Behavioral Science Flashcards
study population in cohort study
Compares a group with a given exposure or risk factor to a group without such exposure.
Ruling in vs. out
Negative test in a highly sensitive test rules out disease. Highly specific test rules in disease. (SNOUT/SPINN)
Other expression for sensitivity
1- false-negative rate
Effect of lowering test cutoff
Increased sensitivity, decreased specificity, increased NPV, decreased PPV (just think about it as increasing the # of FP and decreasing the nuber of FNs)
Effect of raising test cutoff
Increased specificity, decreased sensitivity, increased PPV, decreased NPV (similarly just think of it as increasing the # of FNs and decreasing the # of FPs).
Attributable risk
Difference in risk between exposed and unexposed groups, or the proportion of disease occurrences that are attributable to the exposure (eg, if risk of lung cancer in smokers is 21% and risk in nonsmokers is 1%, then 20% of lung cancer risk in smokers is attributable to smoking).
If 21% of smokers develop lung cancer vs. 1% of nonsmokers, what is the relative risk?
21
absolute risk reduction
Difference in risk (NOT proportion) attributable to interention as compared to a control (eg, if 8% of people who receive a placebo vaccine develop the flu vs. 2% of people who receive a flu vaccine, then ARR = 8% - 2% = 6% = .06
Number needed to harm
Number of patients who need to be exposed to a risk factor for 1 patient to be harmed.
Relation between precision and statistical power
Increased precision, increased statistical power
Statistical power definition
1-beta
random error vs. systematic error
random error decreases precision in a test, systematic error decreases accuracy in a test.
Berkson bias
Study population selected from hospital is less healthy than general population.
Healthy worker effect
Study population is healthier than the general population
Non-response bias
Participating subjects differ from nonrespondents in meaningful ways.
Procedure bias
Subjects in different groups are not treated the same.
Observer-expectancy bias
Researcher’s belief in the efficacy of a treatment changes the outcome of that treatment.
pygmalion effect
observer-expectany bias. AKA self-fulfilling prophecy
Crossover study
Subjects act as their own controls.
Lead-time bias
early detection confused with increased survival
How do you mitigate lead-time bias?
Measure “back-end” survival (adjust survival according to the severity of disease at the time of diagnosis).
normal distribution
mean = median = mode
Variance definition
(SD)^2
Standard error of the mean (SEM)
Estimate of how much variability exists between the sample mean and the true population mean.
Definition of SEM
SD/sqr(n)
Relation between SEM and n
SEM decreases and n increases
positive skew
mean>median>mode (asymmetry with longer tail on right)
negative skew
mode > median > mean
alternative hypothesis
Hypothesis of some difference or relationship.
beta
Probability of making a type II error.
alpha
Probability of making a type I error.
You sAw an error that didn’t exist
False-positive error
Type I error (alpha)
Type I error
Stating that there is an effect or difference when none exists (null hypothesis incorrectly rejected in favor of alternative hypothesis).
What does a p value of 0.05 indicate?
There is less than a 5% chance that the data will show something that is not really there.
Type II error (beta)
False-negative error.
Type II error (beta)
stating that there is not an effect or difference when one exists (Null hypothesis is not rejected when it is in fact false). (beta, you were blind to the truth).
Statistical power
(1-beta). Probability of rejecting the null hypothesis when it is false.
How would you increase power and decrease beta?
1) increase sample size.
2) increase expected effect size.
3) increase precision of measurement.
Confidence interval
Range of values within which the true mean of the population is expected to fall, with a specified probability.
CI equation
CI = mean +/- Z(SEM)
Important points about CI
1) If the CI includes 0, there is no significant difference.
2) If the CI for odds ratio or relative risk includes 1, there is no significant difference.
3) If the CIs between 2 groups overlap, usually no significant difference exists. Conversely, if CIs between 2 groups do not overlap, statistically significant difference exists.
Pearson correlation coefficient (r)
r is always between -1 and +1. The closer the absolute value is to 1, the stronger the linear correlation between 2 variables.
Coefficient of determination
r^2
exceptions to informed consent
1) patient lacks decision-making capacity or is legally incompetent.
2) Implied in an emergency
3) therapeutic privilege
4) waiver (patient waives right)
Conditions for which a minor is legally emancipated
1) Married
2) self supporting
3) in the military
Situations in which parental consent is usually not required.
Sex (contraception, STIs, pregnancy), drugs (substance abuse), and rock and roll (emergency/trauma)
Oral advance directive
Incapacitated patient’s prior oral statements used as a guide for decision-making.
priority of surrogates
spouse>adult children>parents>adult siblings>other relatives.
answer to confidentiality questions
make an exception if its in the patient’s best interests.
Reportable diseases in which patient confidentiality is broken.
