Patients commonly using Internet to obtain health information for themselves and their children

The Internet is commonly used by the lay public for retrieving health information.

According to the Pew Research Center in 2009, 75% of American adults go online and 61% of adults search for health information. In a pediatric-specific study, 60% of families of children with cardiac disease reported using the internet for health information. As nearly any individual, organization or society can establish a specific website, it is not difficult to envision that information obtained from websites can vary considerably, with respect to purpose, content, accuracy, completeness or readability.

The user’s ability to search the internet additionally influences what information is obtained. Patients and their families may be seeking heath information from websites that are visually attractive and easy to understand, desiring clear answers to health problems. Patients and families may not consider whether these answers are current or evidence-based.

Evaluating health information

Many studies have attempted to evaluate the quality of online health information.

In 1998, Kim reviewed web-based and literature-based criteria that were used to evaluate internet health information. One hundred sixty-five specific criteria were examined from 29 rating tools (24 websites and five published papers). The three most frequently used criteria for evaluating health-related websites by these rating tools included content of site, design/aesthetics and disclosure of authors.

In 2000, Berland evaluated internet-based health information in English and Spanish for four specific topics — childhood asthma, obesity, depression and breast cancer — using 14 different search engines. Twenty-five health websites were evaluated by 34 physicians for: 1) accessibility (proportion of links leading to relevant content), 2) quality (coverage and accuracy of key clinical elements) and 3) grade-level readability. Six of the sites evaluated were commonly known (eg, WebMD).

Of the topic first pages identified by these search engines, only 20% led to relevant content (12% for Spanish language sites). Less than 45% of the clinical elements from 10 English-language websites (four Spanish language) were more than “minimally covered” (the clinical element was mentioned at least briefly) and “completely accurate” (22% for Spanish language) for all four topics combined. For childhood asthma, only 36% of clinical elements (23% for Spanish language) were more than minimally covered and completely accurate. The corresponding figure for “therapies and adverse effects” within childhood asthma was 48% (27% for Spanish language).

Fifty-two percent of reviews of websites devoted to asthma contained conflicts (eg, same website stating inhaled corticosteroids do and do not stunt growth). All of the English language websites and 86% of the Spanish language sites required at least a high school level reading ability.

To compare and analyze how studies such as these have evaluated web-based health information, Eysenbach systematically reviewed 79 studies of 5,941 websites. Studies were evaluated for technical criteria (eg, authorship, disclosure), design, readability, accuracy and completeness. Study method rigor was analyzed for search strategy and evaluation quality.

Of the 79 studies evaluated, 70% concluded that information quality was problematic, and 9% concluded that information was good. However, the studies concluding that web health information was good used less rigorous search and evaluation strategies than negative studies. None of the positive studies used evidence-based guidelines as a criterion standard. Most concluded that significant problems in information content were present.

It is important to consider these findings in light of the complex and wide variation of internet site design, purpose and search strategy used by consumers and study researchers, among other factors. The study authors concluded that operational definitions of quality criteria are needed for evaluating online health information.

Pediatric-specific studies

Several studies have evaluated the quality of online health information of specific pediatric health topics.

Italian researchers assessed patient-oriented webpages for information on the management of pediatric fever in 1996 (Impicciatore). A checklist of various clinical elements was used to evaluate relevant pages. This checklist was obtained from a textbook (Fever in Pediatric Practice, El-Radhi AS, 1994) and included minimum temperature defining a fever, optimal site for measuring temperature, when to visit a physician and pharmacotherapeutic treatment (eg, acetaminophen 10-15 mg/kg every four hours).

Forty-one websites (34 sites from the United States or Canada) were identified using the search engines Yahoo and Excite. Of these websites, 32 were commercial and nine pages were developed by individual practitioners, clinics, academic institutions or other educational organizations. Pharmacotherapy was discussed on 34 sites – acetaminophen was recommended in 31 sites and ibuprofen in 14 sites. However, only eight sites recommending acetaminophen- specified dosing and frequency.

