Transcript General Medicine Update

General Medicine Update
Minnesota ACP
November 7, 2008
Steve Hillson
Hennepin County Medical Center
University of Minnesota
[email protected]
Objectives
• At the end of this session you should be
able to:
– Describe the main results of several
important reports from the past year
– Decide how you want to change your
practice in the context of these findings
Disclosure
• I have no direct financial relationships
with any commercial firm having any
interest in any of the reports or topics I
am about to discuss.
Process
• Personally reviewed title of every original
research article from 10/01/07 till 10/22/08 in:
–
–
–
–
–
Annals of Internal Medicine
BMJ
JAMA
Lancet
New England Journal of Medicine
• Reviewed subspecialty updates, scattered
other sources
• Personally reviewed abstract of every article
with “interesting” title.
Process (cont’d)
• Selected “promising” articles by initial
abstract review (about 100)
• Re-reviewed all abstracts, selecting about 60
with medium or high impact potential
• Solicited abstract reviews from colleagues to
select subset of greatest importance
• Critically appraised final subset for
presentation
Limitations on Process
• Personal idiosyncrasies
• Incomplete survey of medical literature
• No claim to comprehensive context for
assessing these articles
• Very simplified presentation of complex
research
• Final slide set available at
– www.paralleltext.net/ppt.html
In Pursuit of the Perfect A1C
• How intensely should we be controlling type 2
diabetes?
• 3 Important Articles
– ACCORD, NEJM, June 2008
• Funded by NIH, CDC, with drugs contributed by many
makers
– ADVANCE, NEJM, June 2008
• Funded by maker of gliclazide
– UKPDS, NEJM, October 2008
• Funded initially by UK government agencies, this followup funded by drug makers
Purpose
• Assess tighter vs looser glycemic
control in type 2 diabetes
• Previously limited information
– None showing mortality or macrovascular
benefit in type 2 DM
• But extensive promulgation of the idea
that lower is better
#1 - ACCORD
• Compare target A1C <6.0 to less tight (7-7.9)
for cardiovascular outcomes
• Clinical Trial, unblinded
– 10,000 US/Canadian patients with DM-2,
A1C≥7.5, and CV disease or risk factor
– Any standard diabetes medications
– More frequent visits and medication adjustments
for intensive therapy group
– Followed 3.5 years for CV death, MI, CVA
#2 - ADVANCE
• Compare target A1C (<6.5) to less tight (local
guideline) for vascular outcomes
• Clinical Trial, unblinded
• 11,000 patients worldwide, type 2 diabetes,
age≥55, no insulin, and pre-existing vascular
disease or a risk factor
• Gliclazide, plus frequent clinic visits and other
drugs as needed, OR
• Usual care, with gliclazide excluded
• Followed 5 years for vascular events
#3 - UKPDS 10-year follow-up
– Compare tight glycemic control (fasting glucose
108), to less tight (fasting glucose < 270) for
macro- and microvascular outcomes
– Clinical Trial, unblinded
– 4000 UK patients with new DM-2, age 25-65
– Received one of several drug-based strategies
OR
– “Usual Care” with diet alone unless FPG>270
– Treated 10 years, then followed additional 10
years on community standard care, for vascular
outcomes
Findings - Achieved A1C
8
7
6
5
Intensive
Standard
4
3
2
1
0
ACCORD
ADVANCE
UKPDS*
Findings - Primary Outcomes
60
Intensive
Standard
50
*
*
40
30
*
20
10
0
ACCORD
ADVANCE UKPDS - All
UKPDS Metformin
Findings - Death
35
Intensive
Standard
30
*
*
25
20
15
10
5
*
0
ACCORD
ADVANCE
UKPDS All
UKPDS
Metformin
Limitations
• ACCORD used a lot of rosiglitazone
• Neither ACCORD nor ADVANCE
achieved target A1C on most patients
• UKPDS “usual care” isn’t
Implications
• Target A1C of 6.5 or less is at best
ambiguous for macrovascular disease,
possibly dangerous
