+12 more

Explain each and every line with reasons with details like you are giving a lecture. Also explain complicated terms that are written here explain as explain to non medico

Reading File

analyse_media

The photos cover several separate Public Health Sciences topics. A full line-by-line lecture on all of them would be very long, so here is the map:

Biostatistics: describing data
- Mean, median, mode
- Range, standard deviation, variance, standard error
- Normal distribution and the 68-95-99.7 rule
- Skewed distributions: positive and negative skew
Hypothesis testing
- Null hypothesis (H₀) and alternative hypothesis (H₁)
- p value
- Type I error (α), Type II error (β), and statistical power
- Statistical significance versus clinical significance
Confidence intervals and statistical tests
- Confidence interval and what 95% confidence means
- t-test, ANOVA, chi-square test, Fisher exact test
- Meta-analysis and common study-design terms
Correlation, causation, and ethics
- Pearson correlation coefficient
- Autonomy, beneficence, nonmaleficence, justice
- Decision-making capacity, consent, confidentiality
- Surrogate decision-maker and advance directives
Communication skills in healthcare
- Patient-centred interview
- PEARLS framework for empathy and rapport
- SPIKES framework for giving bad news
- Gender-sensitive history, cultural formulation, motivational interviewing, trauma-informed care

All these topics are connected by one big idea: healthcare is not just about medicines and diagnoses. It also involves understanding numbers correctly, avoiding wrong conclusions from research, respecting a person's choices, and communicating in a way that helps rather than harms.

Explain each and every line with reasons with details like you are giving a lecture. Also explain complicated terms that are written here explain as explain to non medico

Reading File

This is a large set of pages. To explain every line properly, including the reason behind each point and each difficult word in plain non-medical language, we should do it page by page. Otherwise the explanation will become confusing and incomplete.

Pages in your photos

Page 264-265: Biostatistics and hypothesis testing
- Mean, median, mode
- Standard deviation, variance, standard error
- Normal distribution and skewness
- Null and alternative hypotheses
- p value, Type I and Type II errors, power
- Statistical versus clinical significance
Page 266: Confidence intervals, common statistical tests, and meta-analysis
- Confidence intervals and 95% CI
- t-test, ANOVA, chi-square, Fisher exact test
- Meta-analysis and study quality
Page 267: Correlation and ethics
- Pearson correlation
- Autonomy, beneficence, nonmaleficence, justice
- Capacity, consent, confidentiality, surrogate decisions
Pages 269-271: Communication skills
- Verbal and non-verbal communication
- Patient-centred interview
- PEARLS method for building rapport
- SPIKES method for breaking bad news
- Gender-sensitive care
- Cultural interview
- Motivational interviewing
- Trauma-informed care

I will begin with Page 264: basic statistics and hypothesis testing, explaining it like a lecture for a non-medical learner. After that, we can take the next page in the same depth.

Page 264: Understanding data before making medical conclusions

Statistics is the science of learning from numbers.

In healthcare, imagine we want to know:

What is the average blood pressure in a village?
Does a new medicine reduce fever better than the old medicine?
Are smokers more likely to get lung disease?
Is a test result genuinely meaningful, or could it have appeared only by chance?

Statistics helps answer these questions carefully.

1. Statistical distribution

A distribution means the pattern formed when we arrange data values.

Example

Suppose we measure the heights of ten people:

150, 154, 156, 158, 160, 161, 164, 165, 168, 172 cm.

If we put these values in order or draw them as a graph, we can see:

which height is common,
how spread out the heights are,
whether most people have similar heights,
whether a few people are unusually short or unusually tall.

That overall pattern is called the distribution of data.

Why it matters: before choosing a statistical test or trusting an average, we need to know what the data look like.

Measures of central tendency

The phrase sounds difficult, but its meaning is simple.

Central means middle or typical.
Tendency means the direction in which values gather.

So, measures of central tendency tell us the “typical” or central value of a group.

The three main measures are:

Mean
Median
Mode

2. Mean

Text idea: “Mean = sum of values / total number of values.”

The mean is the ordinary average.

Formula

[ \text{Mean} = \frac{\text{Sum of all values}}{\text{Number of values}} ]

Example

Suppose five students receive these marks:

60, 70, 80, 90, 100

Step 1: Add the marks.

[ 60 + 70 + 80 + 90 + 100 = 400 ]

Step 2: Count students.

There are 5 students.

Step 3: Divide total marks by number of students.

[ 400 / 5 = 80 ]

So the mean mark is 80.

Why do we use the mean?

It uses every value in the group, so it gives a good overall picture when values are reasonably similar.

Limitation: mean is affected by outliers

An outlier is a value that is very far from the rest.

Example: monthly incomes of five people:

₹20,000, ₹22,000, ₹24,000, ₹25,000, ₹10,00,000.

The income of ₹10,00,000 is extremely higher than the others. It is an outlier.

The mean becomes very high and gives the misleading impression that everyone earns a lot. Therefore, the mean can be distorted by unusually large or unusually small values.

3. Median

Text idea: “Median = middle value of a list of data sorted from least to greatest.”

The median is the middle value after arranging all values from lowest to highest.

Example with an odd number of values

Values:

10, 20, 30, 40, 50

The middle value is 30.

Therefore, median = 30.

Example with an even number of values

Values:

10, 20, 30, 40

There are two middle values: 20 and 30.

We take their average:

[ (20 + 30) / 2 = 25 ]

Therefore, median = 25.

Why is median useful?

The median is much less affected by extreme values.

Return to income:

₹20,000, ₹22,000, ₹24,000, ₹25,000, ₹10,00,000.

The middle value is ₹24,000.

So the median shows the income of a typical person better than the mean does.

When do we prefer median?

Use median when data are strongly uneven or contain extreme values, for example:

income,
hospital stay duration,
number of children in families,
waiting time,
very high laboratory values in a few severely ill people.

4. Mode

Text idea: “Mode = most common value.”

The mode is the value that occurs most often.

Example

Blood groups in a group:

A, B, B, O, B, AB, A

Here, B occurs three times, more than any other blood group.

So, mode = B.

Why is mode useful?

Mode is particularly useful for things that are categories rather than numbers.

Examples:

Most common blood group
Most common symptom
Most commonly used contraceptive method
Most common type of cancer in a region
Most common answer in a survey

You cannot calculate an average blood group. Therefore, mean and median are not useful there, but mode is.

Measures of dispersion

“Dispersion” means spread.

Central tendency tells us the middle or typical value. But that is not enough.

We also need to know whether people’s values are close to that typical value or scattered widely.

