Studies analysing the need for health-related information in Germany - a systematic review

Background Exploring health-related information needs is necessary to better tailor information. However, there is a lack of systematic knowledge on how and in which groups information needs has been assessed, and which information needs have been identified. We aimed to assess the methodology of studies used to assess information needs, as well as the topics and extent of health-related information needs and associated factors in Germany. Methods A systematic search was performed in Medline, Embase, Psycinfo, and all databases of the Cochrane Library. All studies investigating health-related information needs in patients, relatives, and the general population in Germany that were published between 2000 and 2012 in German or English were included. Descriptive content analysis was based on predefined categories. Results We identified 19 studies. Most studies addressed cancer or rheumatic disease. Methods used were highly heterogeneous. Apart from common topics such as treatment, diagnosis, prevention and health promotion, etiology and prognosis, high interest ratings were also found in more specific topics such as complementary and alternative medicine or nutrition. Information needs were notable in all surveyed patient groups, relatives, and samples of the general population. Younger age, shorter duration of illness, poorer health status and higher anxiety and depression scores appeared to be associated with higher information needs. Conclusion Knowledge about information needs is still scarce. Assuming the importance of comprehensive information to enable people to participate in health-related decisions, further systematic research is required. Electronic supplementary material The online version of this article (doi:10.1186/s12913-015-1076-9) contains supplementary material, which is available to authorized users.


Background
Providing health-related information to the public and, more specifically, to patients and their relatives, can empower them to make informed decisions concerning prevention, screening and treatment [1]. Public involvement in healthcare decisions is regarded to be an essential element of high quality care [2,3]. Related to this, the concept of empowering patients to participate as active partners in health-related decisions in terms of Shared Decision Making (SDM) has gained more prominence in the last years [4]. Therefore, a crucial prerequisite is the provision of information to the patient meeting his demands. Moreover, not only information targeting ill patients is needed, also 'healthy people' need healthrelated information, e.g. to decide whether they should participate in prevention or screening interventions.
Hence, exploring information needs is necessary to better tailor information to the specific needs of the target population and should thus be regarded as a prerequisite to the development of patient information.
Several systematic reviews exist investigating information needs [4][5][6]. However, prior work mainly concentrated on a single condition, in particular cancer, and involved patients [5,6] or partners and family members, however, did not assess information needs in the general population [4].
Therefore, we conducted a systematic review on information needs independent of certain diseases in patients and relatives as well as the general population in Germany. We were particularly interested in the following questions: 1. For which diseases information needs were investigated 2. Which methodologies were used?
And regarding the findings of the studies 3. What are the health-related topics of information needs that the general population or patients are interested in? 4. How is the extent of (unmet) health-related information needs expressed by the investigated populations or patients? 5. How are personal or disease-related characteristics (e.g. age, gender, course of disease, etc.) or other variables associated with the desire for more information or the need for specific information about single topics?

Methods
Studies were identified by searching the bibliographic databases Medline (via Embase), Embase (via Embase), Psycinfo (via EBSCO) and all databases of the Cochrane Library. Information needs may change over time, for example due to open access to a wide range of information or changes in the health care system. To obtain current information needs, the search was limited to a publication date from January 2000 to August 2012. There was no protocol for this systematic review.

Definitions
The definition of information needs is ambiguous. In this review we define information needs as the 'recognition that their knowledge is inadequate to satisfy a goal, within the context/situation that they find themselves at a specific point in the time' [7].

Inclusion and exclusion criteria
A study was included in the review if: (1) information needs were investigated; (2) the study participants were at least 18 years old; (3) the focus was on patients, parents or other family members of patients or the general population; (4) the study population lived in Germany; and (5) the full-text publication was written in German or English.
Systematic reviews were excluded. Studies that focused on information seeking or information preferences instead of information needs were excluded.

Study selection process
All titles and abstracts were screened independently by two authors. The full-texts of potentially eligible articles were obtained. The references of the included studies were checked for further potentially relevant publications. Two reviewers assessed the eligibility of the full-texts according to the review inclusion criteria. Any disagreements were resolved by discussion. To identify studies performed in Germany we developed a geographic search filter (and applied it to each database) based on a search filter for retrieving studies performed in Spain [8]. The sensitivity of the search filter was increased by using more field descriptors in addition to searching for countryspecific geographic names in the affiliation. The full search strategies for the individual database providers can be found in Additional file 1: Appendix 1.

