Information needs in people with diabetes mellitus: a systematic review

Background The purpose of this study was to identify and analyse currently available knowledge on information needs of people with diabetes mellitus, also considering possible differences between subgroups and associated factors. Methods Twelve databases including MEDLINE, EMBASE and the Cochrane Library were searched up until June 2015. Publications that addressed self-reported information needs of people with diabetes mellitus were included. Each study was assessed by using critical appraisal tools, e.g. from the UK National Institute for Health and Care Excellence. Extraction and content analysis were performed systematically. Results In total, 1993 publications were identified and 26 were finally included. Nine main categories of information needs were identified, including ‘treatment-process’, ‘course of disease’, ‘abnormalities of glucose metabolism’ and ‘diabetes through the life cycle’. Differences between patient subgroups, such as type of diabetes or age, were sparsely analysed. Some studies analysed associations between information needs and factors such as participation preferences or information seeking. They found, for example, that information needs on social support or life tasks were associated with information seeking in Internet forums. Conclusion Information needs in people with diabetes mellitus, appear to be high, yet poorly investigated. Research is needed regarding differences between diverse diabetes populations, including gender aspects or changes in information needs during the disease course. Systematic review registration The review protocol has been registered at Prospero (CRD42015029610). Electronic supplementary material The online version of this article (10.1186/s13643-018-0690-0) contains supplementary material, which is available to authorized users.


Background
Diabetes mellitus (DM) is a highly prevalent chronic disorder. People with DM have to perform comprehensive self-management interventions to achieve good diabetes control [1]. In order to make adequate decisions concerning their illness, a sufficient level of disease-related information is required [2]. In fact, people with DM communicate a particularly high need for information, higher than people with cancer or cardiovascular diseases, for example [3,4]. However, it seems that people with DM do not feel adequately informed about their condition or regarding medication use [3]. Although the importance of an appropriate needs-driven information supply is unquestioned, and a large amount of diabetes information exists, there seems to be limited knowledge about information needs (IN) of people with DM considering different patient subgroups, as well as IN of people with DM in phases of the disease that may affect the need for certain information [5][6][7]. To provide needsdriven information, deeper insight into the perspectives of people with DM is urgently needed. This is particularly true since it has been suggested that information supply, self-management and health outcomes could be improved if more were known about the perspectives and needs of those concerned [8,9].
This systematic review aims to identify and analyse currently available knowledge on the IN of people with DM, also considering possible differences between subgroups and associated factors.

Methods
This systematic review was performed in line with the quality requirements of the PRISMA guideline (available as Additional file 1) [10]. The review protocol has been registered at PROSPERO (CRD42015029610).
As stated in the review protocol, we searched in MEDLINE, EMBASE, CINAHL, ScienceDirect, the Cochrane Library, Web of Science, PsycINFO, CCMed, ERIC and Journals@OVID, Deutsches Ärzteblatt and Karlsruher virtueller Katalog. Publications were included that had been published from the inception of each database up to June 2015 (see Appendix 1 and Appendix 2) [11] with a German or English title and abstract and a full text in any language.

Search strategy
The search strategy was set up using database-specific vocabularies (MeSH, EMTREE) and additional freetext terms (see Appendix 1 and Appendix 2) [11]. The search algorithm was crosschecked by experienced reviewers (S.K., A.S.) and piloted by comparing results of the search strategy in MEDLINE with core references that were identified by pre-search activities. Search terms for IN included ʻinformation needʼ, ʻknowledge needʼ, ʼwish or desire of informationʼ, ʻinformation preferenceʼ and 'request for information'. Search terms for DM included: ʻdiabetesʼ, ʻdiabeticʼ, ʻniddmʼ, ʻiddmʼ, ʻt2dmʼ, ʻt1dmʼ, ʻprediabetesʼ, ʻprediabeticʼ, ʻpre-diabetesʼ, ʻpre-diabeticʼ and ʻimpaired glucoseʼ. For further details see Appendix 1 and Appendix 2.

