Picture a ward in a Swiss sanatorium, around 1925. A patient lies on a veranda in a deck chair, wrapped in two blankets; snow all around, ten below zero. Every two weeks, a doctor comes over, inserts a long needle between the ribs, and pumps two hundred milliliters of nitrogen into the chest cavity. The lung shrivels like a punctured ball and stops breathing. This is not torture. It is treatment. And it was the best medicine had — for thirty years, all over the world, for hundreds of thousands of people.
It was called artificial pneumothorax. The idea was “lung rest”: if a lung ravaged by tuberculosis did not have to expand with every breath, the cavities would stop being eaten away by movement, scarring would begin, and the person might survive.
Thomas Mann, who spent three weeks at a Davos sanatorium in 1912 (caring for his wife Katia, suspected of having tuberculosis), described these patients without the slightest pity: they “formed a group, because a thing like pneumothorax naturally brings people together. They call themselves the Half-Lung Club” (The Magic Mountain, 1924).
Let us put it this way: people with half-dead lungs formed a club. That may be the best metaphor for the entire history of medicine before antibiotics.
The protagonist of this story is Carlo Forlanini (1847–1918), and he is exactly the sort of man a screenwriter would invent if told: “Make an Italian doctor from the nineteenth century, but not too much.”
A Milanese child, the son of a physician. His mother died of pulmonary consumption — so the disease to which he would devote his life took someone from him first thing. He entered the University of Pavia, abandoned his studies in 1866 to volunteer with Garibaldi, fought at Monte Suello and Bezzecca, returned, befriended Camillo Golgi (that very same future Nobel laureate in neuroscience), and received his degree in 1870.
Then comes the important part. Forlanini did not begin with pneumothorax. He began with the opposite: attempts to treat lungs through forced ventilation, breathing exercises, and aerotherapy. Together with his brother Enrico — a brilliant engineer, one of the pioneers of aviation (Milan Linate Airport bears his name) — he built one of the first hyperbaric chambers. Literally: first he tried pumping up lungs, and then realized he had to do the reverse.
The turn came when he observed spontaneous pneumothorax in his patients. A lung ruptured, air entered the pleural cavity, the lung collapsed — and in some patients the tuberculous process afterward... improved. Forlanini read everything there was to read and understood: the point was not the air itself, but rest. Still tissue scars; moving tissue keeps falling apart.
In July 1882, he published the theory. He performed the first intervention on a human being in 1888. That is, six years between the idea and the first needle — with animals and cadavers in between, where he practiced the technique: how to enter with the needle so as to pierce the parietal pleura without piercing the visceral one.
Then comes the engineering, and honestly, this is my favorite part.
At first, Forlanini pumped in atmospheric air — and discovered that the oxygen was absorbed too quickly, the lung re-expanded, and the effect vanished in days. He needed an inert gas. Nitrogen. Which meant he needed an apparatus that produced nitrogen: pyrogallol greedily consumes oxygen from air — there is your nitrogen. A second device — for administration under pressure, with a manometer, needle, syringe. Two instruments, a chemical reaction, and pressure at the outlet.
He worked out the doses empirically and with great care: begin with 200 ml, never more than 400 (otherwise, subcutaneous emphysema — gas escapes into the tissues of the chest wall), and never less than 100. The first session: incomplete collapse. A few days later: a follow-up dose to achieve complete collapse. Then repeat every 15–20 days. And so on — for months, usually up to two years or longer.
Think about that schedule. A person with active tuberculosis, in the age before antibiotics, had to come in every two or three weeks for an injection into the pleural cavity. For two years. This was not treatment; it was a way of life. And people agreed because the alternative was death.
Forlanini’s scientific fate is a separate sad comedy.
The idea of pneumothorax had been in the air long before him. In 1696, the Italian physician Giorgio Baglivi described improvement in a tuberculosis patient after a knife wound to the chest. In 1822, the Scotsman James Carson of Liverpool (the very man who figured out the mechanics of breathing and the elasticity of lungs) tried to induce pneumothorax in rabbits, then in humans — and failed: both patients had extensive pleural adhesions, and the lung would not collapse. The technique had been worked out, the idea was right, but no instruments existed to overcome the adhesions. The subject died for nearly sixty years.