Data collection process
Information on study type and design, population and methods of data collection was extracted from all included studies in standardized summary tables by one reviewer and checked by a second reviewer. To explore our main research questions, the following information was systematically extracted and analysed using/applying a descriptive content analysis (one reviewer coded key study findings, while a second reviewer verified and discussed them for each article and research question) based on predefined categories: 1. Diseases for which information needs were investigated 2. Methodology which has been used, as: qualitative or quantitative study, cross sectional or longitudinal study, assessment by written questionnaire or interview, closed or open-ended questions.
And regarding the findings of the studies 3. health-related topics of information needs. The topics were grouped into 12 categories based on a scheme originally developed by Rutten et al. for cancer patients [6]. The scheme was adopted and applied irrespective of health condition (Table 1).
Regarding the extraction and analysis we distinguished if topics of information or single items were predefined (e.g. rating or choice task, dichotomous questions) or if participants had to state or express their need for information by themselves (e.g. open ended questions). Data was extracted for this question, if at least two categories were analysed in a study. 4. whether and how the extent of information needs was evaluated, i.e. with regard to medical information in general or about specific topics (e.g. ratings of predefined topics or proportion of people who claim to have unmet information needs). 5. whether and how associations between information needs and other variables (e. g, anxiety, depression, disease activity, duration of illness, education) or patient characteristics (e.g., age, gender, course of disease, etc.) were explored.
We did not assess quality of the included studies due to a lack of an established and validated critical appraising tool for the expected study designs.
This study did not need ethical approval nor was individual patient consent needed.
The characteristics of the studies are summarized in Table 2. Fourteen papers were published in the last 6 years which indicates a slightly increasing research interest.

Extent of information needs and unmet needs
Ten studies contained data about the extent of information needs [10,11,14,17,[21][22][23][24][25][26]. The need for more detailed medical information was reported frequently. The demand for information was substantial in the whole population as well as in specific patient groups and their relatives. In a survey, 35 % of the 3008 people from the general population reported that they had some kind of information need while almost half (48 %) declared no specific needs [26]. In another survey, Geraedts and Amhof showed that there was a high demand for information on the quality of health care providers in the general population [11]. Additionally, a total of 89 and 86 % of the surveyed sample wanted   To analyze gender differences in the demand for quality related information on health care providers.
Cross-sectional design. Representative sample of German adults (N = 1523); Health Survey using 5 point Likert scales assessing the demand for quality-related information on physicians and health care providers (10 items) and hospitals (33 items  Not specified a priori. At the beginning of in-patient rehabilitation period, carers requested mainly disease-     Table 2 Overview of extracted studies (Continued) Thon/Ullrich (2009) [24] To assess sources of information and information needs in parents of children with a rheumatic disease.   Younger people and people who received no medical care had significantly higher needs for information on prevention and health promotion. Gender differences were mentioned but were not obvious.
• age • gender • patient status more information on the quality of medical specialists and hospitals, respectively. Analysing the information needs of young women regarding combined oral contraceptives the mean interest rating of all 25 potential counseling items was 5 on a 7-point-scale. All items had a rating of 4 or higher [25]. Similar results were observed in a study about consumers' information needs for informed choice on colorectal cancer screening, where participants rated all potential topics as relevant [21]. Identifying subgroups of information needs among cancer patients using latent class analysis (LCA) Neumann et al. found that 68.6 % of the patients had some kind of unmet information needs [17]. Vogel et al. explored information needs in breast cancer patients and found them to be time-dependent [24]. In a study on the demand for CAM among cancer patients, almost half (48 %) of the interviewed cancer patients demanded consultation about CAM irrespective of whether they already used CAM or not [10]. Only 24 % were not interested in CAM. In a qualitative survey, family members of stroke patients reported a need for detailed disease-specific information by the medical staff in a personal conversation during inpatient rehabilitation [14]. Among parents of children diagnosed with a disease, parents of children with rheumatic diseases had high interest in further information, although they considered themselves as being well informed [22]. Almost all parents (85 %) of children with JIA considered detailed information on the disease as very important. The mean value for importance was 9.52 on a 10-point-scale [23]. A total of 58 % of the parents had specific information needs for at least one theme. One fifth (22 %) did not indicate a specific theme and only 20 % stated that they did not have any specific information needs.