Inclusion and exclusion criteria
Studies that analysed self-reported IN of people with DM (any type) as a primary or secondary research aim were included. IN are defined as: ʻRecognition that their (people's) knowledge is inadequate to satisfy a goal, within the context/situation that they find themselves at a specific point in the time [12].' Original qualitative, quantitative or mixed-methods studies were included. Systematic and narrative reviews, meta-analyses and qualitative meta-syntheses were also included.
Studies reporting the IN of relatives or healthcare professionals were excluded, as well as studies where relatives or healthcare professionals reported IN of people with DM. Publications without available references, letters/short reports, abstracts, editorials, comments or discussion papers were excluded.

Study selection process
Inclusion and exclusion criteria were pre-tested on 380 records and finally discussed (L.B., J.G., S.K.). Then, two reviewers (L.B., J.G.) independently selected the articles, first by title and abstract and thereafter by full text. All decisions were checked by two other reviewers (S.K., M.R.). Unclear decisions were resolved by an additional reviewer (A.I.).
Full texts were screened with the aim of identifying further original studies using backward citation tracking.

Data extraction and synthesis
A data extraction sheet was developed following the requirements of Cochrane [10].We extracted the type of information needed by people with DM and, if investigated, by different subgroups, such as type of DM and age. A content analysis was conducted, developing categories according to the topics of the review questions, in particular, to assess the reported and analysed types of information needed. Furthermore, IN-associated factors were extracted and analysed via a content analysis. Codings were developed inductively (L.B.) [13] using a coding protocol and revised critically (S.K., A.I.). All relevant publications were described according to the following predefined categories: author, date, methods, findings, associated factors and result of the critical appraisal of the study quality. Furthermore, the main categories of IN were described, as well as those related to subgroups of people with DM and the associated factors. We describe the studies stratified for those analysing IN as a primary and a secondary outcome, as well as those with a qualitative, quantitative, or mixedmethod study design.

Critical appraisal
Each study was critically appraised separately (L.B., S.K., J.G., M.R.) by using design-specific critical appraisal tools from the UK National Institute for Health and Care Excellence (NICE) [14]. The study's quality was described as follows: '(++) All or most of the checklist criteria have been fulfilled, where they have not been fulfilled the conclusions are very unlikely to alter. (+) Some of the checklist criteria have been fulfilled, where they have not been fulfilled, or not adequately described, the conclusions are unlikely to alter. (-) Few or no checklist criteria have been fulfilled and the conclusions are likely or very likely to alter.' [14]. Mixed methods were analysed by the Mixed Methods Appraisal Tool (MMAT)-Version 2011 [15]. The critical appraisal for mixed-methods studies includes whether the mixedmethods design was appropriate and whether the integration was relevant to address the research question (objective). The criteria also consider whether limitations are considered, associated with this integration, e.g. whether the divergence of qualitative and quantitative data (or results) in a triangulation design was appropriate [15].

Results
In total, 1993 publications were identified that had been published up to June 2015 ( Fig. 1), of which 26 publications (n = 25 studies) reporting diabetes-related IN of people with DM were finally included ( Table 1). The sample sizes varied from 11 to 1609 participants with DM.
The critical appraisal showed that three of the 25 identified studies met all or most of the NICE checklist criteria. The other studies fulfilled some (n = 14) or a few criteria (n = 8). It was noticeable that within the qualitative studies, most of the criteria were fulfilled but eight of 14 studies did not describe the role of the researcher sufficiently, and six gave no indications concerning ethical approval. None of the included quantitative studies reported how selection bias was minimised, and included studies using mixed-method design reported little about the quantitative part of their study  The study explored the degree of blood pressure control and identified the need for more information about high blood pressure. pp: "Indicates that for that particular aspect of study design, the study has been designed or conducted in such a way as to minimise the risk of bias" p: "Indicates that either the answer to the checklist question is not clear from the way the study is reported, or that the study may not have addressed all potential sources of bias for that particular aspect of study design" m: "Should be reserved for those aspects of the study design in which significant sources of bias may persist" NR (not reported): "Should be reserved for those aspects in which the study under review fails to report how they have (or might have) been considered" NA (not applicable): "Should be reserved for those study design aspects that are not applicable given the study design under review (for example, allocation concealment would not be applicable for case-control studies)" (NICE 2012) * IN focused on a special topic a Data for age, sex and sample size only for participants affected by DM design. The results of the critical appraisals are shown in Table 1.

