I was accused on national radio last week of risking my son's life by giving birth at home. I had been talking on the BBC's Woman's Hour about the threat to independent midwives, who work outside the NHS and offer a natural approach to birth. A change of UK law in October will ban any healthcare professionals from practising without indemnity insurance. That means no more independent midwives, because no insurance provider will cover them. The consultation period for the legislation ends on Friday.

I had explained to the presenter how after a traumatic first labour that ended in an emergency caesarean section, I opted to have my second baby at home under independent care. The next day, one of the comments read out on the programme came from an anonymous, male obstetrician. He said I had taken "a significant risk" and accused me of putting my own satisfaction ahead of my child's safety: "I'm still amazed at this minority of people who place more importance on the experience of childbirth than on the outcome."

He is not alone in his view. The American College of Obstetricians and Gynaecologists has said that home birth prioritises "the process of giving birth over the goal of having a healthy baby". The women who choose it – especially those with "high-risk" pregnancies like me – are regularly accused in the press of being reckless. In the UK, the NHS supports home birth for low-risk women, but advises those with complications – anything from high blood pressure to breech birth, twins and previous c-sections – against attempting it.

I don't believe employing an independent midwife makes me selfish or irresponsible. Before choosing how to give birth I studied the evidence carefully and considered how to balance the competing risks. I came to the conclusion that much of the opposition to home birth is based on a skewed perception of risk that prioritises the benefits of medical intervention while overlooking all other concerns.

Let's look at the death rate first. Randomised trials comparing planned home and hospital births are almost impossible to do, but observational studies suggest that for low-risk women at least, home birth is just as safe: for example a meta-analysis of 24,000 women published in 1997 found the same death rate (0.87%) in both groups. A 2011 study of 64,000 UK women, which included serious injuries as well as deaths, did find that home birth was slightly more dangerous for first-time mums (0.93% vs 0.53%), though for women who'd had children before, the risk was the same.

There are few data on the safety of home birth for complicated pregnancies as so few women attempt it. In my case, the concern was that my c-section scar might rupture during labour. One big meta-analysis of trials done in hospital suggests this happens in around 0.4% of cases, while a more recent UK trial found an overall rate of 0.2%. My risk would be lower than the average, because I had only one previous c-section and a fair gap between pregnancies. Even if I did suffer a rupture, warning signs would probably give me time to get to hospital – a 10-minute drive away. But let's be clear, there was a chance (perhaps an extra 0.1% or so) that by giving birth at home I could lose my baby.

That's very small – substantially lower than the death rate I already faced even in hospital. To be accused of risking my baby's life seems hysterical, especially when you compare it to the dangers of medical interventions routinely carried out during pregnancy, such as amniocentesis or chorionic villus sampling, which kill 1-2% of foetuses. But why chance such a horrendous outcome at all?

Because these death rates don't give the whole story. What the trials also show is that on every other measure, women and their babies are safer at home. The authors of a 2012 Cochrane review – the medical profession's gold standard analysis – pointed out that planned hospital births are more likely to end in complications, which they blamed on "impatience and easy access to many medical procedures at hospital".

Women who choose birth at home are less likely to need drugs to speed up labour or reduce pain, less likely to have an instrumental or surgical delivery, and less likely to be cut open or to tear. Their babies are in better shape when they are born and are more likely to breastfeed.

This is not just about the "process" or "experience" of childbirth; there can be long-term health implications. Breastfeeding protects babies against infection, for example, while complicated deliveries affect how mothers bond with their babies and increase their risk of postnatal depression.

It isn't known whether the trauma of medical interventions during birth directly affects babies in the long term, but it might: stressful events experienced in the womb and in early life make us more sensitive to stress – and more at risk of disease – for the rest of our lives.

Archie Cochrane, the pioneer of evidence-based medicine (after whom Cochrane reviews are named), was an outspoken critic of hospital births. In the 1960s, he awarded "the wooden spoon" to obstetricians for confining pregnant women to hospital, then introducing a whole series of expensive innovations, from induction to ultrasound – all without proper trials.

It seems that little has changed. More than half of all births in England now involve either instruments or surgery. Obstetricians routinely use a range of interventions – including foetal monitoring, episiotomy and early cord clamping – that the evidence suggests have marginal or no beneficial effect, if not to be downright harmful.

So yes, giving birth at home involved a tiny extra risk of losing my son – this kept me awake at night – but the evidence also showed that it would almost certainly give him the best possible start in life. Accepting small risks to enhance our children's lives is what we do every time we allow them to cross the road, ride a bicycle or swim in the sea.

When it comes to birth, not every woman would – or should – make the same decision. We all have different priorities and face different risks. Many feel safest with life-saving technologies on hand, or choose elective c-section, and they have every right to do so. But home birth can be a safe, sensible, evidence-based choice too, even for some of us whom obstetricians write off as high-risk.

Thankfully, the NHS is providing increasing facilities for midwife-led birthing centres and home births, but there are strict guidelines around who can access them. Labour is still expected to progress to a strict timetable or staff will intervene. And there is little focus on emotional support for women going through one of the most painful events it is possible to experience. It is not generally possible to choose or even meet your midwives in advance, for example, or to have the same staff stay with you throughout a birth.

This is where independent midwives come in. They avoid medical interventions unless absolutely necessary, and instead focus on helping women to give birth for themselves, with many (but not all) of their deliveries taking place at home. A guiding principle is "continuity of care" which means the same midwife cares for a woman throughout her pregnancy, as well as during and after birth.

A trial of nearly 9,000 UK women (matched for age and socioeconomic status) found that whether in hospital or at home, women attended by independent midwives suffered fewer complications and were much more likely to breastfeed. Their babies were less likely to be premature, have low birthweight, or be admitted to intensive care. There was a higher neonatal and stillbirth death rate (1.7% vs 0.6%), but the authors concluded that this was because the independent midwives looked after a greater proportion of high-risk cases such as twins. When this was accounted for, the death rate in both groups was the same.

I know from experience how their approach can transform the outcome of a birth. During labour, my baby was in an Iawkward position and I was pushing for over two hours – longer than NHS guidelines allow. NHS staff would have transferred me to hospital and extracted the baby with scissors and forceps, or more likely (given my previous history) another c-section.

That would have meant all the risks of emergency surgery, an extended hospital stay, longer recovery leaving me unable to care for my three-year-old daughter at a critical time in her life, as well as potential difficulties persuading my newborn to breastfeed.

Instead, my independent midwives monitored the baby and – with all looking well – they encouraged me to take my time. It was the reassurance I needed and my son was born in perfect health. I got the natural birth I'd hoped for, but only because I was attended by midwives who supported me rather than rushed to intervene.

At between £2,000 and £4,000 per birth, independent midwives are sadly beyond the reach of many couples, while others save for years to afford them. If the government doesn't rethink its plans, they won't be available at any price.

Jo Marchant is a freelance science journalist based in London

• This article was amended on 15 May 2013. The original stated that the meta-analysis of 24,000 women was published in 2008. This has been corrected.