Introduction

Prior to the introduction of prontosil and then penicillin, childbed, or puerperal, fever was the scourge of childbirth, leaving many babies without mothers ^1,2. Today, puerperal sepsis still kills some100,000 women a year worldwide. The disease is now known to be caused by haemolytic streptococci, but long before the microbial theory of infection was confirmed, the Hungarian gynaecologist, Ignaz Philipp Semmelweis (1818–1865), who practised in Vienna, showed that the incidence of puerperal fever could be reduced by rigorous hand washing ^3,4,5. Semmelweis was largely ignored or ridiculed by his contemporaries and died of a staphylococcal infection in an asylum for the insane. Although Semmelweis is generally regarded as the first to demonstrate the infective nature of puerperal fever, as we shall see, a number of physicians came to this conclusion before him². The aim of this review is to highlight the contributions of these, largely forgotten pioneers and to show that the infectious nature of puerperal fever was long –recognized before the work of Semmelweis.

A few historians have always aired doubts about Semmelweis’ priority. In 1905, C. J. Cullingworth²,⁶ for example stated that, ‘We English-speaking people on both sides of the Atlantic, while giving abundant honour to Semmelweis, have been in danger of forgetting the earlier and equally remarkable contributions to our knowledge of puerperal fever.’ In 1936, J. P. Greenhill also pointed out in that the contagiousness of puerperal fever had been recognized long before Semmelweis even thought about the disease ². One such pioneer in this field was the Manchester (England) based physician Charles White, who even as early as the late 1700s, showed that the occurrence of childbed fever could be drastically reduced isolating its victims and insisting on strict cleanliness⁷. By 1795, Alexander Gordon, of Aberdeen, had arrived at the radical and depressing conclusion that, like many other doctors, he had accidentally spread the disease and had caused the death of a number of women in his care ^8, In 1831, a certain Dr Campbell in the London Medical Gazette suggested⁹ that puerperal fever can be spread from cadavers by doctors attending post mortem examinations, while in 1835, Joseph Comstock of Lebanon, Connecticut ¹⁰ concluded that:

“The disease (puerperal fever) is communicatable from one patient to another cannot be well disputed, and therefore becomes the duty of the physician and nurse to take all precautionary means to prevent the dissemination of the miasms, by changing their garments frequently washing their hands etc., as well as by the employment of fumigants and other well known means. “  Comstock also noted the clear link between puerperal fever and erysipelas. Finally Loudon ¹¹, one of the limited number of critics of the view that Semmelweis’ has priority of demonstrating the cause of puerperal fever, pointed out that the mode of transmission of puerperal fever and its relation to erysipelas was confirmed by the English doctors, John Armstrong, in 1814, William Hey of Leeds, in 1825, and John Robertson of Manchester, in 1831. In 1831¹² , the above mentioned, Dr Campbell made the important point that puerperal fever could be spread by physicians from the dissecting room, but then clouded the issue by stating that it could not be passed from doctor to patient by any other means:

“On the question of infection I am as much as ever impressed by the belief, that unless the practitioner has been engaged in the dissection of bodies of those who have fallen victims, the disease cannot be conveyed by him from females labouring under it to others recently delivered, but if he have been so engaged I have strong reason for believing that he may be the means of propagating it.”

In the same year², John Robertson of Manchester concluded

“That besides being infectious-that is capable of being conveyed in a tangible medium from one woman to another-this disease is propagated by some cause of a more general kind, probably existing in the atmosphere after the fever has prevailed for some time in a locality,(original authors’ italics) cannot be doubted.”

Semmelweis’ Contribution

Semmelweis’ work on puerperal fever was begun in 1846 and first published in 1847 ^13,14,15. He became aware of the disease following a change in hospital practice in the Vienna hospitals where he worked and which had had originally followed the methods of Charles White. These had produced a very low death rate from puerperal fever, but when they were abandoned and cadavers were once again used to demonstrate midwifery techniques, a dramatic increase in mortality resulted. Semmelweis came to the conclusion that childbed fever was spread by a poison found in dead flesh (the so-called ‘cadaveric principle’), but was unaware of the possibility of the involvement of germs until 1861, when the germ theory had been established. Originally, he boldly stated that, ‘Puerperal fever is not a contagious disease, but it is conveyable from a sick to a sound puerpera by means of decomposed organic matter.’ By insisting that the cadaveric principle alone caused childbed fever, Semmelweis invoked the wrath of his critics. For example, the Dublin-based physician, John Denham², exasperatedly pointed out (in 1862) that Semmelweis overlooked the fact that childbed fever often appeared in towns which had no lying-in hospitals or dissecting rooms, as well as in rural districts where medical practitioners were seldom called upon to intervene in childbirth. As we have seen, many of Semmelweis’, forebears and contemporary critics were in fact well aware of the link between puerperal fever and both erysipelas and scarlet fever. Such observations can now be readily explained by the fact that puerperal fever is caused by the haemolytic streptococci which are spread on unwashed hands and on the breath of anyone carrying Streptococcus pyogenes.

