Texture modified diet in German nursing homes: availability, best practices and association with nursing home characteristics

This nationwide cross-sectional survey was part of a larger project on behalf of the German Nutrition Society [14]. The required contact information of 10,589 (97% of all German) NHs was provided by the scientific institute of a large German health insurance company (AOK, Allgemeine Ortskrankenkasse). In order to achieve equal participation throughout Germany, five regions with similar numbers of NHs were predefined according to federal states [15]. In March 2014, a written questionnaire was sent to the management of a random sample of 5000 NHs (1000 for each region). For the participation, cooperation with the head of the nursing service and the kitchen manager was recommended. The questionnaire could be completed online or sent back as hardcopy by postal mail. As the return rate was below 5% after 4 weeks, all addressed NHs with available e-mail information (4156 NHs) received a reminder for the survey. In addition, the sample was enlarged by all NHs in the AOK list with available e-mail information which were not addressed before (3172 NHs). In total, 8172 NHs received an invitation to the survey.

The questionnaire was developed by the authors and pilot-tested in ten NHs. The final questionnaire consisted of 48 questions with mostly predefined answer categories (Additional file 1). For this analysis, two questions regarding TMD and 15 questions covering structural, operational and resident-related NH characteristics were used. An item-selection was made for four multiple-choice questions regarding operational NH characteristics with a comprehensive item set, choosing the most relevant items for this context. Additionally, a question on the knowledge of the national recommendations for nutrition in NHs (“DGE quality standard” known (yes/no)) was posed.

All results are presented for the total group and for subgroups according to the number of available texture modified (TM)-levels and best practices for TMD. The number of available TM-levels was calculated excluding “Smoothfood”, as this is usually not provided as a complete diet but only offered additionally, e.g. for residents relying on a feeding tube. The number of best practices for TMD was also calculated. The data are presented as relative frequencies (TMD, structural and operational NH characteristics), and mean ± standard deviation, median, interquartile range (IQR) (resident-related NH characteristics).

Differences between groups were tested using Chi² tests. A post-hoc analysis was conducted with z-tests and Bonferroni corrections to specify differences. Kruskal-Wallis tests were used to identify differences in resident-related NH characteristics as a function of the number of TM-levels and the number of considered best practices for TMD. Pairwise comparisons with Bonferroni corrections were used to specify differences. Missing values were not included in statistical tests. Statistical significance was set at a p-value of < 0.05.

Statistical analyses were performed with SPSS 24.0.

The availability of TMD is presented in Table 1 for the total group and stratified according to the number of TM-levels and the number of best practices for TMD. Almost all NHs (95.2%) stated to offer “minced & moist” texture. Nearly half of the NHs (43.9%) had two different TM-levels available, and about one quarter offered one (28.4%) or three (27.7%) TM-levels. All combinations of the three TM-levels were found, with soft and “minced and moist” texture being the most common (29.8%) (Additional file 2). “Smooth food” was only offered in a sixth of all NHs (17.1%). The proportion of NHs offering “Smoothfood” increased with the number of available TM-levels (p < 0.001) as well as with the number of best practices for TMD (p < 0.001) (Table 1).

One in five NHs (20.4%) implemented only one of the four best practices for TMD. Two best practices for TMD were considered by 29.5%, three by 37.1% and all four by 13.0% of the NHs. All combinations of the four best practices for TMD were found (Additional file 3). Most commonly implemented was the best practice “components separately visible”, namely by 84.2% of the NHs. “Re-shaped components”, considered by 27.9% of the NHs, was the least common best practice (Table 1).

NHs having three TM-levels available considered each of the best practices for TMD more frequently (all p < 0.02) than NHs with only one TM-level. Furthermore, NHs considering four best practices for TMD offered soft texture (p < 0.001) and pureed texture (p < 0.001) more frequently than NHs considering only one or two best practices for TMD (Table 1). Overall, 7.6% of the NHs stated to offer all of the three TM-levels and all of the four best practices for TMD, whereas 9.5% of the NHs offered only one TM-level and implemented only one best practice for TMD.

The national recommendations for nutrition in NHs are known by 69.8% of the NHs regardless of the available number of TM-levels or the number of best practices for TMD (Additional file 4).

None of the structural NH characteristics were associated with the number of available TM-levels or the number of best practices for TMD (Table 2). Solely the production of hot meals in-house tended to be more common in NHs with three or four best practices for TMD compared to NHs with only one or two (p = 0.061) (Table 2).

The number of residents with swallowing disorders was reported by 478 NHs (84.9%), the number of residents with chewing disorders by 446 NHs (79.2%), and the number of residents receiving TMD by 513 NHs (91.1%).

