Clinical Nutrition (2008) 27, 800e805

available at www.sciencedirect.com

http://intl.elsevierhealth.com/journals/clnu

ORIGINAL ARTICLE

Meeting the nutritional requirements of

hospitalized patients: An interdisciplinary
approach to hospital catering
S. Iff a, M. Leuenberger a, S. Rösch a, G. Knecht b, B. Tanner b, Z. Stanga a,c,*

a
Division of Endocrinology, Diabetes and Clinical Nutrition, Bern University Hospital, University of Bern, Bern, Switzerland
b
Department of Catering, Bern University Hospital, University of Bern, Bern, Switzerland
c
Department of Internal Medicine, Bern University Hospital, University of Bern, Bern, Switzerland

Received 13 May 2008 ; accepted 27 July 2008

KEYWORDS Summary
Hospital food; Aim of study: The study served to assure the quality of our catering, to locate problems, and to
Quality assurance; define further optimization measures at the Bern University Hospital. The main objective was
Hospital food to investigate whether the macronutrient and energy content of the hospital food complies
recommendations; with the nutritional value calculated from recipes as well as with the recommendations issued
Food waste; by the German Nutrition Society (DGE).
Hospital catering Material and methods: Prospective, randomized, single-center quality study. Complete stan-
dard meals were analyzed over seven consecutive days for each seasonal menu plan in one
year. The quantitative and qualitative chemical content of a randomly chosen menu was deter-
mined by an external laboratory.
Results: Sixty meals were analyzed. The amount of food served and the ratio of macronutri-
ents contained in the food satisfactorily reflected all recipes. Not surprisingly, the energy
and carbohydrate content of our meals was lower than in the German recommendations,
because the report of the DGE is based on the sum of meals, snacks and beverages consumed
over the whole day and not only on the main meals, as we analyzed.
Conclusions: Periodic quality control is essential in order to meet recommendations and
patients’ expectations in hospital catering. Members of the catering service should undergo
regularly repeated skills training, and continuous efforts should be made to ensure portion size

* Corresponding author. Division of Endocrinology, Diabetes and Clinical Nutrition, Inselspital, Bern University Hospital, CH-3010 Bern,
Switzerland. Tel.: þ41 31 632 42 46; fax: þ41 31 382 43 60.
E-mail address: [email protected] (Z. Stanga).

0261-5614/$ - see front matter ª 2008 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights reserved.
doi:10.1016/j.clnu.2008.07.008
Food quality surveillance in hospital 801

for all delivered meals. Food provision in the hospital setting needs to be tailored to meet the
demands of the different patient groups, to optimize nutritional support, and to minimize food
waste.
ª 2008 Elsevier Ltd and European Society for Clinical Nutrition and Metabolism. All rights
reserved.

