Hospital foodservice: a comparative analysis of systems and introducing
the ‘Steamplicity’ concept
J.S.A. EDWARDS AND H.J. HARTWELL The Worshipful Company of Cooks
Research Centre, Bournemouth University, Talbot Campus, Poole, Dorset, BH12 5BB
Corresponding author: Heather Hartwell, Tel: 01202 595585 e-mail:
Key words: institution, satisfaction, wastage, food intake
Word count: 7,060
This manuscript has not been published elsewhere and it has not been submitted
simultaneously for publication elsewhere.
Patient meals are an integral part of treatment hence the provision and consumption of a
balanced diet, essential to aid recovery. A number of food service systems are used to
provide meals but recently, the ‘Steamplicity’ concept has been introduced. This seeks,
through the application of a static, extended choice menu, revised patient ordering
procedures, new cooking processes and individual patient food cooked at ward level, to
address some of the current hospital food service concerns.
The purpose of this study was to directly compare selected aspects (food wastage at
ward level; satisfaction with systems and food provided) of a traditional cook-chill food
service operation against ‘Steamplicity’. Results indicate that patients preferred the
‘Steamplicty’ system in all areas: food choice, ordering, delivery, food quality and
overall. Wastage was considerably less with the ‘Steamplicity’ system; although care
must be taken to ensure that poor operating procedures do not negate this advantage.
When the total weight of food consumed in the ward at each meal is divided by the
number of main courses served, results show that at lunch, mean intake with the cook-
chill system was 202g whilst that for the ‘Steamplicity’ system was 282g and for the
evening meal, 226g compared with 310g.
Patient meals are an integral part of hospital treatment and the consumption of a
balanced diet, crucial to aid recovery (Stratton et al, 2006). Even so, it is well
established that up to 40% of patients may be undernourished on admittance to hospital;
a situation which is not always rectified during their stay (McWhirter and Pennington,
1994). The relevance and importance of patient meal service, when compared with
many clinical activities is not always appreciated and it is often seen as an area where
budgetary cuts will have least impact. This is particularly so as nursing staff are under
pressure to follow a medical/technical model of healthcare rather than one focused on
the fundamentals of nursing. Rapid turnover of patients also prioritises clinical
considerations. The provision of a foodservice system that optimises patient food and
nutrient intake together with minimising waste, in the most cost effective manner, is
therefore seen as essential.
Previous research has shown that food preference and acceptance constitutes 50% of the
variability in consumption (Cardello et al, 1996), and is not only a result of the intrinsic
quality of the food; but can also be related to consumer expectations and the degree to
which the food item matches them (Oh, 2000). Sensory characteristics, such as
appearance, flavour, texture and temperature have been found to be most important to
hospital patients when judging food quality (Clark, 1998). Temperature and texture are
key attributes of hospital food that have been shown to indicate patient satisfaction with
the food as served (Hartwell, 2004) with the temperature of hot food an area of patient
dissatisfaction and a regular cause for complaint (Stanga et al, 2003). It should therefore
be the goal of any hospital food service manager to prepare, distribute and serve safe
food of defined standards in respect of nutritional quality, balance, palatability and
temperature (Davis and Bristow, 1999).
Foodservice operations can be classified into three main styles (Jones and Lockwood,
1. Integrated foodservice systems: both food production and foodservice are carried
out as part of a single operation.
2. Food manufacturing systems: production of meals is separate from the service of
those meals, thus there is a decoupling of service from production, such as in
3. Food delivery systems: the operation involves little or no food production and
focuses only on the service of continuously assembled or regenerated meals.
Here there is decoupling and production lining.
This model can be developed and since the mid 1970s, a number of food production
system have been introduced which have sought to maintain current service levels but at
a reduced cost. These have included systems such as ‘Cook-Serve’, ‘Cook-Freeze’,
‘Cook-Chill’ and ‘Sous Vide’. More recently; the ‘Steamplicity’ concept has been
developed which has sought, through the use of a static, extended choice menu, revised
patient ordering procedures, new cooking processes and individual patient food cooked
at ward level, to address some of the current hospital foodservice concerns. Various
systems have been applied to increase profitability through bulk buying power, higher
productivity, better equipment utilisation and process control (Rogers, 2005). However,
selection is dependent on the environment and consumer profile, all physical, financial,
technological and operational issues need to be considered.