Oliver Wendell Holmes’ contribution

In 1843 Holmes wrote an article which appeared in the NewEngland Journal of Medicine and Surgery entitled ‘The Contagiousness of Puerperal Fever’ ^16.17. In this he agreed with White and Gordon that the disease was both contagious and was often transmitted, via an unknown agent, by both physicians and nurses. He also described how, in 1835, an un-named doctor had the ‘good sense to change his clothes after each maternity-related visit and wash his hands in chloride of lime’ – a practice almost invariably credited in accounts on childbed fever, to Semmelweis (in 1848). On the basis of his own observations and those of others, Holmes recommended that physicians should: 1) never attend autopsies prior to examining a pregnant woman 2) always change every article of dress after attending a delivery 3) allow 24 hours or more to elapse before conducting any further midwifery, and 4) always leave a period of at least a month between attending a case of puerperal fever and any uninfected patients. Holmes stated that on, finding three or more closely connected cases of puerperal fever in the practice (with no others existing in the neighbourhood), the physician must assume that they themselves are the prima facie vehicle of the infection; and finally that physicians should regard it as their duty to pass on these warnings to nurses and assistants. Like Semmelweis, Holmes was ridiculed for such views which it should be emphasized, were published in 1843, some three years before Semmelweis began his work.

Robert Storrs-a neglected pioneer

An English country doctor, Robert Storrs, recognised the means by which puerperal fever is spread, some years before Semmelweis even began his studies^2,18. Little is known about Robert Storrs other than he lived in the small Yorkshire village of Sprotbrough, in England and practised in the nearby market town of Doncaster. In the early 1840s, when Storrs did his pioneering work, Doncaster was a quiet market town that had yet to enjoy the boom times brought by the arrival of the railways. From parish records we discover that Storrs was born in Sprotbrough in 1801 and died there in 1847. He was not however, simply a lowly village doctor, but was instead Honorary Surgeon to the Doncaster Dispensary and Medical Officer to the towns Poor Law Union (i.e. workhouse) and in 1824; he was made a Fellow of the Royal college of Surgeons. In medical terms however, the Doncaster that Storrs knew was well off the beaten track and far from the medical power bases of London and Edinburgh, a fact that doubtless explains why he is largely unknown and why his work had less of an impact than it deserved.

By the early 1840s, despite White’s work, the cause of childbed fever remained a mystery to the medical profession; a situation that in April of 1842, Storrs was about to change. On the twenty-third of that month his findings were first published in the Provincial State Medical Journal¹⁸ (the forerunner of the British Medical Journal). The article was then published in the American Journal of Medical Sciences in the following year. In his report, Storrs gives numerous accounts of his tribulations with childbed fever. He states, for example, that on January the seventh, 1841 he attended a Mrs Downes, a washerwoman who was in labour with her tenth child. On the morning of the ninth, thirty-six hours after the delivery of a child she developed rigours with severe abdominal pain, finally succumbing to the disease on the morning on the twelfth. Storrs then described another eight cases of the fever, most of which proved fatal. By the time that a certain Mrs Williams died on February the twenty fourth, it dawned on him that he, the doctor—the man who should be saving the lives of the women in his charge – was in fact spreading the disease!   When he next attended to a further sixteen cases, he took the precaution of changing all his clothes and applied “every means I could think of to prevent the spread of the fever, including “thorough ablutions” (i.e. hand washing). These measures helped reduce, but not eliminate, the disease. From these observations, Storrs concluded that childbed fever must be epidemic in nature and was being spread, not only by himself, but also by other doctors in the district. To avoid his unwitting role in the epidemic, Storrs next took leave of his practise to visit his friends on the borders of Wales. He now believed that the poison somehow “clung to him personally” and in Wales he hoped that the disease would literally be blown off him. On his return, he found that the number of deaths had declined, but still some women were tragically succumbing to childbed fever. Storrs then happened to meet two practitioners from nearby Sheffield, Dr Thompson and Mr Reedall, who had arrived at the same conclusion at about the same time as he had, i.e. that childbed fever was infectious and spread by doctors and midwives. A doctor in nearby Manchester, called John Roberton also came to the same conclusion around this time, while R.Yates Ackerley, a member of the so- called ‘English Contagionists’, had even earlier linked childbed fever and erysipelas; this view was also advocated by the Thomas Nunneley a notable physician from Leeds, famous for his experiments with anaesthetics; Storrs however, went further than his colleagues when he stressed the need for thorough ablution and general hygiene. Clearly, by the 1840s a group of doctors in the Doncaster, Sheffield and Leeds area of England had come to the conclusion that erysipelas and puerperal fever were one and the same disease Storrs soon came to an amazing and worrying realisation. He recognised that before attending the women in his original cases, he been called to a man suffering from erysipelas where he had handled a number of the patient’s weeping abscesses. Realizing that these were the source of the infection, he immediately refused to attend his women patients, leaving them instead with another doctor. On resuming his visits he again changed his clothes between patients and again made ‘“every possible ablution.” Such however, was the scale of the epidemic that Storrs had to continue treating woman despite in his own words having caused “such a great amount of misery.” From his studies, Storrs realised that childbed fever is generally, but not always, associated with erysipelas. He concluded his work with the following statement:

