Term
| Measure of central tendency for NOMINAL data |
|
Definition
Nominal: number assigned, no numerical "value," qualitative value, categorical
Use MODE: most frequently occurring value |
|
|
Term
| Measure of central tendency for ORDINAL data |
|
Definition
Ordinal: number assigned, has value but no steady "difference" between values, rank ordering
Use MEDIAN: when arranged in order, middle value |
|
|
Term
| MoCT MOST susceptible to outliers/skew |
|
Definition
| Mean, use median instead (not as susceptible) if data has outliers |
|
|
Term
| Negative Skew - rank MoCT (lowest to highest) |
|
Definition
(Distribution with tail to the left)
Mean, Median, Mode |
|
|
Term
| Positive Skew - rank MoCT |
|
Definition
(Tail to the right)
Mode, Median, Mean |
|
|
Term
| Standard Deviation & Variance |
|
Definition
Variance: average of square of all distances of values from mean --- doubt we will have to calculate this on exam (?)
S.D.: Sq. root of Variance |
|
|
Term
Percentage within +/- 1 SD of mean
Percentage within +/- 2 SD of mean |
|
Definition
|
|
Term
| Iceberg Diagram for Disease Stages |
|
Definition
Clinical Disease (top of iceberg)
Infection, no Clinical Disease
Exposure, no infection (majority of children teeth) |
|
|
Term
Likert Scale
Bias susceptibility? |
|
Definition
Ordinal scale, 1-5
Susceptible to: CENTRAL TENDENCY AND SOCIAL DESIRABILITY BIASES |
|
|
Term
| Notice of Privacy Practice |
|
Definition
Part of HIPAA, once in a lifetime requirement - POST EVERYWHERE (clinic website, office, in spanish)
Acknowledgment of receipt of NPP allows you to use PHI for TPO |
|
|
Term
| When is patient authorization to release info required? |
|
Definition
| In most cases for reasons other than Treatment, Payment, or Healthcare Operations (TPO) |
|
|
Term
| When are you allowed to repeat to others PHI that you encounter as part of your job? |
|
Definition
| When your job REQUIRES it |
|
|
Term
| Business Associate Agreement required when providing PHI to... |
|
Definition
Company that maintains your office computer system (ex)
Business Associate: any entity with whom you share protected health info: member of staff (not another provider), anyone who uses, discloses PHI, or is connected to providing a service (billing, lab, consulting, management, IT, call service, etc) |
|
|
Term
|
Definition
| WHAT is protected, WHO is permitted to access |
|
|
Term
|
Definition
| HOW private info is protected - administrative requirements |
|
|
Term
| Permitted Uses/Disclosures of PHI |
|
Definition
Individual Access
TPO (treatment, payment, healthcare operations)
Education |
|
|
Term
| Special circumstances dentist can disclose PHI without authorization: |
|
Definition
Public Health Surveillance
Child abuse, neglect, domestic violence investigation
Oversight by Secretary of HHS
Law enforcement with valid warrant, court order or administrative request |
|
|
Term
|
Definition
Don't give out entire medical record (unless to providers for tx), psychotherapy notes...
This doesn't apply to uses and disclosures to patients, HIPAA compliance purposes that are required by law |
|
|
Term
| HIPAA Waiver Criteria for Research (3) |
|
Definition
1. Use or disclosure involves minimal risk to privacy of subject: plans to protect, destroy, or not disclose further
2. Research can't be done without waiver
3. Research can't be done without access and use of PHI |
|
|
Term
|
Definition
Measure of rightness
Close to true value (gold standard, consensus, truth)
Relationship btwn what a test measures and what it is supposed to measure
NOT a property of the measure, but assesses how it measures a particular construct (might be valid for one thing, but not another)
Cannot be valid unless reliable |
|
|
Term
|
Definition
Measure of exactness
Inverse fan of measurement error
Provide consistent measurement from occasion to occasion (how close repeated measures are to each other)
Precision: SD is a measure of this
Sources of error: biological, analytic (measurement technique) |
|
|
Term
|
Definition
How it "appears" to measure what it is supposed to
Not statistical, just what people think
Just bc it has face validity, doesn't make it valid.. but can be valid without face validity |
|
|
Term
|
Definition
Relationship of measure of a construct to scores on measures of other constructs (related)
Does it relate to other measures as it should?