Introduction point disrupts the system. The system on the ward should
be designed to prevent interruption of meals by proce-
The provision of good quality food, fluids, and nutritional dures, rounds, etc. Food waste and intake should be
care is an integral part of the therapeutic care in a hospital. monitored and audited regularly, with modification of
Meeting patients’ nutritional requirements will help them policies as necessary. Furthermore, quality control initia-
get well. Optimizing menu design, adapting menus, and tives should evaluate whether recommendations for
improving the mealtime atmosphere are vital steps to fight adequate hospital nutrition are met. Various aspects of
malnutrition in the hospital. When a menu is planned, food provision should be monitored regularly, including
disease- or therapy-related feeding problems such as loss of portion control, presentation, flavor and texture, temper-
appetite, changes in taste perception, or difficulty chewing ature, timing of service, and patient satisfaction.
and swallowing must be taken into account. On the one Portion control is an important part of quality control,
hand, the hospital’s catering department must be able to and is often poorly addressed or non-existent. There is
deal with individual patients’ needs as well as offering insufficient recognition of evidence that in-patients of
a balanced menu that meets the patients’ nutritional different sexes and ages and with different nutritional
requirements. On the other hand, from an economic point status and disease processes have different energy and
of view, food wastage should be minimized, since wastage protein requirements. In Switzerland there are no specific
can be as high as 67%.1 guidelines for hospital catering. Therefore, because the
The term food chain (Fig. 1) has been adopted to cultural eating habits, the gross national product per cap-
emphasize that all stages in the provision of food must be ita, as well as budgets for meal preparation in the two
adequate, from screening of patients and planning of countries, are comparable,3 we abide by the recommen-
menus to the distribution and serving of the food.2 It is dations issued in 2004 by the German Nutrition Society
desirable for hospitals to appoint a multidisciplinary and (DGE).4
multiprofessional nutrition steering group, including the The main aim of the current study was to investigate
clinical nutrition team, to oversee all aspects of nutritional whether the macronutrient and energy content of the food
care, from catering to artificial nutrition. A failure at any in our hospital complies with the recommendations issued
by the DGE.4,5 Further objectives were to determine
whether the food provided on the plate corresponds to the
amount calculated by the recipes, and to find out whether
Patient our catering service has improved since the last analysis in
1996. The results of the study should help us to initiate, if
necessary, a reform of the catering at our hospital, to
adjust institutional recommendations and guidelines for
menu planning, and to locate difficulties in meal service, in
Screening and monitoring terms of continuous quality assurance.
Nutrition
Steering Material and methods
Group
& Appropriate menu and snacks The hospital kitchen at the Bern University Hospital in Bern,
Clinical Switzerland, provides around 2000 patients meals per day,
Nutrition roughly 75% of which are standard menus. In this prospec-
Team tive, randomized, single-center quality study, complete
Purchase and food preparation standard meals (breakfast, lunch and dinner) of four
different seasonal menu plans produced at the Bern
University Hospital were analyzed. Energy, protein, fat,
carbohydrate and salt (defined as NaCl) content of all
standard menus prepared over seven successive days in the
Food distribution
months of April, June, September and December were
analyzed in an external laboratory (UFAG Laboratories AG,
Sursee, Switzerland). The four test series corresponded to
one calendar week each within a tri-weekly seasonal menu
Food serving plan. This seasonal menu plan was repeated during a whole
season (three months). The April test series evaluated the
Figure 1 The food chain. spring menu, the June series evaluated the summer menu,
802 S. Iff et al.