Content of information needs
The content analysis (Table 2) identified nine main types of IN of people with DM and 28 subtypes. The main categories are 'treatment-process' , 'course of disease' , 'abnormalities of glucose metabolism' , 'diabetes through the life cycle' , 'pathophysiology of diabetes' , 'research' , 'coping' , 'support' , and 'prevention'. IN on the 'treatment process' were reported most frequently throughout the studies, in particular, 'medication' (n = 12), 'diabetes self-management' (n = 11) and 'nutrition' (n = 11). IN regarding the 'course of disease' were the second highest reported, particularly 'consequences of diabetes' (n = 16), e.g. consequences concerning physical health, lifestyle and social life. Only four studies reported IN on 'coping' and 'support' , and two reported on 'prevention'.

Information needs in subgroups and factors associated to IN
Specific comparisons between subgroups or analysis of associated factors using, for instance, regression models were performed minimally in the identified studies. Only one mixed-method study made a comparison between type 1 DM (T1DM) and type 2 DM (T2DM) [3]. Hence, we could only try to find subgroup-specific IN from studies that addressed certain groups as people with T1DM or T2DM or women with gestational DM (GDM). Only two studies investigated related factors such as, for example, socio-economic status [4,8,9]. Duggan et al. was the only study that performed quantitative statistics. The authors found, for example, that higher socio-economic status was positively correlated with the need for drug information [4]. More complex factors such as concepts like participation preferences or seeking behaviour were analysed more frequently than often-investigated associated factors such as age and sex, and these analyses were predominantly performed in qualitative studies, in particular in the study by St. Jean [8,9] and Whetstone [23]. In the following, we describe the main findings.
Only one mixed-method study made a comparison between T1DM and T2DM and showed that contents of patient-identified concerns between these subgroups are similar, apart from ʻnot knowing enoughʼ (T1DM) and ʻconcerns about futureʼ (T2DM) [3]. Overall, no striking differences between the IN of people with different types of DM were identified or noticeable in one certain group. Almost all the studies, with the exception of studies involving people with GDM and maturity onset diabetes of the young (MODY) [22,27], reported IN in the categories 'treatment process' , 'course of disease' , 'pathophysiology of diabetes' ,'research' , 'coping' ,'support' and 'prevention'. With the exception of the studies involving people with T1DM, all the studies reported IN regarding 'abnormalities of glucose metabolism'. Furthermore, 'diabetes through the life cycle' was addressed in most of the studies, but not in the study including people with MODY.

Information needs and age
Overall, only a few differences of IN between people with DM in different age groups were identified in the studies; however, some were reported. Young people with T1DM were particularly interested in 'diabetes through the life cycle' , e.g. 'pregnancy'. There were two populations of young people identified: children with mean age 10 to 13 [19] and adolescents and young adults aged between 14 and 25 [18,22]. Several IN were reported by both groups; however, there were also differences: the study by Olsen Roper et al. reporting IN of children between 10 and 13 years old showed that the population of children was particularly interested in the topic 'course of disease' , especially 'cure' issues [19]. Additionally, they were interested in 'abnormalities of glucose metabolism' , particularly 'pathogenesis' and 'aetiology' of diabetes [19]. In contrast, the population of adolescents and young adults was interested in 'diabetes through the life cycle' , particularly in 'puberty' [18], 'family founding' [22] and 'pregnancy' [22].
Older people with DM were interested in 'treatment process' , the 'course of disease' and the 'abnormalities of glucose metabolism'. A further differentiation in age groups, e.g. 'elderly' and 'very old' , cannot be derived from the studies.

