Note for critical appraisal - Penn Medicine

Report
Developing a Vaccine against
Pneumonia,
“The Captain of Death” -- Osler
History
• 1881 Louis Pasteur discovers the pneumococcus
bacterium
• 1886 Gold mining begins in South Africa
• 1895 Pneumococcus identified as the cause of
epidemic pneumonia that kills 35% of miners who are
infected
• 1904 Miners strike for health reasons, including
pneumonia
• 1910 Treatment with ethylhydrocuprein, a quinine
derivative, is found to be too neurotoxic
• 1911 Mine owners hire the developer of typhoid fever
vaccine to develop a pneumococcal vaccine
Important Early Vaccine Studies
Date
Vaccine composition
Trial type
1914
Whole cell, heat-killed
organisms
Prospective, placebocontrolled, non randomized
trial with 1 yr follow-up
Number of
participants
50,000 South
African
miners
Result
Not reported,
but vaccine
given to all
miners
Therapies Other Than Vaccine
• 1920s through 1940s antiserum
– Gram stain sputum to confirm cause
– Identify strain (Neufield reaction)
– Give strain-specific horse or rabbit antiserum
intravenously
– Problems included severe allergic reactions and
prolonged delays identifying strain and obtaining
antiserum for treatment
• 1938 Sulfapyridine
– Problems included vomiting and severe rash
Therapies Other Than Vaccine
• 1920s through 1940s antiserum
– Gram stain sputum to confirm cause
– Identify strain (Neufield reaction)
– Treat with strain-specific horse or rabbit antiserum
intravenously
– Problems included severe allergic reactions and
prolonged delays identifying strain and obtaining
antiserum for treatment
• 1938 Sulfapyridine
– Problems included vomiting and severe rash
Next Major Vaccine Study
Date
Vaccine composition
1944 Polysaccharide types 1,2,5,7
made by E.R.Squib & Sons
Trial type
Placebo-controlled,
randomized with 6 mo followup
Number of
participants
17,000
members of
the US Air
Force
Result
90% effective
More History
• 1941 First reported use of penicillin to treat
pneumonia in humans
• 1945 FDA approves penicillin
• 1946 FDA approves Squibb vaccine
• 1951 Squibb stops making its vaccine because
no one is using it
The Penn Connection
Robert Austrian, MD, 1917-2007
http://www.historyofvaccines.org/content/robert-austrian-1
• Austrian suspected that serious risks from pneumococcal infection
persisted despite the prevalence of antibiotics, and he produced the
evidence needed to persuade a skeptical medical community. After a
study of patients in New York City’s Kings County Hospital from 1952 to
1962, Austrian concluded that the incidence of pneumococcal pneumonia
was much higher than was thought at the time and that the mortality rate
of 15% in bacteremic cases was unchanged, despite antimicrobial
treatment.
• In 1962, Austrian left the State University of New York College of Medicine
at Brooklyn to join the University of Pennsylvania. There, he developed a
new vaccine and conducted clinical trials among gold miners in South
Africa that found the vaccine safe and efficacious.
• The recent emergence of widespread resistance of pneumococcus to
commonly used antibiotics highlights the incredible importance of the
vaccine. What he did to solve a major human disease problem, often
almost totally by himself, is extremely rare in modern medicine.
– Paraphrased from the obituary by Harvey Freidman in JCI
Current Vaccines
• Pneumococcal conjugate vaccine
– Polysaccharides conjugated with diphtheria
proteins to enhance immunogenicity
– Active against 13 strains
– Abbreviated PCV13
• Pneumococcal polysaccharide vaccine
– Not conjugated
– Active against 23 strains
– Abbreviated PPSV23
Current Vaccine Recommendations for
Infants, Toddlers, and Children
Pneumococcal conjugate vaccine (PCV13) is
recommended for all infants, toddlers, and
children from 2 months through 5 years of
age. It also is recommended for children 6
through 18 years of age with certain medical
conditions regardless of whether they have
previously received a pneumococcal vaccine.
Current Vaccine Recommendations for
Adults
Pneumococcal polysaccharide vaccine
(PPSV23) is recommended for adults 65 years
and older; for younger adults who have a
chronic illness or who smoke; for Alaska
Natives and certain American Indian
populations; for people who had their spleen
removed; and for those with weakened
immune systems.