Forlanini, incidentally, did not know Carson’s work — it resurfaced only in 1909. He arrived at the same conclusion independently, from observing spontaneous pneumothoraces. A classic case: the right idea invented twice because the literature did not travel.
Then came 1894, the XI International Medical Congress in Rome. Forlanini presented his results. And silence. The Italian medical community responded sluggishly, at times with hostility; meanwhile, the end of the nineteenth century was enthralled by bacteriology and vaccines — Koch had already found the pathogen, everyone was looking for an inoculation, and here was some Italian with a needle and a bag of gas.
Then, in 1906, Forlanini opened a German journal and read an article by Ludolph Brauer of Marburg that attributed the invention of the method to... the American surgeon John Benjamin Murphy. Murphy had indeed developed his own version — apparently without knowing the Italian’s work — and described it in 1898 at an AMA meeting. Murphy, to give him his due, was also the first to use X-rays (discovered in 1895) to monitor the size of the pneumothorax.
Forlanini responded immediately — in another authoritative German journal, with results from 25 cases and a claim of priority. He also took a swipe at Murphy’s technique: the American made an incision through skin and muscle rather than a puncture, and introduced 1000–3000 ml of nitrogen, compressing the lung with excessive force.
The response worked, but with a delay of a quarter-century. Recognition came in April 1912, at the VII International Congress on Tuberculosis in Rome, where Forlanini delivered a lecture titled “Artificial Pneumothorax in the Treatment of Pulmonary Consumption.” A standing ovation. Thirty years from the first publication to the moment the hall rose to its feet.
This is where the part begins that made me dig into this story in the first place.
After 1912, Forlanini was nominated for the Nobel Prize in Medicine at least twenty times. Camillo Golgi, his friend and the 1906 Nobel laureate, nominated him three times, calling it “a great benefit to humanity” — almost a word-for-word quotation from Nobel’s will.
He did not receive the prize. And the reasons are a textbook in how not science but scientific politics works:
And now for the most maddening part. The Nobel Prize was not awarded in 1915–1918 — the war. Forlanini died on May 25, 1918. He had a window of exactly three years: 1912, 1913, 1914. In 1913 and 1914, he was considered one of the leading candidates — the committee saw enormous potential in pneumothorax. But the jury ultimately decided that the method was “insufficiently original and brilliant,” and its high risk to the patient was an important argument against it.
He died knowing that his method was being used in clinics around the world, and not knowing that twenty-five years later pills would appear that made it unnecessary.
The moral I take from this: a procedure that genuinely saved people did not receive the prize because it was dangerous. A hundred years later, it is all the same, except now we call it “risk-benefit balance” and write guidelines about it.
There is a detailed clinical account in Applied Sciences (2020), and what makes it good is that it has not been tidied up.
Forlanini recorded: coughing diminishes and disappears, sputum diminishes and disappears, fever falls, sweating goes away, asthenia goes away. The patient “becomes euphoric,” gains weight, and the toxic-infectious condition recedes. The improvement is not merely local: the inflammatory focus shrinks and, possibly, the general immune response improves.
So the procedure did not work mechanically alone. It gave people something they had not had for years: the feeling that something was happening and that they were getting better.
And now — the complications, in the same voice. Empyema. Emphysema. Cerebral embolism, including gaseous embolism — an air bubble in the brain. And “pleural eclampsia” — a fainting episode Forlanini described himself, and described so that it is frightening to read: “the patient suddenly falls, exhausted, onto the bed, unconscious, his head inclined to one side or thrown back; his eyes wide open, the eyeballs motionless or rolled upward; the pupil maximally dilated, rigid; the face extremely pale, at times cyanotic, cold, with bluish streaks on the neck, chest, and limbs; teeth clenched; trunk and limbs tense; breathing absent, pulse imperceptible; — all in all, an impressive syndrome, terrifying in its suddenness and in the thought that the attack follows a medical act and may prove fatal.”
