Abstract
To synthesize findings from previous studies assessing information needs of primary care patients on the Internet and other information sources in a primary care setting. A systematic review of studies was conducted with a comprehensive search in multiple databases including OVID MEDLINE, CINAHL, and Scopus. The most common information needs among patients were information about an illness or medical condition and treatment methods, while the most common information sources were the Internet and patients’ physicians. Overall, patients tend to prefer the Internet for the ease of access to information, while they trust their physicians more for their clinical expertise and experience. Barriers to information access via the Internet include the following: socio-demographic variables such as age, ethnicity, income, education, and occupation; information search skills; and reliability of health information.
Conclusion:
Further research is warranted to assess how to create accurate and reliable health information sources for both Internet and non-Internet users.
Key points for decision makers
If physicians do not satisfy patients’ information needs, patients will turn to other sources of information to satisfy those information needs. Projected physician workforce shortages could make demand for information more acutely. 1
Knowing patient information needs and information-seeking behavior will assist physicians to effectively satisfy patients’ expectations, which could promote informed clinical decision making and lead to increased quality of patient care.
This information can assist in improving the health information systems for personal health information such as patient health records (PHR) and patient portals.
Since patients frequently access health information online, ensuring the availability of valid, usable, and accessible information is a priority.
Introduction
Information needs and information sources of patients
The Institute of Medicine (IOM) recognizes “patient-centered care (PCC)” as one of six major domains of healthcare quality. The Agency for Healthcare Research and Quality (AHRQ) defines PCC as the relationship-based primary care that meets the individual patient and family’s needs, preferences, and priorities. 2 PCC integrates the disease and illness experience while acknowledging the whole person to create a sharing of power, responsibility, and therapeutic alliance. 3 Better understanding of patient information needs is important for providing patients with updated and relevant information to assist in making informed decisions concerning their healthcare and allows patients to be involved in assessing healthcare options available to them. In this review, “information needs of patients” is defined as patients’ desire for more information on a particular subject matter that is expressed verbally or in active information-seeking to assist in taking better care of themselves.4,5
The frequency of patients seeking information has increased due to widespread access to health information mass media and the Internet. 6 With the increased access to the Internet, large amounts of health information are now readily available for consumers. In 2013, 73 percent of American households had Internet access. 7 The limited access to information from traditional channels, such as physicians and books, and the ready availability of trusted online sources, may be driving the demand for online health information. 8 For example, by 2010, the use of print media fell sharply from 33 percent to 18 percent, which may account for most of the decline in overall patient information-seeking. 9 This downward trend in the use of print media for health information may be explained by the declining circulation of print newspapers, magazines, and sales of hardcopy books. 10 Currently, health information is the third most popular online activity measured in the latest Pew Internet study, which is a nationwide survey of 3014 adults living in the United States. Fifty-nine percent of the US adults search online and 52 percent of smartphone owners use their phone to find health information. 11 Identifying patients’ current information needs and information sources will assist with presenting information in a manner that suits the health information needs of the patients. Harris and Dewdney’s principles of information-seeking are employed as an analytic framework and structure for categorizing and examining information sources patients use to satisfy their information needs.12,13
Information-seeking models
Information-seeking models have evolved from researchers in the library and information science disciplines. Marchionini and White describe the process of information-seeking as a set of activities that people undertake in a progressive and diversely iterative manner. The information seeker (1) recognizes the need for information; (2) accepts the challenge to take action to fulfill the need; (3) formulates the problem; (3) uses a search system to express the information need; (4) examines the results from the search, if not satisfied; (5) reformulates the problem and its expression, if satisfied, decides to stop the search; and (6) uses the information found. 14
Kuhlthau describes the information search process (ISP) as a six-stage model of the users’ experience during the information-seeking process: task initiation, selection, exploration, focus formulation, collection, and presentation. The ISP model pinpoints three dimensions of experience that are prevalent to each stage: the cognitive (thoughts), the affective (feelings), and the physical (actions). 15 As the search process evolves, users’ thoughts that begin as ambiguous become clearer. Research has shown that users experience anxiety and doubt in the beginning of the search process but become more confident and certain as the search progresses. Users take action in the beginning of the search process by seeking general information and focus on more relevant information toward the end of their search.16,17