Compare these two groups

Group A: 68, 69, 70, 71, 72
Mean = 70

Group B: 20, 45, 70, 95, 120
Mean = 70

Both groups have the same mean, 70. But Group B is far more spread out.

This is why we need measures of dispersion.

The main ones shown are:

Range
Standard deviation
Variance
Standard error

5. Range

Meaning

The range is the difference between the largest and smallest value.

Formula

[ \text{Range} = \text{maximum value} - \text{minimum value} ]

Example

Values: 10, 15, 18, 25, 30

Maximum = 30
Minimum = 10

[ \text{Range} = 30 - 10 = 20 ]

Why is range useful?

It is easy and quick. It tells us the total span of values.

Limitation

Range depends only on two values: the highest and lowest. One extreme outlier can make the range look large, even if all other values are close together.

6. Standard deviation, SD

Text idea: “Standard deviation tells how much variability exists in a set of values around the mean.”

This is one of the most important ideas in statistics.

Standard deviation, written as SD, tells us how far values usually lie from the mean.

In everyday language, it tells us whether people in the group are mostly similar or very different.

Small SD

A small SD means values are close to the mean.

Example: 68, 69, 70, 71, 72.

Mean = 70. All values are near 70.

So SD is small.

Large SD

A large SD means values are far from the mean.

Example: 20, 45, 70, 95, 120.

Mean = 70, but some values are far below or far above 70.

So SD is large.

Why does SD matter in medicine?

Suppose two medicines reduce blood pressure by an average of 10 mmHg.

In Medicine A, nearly all patients improve by around 8 to 12 mmHg. Small SD.
In Medicine B, some patients improve by 30 mmHg, some do not improve at all, and some worsen. Large SD.

Although the mean improvement is the same, Medicine A gives more predictable results.

7. Variance

Text idea: “Variance = (SD)².”

Variance is simply the square of the standard deviation.

[ \text{Variance} = (\text{SD})^2 ]

If SD is 5, variance is:

[ 5^2 = 25 ]

Why square it?

When we calculate deviations from the mean, some are negative and some are positive. If we simply add them, they cancel out.

For example, relative to a mean of 70:

60 is -10 away from the mean.
80 is +10 away from the mean.

Adding -10 and +10 gives zero, which wrongly suggests no variation.

Squaring removes the minus sign:

[ (-10)^2 = 100 ]

[ (+10)^2 = 100 ]

Now both differences contribute properly to the measure of spread.

Practical point

Variance is used in calculations and statistical tests. SD is usually easier to understand because it remains in the original unit.

For example:

SD of height: centimetres
SD of blood pressure: mmHg
SD of blood glucose: mg/dL

Variance has squared units, such as cm², which is less intuitive.

8. Standard error, SE

Text idea: “Standard error estimates how much variability exists in a theoretical set of sample means around the true population mean.”

This sounds hard, so break it down.

Population

A population means the entire group we want to know about.

Examples:

Every adult in India
Every patient with diabetes in a city
All newborn babies in a hospital in a year

Usually, studying everyone is impossible.

Sample

A sample is a smaller group selected from the full population.

Example: instead of measuring the blood pressure of every adult in a city, we measure 1,000 randomly selected adults.

Sample mean

The average obtained from that one sample is called the sample mean.

Key idea

If we took many different samples from the same population, each sample would give a slightly different mean.

The standard error describes how much these sample means would vary from one another.

So:

SD tells us how much individual people differ from each other.
SE tells us how precisely our sample mean estimates the true population mean.

Formula

[ SE = \frac{SD}{\sqrt{n}} ]

Where:

SE = standard error
SD = standard deviation
n = sample size
(\sqrt{n}) = square root of the number of people studied

Why does a bigger sample lower SE?

Suppose you ask only 5 people about their average daily income. Your result can change greatly depending on whom you select.

If you ask 5,000 people chosen properly, one unusual person has much less influence.

Therefore, a larger sample gives a more precise estimate, so the standard error becomes smaller.

Normal distribution

Text idea: “For normal distribution, mean is the best central measure.”

A normal distribution is a common bell-shaped data pattern.

It looks like this conceptually:

Most values are near the centre.
Fewer values are at the very low and very high ends.
The two sides are mirror images.

Example

Many natural human measurements approximately follow a normal distribution:

height in a large adult population,
some blood-test values,
IQ scores,
measurement errors.

Not every medical measurement follows this pattern, but many do approximately.

In a normal distribution

[ \text{Mean} = \text{Median} = \text{Mode} ]

Why?

Because the distribution is perfectly balanced:

the arithmetic centre is the mean,
the middle-position value is the median,
the most common value is at the highest central peak, the mode.

9. The 68-95-99.7 rule

The diagram on the page shows the normal curve and standard deviations from the mean.

It says:

about 68% of values lie within 1 SD of the mean,
about 95% lie within 2 SDs,
about 99.7% lie within 3 SDs.

Example

Suppose average adult height is 170 cm and SD is 10 cm.

Within 1 SD

[ 170 \pm 10 ]

That means 160 to 180 cm.

About 68 out of every 100 people will lie in this range.

Within 2 SDs

[ 170 \pm 20 ]

That means 150 to 190 cm.

About 95 out of every 100 people will lie in this range.

Within 3 SDs

[ 170 \pm 30 ]

That means 140 to 200 cm.

About 99.7 out of every 100 people will lie in this range.

Why is this useful clinically?

It helps us identify whether a value is unusual.

For example, if a laboratory value lies far beyond 2 or 3 SDs from the expected average, doctors may investigate. However, an unusual number is not automatically disease. It must be interpreted with symptoms, medical history, examination, and repeat testing when appropriate.

Skewed distributions

The page also discusses positive skew and negative skew.

A skewed distribution is not symmetrical. It has a longer tail on one side.

This often happens when a few extreme observations pull the average in one direction.

10. Positive skew

Text idea: “Asymmetry with a longer tail on right. Mean falls closer to tail.”

In a positive skew, also called a right skew, most values lie toward the lower side, while a few unusually high values stretch the graph toward the right.

The long tail points right.

Example: income

Most people may earn between ₹20,000 and ₹60,000 per month, but a few people earn lakhs or crores. Those very high incomes create a long right-side tail.

Order in positive skew

[ \text{Mode} < \text{Median} < \text{Mean} ]

Why?

Mode stays where values are most common, toward the lower side.
Median is in the middle of the ordered values.
Mean is pulled upward by the few very high values, so it lies closest to the right tail.

Practical implication

For positively skewed data, the median is often a better description of the “typical” person than the mean.

11. Negative skew

Text idea: “Asymmetry with longer tail on left. Mean falls closer to tail.”