Dipyrone and aspirin were also recommended in four sites. Dipyrone was removed from the United States commercial availability in the late 1970s due to the potential for agranulocytosis. The country of origin of the websites discussing dipyrone or aspirin was not specified. The study researchers concluded that a minority of websites provided complete and accurate information overall on pediatric fever management.

The same researchers similarly evaluated information on the home management of pediatric cough from internet sites in 1997 using six search engines. Information quality was assessed by comparison with two documents – an AAP document (Use of codeine- and dextromethorphan-containing cough remedies in children, 1997) and a World Health Organization publication (Drugs for Children, 1987).

Nineteen webpages (16 from U.S. sources) were identified and analyzed, mostly representing commercial sources. Three checklists were used to assess webpages – technical (eg, author listing), information completeness and information quality. Dextromethorphan was the most frequently discussed cough suppressant on these sites and was recommended for use by all sites where it was discussed. The use of an expectorant, most commonly guaifenesin, was discouraged in 40% of the sites where it was discussed. When cough suppressants were listed on these webpages, discussion of their use largely failed to state that little evidence exists for their efficacy. Multi-ingredient products were recommended on 32% of webpages, as well as benzonatate (11%), ammonium chloride (5%) antihistamines-decongestants (5%) and decongestants (11%).

Only four sites reminded readers that online health information is not a substitute for professional care, and of the websites scoring highest in technical aspects (eg, author listing, reference listing, date modified), their information was largely incomplete and inaccurate. The authors concluded that overall, only one website (an individual pediatrician’s website) rated high on all three checklists.

Other researchers evaluated internet sites for information quality on another common pediatric problem, diarrhea (McClung, 1998). The first 300 websites were analyzed from internet searching using three search engines, and 60 formally prepared documents from traditional medical sources were selected for analysis. Of these 60 documents, 43 were produced from teaching centers, news services or practitioners. The AAP practice parameter on the management of pediatric acute gastroenteritis (1996) was used as the standard to assess website information. The authors concluded that only 12 of the web documents evaluated conformed to the AAP guidelines for treatment, and that much information from these online documents was erroneous or outdated.

Conclusions

As the internet is a commonly used information source, the quality of information obtained from it is important to consider. Many studies have analyzed this information, often concluding that health information is largely incomplete, inaccurate and requires relatively high reading abilities. However, these studies have mostly used descriptive analysis, and their methodologies have varied considerably. Many variables exist when analyzing how online information is retrieved and how it affects health outcomes, including internet search ability. Very little evidence is available that use of online health information has actually resulted in patient harm. More studies are needed to measure how online health information affects patient health – both negatively and positively. Practitioners should be cognizant that consumers are likely using the internet to retrieve health information. Practitioners may wish to discuss this with their patients and families and offer recommended internet sites.

For more information:

• Berland GK. Health information on the Internet, accessibility, quality, and readability in English and Spanish. JAMA. 2001;285:2612-21.
• Eysenbach G. Empirical studies assessing the quality of health information for consumers on the world wide web. JAMA. 2002;287:2691-2713.
• Impicciatore P. Reliability of health information for the public on the world wide web: systematic survey of advice on managing fever in children at home. Br Med J. 1997;314:1875-1885.
• Kim P. Published criteria for evaluating health related Web sites: review. Br Med J. 1999;318:647-9.
• McClung HJ. The Internet as a source for current patient information. Pediatrics. 1998;101:10-14.
• Pandolfini C. Parents on the Web: risks for quality management of cough in children. Pediatrics. 2000;105:1-8.
• Purcell GP. The quality of health information on the Internet. Br Med J. 2002;324:557-558.

Edward A. Bell, PharmD, BCPS is Professor of Pharmacy Practice at Drake University College of Pharmacy in Blank Children’s Hospital and Clinics in Des Moines, Iowa.