Unit 12 · How Medicine Knows

Before Lind, most doctors judged a treatment by memory and reputation. A famous doctor might say vinegar helped scurvy, and others would believe him. Nobody had checked it against anything else. If a sailor got better after vinegar, it was easy to give the vinegar the credit.

Lind’s test was different because it was a comparison. He chose patients with similar illness, kept them in the same place and gave them the same diet. The only planned difference between the pairs was the extra treatment. So when the citrus pair recovered fastest, the fruit was the most likely reason.

Scientists call the thing being changed the variable. A fair test changes one variable and holds the others steady. Lind’s trial had flaws. Two patients per group is very few, and he chose who got what himself. Still, the plan of comparing like with like became a model for medical research.

The lesson took time to spread. The British Navy began giving sailors lemon juice in 1795, more than 40 years later. In the 1930s scientists identified the missing substance as vitamin C. Lind had found the answer through a fair test long before anyone could explain it.

Imagine a doctor testing a new medicine. Without meaning to, she might give it to the patients who seem strongest. Those patients might have recovered anyway. The medicine would look better than it really is. This hidden unfairness is called bias. Researchers work hard to prevent it.

The fix is to let chance choose. In a randomized controlled trial, or RCT, each patient is assigned to a group by something like a coin toss or a computer list. Nobody picks. With enough people, the groups end up similar in age, health and habits, even in ways no one thought to measure.

One group gets the new treatment. The other, the control group, gets the usual care or a dummy treatment. Then researchers compare what happens. A famous early RCT was run in Britain in 1948. It tested the antibiotic streptomycin for tuberculosis, a lung infection, and helped make randomization standard in medicine.

Randomization does not make a trial perfect. People may drop out, or the groups may be too small. But it removes one of the biggest sources of error. That is why RCTs are trusted so much when the question is whether a treatment causes a result. Doctors still rely on them today.

Suppose you take a pill and believe it will ease your headache. You may feel better partly because you expect to. A dummy treatment with no active medicine, such as a sugar pill, is called a placebo. Trials often give a placebo to the control group so both groups take a pill that looks the same.

Expectations can also affect the people running a trial. A doctor who knows a patient got the real medicine might rate that patient as doing better. To prevent this, trials use blinding. In a single-blind trial, the patients do not know which group they are in.

In a double-blind trial, neither the patients nor the researchers who measure results know who got what. A separate team keeps the code. The code is opened only at the end. This way, hopes and hunches on both sides cannot tip the results. Everyone finds out together, at the same time.

Blinding is not always possible. You cannot easily hide whether someone had surgery or did an exercise program. In those cases, researchers may blind the people who measure the results, even if the patients know. Good studies report clearly who was blinded and who was not.

In many trials, some people in the placebo group report feeling better. This is called the placebo effect. It is strongest for things people feel and report, such as pain, nausea or tiredness. It is weaker or absent for things measured directly, such as the size of a tumor.

Several things can cause this. Expecting relief can change how we notice pain. Caring attention from a nurse or doctor can help people feel calmer. And many illnesses get better on their own over time. Doctors call that last one the natural course of an illness.

The opposite can happen too. Some people in placebo groups report side effects like headaches, just from expecting them. This is called the nocebo effect. It shows how strongly the mind and body are connected. Doctors take these reports seriously. They count as real experiences, even though the pill had no active medicine.

For researchers, the lesson is clear. A new medicine is not proven just because people feel better. It must help clearly more than a placebo does. That is why so many trials compare the treatment with a placebo instead of with nothing at all. The placebo sets the bar a medicine must clear.

Flip a coin four times. You might get three heads, even though the true chance is one in two. Flip it a thousand times and you will land very close to half heads. Small samples bounce around by luck. Large samples settle near the truth. The same rule holds for studies of medicine.

Trials work the same way. The number of people in a study is called the sample size. If only ten people take a new medicine, a few lucky recoveries can make it look amazing. With thousands of people, luck evens out and the real effect shows more clearly.

Before a trial starts, researchers calculate how many people they need. The smaller the effect they hope to detect, the more people they need. A result that is unlikely to be due to chance alone is called statistically significant. That phrase does not mean the effect is large or important. It only means chance is an unlikely explanation.

This is why Lind’s trial, with two sailors per group, could not prove much on its own. The citrus result was dramatic, which helped. But later, larger studies were needed to be sure. Size matters most when effects are modest, which is true of many modern medicines.

When researchers finish a study, they write a paper and send it to a science journal. The editor sends it to two or three other experts in the field. These reviewers read closely. They ask whether the methods were fair, whether the numbers add up and whether the conclusions go too far. This process is called peer review.

Reviewers often ask for changes or more analysis. Sometimes they recommend that the journal reject the paper. Only after the editor is satisfied is the study published. Many journals now also ask researchers to register their trial plans before starting, so they cannot quietly change their goals later.

