First results from the study ‘Disease knowledge and information needs - Diabetes mellitus (2017)’

Rebecca Paprott; Christin Heidemann; Lena M Stühmann; Jens Baumert; Yong Du; Sylvia Hansen; Marie-Luise Zeisler; Johannes Lemcke; Silke Beyhl; Ronny Kuhnert; Christian Schmidt; Lars Gabrys; Andrea Teti; Thomas Ziese; Patrick Schmich; Paul Gellert; Daniela Zahn; Christa Scheidt-Nave

doi:10.17886/RKI-GBE-2018-064

. 2018 Jun 6;3(Suppl 3):22–60. doi: 10.17886/RKI-GBE-2018-064

First results from the study ‘Disease knowledge and information needs - Diabetes mellitus (2017)’

Rebecca Paprott ¹, Christin Heidemann ¹, Lena M Stühmann ², Jens Baumert ¹, Yong Du ¹, Sylvia Hansen ³, Marie-Luise Zeisler ¹, Johannes Lemcke ¹, Silke Beyhl ¹, Ronny Kuhnert ¹, Christian Schmidt ¹, Lars Gabrys ^1,⁴, Andrea Teti ⁵, Thomas Ziese ¹, Patrick Schmich ¹, Paul Gellert ², Daniela Zahn ³, Christa Scheidt-Nave ^1,^✉

PMCID: PMC8852783 PMID: 35586544

Abstract

Very little research has been undertaken into what people in Germany know about diabetes, the information they may require about the condition, where they look for such information and how they rate the information currently available. In 2017, the Robert Koch Institute (RKI) carried out a nationwide telephone survey aimed at answering these questions. The study entitled ‘Disease knowledge and information needs - Diabetes mellitus (2017)’ focused on people aged at least 18 years. A total of 2,327 people without diabetes and 1,479 people with diagnosed diabetes were interviewed for the study. First results show that 56.7% of people without diabetes and 92.8% of those with diabetes rate their knowledge about the condition as ‘very good’ or ‘good’. People without diabetes were found to have the strongest need for information in terms of ‘lifestyle changes, health promotion and disease prevention’, whereas respondents with diabetes stressed the strongest need for information about ‘treatment and therapy’. Almost a third of respondents without diabetes have actively sought information about diabetes at least once, mostly via print media. Patients with diabetes stated that their general practitioner was their most frequent source of information about the condition. In both groups, about half of respondents reported that they found it difficult to judge the trustworthiness of the information published in the media about diabetes. The results of the study form part of the German National Diabetes Surveillance, which is coordinated by the RKI. The data are also intended to be used by the Federal Centre for Health Education to develop a strategy to improve the information provided about diabetes.

Keywords: DIABETES MELLITUS, TELEPHONE SURVEY, DISEASE KNOWLEDGE, DISEASE PERCEPTION, INFORMATION-SEEKING BEHAVIOUR

1. Introduction

Diabetes mellitus is a chronic metabolic disorder characterised by elevated blood-sugar levels [1]. The condition can be divided into a number of different types. Type 1 diabetes is an autoimmune disease that is thought to be primarily caused by a genetic predisposition and environmental factors [2, 3]. In contrast, genetic but also individual, social and contextual factors such as the person’s living environment or health care conditions as well as the way in which these factors interact with one another play a role in the development of a much more common form: type 2 diabetes [4-6]. A family history of diabetes, lifestyle and individual patterns of behaviour such as smoking, a lack of physical activity and dietary aspects are associated with a higher risk of type 2 diabetes [4-6].

Info box 1: Study ‘Disease knowledge and information needs - Diabetes mellitus (2017)’.