STIs, TB, hepatitis, food poisoning
Tarasoff decision
California Supreme Court decision requiring physician to directly inform and protect potential victim from harm.
nonintuitive exceptions to patient confidentiality
1) child and/or elder abuse
2) impaired automobile drivers
3) suicidal patients
patient has difficulty taking medications
1) Provide written instructions
2) Attempt to simplify treatment regimens
3) Use teach-back method (ask patient to repeat regimen back to physician) to ensure comprehension.
Patient’s family member asks you not to disclose results of a test if prognosis is poor.
1) ask for more info
2) explain that as long as patient has decision making capacity and doesn’t indicate otherwise, information will not be withheld.
Minor wants an abortion…
1) Many states require parental notification or consent.
2) Unless there are specific medical risks, physician shouldn’t affect patients decision.
parents and child are at odds about what to do with a pregnancy…
1) patient makes decision
2) Provide info to teenager about practical issues of caring for a baby
3) Discuss options, IF requested.
4) encourage discussion between teenager and her parents.
physician assisted suicide
In large majority of states, answer is refuse involvement. You can give analgesics though that coincidentally shorten patient’s life.
patient is suicidal…
1) assess seriousness
2) If serious, suggest patient remain in hospital voluntarily
3) if resistant, pt can be hospitalized involuntarily
Patient states he/she finds you attractive…
1) Ask direct, closed-ended questions
2) Use a chaperone
3) relationship never okay.
Pt angry about delay
Don’t explain it, just apologize and acknowledge patient’s anger.
Pt angry with another physician
Suggest patient talk to that physician.
Pt angry with staff member
Tell pt. you will speak to that person.
Patient requires a treatment not covered by insurance
1) discuss all options even if some aren’t covered.
2) never limit or deny care.
When do kids start to realize death is permanent?
ages 5-7
presbycusis
Sensorineural hearing loss (often of higher frequencies) due to destruction of hair cells at the cochlear base (preserved low-frequency hearing at apex).
Libido changes in men and women.
Stable in men, decreases in women after menopause.
Sexual changes in men and women
Men–slower erection/ejaculation, longer refractory period.
Women–vaginal shortening, thinning, and dryness.
Changes in sleep patterns in elderly
Decreased REM and slow-wave sleep + increased sleep onset latency + increased early awakenings.
Immunology in elderly
immune response declines
Secondary prevention
Screen and manage existing but asymptomatic disease.
Quaternary prevention
Identify patients at risk of unnecessary treatment, protecting from the harm of new interventions.
People who qualify for medicare
1) >65 years old
2)
Medicaid
Joint federal and state health assistance.
Medicare part A
Hospital insurance
Medicare part B
Basic medical bills (eg, doctor’s fee, diagnostic testing)
Medicare part C
Parts A+B delivered by approved private companies.
Medicare part D
Prescription drugs
Top 3 causes of death in less than 1 year olds
1) congenital malformations
2) preterm birth
3) SIDS
Top 3 causes of death in 1-14 year olds
1) Unintentional injury
2) Cancer
3) Congenital malformations
Top 3 causes of death in 15-34 age group
1) Unintentional injury
2) Suicide
3) Homicide
Top 3 causes of death in 35-44 age group
1) Unintentional injury
2) Cancer
3) Heart disease
Top 3 causes of death in 45-64 age group
1) Cancer
2) Heart disease
3) Unintentional injury
Top 3 causes of death in 65+ age group
1) Heart disease
2) Cancer
3) Chronic respiratory disease
Hospitalized conditions with frequent readmissions –> Medicare
1) CHF
2) Septicemia
3) Pneumonia
Hospitalized conditions with frequent readmissions –> Medicaid
1) Mood disorders
2) Schizophrenia/psychotic disorders
3) DM with complications
Hospitalized conditions with frequent readmissions –> Private insurance
1) Maintenance of chemo or radiotherapy
2) Mood disorders
3) Complications of surgical procedures or medical care
Hospitalized conditions with frequent readmissions –> Uninsured
1) Mood disorders
2) Alcohol-related disorders
3) DM w/ complications
Forcing functions
Functions that prevent undesirable actions. Most effective approach in human factors design.
Human factors design
Standardizing processes, simplifing shit to reduce waste, etc.
PDSA model
Plan (define problem and solution), Do (test new process), Study (measure and analyze data), Act (integrate new process into regular workflow)
“Process” as a quality measurements
Measures performance of system as planned.
Run charts
plot quality measurements. graphs of data collected over time that can help determine whether an intervention or enhancement in patient care process has resulted in true improvement over time. /time on x-axis, quality measure on y-axis.