Peer review is a filter, not a guarantee. Reviewers cannot redo the experiment. Mistakes and, rarely, fraud can slip through. Sometimes a published paper is later retracted, meaning the journal formally withdraws it. When that happens, the journal posts a notice so readers know the paper should not be trusted.

The strongest check comes after publication, when other scientists try to repeat the study. When independent teams get similar results, confidence grows. This is called replication. A single study, even a peer-reviewed one, is one piece of evidence, not the final word.

A new medicine starts in the lab and is usually tested on cells and animals first. If it looks promising and safe enough, the company asks the U.S. Food and Drug Administration, the FDA, for permission to test it in people. Human testing happens in steps called phases.

Phase 1 trials use a small group, often 20 to 100 people, mainly to check safety and find a safe dose range. Phase 2 trials include up to several hundred people who have the condition. They look for signs the drug works and watch side effects. Phase 3 trials are larger, often 300 to 3,000 people or more, and compare the drug with a placebo or current treatment.

If phase 3 goes well, the company sends all the data to the FDA. FDA scientists review it and decide whether the benefits outweigh the risks. After approval, doctors keep reporting problems. This watching of a drug in everyday use is sometimes called phase 4, or post-marketing surveillance.

The rules grew stronger over time. In the early 1960s, the drug thalidomide caused serious birth defects in thousands of babies in other countries. An FDA reviewer, Frances Kelsey, had held back its approval in the United States. In 1962 Congress passed a law requiring companies to show that drugs are effective, not only safe.

In summer, ice cream sales rise. So do drownings. Does ice cream cause drowning? No. Hot weather causes both: more people buy ice cream, and more people swim. When two things rise and fall together, they have a correlation. That alone does not show causation, where one thing actually makes the other happen.

The hidden third factor, here the hot weather, is called a confounder. Confounders are common in health. For example, people who take vitamins may be healthier. But people who take vitamins may also exercise more and smoke less. The vitamins might not be the reason at all.

Randomized trials help because chance spreads confounders evenly between groups. When a trial is not possible, researchers look for other clues. Does the effect appear in many studies and places? Does the cause come before the effect? Does more exposure lead to more effect? Is there a sensible way it could work?

The British researcher Austin Bradford Hill listed clues like these in 1965. He said no single clue proves cause. Taken together, they help scientists judge. His ideas are still taught in medical schools. Later in this unit, you will see how Hill used them to study smoking and lung cancer.

Not all evidence is equal. At the bottom are single stories, such as a friend saying a remedy worked. These are called anecdotes. They can spark ideas, but they cannot rule out chance, bias or the natural course of an illness. A story is a starting point for a question, not an answer.

Next come observational studies. Researchers watch groups of people without assigning treatments. They can follow thousands of people for years and find important links. But confounders can fool them. Famous examples include studies that follow nurses or doctors for decades to see who develops which illnesses.

Above those sit randomized controlled trials, where chance assigns treatments. Near the top are careful summaries of many good trials. Doctors often draw this ranking as a pyramid, called the hierarchy of evidence. Each step up usually means better protection against chance and bias.

The pyramid is a guide, not a strict rule. A large, well-done observational study can be more useful than a tiny, sloppy trial. And some questions can only be studied by observation. No one would randomize people to smoke for 30 years. Good judgment asks both what kind of study it is and how well it was done.

Suppose ten trials have tested the same medicine. Six show a small benefit. Four show none. Which should a doctor believe? Looking at only one study could mislead. The better approach is to gather and compare all of them. Each study might have used different people, doses or lengths of time.

A systematic review does this with a strict plan. Researchers search every database for studies on one question. They judge the quality of each study. Then they summarize what the best evidence shows. When they combine the numbers from many trials into one overall result, that is called a meta-analysis.

The Scottish doctor Archie Cochrane argued in the 1970s that medicine should use trial evidence more fully. He criticized doctors for not organizing what was already known. In 1993, researchers founded the Cochrane Collaboration, named after him. Now called Cochrane, it is an international network that publishes systematic reviews on health questions.

Reviews are only as good as the studies they include. If many trials are small or biased, the review will say so. Honest reviews often conclude that the evidence is uncertain and more research is needed. That is a useful answer too. Knowing what we do not know helps doctors and patients decide wisely.

You scroll past a headline: Chocolate Cuts Heart Disease Risk. It sounds great. But headlines are short, and they are written to be clicked. The study behind them may say something much more careful. A careful reader learns to look past the headline to the study itself.

Start with a few questions. Was the study in people, or in mice or cells? Results in mice often do not hold up in humans. Was it a randomized trial or an observational study? If observational, confounders may explain the link. How many people took part? Who paid for it?