the September series evaluated the autumn menu, and the these ranges should be as small as possible to meet the
December series evaluated the winter menu. needs of different hospital patients; on the other hand they
The nursing staff was responsible for the ordering of should take menu organization and food preparation into
food. When a patient entered the hospital he was asked consideration. A deviation of 10% from the DGE recom-
about his appetite and eating habits. Food was ordered mendations was therefore considered as acceptable.
according to each patient’s individual requirements. Moreover, individual fluctuations in the metabolic rate also
Patients can choose between different standard menus, support a range of tolerance. The range of tolerance refers
which can be ordered in five different portion sizes: one to the total content of energy, protein, fat, carbohydrates,
quarter, one half, one, one-and-a-half and two portions. and salt for three main meals. Beverages and snacks were
The standard size is one portion. In addition, each patient not included in the tolerance values.
can order individual snacks from a variety of fruits, yogurts,
sandwiches, or small snacks. This easy standardized Statistics
ordering process results in ‘‘tailored’’ food.
A standard breakfast comprised one slice of bread and
The analyzed data (daily total per test day) are indicated as
one bun, 25 g jam, 20 g butter, and 2 dl coffee with 1.5 dl
mean and standard deviation. Further data were tested with
milk. The complete lunch consisted of a soup, a main
the unpaired Student’s t test. Data analysis was performed
course with a carbohydrate source (pasta, potatoes, rice,
with SPSS for Windows, Version 12.0 software (SPSS Inc.,
corn, cereals) and a protein source (meat, fish, eggs or
Chicago, IL, USA). An acceptable level of statistical signifi-
tofu), seasonal vegetables, and a dessert. Dinner had the
cance was established in advance at p < 0.05. SD Z Standard
same components as lunch. These meals corresponded to
deviation; and n.s. Z not significant.
the standard offering prepared by the hospital kitchen for
patients without any special diets. The basis of the meals
was defined to be a Mediterranean diet. Therefore high- Results
quality fat, e.g., olive oil, was preferred in the menu
composition, and foods with a naturally high proportion of A total of n Z 60 meals were analyzed between February
starch (polysaccharides) were favored, such as cereal 2004 and February 2005. The analysis consisted of one
products and potatoes. Meals were prepared in advance in standard breakfast, seven standard lunches and seven
the kitchen on individual trays and re-heated on the ward standard dinners for each season.
immediately before consumption. The usual beverages
accompanying the meals were water and tea with little or
no sugar. Snacks and other beverages consumed on an
Comparison of the chemical analysis 2004/2005
individual basis were intentionally not included in the and 1996
analysis.
The head of the hospital logistical support specified the The results of the chemical analysis of the current study are
test weeks together with the laboratory in advance. The compared to the results of the analysis made in 19969
kitchen staff was not aware of which menus was intended (Table 1). The analysis showed a significantly lower fat and
for analysis. One standard lunch menu and one standard salt content for 2004/2005. The reduced amount of fat of
dinner were randomly chosen for analysis from approxi- 21% results in a significantly decreased energy content in
mately 700 menus by a co-worker who was not involved in 2004/2005 (8%, p < 0.001). The salt content of the meals
the kitchen or in the study. The sample meals were locked in 2004/2005 was significantly lower (13%, p Z 0.004).
into a plastic box provided by the laboratory, and frozen
immediately. The samples were kept on dry ice on the way Menu planning precision
to the laboratory. Before the analysis, each meal was
homogenized and the net weight of the whole meal was The chemical analyses of macronutrients and energy
determined. content of hospital food in comparison with the calculated
nutritional values from recipes are shown in Table 1. The
analyzed amounts of fat, protein and energy were not
Chemical analysis significantly different from the nutritional profile calcu-
lated from the recipe. The analyzed amount of carbohy-
The nitrogen content was determined according to the drates was significantly higher (þ4%, p Z 0.017) than the
method of Kjeldahl,6 and the protein content was calcu- amount in the recipe. Despite the higher carbohydrate
lated as nitrogen in grams  6.25. Total fat was determined content on the plate, the energy content did not differ
by Soxhlet’s extraction.7 Carbohydrate and energy content significantly.
were calculated from the analytically determined energy-
supplying food components. All chemical analyses were
performed according to the Swiss Food Handbook.8 Comparison of chemical analysis 2004/2005 with
Energy, fat, protein and carbohydrate contents of DGE recommendations
a breakfast, lunch and dinner for each test day were added
together to create a daily total. The results of the analysis of macronutrients, salt and
Following the DGE recommendations,4,5 the clinical energy content (mean value of the seven days of all four
nutrition team of our Hospital also defined a range of test series) are compared to the DGE recommendations
tolerance for the daily nutrient supply. On the one hand (Table 1). The table shows that the mean energy content of
Food quality surveillance in hospital 803

