Scientific terminology

1
Scientific terminology

• Empirical Fact
• Obvious observation from experiment.
• Assumption
• a fact or statement taken for granted
• Hypothesis
• An educated prediction of the outcome of an
  experiment
• Theory
• Assumption used to explain empirical facts
• Law
• Hypothesis that is well confirmed or established
• Cause
• entity that causes events to happen

2
Inductive/deductive reasoning

• Induction moves from the specific to general
• I've noticed previously that every time I kick a
  ball up, it comes back down, so I guess this next
  time when I kick it up, it will come back down,
  too.
• Deduction moves from general and to specific
• That's Newton's Law. Everything that goes up
  must come down. And so, if you kick the ball up,
  it must come down.

Arguments based on experience or observation are
best expressed inductively, while arguments based
on laws, rules, or other widely accepted
principles are best expressed deductively.
3
Inductive/deductive reasoning in chiropractic

• Chiropractic prefers deductive reasoning to
  inductive reasoning.
• Chiropractic is based upon a major philosophical
  premise . . . the starting point from which all
  deductions are made.
• . . . there is a Universal Intelligence in all
  matter constantly giving to it all of its
  properties and activities, thus maintaining it in
  existence. If this is a truth, then all the
  logical conclusions drawn from it are necessarily
  true.
• As your chiropractor I feel a responsibility to
  make you think, and help you to follow your own
  logic to discover truth . . . your reasoning goes
  into every decision that you make about life,
  particularly your health care.

http//www.drgreganderson.com/doc18.htm
4
Quantitative and qualitative research
5
Cause as risk factor

• A cause is a factor (or member of a set of
  factors) which results in a sequence of events
  that eventually result in an outcome.
• Exposure ? Outcome
• Cause ? Effect

6
Causation

• Causation is investigated by determining
  association (link) between exposure and outcome
• Examples
• Smoking / lung cancer
• Chiropractic care / optimal health
• Study of causation leads to inference
• passing from statistical sample data to
  generalizations (as of the value of population
  parameters) usually with calculated degrees of
  certainty

7
The Building Blocks of Theory

• Concepts
• Operational Definitions
• Propositions

8
Concepts

• Abstractions that allow classification of
  observations
• When scalar values can be assigned, they may
  become variables
• Variables must be operationally defined

9
Operational Definitions

• Description/delineation of the exact procedures
  for measuring or observing the phenomenon, event
  or behavior.

10
Propositions

• Propositions state the nature of the relationship
  between variables (concepts).
• An hypothesis is a statement about the expected
  relationship between two or more concepts that is
  based on a theory and that can be tested.

11
ExampleChiropractic Concepts

• Variables
• Adjustment
• Subluxation
• Health

12
Chiropractic Proposition I

• Between subluxation and health
• The gt the quantity, quality, severity of
  subluxations, the lt health.

13
Chiropractic Proposition II

• Between adjustment and subluxation
• The gt the quantity, (etc) of adjustment, the lt
  subluxation.

14
Chiropractic Proposition III

• Between adjustment and health
• The gt the quantity, (etc) of adjustment, the gt
  the health.

15
Chiropractic Theory and Clinical Epidemiology

• Subluxation assessment performance
• Adjustment treatment performance
• Health outcome performance

16
Epidemiological Reasoning

• Derive inferences regarding possible causal
  relationships
• Determine whether these relationships are
  spurious or true
• Discussion
• Associations
• Causal relationships
• Threats to validity
• Play of chance (statistical association)

17
Estimating Risk is there an association?

• Compare the risk of outcome in exposed to the
  risk of outcome in the nonexposed
• Relative Risk
• Calculated in Cohort Studies
• If RR1, risk in exposed equal to risk in
  non-exposed (no association)
• If RRgt1, risk in exposed greater than risk in
  nonexposed (positive association, possibly
  causal)
• If RRlt1, risk in exposed less than risk in
  nonexposed (negative association, possibly
  protective)

18
Basic biostatistics and epidemiological terms

• Populations and samples
• Prevalence and incidence
• Statistics
• Reliability
• Validity
• Bias

19
What do you mean, sample?

• One is interested in the characteristics of a
  population but must, for practical reasons,
  estimate them by describing a sample
• A subset of a population
• Selected from the population
• 100,000 randomly selected US residents
• LBP patients recruited for a clinical trial
• HMO members randomly selected from files
• Students enrolled in the EBC class

20
Populations and Samples
Figure 1.3
21
Who was the intended population represented by
the sample?

• Population
• Large groups of people in a defined setting or
  with a certain characteristic
• The general population
• Adults with low back pain
• Residents of North Carolina
• Members of a California HMO
• Chiropractic students
• Is the sample representative?

