Unwarranted variations in Swedish childbirth care

Meet Johan Mesterton, doctoral student at the Department of Learning, Informatics, Management and Ethics at Karolinska Institutet. For his thesis, he analysed registry data from almost 140,000 births in seven Swedish regions between 2011 and 2012.

What are your most interesting results?

That there are large differences in outcomes in Swedish childbirth care, differences that can’t be explained by different hospitals treating patients with different characteristics. We’ve adjusted for differences in patient populations between the hospitals, and still there are large variations in how the patients are treated and in health outcomes. It’s also interesting that medical staff feel that data feedback on their results is motivating and can support their quality improvement efforts.

What would have been the effect if all birth clinics had been as good as the best?

During the period studied, 2,200 caesarean sections, 900 severe perineal tears, 1,500 infections and 2,700 haemorrhages, could have been avoided. That would have been the case if all clinics had performed as well as the 20 per cent that performed best on each quality indicator.

What can be done to eradicate the differences?

I think the key is to systematically measure healthcare outcomes and to feedback the information to healthcare staff. If you don’t know how many infections there are at your clinic, and even less how it compares to other clinics, it’s hard to motivate yourself to initiate a systematic change programme. Everyone in healthcare wants to give their patients the best possible care. With better feedback on outcomes and improvement potential to the staff, I think everyone will perform better.

Why did you choose this particular concern for your thesis?

I’ve been working for a long time with analysis of healthcare data and am really interested in how healthcare improvements can be supported. My research was linked to Sveus, a collaboration project involving seven regions in Sweden on the development of methods for monitoring of results in healthcare. Childbirth care was one of the patient groups included in Sveus and I found it fascinating.

What will you do next?

I’m an industry-financed doctoral student and will continue to work at my company, which develops analytical products for the healthcare sector. In the future I hope to carry on researching in one way or another.