Table 1 Chemical content in 2004/2005, compared with 1996, the DGE recommendations, and the estimated amount of
macronutrients of the recipes
Chemical content 2004/2005 DGE recommendations  10% p-Value
(per day) (mean  SD) (per day) (mean  SD)
Energy (kcal) 1561.4  129.1 1850  185 p < 0.001
Protein (g) 66.9  10.4 67.7  6.8 n.s.
Fat (g) 60.9  9.3 61  6.1 n.s.
Carbohydrates (g) 185.7  17.2 248.2  24.8 p < 0.001
Estimation by recipe (per day)
Energy (kcal) 1561.4  129.1 1553.5  75.2 n.s.
Protein (g) 66.9  10.4 64.3  6.3 n.s.
Fat (g) 60.9  9.3 63.4  6.2 n.s.
Carbohydrates (g) 185.7  17.2 178  9.8 p Z 0.017
NaCl (g) 9.4  1.9 9.2  1.7 n.s.
Chemical content 1996 (per day)
Energy (kcal) 1561.4  129.1 1698  123.3 p < 0.001
Protein (g) 66.9  10.4 66.5  9.7 n.s.
Fat (g) 60.9  9.3 77.3  16.2 p < 0.001
Carbohydrates (g) 185.7  17.2 180  10.8 n.s.
NaCl (g) 9.4  1.9 10.8  1.6 p Z 0.004

the meals was significantly lower than the recommenda- Discussion

tions issued by the DGE (16%, p < 0.001). This was due to
a significantly lower carbohydrate content (25%, The daily full portion menu for each season was analyzed
p < 0.001). It should be mentioned that our results do not for total energy, protein, fat, carbohydrate and table salt
include any snacks or beverages e in contrast to the DGE content. These values were compared to our values recor-
recommendations. ded in 1996, to the recipes, and to the recommendations of
The average daily energy content of the four test series the DGE.10
is shown in Fig. 2. The mean energy content of the autumn Comparing the chemical content in 2004/2005 with the
test series was lower (9.6%) than that of the other analysis made in 1996,10 we found that the fat content was
seasonal menus, although not significantly. 21% lower in the later period, resulting in a significantly
Average daily macronutrient content of the four test decreased energy content (8%, p < 0.001) (Table 1).
series analyzed in 2004/2005. According to the international DACH recommendations, the
The mean values per test series of each macronutrient daily energy requirements should be made up 15% of
are shown in Fig. 3. Although not significant, the carbohy- proteins, 30% of fat and 55% of carbohydrates.10,11 In 1996
drate (13%), protein (13%) and fat (4%) contents were the fat content of 41% was too high and may have promoted
all lower in the autumn test series. heart disease, cancer and metabolic diseases.10,12 As
a consequence we adapted our recipes, and the analysis for
2004/2005 showed a fat content of 35%. Another nutritional
component we analyzed was salt content. We found
3000

2500
Protein (g) Fat (g) Carbohydrates (g)
250
Energy (kcal)

2000
Nutrient amount (g)

200
1500
150
1000 1850
1598 1634 100 191 201
1569 1446 183 167
500 50
71 72 65 63
61 61 60 59

0 0
Spring Summer Autumn Winter DGE Spring Summer Autumn Winter
Test series Test series

Figure 2 Average daily energy content of the four test series Figure 3 Average daily macronutrient content of the four
analyzed in 2004/2005 (mean  SD). test series analyzed in 2004/2005 (mean  SD).
804 S. Iff et al.

Table 2 Measures taken following the quality control

Standard Results Measures
DGE recommendations: 1561.4  129.1 kcal 1. Provide snacks, energy- and protein-rich supplements,
energy content to be provided fortified meals if nutritional requirements are not covered
by standard meals per day: 2. Selection of appropriate menu (e.g., age, severity of
1850  185 kcal (mean  SD) disease, cultural origin)
3. More frequent and energy-dense meals and smaller portion
size for selected patients
4. Optimizing food distribution and service in general

Normalized management to Macronutrient content 1. Regularly repeated skills training for food services and
plan the menu and energy was consistent with continued efforts to ensure portion size
content which is calculated the recipes except 2. Standardization of the ladles
from recipes for carbohydrates 3. A sample plate is shown to every staff member before
food preparation and portion size is regularly controlled
by senior staff members
4. Periodic control of portion size and nutrient content by
chemical analyses (quality assurance)