Shewhart (control chart)
Applies formal statistical calculations to determine whether observed rise and fall of a quality measure over time is within a predictable range of variation or is an indication of a significant change in the system.
How do you fix problems identified in a root cause analysis?
A corrective action plan
Fishbone (Ishikawa)
Used to identify possible causes for an effect or problem.
Failure mode and effects analysis + important caveat
1) Uses inductive reasoning to identify all the ways a process might fail and prioritize these by their probability of occurrence and impact on patients.
2) Forward-looking approach applied BEFORE process implementation to prevent failure occurrence.
managing child abuse
First protect, then report. Separate parents from child, then report.
Anchoring heuristic
premature closure. Relying on initial diagnostic impression despite subsequent information to the contrary. Eg. Repeat positive blood cultures with a rare bug are dismissed as contaminants.
Availability heuristic
diagnosis biased by experience by past cases
Blind obedience
placing undue reliance on test results or expert opinion. Eg. False-negative
Framing effects
Diagnostic decision-making unduly biased by subtle cues and collateral information. Eg. Opioid user with abdominal pain treated for opiate withdrawal but later diagnosed with bowel perforation.
DRGs
Diagnostic-related groups: payment categories used by the government to classify patients (especially Medicare patients) for the purpose of reimbursing hospitals for each case in a give category. DRGs essentially determine how much medicare pays the hospital for services. They limit what the government will pay for a given diagnosis.
Incident reports
1) legal protection
2) voluntary
3) passive form of surveillance
how do you involve patients in care?
1) encourage patients to detect and report adverse events. 2) stress patient involvement
HMOs
1) can’t see out of network physicians.
2) lower cost.
3) need referral from PCP.
4) file reimbursement claim patients see in-network providers only and generally pay them directly.
Preferred provider organizations (PPOs)
1) NO requirement for a PCP. 2) no need for a referral.
3) higher cost.
4) can visit providers outside of network.
5) patients may pay a doctor for services directly and then file a claim to get reimbursed.
Point-of-service (POS)
1) combination of HMO and PPO.
2) ”point of service” reflects the fact that patients are able to make choices of whether to use HMO or PPO services each time they see a provider. 3) policies similar to HMO insurance but a POS will also permit patients to see an out-of-network physicians for a higher fee and/or with a referral from a PCP.
4) cost = lower than a PPO but likely higher than an HMO.
5) need a PCP.
6) need a referral to see a specialist.
Negligent nondisclosure
requirement that physician acts in the same manner as another competent, reasonable clinician in similar circumstances would act when disclosing information + breach of duty + injury + causal relationship established.
breach of duty
failure to meet minimal standard of care
Pareto chart
used to describe a large proportion of quality problems being caused by a small number of causes. /based on classic 80/20 rule: 80% of world’s wealth is distributed in hands of an elite 20%. /majority of patient safety errors stem from only a few recurring contributing factors.
husband is talking for wife patient and she looks helpless…
ask the husband to wait outside while you obtain a complte history and examine the patient in private. If husband objects, offer to have a female staff member present while he is out of the room. If wife objects, ask what would make her most comfortable
signout
act of transmitting information about the patient.
handoff
process of transferring responsibility for patient care.
STEEEP
Safe, Timely, Effective, Efficient (avoid waste) Equitable, Patient-centered. /Principles intended to serve as a blueprint for redesign of a given health care system.
slip
action not carried out as intended or planned
lapse
missed actions and omissions (eg forgetting to order potassium in a pt receiving furosemide).
mistake
intended action is wrong.
errors
accidental mistakes
violations
deliberate actions whereby someone does something and knows it to be against the rules, such as deliberately failing to follow proper procedures (eg foregoing entering a patient’s PMH into an EHR due to time constraints).
5 R’s of patient safety
o Right drug, right patient, right dose, right route, right time.
o 3 other R’s Right to know information about the drug, right to refuse the drug, right documentation.
1 SD
68%
2 SD
95%
3 SD
99.7%
sleep onset latency
length of time it takes to accomplish transition from full wakefulness to sleep.
What does a case-control study compare and what is the point?
A group of people with a disease to a group of people without the disease to look for a prior exposure or risk factor. Asks, what happened?
What does a cohort study compare and what is the point?
A group of people with a given exposure or risk factor a group without such exposure to see if the exposure increased the likelihood of the disease.
Caveat about cohort studies…
Can be prospective or retrospect.
If you were doing a cohort study of kids with neurologic disability who’s parents worked in a factory. Who’d you want to serve as your control?
Kids who’s parents
Control group selection in case-control study
Control group = no disease.
**independent of the exposures of interest.
Control group selection in cohort study
Control group = no exposure to disease.
Z values for 95% CI and 99% CI
Z = 1.96 for 95% CI Z = 2.58 for 99% CI