In a negative skew, also called a left skew, most values lie toward the high side, but a few very low values stretch the graph toward the left.

Example

Suppose an easy exam is given to a class. Most students score 80 to 100, but a few score very low. The low marks create a long left tail.

Order in negative skew

[ \text{Mean} < \text{Median} < \text{Mode} ]

Why?

The few very low values pull the mean toward the left-side tail.

Hypothesis testing

Hypothesis testing is a structured method used to decide whether an observed difference is probably real or might simply be due to chance.

Example question

A doctor compares two fever medicines:

Medicine A: standard treatment
Medicine B: new treatment

The researcher asks:

Does Medicine B truly reduce fever more than Medicine A?

One small study cannot give perfect certainty. Hypothesis testing helps decide what the data support.

12. Null hypothesis, H₀

Text idea: “A hypothesis with no difference or association.”

The null hypothesis, written as H₀, is the starting assumption that there is no real effect, no difference, or no relationship.

Examples

The new medicine and old medicine work equally well.
Smoking has no association with lung cancer.
The average blood pressure is the same in men and women.
A health education programme does not improve vaccination rates.

Why begin with “no difference”?

Because random chance can create apparent differences in small samples.

Suppose one group has an average recovery time of 6 days and another 7 days. That difference could be real, but it could also happen merely because different patients happened to enter the two groups.

The null hypothesis forces us to ask:

Is the observed difference large enough that chance is an unlikely explanation?

13. Alternative hypothesis, H₁

Text idea: “A hypothesis with at least one difference or relationship.”

The alternative hypothesis, written as H₁ or sometimes Hₐ, says that a real difference or relationship exists.

Examples

The new medicine works differently from the old medicine.
Smoking is associated with lung cancer.
The health education programme changes vaccination rates.

Important point

The alternative hypothesis is not automatically proven just because we want it to be true. Research data must provide enough evidence against the null hypothesis.

14. p value

Text idea: “Probability of getting results at least as extreme as those observed in the sample, assuming H₀ is correct.”

This is one of the most misunderstood concepts in statistics.

A p value asks:

If there were truly no difference in the real world, how likely would it be to obtain results like ours, or more extreme results, just by chance?

Example

Assume a new medicine truly has no advantage over the old medicine.

If, despite that, your study finds a very large difference, this could be:

a real effect, or
a rare chance event.

The p value estimates how surprising your observed result would be if the “no real difference” assumption were correct.

p < 0.05

A conventional cutoff is:

[ p < 0.05 ]

This means the observed result would occur less than 5 times out of 100 by chance alone, if the null hypothesis were true.

Then we usually call the result statistically significant and reject H₀.

What p < 0.05 does not mean

It does not mean:

there is a 95% chance that the research conclusion is true;
the treatment definitely works;
the result is clinically important;
only a 5% chance exists that the null hypothesis is true.

Those are common but incorrect interpretations.

15. Type I error, alpha error, α

Text idea: “Concluding that an effect or difference exists when none exists. Also called false-positive error.”

A Type I error happens when we reject H₀ even though H₀ is actually true.

In simple words:

We say there is an effect, but in reality there is no effect.

This is a false positive.

Example

A screening test says a person has a disease, but the person actually does not have the disease.

Or a study concludes that a medicine works better, but in reality it does not. The difference arose by chance.

Alpha, α

The symbol α represents the planned probability of making a Type I error.

Usually:

[ \alpha = 0.05 ]

That is why p < 0.05 is often used as the threshold for statistical significance.

Why is it called “false positive”?

“Positive” means we claim an effect or difference exists.
“False” means that claim is incorrect.

16. Type II error, beta error, β

Text idea: “Stating that there is no effect or difference when an effect exists. Also called false-negative error.”

A Type II error happens when we fail to reject H₀ even though H₀ is false.

In simple language:

We say there is no effect, but in reality an effect does exist.

This is a false negative.

Example

A study concludes that a new medicine is no better than the old medicine. But the new medicine actually is better. The study failed to detect the benefit.

Why could this happen?

Too few participants were studied.
The true benefit was small.
Measurements were inaccurate.
There was too much variation between people.

17. Power

Text idea: “Power = 1 - β.”

Power means the ability of a study to detect a real effect when it truly exists.

[ \text{Power} = 1 - \beta ]

Example

If β = 0.20:

[ \text{Power} = 1 - 0.20 = 0.80 = 80% ]

An 80% power means that if a true difference exists, the study has an 80% chance of detecting it.

What increases power?

Larger sample size
- More participants usually give more reliable information.
Bigger true effect
- It is easier to detect a treatment that reduces deaths by 50% than one that reduces them by 1%.
Better measurements
- Accurate instruments and clear definitions reduce noise.
Less variation
- If patients are extremely different from each other, a true effect becomes harder to identify.

18. Statistical significance versus clinical significance

This is a very important distinction.

Statistical significance

A result is statistically significant when it is unlikely to have occurred by random chance alone according to a preset cutoff, often p < 0.05.

Clinical significance

A result is clinically significant when the difference is large enough to matter in real life to patients.

Example

A pain medicine reduces pain score from 7.0 to 6.8 out of 10 in 100,000 people.

Because the sample is huge, the difference may be statistically significant.

But would a patient feel a difference of only 0.2 points? Probably not.

So it may be statistically significant but not clinically meaningful.

Opposite situation

A new emergency treatment could reduce death from 10% to 7%, which is clinically very important.

But if the study has only 20 patients, it might not reach statistical significance because the sample is too small.

So a non-significant result does not always prove “no benefit.” It may mean the study lacked sufficient power.

Quick memory table

Term	Meaning in everyday language
Mean	Ordinary average
Median	Middle value after sorting
Mode	Most common value
Range	Highest value minus lowest value
SD	Usual spread of individual values around the mean
SE	Precision of the estimated sample mean
Normal distribution	Symmetrical bell-shaped pattern
Positive skew	Long tail on the right due to a few high values
Negative skew	Long tail on the left due to a few low values
H₀	“No true difference/effect” starting assumption
H₁	“A true difference/effect exists” claim
p value	How surprising the data are if H₀ is true
Type I error	False positive: claiming an effect that is not real
Type II error	False negative: missing an effect that is real
Power	Ability to detect a real effect
Statistical significance	Unlikely to be chance alone
Clinical significance	Important enough to matter to patients

Explain rest of the pages like this

Reading File

Finding Sources

Searching PubMed

patient centered communication AND systematic review

Searching the Web

AHRQ patient centered communication shared decision making health literacy official

Page 266: Confidence intervals, statistical tests, and meta-analysis

This page continues the statistics topic. The main idea is:

A study gives us an estimate, but an estimate is never perfectly exact. Statistics helps us show how uncertain that estimate is and decide whether groups are truly different.