“I believe it (childbed fever) may also readily be propagated from one individual to another, so that if a person be unfortunate to attend a case of the kind, without great precaution, he may be liable to transmit it to others” and that “In many cases there is good reason to believe that it arises from attendance at post mortems, especially where there has been peritoneal inflammation.”

Storrs went on to extend his ideas on contagion to suggest that other fevers might be spread, not by miasma, but by person to person contact¹⁹. As to the possibility of a cure for puerperal fever, Storrs was very pessimistic; noting that everything he tried had little effect on the outcome; when a woman was infected then the likely outcome was death. He found bloodletting to be of absolutely no use, although at one time, he was encouraged to believe that calomel might be a useful treatment; but to no avail.

Let us try and put Storrs’ work in context? Firstly why is his work groundbreaking? Well, his paper is dated 1842, that is before either Semmelweis or Oliver Wendell Holmes reported on childbed fever. While Holmes was clearly aware of Storrs’ work, Semmelweis probably knew nothing of it when he came to publish his own studies in 1846. The importance of Storrs work can be summarised as follows: Firstly, he recognised that doctors passed on puerperal fever and that its spread could be prevented by removing all soiled clothing and by thorough hand washing. Secondly, he recognised that the disease was carried to mothers to be by doctors who had previously attended autopsies and had failed to wash their hands or changed their clothing. Storrs, in fact, recognised everything that Semmelweis is usually credited with having discovered! To his credit, Semmelweis did back up his observations with statistical evidence in lying in hospitals (data that was unavailable to Storrs), even though many of his obtained similar statistics to come to the opposite conclusion. There is however, a particularly important way in which Storrs work excelled over that of Semmelweis. The Hungarian doctor believed that the childbed fever was actually spread by the dead flesh (the so-called cadarveric principle) carried from autopsies to the pregnant woman. Storrs however, recognised that the disease was epidemic in nature and was generally, but not exclusively, associated with erysipelas, a disease which is caused by the bacterium Streptococcus the same bacterium that we now know causes childbed fever. As a result, Storrs confirmed the important etiological connection between these two diseases, a connection that Semmelweis missed. By 1865 the notably English obstetrician Robert Barnes confidently asserted that puerperal fever was associated with scarlet fever and erysipelas². He also criticised Semmelweis for believing that “puerperal fever, without exception, is a form of absorption—that is it arises from the generation of decomposed organic matter.” Many of Semmelweis’ contemporaries also noted that they could attend women after doing autopsies without causing ill effects. The Russian obstetrician Hugenberger noted that cadarveric fever is the rarest cause of childbed fever and that an epidemic ranged through Prague in 1849, even though all staff were forbidden to touch dead bodies. Even chlorine hand- washing he claimed, often failed to measurably alter the outcome of epidemic childbed fever (Barnes, 1865).

How can these observations be explained? Well, the situation regarding the spread of puerperal fever is obviously more complex than is generally portrayed. Puerperal fever is caused by Streptococcus that can be carried on hands; hand washing does therefore reduce the incidence of the disease. However, childbed fever is also spread from the mouth, particularly during epidemics (that is why face masks were later used by doctors and midwives).

Semmelweis obtained his data outside an epidemic, when the most of the cases would have been infected by unwashed hands, rather than from the throats of infected doctors and nurses, or merely anyone carrying streptococci. As a result, Semmelweis’ (by no means novel) suggestion that hand washing would always reduce the incidence of the childbed fever proved incorrect. Storrs in contrast, realised that the disease was not only spread by the unwashed hands of doctors, but also that puerperal fever was related to other epidemic diseases, notably erysipelas and scarlet fever (diseases both also caused by Streptococci).

In 1845, Storrs showed that husbands, children and doctors, in fact anyone in attendance, could acquire a generalised fever or erysipelas from childbed fever victims, proving that the diseases were one and the same. In the following year, the Edinburgh- based physician, Alexander Peddie suggested that childbed fever was caused by a “virus of animal nature” that originates from erysipelas inflammation².                