“May prove fatal.” The doctor who invented the procedure wrote that himself about his own procedure. To me, that is a much more honest document than any modern press release.
Then doctors began fixing the problems created by their own method. Adhesions kept the lung from collapsing — the Swede Hans Christian Jacobaeus came up with inserting a thoracoscope and cutting the adhesions from within by cauterization. 1906–1907: the Dane Christian Saugman of Vejlefjord Sanatorium built a water manometer into the apparatus for precise pressure control. 1912: Maurizio Ascoli proposed partial collapse so that bilateral pneumothorax could be performed. 1913: Ernst Sauerbruch introduced phrenicotomy — severing the phrenic nerve to limit the diaphragm’s movements.
And in parallel — an entire staircase of surgery that today reads like a list of horrors: 1891 — Théodore Tuffier removed the apex of a lung in cavernous tuberculosis. 1893 — David Lawson proposed pneumonectomy. Two years later, William Macewen removed a left lung because of tuberculosis (the patient was alive in 1940 — honestly, that alone is a miracle). 1908 — Ludolph Brauer published the first radical thoracoplasty. And finally — plombage: a cavity is created in the chest wall and filled with inert material. Fat, paraffin, rubber balloons, oil, methyl methacrylate balls (Lucite). In 2022, Respirology Case Reports described an 86-year-old man treated 68 years earlier: on X-ray, a large opacity with multiple ring-shaped shadows; on CT, extrapulmonary spheres compressing the upper lobe. Lucite balls. Still inside. The man is alive.
Before going further with pneumothorax, we need to sort out what surrounded it.
The sanatorium movement began in 1854 in Görbersdorf (Silesia, 650 meters above sea level), and — pay attention — with an utterly wrong theory. Hermann Brehmer believed tuberculosis arose because the heart had difficulty properly supplying the lungs with blood, and therefore one had to go higher up, where atmospheric pressure was lower and the heart had an easier time. Hence: mountains, clean air, rest, food. Legend has it that Brehmer cured his own tuberculosis in the Himalayas — historians have found no confirmation that he ever traveled to India, but the legend stuck because it made sense.
By 1877, sanatoriums had spread across Europe. Edward Livingston Trudeau founded his in the Adirondacks.
And now — the reason I started this whole conversation. I am reading the BBC archive, an interview with a former head porter at the Joint Cheshire Sanatorium in Loggerheads (Staffordshire, 300 beds, the 1930s–40s, Dr. Peter Edwards’s regime: “fresh air and rest,” at least 18 months, often longer):
“On a good fresh day we used to wheel all the patients’ beds out into the open air so they could get fresh air into their lungs. We’d line them up in nice straight rows.”
Ted Parton, that very porter, adds: that day in Staffordshire, snow was on the way, and the temperature was “let’s say, bracing.” And then:
“Of course, they were wheeled out in frost and snow, and snow would blow onto their beds — but it was thought to be good for them.”
And one more thing: “We also put sandbags on their chests while they were lying there.” Patients could spend hours outdoors with sandbags strapped to their chests. The aim was to “give the lungs something to fight against, to increase their strength and breathing power.”
Think about it. A person with tuberculosis — that is, necrotizing tissue in their lungs, coughing blood, night sweats, fifteen kilos down — lies in the frost under snow, with sand laid on their chest so they can breathe “more deeply.” The grounds were planted with pines — Dr. Edwards believed pine trees purified the air.
This is what the real history of evidence-based medicine looks like: not malice, but sincere conviction. All of it was done by people who wanted to help, and they did it on the scale of hundreds of thousands.
And here comes the most important thread in the whole text.