Important Observations
• In elderly adults, PPSV23 protects against
invasive pneumococcal disease (IPD) but
perhaps not against nonbacteremic
pneumococcal disease (NPD)
• In children, PCV13 protects against IPD and
NPD
• The effect on adult disease from childhood
vaccination is unknown
Cost-effectiveness of Adult
Vaccination
Strategies Using Pneumococcal
Conjugate
Vaccine Compared With
Pneumococcal
Polysaccharide Vaccine
JAMA. 2012;307(8):804-812
What Do We Know about the Journal
and the Authors?
Vaccination Strategies Modeled
•
•
•
•
•
•
No vaccination
PPSV23 at age 65 (current recommendation)
PCV13 at age 65
PCV13 at age 50, and PPSV23 at age 65
PCV13 at age 50 and again at age 65
PCV13 at age 50 and again at age 65 plus
PPSV23 at age 75
Model Features
•
•
•
•
Life-time horizon
Societal perspective
3% discount rate for costs and benefits
Quality of life measured on 0 to 1 scale, and
QALYs calculated by multiplying the utility of a
state times the duration in that state
Model Inputs from Table 1
•
•
•
•
•
Vaccine effectiveness
Vaccine adverse events
Disability – risk and mortality
Utility weights by age and level of risk
Hospitalization rates and costs by type of
disease (IPD vs. NPD)
• Vaccine costs
Results in Table 3
Henry’s Efficient Algorithm
Results in Table 3
Deterministic Sensitivity Analyses
One-Way Only
More Results in Table 3
How to Describe the Results of
Probabilistic Sensitivity Analyses:
Could these results have occurred by
chance alone?
• Plot results of separate Monte Carlo runs as a
cloud in the cost-effectiveness plane
• Calculate p-values or confidence intervals
• Create cost-effectiveness acceptability curve
Step 4. Analyze the “Stochastic” Tree
Sampling
Cost-Effectiveness Plane
Incremental CE Plot Report
QUADRANT
INCR
EFF
INCR
COST
C1
IV
IE>0
IC<0 Superior
C2
I
IE>0
C3
III
C4
FREQ
PROPORTION
0
0
IC>0 ICER<50k
106
0.0212
IE<0
IC<0 ICER>50k
0
0
I
IE>0
IC>0 ICER>50k
1 2.00E-04
C5
III
IE<0
IC<0 ICER<50k
0
0
C6
II
IE<0
IC>0 Inferior
4893
0.9786
origin
IE=0
IC=0 0/0
0
0
Indiff
Do You Agree with the Authors’
Conclusion?
Overall, PCV13 vaccination was favored
compared with PPSV23, but the analysis was
sensitive to assumptions about PCV13
effectiveness against nonbacteremic
pneumococcal pneumonia and the magnitude
of potential indirect effects from childhood
PCV13 on pneumococcal serotype
distribution.
My Translation of the Authors’
Conclusion
In the base case analysis and in most of the
deterministic sensitivity analyses, PCV13
vaccination was favored compared with PPSV23,
but the analysis was sensitive to assumptions
about PCV13 effectiveness against nonbacteremic
pneumococcal pneumonia and the magnitude of
potential indirect effects from childhood PCV13
on pneumococcal serotype distribution.
Two Problems with This Conclusion
In the base case analysis and in most of the deterministic sensitivity analyses,
PCV13 vaccination was favored compared with PPSV23, but the analysis was
sensitive to assumptions about PCV13 effectiveness against nonbacteremic
pneumococcal pneumonia and the magnitude of potential indirect effects
from childhood PCV13 on pneumococcal serotype distribution.
1. This conclusion does not help readers
decide whether there is a preferred vaccine
strategy, and if so, which one it is.
2. This conclusion ignores the probabilistic
sensitivity analysis.
What about this Alternative
Conclusion?
In the base case analysis and in most of the
deterministic sensitivity analyses, the strategy of
giving PCV13 at age 50 and again at age 65 was
favored over other strategies, but our
probabilistic sensitivity analysis showed that the
differences between this strategy and other
strategies could have occurred by chance alone.

similar documents