A cook-serve system is a ‘traditional’ catering operation where food is prepared and
cooked on site and distributed at the appropriate temperature to the wards, either already
plated or in bulk. This system allows for batch cooking which minimises hot-holding
and nutrient losses and optimises the food’s sensory characteristics as it can be prepared
close to the time required. However, in practice there can be a substantial time delay
between production and consumption as wards are often situated a long way from the
kitchens. The result is that many of the potential advantages are not realised.
In this system, food is cooked and held at a temperature of 70 - 750 C or more for at
least two minutes. Chilling occurs within 30 minutes of cooking and the temperature of
the food is reduced to 0-30C within 90 minutes. This temperature is maintained
throughout the storage and distribution cycle until regeneration occurs. Regeneration
can either be centrally controlled or carried out at ward level. However, a core
temperature of 70 - 750C must be reached for a minimum of 2 minutes for
microbiological reasons. In this system, dishes may be stored chilled for up to 5 days,
however, after reheating the food should be consumed immediately (Department of
Health, 1989). Advantages are higher efficiency and lower food costs based on bulk
buying and centralised purchasing while disadvantages number temperature control
which may compromise food safety and nutritional content (Hwang et al, 1999).
Sous vide is a variation of a cook-chill operation. Systems based upon large scale
production methods and the use of vacuum packaging, either before or after cooking, in
combination with the chilling techniques of cook-chill, were developed initially for the
institutional catering sector in Sweden (Schafheitle and Light, 1989).
Sous vide involves placing the food into heat stable, air and moisture high barrier plastic
bags or pouches. Air is then removed creating a vacuum with subsequent sealing of the
pouch. A pasteurising cooking process takes place followed by immediate rapid chilling
within 90 minutes to 0-30C. The product must then be stored within this temperature
range until required for consumption, but within five days of the date of production
(Department of Health, 1989). Both the quality and microbiological safety of sous vide
foods with extended shelf lives, requires good control and monitoring of critical points
throughout manufacture and distribution. The health risk of these products is small as
long as the temperature during storage is low (4±0.5°C) (Nissen et al, 2002).
This system is similar to cook-chill, except the food is frozen rather than chilled. After
cooking, dishes are blast-frozen to a temperature of –200C and kept at this temperature
until required. Storage at frozen temperatures can be more prolonged, for up to two
years. When required the food is defrosted and regenerated to a core temperature of at
least 70 - 750C (Department of Health, 1989). The disadvantage of this system is a loss
of texture owing to the freeze/thaw process involved and subsequent regeneration and
distribution to the wards (Hwang et al, 1999).
One of the most radical developments in hospital catering in recent years is the
introduction of this new technology which relies on a sealed pack incorporating a valve.
The food, both raw and partially cooked, is plated in a centralised production unit,
chilled (<5°C) and distributed to satellite kitchens where it remains chilled with an
expiry date currently of four days. As required, meals are heated/cooked individually in
a microwave to >75°C which allows patient choice at short notice and ensures better
quality food. A further advantage is the ability to control the cooking environment,
allowing consistent regeneration of the food with the right climate of moist heat, thus
avoiding drying out and therefore enhancing texture.
Comparison of Systems
Food temperature and texture are the statistically significant factors in the selection of a
system (Nettles et al, 1997) and also relate to patient satisfaction (Hartwell, 2004).
Notwithstanding, there is no agreement as to which system provides the best food
quality as in most cases, the effect is product specific (Rogers, 2005). It is suggested
that chilling is less damaging to texture and juiciness than freezing (Tansey et al, 2003)
while sous vide is superior in terms of vitamin retention but detrimental to sensory
quality (Church and Parsons, 2000; Pontin, 2005).
In all foodservice systems, food preparation and cooking can cause substantial and
unavoidable nutrient losses. The vitamins with the greatest losses during hot-holding of
food (> 10% after 2 hours) are vitamin C, folate, and vitamin B6; retinol, thiamin,
riboflavin and niacin appear to be relatively stable. Under normal operating conditions
with hot-holding limited to less than 90 minutes, vitamin retention is better in traditional
food service (cook-serve) than in a cook-chill system (Lawson et al., 1983). If chilled
food is stored for longer than 3 days or if food is held hot for long periods after bulk
reheating vitamin losses can be high (Williams, 1996; McErlain et al, 2001).
Traditional systems also give the opportunity for the patient to select portion size and to
decide if gravy is required with the meat. However, it has been suggested that menus
from hospitals using cook-chill systems provide a greater choice of hot menu items
(Williams, 2002) although, do not necessarily support improved dietary intake by
patients (McClelland and Williams, 2003). Cook-chill systems are more likely to have
trays delivered by food service employees whereas traditional food production systems,
where delivery is by trolley, tend to use nursing personnel (Jackson, 1997).