In 1948, Britain published the results of a trial that would later be called the first randomized controlled trial in medical history — not of a procedure, but of a drug. Streptomycin against pulmonary tuberculosis; the Medical Research Council; statistician Austin Bradford Hill. Patients were assigned to groups using random-number tables and sealed envelopes. The idea was not only scientific but ethical: postwar Britain had a catastrophic shortage of streptomycin (the government was able to purchase only 50 kg in the United States), and random allocation spared doctors the monstrous choice of who should receive the drug and who should not.
Now look at the tuberculosis mortality graph for England and Wales from the mid-nineteenth century onward. The main decline happened long before streptomycin. Thomas McKeown showed in the 1970s that TB mortality had fallen by 90–95% before antibiotics and vaccination became widespread. And within the 1948–1971 period, streptomycin reduced mortality by 51% — that is, from an already low baseline. Over the entire period since causes of death began to be recorded, he estimated the contribution of antibiotics at 3.2%.
McKeown himself explained this through improved nutrition, living standards, and hygiene. He was immediately attacked — and rightly so. Historian Simon Szreter showed that it was not so much about food as sanitation measures, clean water, and public health. Vanessa Heggie in The Guardian (2013) added a third, deeply inconvenient correction: diagnosis changed. In 1850, tuberculosis was identified by symptoms — night sweats, menstrual irregularities, coughing. From the 1820s onward, doctors listened with stethoscopes. From the 1880s onward — bacteriology and immunological tests. Strictly speaking, these are different diseases under one name, and piling them into a single statistic is a methodological error.
But whatever the correct interpretation may be, there is one fact that cannot be pried out of this story: sanatoriums did not close because someone proved they were useless. Antibiotics swept them away. And there is a separate irony in that. The Madras study (Tubercle, 1959, India) compared home treatment and sanatorium treatment: with drugs available, the outcomes were comparable. So the sanatorium as an institution was never tested — it was bypassed.
No proper RCT of pneumothorax was conducted in the 1920s either. There were observational series: 191 patients in Denmark (1925–1931), 557 patients at Trudeau Sanatorium in Saranac Lake (1930–1939). And there were articles like “A Survey of Artificial Pneumothorax in Representative American Tuberculosis Sanatoria, 1915–1930” — that is, retrospective work from archives. All of this is called “case series” and, by modern standards, counts as evidence of the lowest grade.
Doctors in the 1930s did not know that. They had neither the methodology nor the word “evidence” in its present-day meaning. They had patients who were dying, and a procedure after which some of them felt better.
Literature here is not decoration — it gives us what the papers do not: subjective experience.
Thomas Mann described the sanatorium from the inside and without illusions: a closed little world on a mountain, where illness becomes social status, where everyone has their own thermometer and their own temperature, where people live for years in suspended animation. The “Half-Lung Club” is not just about pneumothorax; it is about an entire subculture of people whose main occupation is being ill.
A. J. Cronin, in The Citadel (1937), made pneumothorax central to the plot. The young doctor Andrew Manson takes a patient dying of tuberculosis to the American specialist Richard Stillman, who has no British diploma — and Stillman induces a pneumothorax with nitrogen. The patient recovers. Then Manson’s enemies in London seize on Stillman’s lack of British credentials and report Manson to the General Medical Council. The climax of the novel is Manson’s speech at the hearing, where he defends scientific progress “from whatever source” and attacks the closed nature of the profession.
Notice what is happening here. Innovation comes from a man without formal credentials. The profession responds not by testing the method, but with disciplinary proceedings. This is not literary fantasy — it is exactly what happened to Forlanini, when his method was first ignored, then stripped of its priority, then denied the prize twenty times.
George Orwell is a separate document — the cruelest of all.
December 1947. Orwell is on Jura, has a cold, is coughing blood. A specialist summoned from Glasgow says: seriously ill. A week before Christmas, he is in Hairmyres Hospital. The diagnosis: tuberculosis. He is 44.
The treatment of the day: collapse therapy, “rest” for the lung, painful surgical procedures, vitamins, fresh air, bed rest. They confiscated his typewriter and told him to stop working — but, as the authors of an analysis in J R Soc Med (2006) note, nobody thought to advise him to stop smoking.