Sense-making is producing situational awareness and understanding in highly complicated and uncertain circumstances in order to make a decision. 18 Dervin et al. describe information-seeking as a situation-sensitive sense-making process. The sense-making model focuses on the cognitive gap that hinders a user’s momentum when trying to make sense of information he or she observes. The user needs to conquer that gap by getting help or making sense of the situation in order to achieve their desired outcome.18,19
Harris and Dewdney combined past research of DePaulo’s 20 steps in help-seeking, Krikelas’ 21 model of information-seeking, and Dervin et al.’s sense-making model to create six general principles of information-seeking behavior. Harris and Dewdney made these principles generalizable in order to apply to ordinary people, such as primary care patients. The principles are as follows: (1) information needs arise from the help-seeker’s situation; (2) the decision to seek help or not seek help is affected by many factors; (3) people tend to seek information that is most accessible; (4) people tend to first seek help or information from interpersonal sources, especially from people like themselves; (5) information seekers expect emotional support; and (6) people follow habitual patterns in seeking information.12,13
Information overload
There are many sources of information available today for patients to find answers to their health-related questions, which may cause information overload. Bawden et al. 22 suggest that information overload can occur when information received becomes more of an impediment rather than benefit. Information overload may cause the following: incorrectly processing information, delay in processing information, accepting lower quality information, and giving up the search for needed information. 23 In order to present patients with information effectively, it is necessary to assess their information requirements as the clinical environment moves toward a patient-centered model and health information becomes more attainable through multiple sources.
Objective of the study
This review seeks to identify the information needs and sources that patients use frequently when seeking information. In this review, information-seeking is defined as “a basic activity in which a person identifies their information needs and searches for the information.”24,25 In this study, information-seeking is not limited to seeking information from just technology sources. Information seekers may use other sources, such as people and books, to satisfy their information needs. The first objective of this study was to identify and analyze the research on information needs of patients in a primary care setting. A primary care setting was chosen because primary care physicians are the initial contact with patients and play an important role in preventive and ongoing healthcare 26 and account for the majority of patient visits for conventional illnesses. 27 The second objective was to determine the information sources used to satisfy those information needs especially the Internet because of its accessibility. The third objective was to identify barriers to accessing the identified information sources, specifically Internet-based sources. Knowing what information patients need and what source(s) they use to satisfy these information needs will assist physicians to effectively satisfy patients’ expectations, which may improve informed clinical decision making and potentially raise the quality of patient care.
Method
Search strategy and study selection
This review selectively adopted the Preferred Reporting Items for Systematic reviews and Meta-Analyses (PRISMA) guidelines to format the contents of the review. PRISMA is a validated method to conduct literature reviews which ensures the transparent and complete reporting of systematic reviews. 28 PRISMA includes a checklist of items to include when reporting a systematic review and a flow diagram of showing the different phases of a systematic review. Both the checklist and flow diagram were used as references to guide the structure of this article. The checklist has 27 items, which include the title, abstract, methods, results, discussion, and funding, that should be included in the content of a systematic review. Figure 1 illustrates the flow of information through the different phases of a systematic review. 28

Flow chart of the review of literature to finalize selected articles to include in this review. After examination of full articles, 46 articles were chosen to be a part of the review.
Data sources
To identify pertinent studies, a literature search was conducted with assistance from two professional health librarians for the combined concepts indicated below. There were searches in OVID MEDLINE, OVID MEDLINE In-Process, Cochrane Controlled Trials Register (limited to Embase Records), Cochrane Database of Systematic Reviews, PsycINFO, CINAHL, Scopus, Communication and Mass Media, Library Literature and Information Science Full Text, ERIC, and Library Information Science and Technology Abstracts. We began searching databases for articles on 27 August 2011 and we continued searching until the submission of this article. A total of 607 citations were initially retrieved from these databases.
Inclusion criteria
Retrieval was limited to original articles in primary care of adult patients in outpatient setting, written in English, and which were published from 2000 to 2015. Medical Subject Heading (MeSH) terms used were as follows: patient education as a topic, health education, information-seeking behavior, primary healthcare, and family practice. Search terms used that were not included in MeSH were as follows: information needs and information source and over 50 other terms. A detailed search strategy, including all the terms used, is found in Supplementary Appendix 1.
