Heart diseases are still chronically misdiagnosed or underdiagnosed in women. With depressing regularity, we see stories of women failed by the health system when they come to hospitals with the symptoms of a heart attack. As a professor of cardiac science with 40 years’ experience, for me it has been a frustrating journey to get to the real cause of this problem: a combination of professional, systemic and technical biases. The experiences of individual patients are complex to analyse and interpret, but now we can view these effects on a much bigger scale.
Women are 50% more likely to receive a wrong initial diagnosis; when they are having a heart attack, such mistakes can be fatal. People who are initially misdiagnosed have a 70% higher risk of dying. The latest studies have similarly shown that women have worse outcomes for heart operations such as valve replacements and peripheral revascularisation. As well as being misdiagnosed, women are less likely to be treated quickly, less likely to get the best surgical treatment and less likely to be discharged with the optimum set of drugs. None of this is excusable, but is it understandable?
The first excuse commonly offered is that women don’t develop heart disease as much as men and so seeing a woman with a heart attack is “unexpected”. It’s true that women themselves may not expect to have a heart attack and so may overlook the first symptoms. But I’m unconvinced by the justifications I often hear from clinicians. While it may also be true that the heart disease rate is lower in younger women, it is far from a rare event. Each year more than 30,000 women are admitted to hospital in the UK with a heart attack. For every 10 young men with heart disease, a doctor can expect to see between three and five young women, and the numbers are even closer in the older age brackets. Compare this with meningitis, for example, where a general practitioner may see only one or two cases in their entire career. Taking all heart disease types over an individual’s lifespan, about 21% of women die from heart disease – not far off the figure for men, at 24%. For a clinician, seeing a woman with heart disease in casualty cannot be called unexpected by any stretch of the imagination.
A second popular excuse is that women’s symptoms are strange and unpredictable. But the reality is that there is a great deal of overlap between the sexes in the symptoms they experience. Feeling sick, sweaty, or lightheaded are symptoms common to both, as is the classic symptom of crushing chest pain, often radiating up the arms and to the jaw. This pain is the most common symptom in men and women, although women are more likely than men to experience it in their back. Shortness of breath and tiredness are again common, but there is more probability that women will be experiencing breathlessness, fatigue or nausea when they come to A&E. Since there will be a significant number of women coming to the hospital with heart attacks, there should be no excuse for remaining ignorant of this range of symptoms.
Once a heart attack is suspected in a patient, the standards and guidelines for treatment are well defined. Doctors should be recognising heart disease in women and giving them the optimum standard of care. However, this is not happening. Clinicians are less likely to stick to the guidelines when treating women, sending them home with painkillers rather than the armoury of therapeutics we have now. Women are less likely to receive the gold-standard treatment, where the blood vessels are opened using catheters to restore the blood flow. One study of more than 100,000 hospital patients found that men have 20% more of these reperfusion treatments than women and men were nearly twice as likely to survive while in hospital. Even when women do get treated, there is not as much haste in doing so. The time from first contact with a doctor to reaching the catheter laboratory for reperfusion therapy is vital: for every five-minute delay there is a 5% increase in the risk of death. The study also found that women were moved to the catheter laboratory significantly less promptly than men and this contributed to the higher death rate.
However, the study’s most shocking statistic was that this only happened if the doctor was male. Why should this be?
The biggest study on physician gender and treatment came from the experience of 1.3 million Florida residents who had been admitted to hospital for a heart attack. Survival rates were two to three times higher for female patients treated by female physicians compared with female patients treated by male physicians. Male physicians who had good prior experience of treating women did improve their success rate – there was a measurable increase in survival with every new female patient they saw. Even more interesting, the number of women clinicians in the team made a big difference to the men they worked with. A higher proportion of female doctors improved both the success of the team in general and the competence of men in the team for treating women. The study concluded that the best way to help female patients was to have a gender-balanced team, rather than waiting for individual male doctors to gain experience at the expense of their early failures.