1. Confidence interval

Line: “Range of values within which the true mean of the population is expected to fall, with a specified probability.”

Let us break this sentence into small parts.

What is a “true mean”?

A mean is an average.

Suppose we want to know the average blood pressure of every adult in a city. Testing every adult is usually impossible. Therefore, researchers test a smaller group, called a sample.

The average blood pressure found in that sample is an estimate of the city’s real average.

The real average of all adults in that city is called the population mean.

Population = the whole group we want to understand.
Sample = the smaller group actually studied.
Population mean = the real average of everyone.
Sample mean = the average in the people we tested.

Because a sample contains only some people, its average may not match the population’s real average exactly.

A confidence interval, or CI, gives a reasonable range in which the real population value is likely to lie.

Example

Suppose 100 patients are studied. Their average systolic blood pressure is 130 mmHg.

The report says:

[ \text{Mean blood pressure} = 130 \text{ mmHg, 95% CI: 126 to 134} ]

This means the study’s best estimate is 130 mmHg, but the true average in the entire population is plausibly somewhere between 126 and 134 mmHg.

It does not mean that 95% of individual people have blood pressure between 126 and 134. It refers to uncertainty about the average, not the spread of individual measurements.

2. What “95% confidence interval” really means

Line: “For a 95% CI, the true mean lies within the interval in 95% of repeated samples.”

This sentence is often explained incorrectly.

Imagine we repeat the same study 100 times. Each time, we take a new properly chosen sample of 100 people and calculate a 95% CI.

About 95 out of those 100 intervals would contain the true population mean. About 5 would miss it.

The true population mean itself does not keep moving. It is one fixed, though unknown, number. The intervals change because each sample is a little different.

Why use 95%?

95% is a convention. It gives a balance:

A wider interval gives more confidence, but it is less precise.
A narrower interval is more precise-looking, but it gives less confidence.

For example:

99% CI: greater confidence, but wider.
95% CI: commonly used balance.
90% CI: narrower, but less confidence.

3. Precision and width of the confidence interval

A narrow CI means the estimate is more precise.

Example:

Drug A reduces blood pressure by 10 mmHg, 95% CI 9 to 11.
Drug B reduces blood pressure by 10 mmHg, 95% CI 1 to 19.

Both studies have the same estimated effect: 10 mmHg. But the first estimate is much more precise because its interval is narrow.

What makes a confidence interval narrow?

A larger sample size
- More people generally provide more reliable information.
- Random variation has less effect.
Less variation among participants
- If people’s measurements are very scattered, uncertainty is greater.
More accurate measurement
- A reliable measuring method produces a more precise estimate.

4. Confidence interval and statistical significance

For many comparisons, especially differences in means:

If the 95% CI for a difference does not include 0, the result is usually statistically significant at the 5% level.
If the 95% CI includes 0, the study cannot rule out “no difference.”

Example

A medicine lowers blood pressure by 5 mmHg compared with placebo.

Result A

[ \text{Difference} = -5 \text{ mmHg, 95% CI: -8 to -2} ]

Zero is not within the interval. The medicine probably has a real effect.

Result B

[ \text{Difference} = -5 \text{ mmHg, 95% CI: -12 to +2} ]

This interval includes zero. The medicine may help, but the result is uncertain. It is also possible that there is no real difference.

For ratios such as odds ratio, risk ratio, or relative risk, the “no effect” value is 1, not 0.

Common statistical tests

A statistical test is a mathematical method that helps answer a question such as:

Is this observed difference likely to be real, or could it have happened by chance?

The correct test depends on:

What kind of data we have
How many groups are being compared
Whether the data are numerical or categorical
Whether the same people are measured twice or two separate groups are compared
Whether the data are normally distributed

5. t-test

Line: “Checks differences between means of two groups.”

A t-test compares the averages of two groups.

Example

A researcher compares average haemoglobin levels in:

Group 1: people receiving iron tablets
Group 2: people not receiving iron tablets

If the groups have different average haemoglobin values, a t-test helps decide whether that difference is likely to be real.

“Mean” in simple language

The mean is the usual arithmetic average:

[ \text{Mean} = \frac{\text{sum of all values}}{\text{number of values}} ]

Two common types

Independent t-test

Used when the groups contain different people.

Example:

50 patients receive Drug A.
Different 50 patients receive Drug B.

The two groups are independent because one person belongs to only one group.

Paired t-test

Used when the same people are measured twice.

Example:

A patient’s blood pressure is measured before treatment.
The same patient’s blood pressure is measured after treatment.

The values are paired because each “before” value belongs to the same person as the “after” value.

Why does this distinction matter?

People differ naturally. A paired test accounts for the fact that the same person is being compared with themselves. This removes much of the person-to-person variation.

6. ANOVA

Line: “Checks differences between means of three or more groups.”

ANOVA stands for Analysis of Variance.

The name sounds strange because ANOVA is often used to compare averages. It does this by examining whether variation between groups is bigger than the normal variation within groups.

Example

A researcher wants to compare average weight loss from three diets:

Diet A
Diet B
Diet C

A t-test compares two groups. If we repeatedly use t-tests for many groups, the chance of a false-positive result increases. Therefore, ANOVA is used first.

What does ANOVA tell us?

ANOVA can tell us:

At least one group differs from the others.

But it does not immediately tell us which specific groups differ.

If ANOVA is significant, researchers often do a further comparison called a post-hoc test.

Post-hoc means “after this.”
It is an additional test done after ANOVA to identify where the difference lies.

7. Chi-square test

Line: “Checks differences between two or more proportions or categorical outcomes.”

A chi-square test, written as (\chi^2), is used when data are categories rather than numerical measurements.

What are categorical data?

Data placed into labels or groups.

Examples:

Male / female / another gender identity
Smoker / non-smoker
Disease present / disease absent
Vaccinated / not vaccinated
Improved / did not improve

What is a proportion?

A proportion is a part of the total.

Example:

Out of 100 people, 40 are vaccinated.

[ 40/100 = 0.40 = 40% ]

Example of chi-square

Suppose 100 smokers and 100 non-smokers are studied.

Group	Lung disease present	Lung disease absent
Smokers	30	70
Non-smokers	10	90

The chi-square test asks:

Is the difference in disease frequency between smokers and non-smokers probably greater than chance would explain?

It does not, by itself, prove that smoking caused the disease. Association and causation are different ideas.

8. Fisher exact test

Line: “Checks differences between two proportions when sample size is small.”