The traditional system of food production, cook-serve, is the most popular system used
in US medical food service (Silverman et al, 2000) and especially with smaller hospitals
(<100 beds) (Gledhill, 1993; Mibey and Williams, 2002). It is also considered by the
Audit Commission (2001) to be the cheapest at £2.20 per average spend compared to
£2.40 for a NHS operated cook-chill/freeze service. In Australia there has been a large
increase in the use of cook-chill systems from 5% in 1986 to 42% in 2001, despite the
fact that managers of such systems report lower levels of patient satisfaction (Williams,
2002). Conversely, other research has demonstrated little difference in satisfaction
between production systems (Edwards et al., 1998).
Increasingly in the U.K., a meal assembly foodservice system is being used where no
food preparation takes place on site, leaving the operation to focus on assembly,
regeneration and service. About a fifth of hospitals in the NHS operate in this way
purchasing meals from specialised food manufacturers (West, 2001). This trend looks
likely to continue as there is a cost implication. Contrary to the earlier assertion (Audit
Commission, 2001), Trusts that use the system of cook-serve may well spend more per
patient per day if overhead costs are included than those that buy in ready-made
(Deeming, 2002). However, vitamin C retention in vegetables in the meals assembly
system has been found to be between 17-80% for chilled vegetables and between 27-
83% for frozen vegetables after regeneration (West, 2001). Inevitably, the retention of
vitamin C in a meal assembly system would be lower than in a cook-serve due to the
number of processing stages involved, a well controlled cook-serve system will always
give better retention values for the heat labile vitamins. The benefit of ‘Steamplicity’ is
that the pre-prepared food is sealed and a vacuum created, stored for two days at chilled
temperatures and then reheated individually just prior to consumption, thereby reducing
potentiality for nutrient degradation while coincidently allowing for appropriate texture
The aim of this research, therefore, was to directly compare selected aspects of a
traditional cook-chill foodservice operation against ‘Steamplicity’. Specifically, the
goals were to measure food wastage at ward level; ‘stakeholders’ (i.e. patients, staff,
etc) satisfaction with both systems; and patients’ acceptability of the food provided.
Materials and Methods
Context and overview
A large National Health Service (NHS) teaching hospital was selected where
developmental work for Steamplicity was taking place. Permission was sought and
granted by hospital management to conduct this research and consent was given by
Data were collected from a post operative surgical ward where patients (n=52, 48%
male and 52% female) presented a mixture of clinical conditions. Over the study, 11
patients had been in this hospital previously 42 had not and the number of days as an ‘in
patient’ ranged from 1 to 240 (thirty-four weeks) with a mean of 28 for the cook-chill,
and 24 for the ‘Steamplicity’ system.
This ward was identified with the help of catering and medical staff as the most suitable
in that; these patients are more likely to stay longer, their medical condition would not
interfere with food consumption, they are capable of independent judgement and a
surgical ward represented a typical realistic hospital situation. The ward was not part of
the pilot developmental work for ‘Steamplicity’.
The study was conducted in two phases over two, two-week periods; the first phase
used the existing cook-chill food service operation, and the second, the ‘Steamplicity’
system. In both phases, the total amount of food sent to the ward and the total amount
remaining once the meal service had been completed was ascertained, over three
consecutive days, each week (Tuesday, Wednesday and Thursday) using appropriate
A mixed methodological approach was used to elucidate the complex nature of
satisfaction with the food service system and food. Patient satisfaction with both
systems was initially evaluated using a questionnaire. This was administered at the
midday meal on the first day of the study and at the evening meal on the last day of the
study. To enhance and validate this information, the research was extended to
incorporate observation, focus groups and interviews with both patients and pertinent
stakeholders such as medical staff, food service staff, and visitors. Food wastage, both
bulk and plate, was directly measured using appropriate balances over three days each
week. This approach permitted a direct comparison between the two meal systems.
Phase 1 (2 weeks) - Cook-Chill System
This phase utilised a ‘traditional’ cook-chill system, which was in operation in the ward.
Here a cyclical menu was used with food being pre-ordered the day before. On the day
of consumption, cold bulk food was loaded into the trolley and transported to the ward.
