Searching for health information online: characteristics of online health seekers

Authors


Correspondence
S Gallagher, Research and Information Analyst, Mental Health Research Unit, Health Research Board, Knockmaun House, Lower Mount Street, Dublin 2, Ireland. Tel: 01-2345-220 Fax: 01-661-2214 Email: silviaelenagallagher@gmail.com

Abstract

Objective To investigate the use of, and willingness to use, the internet for health information purposes in Ireland, with a particular focus on the demographic and mental health of users and non-users.

Methods Data from the Health Research Board National Psychological Wellbeing and Distress Survey were analyzed. This telephone survey included 2711 adults aged 18 years and over living in private households in Ireland. Internet use was measured using three items: past use, use specifically for health information, and willingness to use the internet in the future. Current psychological wellbeing was measured using the 12-item General Health Questionnaire. In addition, mental health problems in the previous year were measured. Sociodemographic variables examined included age, gender, employment status, and education.

Results Online health-seeking differed significantly by gender, age, employment status, and educational level. A substantial proportion of internet users with poor mental health were found to have searched for health information online. A high willingness to use the internet for health information in the future by non-users was reported.

Conclusion This study has shown that those with mental health problems use the internet to search for health information more often than those who have not experienced such problems. It has also shown that there is still a digital divide in terms of online health seeking, which is influenced by age, gender, education, and employment status. These findings highlight the need to address access, technological, psychological, and perceptual issues that may present barriers to the use of online health seeking.

inline image

Introduction

E-health is the delivery or enhancement of health services and information through the internet and related technologies (1). An important aspect of e-health is the growing tendency for individuals to use the internet for health information. These internet users, who use search engines, websites, message boards, and other online resources to inform themselves on health matters, are known as online health seekers (2, 3). The most frequent online health information searches relate to specific diseases and illnesses. However, many other topics are commonly searched for, including advice on diet and nutrition, and information on prescription drugs, health insurance, and mental illness (3).

Online health information can provide users with anonymity and information to increase the possibility for self-care, and a means for information exchange and community support (4, 5). It can also empower users to seek help and increase understanding of their medical conditions (6). Most individuals do not consider the internet as a replacement for health professionals (5–7), but it is becoming ever more important to view the internet as a source of health information because of its increased use at all levels of society (8–10).

Even with all the positive outcomes that the internet offers to consumers of online health information, searching for specific information can cause many difficulties and disadvantages in relation to the reliability and quality of health websites (3). Finding high-quality sites can be problematic (11), and there are difficulties sourcing information within websites, and problems with the completeness and accuracy of information (12, 13). Additionally, there have been user issues with trust and the quality of information provided on health websites (14). More worryingly, using health information from the internet for decision-making without expert advice could potentially have a negative impact on a patient's health (15).

International studies have shown significant differences in the demographic make-up of online health seekers. More women than men are reported to use the internet for health information (16–19), and individuals who are better educated and of a higher social class are more likely to use the internet as a source of health information than individuals in a lower social class and educational level (16, 20–23). A study into the online health seeking of low-literacy adults found that low literacy levels inhibit information seeking online (24). This study reported that many health websites have content that requires high reading levels, which makes online health seeking problematic for those with low reading ability.

Health status has also been found to be a significant factor in online health seeking. Houston and Allison (25) reported that those who disclosed fair or poor health status were more frequent users of the internet for health information. This was consistent with previous studies which found those who defined themselves as in either poor or fair health, those with chronic conditions, and those suffering from a long-term illness or disability were more likely to visit health sites than those who reported better health (23, 26). Individuals with even moderate levels of anxiety tend to search for health information more frequently online than those without anxiety (27). Powell and Clarke's (18) UK study on online information searching found that 18% of their sample (n= 917) who had ever used the internet had used it to find out about a mental health issue.

It is argued that the internet allows anonymity and can provide support and information for those with stigmatized illnesses such as mental health problems (28, 29). Respondents with stigmatized illnesses such as anxiety, depression, or herpes were significantly more likely to use the internet for health information than respondents with non-stigmatized illnesses such as back pain and diabetes (29). More strikingly, results found that they had increased their use of health care and had better communication with medical staff based on information they had found on the internet. Powell et al (28) found that 49% of people who were using the internet for mental health problems had not sought formal support from health services. In addition, of those who did seek help, more than one-third reported that the internet had been an important factor in deciding to seek help. Help seeking for mental health is often delayed because of public or self-stigma (30). The power of the internet as a potential means of encouraging help seeking was revealed by these studies.

