Transcript General Medicine 1

General Medicine I
Jarrett J Weinberger M.D
High Value Care
• We spend too much $$ (18% of GDP)
• Misuse and overuse account for most waste
• HVC balances potential benefits with potential
harms and costs with focus on outcomes
• www.choosingwisely.org
• http://hvc.acponline.org/index.html
High Value Care
Medical Literature- Study Designs
•
Experimental (RCT, cluster-randomized, quasi-experimental)
• Subjects and interventions determined at study onset and blinded to
minimize bias
• S: can determine causation
• W: expensive, time-consuming, generalizability
•
Observational (cohort, case-control, cross-sectional, case series)
• Compare 2 or more naturally existing groups
• S: inexpensive, used for rare diseases, mimics practice settings, quick
• W: risk of bias, requires complicated statistical analysis
•
Systematic Reviews (comprehensive synthesis of current literature)
• S: comprehensive review using large amounts of data
• W: high variability of studies included in analysis
Medical Literature- Study Designs
Medical Literature- Study Designs
Medical Literature- Validity
• Validity (trustworthiness) has multiple threats
• sampling errors
• measurement errors
• data analysis errors
• Internal Validity
• extent that study results are true and supported by
study
• External Validity
• generalizability to other settings
Medical Literature- Validity
• Random Errors
• due to chance
• reduced by increasing
measurement precision
sample
size
and
improving
• Systematic Errors
• results from bias which influences study results
• can NOT be reduced by sample size, must eliminate bias
• Confounding Effect
• outcome influenced by hidden extraneous variable
Medical Literature- Statistics
•
Confidence Intervals (95%)
•
•
•
•
•
every research finding reflects some error
“I am 95% sure the true measure lies within the range”
affected by sample size
sample size is proportional to a study’s “power” which can determine
minimum sample size for detecting association
P Value (<0.05)
•
•
•
•
indicates likelihood that result of study happened by chance alone
represents a “1 in 20” chance of obtaining results by chance
assumes no difference in study groups
STATISTICAL SIGNIFICANCE AND CLINICAL IMPORTANCE
ARE NOT THE SAME
Medical Literature- Statistics
P Value example:
• study showed association between flu vaccine and GBS (p=
0.02)- YIKES
• relative incidence was 1.45 (CI= 1.05-1.99)
• baseline GBS risk = 10 per million
• CI indicates increased risk of 5% to 100% or “double”
• worst case risk= 20 per million, NNH= 100,000
• best case risk= absolute increase of 0.5 per million, NNH=
2,000,000
Medical Literature- Statistics
• Sensitivity- “SNOUT” “few false Negatives”
• ability of test to detect disease when truly present
• =TP / (TP+FN)
• independent of prevalence
• Specificity- “SPIN” “few false Positives”
• ability of test to exclude disease when truly not present
• =TN / (TN=FP)
• independent of prevalence
• Predictive Values (PPV, NPP)
• reflect likelihood that disease is present/absent
• dependent upon prevalence
Medical Literature- Statistics
Medical Literature- Statistics
Medical Literature- Statistics
• LR’s
simplify application of diagnostic tests
+LR of 2, 5, 10= 15%, 30%, 45% increase
-LR of 0.5, 0.2, 0.1= 15%, 30%, 45% decrease
ratio of the probability of a given test result (+/-)
among patients with a disease to the probability of the
same test result among patients without the disease
• can be used for any test where sensitivity and specificity
are known
•
•
•
•
Medical Literature- Statistics
• Relative Risk
• compare rates of events in two study groups
• may exaggerate outcomes
• Absolute Risk
• represent total differences in outcomes between the two groups
• Example
• intervention reduces rate from 40% to 20% and another reduces
the rate from 4% to 2% both have RRR= 50%
• If a RRR of 50% is reported how can you determined which
intervention is better?