Exclusion criteria
Studies excluded from this review were opinion pieces, editorials, articles using secondary data, review papers, articles on patients seen by specialist physicians, and inpatient hospital setting. Figure 1 shows a flow chart of the review of literature to finalize selected articles to include in this review. The examination of article abstracts revealed 141 articles for full article review. After examination of articles, 98 were excluded because they were not directly related to information needs or information-seeking behavior of patients, were concerning pediatric patients, or were from studies not in a primary care setting. Forty-six relevant articles then became a part of the review.
Data extraction
Titles and abstracts were initially reviewed to select articles for full-text analysis by three authors (M.A.C., M.S.K., and J.L.B.). A literature search was conducted with assistance from two professional health librarians and relevant articles were located by one author (M.A.C.). When the author was unsure whether the article met inclusion criteria, two authors (M.S.K. and J.L.B.) were contacted to assess the article’s relevance. One author, a family medicine physician (J.L.B.), assisted in identifying articles that were related to primary care and identifying those that were not. Articles were examined for instances of codes for information needs and information resources. Information needs and sources were coded only if the article specifically mentioned them in their results. Data extracted from the articles were as follows: author name, year of publication, title of article, disease (if mentioned), research method used to collect data, country study took place, sample population, aim of the study, main findings of the study, information needs and sources mentioned, and barriers to information access.
Results
Characteristics of reviewed studies
Table 1 summarizes the studies selected for evaluation on information needs and information-seeking behavior of patients sorted alphabetically. The table contains the authors, research method, location of study, sample population, aim of the studies, and important findings. The studies were conducted in multiple countries including the United States,31,33,37,38,41,42,48,51,57,65,67,69,70,72,74 United Kingdom,29,30,34,43,46,49,52,53,55,58–60,71,73,75 Canada, 40 Australia,35,44,54 The Netherlands,56,62,63,68 Croatia, 36 Denmark, 32 Greece,47,66 Japan, 61 and Scotland. 39 Common research methods included interviews,29,30,32–35,40,49,54,58–60,71–73,75 focus groups,48,50,51,67 questionnaires,31,36,39,44,46,52,53,55,60,62,63,66,67 and computer logs.65,70,74
Summary of studies selected for evaluation on the information needs and information-seeking behavior of patients sorted alphabetically.
Information needs of patients
The first objective of this article was to identify the information needs and information-seeking behavior of patients in a primary care setting. The following information needs were commonly mentioned in more than one of 23 articles pertinent to information needs (Table 2): information for an illness or medical condition,29,36,38,41,43–45,49,53,54,58,59,61,62,64,67,72,75 nutrition,36–38,40,45,47,56,61,62,64,67,71,72 and alternative medicines or new/experimental treatment available.29,37,38,59,61–64,67 The complete list of information needs is found in Table 2.
List of information needs of patients commonly mentioned in pertinent articles from the review of literature (percentages of articles overlap).
Information sources of patients
The second objective of this study was to identify the sources used by patients when satisfying their information needs. Most researchers reported patients using multiple sources to address information needs; however, the following information sources were identified most frequently in the 19 articles pertinent to information sources (Table 3): Internet,30–32,37,39,44,45,47,51,53,56,58,59,61,64,66,67,73,75 physicians,30,31,37,39,42,44,45,47,51,53,56,59,61,66,67 television,32,37,39,44,45,47,50,51,53,56,61,66,67,73 and magazines.31,32,37,47,50,51,53,56,59,61,66,67 Television and magazines were more found in the older studies, with Internet becoming more predominant in the newer studies. The full list of information sources is found in Table 3. Four articles mentioned specific websites accessed by patients when satisfying information needs.30,50,51,67 These websites were as follows: an American support group; 30 chat rooms; 30 email lists; 30 Parkinson’s Disease Foundation; 50 Ostomy Support Group; 50 diabetes.com; 51 WebMD; 51 American Diabetes Association; 51 Mayo Clinic; 51 websites devoted to cooking, exercise, and stress reduction; 51 Alzheimer’s Association; 67 and well-known hospitals and established organizations. 67
List of information sources utilized by patients commonly mentioned in pertinent articles from the review of literature (percentages of articles overlap).