The Fisher exact test is used for categorical data, like the chi-square test.

The main difference is that it works better when numbers are very small.

Example

Suppose a rare side effect occurs in a small trial:

Group	Side effect	No side effect
Drug A	1	9
Drug B	0	10

The numbers are too small for the usual chi-square approximation to be dependable. Fisher exact test calculates the probability more accurately.

Why is it called “exact”?

It does not rely on the approximations that chi-square commonly uses. It calculates the probability of the possible arrangements of the observed data more directly.

Meta-analysis

Line: “A method of statistical analysis that pools summary data from multiple studies.”

A meta-analysis combines the numerical results of several studies addressing the same question.

Example

Suppose 12 studies compare a new vaccine with an old vaccine. Every study is relatively small. Some show benefit, some do not.

A meta-analysis combines their results to obtain a larger overall estimate.

“Pools data” means

It combines information in a planned statistical way. It does not merely count how many studies are positive versus negative.

Usually, better and larger studies are given more importance, called weight.

9. Why meta-analysis can be useful

Meta-analysis may:

Increase the total number of participants
Improve the ability to detect a small true effect
Provide a more precise estimate
Help explain why different studies have apparently different results

A systematic review that includes meta-analysis is often high on the evidence hierarchy, but only if the included studies are good quality and sufficiently similar.

10. Systematic review versus meta-analysis

These terms are related, but not identical.

Systematic review

A systematic review is a careful, pre-planned method of finding, selecting, assessing, and summarising all relevant studies on a question.

It asks:

What does all the reliable research say about this question?

Meta-analysis

A meta-analysis is the mathematical combination of numerical results from studies.

A systematic review may exist without a meta-analysis. This happens when studies are too different to combine fairly.

Why should studies be similar?

Imagine mixing together studies of:

children and adults,
mild disease and severe disease,
low doses and very high doses,
entirely different outcomes.

Combining such studies may create a misleading average. This variation between studies is called heterogeneity.

Page 267: Correlation, causation, and ethics

Pearson correlation coefficient

Line: “A measure of linear correlation between two variables.”

A correlation asks whether two things tend to change together.

A variable is anything that can differ from one person, place, or time to another.

Examples:

Height
Weight
Age
Blood pressure
Number of cigarettes smoked per day
Hours of exercise per week

The Pearson correlation coefficient, usually written as r, measures the strength and direction of a straight-line relationship between two numerical variables.

Its value is from -1 to +1.

11. Positive correlation

Line: “Positive r value means a direct relationship.”

A positive correlation means that as one variable increases, the other tends to increase too.

Example

Height and weight often have a positive correlation:

Taller people tend, on average, to weigh more.
Shorter people tend, on average, to weigh less.

This does not mean every tall person weighs more than every short person. It only describes the general pattern.

12. Negative correlation

Line: “Negative r value means an inverse relationship.”

A negative correlation means that as one variable increases, the other tends to decrease.

Example

In many settings, more hours of physical activity may be associated with lower body-fat percentage.

One value rises, the other tends to fall.

13. Meaning of r values

Value of r	Meaning
+1	Perfect positive correlation
0	No linear correlation
-1	Perfect negative correlation
Around +0.8	Strong positive correlation
Around -0.8	Strong negative correlation
Around +0.4 or -0.4	Moderate relationship
Around +0.1 or -0.1	Weak relationship

The book shows examples such as:

strong negative correlation, (r = -0.8)
weak negative correlation, (r = -0.4)
no correlation, (r = 0)
weak positive correlation, (r = +0.4)
strong positive correlation, (r = +0.8)

The closer the value is to either +1 or -1, the tighter the points fall around a straight line.

14. Correlation does not prove causation

This is a central rule.

Correlation means association

Two things happen together more often than expected.

Causation means one thing actually produces a change in the other.

Example

Ice-cream sales and drowning deaths may both rise during summer. They are correlated.

But ice cream does not cause drowning.

The hidden factor is hot weather:

Hot weather leads to more ice-cream consumption.
Hot weather leads to more swimming.
More swimming increases the opportunity for drowning.

A third factor that affects both variables is called a confounder.

In health research

If people who drink more coffee have more heart disease, that does not automatically mean coffee causes heart disease. Perhaps coffee drinkers in that group were also more likely to smoke, sleep less, or have stressful jobs.

Therefore, correlation is a clue for further research, not proof of cause.

Core ethical principles

Ethics means deciding what is right and fair in how we treat people.

Healthcare ethics protects patients from being treated merely as “cases” or “data.”

The four major principles shown are:

Autonomy
Beneficence
Nonmaleficence
Justice

15. Autonomy

Line: “Obligation to respect patients as individuals with their own beliefs, values, and preferences.”

Autonomy means a person’s right and ability to make decisions about their own body and life.

A doctor may give medical advice, explain risks and benefits, and recommend a treatment. But a capable adult patient has the right to accept or refuse it.

Example

A patient understands that surgery may be life-saving but refuses it because of their own values. If the patient has decision-making capacity and is not being forced, that decision must generally be respected.

Autonomy does not mean patients must make decisions alone. They may involve family or trusted people if they wish.

Why it matters

The patient is not a passive object. Medical care should be based on the person’s goals, values, and informed choice. This is the basis of informed consent and shared decision-making.

16. Beneficence

Line: “Do good. Physicians have a special ethical duty to act in the patient’s best interest.”

Beneficence means actively trying to benefit the patient.

Examples:

Treating pain
Preventing disease
Explaining options honestly
Helping a patient receive appropriate care
Recommending a treatment that is likely to help

“Patient’s best interest” does not mean doctor’s preference

A doctor may believe one treatment is medically best. But the patient may value quality of life, religious beliefs, work responsibilities, or avoiding certain side effects differently.

Good care joins medical evidence with the patient’s own priorities.

17. Nonmaleficence

Line: “Do no harm.”

Nonmaleficence means avoiding unnecessary harm.

The traditional phrase is primum non nocere, meaning “first, do no harm.”

This does not mean doctors can never use a treatment that has side effects. Almost every treatment has some risk.

It means:

Do not expose a patient to harm unless there is a sound reason to believe the likely benefit justifies that risk.

Example

Chemotherapy may cause nausea, weakness, infections, and hair loss. It can still be ethical if it offers a meaningful chance of treating cancer or extending life, and the patient understands and accepts the trade-off.

18. Justice

Line: “Treat persons fairly and equally.”

Justice means fairness.

In healthcare, this includes fair access to care and fair distribution of limited resources.