Factors associated with information needs
Associations between information needs and other variables or characteristics were analysed in eleven studies ( Table 2) [9-12, 14, 17, 22, 23, 25-27]. We found analyses concerning the association between information needs and gender, age, education, duration and course of the disease, knowledge, emotional factors (e.g. anxiety, depression) and the relationship between patient and health professionals.
In two studies, women expressed higher information needs than men [11,14]. Assessing the need for qualityrelated information on health care providers, significant gender differences were found in 17 of 41 items [11]. In this study, gender was most strongly associated with the need for information among all considered sociodemographic variables. However, in other studies analysing information needs of cancer patients [17] and patients with inflammatory bowel diseases [9], hardly any gender differences were found.
There is some evidence that age influences type and extent of information needs. Conrad et al. [9], Eustachi et al. [10] and Neumann et al. [17] found higher information needs in younger patients. A trend for higher information needs in younger people was also observed in a survey of the general population [26]. In addition, younger people showed higher interest in disease prevention and health promotion [26]. Information needs of parents of children with JIA differed depending on the age of the child [23].
No conclusive association was found between information needs and education [9,25], as hardly any significant differences were found between people with different levels of education.
Regarding the duration of disease, a negative correlation between information needs and the duration of an illness was found in two studies [9,17]. The course of disease had a significant impact on the amount or type of information needs in several assessed studies. Neumann et al. stated that cancer patients in early disease stages had more unmet needs for information on results of medical examinations and treatment options, whereas patients in more progressive stages had more unmet needs for information on social issues and health promotion [17]. Vogel et al. have observed decreasing scores of information needs for all topics except 'aftercare' in the course of breast cancer treatment. Only one specific topic showed no significant change (p > 0.05) of interest in the course of treatment ('examination and medical test') [24].
Cancer patients with a poorer health status had higher information needs about CAM [10]. Heesen et al. detected considerable differences in information needs between primary-progressive (PP) and relapsing-remitting (RR) patients with multiple sclerosis. RR patients expressed a higher need for information regarding magnetic resonance imaging, relapses and CAM, while PP patients were more interested in the treatment of gait disturbances, physiotherapy and experimental therapies [12]. Patients with inflammatory bowel diseases had significantly higher information needs with increasing disease activity scores [9].
For the association between knowledge and information needs ambiguous results exist. According to Ullrich et al. parents of children with JIA had lower information needs when they had higher knowledge scores and were more satisfied with the information provision [23]. Likewise, cancer patients who felt informed had significantly lower information needs about CAM [10]. In a small sample of young women, however, the relationship between interest and knowledge ratings about risks and benefits of combined oral contraceptives showed no clear trends [24]. Assessing the information needs in parents of children with a rheumatic disease, the level of prior information (self-rating) and the interest in further information were compared [22]. All topics which were covered by prior information were nonetheless of high interest. No response pattern for topics covered by prior information and low current interest was detected. Patients with multiple sclerosis and knowledge scores in the upper quartile rated the item 'validity of studies' among the first three of interest while the remaining patients did not mention this item among the first ten [12].
Anxiety, depression and the requirement for psychological support were correlated with increased needs for information in three studies [9,17,27].
Patients had less unmet need for information when they were satisfied with their actual treating physicians [23], trusted their nurses highly, received more caring attention from nurses or were treated empathically by their physicians [17].

Discussion
This systematic review investigated the methodologies used to assess information needs, and identified findings from existing studies regarding topics of healthrelated information needs, the extent of health-related information needs, and associations between diseaserelated characteristics or other variables and the desire for more information or the need for specific information about single topics.

Methodological aspects of the assessment of information needs in the identified publications
Studies on information needs are rare for many diseases (except for cancer) but a recent increase in publications indicates rising interest.
In the identified articles a wide range of methods was described: qualitative research employing focus groups [21], quantitative methods using questionnaires and Latent Class Analysis (LCA). The most frequently used measure to assess the need for information was the Likert scale [10, 11, 15, 21-23, 25, 27]. Some methodological issues regarding the usage of Likert scales in this context have been discussed [28,29]. Ratings for every single item may lead to biased results because participants tend to rate every item as important [30]. Hence, mean values of single information categories do not differ much within single studies and tend to reach high levels. Such ceiling and cluster effects can also be found in several of the assessed studies in this review [14,18,24,26] and make it challenging to identify specific topics of interest. Even if high ratings for all defined topics reflect the true preferences, individual information need should guide health care professionals in delivering the most important or urgent information in the limited consultation time [31]. Therefore, studies assessing information needs should include a prioritization of the topics by the participants. Prioritization can be achieved in several ways. One opportunity is to let them select a limited number of relevant topics out of a list of predefined topics [22]. Another way to avoid ceiling effects is the use of differential scaling to rate relative levels of importance [30]. Choi et al. have tested the consistency of breast cancer patients in prioritizing information needs making paired comparison judgments, another method to assess the priority of information needs [31]. However, the number of comparisons ascends considerably with a rising number of predefined topics.
Another potential source for differences in the results was observed between the usage of scales and other methods. The kind of questions differed in most cases. In scaling tasks participants were often asked to rate how important [13,14,30], interesting [18] or relevant [24] different topics are in general. Using other methods (e.g. open-ended questions, dichotomous questions) participants usually expressed their personal need for further information regarding specific topics [12,16,17,21,26]. Perhaps people rate topics as important, interesting or relevant but do not need further information because they are already well-informed or satisfied with information provision.