It is evident from previous literature that a substantial percentage of internet users are searching for health information online; however, it is also important to examine those who are not using the internet for this purpose, or those who do not use the internet, such as elderly people. Their willingness to use the internet for health seeking is an important issue for future online health provision and promotion. Studies have reported that elderly patients were reluctant to use the internet because they felt they were too old to understand it or because of cost and education concerns (25, 31). However, increased education on internet use and health seeking for the elderly was found to boost willingness to use the internet as a source of information on health (32).

The aims of this paper are to examine the demographic and mental health status of online health seekers and to examine the willingness of non-online health seekers to search for health information online in the future.

Methods

The Health Research Board National Psychological Wellbeing and Distress Survey (HRB NPWDS) was a telephone survey of the prevalence of psychological wellbeing and distress in the Irish population. It was conducted using a random sample of the population aged 18 years and over and living in private households (n= 2688). The data were re-weighted in line with best practice. For more information on the HRB NPWDS and the re-weighting procedure see Doherty et al (33).

For the purposes of this paper, only items relating to use of the internet were analyzed in conjunction with the demographic and self-reported mental health status of participants. Figure 1 illustrates the sequence of questions in the HRB NPWDS using three items: past use of the internet, internet use specifically for health information, and willingness to use the internet in the future.

Figure 1.

Sequence of questions on the internet in the Health Research Board National Psychological Wellbeing and Distress Survey (HRB NPWDS).

Internet use

All respondents were asked if they had ever used the internet for any purpose. Those who answered positively were subsequently asked if they had ever used the internet as a source of information on health. Respondents who answered that they had never used the internet for any purpose, and respondents who answered that they had never used the internet for health information, were then asked if they would be willing to use the internet as a source of information on health (Figure 1).

Sociodemographic variables

The sociodemographic variables used in the analysis included gender, age (five categories), employment status (seven categories), and educational level (thee categories). Table 1 shows the specific categories in each variable.

Table 1.  Characteristics of internet users who had, or had not used the internet as a source of information on health in Irelanda
 Used the internet as a source of information on healthNot used the internet as a source of information on health
  1. aBase: internet users in sample (n= 1566).

  2. *P < 0.001.

  3. GHQ-12, 12-item General Health Questionnaire.

Gender*
Male40.7% (325)59.3% (473)
Female58.0% (441)42.0% (319)
Age group*
18–2945.5% (320)54.5% (384)
30–3958.5% (210)41.5% (149)
40–4955.3% (136)44.7% (110)
50–6441.2% (87) 58.8% (124)
65+35.9% (14) 64.1% (25) 
Employment status*
Employed52.9% (459)47.1% (408)
Self-employed48.6% (71) 51.4% (75) 
Unemployed32.7% (33) 67.3% (68) 
Training/education38.6% (93) 61.4% (148)
Domestic duties55.0% (71) 45.0% (58) 
Retired39.1% (18) 60.9% (28) 
Long-term sickness and disability75.0% (18) 25.0% (6)  
Educational level*
Primary52.7% (48) 47.3% (43) 
Secondary44.5% (447)55.5% (558)
Third level58.6% (272)41.4% (192)
Self-reported mental health problems*
Yes59.8% (101)40.2% (68) 
No47.7% (654)52.3% (717)
GHQ-12 measure of current psychological distress (P = 0.242)
Case51.9% (150)48.1% (139)
Non-Case48.1% (587)51.9% (634)

Psychological distress in the last few weeks

The 12-item General Health Questionnaire (GHQ-12) was used to measure psychological distress. This questionnaire has been used extensively in previous research to assess psychological distress in a population (34, 35). Scores were calculated using a bimodal scoring method, which involved scoring items on a scale of 0–0-1–1 with a score range of 0 to 12. In line with previous studies, those with a score of 2 or above were defined as ‘cases’ of significant psychological distress.

Self-reported mental health problems

Respondents were also asked if they had had any experience of mental, nervous, or emotional problems (eg anxiety or depression) in the last 12 months; yes/no response categories were provided.

Data analysis

Relationships between the demographic variables and internet use variables were examined using Chi2 tests. Sociodemographic predictors of internet use were examined using logistic regression analysis. To reduce the standard error, employment status was recoded into three categories: employed, in training, and not in employment or training. Logistic regression models were developed for each of the internet use variables, and only variables that were statistically significant in the Chi2 analysis were included in the regression analysis. Only those variables that added significantly to the model were included in the final model (ie P less than or equal to 0.25) (36).