9.4  1.9 g salt per day in 2004/2005, significantly lower Providing 1600 kcal a day has several advantages: We
(13%, p Z 0.004) than the previous analysis in 1996. The can avoid an overloaded plate, which does not promote
daily average consumption of salt by the Swiss population food consumption. Further, the waste of food eventually
was 8e16.8 g/person/day in 2005.10 According to these results in higher costs.14 According to unpublished data
data, our results are situated in the lower range of the per from our hospital, food is wasted at a rate of 22e30%,
capita salt use. which agrees with the findings of Dupertuis et al. (27%),
The mean values of the daily macronutrient content of Barton et al. (32e42%) and Suominem et al. (27%).14e16
the four test series analyzed in 2004/2005 are shown in Providing larger portions is not a valid strategy to improve
Figs. 2 and 3. Although not significant, the carbohydrate energy intake. Despite providing 2007  479 kcal per day
(13%), protein (13%), fat (4%) and consequently the with menus and snacks, the group of Pichard et al. found
energy (9.6%) content were all lower in the autumn test the mean energy intake to be 1536  599 kcal. Forty-one
series; the reason for this is not evident. A possible expla- percent of patients received less than the recommended
nation for the slightly reduced portion size served could be amount of energy of 1850 kcal, and 31% were below the
an irregularity of a member of the kitchen staff in scooping minimal energy needs defined by the HarriseBenedict
the portion. In response, we adopted the following formula. Another study showed a mean daily energy
measures: skills training for the food services department is content of 1838 kcal for meals evaluated. However, the
repeated regularly, the ladles have been standardized, mean energy intake was at best 1463  393 kcal per day.15
a sample plate is shown to every staff member before food Providing snacks and enriching the food with energy is
preparation, and the portion size is regularly controlled by a better strategy to meet the needs of patients at risk of
senior staff members. malnutrition.17 Using fortified meals and snacks, Gall et al.
The DGE recommendations state that hospital food were able to achieve energy targets and enhanced energy
should aim to provide 1850 kcal per day for in-patients, intake in 82% of the elderly patients they studied; they
based on an estimate of a physical activity level of 1.2 for recommended that food fortification and provision of
an acute care hospital. However, the DGE recommenda- snacks be considered as a key strategy to achieve improved
tions apply not only to the main meals, but to the sum of energy and protein intake in hospitalized patients.13
meals, snacks and beverages consumed over a whole day. Further, people who are unwell or have a low appetite
Not surprisingly, we found significantly lower energy prefer numerous small portions to three large meals a day.
(16%) and carbohydrate (25%) contents in our meals, In addition to standard meals, patients in our hospital have
which can be attributed to the fact that we analyzed the the possibility to order bigger portions if they wish to.
main meals only (Table 1). A mean energy content of Snacks and fortification of the food can be offered routinely
1561  129 kcal (median 1568 kcal) was obtained with our by the nursing staff and dieticians to complement the main
main meals alone. If we add two standard snacks for a total meals. Food service provision in the hospital needs to be
of 350 kcal, such as a small sandwich (whole grain bread) tailored to meet the demands of these different groups, as
and a fruit (banana), the recommendations are easily met, recommended by the German guidelines, in order to mini-
and the difference in the carbohydrates (þ60 g/day) can be mize waste and indirectly reduce costs.4
explained. With a mean protein content of 67  10 g per
day, we reached the recommendations with the main meals
alone. In our opinion, the total daily requirement of protein Conclusion
should be supplied by the main meals, since common snacks
usually are not rich in protein.13 The total daily require- Our results show that the amount and ratio of macronutrients
ment of fat was covered with the main meals as well. in the food we served in 2004/2005 were satisfactorily
Food quality surveillance in hospital 805

