia, electrolyte disturbances and
associated medical charges.

1. Menzin, J. et al, Diabetes Care, Vol 24, No 1,
   Jan 2001, 51-55
2. Stephens, J. M. JMCP, Vol 12, No 6, Mar 2006

34
Managed Care Benefits of
Improved Glycemic Control ( cont)

• Sustained reduction in A1C level among adult
  patients with diabetes in Washington State HMO is
  associated with significant cost savings within 1
  to 2 years of improvement.
• Mean Adjusted Savings in Total Health Care Costs
  by Baseline Complication. (1992-1997)
• Cardiovascular diseas e 882
• Other Complications 802
• No Complications 438

Health Care Costs by Year (n4744)
Baseline A1C 1992 1993 1994
lt8 unimproved 5400 4692 6363
lt8 improved 5121 Dif -279 3683 Dif -1009 4475 Dif -1885
8-10 Unimproved 5161 4546 5898
8-10 Improved 4211 Dif -950 6186 1640 5898 Dif -692
gt10 Unimproved 4325 6271 7947
gt10 Improved 4325 Dif 722 6271 Dif -378 7947 Dif 141
Wagner, E.H., JAMA, Vol 285, No 2, Jan 2001,
182-189
35
Significant Reductions in Complications Result
from Every 1 Percentage-point Decrease in A1C1

• Type 2 Patient in the UKPDS intensively treated
  with sulfonylureas or insulin achieved an A1C
  significantly lower than those treated
  conventionally, were less likely to develop
  microvascular complications, but had increased
  risk of hypoglycemia.1
• Obese patients with type 2 diabetes who were
  originally diet treated were intensively treated
  with metformin and achieved an A1C significantly
  lower than those treated conventionally and were
  less likely to die or develop any diabetes
  complications including myocardial infarction. 2

UKPDS Benefits of 1 reduction in HbA1C3
1. UK Prospective Diabetes Study (UKPDS) Group.
Intensive blood-glucose control with
sulphonylureas or insulin compared with
conventional treatment and risk of complications
in patients with type 2 diabetes (UKPDS 33).
Lancet. 1998352837853 2. UK Prospective
Diabetes Study (UKPDS) Group. Effect of intensive
blood glucose control with metformin on
complications in overweight patients with type 2
diabetes (UKPDS 34). Lancet. 1998352854865 3.
Braunstein S et al. J Manag Care Pharm. 2005
11S1-11.
36
SMBG is a Cost-effective Method of Helping
improve Glycemic Control

• A cost-utility found that SMBG was cost-effective
  in a UK setting for oral agents.1
• Incremental cost-effectiveness ratios (ICERs) for
  SMBG vs. non SMBG produced values of
• 15515 per QALY for PTs on Diet and Exercise
• 4508 per QALY for PTs taking OADs
• 4593 per QALY for PTs taking insulin
• The additional treatment costs associated with
  SMBG were, for a large part, offset by the
  reduced cost of diabetes related complications.

Mean Life Expectancy (Yrs) QALY Lifetime costs () SMBG Lifetime Costs () NO SMBG Likelihood of being cost effective
Diet Exercise 12.3 (6.34) 20668 18105 51
OADs 11.80 (6.15) 21650 20636 51
Insulin 10.33 (5.26) 23712 22541 55
At a willingness to pay threshold of 30000 for
each QALY gained.
QALY Quality Adjusted Life Year OADs Oral
Antidiabetes Medications
1. Palmer, A. J. et al, Curr Med Research and
Opinion, Vol 22, No 5, 2006, 861-872
37
SMBG Impact on Patient Motivation
Pros
Cons

• Display of blood glucose figures appeared to
  render their invisible and imperceptible illness
  visible.1
• Patients often report personal gratification on
  obtaining low readings.1
• Monitoring appeared to bolster a perception that
  diabetes management is the patients
  responsibility and cultivated independence from
  health services and enhanced self-regulation.1
• Some patients conveyed the direct impact glucose
  monitoring had on their diet by encouraging
  appropriate modifications in light of higher
  readings.1
• Decreases in A1C levels over time was present
  only among individuals who changed their
  treatment.2
• Patients performing SMBG with a frequency of gt
  1/day or gt1/week were more likely to report a
  decrease in the frequency of hypoglycemic
  episodes during follow-up than patients not
  performing SMBG.2