Next, look at size. A headline may say a food cuts risk by half. That is relative risk. If the risk was 2 in 1,000 and fell to 1 in 1,000, the absolute risk dropped by only 1 in 1,000. Both numbers are true, but the absolute number shows how much difference it makes to one person.

Finally, see whether other studies agree, and check trusted sources such as public health agencies or a review. Talk with a doctor, nurse or pharmacist before changing medicines or trying supplements. One exciting study is a reason for curiosity, not a reason to change your life overnight.

The science of how diseases spread and affect groups of people is called epidemiology. Its first job is counting. Epidemiologists ask how many people are sick, where they live and when they got sick. Two counts matter most. The first is how many new cases appear. The second is how many people have a condition right now.

Incidence is the number of new cases in a group over a set time, such as new flu cases in Illinois during one week. It tells you how fast a disease is appearing. A sudden jump in incidence is often the first sign of an outbreak. Health departments watch incidence closely for this reason.

Prevalence is the number of people who have a condition at one point in time, whether they got it last week or ten years ago. Long-lasting conditions like diabetes can have high prevalence even if incidence is steady, because people live with them for years.

Think of a bathtub. Incidence is the water flowing in from the tap. Prevalence is the water already in the tub. Recovery and death are the drain. A good treatment that helps people live longer can actually raise prevalence, because fewer people leave through the drain.

An outbreak is more cases than expected in a limited place, such as a school or a town. In 1854, a cholera outbreak struck the Soho area of London. The doctor John Snow mapped the deaths and found they clustered around one water pump on Broad Street. His work helped show that cholera spreads through dirty water.

An epidemic is a bigger rise in cases across a larger area, such as a country. A pandemic is an epidemic that spreads across many countries and continents. The words describe how far a disease spreads, not how deadly it is. A mild disease can cause a pandemic, and a deadly one can stay a small outbreak.

The 1918 influenza pandemic spread around the world in waves. The CDC estimates that about 500 million people were infected, roughly a third of the world’s population then. At least 50 million died. There were no flu vaccines or antibiotics for the lung infections that often followed.

In 2020, the World Health Organization described COVID-19, caused by a new coronavirus, as a pandemic. It spread quickly through air travel and close contact. Scientists worldwide shared data and developed vaccines within about a year. Each pandemic has taught public health workers new lessons about preparing.

In June 1981, the CDC published a short report about five young men in Los Angeles with a rare lung infection. Their immune systems were not working well. Doctors soon saw more cases. The illness was named AIDS, short for acquired immunodeficiency syndrome.

By 1984, scientists had identified the cause: a virus later named HIV, the human immunodeficiency virus. HIV attacks cells of the immune system. It spreads through certain body fluids, such as blood. It does not spread through hugs, sharing a desk or touching a doorknob.

For years there was no effective treatment. Then, in the mid-1990s, doctors began combining several medicines at once. This antiretroviral therapy keeps the virus under control. It does not cure HIV. But many people who take it every day can live long, healthy lives and have children.

HIV became a worldwide epidemic, hitting parts of Africa especially hard. UNAIDS and the World Health Organization track it across the globe. Public health work on testing, treatment and prevention continues. Anyone with questions can talk privately with a doctor, nurse or school nurse.

Imagine a student is diagnosed with measles. Measles spreads very easily through the air. Public health workers call the family. They ask where the student has been and who they were near in the days before getting sick. This process is called contact tracing.

Each person identified is a contact. Workers reach out to the contacts, explain the risk and check whether they are protected, for example by vaccination. Some may be asked to stay home for a set time, called quarantine, to see if they get sick without spreading it further.

Contact tracing is old but powerful. It has been used for tuberculosis, sexually transmitted infections, Ebola and COVID-19. It works best when cases are caught early and people trust health workers enough to share information. Each new case that is caught early can prevent several more.

Tracing depends on privacy. Health workers do not tell contacts the name of the sick person. Protecting privacy helps people feel safe cooperating. That trust is one of public health’s most important tools. Contacts also get clear, calm advice about what to watch for and whom to call.

The Framingham Heart Study began in 1948 and still continues. It is a cohort study: researchers pick a group of people, called a cohort, and follow them over time. Because they measure people before anyone gets sick, they can see what came first. That order matters when you want to find causes.

Over years of checkups, patterns emerged. Volunteers with high blood pressure were more likely to have heart attacks and strokes. So were people with high cholesterol, a fatty substance in the blood, and people who smoked. Later findings added lack of exercise, diabetes and obesity.

Each of these is a risk factor: something that raises the chance of a disease. A risk factor is not a guarantee. Many smokers never have a heart attack, and some nonsmokers do. But across thousands of people, the pattern is clear. It raises the odds for a group, but it does not decide what happens to one person.

