Baby talk: survey claims third of infant health webs...

“Be careful what you Google for: Parents warned half of baby health advice online is wrong,” is the startling headline in the Daily Mail.

This story is based on a US survey looking at how well 1,300 websites identified by Google searches agreed with the American Academy of Pediatrics (AAP) recommendations on infant sleep safety.

Specific questions that the researchers were looking at included:

What’s the safest position for a baby to sleep? (On their back, according to the AAP.)
Should babies share their beds with others? (No – ideally a baby should sleep by themselves.)
What type of surface should a baby sleep on? (A firm mattress covered by a fitted sheet.)

Researchers found that, once non-relevant hits were excluded, over a third of the websites (39.2%) provided inaccurate information.

The researchers found that the most accurate websites were government-funded and non-profit websites, while the least accurate were blogs and product-review sites.

A UK audience should note that the survey was conducted using US-centric search terms (such as “pacifier” instead of “dummy”), which will have produced results for mainly US-based websites. Therefore the findings may not apply to all baby health websites worldwide.

However, as the researchers pointed out, “general trust in the reliability of internet information” can never be taken for granted, wherever you are in the world.

Where did the story come from?

The study was carried out by researchers from the University of South Carolina and other research centres in the US. The lead author was supported by the American Pediatric Society/Society of Pediatric Research Student Research Program and a National Institutes of Health grant.

The study was published in the peer-reviewed Journal of Pediatrics.

The coverage of this story by the Daily Mail was generally appropriate. However, its headline suggested that half of all baby health advice on the internet is wrong, which is not what the survey found.

The survey only looked at infant sleep safety, not at other health issues. Even in terms of infant sleep safety only just over a third of the websites surveyed (39.2%) had inaccurate information, not the "half" quoted in the headline.

Furthermore, accuracy was based on agreement with a US guideline on infant sleep safety; guideline recommendations may differ between countries.

What kind of research was this?

This was a cross-sectional survey looking at the accuracy of information relating to infant sleep safety on websites.

Many people use the internet for health information and assume that it is accurate. A US survey reportedly found that almost three-quarters of adults (72%) agreed that most or all of the health information on the internet can be believed. This study wanted to ascertain whether this belief is misguided by looking at internet information on a particular issue for which there is guidance from the American Academy of Pediatrics: infant sleep safety. These recommendations aimed to reduce the risk of sudden infant death syndrome (SIDS), suffocation, strangulation and other sleep-related infant deaths.

What did the research involve?

The researchers used 13 different phrases relating to the American Academy of Pediatrics (AAP) recommendations on infant sleep safety. This included phrases relating to AAP recommendations on:

sleep position
sleep surface
bedding
smoking
room-sharing
pacifier (dummy) use
overheating
products aimed at reducing SIDS
home monitors
bed-sharing

The researchers searched the internet for these phrases using Google search engine. They analysed the first 100 website hits identified for each phrase and categorised the type of website as:

government (.gov or .state address)
organisation (.org address)
educational (.edu address, online books, peer-reviewed articles)
company or interest group (involving a compilation of ideas from various parties or involving content not confirmed by health professionals or government officials)
sponsored link (site containing links from domain sponsors and advertisements)
news
blog
retail and product review
individual (representing one person’s ideas or being funded by one person, such as a health professional)

Using strict definitions of what was acceptable advice based on AAP recommendations, the website information was classed as:

accurate (consistent with current AAP recommendations)
inaccurate (not consistent with current AAP recommendations)
not relevant (did not address topic; did not give advice; the website didn’t work or was unrelated to the key phrase used to identify it)

The websites were analysed in July and August 2011 and re-analysed for accuracy after the most recent AAP recommendations were released in October 2011.

What were the basic results?

The researchers found that of the 1,300 websites they looked at:

43.5% provided accurate information on infant sleep safety.
28.1% provided inaccurate information on infant sleep safety.
28.4% were not relevant to sleep safety.

If the irrelevant websites were excluded, 60.8% of websites were accurate based on AAP recommendations. Looking just at the first page of hits on Google, 67.3% were accurate.

The level of accuracy varied across the different phrases being searched for. Websites relating to advice on smoking, sleep position and sleep surfaces was the most likely to be accurate (73% to 82% accurate). Websites relating to advice on co-sleeping, home monitors and pacifiers were the least likely to be accurate (14% to 20% accurate).

The most common types of websites identified were company or interest group sites, retail and product review sites, and educational sites. The websites providing the most accurate information were government websites, while blogs were found to have the most inaccurate information. The accuracy figures in the survey (excluding irrelevant websites) were:

80.9% of government websites were accurate (12.4% were inaccurate).
72.6% of organisational websites were accurate (14.7% were inaccurate).
52.4% of company or interest group websites were accurate (21.6% were inaccurate).
50.9% of news websites were accurate (36.8% were inaccurate).
50.7% of sponsored link websites were accurate (29.4% were inaccurate).
50.2% of educational websites were accurate (41.1% were inaccurate).
30.3% of individuals’ websites were accurate (36.4% were inaccurate).
25.7% of blog websites were accurate (57.5% were inaccurate).
8.5% of retail and product review websites were accurate (15% were inaccurate).

How did the researchers interpret the results?

The researchers concluded that the internet contains a lot of information which is inconsistent with the American Academy of Pediatrics’s (AAP) recommendations on infant sleep safety. They said that healthcare providers should be aware that parents may use the internet for information on infant sleep safety.

Conclusion

This study has addressed the issue of accuracy of health information on the internet. It looked specifically at the issue of infant sleep safety and rated the websites’ accuracy based on their compliance with the American Academy of Pediatrics (AAP) recommendations.

There are some points to note:

The researchers were only looking at information related to issues surrounding infant sleep safety, comparing it with the recommendations from the AAP. The websites may have more (or less) accurate results for different topics, or if based on recommendations from different bodies or countries, which could differ from US recommendations.
It wasn't clear whether the survey assessed only US-based websites or those from other countries. Websites from other countries might be less likely to comply with US recommendations.
Some of the recommendations related to controversial topics such as breastfeeding and bed-sharing, so the authors were not surprised to find variation from the AAP recommendations.

Overall, this study highlights the fact that internet users should be aware that not all health information on the internet is accurate or consistent.

In England, the Department of Health set up the Information Standard programme in 2009 to address this issue. The Information Standard aims to help the public and patients quickly identify reliable sources of quality, evidence-based information using an easily recognised quality mark. It offers certification for organisations producing evidence-based health and social care information for the public.

For reasons of transparency, it is worth noting that Capita, which was awarded the contract to deliver the Information Standard programme, also has a contract to deliver the NHS Choices website.