22
Sample size . . .

• Affects the probability of detecting a real
  difference between groups if there is one
• Affects the probability that a difference seen
  between samples reflects a real difference
  between the groups, and is not just a random
  occurrence

23
Sample size is not happenstance!

• To draw conclusions about the effectiveness of
  treatment (i.e. the difference between 2 groups
  outcomes) the RCT must have the statistical power
  to detect a real difference
• Drawing conclusions about a population based upon
  a sample
• Study says A B
• Caution - Small numbers increase the chance of a
  Type II error
• Study says A is not B
• Caution - Small numbers increase the chance of a
  Type I error

24
Prevalence and incidence

• Prevalence proportion of a defined population
  that has a condition at a given point in time
• Estimated with cross-sectional study
• Incidence proportion of a defined population
  that develops a condition over a defined period
  of time.
• Estimated with longitudinal study

25
Measures of morbidity

• Prevalence
• The number of affected persons present in the
  population at a specific time divided by the
  number of persons in the population at that time
• A slice through the population at a point in time
  at which it is determined who has the disease and
  who does not
• The numerator includes a mix of people with
  different durations of disease, and as a result
  we do not have a measure of risk
• Incidence
• The number of new cases of a disease that occur
  during a specified period of time in a population
  at risk for developing the disease
• A measure of risk
• The time period must be clearly stated

26
Prevalence and incidence illustrated
27
Statistical Tests

• Employed in explanatory studies
• Assess the role of chance as explanation of
  pattern observed in data
• Most commonly assesses how 2 groups compare on an
  outcome
• Is the pattern most likely not due to chance?
• The difference is statistically significant
• Is the pattern likely due to chance?
• The difference is not statistically significant
• No matter how well the study is performed, either
  conclusion could be wrong

28
Descriptive Statistics (examples)

• Simply describe what is (or was)
• Patient characteristics at baseline
• Examples mean, standard deviation, median,
  range, percentage
• Assess group comparability on baseline
  characteristics
• Assess generalizability of results to target
  population

29
Analytical (inferential) statistics

• Assess statistical significance with confidence
  intervals and p-values
• Within and between group differences
• Make inference about target population
• Must be appropriately interpreted in the context
  of the research question and the study design

30
Difficulties in the process of drawing inference

• Exposure of healthy subjects to suspected agents
  ethical?
• Withholding suspected beneficial agents ethical?
• Thus, epidemiologic evidence from observational
  studies very valuable

31
Bradford Hill Criteria for assessing causality

• Temporal relationship
• Exposure to the factor must precede the outcome
• Strength of association
• correlation
• Dose-response relationship
• As dose of exposure increases, so does risk of
  outcome
• Replication of findings
• Relationship consistent in different studies in
  different populations

32
Bradford Hill (cont.)

• Biologic plausibility
• Seek consistency of epi findings with existing
  biologic knowledge
• If absent, the meaning of the association may be
  difficult
• Consideration of alternate explanations
• Investigators should take other possible
  explanations in to account and rule them out
• Cessation of exposure
• Risk declines when exposure is reduced
• Consistency with other knowledge
• Ex increase in lung cancer rates following
  increase in cigarette sales

33
Smoking causes lung cancer

• Where is the RCT that gave evidence to this?
• Decades of observational research
• Observing the same thing in a wide variety of
  settings
• Some free of some types of bias, others free of
  other types of bias
• It takes a lot of observational data to even
  begin to suggest causation!
• Strength of association Consistency
  Specificity Temporality Biologic gradient
  (dose/response) Biologic plausibility
  Coherence of evidence

34
Smoking and lung cancer Example of causality
guidelines

• Cohort studies clearly demonstrate that smoking
  precedes lung cancer
• Risk ratios between smoking and lung cancer are
  high in many studies
• More cigarette smoke inhaled over a lifetime, the
  greater the risk of cancer
• Association found in both sexes, all races, all
  SES, etc.
• Burning tobacco produces carcinogenic compounds
  which contact pulmonary tissue

35
Key analytical statistical concepts p and r

• P-values and confidence intervals
• Reflects measure of effect relative to variation
  and sample size
• Variation consistency
• Example Mean value of 5
• 1, 10, 2, 11, 1
• 5, 4, 3, 6, 7 (less variation)
• R-values and correlation
• Pearsons correlation coefficient

36
P-values (pprobability)

• A statistical value that indicates the
  probability that the observed pattern is due to
  chance alone
• How confident we can be in the conclusion?
• this result was significant at plt0.05
• Example 50 patients each treatment group
• 25 get better with tx. A 35 get better with tx.
  B
• Statistically speaking, and all other things
  being equal, we could expect this result to occur
  by chance no more than 5 times in every 100
  trials
• Statistically, it is possible, but unlikely, that
  the groups could actually be the same, and a
  difference of this magnitude would be seen

37
R-values (Pearsons Correlation coefficient)

• Whether two ranges of data move together
• are large values of one set associated with large
  values of the other (positive correlation?
• Are small values of one set associated with large
  values of the other (negative correlation)?
• Are values in both sets unrelated (correlation
  near zero)?
• R2 percentage of variable 1 explained by
  variable

38
Significant Findings

• jargon term
• Need to consider BOTH statistical and clinical
  significance

39
Clinical Significance

• i.e. clinical importance
• is defined before study is conducted
• assessed with descriptive statistics (e.g. mean
  improvement in outcome measure)

40
Statistical Significance

• Interpreting p-values
• plt0.01 ? statistically significant difference!
• 0.01?p?0.05 ? statistically significant
  difference
• 0.05ltp?0.10 ? borderline statistically
  significant difference (but not significant)
• pgt0.10 ? no statistically significant difference

41
Possible Scenarios

• statistically and clinically significant findings
• clinically significant, but not statistically
  significant
• statistically significant, but not clinically
  significant

42
Reliability and Validity

• Reliability is consistency
• Lack of reliability is a problem with random
  error
• CHANCE
• Validity is TRUTH or ACCURACY
• Lack of validity is a problem with systematic
  error
• BIAS