– May depend on drug choice
– Death (NNH of 100) trumps improved
nephropathy/retinopathy (NNT of 70)
• Metformin, without a tight target A1C, is
useful for survival in obese diabetics (NNT
about 15 over 20 years)
• I will not seek extremely tight A1C
• I will use still more metformin
Preventing the Clot
• There’s a new perioperative
anticoagulant on the block - 2 studies
– RECORD1, NEJM, June 2008
– RECORD3, NEJM, June 2008
Purpose
• Compare rivaroxaban to enoxaparin for
preventing post-op VTE
– Total Hip Arthroplasty (RECORD 1)
– Total Knee Arthroplasty (RECORD 3)
• Funded by makers of rivaroxaban
– Orally administered, fixed dose factor Xa inhibitor
– Reportedly out in January
• Related drugs
– Argatroban - parenteral
– Ximelagatran - oral, withdrawn due to liver toxicity
– Dabigatran - oral, possibly out in 2010
Method
• Clinical trials, blinded
• 2500 (knee) and 4400 (hip) patients, age≥ 18
with no hepatic or renal disease
• Given rivaroxaban 10 mg orally each day, OR
• Enoxaparin 40 mg SC each day
– KNEE study: 10-14 total days
– HIP study: 35 total days
• Followed 2-6 weeks for venographic DVT and
symptomatic VTE or death
Findings - Detectable Venous
Thromboembolism
20
18
16
14
12
10
8
6
4
2
0
Rivaroxaban
Enoxaparin
Hip
Knee
Bleeding
Limitations
• Symptomatic VTE was rare (about onetenth of all VTE events)
• Industry-funded research has many
opportunities to mislead
• Issue of spinal catheter management
not clarified
Implications
• I’m usually a turtle, but…
– I will start using perioperative rivaroxaban when it
is released
• Easier for everyone
• Question of pricing
– Not for frail or otherwise high-risk patients
– Does not replace heparin
– Watch for studies comparing it to chronic
coumadin for long term anticoagulation
– Look for dabigatran
The Infected Respiratory Tract
• Two studies of antibiotics
– BMJ, October 2008
– JAMA, December 2007
Purpose
• Assess the value of antibiotics (and
steroids) for common respiratory tract
infections
• Many guidelines and some prior
evidence
– Largely recommend against antibiotics for
most conditions in absence of pneumonia
– Acute bacterial sinusitis more equivocal
#1 - Antibiotics for common
respiratory infections
• Historical cohort study
• 1.1 million episodes of respiratory infection
(URI, “chest infection,” sore throat, otitis,) in
UK
• Record assessed for antibiotic prescription
• Followed 1 month for diagnosis-specific
complications (pneumonia, quinsy,
mastoiditis)
• Funded by UK Department of Health
Findings - Complications of
Respiratory Infections
4.5
4
3.5
3
2.5
2
1.5
1
0.5
0
Treated
Untreated
URI
Sore
Throat
Otitis
Chest
Infection
(Elderly Patients Only)
#2 - Antibiotics and topical
steroids for maxillary sinusitis
• Clinical trial, blinded
• 240 adults with < 4 weeks acute
bacterial sinusitis (purulent discharge,
local pain, pus on exam), no diabetes
• Treated with amoxicillin, budesonide
spray, both or neither
• Followed for clinical cure at 10 days
• Funded by UK Department of Health
Findings - Resolution of
Sinusitis
100
90
80
70
60
50
40
30
20
10
0
10 Day Cure
Amoxicillin
Budesonide
Nothing
Limitations
• The respiratory complication study was
not a trial
– Many ways that treated and untreated
groups may have differed
– Including getting diagnosis of complication
• The sinusitis study was small
– Could have missed difference in serious
complications
Implications
• Despite limitations
– Antibiotics don’t seem important for bacterial
sinusitis, otitis, sore throat, URI
– BUT, may be quite useful for “Chest Infection”
• Acute bronchitis?
• NNT 40 to prevent pneumonia
– I will try to use less antibiotic for sinusitis (even
acute bacterial) and otitis
– I will try to distinguish “chest infection” in older
patients and treat
How Do You See the Colon?