ex) scores of self esteem should positively correlate with measures of confidence/optimism and negatively with insecurity/anxiety |
|
|
Term
| Criterion-Related Validity |
|
Definition
How measure distinguishes participants based on criterion
ex) SAT distinguishes students who do well in college
Concurrent V: High SAT = better prepared for college
Predictive V: High SAT = will do better in college
Marital conflict should correlate with behavioral observation (fights)
Measure of depressive symptoms should distinguish those in treatment from those who are not |
|
|
Term
Reliability Estimates:
Total Variance |
|
Definition
TV = True variance + Error variance
Reliability = True Variance / TV
Can range from 0-1.0
>70% have acceptable reliability |
|
|
Term
|
Definition
| Consistency of participant's response over time |
|
|
Term
|
Definition
|
|
Term
| Internal Consistency Reliability |
|
Definition
Inter-item reliability - consistency of items in a scale/observations
Want to make sure all items are measuring the same construct |
|
|
Term
|
Definition
#cases/#in population at a specific time (date, age, life-time event, period)
Point (one time) & period prevalence (cases at start and those developed during the period)
Measures BURDEN of disease
P = Incidence*Duration of Disease |
|
|
Term
|
Definition
#new cases/#persons at risk during a specific time period
Measures RISK of disease
Incidence rate = #Dz/ (#at risk)*(#yrs) |
|
|
Term
| Gingival Inflammation Indices |
|
Definition
0 - Absence of visual signs of inflammation
1 - Slight change in color, no bleeding
2 - Visual inflammation, bleeding from GM on probing
3 - Overt inflammation, spontaneous bleeding |
|
|
Term
|
Definition
3rd year 80s rock band... jk
Occurs when epithelial lining of inflamed tissue is thin/micro-ulcerations
From repeated probe insertions at single site, excessive force (>.25N) |
|
|
Term
|
Definition
Measures from free gingival margin to tip of probe (1-3mm is normal)
Healthy - probe tip will stop coronal to the apical termination of junctional epithelium
Walk probe around to find deepest point at 6 tooth surfaces (DF, F, MF, DL, L, ML) |
|
|
Term
| Clinical Attachment Level (CAL) |
|
Definition
| Distance from CEJ to tip of probe |
|
|
Term
|
Definition
| Assessed with curved Nabers probe |
|
|
Term
|
Definition
Grade 1: <1 mm in BL or MD direction
Grade 2: >/=1mm in BL or MD direction
Grade 3: >/=1mm in BL, MD, or vertical direction! |
|
|
Term
| Reducing measurement errors |
|
Definition
Double-blind
Standardize measures/devices
Practice measure techniques
Have a "gold standard" |
|
|
Term
|
Definition
D1: Clinically detectable enamel lesion
D2: Cavity limited to enamel
D3: Dentin lesion (open/closed)
D4: Pulp lesion
Caries free until the D3 threshold, Preventive care |
|
|
Term
| Most prevalent lesion in children and young adults |
|
Definition
| noncavitated occlusal carious lesions |
|
|
Term
|
Definition
Adjusts for degree of agreement expected purely by chance
Below 0.4, poor agreement
0.4-0.75, fair agreement
Above 0.75, excellent! |
|
|
Term
| Signs & Symptoms of Oral Cancer |
|
Definition
Non-healing ulceration
Red Plaque, white plaque that doesn't rub off
Tumor
Mixed picture
Dysphagia (can't swallow)
Pain
Hemorrhage
Swelling or overgrowth
Red/white patches |
|
|
Term
|
Definition
| Time from screening detection to when they would have otherwise been diagnosed... Lead time bias: detecting earlier makes patients SEEM to live longer, but really just detecting earlier |
|
|
Term
|
Definition
= #new cases/#at risk at the start over a time period
The average risk/probability of developing disease during a period of time |
|
|
Term
|
Definition
= #New cases/#Non-cases
How many times more likely is it to develop a disease over a specified time period? |
|
|
Term
|
Definition
Risk = probability of developing disease (3 in 10 = 0.3)