Data holder: Robert Koch Institute

Objectives: To provide reliable information from adults in Germany about their perception of the risk associated with diabetes mellitus, their perception and knowledge of the condition, their information needs and information-seeking behaviour, the subjective burden of diabetes and the quality of care

Survey method: Telephone interview

Population: The German-speaking general population aged at least 18 years with or without a diagnosis of diabetes living in private households who were reachable by telephone

Sampling: Individual private households in Germany that are reachable using randomly selected landline or mobile telephone numbers

Sample size: 2,327 people without diabetes (1,313 women, 1,014 men) and 1,479 people with diagnosed diabetes (740 women, 739 men)

Response rate: 17.9% (of all households reachable by telephone in Germany)

Survey period: September - November 2017

Data protection: the participants were informed about the study’s aims and content and provided their informed consent to participate in the study

More information in German is available at:

www.rki.de/krankheitswissen_diabetes

Individual patterns of behaviour can be influenced by a variety of subjective factors which may play a key role in the condition but have often been under-studied. These factors include an individual’s subjective perception of their risk of developing diabetes [7], their knowledge about the disease [8], their perception of the condition [9] and their health-related information behaviour [10, 11]. In the context of this study, health-related information behaviour refers to the issues of whether people actively seek information about health related topics and, if so, the sources of information that they turn to. Ultimately, contextual factors such as social support and the conditions associated with care also affect people’s risk of developing diabetes [12]. Moreover, international studies have shown that the perception of the risk associated with complications linked to diabetes, the way in which people deal with the condition, the social support they receive, as well as various aspects linked to care and access to the information required to deal with the disease can influence the way in which the disease develops over time [13-19].

According to data from the Robert Koch Institute (RKI), an estimated 6.7 million adults in Germany are affected by diagnosed or undiagnosed diabetes [20, 21]. Although diabetes can be treated with medication, the disease is still associated with serious complications and increased risk of mortality [22, 23]. In addition, diabetes also incurs a large amount of direct and indirect costs to society, mainly because of these complications [24]. In order to tailor healthcare services towards people’s needs and requirements, it is essential that data be collected on the burden caused by the disease, on people’s knowledge about the condition, and the need for information and support among people with diabetes. Furthermore, data from the general population - in particular on knowledge about the disease, disease perception and, above all, information needs and information-seeking behaviour - are vital if measures in population-wide health promotion and the primary prevention of type 2 diabetes are to be effective.

At the present time, only a limited number of population-representative data are available on these issues at the national level for people with or without diagnosed diabetes [25-29]. Moreover, the data on information-seeking behaviour and information needs [26, 28] is fragmentary and only a small proportion of participants with diabetes took part in these studies [26, 29]. The ‘Disease knowledge and information needs - Diabetes mellitus (2017)’ study (Info box 1), therefore, aimed to close these gaps in the data. In order to do so, the study collected data on knowledge about the disease, disease perception, the risk of diabetes and the risk associated with diabetes-related complications, subjective aspects of the disease as well as specific diabetes-related information (Info box 2). This article aims to set out the study’s first descriptive findings for the key topics of knowledge about disease, disease perception and diabetes-related information from people with and without diabetes.

Finally, the study also provides a foundation with which to develop the educational and communications-related strategy on diabetes that is currently being drawn up by the Diabetes Office of the Federal Centre for Health Education (BZgA). This strategy is to reflect people’s current needs and will be tailored to specific target groups [30]. The goal is to improve diabetes prevention, including primary prevention (reducing the incidence of diabetes), secondary prevention (early detection of previously unknown diabetes) and also tertiary prevention (prevention of the progression of diabetes). Furthermore, the results of the study have been integrated into the National Diabetes Surveillance which is currently being established at the RKI [31].

Info box 2: Diabetes-related information.

Diabetes-related information includes the following aspects:

Perceived level of information

How informed do respondents feel about different diabetes-related topics?
Subjective information needs

Which diabetes-specific topics do respondents require more information about?
Information-seeking behaviour

Are respondents actively looking for information and, if so, which sources do they turn to?
The perceived trustworthiness and clarity of information about diabetes published in the media

2. Methods

2.1 Study design and sample

The study ‘Disease knowledge and information needs - Diabetes mellitus (2017)’ (Info box 1) was undertaken in two parts that were conducted using a standardised telephone survey (Figure 1). The first part of the study focused on the German-speaking general population aged at least 18 years (including people with diabetes); the second part only included people with diabetes. Sampling was undertaken using the ADM Telephone Survey System, which is provided by Arbeitskreis Deutscher Markt- und Sozialforschungsinstitute e.V. (ADM) [32]. This system can be used together with a special procedure to generate all usable German landline and mobile phone numbers. This even includes numbers that are not registered in public telephone directories. The study applied the dual-frame principle, which provided a sample of 60% landline numbers and 40% mobile telephone numbers. This method ensured a representative sample of all potentially reachable private households in Germany. A detailed description of the methodology used by the RKI for telephone interviews and the sampling applied in this case will be provided in an article that is due to be published in 2018 in the Journal of Health Monitoring [33].