Barriers to information access to the Internet
The third objective of this study was to identify the barriers that affected patients’ use of the Internet and other information sources. Factors affecting information-seeking behavior included the following: age,57,61,75 education,37,55,57,61 and household income.37,55,57,61 Younger patients,57,61,75 patients with higher educational background,37,55,57,61 and patients with higher household income37,55,57,61 were also more likely to seek information and to use the Internet as an information source. Sources may be available to patients, but there may be barriers to accessing relevant information such as deficiency in reliable, credible information, and Internet search skills.30,48,51,67,76–79
Discussion
We are using Harris and Dewdney’s six principles of information-seeking behavior as an analytical framework to explain our results.
Harris and Dewdney Principle 1: information needs arise from the help-seeker’s situation
In the context of this study, primary care patients’ visits to their primary care physician are often driven by the need to receive diagnostic and/or therapeutic information from their physician based on symptoms the patient may be experiencing. Patients try to satisfy their information needs during initial contact with healthcare providers, 80 and patients prefer having background knowledge about their condition along with an understanding of the diagnosis. Hence, it was not surprising that information on an illness or medical condition, especially chronic illnesses, was a common information need. In 2013, 7 of the 10 leading causes of death in the United States were chronic diseases, and almost 50 percent of Americans had at least one chronic illness. 81 A report by World Health Organization (WHO) calls for simple lifestyle changes, such as nutrition and physical activity, to reduce the prevalence of chronic diseases. 82 Both proactive patients, who wish to prevent chronic diseases, and chronic disease patients looking for natural remedies to complement their current treatment seek information on nutrition and exercise. 57 For some chronic conditions, regular exercise can help manage symptoms and improve health. Regular physical activity reduces the risk of coronary artery disease, stroke, type 2 diabetes, colon and breast cancer, osteoporotic fractures, osteoarthritis, depression, and erectile dysfunction. 83 The health benefits from physical exercise may be the driving force that influences patients’ information need for exercise/physical activity.
It is difficult to completely separate the information sources patients use from their information needs. For example, patients initiate pre-consultation information-seeking to determine the type of physician they need to see29,38,56,58,61,64,67,72 and to prepare for a consultation.29,35,41,49,56,72 For this information, they frequently use the Internet as an information source.41,56,58,61,64,67 This is usually motivated by a concern not to waste healthcare resources or their own time.29,33,35,41,47,49,84 During a consultation, patients attempt to validate the information found on other resources with their physician’s knowledge.33,51 After consultation, patients also seek information as a second opinion to validate the information provided at consultation29,33,56,67 and to research alternative medicines or new/experimental treatment available to them.29,37,38,59,61–64,67 Also, patients may not be satisfied with the information provided by their physician and feel the need to seek out information on their own, perhaps to fill in gaps in the information provided.35,49,57,72 For these needs, patients use the Internet as a source of information.
Harris and Dewdney Principle 2: the decision to seek help or not seek help is affected by many factors
In the context of this study, many factors affected the decision to seek information or not seek information through the use of Internet for health information. For example, in countries lacking a publicly funded healthcare system, the uninsured patients use the Internet as an alternative expert for medical advice due to costs and their limited access to physicians as an information source. 48 Younger people with Internet access are more likely to search the Internet for health information because of the ease of use and privacy when searching for sensitive health information.47,85 Similarly, health information seekers with higher educational background or household income were more likely to use the Internet to look for medical information.37,47,55,57,61 Patients who are independently searching for health information, especially on the Internet, may face challenges due to lack of effective search skills and issues with reliability of information.34,41,67,86 Usability was a major deterrent for patients when using electronic health information resources. For example, female patients seeking information on human papillomavirus and female patients with polycystic ovary syndrome stated that the information found on the Internet was useful but not easy to acquire. 54 Searching skills are required to find quality information. Therefore, the issues with technical search terms that are not used in everyday language and complicated website design may result in navigation difficulties that may discourage patients. 86 Natural language processing may be beneficial in the case of technical search terms to assist patients in finding information they need. Health information websites could improve ease of use by not requiring patients to enter complex medical terms when searching and by performing more usability testing to detect and resolve common navigation and user input errors.
