During the debate for healthcare reform in the US there was a lot of media coverage on the UK’s National Health Service. The NHS was disparagingly used as an example of what the future US system might look like. Most of this coverage focused on emotive topics like the so-called ‘death panels’ and the type of healthcare that somebody like Steven Hawking might receive under the NHS. The flaws of these sweeping statements are easy to unpick. However, other examples used to show varying health outcomes across different systems, such as cancer survival rates in the US v.s. the UK, are less easy to understand. The way they were covered in the media gave the appearance of potential deficiencies in the NHS.
International comparisons of outcomes in healthcare are complex. To grasp them fully it is first necessary to understand exactly where the data has come from and how it was analysed. This article in the Journal of the Royal Society of Medicine does a good job looking at some of the issues behind such comparisons between the UK and USA.
The authors looked at life expectancy, neonatal and infant mortality, cancer survival and death rates due to diabetes. Their findings are that the NHS actually does a pretty good job. Life expectancy at birth measures how long someone born in a certain year would be expected to live based on the predicted death rates into the future. Currently someone born in the UK can expect on average to live 2 years longer than someone in the USA. Neonatal and infant mortality rates are both higher in the US than in the UK, which may in part explain why life expectancy is higher in the UK, as these figures influence life expectancy numbers. An important point with all of the data used to calculate these statistics is that it is mainly based on birth and death certificates making it reasonably comparable and accurate due to the systematic and full coverage of such data collection.
Cancer survival statistics are more complicated. Looking at the raw numbers, the UK appears to lag behind the USA in common diagnoses like breast cancer. In their article, Desai et al. point out that the percentage of women still alive 5 years after a breast cancer diagnosis in 2000-2002 was 77.8% in the UK and 90.1% in USA. Does this mean that women in the USA get better treatment and therefore survive for longer as a result of this? Not necessarily. Whilst this might be true in part, a large portion of this difference can be explained by the sources of the data used in the two countries.
To calculate survival rates it is necessary to know how many people are diagnosed with the condition each year, and then how many of these people survive until the time point you’re interested in – 5 years is a standard length of time quoted for these figures.
Deaths due to breast cancer can be estimated using death certification. Although this isn’t always 100% accurate because the cause of death might not always be clear to the doctor completing the death certificate. In the case of breast cancer most cases should be picked up, as it’s fairly unlikely to be missed as a cause of death.
Calculating the number of new diagnoses of breast cancer in a year is a more difficult task. In the UK a national registry collects data on all new diagnoses, and although this may not be 100% accurate, this registry is fairly comprehensive and covers most of the population. In the USA, up until 2000, data were not collected for the whole population. So data for this analysis were based on a sample that covers approximately 10% of the population. This doesn’t necessarily have to be a problem – if this 10% sample were representative of the rest of the country. If that were the case, the comparison between the UK and US would be valid. However, as other studies have shown, those who tend to have the lowest survival rates (typically the poor and African American’s) are under-represented on the sample. This then has the overall effect of artificially inflating the 5-year survival rates because those people who are diagnosed with breast cancer and are more likely to have a poor survival rate aren’t included in the figures. It may not fully explain the difference in rates, but it will certainly reduce the difference.
As the authors of the article conclude, many commentators in the US condemn healthcare reforms for creating a system like the NHS because it represents ‘socialised medicine’ with poor healthcare outcomes for its patients. However, comparisons of such outcomes are not always as simple as they seem at first glance. Looking at the data more closely brings to light that actually the NHS has many very good outcomes when compared to the system in the USA.