These findings changed medicine. Doctors began checking blood pressure and cholesterol at routine visits. They started talking with patients about smoking and exercise. Heart disease is still common, but U.S. death rates from it have fallen a great deal since the 1960s, thanks to prevention and better treatment.

In the early 1900s, lung cancer was rare. By the 1940s, doctors in Britain and the United States were seeing far more of it. Some blamed car exhaust or road tar. Two British researchers, Richard Doll and Austin Bradford Hill, set out to find the cause.

In 1950 they published a case-control study. They interviewed patients with lung cancer, the cases, and similar patients in the same hospitals without it, the controls. Nearly all the lung cancer patients smoked, and heavy smokers were most at risk. In the United States, Ernst Wynder and Evarts Graham reported similar findings the same year.

Doll and Hill then began a cohort study. From 1951 they followed tens of thousands of British doctors and their smoking habits. Smokers died of lung cancer far more often than nonsmokers. The more they smoked, the higher the risk. Doctors who quit lowered their risk.

The evidence met Hill’s clues for causation: strong, repeated, dose-related and with the cause coming first. In 1964, the U.S. Surgeon General’s report concluded that smoking causes lung cancer in men. Smoking rates in the United States have fallen greatly since then.

When a new illness appears, someone has to notice, count and respond. In the United States, that job falls largely to the CDC, the Centers for Disease Control and Prevention. It began in Atlanta in 1946, first working to control malaria in the South.

The CDC gathers reports from state and local health departments, such as the Illinois Department of Public Health and the Chicago Department of Public Health. This steady watching for disease is called surveillance. It helps spot outbreaks early. The CDC also studies injuries, chronic diseases and environmental health.

Diseases do not stop at borders. The WHO, the World Health Organization, was founded in 1948 as part of the United Nations. It helps countries share data, sets health guidelines and coordinates responses to global threats. April 7, the day its constitution took effect, is marked as World Health Day.

Neither agency can do everything alone. They depend on local nurses, doctors, labs and ordinary people who report illness and follow guidance. Public health is a team effort, from a school nurse to an international agency. Everyone who washes hands, stays home when sick or reports an illness is part of it.

Smallpox was a disease caused by a virus. It caused fever and a rash of blisters. For thousands of years it killed or scarred many millions of people. In 1796, the English doctor Edward Jenner showed that material from cowpox, a mild related disease, could protect people. This was an early vaccine.

A vaccine trains the immune system to recognize a germ without causing the full disease. When enough people in a community are protected, a disease has trouble spreading. This is called herd immunity. It also shields people who cannot be vaccinated, such as very young babies.

In 1967 the WHO began an intensified campaign to end smallpox. Vaccinating everyone everywhere was not possible. So teams used a strategy called ring vaccination: find each case, then vaccinate the people around it. Surveillance and contact tracing made this work.

The last natural case was in Somalia in 1977. In 1980 the WHO declared smallpox eradicated, gone from nature worldwide. It remains the only human disease to be eradicated. Health workers are working to do the same for polio. Polio cases have fallen by more than 99 percent since the late 1980s.

Two children live a few miles apart. One lives near parks and grocery stores with fresh food. The other lives near a busy highway, with few stores selling fresh food and older housing. Their doctors may be equally good. But their health chances may differ.

The conditions in which people are born, grow, live, work and age are called social determinants of health. They include income, education, housing, safe neighborhoods, clean air and water, and access to care. The WHO and the CDC both use this idea.

Researchers have found that life expectancy can differ by years between neighborhoods in the same city, including Chicago. These gaps are called health disparities. Their causes are debated. Most researchers point to several factors working together, such as income, pollution, access to care and past policies about where people could live.

Public health tries to improve conditions for everyone. Examples include removing lead paint from old homes, adding clinics where few exist, and improving school meals. Choices about how to do this, and who should pay, are often debated in politics. But the goal is shared: a healthy start for every person.

Many things we take for granted came from public health work. Clean tap water, sewers, seat belts, food inspections and smoke-free indoor spaces all prevent illness or injury for whole populations. Most people never notice the diseases that did not happen.

Some measures ask something of individuals. Examples include school vaccine requirements, smoking bans and, during COVID-19, rules about masks and closures. Supporters say these protect the vulnerable and the whole community. Critics say some go too far into personal choice, or cost too much for families and businesses.

These are real debates about values as well as facts. Evidence can show what a measure does. It cannot alone decide how to weigh safety against freedom or cost. Communities decide that through elected leaders, courts and public discussion. Knowing the difference between a fact question and a values question helps everyone discuss health calmly.

Good public health leaders explain their reasoning, share data openly and change guidance when new evidence arrives. That honesty builds public trust. When you hear health advice, you can ask what the evidence shows, who it protects and what it costs.