congruent with the nutritional profile calculated from all 7. Soxhlet F. Die gewichtsanalytische Bestimmung des Milchf-
recipes. Table 2 gives an overview of the measures taken at ettes. Polytechnisches J 1879;232:461.
our hospital in response to the findings of the present study. 8. Swiss Federal Office of Public Health. Schweizerisches Leb-
Periodic quality assurance is essential in hospital food ensmittelbuch (Swiss Food Manual). Bern: Eidg. Drucksachen
und Materialzentrale; 2005.
preparation in order to meet government recommendations
9. Jost C, Knecht G, Sterchi A, Huber P, Lindenmann U, Bürgi U,
and patients’ expectations. If the findings of the measure- et al. Wie stark weicht die chemisch analysierte
ment and monitoring activities diverge from the control Nährstoffzusammensetzung einer typischen Spitalkost von den
standards, immediate action should be taken. With this DGE-Empfehlungen einer gesunden Ernährung ab? Aktuelle
combined approach, the hospital catering service can Ernährungsmedizin 1998;23:234e9.
improve quality standards and provide cost-effective nutri- 10. Eichholzer M, Camenzind-Frey E, Matzke A, Amadò R, Ballmer P.
tional support. Beurteilung des Verbrauchs und angenäherten Verzehrs an
Nahrungenergie und Nährstoffen. In: Camenzind-Frey E, Sutter-
Leuzinger A, Schmid A, Sieber R, editors. Fünfter Schweizerischen
Conflict of interest Ernährungsbericht. BAG; 2005. p. 52e67.
11. DACH. Deutsche Gesellschaft für Ernährung ÖGfE, Schweizeri-
sche Gesellschaft für Ernährungsforschung, Schweizerische
The authors have no conflict of interest.
Vereinigung für Ernährung. Referenzwerte für die Nähr-
stoffzufuhr. Frankfurt am Main: Umschau Braus; 2000.
12. Wahrburg U. What are the health effects of fat? Eur J Nutr
References 2004;43(Suppl. 1):I/6e11.
13. Gall MJ, Grimble GK, Reeve NJ, Thomas SJ. Effect of providing
1. Edwards J, Nash A. Catering services. Measuring the wasteline. fortified meals and between-meal snacks on energy and
Health Serv J 1997;107:26e7. protein intake of hospital patients. Clin Nutr 1998;17:259e64.
2. Allison SP, Stanga Z. Organization of nutritional care. In: 14. Barton AD, Beigg CL, Macdonald IA, Allison SP. High food
Sobotka L, editor. Basics in clinical nutrition. Prague, Czech wastage and low nutritional intakes in hospital patients. Clin
Republic: Publishing House Galén; 2004. Nutr 2000;19:445e9.
3. OECD. OECD Factbook: economic, enviromental & social 15. Dupertuis YM, Kossovsky MP, Kyle UG, Raguso CA, Genton L,
statistics: organisation for economic co-operation and devel- Pichard C. Food intake in 1707 hospitalised patients: a prospec-
opment OECD; 2007. tive comprehensive hospital survey. Clin Nutr 2003;22:115e23.
4. Kluthe R, Dittrich A, Everding R, Gebhardt A. Das Ration- 16. Suominem M, Laine T, Routasalo P, Pitkala KH, Rasanen L.
alisierungsschema der Deutschen Gesellschaft für Ernährungs- Nutrient content of served food, nutrient intake and nutri-
medizin (DGEM) 2004. Aktuel ErnaehrMed 2004;29:1e9. tional status of residents with dementia in a finnish nursing
5. Referat DGE. Umsetzung der Referenzwerte für die Gemein- home. J Nutr Health Aging 2004;8:234e8.
schaftsverpflegung; 2000. Bonn, Germany. 17. Olin AO, Osterberg P, Hadell K, Armyr I, Jerstrom S, Ljungqvist O.
6. Kjeldahl J. Neue Methode zur Bestimmung des Stickstoffs in Energy-enriched hospital food to improve energy intake in
organischen Körpern. Anal Bioanal Chem 1883;22:366e82. elderly patients. J Parenter Enteral Nutr 1996;20:93e7.