Results

Use of the internet as a source of information on health

Of the total sample (n= 2688), 58.3% (n= 1566) had used the internet at some stage for any purpose. Almost half of these internet users had used the internet to source information on health (49.0%, n= 767). Table 1 describes the characteristics of these internet users in terms of their use of the internet for health information. Note that only valid responses to each question were used in the analysis.

Internet users with a long-term sickness or disability had the highest levels of using the internet for health information (75.0%, n= 18), followed by those in domestic duties (55.0%, n= 71). Unemployed internet users were the least likely to seek online health information (32.7%, n= 33) (Chi2(6) = 36.78, P < 0.000). Significant differences were also shown in educational level with the highest percentage having completed third-level education (58.6%, n= 272) (Chi2(2) = 25.90, P < 0.000). There were significantly higher numbers of women searching for health information online than men (Chi2(1) = 46.61, P < 0.000). Users between 30 and 39 years were most likely to use it as a source of information on health, with those aged 65 years and over least likely.

A significantly higher percentage of internet users who reported mental health problems in the past 12 months had used the internet for health information (59.8%, n= 101) than those who did not report mental health problems (47.7%, n= 654) (Chi2(1) = 8.75, P < 0.00).

Logistic regression analysis was used to examine the demographic predictors of internet use for health information by internet users (Table 2). Only the variables that were significant in the Chi2 analysis were included in the initial model, and three variables remained in the final model: employment status, education, and gender. The Chi2 goodness-of-fit test was significant (Chi2(5), 89.63, P < 0.000), and the Hosmer–Lemeshow test result not significant (P > 0.05) which suggested a good fit. The final model predicted 7.5% of the variance (Nagelkerke r2= 0.075, Cox and Snell r2= 0.056).

Table 2.  Logistic regression model of internet use for health information
VariableBOdds Ratio95% CISignificance (p)
  1. Base: internet users in sample (n= 1566)

  2. B, logistic regression coefficient; CI, confidence interval.

Gender (Reference males)
Female 0.7762.1741.760, 2.6860.000
Employment   0.000
 status (Reference employed)
Training−0.5780.5610.417, 0.7550.000
Not in−0.3900.6770.515, 0.8900.005
 employment
 or training
Education   0.000
 (Reference primary)
Secondary−0.4850.6160.394, 0.9620.033
Third level−0.0010.9990.624, 1.5980.995

The regression analysis showed that women were twice as likely as men to use the internet for health information (odds ratio (OR) 2.174, 95% CI 1.760 to 2.686), and employed respondents were more likely to be online health seekers than those in training (OR 0.561, 95% CI 0.417 to 0.755) or those not in employment or training (OR 0.677, 95% CI 0.515 to 0.890).

Willingness to use the internet as a source of health information

Respondents who had never used the internet for any purpose and internet users who had never used the internet for health information were asked whether they would be willing to use the internet as a source of health information (Table 3).

Table 3.  Characteristics of respondents who were willing or unwilling to use the internet as a source of information on healtha
 Willing to use the internet as a source of information on healthUnwilling to use the internet as a source of information on health
  1. aBase: Respondents in the sample who had never used the internet and internet users who had never used the internet as a source of health information (n= 1886).

  2. *P < 0.001.

  3. GHQ-12, 12-item General Health Questionnaire.

Gender (P = 0.76)
Male49.3% (479)50.7% (493)
Female48.6% (444)51.4% (470)
Age group*
18–2970.7% (325)29.3% (135)
30–3964.4% (204)35.6% (113)
40–4952.0% (168)48.0% (155)
50–6440.5% (170)59.5% (250)
65+15.5% (57)84.5% (310)
Employment status*
Employed59.9% (440)40.1% (294)
Self-employed49.4% (85)50.6% (87)
Unemployed54.0% (67)46.0% (57)
Training/education83.9% (135)16.1% (26)
Domestic duties39.2% (124)60.8% (192)
Retired15.4% (43)84.6% (237)
Long-term sickness and disability30.8% (28)69.2% (63)
Educational level*
Primary22.9% (111)77.1% (374)
Secondary57.0% (660)43.0% (498)
Third level62.3% (152)37.7% (92)
Self-reported mental health problems (P = 0.429)
Yes46.6% (125)53.4% (143)
No49.3% (788)50.8% (812)
GHQ-12 measure of current psychological distress (P = 0.347)
Case50.3% (699)49.7% (692)
Non-case47.6% (202)52.4% (222)

There was a high willingness by respondents with no experience of using the internet and by those who had never used the internet for health seeking purposes to do so in the future (48.9%, n= 923). Significant differences in age group, employment status, and educational level were observed. No significant differences were found in gender, self reported mental health problems, and GHQ-12 cases.