Validity of healthcare information was also a big concern when determining satisfaction with a health information resource. Identifying quality health information is not always easy because of the difficulties in distinguishing credible sources from unreliable sources. Although the Internet contains a large amount of readily available information, it is unregulated and may be inaccurate and contain bias.87–89 Adelhard and Obst 90 indicated that more than half of health information websites offer unreliable information. Thus, patients who engage in their healthcare by seeking online health information may receive inaccurate information that can harm their health. The increase in patient or personal health record technologies (e.g. patient portals) as additional information sources may help in addressing issues of the inaccuracy and bias associated with the Internet as a patient information source.
Harris and Dewdney Principle 3: people tend to seek information that is most accessible
According to Harris and Dewdney, 13 “… information should be physically, psychologically, and intellectually accessible.” In a number of studies, the Internet was used as a second opinion or as information supplement after an office visit. A study by Tustin 91 found an association between unmet information needs provided by their physician and patients’ reliance for health information from the Internet, which shows there is a need for improved patient–provider communication. This finding is consistent with the results of other studies that identify the Internet as being used as a major communication tool when individuals felt a lack of patient-centered communication with their healthcare providers.91,92 Patients may also use the Internet when new questions arose after the consultation. 33 For example, female patients searched for information on human papillomavirus 43 on the Internet after consultation, because they felt inadequately informed, needed a better understanding, and wanted to find more updated information on treatment options. 43 Diabetes patients searched for information on the Internet to increase their knowledge of self-management of their chronic disease. 51 The Internet also has the potential to influence some patients’ decisions about treatment;84,92,93 thus, it is important that patients receive reliable online health information to reduce the risk of making incorrect clinical decisions based on Internet content. Beyond the more general publicly accessible Internet sources, patients are more frequently gaining online access to clinicians’ visit notes. These patient-specific web-based resources can help to increase patient engagement in healthcare, specifically, patient portals, which are secure online websites providing patients convenient 24 hour access to personal health information. OpenNotes is a national initiative in the United States that encourages providers to give patients access to their healthcare provider’s visit notes. 94 Access to patient portals could provide patients with accurate information from their provider’s office. 95 For example, the US Department of Veterans Affairs invited patients and their delegates to review and download content from their electronic health record through the online portal, My HealtheVet. 96 Although some patients found access to their records overwhelming, the results show that patients reviewing their records positively affected patient–provider communication, improved self-care, and increased patient participation improving quality of care.
Harris and Dewdney Principle 4: people tend to first seek help or information from interpersonal sources, especially from people like themselves
Our results show that family and friends were an important source of information for patie-nts.31,32,37,42,44,51,53,61,66,67,71 Primary care patients receive various information, such as advice on diets and weight control, information on medical problems, and medications, especially from people who have the same disease.32,44,51 Studies by Wilkinson et al. 71 and Longo et al. 51 found that relatives and friends were frequently cited as sources of information, especially in families with a history of diabetes. Patients in Stavropoulou’s 66 study regarded family and friends as information sources on hypertension and medication for hypertension. Malone et al. 53 found that patients were also using interpersonal sources for advice on alternative forms of healthcare. Patients who do not use the Internet as their main health information sources rely on their family and friends for information than patients who use the Internet. This suggests that Internet users may be less reliant on these other sources as principal sources of health information.37,67
Principle 5: information seekers expect emotional support
Findings from this review show that patients sought information from email and support groups to satisfy their information need.30,32,51 Women diagnosed with polycystic ovarian syndrome sought out information from support groups and involved themselves in activities that provided contact with other sufferers. This not only allowed women to share their experiences with each other but also they were able to gain information concerning their condition on both an individual and a more general level. 30 Patients diagnosed with diabetes in Longo et al.’s 51 study used relatives and friends as a sounding board to discuss the disease and its management, as well as clarify any contradictory information received from health professionals and the media. Healthcare providers should make emotional support available to all patients.