Odds = how much more likely (3 with, 7 without = 0.43) |
|
|
Term
Incidence RATE
(incidence density, force of mortality/morbidity) |
|
Definition
| = #New cases in period/ Total person-time at risk during period |
|
|
Term
|
Definition
Subjects: random/convenience (NOT chosen based on dz or exposure status)
Estimates prevalence of disease at one point
Advantage: short, cheap, can look at many exposures
Disadvantage: Can't get incidence, temporality cannot be established
Measure: Odds Ratio |
|
|
Term
|
Definition
Group with dz compared to controls
Sample is SELECTED, so cases and controls are similar on all other levels except dz status
A: short, cheap, study many exposures/RFs
DA: temporality/incidence rate cannot be assessed
Bias: Recall
Measure: Odds Ratio |
|
|
Term
|
Definition
CREME DE LA CREME/Gold Standard of Observational Studies
Prospective & Retrospective
Sample SELECTED based on exposure status, followed fwd in time
Good for rare exposures or when diff of exp time and dz development is short
A: incidence rates, temporality, multiple outcomes can be looked at
DA: expensive, large sample needed, long-term follow up
Biases: Selection bias (determining exp status not standardized), Loss to follow up, Misclassification of outcome/investigator bias, confounders
Measure: Relative Risk |
|
|
Term
|
Definition
Study & control arm - field trial or quasi-eperimental study
Subjects: eligibility & exclusion
Randomly assigned to groups, not haphazard: decreases bias, each has equal chance of being in each group
Best way - double blind (least bias, but ethical/practical considerations)
Bias: examiner, loss to follow up, selection
Measure: Relative Risk |
|
|
Term
| Strength of Association (Relative Risk) |
|
Definition
RR = (incidence among exposed)/(incidence among unexposed) (a/a+b)/(c/c+d)
Exposed are __x more likely to develop ___
RR=1, no association
RR<1, protective factor
RR>1, risk factor |
|
|
Term
| Strength of Association (Odds Ratio) |
|
Definition
OR = ratio of odds of exposure in diseased to the odds of exposure in controls = ad/bc
Good approx of RR when cases/controls rep the population, dz does not occur frequently |
|
|
Term
| Attributable Risk (how much of the disease was due to the exposure?) |
|
Definition
Difference in incidence of exposed versus unexposed
AR = (I in exposed) - (I in unexposed)
AR = a/(a+b) - b/(b+c)
AR% = AR / Incidence of exposed |
|
|
Term
| Population Attributable Risk (PAR) |
|
Definition
PAR = I in total population - I in unexposed
PAR% = PAR/Incidence in Total Population |
|
|
Term
| Pyramid with study designs (weak to strong) |
|
Definition
| Test-tube research, animal research, ideals/editorials/opinions, case reports, case series, case control studies, cohort studies, RCT, RCT w/ Double blind! |
|
|
Term
|
Definition
Systemic review with statistical analysis - answers one specific question
Considered primary research with included studies treated as data
Statistical methods to combine and summarize the results of several studies |
|
|
Term
|
Definition
Summary of systematically gathered and analyzed primary research, saves clinicians time
Assembly, critical appraisal, and synthesis of all relevant studies to a topic |
|
|
Term
|
Definition
Patient Population
Intervention
Comparison
Outcomes |
|
|
Term
|
Definition
Primary: pre-diagnosis - prevent disease from occurring/reduce chances
Secondary: post-diagnosis - halt or slow progression in early stages
Tertiary: managing disease, preventing further deterioration |
|
|
Term
Screening biases:
Length-time bias
Lead-time bias
Selection bias |
|
Definition
*screening identifies slower growing or less progressive cases that have a better prognosis
*apparent increased survival duration introduced by lead time
*volunteers more likely to be healthier and adherent |
|
|