• Blood glucose parameters were problematic when
  the felt they were receiving either contradictory
  information about upper thresholds or no guidance
  about ideal parameters.1
• Most patients were clear as to how to counteract
  hypoglycemia but many lacked awareness of how to
  manage hyperglycemia.1
• Increased self-responsibility was often
  accompanied by self-blame and negative reactions
  to high glucose readings.1
• Not knowing a reason for high blood glucose
  readings were reported as sources of distress and
  anxiety and sometimes effected adherence to
  diabetic regimens by promoting nihilistic
  attitudes.1
• Some patients felt SMBG was a waste of time if
  health care professionals were not interested in
  their readings.1
• In a large sample of non-insulin treated type 2
  patients SMBG did not predict better metabolic
  control over a 3 year period.2

In all studies showing a positive impact of SMBG
on metabolic control, the practice was part of a
comprehensive educational program. 2 SMBG could
result as ineffective because of the inability of
the patients to use the SMBG information to
undertake specific actions.2 Physicians do not
fully utilize the information deriving from SMBG
to adjust treatment.2

1. Peel, E., et al, Brit Journal of General Practice
   Med, Mar 2004, 183-188
2. Franciosi, M. et al, Diabetes UK. Diabetic
   Medicine, 22, 2005, 900-906

38
SMBG in Type 2 Diabetes
Long-Term Outcomes
In a German Multicentre Retrolective Study
following 3,268 patients between 1995 and 2003
All Patients n3268
Experienced Non-fatal endpoint Experienced Fatal endpoint
With SMBG 7.2 2.7
Without SMBG 10.4 4.6

• SMBG was associated with a 32 reduction in
  combined non-fatal endpoints, despite an increase
  in microvascular events, and a 51 reduction in
  mortality over the period.
• Possible explanation for the clinical impact on
  mortality
• Immediate feedback on the effects of diet and
  exercise SMBG may provide could enhance patient
  empowerment.
• SMBG has been shown to be associated with
  improved patient compliance.
• Patients using SMBG visited the treating
  physician more frequently demonstrating that SMBG
  has the potential to change patient attitudes.
• SMBG may alter physician attitudes and may
  strengthen their ability to teach self-management
  skills and motivate patients to make behavioral
  changes.

p0.002 p0.004
Of Patients that did not receive insulin n2515
Experienced Non-fatal endpoint Experienced Fatal endpoint
With SMBG 6.7 2.5
Without SMBG 10.4 4.3
p0.002 plt0.001
Martin, S. et al, Diabetologia, (2006) 49271-278
39
The Impact of
Disease Management Programs
Study Study Type No. of patients or studies analyzed Study Conclusions Study Setting
Chodosh, J. et al Ann Inter Med 2005 Meta-Analysis of trials comparing self-management interventions vs. controls 26 Diabetes 13 hypertension Self-management programs produce clinically important and statistically significant benefits. Mean change in A1C of -0.81 Patients are more likely to derive benefit from a cycle of intervention and individual review with the provider than from interventions with no review. Various
McMurray, S. D. et al Diab Care Mgmt in Dialysis Unit 2002 Randomized Controlled Diabetes care manager N91 Study group 45 Control Group 38 Study group showed statistical significant improvement in quality of life, improvement in A1C at 6 and 9 months but not 12 months. Control group experienced significant progression of diabetes related neuropathic and vascular disease from baseline. Dialysis unit
Choe, H. M et al Am J of Managed Care Vol 11 No 4 Apr (2005) Randomized Controlled Pharmacist based management N80 Completed trial Intervention group N36 Control group n29 Mean decrease in A1C in the intervention group was -2.1 vs. -0.9 in the control group. Primary Care Clinic
Varroud-Vial, M et al Diabet. Med 21, 592-598 (2004) Randomized Controlled Education French GPs on the SDM program N340 Intervention group n192 Control group n148 Educating general practitioners on the adaptation of the Staged Diabetes Management improved glycemic control in the primary care setting. Screenings for complications was performed more often,. A1C values decreased by 0.31 in the intervention group vs. an increase of 0.56 in the control group. Incremental cost between the groups was not significant. Primary Care settings
Menard, J. et al CMAJ 173(12) Dec 2005 Randomized Controlled Team Education N 72 Intensive multitherapy n36 Usual care n36 At 12 mo. A higher proportion of patients in the intervention group had achieved Canadian Diabetes Association Goals for A1C (35 vs ), diastolic BP (64 vs 37), LDL cholesterol (53 vs 20)and triglyceride levels (44 vs 14) than the control group. No difference was seen if the fasting glucose levels. All patients resumed normal care after 12 months and at 18 mo there were no significant differences in results other than LDL chol. Hospital based Clinical Research Centre
Gaede, P. et al NEJM 3485 Jan 2003 Open parallel trial Multifactorial intervention N160 Completed Intensive treatment n67 Standard treatment n63 Decline in A1C, systolic and Diastolic BP, serum Chol, triglycerides, and urinary albumin were all significantly greater in the intensive treatment group vs the control group. General Practitioners
Gaede, P, et al Lancet Vol 353 Feb 1999 Open parallel trial Multifactorial intervention N160 Completed Intensive treated n73 Standard treatment n76 4 years intensive multifactorial treatment slowed the progression to nephropathy and autonomic neuropathy in patients with type 2 DM and microalbuminuria. General Practitioners
40
The Impact of
Disease Management Programs (cont)