What is it about female patients that makes the male doctors treat them differently? What behaviours or characteristics trigger this response in the clinician? This is where the difference between sex and gender plays a part. Each of us, independent of our biological sex, has a range of gendered attributes that are traditionally thought of as male or female and, importantly, that might be valued differently if displayed by a man or a woman. Are you shy, gentle and compassionate or assertive, risk-taking and individualistic? There is a test you may like to try called the Bem sex-role inventory that assesses how “male” or “female” your behaviour is – almost all of us will fall somewhere between the two extremes.
Our home circumstances also affect how we are seen: factors such as being the primary wage earner, having a high income or doing most of the housework. All these add up to how male or female we appear. When gender and biological sex were compared for how they influenced treatment, it was the perceived gender – the strength of the “female” score compared with the “male” – that made the difference in treatment and outcome. For example, “female” patients (men or women) were more than four times as likely to return to the hospital with recurrent symptoms after being discharged. Essentially, behaving in a manner perceived as traditionally female downgrades you in the eyes of a male physician – there is a higher chance that your distress will be seen as overblown, inaccurate or hysterical.
Uncontrollable emotional excess has long been associated with women and has alternately been classified as a disease of either the body or mind. The Greeks termed it hysteria (or wandering womb, hystera being the word for womb) and only at the time of Freud was the same behaviour pattern recognised in men. In her book Sex Matters, the US physician Alyson McGregor describes how women who are in pain often have trouble convincing the doctor treating them of how serious that pain is. The more they protest and try to convince the physician, the more their behaviour is perceived as hysterical. Women from more demonstrative cultures have a particularly hard time. If they have grown up always encouraged to be very vocal about their emotions, then this can work against them in casualty. As McGregor says, the best thing you can do as a woman is to bring a man with you to explain.
On a side note, one clue to understanding whether these observations are the result of bias is to observe whether the same is true for other disadvantaged groups as well as women. As it happens, the same phenomenon for doctor-patient matching occurs for race, with patients from minorities doing better with a physician of the same race, or in a team with a good proportion of minority doctors. The use of healthcare resources and satisfaction with the outcome both rise when there is good matching. This is part of a much wider appreciation that there are numerous inequalities in medical care by race in the US and the UK. It is not hard to predict the pattern for minority women, who are doubly disadvantaged in terms of healthcare.
If gender-balanced teams are an answer to the problem, why is this not happening in cardiology? Clinical cardiology has traditionally been a predominantly male occupation – sometimes referred to as “boys and toys” because of the many and varied devices that can be implanted in the heart, purportedly attracting male clinicians to the discipline. The UK Athena Swan gender equality scheme for universities was operating in my own institution, and over about 10 years made many adjustments to reduce bias in hiring and promotion practices. By 2020 we had raised the number of female science professors in cardiology to be about equal with the number of males. Female clinical cardiology professor numbers in our associated hospitals, however, remained stubbornly at 10% of the number of males. In the US, more than 50% of medical school students are women but this figure drops to only 4.5% for the practising “interventional” cardiologists (the ones using catheters to treat heart attacks). This difference seems to be crucial in the poor treatment of women who come into hospital with cardiac symptoms.
What can we do while we wait for the world of cardiology to achieve gender parity? A recent study used data science to combine the clinical signs, heart measurements and blood test results from 13,000 people coming into hospital with cardiac symptoms. With the help of artificial intelligence (AI) the researchers produced an algorithm that was significantly better at diagnosing heart attacks in general (achieving almost 84% accuracy in sending patients for further tests, compared with about 50% accuracy for conventional tests) and levelling up the field between men and women. We just have to hope that advances such as this will make some inroads into shocking statistics such as the estimated 8,200 women in England and Wales who died of heart problems between 2002 and 2013 because of misdiagnoses. Hopefully, data science, with its analysis of large numbers of patients, will give us new insights and will show the potential for AI-generated algorithms to provide a solution.
Sian Harding is emeritus professor of cardiac pharmacology at Imperial College London. Her new book, The Exquisite Machine: The New Science of the Heart, will be published on 20 September by MIT Press (£25). To support the Guardian and Observer, order your copy at guardianbookshop.com. Delivery charges may apply