Flowchart for sampling in the study ‘Disease knowledge and information needs - Diabetes mellitus (2017)’

Own diagram

Interviewees (target persons) for the first part of the study were randomly selected from multi-person households using the Kish Selection Grid method. For this procedure, all potential target persons have the same probability of selection. The number of adults in the household in question and their age are first determined in order to draw up a particular sequence of these individuals. An algorithm is then employed as part of computer-assisted random selection to select the target person for interview. In the case of this study, the aim was to gain a sample of 2,500 people without diabetes with which to assess the different groups within the general population and to provide stratified analysis with the lowest possible level of error tolerance.

The second part of the study focused on people with a physician diagnosed diabetes. The aim was to gain a sample of 1,500 individuals (including those with diabetes from the first interview part) also in order to ensure that stratified analysis had the lowest possible tolerance of error. The study was implemented using a direct screening procedure to find participants for the category ‘people with diabetes’. In these cases, the person who answered the call (the contact person) was asked whether they or another person in the household had ever been diagnosed with diabetes by a physician (the target person). If a household had more than one potential target person, in other words, more than one person with a physician diagnosed diabetes, the target person was randomly selected using the Kish selection grid [33].

2.2 Study procedure

Two questionnaires were developed for the study: one for people with a physician diagnosed diabetes, and one for those without. The two questionnaires deliberately used overlapping questions so that the results for people with diabetes could be compared to those without the condition. Wherever possible, the questions were selected using validated German-language instruments. In case only English-language instruments were available they were translated into German before translating them back into English (forward-backward translation) [34, 35]. The translations were shown to be very similar to the original texts.

A number of further questions had to be newly developed for the study to cover other topics of interest. A detailed description of this process and the psychometric properties of the selected constructs and instruments is currently being summarised for a methodological publication. Cognitive testing was conducted to assess how understandable the questions and certain terminology were. Furthermore, in order to assess the completeness of the questionnaires, find out how long the interviews would take, and address any remaining ambiguities, a standard pretest was carried out in the field between the 23 August and 17 September 2017 (203 participants: 28 with diabetes, 175 without). The results from the pretest were used to shorten the questionnaires to an average of 32 minutes (for respondents without diabetes) and 43 minutes (for respondents with diabetes) and to adapt the questions.

The first part of the main study was conducted between 18 September and 31 October 2017, with the second taking place between 1 November and 30 November 2017. 2,592 interviews were conducted for the first part of the study (the representative part). This figure also includes 263 people who had been diagnosed with diabetes. The response rate, which was calculated using criteria from the American Association for Public Opinion Research (AAPOR) (in this case: response rate 3, proportion of interviews conducted in relation to all probable households in the population) was 17.9% [36]. 1,216 interviews were conducted for the second part of the study.

Two respondents were excluded from the first part of the study due to an unclear answer to the question ‘Have you ever been diagnosed with diabetes by a doctor’ (‘don’t know’) or missing information about the federal state of residence (this was needed to calculate the weighting factors). The final sample, therefore, consisted of 2,327 people without a diagnosis of diabetes and 1,479 people with a physician diagnosed diabetes.

2.3 Definition of diabetes

Participants with diabetes were identified using the question ‘Have you ever been diagnosed with diabetes by a doctor?’. This group was further subdivided into: 1) Diabetes in the last 12 months: in this case, a participant either answered ‘Yes’ to the follow-up question as to whether they had had diabetes in the last 12 months, or stated that they were undergoing medical treatment for diabetes (n=1,396); 2) No diabetes in the past 12 months: the respondent answered ‘No’ to the question about whether their diabetes had been present in the last 12 months, and also provided no indication that they were currently undergoing medical treatment for diabetes (n=89); 3) Gestational diabetes: the respondent answered ‘Yes’ to the question as to whether they were pregnant at the time of their diagnosis (n=52). A current case of gestational diabetes was identified by asking whether diabetes had been diagnosed during a current pregnancy (n=1). Of the female participants who had ever been diagnosed with gestational diabetes, twelve women had type 2 diabetes at the time of the study.