• Teaching patients to monitor their risk factors
  and providing therapeutic targets and goals
  yields well informed and motivated patients that
  are more insistent to reach and maintain target
  values of the main risk factors of diabetes
  complications.1
• Compared to standard care at the end of a 4 year
  study a greater percentage of these patients1
• Were taking ACE-I or ARBs (100 vs. 54)
• Were taking Statins (81 vs. 23)
• Were taking Fibrates (20 vs. 6)
• Were taking ASA (94 vs 57)
• Had statistically significant lower BMIs,
  Systolic and Diastolic BP, LDL and A1C values.
• Providing intensive group education with skills
  training melded into daily medical management
  provides significantly greater drops in A1C than
  standard care with quarterly educational
  mailings.2
• Patient receiving intensive treatment also
  performed more frequent SMBG and paid more
  attention to dietary fats and carbohydrates.
• Frequency of nurse case manager follow-up
  contacts is positively linked to better A1C
  outcomes.2 Addition of insulin does not appear to
  be a significant contributor to glycemic change.2

1. Rachmani, R, Diabetes UK. Diabetic Medicine, 19,
   385-392
2. Polonsky, W. H., Diabetes Care, Vol 26, No 11 Nov
   2003

41
The Effects of SMBG and Timely Follow-up

• In one study where patients at an outpatient
  clinic where treated using an Internet Based
  Glucose Monitoring System (IBGMS) for 3 months
  vs. the usual outpatient management over the same
  period, A1C levels were significantly lower in
  the intervention group.1 (-0.54 vs 0.33)
• Advantages of this SMBG reporting system1
• More frequent reporting of SMBG results
• Patients ability to ask questions and receive
  feedback in a timely manor
• More frequent contact with physicians via the
  internet with medical advice provided based on
  most current data.

1. Kwon, H, et al, Diabetes Care, Vol 27, No 2,
Feb 2004, 478-483
42
Diabetes Education with Care Management
Communication

• In a study involving patients in a dialysis unit,
  diabetes care managers performed the following
  activities in the study group1
• Self management education
• Diabetes care monitoring and management
• Motivational coaching
• Initial nutrition counseling
• The major changes made in glucose management
  were
• Increasing SMBG
• Identifying the physician managing the diabetes
• Forwarding results from SMBG
• Reductions in A1C and BG levels were not
  associated with increased hypoglycemia.
• Statistically significant reductions in A1C were
  seen at 6 and 9 months but not 12 months.
• The control group (standard care) experienced a
  significant progression in diabetic-related
  peripheral vascular disease from baseline to 12
  months where as the study group did not.

• A systematic review of the research literature by
  van Dam , et al suggests2
• Consider more focus on programs for directly
  enhancing patient participation
• Assistant-guided preparation for patient-provider
  encounters, patient empowering group education,
  group consultations or automated telephone
  management with nurse support.
• Combine support to improve provider participatory
  behavior in consultations and education.

1. McMurray, S. D et al, Am Jour Kidney Diseases,
   Vol 40, No 3, Sept 2002 566-575
2. Van Dam H. A. et al, Patient Education
   Counseling, 51, 2003, 17-28

43
Diabetes Education with Care Management
Communication (cont)
A meta-analysis of 63 articles on Randomized
Educational and Behavioral Interventions in Type
2 diabetes suggests

• Educational and behavioral interventions in
  patients with type 2 diabetes produced a moderate
  decline in A1C of 0.43 which was statistically
  significant.
• Studies with physician interventions produce
  larger declines in A1C than those with nurses or
  dietitians -0.71 and -0.88 respectively.
• Intervention type and effect
• Physician -1.8 (p.229) Group setting -0.70 (p.
  015)
• Nurse -0.71 (p.022) Individual -0.62 (p.005)
• Dietitian -0.88 (p0.43)
• Topic Areas
• Medication -0.71 (p.032) Exercise -0.69 (p.00
  7)
• Diet -0.51 (p.008) SMBG -0.20 (plt.001)
• It is possible the physician interventionists
  were able to manipulate medication regimens,
  therefore producing better effects on glycemic
  control.

Gary, T. L. et al, Diabetes Educator, Vol 29, No
3, May/June 2003