Streptomycin was unavailable in Britain: 50 kg for the entire country; patients and doctors “besieged” the government, the BBC broadcast urgent appeals, a black market emerged. Orwell was not eligible for any of the trials (Scottish hospitals had none at all, and he was also too old under the criteria). Then his doctor suggested using connections. Orwell wrote to David Astor. Astor contacted Health Minister Aneurin Bevan (Orwell’s former editor at Tribune), secured a license, and the money came from an American account containing royalties from Animal Farm. Within a few weeks, the drug was there. Orwell became the first person in Scotland to be offered streptomycin.
Then come his own notes. I cannot tell it better:
“At first, although the streptomycin seemed to produce an almost immediate improvement, there were no side-symptoms except some discolouration at the bases of the finger and toe nails. Then the face became noticeably redder, the skin began to peel, and a rash appeared all over the body, particularly the back. There was no itching. After about three weeks a bad sore throat began, which did not go away and did not respond to penicillin lozenges. It was very painful to swallow, and for several weeks I had to have a special diet. Now ulcers with blisters have appeared in the throat and on the inside of the cheeks, and blood keeps appearing in small bubbles on the lips. At night they burst and bleed freely, so that in the morning the lips are always stuck together with blood, and I have to soak them before I can open my mouth. Meanwhile the nails have disintegrated at the roots, and the disintegration has, as it were, gone up the nail, while new ones were growing beneath. The hair has begun to fall out, and on the back of the head there are one or two patches of completely white hair (previously there was only grey).”
After 50 days, streptomycin was stopped. The lips and ulcers healed almost immediately, the rash went away, the hair stopped falling out. The nails did not recover: “They are much more ribbed than before, and much thinner, with a constant tendency to split unless I keep them very short.”
Orwell gave the remaining drug to the hospital — it was administered to two doctors’ wives, and both recovered. He himself wrote about it in a letter to Julian Symons: “I am much better, but had a bad couple of weeks with the side effects of streptomycin. I suppose with all these drugs it is rather a case of sinking the ship to get rid of the rats.”
Then came the sanatorium at Cranham, Gloucestershire — small wooden huts in a remote corner of the Cotswolds. Then PAS, which made him nauseous and did not help. He died in January 1950, aged 46, seven months after the publication of Nineteen Eighty-Four.
The twentieth century in a single plot: the man who wrote the most famous novel about total control gained access to the first real medicine not by the rules but through personal connections — via a publisher, a minister, and an American currency account funded by his own book.
It seems to me this story invites two equally stupid morals.
Moral one, the popular version: “See, old medicine was barbarism, and now we have science.” This is untrue in one important sense. Barbarism was not the alternative to science; it was an early stage of science. Forlanini was not a charlatan — he was a man who noticed a pattern, built an apparatus, practiced the technique on cadavers, described the complications of his own procedure, and spent thirty years trying to get his colleagues to accept it. That is exactly what a good research engineer does. He simply did not have RCTs, because RCTs had not yet been invented.
Moral two, also popular: “Look, sanatoriums and pneumothorax achieved nothing; antibiotics solved everything.” Also incomplete. Pneumothorax did achieve something — at the very least, some patients genuinely went into remission. The problem is that we will never know what proportion, because there were no control groups. And sanatoriums did not provide medicine but isolation: a person with open tuberculosis was removed from their family and from shared housing. From a public-health perspective, that worked. From the standpoint of treating the individual person — the question remains open to this day.
And what is truly horrifying is the deployment failure. The idea is right, the technique refined, but the method spreads around the world, and doctors who do not know the technique take it up, selecting patients by arbitrary criteria. The number of severe and fatal complications rises sharply. The authors of the 2020 review say this outright: “With the spread of the operation and its application by physicians who did not know the method very well, and who used inadequate criteria for the selection of candidate patients, the number of serious and even fatal complications greatly increased.”
This is literally the story of any modern tool rolled out to production without documentation and with arbitrary access permissions. The same pattern. The same consequences.