2.4 Concepts and instruments

The study collected information on disease knowledge, disease perception, diabetes risk, the risk of complications, diabetes-related information, health-related behaviour and lifestyle, health care services utilisation, the characteristics of disease, diabetes self-management, the subjective burden of disease and mental health. An overview of all of the subject areas and instruments employed in the survey can be found in Annex 1. This article presents the first results for selected topics, including disease knowledge, disease perception and diabetes-related information.

Survey instruments used for respondents without diabetes

In order to study knowledge about disease, data was collected about the respondents’ perceived level of knowledge about diabetes [37] and their objective level of knowledge about the disease. Annex 2 provides a list of the questions that were posed and acceptable response options.

The topic of disease perception included an assessment of whether the respondents’ considered diabetes to be a serious disease [38], and whether they believed that diabetes was a condition that people had for the rest of their life [39]. In addition, data was also collected on various aspects of stigmatisation [26, 40].

In the case of diabetes-related information, data was gathered on the respondents’ perceived level of information and personal need for information in terms of the causes of diabetes, the course of the disease, its treatment and therapy, the complications it can lead to, lifestyle changes as well as health promotion and prevention [41]. With regard to information-seeking behaviour, participants who had previously informed themselves about diabetes were asked which sources of information they had used to do so [29]. The remaining participants were asked where they gained information about general health issues [29]. In each case, several sources of information could be provided.

Respondents who had already actively sought information about diabetes were also asked about their view of the trustworthiness and clarity of the information about diabetes that is published in the media [42].

Survey instruments used for people with diabetes

Respondents with diabetes were asked the same questions about their level of knowledge about diabetes and about the severity of diabetes that were posed to respondents without the condition. However, the question as to whether a respondent believed that they would have diabetes for the rest of their life [43] and the questions linked to stigmatisation [26, 40] were only posed to respondents who had stated that they had had diabetes in the last 12 months. These questions only differed in perspective from those posed to people without diabetes: people with diabetes were not only asked about diabetes in general, but also about their own condition. Moreover, the questions on their perceived level of information, individual information needs and sources of information were posed for a broader range of topics than for people without diabetes. For example, data was also collected on support services, helplines and sources of information [41]. Finally, both groups were also asked the same questions about the trustworthiness and clarity of the information published in the media.

2.5 Statistical analysis

In order to ensure that a sample is representative of the general population, it is important to take into account all relevant population groups, especially those who are less willing to participate in the study. As such, this study also employed a weighting factor. First, the distribution of household sizes in the sample was adjusted to reflect the known distribution within the population in Ger many (as of 31 December 2016) [44]. The sample was then adjusted to take into account the different probabilities of selection. This was followed by a standard method of calculation that is recommended by the ADM and that takes into account the study’s dual-frame design [45]. Finally, adjustment weighting was applied. For the representative part of the study (the first section with 2,591 interviews including the 263 interviews of people with diabetes), a gradual weighting in terms of age × gender (in 7 × 2 steps), education x gender (in 3 × 2 steps) and federal state (in 16 steps) was implemented to ensure that these characteristics were distributed in a manner that was comparable to the German-speaking, resident population aged at least 18 years. Data supplied by the Federal Statistical Office were used (as of 31 December 2016) to make these adjustments [44].

As the Federal Statistical Office’s data do not enable conclusions to be drawn about individuals with diabetes among the German-speaking resident population aged at least 18 years they could not be applied to the sample of respondents with diabetes (1,479 interviews including 1,216 from the screening conducted for diabetes and 263 interviews from the representative part of the survey). Instead, a gradual weighting was implemented for age x gender (in 6 × 2 levels) and education × gender × age (in 3 × 2 × 2 levels) to reflect the distribution of people with diabetes identified by the German Health Update study (GEDA) 2012 (n=1,828) [46]. The GEDA 2012 sample was used because the study applied a comparable methodology. The distributions identified by GEDA 2012 were assumed to correspond to those within the current German-speaking resident population aged at least 18 years.