Information and participation preferences, knowledge about and experience with diabetes
Three studies addressed IN in association with participation preferences. All three studies analysed IN in relation to decision-making and involvement in the decision process [3,8,9,25]. St. Jean pointed out that an involvement in decision-making is related to higher ratings for the usefulness of information [8,9]. Two studies addressed IN in association with ʻknow-ledgeʼ and ʻ(feelings about) diabetes experienceʼ [8,9,19]. Knowledge, diabetes experience and IN are strongly related. Therefore, both studies recommended identifying the individual level of knowledge and whether this is correct [8,9,19]. Furthermore, clear, optimistic, lessuncertain feelings, or support in diabetes experience were associated with different IN, e.g. 'diabetes management' , 'causes of diabetes' , 'diabetes-related complications' [8,9].

Stage of the disease
One study that focused on people with DM who had recently started treatment with oral anti-hyperglycaemic drugs addressed IN in association with the stage of the disease. It pointed out that people with DM who recently started treatment with anti-hyperglycaemic drugs are in need of diabetes medication information such as drug-related issues [16].

Discussion
This is the first systematic review of studies dealing with IN of people with DM. We identified 25 studies (26 publications). This is a limited number compared with, say, cancer, where a large number of quantitative and qualitative IN studies (n = 112) already exist [37]. This is surprising, since it is known that people with DM have a higher or similarly high need for information compared with people with other chronic diseases [3,4]. Looking for the content of IN, it was comparable to those found in people with cancer, such as ʻprognosis of diseaseʼ, ʻdiagnostic testsʼ, ʻtreatmentʼ, ʻself-careʼ, and ʻemotional and psychological needsʼ [37]. However, Duggan et al. found that people with DM have a higher need for information about drugs than people with cancer or cardiovascular disease [4].
It would be interesting to look for differences between people with diabetes and cancer or other chronic diseases in more detail.
Besides the low number of studies, it became obvious that differences between patient groups such as male and female patients, different age groups or types of diabetes have not been analysed so far. Factors associated with IN are rarely investigated. In cases in which analysis was performed, mainly the more complex factors such as participation preferences or information seeking were investigated, and it was particularly done in qualitative studies, which were highly heterogeneous. Also, changes in IN during the course of the disease are poorly investigated, although they may be expected.
Instruments for collecting IN data from individuals with DM have not been validated. Similar results were identified in cancer studies and showed that only a minority of instruments for the collection of IN are validated [37].

Implications for research
There is a need for research on several levels. First, compared with other chronic diseases, there is a huge lack of studies addressing IN of people with DM, although DM is one of the largest public health issues [38]. Due to the low number of studies dealing with unmet needs, and considering the relevance of this topic mentioned in the studies, there is a need for further research. Second, differences between the types of DM populations including gender aspects and patient characteristics should be analysed. Third, there is a need for research to show associations of variables with IN and to amend the findings from qualitative studies via qualitative analyses. Available knowledge about the IN and associated factors and concepts can be used in targeted counselling and to strengthen the health literacy of people with DM. Finally, methods and instruments should be further developed against a theoretical background and validated.

Limitations
We conducted a comprehensive and sensitive search that was also pre-tested. The study selection and critical appraisal were performed by two reviewers. Two other reviewers checked all the decisions. The critical appraisals were performed for each study design. However, there were still difficulties to provide a clear decision based on the reports.
The identified studies have several limitations: only three qualitative studies met all the quality criteria. None of the quantitative studies met all the criteria. The comparability of the IN categories is restricted because the IN were collected by different study designs. Some of them were collected in the context of a specific health-related topic or by using different methods or instruments.

Conclusion
There is a limited number of studies analysing IN in DM, and there is a low number of studies investigating differences between subgroups of DM populations, including gender aspects or changes of information needs during the disease. This should be further investigated.