• intervention 40% to 20% has ARR of 20% versus 2% for the 4%
to 2% intervention
Medical Literature- Statistics
• Numbers Needed (NNT, NNH)
• estimates number of patients requiring treatment until
benefit (NNT) or harm (NNH) is achieved
• provide a sense of magnitude expected from an
intervention
• reciprocal of the change in absolute risk
• Example
• Intervention 40% to 20% (ARR=20%), NNT= 1/.2= 5
• Intervention 4% to 2% (ARR=2%), NNT= 1/.02= 50
Medical Literature- Statistics
Routine Care of Healthy Patient
• Multiple trials have failed to show benefit of periodic
health exam on morbidity and mortality (there is in
increase in delivery of preventative care)
• old trials, limited in scope, assessed out-of-date interventions
• Build rapport and improve adherence
• No evidence to support a routine panel of laboratory tests
in all adult patients
• USPTF:
• DON’T do screening abdominal/testicular exams for cancer
• DON’T do screening pelvic exams (ACP also supports)
Routine Care of Healthy Patient
• Prevention & Screening
• Primary= Prevent
• Secondary= Screen asymptomatic patients with RF’s
• Tertiary= Treat those with disease to reduce
progression/complications
• Effective screening relies on the condition being fairly
common, having a pre-clinical stage allowing detection,
and an effective treatment
• Screening trials must use intention-to- treat analysis
which reduces bias in the volunteer population
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
• Screening Tidbits:
• DON’T perform cancer screening in anyone with <10
years of quality life predicted
• For every 1000 males screened for prostate cancer
approximately 0 to 1 death will be prevented
• For females ages 40-49 one cancer death is prevented
for every 1900 patients invited to screen (RRR=15%)
• For females ages 50-59 one cancer death is prevented
for every 1300 patients invited to screen (RRR=14%)
• When screening controversy exist use a patientcentered shared decision model
Routine Care of Healthy Patient
• Screening Tidbits:
• The USPTF doesn’t recommend for or against CBE
• The USPTF recommends against SBE
• ASCVD RF’s include: DM2, personal history f CAD or
atherosclerosis, FH at <50 males and <60 females, tobacco
use, HTN, and obesity BMI >30
• The USPTF and AUA do NOT recommend for or against
DRE for screening purposes
• PHQ2 (sens= 83%, spec= 90%)- during past 2 weeks how
often have you been bothered with (1) little interest or
pleasure in doing things, or (2) feeling down, depressed, or
hopeless?
Routine Care of Healthy Patient
• Screening Tidbits:
• increased risk of
infection= previous STI,
new/multiple partners, inconsistent condom use, CSW
• Use Alcohol Use Disorders Identification Toll
(AUDIT) for screening problem alcohol use in primary
care setting
• USPTF does not recommend for or against illicit drug
use screening but may use validated Drug Abuse
Screening Test (DAST-10)
Routine Care of Healthy Patient
• Genetics and Genetic Testing
• 3 generation FH is appropriate
• patients are better at recalling the absence of disease,
allow them time to confer with family
• ALWAYS REFER TO GENETIC COUNSELOR
PRIOR TO ANY GENETIC TESTING
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
• Immunization
• Extremely cost effective but vaccination rates remain
unacceptably low
• ACP IA app available
• DO NOT give vaccinations at < recommended
intervals
• DO resume series vaccines at the point of interruption
• Defer vaccination during moderate to severe illness or a
history of anaphylaxis exists
Routine Care of Healthy Patient
• Immunizations
• Avoid influenza vaccine in patients who developed
GBS within 6 weeks of previous administration
• Birth before 1980 confirms immunity against VZV
except for health care professionals, pregnant women,
and immune compromised patients (confirm
serologically)
• If pregnant female not immune wait until after
pregnancy to vaccinate
• Smokers should receive pneumococcal and influenza
vaccines
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
• Healthy Lifestyle Counseling
• 5 leading causes of death: heart disease, cancer, chronic
lower respiratory disease, stroke, and accidents
• Counsel behaviors which reduce these reported causes
of death.. diet, exercise, tobacco cessation, seatbelts,
helmets, stress reduction, firearm handling, alcohol use,
smoke detectors etc.
• Ask, Advise, Assess, Assist, Arrange for tobacco
cessation
• Very brief interventions can be effective but longer
intervention periods are superior
Routine Care of Healthy Patient
• Healthy Lifestyle Counseling
• USPTF recommends offering dietary and exercise behavioral
counseling on individual basis (grade C)
• Products labeled as “dietary supplement” DO NOT undergo
the same level of FDA scrutiny as OTC medications (lack of
standardization)
• USPTF found insufficient evidence to recommend calcium /
vitamin D supplementation for the prevention of fractures
• Age related macular degeneration may benefit from MVI
containing copper, zinc, B-carotene (vitamin A), vitamin C,
and vitamin E (avoid B-carotene and vitamin A in smokers
as high levels increase risk of lung cancer)
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Routine Care of Healthy Patient
Patient Safety and Quality
Improvement
• Quality Improvement (QI) consists of systematic
and continuous actions that lead to measureable
improvement in the quality and safety of patient
care
• Errors
• diagnostic- bias or failed heuristics (reasoning
shortcuts)
• medication- most common (500K to 1.5 million AE
annually)
• transitions of care
Patient Safety and Quality
Improvement
Patient Safety and Quality
Improvement
Patient Safety and Quality
Improvement
• QI Models
Patient Safety and Quality
Improvement
• QI Models
Patient Safety and Quality
Improvement
• QI Models – Root Cause Analysis (RCA)
• involves all stakeholders involved in error
• 5 “why?” technique leads to fish-bone diagram
(Ishikawa diagram)
• Core Measure Sets
• Meaningful Use
• Patient Centered Medical Home (PCMH)comprehensive care by primary provider in teambased medical practice
MKSAP 17 Questions
• 1, 2, 6, 9, 21, 22, 26, 29, 30, 38, 44, 47, 48, 54, 57,
66, 68, 76, 78, 81, 83, 89, 90, 93, 97, 100, 102, 112,
126, 129, 144, 156, 162, 168