Although Storrs’ work was highlighted in the major medical journals, both here and in the States, and was also mentioned by Oliver Wendell Holmes, it seems to have made no impact on medicine of the time; subsequently, even English experts were content to give priority on his findings to Semmelweis. As a result, Storrs died largely unknown and unheralded on the fourteenth of September 1847. Like Semmelweis and many of the women he tried to help, Storrs died of fever and was buried in the churchyard of his birth in Sprotbrough, where there is a very simple, but apt, inscription on his gravestone which reads:“the memory of the just is blessed”.

Conclusion

The idea that Semmelweis originated, and was first to demonstrate the view idea that puerperal fever is a) infective, b) that it can be transmitted by hand to patients and c) that it can be prevented by hand washing is so ingrained in the minds of medical historians and the writers of popular accounts of the history of medicine that it seems almost impossible to dislodge even in the face of large amounts of evidence to the contrary ^20,21,22,23,. This review together with other literature sources has provided this evidence in abundance, yet the myth continues to be perpetrated that Semmelweis was the first person to show that disease, both in the particular and in general, could be spread by hand and that hand washing could prevent this spread. The reason why this myth continues to be peddled is simple-it provides a wonderful story, a kind of moral tale that is simply too good to through away in the face of historical reality. While such devotion to storytelling is almost understandable in the case of popular science writers it is completely unacceptable that it should be continued by historians. Hopefully, this uncritical telling of the Semmelweis myth will now end in the face of such overwhelming evidence which clearly denies Semmelweis such a critical role in the history of puerperal fever.

References

1. Colebrook, L. The story of puerperal fever-1800-1850. Br Med J.,1956; 1: 247-252.
2. Wainwright, M. Childbed fever-the Semmelweiss myth. Microbiolo. Today 2001; 28:173- 174.
3. Ricci, J.V. The Genealogy of Gynecology: History of the Development of Gynaecology through the Ages. Philadelphia, the Blakiston Company, 1943; pp1800-1900.
4. Raju, T.N. Ignaz Semmelweis and the etiology of fetal and neonatal sepsis. J Perinatol., 1999; 19: 307-310.
5. Cullingworth, C.J. Oliver Wendell Holmes and the contagiousness of puerperal fever. B.r Med. J., 1905; 2:1161–1167.
6. Burgess, A. H. Charles White of Manchester. Lancet 1941;1: 235–240.
7. Gordon, A. Puerperal sepsis and modern theories of infection control-Semmelweis in perspective. Lancet Infect Dis., 2010; 10: 275-278.
8. Kneeland, S. Is Puerperal fever Infectious? Dublin Med. Pr., 1846 15; 325-328.
9. Comstock, J. Puerperal fever. Amer. J. Med. Sci., 1839; 15: 447-450.
10. Loudon, I. The causes and prevention of puerperal sepsis. J. R. Soc. Med., 2000; 93:394-395.
11. Anon. Puerperal fever infectious. Boston Med. Surg. J., 1832; 6: 92-5.
12. Semmelwiess, I .The aetiology, cause and prophylaxis of puerperal fever. Med. Clas., 1941; 5: 350-357
13. Eskes, L.D., Longo, T.K. Classics in Obstetrics and Gynecology. Carnforth and Pearl River, Parthenon Publishing, New York. 1994;pp 339-359.
14. Gortvay, G., Zoltan, I. Semmelweis His Life and Work. Akad. Kiado Budapest,1966; pp 32-35.
15. Holmes, O.W. The contagiousness of puerperal fever. Medi. Clas., 1843;1:19-36.
16. Cullingworth, C.J. Introductory address on puerperal fever, a preventable disease: a plea for the more general adoption of antiseptics in midwifery practice. Br. Med. J., 1888; 2:743-747.
17. Storrs, R. On the contagious effects of puerperal fever on the male subject. Prov.Med.J., 1845; 19:289-294.
18. Storrs, R. On the contagious effects of puerperal fever on the male subject or on persons not childbearing. N. York. J. Med., 1845; 5:122-124.
19. Adriaanse, A,H., Pel, M., Bleker, O.P. Semmelweis the combat against puerperal fever. Eur. J. Obstet. Gynecol. Reprod. Biol., 2000; 90:153-158.
20. Codell, K. Carter. Puerperal fever. In: The Cambridge World History of Human Disease. New York: Cambridge, 1993; pp 33-53.
21. Bendiner, E. Semmelweis Lone ranger against puerperal fever; Hosp.Pract., 1987;22: 194-200.
22. Best, M., Neuhauser, D. Ignaz Semmelweis and the birth of infection control. Qual. Saf. Health. Care., 2004; 13:233-234
23. Carter, B.R. Childbed fever A Scientific Biography of Ignaz |Semmelweis. Geenwood, New Brunswick, NJ: Transaction Publishers, 2nd edition.2009; pp1818-1865.