Topics and extent of information needs, and associated factors
Most frequently described information needs were related to predefined topics such as treatment, diagnosis, prevention and health promotion, etiology and prognosis. However, the same topics were explicitly specified as items in qualitative studies [16,17,23,24,26].
Information needs were found in all kinds of surveyed patients as well as in samples of the general population. They were found for all the above mentioned more general topics. Regarding more specific topics such as CAM and nutrition, they were found to be of high interest in every survey containing these predefined topics. However, information about the medical system, financial and legal issues, rehabilitation and impact on social life was desired less frequently.
Several variables were found to be associated with the extent and type of information needs. A younger age, shorter duration of disease, poorer health status or higher degree of severity of a disease, higher anxiety and depression scores appear to be associated with higher unmet information needs in at least two studies, respectively. The results for the association with gender, education level and knowledge were less consistent. Knowledge or the amount of prior information does not seem to be a reliable explanatory variable for the need/desire for further information. Two studies reported that patients with higher knowledge had fewer information needs [10,23], while other studies detected no association between knowledge and information needs [22,24]. Emotional traits and conditions like trust, attention of nurses or the empathy of the health care professional had a strong association with the need for information [9,17,27]. There is some evidence of a change in the type and extent of information needs over the course of disease [26,32].

Comparison with other reviews
Recent reviews were restricted to certain diseases, mainly cancer, and to patients and relatives [4][5][6]33]. According to our review, they found high information needs, in particular for general topics as treatment. Rutten et al. and Puts et al. also analysed the association between sociodemographic factors and information needs, and reported higher information needs in younger people in the included studies, in line with our finding [5,6].

Limitations
Our review has some limitations. We have no information about the specificity and sensitivity of our geographical search filter. In addition, the definition of information needs is ambiguous in relation to the terminology used. Thus, we cannot preclude having missed relevant studies. Furthermore, we did not use a definition for "health". The choice of our databases might have resulted in a stronger focus on medical and psychological dimensions, while neglecting social dimensions of health. However, we did not experience any problems in deciding whether a study dealt with health information needs or not.
The extension of our search strategy to other databases might have yielded additional studies for inclusion. For example, CINAHL is commonly used by qualitative researchers. We also did not search for grey literature.
We only searched studies until August 2012. Further relevant studies meeting our inclusion criteria might have been published after this search date. However, following the idea of qualitative research, we feel that we reached a point of data saturation with respect to the methodological aspects of studies investigating information needs and further studies are unlikely to change our conclusion unless the methodological approaches change.
The evidence synthesis and the interpretation of our findings is hampered by the inclusion of very heterogeneous studies in terms of their objectives, methods and the investigated diseases.
There is some evidence that patients' participation preferences, their satisfaction with health care or care provision, patients' involvement in care decisions and their need for information differ between countries and health care systems [28,34]. As we included studies performed in Germany, the transferability of our findings to other countries may be limited. We compared our findings to those of the few existing reviews regarding IN. However, we are not aware of any other reviews or relevant studies for direct comparison with our nation-focused results.
We did not assess quality of the included studies due to a lack of an established and validated critical appraising tool for the included study designs. However, we have considered methodological issues in the discussion section.

Conclusion
The consideration of individual information needs is needed to provide tailored information. This is e.g. a crucial precondition of shared decision making.
Empirical findings regarding patients' information needs are lacking for the most widespread diseases in Germany. However, our study found a considerable extent of information needs in all assessed target groups. Apart from common pre-defined topics of information needs, e.g., treatment and diagnosis, etiology, researchers, care providers and developers of patient information and decision aids should consider a wide range of potential individual topics when assessing information needs or providing information to patients, relatives, and the general population. In addition, individual characteristics should be considered when providing information. We found no study investigating the preferred formats for health-related information.
Concerning methodology, studies should not rely exclusively on ratings of predefined topics but should also ask open ended questions, seek for prioritization and apply techniques appropriate to discriminate between people's needs for the most relevant information and different disease specific topics.