• Two studies of CT Colonography
– NEJM, October 2007
• Funding not reported,investigators receive
money from makers of the colonography
processing software
– NEJM, September 2008
• Funded by National Cancer Institute and
American College of Radiology
Purpose
• Determine whether a relatively non-invasive
colonic imaging technique can approach the
ability of colonoscopy to detect pre- and early
malignancies
• Colonoscopy never proven to reduce colon
cancer mortality, but almost certainly does
(FOBT does)
• Colonoscopy is expensive, inconvenient, and
not completely safe
– 1-3/1,000 have serious consequences, usually
associated with biopsies
• CT Colonography uses similar prep,
insufflation, plus fluid tagging
#1 - CT Colonography for
advanced neoplasia
– Cohort study, sort of
– 6300 adults with no bowel disorder
• Half had enrolled in a CT colonography
screening program (why?), with colonoscopy
follow-up for selected findings
• Half were getting ordinary colonoscopic
screening
– Assessed number and pathology of lesions
found
– No follow-up
#2 - Accuracy of CT
colonography
• “Test of a Test”
• 2600 adults over 50, asymptomatic,
referred for ordinary colonoscopic
screening
– First received CT colonography
– Follwed by immediate colonoscopy
• Assessed concordance for important
polyps
Findings - Cohort Study
3.5
3
2.5
2
CT
Scope
1.5
*
1
0.5
0
Advanced Adenomas
Cancers
Findings - Sensitivity Study
• CT detected
– 90% of advanced lesions ≥ 1 cm
– 65% of advanced lesions ≥ 5 mm
• CT incorrectly called abnormalities in
14% of subjects
Limitations
• First study had no direct comparison of CT to
scope in the same patient
– Why the excess of cancers in colonography?
• In both studies, CT found extracolonic stuff in
majority of patients
– Mostly trivial, often requiring further assessment
• In practice, unlikely to get immediate
colonoscopy after positive CT
– Requires repeat preps, other inconvenience
Implications
• CT Colonography still not ready for
prime time
– Difficult prep
– Lots of follow-up colonoscopies
– Lots of irrelevant findings
• I won’t be doing it
• Fecal Occult Blood for my patients who
don’t want colonoscopy
After the Fall
• Prevention after a hip fracture
• NEJM, November 2007
• KW Lyles et al.
Purpose
• Determine whether annual infusion of
zoledronic acid reduces subsequent
fracture after hip fracture repair
• Inconclusive prior evidence about
bisphsphonates following hip fracture
• Funded by the maker of zoledronic acid
Method
• Clinical Trial, blinded
• 2100 adults with recent “minimal trauma” hip
fracture, previously ambulatory, no kidney
disease, and refusing oral bisphosphonate
• Received Calcium and Vitamin D, plus
– 5 mg IV zoledronic acid or placebo infusion
annually
• Followed 2 years for new clinical fractures
and survival
Findings
14
12
10
Zoledronic Acid
Placebo
8
6
4
2
0
Hip Fx
Vertebral Fx
Any Fx
Death
Limitations
• Mortality benefit unexpected and
unexplained
• Industry-funded research has many
opportunities for misleading reports
Implications
• Bisphoshonates reduce subsequent fractures
and possibly mortality following hip fracture
repair
– NNT for another hip fx = 70 over 2 years
– NNT for death = 27 (!)
• If oral bisphosphonates aren’t an option,
zoledronic acid can be given IV yearly
– Alendronate $100/month
– Zoledronic acid $1200/year
Is the Blockade Working?
• Perioperative beta blockers
• The Lancet, May 2008
• The POISE study group
– Funded by governments of Canada,
Australia and Spain, with some support
from maker of the study drug
Purpose
• Reassess perioperative beta-blockade for
preventing cardiac complications after noncardiac surgery
• Several prior studies indicate improved postoperative cardiac outcomes with betablockade
• “Standard of care” for higher risk patients for
at least 5 years
– Some doubts due to study limitations and some
conflicting results
Method
• Clinical trial, blinded
• 8300 adults worldwide, age ≥ 45, either
existing major vascular disease or at least 3
risk factors
– Age>70, TIA, DM, CRF (2.0), CHF history,
emergent or high-risk surgery
• Received metoprolol, starting 4 hours pre-op,
or placebo
– Held for P<45 or SBP < 100
• Followed 1 month for major vascular
outcomes and death
Findings
7
Metoprolol
Placebo
6
5
4
3
2
1
0
Composite
Endoint
MI
Stroke
Death
Limitations
• Beta-blocker started immediately pre-op
• Drug held only for “consistent” severe
bradycardia or hypotension
• Excluded patients whose physicians
had planned to beta-block
Implications
• Perioperative beta-blockade, at least as done
in this study, may be dangerous
• I’m limiting my use
– Only beta-block if otherwise indicated
– Only with plenty of advance time for slow uptitration (a month!)
– Not in higher stroke risk setting
• (Sad sigh…)
All you need is…Salt?
• Saline or bicarbonate for preventing
contrast nephropathy
• JAMA, September 2008
• SS Brar et al.