In almost all NHs “minced and moist” texture is available, which is easy to produce as only a blender and no special knowledge is needed. Soft texture, offered by only 60% of the NHs, requires no special processing, but only a special range of products can be used that is naturally soft or become soft (fork-mashable) by cooking. The production of pureed texture is most elaborate as it should be lump free and requires special equipment (e.g. a bowl cutter) for several types of food (e.g. meat) because of the natural fiber content. The different requirements on time and equipment and insufficient knowledge – e.g. about the required consistency of each TM-level, and about the food and processing methods required to achieve the right texture – might explain their varying availability. Additionally, the necessity to offer different TM-levels to be able to meet residents’ individual needs might not be present. Furthermore, it is possible that some institutions did not state to offer a specific texture because none of the residents had the respective needs at that moment although the questionnaire asked for general availability of TM-levels.

Differing terminology, definitions and levels of TMD are described in the literature of individual countries, and there are even differences within small geographical regions within countries [6, 17]. Therefore, it is unclear whether the definitions used in the questionnaire were understood in the same way by all NHs although a short explanation was given.

Residents with swallowing and/or chewing disorders need TMD for a safe and adequate energy and nutrient intake. Texture- modified meals often look less appealing, taste different and can also have psychological impacts, e.g. embarrassment which can affect acceptance of TMD and quality of life. Therefore, it is important to consider best practices for TMD in addition to the availability of different TM-levels [2, 7, 18, 19]. The basic recommendation is to serve the meal with separately visible components that allow the residents to recognize the food and to eat and taste it as they are used to with a regular diet. Re-shaping the components enhances the visual appeal of the meals significantly. In an intervention study conducted in a long-term care facility in Canada, increased body weight, energy and nutrient intake were observed after 12 weeks in residents receiving re-shaped TMD compared to a control group receiving unshaped TMD [7]. Another study in the USA showed an increase in food intake by 15% after changing to 3-D preparation of pureed foods [19]. Missing knowledge, equipment, and especially time and financial resources but also lacking awareness of the importance of this aspect might be reasons for not re-shaping or separately presenting the meal components.

The consideration of residents’ individual capabilities is important for menu planning in order to be able to offer residents meals that they are able to eat safely and at the same time process these meals as little as possible to minimize the impairment of the quality of life due to texture-modification [20]. At least one study indicated that the majority of NH residents (91%) received overly restrictive TMD [20]. One in five NHs that stated considering residents’ individual capabilities offered only one TM-level (Table 1). Therefore, it is unclear how that aspect was understood and implemented.

The positive association between availability of TM-levels and best practices for TMD indicates that NHs with a stronger focus on TMD considered both aspects simultaneously.

Structural NH characteristics are given facts that are not related to individual people and cannot be changed easily or in the foreseeable future. The structural characteristics examined here, do not seem to influence the implementation of TMD recommendations. Solely in-house meal production tended to be more common in NHs considering four instead of one best practices for TMD (Table 2). This tendency might be explained by the direct contact of staff and residents to the kitchen, allowing for immediate feedback and the possibility to change preparation processes on short notice.

Associations between NH size, as one of the structural NH characteristics, and single best practices for TMD, but no TM-level, were shown in a previous publication [15]. Components were found to be re-shaped more frequently in large than in small NHs, and individual capabilities were more frequently considered in small than in large NHs [15]. The size and the affiliated resources of the NH seem to influence the implementation of single best practices for TMD contrariwise. This might be one reason why the size of the NH was not associated to the number of implemented best practices for TMD.

Operational NH characteristics describe processes and the availability of special offers. Their implementation is usually dependent on a few persons in the NH and can be changed in the foreseeable future. The analyzed characteristics focus on nutrition aspects and their implementation may reflect a general nutritional awareness.

The majority of operational NH characteristics studied (13 out of 16) were associated either with the number of TM-levels or the number of implemented best practices for TMD, mostly with both (Table 3).

The availability of the analyzed special diets and delivery forms are also part of the national recommendations for nutrition in nursing homes [12]. This might be one reason for the positive association of them being offered and TMD. NHs following the recommendations for nutrition in NHs of the German Nutrition Society intentionally will probably consider all recommended aspects regarding special diets including TMD.

The calculation of energy and nutrient content is important as texture modification can lead to a dilution. A significantly lower mean energy provision of TMD compared to regular diet was shown in previous studies [21, 22]. NHs that do not calculate energy and nutrient content of TMD might not be aware of that and are at higher risk to provide insufficient amounts of energy and nutrients. The basis for calculating energy and nutrient content are written recipes, which also ensure a standardized preparation of meals. Those recipes are missing in 40% of the NHs. It is questionable how precise the calculation is made in some institutions as 32.9% of the 368 NHs that reported calculating energy and nutrient content do not have recipes for most/all dishes.