Respondents who were in training and education (83.9%, n= 135) were the most willing to use the internet as a source of information on health, and retired respondents (15.4%, n= 43) the least (Chi2(6) = 265.64, P < 0.000). Those who had completed third-level education (62.3%, n= 152) were more willing to use the internet as a source of information on health than those who had completed primary (22.9%, n= 111) and secondary levels (57.0%, n= 660) (Chi2(2) = 179.228, P < 0.000). A much higher proportion of younger respondents reported willingness to use the internet as a source of information on health than older respondents (Chi2(4) = 294.09, P = 0.000).

A logistic regression analysis was performed on the willingness to use the internet as a source of information on health. Only variables that were statistically significant in the Chi2 analysis were introduced to the model, and the final model included three of the four variables: age, employment status, and education (Table 4). The Chi2 goodness-of-fit test for the final model was significant (Chi2(8) = 389.929, P < 0.000), and the Hosmer–Lemeshow test result was not significant, (P > 0.05), which suggested a good fit. The final model predicted 25% of the variance (Nagelkerke r2= 0.25, Cox and Snell r2= 0.185).

Table 4.  Logistic regression model of the willingness to use the internet for health informationa
VariableBOdds Ratio95% CISignificance (p)
  1. aBase: Respondents in the sample who had never used the internet and internet users who had never used the internet as a source of health information (n= 1886).

  2. B, logistic regression coefficient; CI, confidence interval.

Age (Reference 18–29)   0.000
30–39  0.0871.0550.755, 1.4750.754
40–49−0.3210.7060.507, 0.9840.040
50–64−0.6180.5210.375, 0.7230.000
65+−1.6580.1910.125, 0.2920.000
Employment   0.000
 status (Reference employed)
Training  0.9722.6421.627, 4.2920.000
Not in−0.3990.6710.526, 0.8560.001
 employment or training
Education   0.000
 (Reference primary)
Secondary  0.7162.0461.552, 2.6980.000
Third level  0.8412.3181.585, 3.3890.000

Respondents aged over 65 years were the least willing to use the internet for health information (OR 0.191, 95% CI 0.125 to 0.292) and respondents not in training or employment were less likely to be willing to use the internet for health information (OR 0.671, 95% CI 0.526 to 0.856) than those in employment. Respondents who had completed either secondary (OR 2.046, 95% CI 1.552 to 2.698) or third level (OR 2.318, 95% CI 1.585 to 3.389) education were both twice as likely to be willing to use the internet for health information as primary level educated respondents.

Discussion

Our analysis has shown that there are significant differences in gender, age, employment, and educational level between internet users who search for health information online and those who do not. Internet users who defined themselves as having a long-term sickness or experiencing disability were noticeably more likely to use the internet as a source of information on health than any other category. This is an interesting result that provides evidence of high usage levels by those experiencing sickness or disability.

The gender difference in online health seeking has been attributed to women's tendency to have more of an interest in dieting and healthy eating than men (3), and their role as primary carers in the home (2). There was no gender difference in the willingness to use the internet as a source of information on health, suggesting that men are as willing as women to use the internet for online health seeking. As men are more likely to use the internet in general than women, (37) further exploration of why they do not use it proportionally for health information is required. The findings also suggest that sociodemographic factors are more predictive for willingness to use the internet for health information than for actual behavior (25% of variance predicted compared to 7%), suggesting that other factors may influence use. These may include psychological factors, such as perceived current mental or physical health status, or interest in promotional and preventative health measures.

A greater proportion of internet users who had reported mental health problems in the last year had used the internet as a source of information on health than those who had not experienced mental health problems. This suggests that the mental health status of respondents influences their online health seeking behavior. This raises issues around the potential use of the internet for mental health promotion or as a mental healthcare delivery system in the future. These findings replicate previous work on the health of online health seekers (27). Studies have revealed that respondents with stigmatized illnesses (such as anxiety and depression), were more likely to look up health information online than those with non-stigmatized conditions, and that the internet was somewhat beneficial for them (7, 28, 29). While individuals do not see the internet as a replacement for a medical professional (6), it may provide extra information that can be used in the health decision-making process.