In Stalinist Russia in the 1930s, pneumothorax was declared an “aristocratic therapy,” meaning counterrevolutionary, and propaganda called the apparatus itself a “killing machine.” One of the method’s leading supporters, Wolf S. Holzman, director of the Moscow Regional Institute of Tuberculosis, was arrested in 1939 and shot two years later.
A medical procedure became a political marker. Not because anyone seriously believed that gas in the pleura was class-alien, but because the method required an individualized approach, expensive equipment, and a trained doctor — precisely everything that fit poorly into the rhetoric of equal access.
Meanwhile, in occupied Poland, SS-Obergruppenführer Wilhelm Koppe organized the extermination of more than 30,000 Polish tuberculosis patients — “knowing little or caring less that a cure was almost at hand.” Streptomycin was isolated in 1943–44.
This is not a curiosity. It is the background against which this entire story unfolded. While doctors argued over priority and manometers, people with the same diagnosis were being murdered for political reasons.
Streptomycin was isolated in 1943 (Waksman, Schatz, Bugie), released onto the market in 1946, and subjected to the first RCT in 1948. PAS appeared in 1946 and began to be used in combination so resistance would not develop. In 1952 came isoniazid — the first oral bactericidal drug. The real revolution happened not with streptomycin, but with the combination.
Pneumothorax, meanwhile, did not vanish overnight. It was used into the 1970s — in cases where large cavities made the patient especially vulnerable to stagnant infection. The conclusion of the 2020 review even credits Forlanini with the fact that the method did not die completely: it is still used in special cases, for example in pre-thoracoscopic artificial pneumothorax. I checked contemporary guidelines — yes, “creation of artificial pneumothorax” appears in consensus recommendations for medical thoracoscopy as a standard part of the procedure, one that can itself cause hypoventilation because of sedation and lateral positioning.
So in the 2020s, a surgeon introduces gas into the pleural cavity to gain access to the chest. Technically — Forlanini, only the purpose is different. Pure irony: the method outlived its disease and changed professions.
And now, today. According to the WHO, 1.23 million people died of tuberculosis in 2024, and 10.7 million fell ill. TB remains the world’s leading cause of death from a single infectious agent. Around a quarter of the planet’s population is infected, and 5–10% of them will develop the disease. Multidrug-resistant TB is a “public health crisis”: only around 2 out of 5 people with drug-resistant disease received treatment in 2024.
We traveled from sandbags to isoniazid, and still have 1.23 million deaths a year from a disease Wilhelm Koppe used as a pretext for mass murder, and Carlo Forlanini tried to treat with a needle and nitrogen.
Three things I consider honest conclusions rather than pretty phrases.
First: efficacy and correctness are different questions. Pneumothorax most likely helped some patients. Nobody knows what proportion, because there were no controls. We still work this way in places where control is impossible or unethical: interventions in critical situations, rare diseases, surgery. The only difference is that now we can count and say honestly: “We do not know.”
Second: the danger of a method does not negate its benefit, but it demands documentation. Forlanini himself described “pleural eclampsia” in all its horrifying detail. This is not masochism; it is engineering: if you have not described the failure mode, you are not an engineer but a preacher. His method ultimately did not bring him the Nobel Prize, in part because of its risks. And that was probably right — but the price of that correctness was that he died without seeing recognition.
Third: the most dangerous thing in medicine is not charlatanism, but mass good-faith error. The people who wheeled beds out into snow and placed sandbags on patients’ chests were not villains. They were certain. And they had no way to test their certainty. Which is why the first thing to look for in any new technique is not “does it work?” but “how will we know that it works?”
Forlanini died in 1918, twenty-five years before his method became unnecessary. Thomas Mann died in 1955, having outlived both pneumothorax and streptomycin. Orwell died in 1950, two years after he was lucky enough to receive the drug outside the rules — and it nearly killed him with side effects.
Medicine is not a history of victories. It is the history of a very long series of well-documented attempts, some of which did not work.
Special thanks to the BBC archive for the interview with the porter — it is more honest than half the medical papers of that era.