Dieticians are the food and nutrition experts in the institutions, but their roles can vary considerably [23]. Dieticians are, even in the group of German NHs offering three TM-levels (55%), clearly less common compared to NHs in other countries like the Netherlands (92%), Austria (84%) [24] and Italy (88%) [25].

A dietetic counseling service was common in participating NHs, however only half of the NHs (52%) offering the dietetic counseling service had a dietician available. Therefore, further information about the contents of the service and the qualification of the counselor would be of interest. The assessment of residents’ personal nutrition history is common as well, especially in NHs offering three texture-modified levels or four best practices for TMD. Although it is a good way to improve the quality of life by offering meals that meets the residents’ wishes, it is unclear how the collected information is used in daily practice.

The implementation of periodic malnutrition screening that was recommended and controlled by the medical service of health insurances in Germany, an institution controlling nursing care routinely, was not associated with the implementation of TMD. Interestingly, in spite of annual controls, 15% of the NHs do not periodically screen for malnutrition.

The assessed management and quality assurance aspects describe tools and processes to improve special knowledge, teamwork and communication. Except for the regular resident satisfaction survey, all aspects are associated with the offer of TMD (Table 3), leading to the assumption that an interdisciplinary approach with established processes and lines of communication influences the offer of TMD, probably in different ways. One aspect might be the general nutritional awareness that initially promotes the establishment of these management and quality assurance aspects. Additionally, different views and feedback resulting from these structures might directly influence the offer of TMD.

Regular resident satisfaction surveys that are assumed to influence meal quality through direct feedback were not associated with both TMD aspects. Reasons might be that TMD is not part of these surveys, that residents on TMD cannot take part because of cognitive or physical impairment, or that the results did not induce any changes in TMD.

Swallowing and chewing problems are the main reasons for the implementation of TMD [26‐28]. Cognitive deficits, refusal to eat, chew or swallow, and efficiency in feeding were other common reasons reported in combination with swallowing or chewing problems [26, 27]. The prevalence rates reported by the NHs varied widely. The median prevalence of swallowing problems (8%) was slightly below the results of two large studies assessing the prevalence at resident level, the LPZ study in the Netherlands including 3220 NH residents (11%) [29] and the worldwide nutritionDay study including 23,549 NH residents (13%) [30]. Other studies using specific dysphagia screening tests instead of asking staff or residents, reported markedly higher prevalence numbers (38 to 70%) [31‐33]. The median prevalence of chewing problems (9%) was lower compared to 26% in the LPZ study and the nutritionDay study [29, 30] and is probably underestimated by those completing our questionnaire.

Information on the frequency of TMD use is rare. In our study, the median percentage of residents receiving TMD was 14% (range 0–83%). In a study conducted in the USA in 1989, 292 long-term care institutions asked for the percentage of residents receiving TMD reported an average of 42% [27]. Other studies also reported a more frequent use of TMD with 26% [26], 29% [8] and 67% [22] of the residents receiving this diet, but these studies were limited to only one or two NHs. In NHs participating in the nutritionDay project, 13% of the residents received TMD which is similar to our result [30].

In NHs offering three TM-levels the proportion of residents receiving TMD was significantly higher than in NHs with one TM-level, due to a few institutions with more than 40% of residents receiving TMD. Regardless of the number of TM-levels, about half of the NHs reported between 10 and 20% of their residents being on TMD. Nevertheless, there are NHs with one out of five residents being on TMD offering only one TM-level. Even if the proportion is low, individuals might need a specific TM-level. Since we do not have detailed information on the severity of swallowing or chewing problems and the TM-level received, an evaluation of the quality of their nutritional care is not possible.

The strength of our study is the large number of assessed aspects, which allows good characterization of the NHs on institutional level. Additionally, the distribution of existing and participating NHs throughout Germany is comparable. The sample size is large compared to the only other study found with focus on TMD on institutional level including 292 institutions [27]. Other studies, reporting about prevalence of TMD on resident level, were limited to one or two institutions [8, 22, 26]. Nevertheless, the response rate is low (7.2%).

The length of the questionnaire might have limited motivation to participate. The accuracy of completion (e.g. regarding the estimation of the proportion of residents receiving TMD) was dependent on the knowledge and motivation of one or two people of the NH. Another constraint of the study is the limited information that can be assessed by a survey. The data collection with the help of a questionnaire might have led to more socially desirable than true answers.

The voluntary participation in the study presumably led to a bias of NHs with special interest in nutrition and comprehensive data on German NHs for a comparison is missing. Furthermore, the information on resident characteristics was limited to 79–91% of the participating NHs.