Purpose
• Reassess whether bicarbonate infusion
reduces contrast nephropathy
• Prior evidence that contrast nephropathy is
common, around 25% of high-risk patients
• A few prior reports showed reduced
nephropathy with pre-procedure bicarbonate
hydration
• Funded by Kaiser Permanente
Method
• Clinical Trial, unblinded
• 350 adults having non-emergent cardiac
catheterization, with GFR ≤ 60 and at least 1
of:
– DM, CHF, HBP, Age > 75
– Received either Sodium Bicarbonate, 150 meq in
1 liter D5, OR Normal Saline.
• 3 ml/kg/hour for 1 hour pre-procedure, then 1.5
ml/kg/hour during and 4 hours after
– Followed 4 days for 25% fall in GFR
Findings
16
14
12
10
Bicarbonate
Saline
8
6
4
2
0
25% Fall in GFR
0.5 Cr Rise
Limitations
• Relatively small study
• Only coronary angiography patients
• Relatively good baseline GFR
Implications
• Bicarbonate might not be necessary for
renal protection from contrast dye
– Saline hydration probably acceptable
substitute
• However
– Bicarbonate is not hard or apparently
dangerous to use
– Should certainly use some form of
hydration
Staying Off the Sauce
• Baclofen to maintain alcohol abstinence
• The Lancet, December 2007
Purpose
• Assess whether baclofen can help achieve
and maintain alcohol abstinence in cirrhotic
alcoholics
• Growing interest in several drugs to help
prevent alcohol craving and relapse
– Naltrexone, acamprosate, topiramate
• Limited information, particularly in cirrhotic
patients
• Funded by Italian government
Method
• Clinical Trial, blinded
• 84 adults, age 18-75, with alcoholic cirrhosis,
at least 14 (women) to 21 (men) weekly
drinks, and no other major system disease
• Admitted, given baclofen 5-10 mg tid, for 12
weeks, or placebo
– Also frequent visits with counseling
• Followed 4 months for self- and familyreported abstinence
– Dropouts assumed to be relapsed
Findings - Abstinence from
Alcohol
80
70
60
50
Baclofen
Placebo
40
30
20
10
0
Total Abstinence
Limitations
• Small study
• Many dropouts, assumed relapsed
– But similar results if assumed abstinent
• Duration only 3 months
• Used in context of additional support for
abstinence
Implications
• I will try using it
– High gain, low risk (NNT 2.5)
– Avoid in renal dysfunction, epilepsy
– Attempt to provide broader treatment
context
• But I’m not pushing this
Also Noted
• N-3 Polyunsaturated fatty acid
supplementation may reduce 3-year mortality
in CHF, NNT=60.
– Lancet, 10/4/2008
• Telling smokers their “lung age,” derived from
FEV1, may improve quit rates, NNT=14
– BMJ, 3/6/2008
• Arthroscopic debridement and lavage does
not help the osteoarthritic knee more than
medicine and PT
– NEJM, 9/11/2008
More “Also Noted”
• In new type 2 (Irish) DM on oral treatment,
home glucose monitoring did not improve
A1C but did worsen depression and anxiety
– BMJ, 4/17/2008
• Low-dose risperidone may improve response
in depression refractory to monotherapy,
NNT=7 (Industry funded)
– AnnIntMed 11/6/2007
• The US Preventive Services Task Force still
does not recommend prostate cancer
screening, and recommends against it after
age 75
– AnnIntMed, 8/5/2008
And Last • Coffee might decrease cardiovascular and
overall mortality
• At 6 cups per day, over 25 years:
– Men were 20% less likely to die
– Women were 17% less likely to die
– Independent of caffeine
• WARNING: Brought to you by the Nurses’
Health Study
– Remember HRT?
• AnnIntMed, 6/17/2008
Summary
• Reconsider the A1C goal, use more
metformin
• Oral thrombin inhibitors for perioperative DVT
prophylaxis look promising
• Avoid antibiotics for most non-pneumonia
respiratory infections; “chest infection” in the
elderly may be an exception
• CT Colonography is pretty good, not yet
ready
• Bisphosphonates may be important after hip
fracture
Summary, cont’d
• Perioperative beta-blockade looks more
risky than helpful
• Saline may be as good as bicarbonate
for IV dye renal protection
• Baclofen may help alcohol abstinence in
cirrhotics
• Coffee?
Remember:
• Before acting on anything you heard
here, you may wish to study the original
research, and discuss with colleagues
or domain experts