Examples:

Do not deny care because of religion, caste, ethnicity, disability, income, sexual orientation, or gender identity.
Do not give scarce resources simply to the loudest or most influential person.
Use clear, medically relevant criteria when resources are limited.

Equality versus equity

Equality means giving everyone the same thing.
Equity means giving people the support they need to have a fair chance.

For example, providing a sign-language interpreter for a deaf patient is equitable. It is not “special treatment.” It allows meaningful access to the same healthcare information.

Page 268: Consent, capacity, confidentiality, and advance care planning

Informed consent

Line: “A process, not just a document/signature.”

Informed consent is not merely obtaining a signature on a form.

It is a conversation in which the patient:

Receives understandable information.
Has the chance to ask questions.
Understands the main choices.
Makes a voluntary decision.
Has capacity to make that decision.

A signed form without understanding is not meaningful informed consent.

19. Essential parts of informed consent

A. Explanation of the intervention

The patient should know what is being proposed.

Example:

“We recommend removing the appendix through an operation.”

B. Benefits

What good might the treatment do?

Example:

“The operation removes the infected appendix and reduces the risk of rupture.”

C. Risks

What harm or complications may occur?

Example:

“Possible risks include bleeding, infection, reaction to anaesthesia, or injury to nearby structures.”

D. Alternatives

What other reasonable choices exist?

This can include:

Another treatment
Waiting and watching
No treatment

E. Voluntariness

The patient should not be threatened, forced, tricked, or made to feel that care will be withheld unless they agree.

F. Capacity

The patient must be able to understand and decide.

Decision-making capacity

Line: “Patients generally have capacity unless shown otherwise.”

Capacity is the ability to make a specific medical decision at a particular time.

It is not the same as intelligence, education level, or agreeing with the doctor.

A patient can have capacity even if the doctor thinks their choice is unwise.

A person usually has capacity if they can:

Understand the information
Appreciate how it applies to their own situation
Reason about the options
Communicate a stable choice

Why is it decision-specific?

A person may be able to choose between two simple pain medicines but may not currently be able to understand a highly complex operation.

Capacity can also change over time. Delirium, severe illness, intoxication, medication effects, or a psychiatric crisis may temporarily impair it.

Confidentiality

Line: “Keep patient information private.”

Confidentiality means health workers must protect private patient information.

This includes:

Diagnosis
Test reports
Medication details
Sexual history
Mental-health history
Conversations during appointments

Why is it important?

Patients must feel safe to reveal private information honestly. If they fear gossip or unnecessary disclosure, they may avoid care or hide important facts.

Are there exceptions?

Yes, but they are limited and depend on law and context.

Examples may include:

Serious immediate risk of harm to the patient or another person
Certain legally reportable infectious diseases
Suspected abuse where reporting is legally required
A valid court order

Information should be shared only with the necessary people and only to the extent required.

Advance directives

Line: “Instructions given by a patient in anticipation of the need for medical decisions at a later time.”

An advance directive is a person’s written or clearly expressed wish about future healthcare, made while they have decision-making capacity.

It becomes relevant if they later become unable to communicate or decide.

Example

A person may write:

“If I have irreversible brain damage and cannot recover awareness, I do not want prolonged life support.”

Rules differ by country and region, but the ethical purpose is the same: respect the person’s own choices even when they cannot speak for themselves.

Living will

A living will is a type of advance directive. It describes what treatments a person would or would not want in specific future circumstances.

It is not a will about money or property. It is about medical care.

Durable power of attorney for healthcare

This is a document through which a person appoints another trusted person to make healthcare decisions if they lose capacity.

That person is sometimes called:

Healthcare proxy
Healthcare agent
Medical power of attorney
Surrogate decision-maker

Their role is not simply to choose what they personally want. They should try to represent what the patient would have wanted.

Surrogate decision-making

Line: “If a patient lacks capacity, decision-making may be made by a legally designated surrogate.”

A surrogate is someone who makes decisions for a patient who cannot decide for themselves.

Examples include a person chosen in an advance directive or, depending on local law, a spouse, parent, adult child, or another close relative.

Two standards guide the surrogate

1. Substituted judgment

The surrogate asks:

What would this patient have chosen if they were able to speak now?

This is preferred when the patient’s previous wishes are known.

2. Best-interest standard

If the patient’s wishes are unknown, the surrogate asks:

What option is most likely to benefit this patient and minimise suffering or harm?

Page 269: Respectful communication, confidentiality, and delivering difficult information

Gender-assigned sexuality history-taking

This section teaches respectful communication when discussing sexual health, gender, and relationships.

These topics may be private, sensitive, and connected with fear of judgement. A careless question can make a patient stop sharing important health information.

20. Avoid assumptions

Line: “Avoid making assumptions about sexual orientation, gender identity, and behaviours.”

A healthcare professional should not assume:

A person’s gender identity from appearance
Their sexual orientation from gender or marital status
The sex of their partner
Their sexual practices
Their pregnancy risk
Their risk of sexually transmitted infections

Why?

Assumptions can be wrong and may cause missed diagnoses or poor care.

For example, asking only a married woman whether she uses contraception with her husband may overlook her actual relationships, reproductive plans, or health risks.

Neutral, respectful questions work better.

21. Use the patient’s name and pronouns

Line: “Ask the patient their name and how they would like to be addressed.”

A pronoun is a word used in place of a name, such as he, she, or they.

A respectful approach is:

“What name would you like me to use?”
“What pronouns do you use?”

Why does this matter medically?

Respect reduces stress and builds trust. Trust helps patients disclose accurate information, return for follow-up, and participate in care.

The aim is not to make assumptions or turn the conversation into a debate. It is to address a person respectfully.

Culturally informed interview

Line: “Identify the problem through the patient’s perspective.”

A culturally informed interview means exploring how the patient understands their illness in the context of their family, community, language, religion, financial situation, and personal experiences.

Culture does not mean only ethnicity or religion. It can include values, family roles, education, migration experience, and beliefs about illness.

Useful questions

“What do you think is causing this problem?”
“What worries you most about it?”
“How has this affected your daily life?”
“What treatments have you already tried?”
“Are there traditional remedies or beliefs that are important for us to know about?”
“Who should be involved in decisions about your treatment?”

Why this is useful

A patient may use a home remedy, worry about stigma, need family approval, or have difficulty paying for medicines. These factors may affect whether a plan is realistic.

The goal is not to judge beliefs. It is to understand the person and make a safe, acceptable care plan.

Motivational interviewing

Line: “Counselling technique for facilitating behaviour modification by helping patients resolve ambivalence.”

Motivational interviewing is a respectful conversation method used when someone feels two opposing things at once.

That mixed feeling is called ambivalence.