Finally, a third weighting factor was employed to enable comparisons to be made between respondents with and without diabetes in the overall data set. This was necessary because the screening method used in the study led to a significant overrepresentation of respondents with diabetes in the overall sample. The weighting factor adjusted the ratio of respondents with and without diabetes to ensure that it reflected the level identified by the representative part of the study; as such, the proportion of respondents with diabetes in the overall data set was the same as the proportion identified in the representative part of the study (9.7%).

All results presented here were calculated using the respective weighting factor. Averages or percentages (including 95% confidence intervals, 95% CI) were calculated using SAS (Version 9.4, SAS Institute Inc., Cary, NC). The results from both samples are presented as totals and separately for women and men (Annex 3). In addition, differences in terms of educational group (low, middle and high) were examined (Annex 4). These groups were classified using the CASMIN scale (Comparative Analysis of Social Mobility in Industrial Nations) [47], which takes both general and vocational training into account.

3. Results

3.1 Population without diabetes

In total, 2,327 people (1,313 women, 1,014 men) between the age of 18 and 97 years without a previous diagnosis of diabetes participated in the telephone survey. Their average age was 50.4 years (women 51.7 years, men 48.3 years).

Knowledge about the disease

More than half of the participants without diabetes perceived their knowledge about the disease as very good or good (Figure 2). The proportion of people who perceived their own level of knowledge as very good was almost twice as high among women as men. Respondents with low levels of education were more likely to rate their knowledge of diabetes as ‘non-existent’ than people with middle or high levels of education. Furthermore, respondents with low or middle levels of education also perceived their knowledge of diabetes as very good less frequently than those with high levels of education (Annex 4 Table 3).

Self-assessed level of knowledge about diabetes among adults without diagnosed diabetes according to gender (n=1,305 women, n=1,008 men)

Source: Disease knowledge and information needs - Diabetes mellitus (2017)

A total of six questions were used to collect data about the respondents’ objective level of knowledge about diabetes (Annex 3 Table 1). The participants were most frequently aware of the fact that people with diabetes had elevated blood-sugar levels (66.2%) and that the symptoms of diabetes usually develop gradually (62.6%). Almost half of the respondents (49.6%) were aware that more people are affected by type 2 diabetes in Germany than type 1. By contrast, less than a third of respondents (31.3%) knew that diabetes was not a blood cell disorder. Respondents were the least familiar with type 1 diabetes. Only 26.9% of respondents knew that type 1 diabetes does not disappear with puberty and only 17.2% knew that it cannot be treated with tablets. An educational gradient was identified among answers to the question as to whether diabetes was a blood cell disorder. Four further questions were answered correctly more frequently by respondents with high levels of education than those with low levels of education (Annex 4 Table 1).

Disease perception

Almost two-thirds of respondents (65.0%) considered diabetes to be a serious or very serious disease (Annex 3 Table 2). In contrast, only 1.2% of respondents felt that diabetes was not a serious condition, while 12.6% of respondents expressed no opinion on this issue. The results were similar among women and men. Respondents with low levels of education rated diabetes less often as not or moderately serious compared to the two higher educational groups (Annex 3 Table 3).

60.1% of respondents believe that diabetes is a condition that people have for the rest of their lives. However, 28.2% rejected outright or partially disagreed with this statement (Annex 3 Table 2). The results were similar for women and men and across the three educational groups (Annex 4 Table 2).

A small proportion of respondents believed that people with diabetes are discriminated against (Annex 3 Table 2). Overall, 20.6% felt that people with diabetes are often unable to meet their everyday needs, and 28.8% believe that individuals with diabetes cause the disease themselves through an unhealthy lifestyle. Only 14.0% of respondents felt that people with diabetes face disadvantages. No differences according to gender where identified in attitudes towards people with diabetes. Participants with middle or high levels of education were more likely than those with low levels of education to believe that people with diabetes can meet their everyday needs (Annex 4 Table 2).