Principle 6: people follow habitual patterns of seeking information
When an information source has been helpful in the past, users will revisit that source for another need, unless some barrier intervenes. People develop information habits in relation to the sources they consult. 12 In the context of this study, patients have confidence in the knowledge of their physician, which may be the reason why patients use their physicians as one of their primary sources for health information.30,59,67,76–79,97,98 Although patients are actively seeking information on their own, a national survey reported that 72 percent of US adults still receive information, care, or support from a doctor or other healthcare professional. 11 However, because of physicians’ lack of accessibility, patients have addressed this deficiency using the Internet as an information source. Overall, patients tend to prefer the Internet for the ease of access to information, while they trust their physicians more for their clinical expertise and experience.59,67,76,78,79,97
Validity of selected articles’ research methodologies
We must recognize that our findings are heavily dependent on the research methodologies employed in the articles included in this review. A range of research methodologies was used in the selected articles. Questionnaires and surveys were a very common method for collecting information needs of patients. Out of 46 articles, 21 articles utilized the questionnaire or survey method to collect data, 14 articles utilized interviews as a method of data collection, 8 articles utilized focus group as a part of their data collection, and 3 articles employed log analysis as their data collection method. However, questionnaires, interviews, and focus groups are implemented after an event, which could be a weakness to these methodologies because of the dependency on the participant’s ability to accurately and completely recall information. Also, recall using these methods may also introduce social desirability bias. 99 For example, Hogue et al. 44 used open-ended questions that required people to recall past events and behaviors when investigating media influence on consumers’ health-related behaviors and participants may have forgotten important information that could have been pertinent to the study. A study by Zach et al., 73 Budtz and Witt, 32 and Caiata-Zufferey et al. 33 used face-to-face anonymous interviews with patients waiting for appointments. Patients being approached in a waiting room may feel trapped and may not answer questions truthfully when asked questions during the interview, including a bias toward social desirability in their answers. Five articles adopted a mixed methods approach to their data collection. Using a mixed methods approach, such as triangulation, can increase the validity of research findings. For example, to determine older adults’ health information needs and perceived usefulness of health information sources, Taha et al. 67 used both focus groups and a questionnaire to gather data. Future research should therefore consider a mixed methods methodology to assist in increasing the comprehensiveness of overall findings by including both quantitative and qualitative methods. 100
Limitations of this review
This review concentrates only on published studies. A possibility of publication bias can be included and was not investigated. This review also excludes non-English language studies, which may suggest different results. Some studies were Internet-based, which may have biased toward the Internet as a preferred source among patients. The samples used within the articles need to include more diverse patient groups with larger sample sizes for generalization of the results. This review excluded the information needs of children, specialist patients, and caregivers because of their unique information needs, and in the case of children and adolescents, the unique roles of cognitive development and parents as information proxies. Further investigation should be conducted on information needs of pediatric patients and caregivers for results to be generalizable to all patient groups. This review focused on identifying information needs and sources for primary care patients. However, it may be important to better understand the health-related information needs of a more general population. By focusing on primary care patients, we may miss information needs or sources of people who are not yet patients. People may use the Internet to avoid being a “patient.” That is, they may want to avoid visiting the doctor, because (1) they do not have time to visit the doctor, (2) they consider or suspect that their problem is a minor complaint not worthy of a doctor’s visit, or (3) they do not have a doctor because they have previously been healthy.
Conclusion
Health information needs in primary care is an under-investigated area. This study provides a foundation for determining the most common information needs among primary care patients and understanding how the lack of physician accessibility can awaken these information needs. The breadth of information in primary care is unique and primary care patients have different information needs than the general population who are not actively patients, so this review contributes to the literature by focusing on actual patients. While there is some literature on patient information needs, this review adds to the discussion of patient information needs by utilizing an information-seeking framework to structure the discussion in order to help to identify types of user-centered behavior which may affect when and how patients satisfy their information needs. In addition, the findings from this review point to the importance of Internet-based information sources to supplement or extend the information gleaned from physician–patient interactions. Improving patient engagement in healthcare through online access to clinicians’ visit notes and personal health record may improve the quality of patient information by reducing their reliance on less reliable sources. In spite of the increase in available information through patient-centered online health records and information sources, it is challenging to distribute accurate and useful health information to patients without Internet access. However, it is essential to meet that challenge and facilitate quality health information exchange to increase patient awareness and achieve better health outcomes.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