Example

A person who smokes may think:

“I know smoking is harming me.”
“But it helps when I feel stressed.”
“I want to stop, but I am afraid I will fail.”

They are not simply “non-compliant.” They are conflicted.

Motivational interviewing helps them express their own reasons for change rather than being pushed, shamed, or lectured.

22. The spirit of motivational interviewing

It involves:

Partnership rather than commanding
Curiosity rather than judgement
Supporting the patient’s own reasons for change
Respecting that the patient has a choice

Example of unhelpful communication

“You must stop smoking immediately. You are destroying your lungs.”

This may make the person defensive.

Better approach

“What do you enjoy about smoking, and what concerns do you have about it?”
“On a scale from 0 to 10, how important is quitting to you?”
“Why did you choose 5 rather than 1?”
“What might make moving from 5 to 6 easier?”

This lets the person identify their own motivations.

23. SMART goals

The book highlights that a patient’s readiness for change should be assessed and goals should be SMART.

SMART means:

S - Specific: clear goal
M - Measurable: progress can be counted
A - Achievable: realistic for the person
R - Relevant: connected to the person’s own priorities
T - Time-bound: has a time frame

Vague goal

“I will exercise more.”

SMART goal

“For the next two weeks, I will walk for 20 minutes after dinner on Monday, Wednesday, and Friday.”

Why is this better? It tells the person exactly what to do, when to do it, and how success will be measured.

Trauma-informed care

Line: “Patients with a history of psychological trauma should receive care that avoids retraumatization.”

Psychological trauma is an experience that overwhelms a person’s sense of safety or control.

Examples can include:

Violence
Sexual assault
Childhood abuse
War
Serious accidents
Torture
Severe discrimination
Medical trauma

Trauma-informed care means providing care in a way that recognises these experiences may affect a person’s trust, reactions, and ability to tolerate examinations or procedures.

It does not require the patient to disclose trauma.

24. Main principles of trauma-informed care

Safety

Help the person feel physically and emotionally safe.

Example:

“Before I examine you, I will explain each step. Please tell me if you want me to stop.”

Choice

Give choices whenever possible.

Example:

“Would you prefer to have a support person in the room?”
“Would you prefer a clinician of a particular gender, if available?”

Collaboration

Work with the patient rather than doing things to them.

Trustworthiness

Be clear, honest, and predictable.

Empowerment

Help the patient regain a sense of control.

Why it matters

A sudden examination, unexpected touch, or being ignored during a procedure can trigger intense fear in a trauma survivor. Respectful explanation, permission, privacy, and choice can prevent unnecessary distress.

Page 270: Patient-centred interview, PEARLS, and SPIKES

Patient-centred interviewing techniques

A patient-centred interview is a healthcare conversation that focuses not only on disease, but also on the person living with it.

It asks:

What is the medical problem?
What does the patient think is happening?
What are they worried about?
What do they hope for?
What is realistic in their life?

This is not “being nice instead of being scientific.” It improves the accuracy of information and makes treatment plans more practical.

25. Introduction

Line: “Introduce yourself and ask the patient their name and how they would like to be addressed.”

The clinician should introduce themselves, state their role, and confirm the patient’s preferred name.

Example:

“Hello, I am Dr Khan, one of the doctors working with the team. How would you like me to address you?”

Why?

It establishes basic respect, prevents errors, and reduces the power imbalance between clinician and patient.

26. Agenda setting

Line: “Identify concerns and set goals by developing a joint agenda.”

An agenda is a shared plan for what will be discussed during the visit.

Example:

“What are the main things you would like us to talk about today?”
“I also want to discuss your blood test result. Shall we start with your pain and then discuss the result?”

Why?

Patients may have several worries but may mention the most important one only at the end. Agenda setting helps prevent this and makes the visit organised.

A joint agenda means both patient and clinician contribute to the plan.

27. Reflection

Line: “Actively listen and synthesize information offered by the patient, particularly with respect to the patient’s primary concern.”

Active listening means paying close attention and demonstrating that attention.

Synthesise means putting separate pieces of information together into a clear summary.

Example:

“You have had headaches for three months. They have become more frequent, you are worried it might be serious, and they are making it difficult to work. Is that correct?”

Why?

It verifies that the clinician understood correctly, lets the patient correct errors, and shows the patient they have been heard.

28. Validation

Line: “Legitimise or affirm the patient’s perspectives.”

To validate is to recognise that a person’s feeling or concern makes sense.

Example:

“It makes sense that you are anxious after seeing blood in your stool.”
“Anyone in your situation might feel frightened.”

Validation does not mean automatically agreeing with every belief. It means respecting the emotion and experience.

29. Recapitulation

Line: “Summarize what the patient has said so far to ensure correct interpretation.”

Recapitulation means a short summary of what has been discussed.

Example:

“Let me check that I have understood. The cough started two weeks ago, is worse at night, you have no fever, and your main worry is whether it could be asthma.”

Why?

It prevents misunderstanding before moving on to advice, tests, or treatment.

30. Facilitation

Line: “Encourage the patient to speak freely without guiding responses or leading questions.”

A leading question pushes someone toward a particular answer.

Example of a leading question:

“The pain is not very severe, right?”

This may make a patient minimise symptoms.

A better open question:

“How severe is the pain?”
“Can you describe what the pain feels like?”

Facilitation means using silence, nodding, and open questions to help the person speak freely.

PEARLS: building rapport

Rapport means a trusting, comfortable working relationship.

PEARLS is a memory aid:

P - Partnership
E - Empathy
A - Apology
R - Respect
L - Legitimation
S - Support

31. Partnership

Line: “Work together with the patient to identify primary concerns and develop preferred solutions.”

Example:

“Let us look at the options together and decide what fits your goals.”

The patient brings their experience, values, and goals. The clinician brings medical knowledge. Good decisions need both.

32. Empathy

Line: “Acknowledge the emotions displayed and demonstrate understanding.”

Empathy means recognising another person’s emotion and showing that you understand it.

Example:

“This diagnosis sounds overwhelming.”
“I can see that this has been very stressful for you.”

Empathy is not the same as pity. It does not mean “I feel sorry for you.” It means “I recognise your experience and I am taking it seriously.”

33. Apology

Line: “Take personal responsibility when appropriate.”

An apology may be needed for a delay, a mistake, poor communication, or an upsetting experience.

Example:

“I am sorry you waited so long without an update. You deserved clearer information.”

A good apology acknowledges what happened and does not blame the patient.

34. Respect

Line: “Commend the patient for discussing a problem or making constructive behaviours.”

Example:

“It took courage to bring this up.”
“You have been working hard to take your medicines regularly.”

