Transcript Treatment of Type 2 Diabetes and Outcomes in Patients With Heart

Macdonald, M. et al. Diabetes Care; Jun 2010; 33, 6. 1213-1218
Carina Signori, DO
Journal Club
August 2010
Introduction
 Treatment guidelines for diabetes (DM) in heart
failure (CHF) patients is controversial.
 Evidence is insufficient:
 CHF patients have been excluded from trials of glucose
lowering drugs
 Safety of diabetic medications in CHF patients still
unclear.
Introduction
 Several observational studies have shown prognostic
differences between various diabetic meds in CHF
patients
 However, no placebo group is available to compare
prognosis
 Also, no data was available for potential confounders
in CHF
Objective of Study
 To examine mortality risk in patients with diabetes
and heart failure
 To examine whether outcomes were associated with a
particular antidiabetic drug therapy.
Methods
 Case control study
 Using the U.K. General Practice Research Database
 Cohort with information on comorbidities and
treatments
 Collected from >450 general practictioners in U.K.
 Often used for studies of benefits and harms for
prescription medications.
 Includes patient demographics, physical and lab data,
diagnoses, outpatient medications.
 Clinical diagnoses assigned by PCP .
Methods
 Clinical comorbidities were coded from the time
patient entered into the database up until index date.
 Code selection for each diagnosis was done by 2
researchers independently and cross checked by a
third.
Methods
 Study sample
 Inclusion criteria: >35yo with newly dx’d type 2 DM and
CHF between January 1988-2007 with >1year of data
 Exclusion: Dx of DM or CHF <1988, type I DM,
gestational DM, or drug induced DM
 All patients were followed until death, termination from
database, or October 2007.
Methods
 Case subjects = patients in cohort with DM and CHF
who died (all cause mortality)
 Control subjects:
 Matched to cohort patients on age (+-5 years), sex,
general practice, calendar year, years of follow up within
GPRD.
 Patient had to be alive on index date (date case subject
had died).
Methods
 Definition of medication exposure:
 Current use = 1 prescription recorded in 90 days prior to index
date.
 Classification of antidiabetic drug exposure

No drug tx, sulfonylurea alone, metformin alone, TZD alone, insulin
alone, combination therapy with insulin, combination oral without
insulin.
 Evaluated any use of diabetic medication in a sensitivity
analysis 90 days prior to index date.
 Adjusted for duration of DM, CHF and which developed
first.
 Drug use was examined 6 months and 12 months prior to
index date.
Statistical analysis
 Conditional logistic regression used to estimate crude
and adjusted odds ratios.
 Numerous potential confounding variables were
adjusted for.
 Analysis was conducted using SAS version 9.2
 P value of 0.05 was used.
Results
 8404 patients found to have DM and CHF dx’d between




1988 and 2007.
1633 patients who died were matched with 1633 control
subjects.
Mean living with DM and CHF was 2.8, 2.9 years in case
subjects and control subjects respectively.
DM diagnosed first in 54% of case patients and 41% of
control subjects (mean time of dx: 3.9, 3.3).
CHF diagnosed first in 44% of case subjects and 58% of
control subjects (mean time of dx: 3.1, 3.2 respectively)
Results









Mean age was 78 years at index date.
1738 male (53%)
Average time followed in GPRD was 11 years.
Case and control subjects matched well in age, sex and time
within GPRD.
Case subjects had higher rates of comorbidities and abnormal
lab values
Poor prognostic factors evaluated: hypotension, elevated cr,
anemia, copd, cancer, dementia, cva, prior MI
Elevated BMI was associated with lower mortality risk.
644 (18%) of patients received ACEI/ARB and beta blockers. 271
(12%) received ACEI/ARB, BB, ASA, statin.
ACEI/ARBS, BB, ASA, Dig, statins all independently associated
with reduced mortality risk in pts with CHF and DM.
Results
 Diabetes treatment
 Off medications


1306 patients (40%) were not exposed to antidiabetic meds in
the 90 days prior to the index date.
It was presumed they were using diet and lifestyle
 On medications

Sulfonylurea was most common (n-753, 25%), then
combination oral therapy without insulin (n=470, 14%), then
metformin monotherapy (n=378, 12%)
Results
 All cause mortality
 Unadjusted analyses= compared to patients not on
antidiabetic therapy, users of sulfonylureas, metformin,
or combination antiDM meds exhibited lower mortality
risk.
 Adjusted analysis = only current use of metformin
monotherapy associated with covariates (adjusted OR
0.65 [0.48-0.87]
Results
 Metformin alone had lower mortality even when
 duration of DM and CHF were adjusted for (0.63 [0.470.86],
 GFR was included as a continuous variable (0.68 [0.490.93]
 If DM was adjusted for diagnosed first (0.65 [0.48-0.88].
 No association between current use of other diabetic
medications and all-cause morality
Results
 Outcomes for “any use” of drugs=
 Only metformin was associated with all-cause mortality
reduction (OR .72 [-.59-0.9], p=0.003).
 No association seen for any of the other medications
(p>0.2)
Conclusions
 Compared with individuals not exposed to antidiabetic
drugs, metformin use is associated with lower
mortality risk than other medications even after
adjusting for other prognostic factors.
 This is consistent with other study findings suggesting
metformin has a positive effect instead of harm from
other agents.
 Metformin may improve heart failure based on
improving insulin resistance.
 There was no association between A1C and mortality
Conclusions
 Mortality risk was improved in diabetic and heart
failure patients on ACEI and BB.
 However only 18% of study patients were on both.
Conclusions
 Weaknesses:
 High amount of patients not on antidiabetic
medications (40%).

Authors presumed these patients were doing dietary or
lifestyle modifications
 Data collected from UKGPRD depends on physician
diagnoses or documentation of heart failure
 There could be a selection bias- metformin patients may
have less severe diabetes
 Observational study so confounding variables may be
present.
Conclusion
 The results suggest metformin has better outcomes
(compared to other DM meds) in patients with DM
and CHF.
 This is consistent with observational studies
 Metformin reduces mortality risk
 compared to age and sex matched diabetics on no
antidiabetic medications
 Independent of glycemic control, BMI and other
prognostic factors.
 Metformin can be used patients with heart failure to
treat diabetes.