Transcript How to Search PubMed

How to Search PubMed
Rick Wallace, MA, MDiv, MAOM,
MSLS, EdD
QUILLEN College of Medicine Library
Library Home Page
You are a second year resident picking up a ward service at the VA. On rounds your first morning,
your medical student presents the following patient to you:
The patient is a 73-year-old white man who was admitted with an exacerbation of CHF. He has a
history of past MI and an EF by Echo of 25%. He is very compliant and seldom, if ever, misses
his medications. On admission, his medications were furosemide 40 mg po qd, enalapril 20
mg po bid, simvastatin 10 mg po qd and aspirin 325 mg po qd. He usually functions at a NYHA
class II level, but he now has had two admissions for CHF in the past 4 months. On this
admission, he has already been ruled out for MI and remains in sinus rhythm. He has a
normal renal function with a creatinine of 0.7. The patient is particularly concerned about
staying out of the hospital because he lives alone and has no one to care for his pets or
plants. He is very active in his community and is eager to get out and stay out of the hospital.
The medical student asks whether Digoxin would help this man with his heart failure. The student
was told by one of his prior residents that Digoxin was a good choice for a patient with CHF.
Your intern states that she feels that Digoxin does not improve mortality in patients with
heart failure and that we must consider potential toxicities, especially in a 73-year-old man.
They look to you for guidance.
PICO
•
•
•
•
P- elderly, male, Caucasian pts with CHF
I- digoxin
C- no digoxin
O- reduce hospitalization, mortality
PICO to MESH
•
•
•
•
P- Heart Failure; Aged (male, Caucasian?)
I-Digoxin
C- none
O-
How large was the treatment effect?
Mortality was the primary outcome measure, however the hospitalization outcomes were also important. Results are summarized in the table below:
Outcome
Mortality
Total hospital
CHF hospital
CV hospital
Digoxin
34.8%
64.3%
27%
49.9%
Placebo
35.1%
67.1%
35%
54.4%
RRR
ARR
Nonsignificicant (p=0.8)
4.1%
2.8%
23%
8%
8.3%
4.5%
NNT
36
13
22
Note: Of course, suspected digoxin toxicity was a more frequent cause of hospitalization in the treated group.
How precise was the treatment effect?
95% Confidence Intervals are reported in the tables surrounding the risk ratios. Because the sample size is large for this trial, the CI’s are fairly
tight in general.
Table 2 (page 528) shows CI’s for the Risk Ratios. The major outcomes (all cause, cardiovascular and heart failure) all include 1 (i.e. no difference)
For hospitalization outcomes in Table 3, however, (page 529) several important outcomes have CI’s that do not include 1 and favor the treated group.
How can I apply the results to my patient care?
Were the study patients similar to my patient?
In some ways, yes: 86% of the patients in the trial were white, 84% were either NYHA class II or III, and 70% of patient had ischemic
cardiomyopathy. However, only 27% were older than 70 years. All in all, however, it is likely that our patient is close enough in characteristics
to the study population that the results may be applied to him. In addition, there are some ways in which the study results are particularly
helpful to our patient. Notably, in our patient the outcome of preventing hospitalizations may be of great interest to him. She explicitly reports
that staying out of the hospital is an important goal for her (and her pets!)
Were all patient- important outcomes considered?
Yes. Most (if not all) clinically relevant outcomes were considered.
Are the likely benefits worth the potential harms and costs?
This may be a matter of weighing patient’s values. For a patient who elects to be treated with digoxin, there will be the need to monitor
therapy, draw frequent drug levels, and hold the risk of toxicity. However, for some, these issues will be offset by the possible benefit
of avoidance of hospitalization. From a resource utilization point of view, the cost of the drug (cheap) and monitoring (not quite so cheap) will
still be less than a single hospitalization. It is fairly convincing from this large, well-designed trial that there is not a large difference in mortality
overall when using digoxin. Thus the quality of life issues (including hospitalization) should predominate the discussion.
Can edit from here
http://www.ncbi.nlm.nih.gov/pubmedhealth/
• PubMed Health is a US National of Library of Medicine
(NLM) resource to make systematic reviews of
clinical effectiveness easier to find and use.
• A comprehensive collection of systematic reviews and
their knowledge translation versions in one place is the
basic premise of PubMed Health. The website
integrates DARE (Database of Reviews of Effects),
Cochrane Reviews and systematic reviews published
directly by health technology assessment agencies.
• PubMed Health has almost 20,000 systematic reviews
from the last 10 years now in the collection.
Finally
• Library (Building 4) open to midnight every day
(July we close at 10 pm). Basement opened 24/7
• We will load clinical software on your
Smartphone/PDA- DynaMed, ePocrates, AHRQ
clinical screening tool, StatRef!, etc
• Search help• Nakia Woodward- [email protected]
• Rick Wallace- [email protected]
• Katie Wolf