Respect encourages honesty and effort. It should be genuine, not patronising.

35. Legitimation

Line: “Assure the patient that certain responses are understandable or common.”

Example:

“Many people feel nervous before surgery.”
“It is understandable to feel upset after receiving this news.”

This reduces shame and isolation.

36. Support

Line: “Reassure the patient that you will work together through difficult times and offer appropriate resources.”

Example:

“You do not have to manage this alone. We can involve a counsellor, social worker, and your family if you want.”

Support should be realistic. Do not give false promises such as “Everything will be fine” when the outcome is uncertain.

SPIKES: delivering bad news

Bad news is any information that seriously changes a person’s view of their future.

Examples:

A cancer diagnosis
A serious disability
A poor prognosis
Treatment failure
Death of a relative

SPIKES is a structured method for delivering difficult news with sensitivity.

S - Setting
P - Perception
I - Invitation
K - Knowledge
E - Emotions
S - Strategy and summary

37. Setting

Line: “Offer in advance for the patient to bring support. Eliminate distractions. Ensure privacy.”

Bad news should be discussed in a private, quiet place whenever possible.

The clinician should:

Sit down rather than stand over the patient
Avoid interruptions
Switch off unnecessary alerts
Ask whether the patient wants a family member or friend present
Allow enough time

Why?

Privacy communicates respect. A quiet setting allows the patient to react and ask questions without feeling exposed.

38. Perception

Line: “Determine the patient’s understanding and expectations of the situation.”

Before giving new information, ask what the patient already knows.

Example:

“What have you understood from the tests so far?”
“What are you expecting the scan may show?”

Why?

This prevents talking far above or below the patient’s understanding. It also identifies misunderstandings.

39. Invitation

Line: “Obtain the patient’s permission to disclose the news and what level of detail is desired.”

Not every person wants the same amount of information at once.

Example:

“Would you like me to explain all the details now, or would you prefer a general overview first?”
“Would you like your family member to be present?”

Respecting preferences is part of autonomy.

40. Knowledge

Line: “Share the information in small pieces without medical jargon, allowing time to process.”

Medical jargon means specialised technical words that may be difficult for a person without medical training.

Instead of saying:

“The biopsy confirms metastatic adenocarcinoma with a poor prognosis.”

Say:

“The tissue test shows cancer. It has spread beyond where it started. This makes it harder to cure, but we can discuss treatments that may slow it down and control symptoms.”

Then pause.

Why give information in small pieces?

Bad news can cause shock. A person may not remember everything after the first difficult sentence. Pausing allows them to absorb, react, and ask questions.

41. Emotions

Line: “Acknowledge the patient’s emotions and provide opportunity to express them.”

After bad news, a patient may cry, become silent, become angry, deny it, or ask the same question repeatedly.

The clinician should not rush to fill silence with more facts.

Helpful response:

“I can see this is a lot to take in.”
“Take your time. I am here.”
“What is going through your mind right now?”

The goal is not to remove emotion. The goal is to provide a safe space for it.

42. Strategy and summary

Line: “Give ready, discuss treatment options and goals of care, offer an agenda, and ask what the patient would like to do about the plan of action.”

Once the patient is ready, explain the next steps.

Example:

“The next step is to meet the cancer team. They will discuss whether treatment can shrink or control the cancer. We will also focus on pain, appetite, and support at home. Would you like to discuss the options today, or would you prefer time and another appointment with your family present?”

A goal of care means what matters most in treatment.

Examples:

Cure, if possible
Longer life
Less pain
Maintaining independence
Staying at home
Avoiding hospital admissions
Being able to attend an important family event

Page 271: Additional communication principles

This page reinforces the methods above. Its central message is:

Good healthcare communication is respectful, clear, patient-led, and adapted to the person’s circumstances.

Here are the practical rules in plain language.

43. Do not use unnecessary medical language

If technical terms are needed, explain them immediately.

Instead of:

“You have hypertension.”

Say:

“You have high blood pressure. It means the pressure of blood against your artery walls is higher than it should be over time.”

Then ask:

“Could you tell me in your own words what you understood?”

This is called teach-back. It checks whether the explanation was clear, not whether the patient is intelligent. The AHRQ health-literacy guidance similarly recommends clear communication and checking understanding.

44. Invite questions

Patients may remain silent because they are frightened, rushed, embarrassed, or feel they should not “bother” the doctor.

Instead of asking only:

“Do you have any questions?”

Try:

“What questions do you have?”
“Many people ask about side effects or what happens next. What would be most helpful for you to discuss?”

This makes questions feel expected and welcome.

45. Shared decision-making

Shared decision-making means patient and clinician make decisions together.

It is most useful where more than one medically reasonable option exists.

Example: a patient with knee arthritis may choose between:

Exercise and physiotherapy
Pain medicines
Injection
Surgery
Waiting and monitoring

The clinician explains benefits, harms, and uncertainty. The patient explains what matters to them, such as ability to work, avoid surgery, reduce pain, or remain active. The decision is made together. The AHRQ shared-decision framework describes this as patient-centred decision-making.

Final revision table

Topic	Plain meaning	Key reason
Confidence interval	A likely range for the true population value	Shows precision and uncertainty
t-test	Compares averages of two groups	Determines whether an observed difference may be real
ANOVA	Compares averages of three or more groups	Avoids repeated t-tests and excess false positives
Chi-square test	Compares category-based proportions	Useful for yes/no or group data
Fisher exact test	Categorical comparison for small samples	More reliable than chi-square for very small numbers
Meta-analysis	Combines results from multiple studies	Can improve precision and power
Correlation	Measures whether two variables change together	Does not prove one causes the other
Autonomy	Patient’s right to make informed choices	Respects the person’s values and control over their body
Beneficence	Act to help the patient	Healthcare should provide benefit
Nonmaleficence	Avoid unnecessary harm	Benefits should justify risks
Justice	Treat people fairly	Prevents discriminatory or unfair care
Informed consent	Understanding plus voluntary agreement	A signature alone is not enough
Capacity	Ability to understand, reason, and communicate a choice	Allows valid medical decisions
Confidentiality	Protect private health information	Builds trust and protects privacy
PEARLS	Partnership, Empathy, Apology, Respect, Legitimation, Support	Builds rapport
SPIKES	Method for giving bad news	Helps clinicians communicate difficult information humanely
Motivational interviewing	Helps patients resolve mixed feelings about change	Encourages lasting change without shame or force
Trauma-informed care	Care that promotes safety, choice, trust, and control	Reduces the chance of retraumatising a patient

This is a shared conversation. Sign in to Orris to start your own chat.