cbse equations Flashcards
filtration fraction
FF = GFR / RBF
GFR is about 120
RPF is about 600 (renal plasma flow)
in healthy FF is about 20% of RPF
clearance
Cs + [urine concentration of S] x [urine flow rate] / [plasma concentration of S]
GFR can be obtained by ?
RPF?
clearance of creatine
clearance of PAH
absolute risk increase?
number needed to harm?
adverse event rate in control and experimental then subtract adverse in experimental from control
NNH = 1/ the absolute risk increase
sensitivity
people with the disease that test positive
want this number high to know picking up on diagnosis
higher = better at ruling out disease
SCREENING diseases
A / A+C
true positive / true positive + flase negative
rule OUT disease
specificity
true negative
without the disease who test negative
closer to 100 p = better at ruling in
low false positive
confirm after positive screen
SPIN
D/D+B
true negative/ true negative + false positive
type 1 error
shows a relationship tht does not really exist
type II error
study fails to show a relationship that does exist
beta error
Vd=
amount of drug given (IV)/ [drug] plasma
low Vd = intravascular space and large/ charged molecules + bound
medium = intrvascualr and extracellular
large = able to get to all tissues and including fat
- usually small and lipophilic molecules
clearance of a drug
0.7 X Vd/ t 1/2
1/2 life =
.7 X Vd / clearance
hardy weinberg equations
p+q=1
P^2 +2pq+q^2=1
p^2= frequency of homozygous for p
q^2 = freq for homozygous for q
2pq= frequency of heterygosity
positive skew graph in terms of mean median and mode
negative?
shifted to the left
tail to the right
mean > median
median > mode
mode > median
median >mean
case control study
w/ disease and w/out groups
then look back at it
look at some exposure risk
retrospetive and observational
purely observational
- no intervention
*used to identify risk factors for diseases
NOT causal -
yields an odds ratio!!
odds ratio
study design and equation?
AxD divided by BxC
cohort study
group of people that has something in common
compare them to group that have not had that exposure and then follow them
*purely observational
either retrospective or prospective
formula associated with cohort study
relative risk
or risk ratio
relative risk equation
A/A+B divided by C/C+D
cross sectional study
looks at a population at a point in time
(ex dx of COPD at a time)
or ask about a risk factor
- exposed to 2nd hand smoke on that day
known as PREVALENCE
observational study
what does cross sectional show
prevalence
controlled clinical trial
investigator intervenes
so not purely observational - there is intervention
controlled = placebo and then an experimental group
randomized
limit bias
double blinded
particpant or investigator doesnt know who gets what
meta analysis
combines data from many studies together
increase statistical power
quality depends on quality of individual studies
true positive location
upper left (A)
false positive location
upper right (B)
false negative location
lower left (C)
true positive location
lower right (D)
sensitivity equation
+ acronym
A/ A+C
true positives divided by everyone who has the disease (true positives plus the false negatives)
1- the false negative
PID
- positive in disease
specificity
+ acronym
proportion without disease with a negative test results
true negative / true negative plus false positives
D/ (D+B)
1- false positive
NIH acronym
negative in health
PPV
positive predictive value
A/ (A+B)
true positives / everyone that tested positive
NPV
negative predictive value
D/ (C+D)
PPV change with increasing prevalence
PPV increases with increases in prevalence of disease
so increase true positives and false negatives
NPV change with increasing prevalence
increasing flase negatives
so decreasing the NPV
low disease prevalence changes to PPV and NPV?
PPV decreases and NPV increases
prevalence
of people with the disease / total population
certain amount with disease at a certain time
incidence
# of new cases diagnosed / total # of people at risk for that illness
people with the disease not used in this calculation
low prevalence effect on relative risk and odds ration
in low prevalence situations
RR will EQUAL OR
attributable risk equation
rlative risk equation but instead of dividing you subtract
A/(A+B) - C/(C+D)
absolute risk reduction
looks at how much an intervetnion will reduce risk of disease
opposite attributable risk equation
so C/(C+D) - A/(A+B)
number needed to treat
1/ absolute risk reduction
needed to treat to save a life or avoid a bad outcome
Number needed to harm
1 divided by the attributable risk reduction
which is A/(A+B) - C/(C+D)
so 1 divided by that
standard distribution curve / graph
% that fall within SD?
68% fall within 1 SD (34% +/-) 95% fall within 2 SD (13.5% +/-) 99.7% fall within 3 SD (2.35% +/-)
then .15%
small p value?
more likely to be
less than 0.05 - can reject null and shows an association
standard error of the mean?
standard deviation divided by / square root sample size (n)
confidence interval
range from
[mean-Z(SEM)] to [mean+Z(SEM)]
Z is specific to confidence interval
Z in confidence interval if
90% CI?
95% CI?
99% CI?
90% CI = 1.645
95% CI =1.96
99% CI = 2.57
if CI crosses 0?
accept null hypothesis
chi ^2 vs t test
t test looks at the means
chi-square looks at percentaes or proportions of categorical outcomes in 2 or more groups
correlation coefficient that is perfect
1= perfect 0= none
greater than 0 = positive correlation
less than 0 = negative