Diabetes-related information

Perceived level of information

44.0% of participants felt that they were well or very well informed about the ‘causes of diabetes’; between 36.0% and 41.2% felt the same way about the ‘course of the disease’, ‘treatment and therapy’ and ‘complications’. Furthermore, this was the case with 48.9% of respondents when it came to ‘lifestyle changes, health promotion and disease prevention’. Women considered themselves better informed than men on all issues. Participants with lower and middle levels of education more frequently felt completely uninformed about these issues than those with a high level of education (Annex 4 Table 3).

Subjective information needs

The greatest need for information was identified for the issues of ‘lifestyle changes, health promotion and disease prevention’ (Figure 3). No differences were found between women and men in this regard. The need for information on ‘causes of diabetes’, ‘course of the disease’ and ‘complications’, was higher among people with a middle level of education than among those with a high level of education (Annex 4 Table 4). Participants who felt uninformed about the ‘causes of diabetes’, ‘course of the disease’, and ‘complications’ also expressed a greater need for information than those who saw themselves as well or very well informed (data not shown).

Information-seeking behaviour

Almost a third (31.0%) of respondents stated that they had actively sought information about diabetes at least once. However, this was more common among women (37.0%) than men (24.7%). Printed information such as brochures, newspapers, magazines and books were the most commonly used source of information (83.9%), followed by websites on health and diabetes (49.3%), general practitioners and other doctors (47.2%) as well as radio or television broadcasts (43.4%). In this context, women more frequently cited printed information than men (88.9% compared to 75.9%). Respondents with low levels of education more frequently sought information from their general practitioner or other doctors than respondents with middle or high levels of education (65.9%, 43.4% and 41.8% respectively; data not shown). The most frequently mentioned sources of information on health issues in general were general practitioners or other doctors (72.0%), followed by printed information (69.6%), radio or television broadcasts (60.0%) and websites covering health issues (53.7%). Participants with middle or high levels of education use websites more frequently as sources of information than those with low levels of education (70.6%, 60.0% and 35.5% respectively; data not shown).

The perceived trustworthiness and clarity of information

More than half (54.7%) of people who had previously actively sought information about diabetes found it difficult or very difficult to judge whether the information published in the media about the condition was trustworthy (Figure 4). In contrast, 69.3% of respondents found it very easy or fairly easy to understand this information. No gender or educational differences were identified (Annex 4 Table 7).

**Perceived level of trustworthiness and clarity of information among adults without diagnosed diabetes according to gender.** Only participants who had already actively sought information about diabetes were asked this question **(n=550 women, n=318 men)**

Source: Disease knowledge and information needs - Diabetes mellitus (2017)

3.2 Population with diabetes

In total, 1,479 people (740 women, 739 men) aged between 18 and 96 years with a physician diagnosed diabetes participated in the telephone survey; 263 of these respondents had taken part in the representative part of the study. Data for the prevalence of diabetes in the representative survey was gathered by asking respondents whether they had ever been diagnosed with diabetes; the prevalence rate was 9.7% (8.2%-11.5%). The average age of all participants with diabetes was 64.9 years (women 65.3 years, men 64.5 years). The mean duration of diabetes was 14.3 years (women 13.7 years, men 15.0 years). In total, 1,386 participants (92.1%) who had ever been diagnosed with diabetes had had diabetes during the last 12 months. The prevalence of diabetes in the last 12 months was 8.7% (7.2%-10.5%). Respondents who had diabetes within the last 12 months were also asked about the type of diabetes they were affected by: 14.1% of respondents who had had diabetes over the last 12 months stated that they had type 1 diabetes; 80% that they had type 2 diabetes; 1.4% another type; and 4.6% did not know which type of diabetes they were affected by. In terms of treatment, 90.2% of these respondents stated that they were receiving medication such as tablets, insulin or other injectable hypoglycaemic drugs. 7.4% were only in receipt of treatment through diet and sport, and 2.4% of these respondents said they were not receiving any treatment at all.

Knowledge about diabetes

92.8% of respondents who had received a diagnosis of diabetes at some point viewed their knowledge of the condition as good or very good (Figure 5). The self-assessed level of knowledge was similar between women and men and across the various educational groups (data not shown).