Use of the internet to search for health information has many benefits, but there are also negative consequences associated with searching for health information online. For example, extreme communities have been formed online, such as the ‘pro-ana’ or pro-anorexia movement, and the phenomenon of online suicide pacts (38). As it is impossible to censor such information on the internet, little can be done to prevent the creation of these sites. However, it is important to promote websites that give information and support positive mental health.

A major disadvantage to health seeking on the internet is that the quality of the health websites remains a challenge and is difficult to ensure. A number of steps have been taking in recent years aimed at ensuring the quality and accuracy of health information on the internet. These include Health on the Net Foundation (HON) which oversees and assesses the content of health websites, and DISCERN, a tool to assess the quality of the information on internet sites. An Irish review conducted in 1998 found that there was only a small number of Irish healthcare websites, none of which satisfied the internet health information Code of Conduct (11). Given the expansion of healthcare websites in the past ten years, there is an obvious need to carry out another similar review. The HON code (39) could also be used in public health promotion to guide users to the most accurate e-health sites on the internet, specifically in relation to mental health, since results from this and other studies have shown that mental health issues are commonly searched for by general internet users. Concerns have been raised about the quality of mental health information online, with studies finding an overall poor level of information for disorders such as depression, schizophrenia, attention deficit hyperactivity disorder, and eating disorders (38). Again there is a need to review mental health information websites to ensure that they are of a sufficient standard.

A high percentage of respondents reported willingness to use the internet for health information. Willingness to use the internet for health information decreased with age. Respondents aged 30–39 years, those in training, and those with a second or third-level education were the most willing to use the internet as a source of information on health in the future. The respondents who were least willing to use the internet as a source of information on health were those in the older age groups, those who were not in employment or training, and those with only primary-level education. These findings point to demographic factors that influence the use of the internet specifically as a source of information on health. Those who do not have access to the internet or are unable to access the internet for health information may be at a disadvantage as more health information and even delivery of health services become available through the internet. Training and education in the use of the internet and online health seeking is required for these individuals.

A limitation with the current survey was that it was a telephone survey of people aged 18 years and over and living in private households with a telephone line. As a consequence, it is likely that those most vulnerable to exclusion from an ‘e-society’ have been excluded from this survey. This includes immigrants, refugees, homeless people, and people living in institutions. In addition, there may be a substantial number of people who do not have landlines and rely on mobile phones or internet connections (eg voice over internet protocol technologies) to keep in touch. The latter individuals are obviously more likely to use the internet for a range of purposes and would have been excluded from this survey.

Conclusion

This analysis has added to the literature by providing information on the mental health of internet users. It is clear from the findings that there is still an existing digital divide in terms of online health seeking. The findings highlight the need to design and develop further initiatives aimed at addressing this divide. It is also important that initiatives are developed that target specific groups that are least likely to use the internet. Initiatives need to address access and technological issues, but also psychological and perceptual issues that may impede the use of information technology.

In line with previous studies, this survey showed that a substantial number of people are using the internet for health seeking or are willing to use the internet for this purpose in the future. In addition, a large number of internet users with disabilities or long-term physical or mental health problems were using the internet as a source of information on health. This highlights the opportunity for using the internet as a source of health promotion and health service information.

Of primary concern is the quality of the information currently provided on healthcare websites. There is an urgent need to evaluate and assess current sites for quality of information. Evaluation results then need to be circulated widely so that health professionals and the public are made aware of the sites that meet appropriate standards. In addition, the public need to be made aware of the dangers of purchasing medicines or health diagnostic kits online. Further research into the specific activities of online health seekers would be useful for future health promotion; for example, the type of health information sought, the procedure involved in online health seeking, opinions about the information retrieved, and the outcomes of seeking health information online.

For those who do not use the internet as a source of information on health, or are less willing to do so, there is a need to increase awareness of the benefits of using this medium, either for health and wellbeing, or for information on specific illnesses.

The findings highlight the need for a comprehensive study of online and other forms of health seeking in Ireland, and an analysis of the mental health and attitudes of internet users in order to fully understand their needs. This survey has provided an insight into the use of the internet by people with mental health problems, discussed the potential of the internet to be used for mental health promotion, support, and information, and highlighted the need for further research into this topic.

Ancillary