Here it was regenerated, plated and taken to patients’ beds. All food loaded into the
trolley was weighed by individual food item using ‘Teroaka Digital Weighing Scales’
(Model DS-50; ± 2g; max 6kg) once it had arrived at the ward, and details recorded.
When the meal service had been completed, all food remaining on the trolley was
weighed by individual food item and recorded.
Once patients had finished their meals, or all they could eat, any food remaining on the
plate was weighed by food item using either the Teroaka Scales or a ‘Soehnle Balance’
(± 1g; max 1kg), out of sight of patients and recorded. Where foods had become
‘mixed’, individual components were separated where possible; otherwise, a value
judgement was made as to what the food item was. Note was made of the number of
patients who had been served meals at each mealtime and also the number of meals, by
mealtime, which were served but not consumed for reasons such as Nil-by-Mouth,
discharged or absent from the ward for any other reason. Unfortunately due to
unforeseen circumstances (London bombing) data could not be collected on the final
study day and therefore results are presented for five days for this system.
Phase 2 (2 weeks) – ‘Steamplicity’ System
This phase utilised the ‘Steamplicity’ system. Patients ordered meals approximately
two hours prior to meal service from an extended choice menu, which remained
unchanged from midday to evening meals, and throughout the study. Individual, ready-
plated chilled meals (incorporating both fresh and partially cooked items) were
transported to the ward and cooked, as required, using microwave ovens. The average
portion sizes of individual foods and meals were taken from existing production
records. All meals (main courses and desserts) sent to the ward were then recorded.
Any food items left on the plate, once patients had completed their meals, were weighed
out of sight of patients, providing the average weight of food served to patients. Note
was also taken of all meals which were served but not consumed for reasons such as
Nil-by-Mouth, discharge or absent from the ward for any other reason and where the
meal had remained refrigerated and could therefore be used for a subsequent meal.
Patient acceptability of the food service systems and food
In order to assess the overall acceptability of the food service systems and of each food
item, a questionnaire was administered on two occasions to patients on the research
ward during each week of the study.
Developing the Questionnaire
A questionnaire was developed, informed from the literature and previous hospital
research, to ascertain patients’ opinions of the food service system and the food
provided. Prior to the main research, a small pilot study was conducted to establish the
validity and ease of completion of the questionnaire for patients. Two versions of the
questionnaire were developed and distributed to eight individuals (male and female age
range 24-55), with recent prior experience of being in hospital. The purpose here was to
ensure that both the questions and questionnaire could be understood and had been
interpreted correctly. These issues were confirmed by talking with participants
immediately after they had completed the questionnaire. Resulting from the responses
received, the questionnaire was revised and a further questionnaire distributed. This
then only required minor modifications and is given in Appendix A.
Administering the Questionnaire
Questionnaires were administered to each patient at the beginning of their stay (earliest
Tuesday midday) and at the end of their stay (or Thursday evening meal). They were
not administered to patients who for any reason were not eating, or who were not
‘mentally’ capable of completing them. Questionnaires were administered immediately
before the meal and collected once the meal had been completed and within an hour.
Assistance was given where necessary to help patients complete the questionnaires,
although those providing assistance had been appropriately trained and did not attempt
to ‘lead’ the responses in any way.
Focus groups/semi-structured interviews
Hospital food service does not operate in isolation but requires the co-operation and
integration of several disciplines to provide the ultimate patient experience. It is
accepted in the literature that patient assessment of meal service is multidimensional
(Gregoire, 1994) and that the hospitality experience is essentially interactive (Hepple et
al, 1990). Accordingly, stakeholders such as medical staff, food service staff, patients
and visitors were consulted after encountering both service systems to identify factors
contributing towards patient satisfaction and to elucidate each patient meal experience.
Sampling was purposive, that is directed, and data collected until saturation point,
thereby giving credibility to the study. A research protocol informed from a review of
the relevant literature and past studies was developed, with the main issues around
patient satisfaction and meal experience being explored. Perceived temperature and
texture have been previously identified (Hartwell, 2004) as the two most significant
factors in the evaluation of patient satisfaction with hospital food service and therefore
food quality issues were expanded.
Patients were representative of the patient population and included males and females in
the age range 25-68 years with a length of stay ranging from four days to seven weeks.
Views and opinions were eagerly expressed and recorded where possible, other wise
notes were written-up immediately afterwards.
Food wastage data were entered directly into a pre-prepared spreadsheet for analysis.
Both trolley waste and plate waste were calculated for the cook-chill system and plate
waste for ‘Steamplicity’ using formulae within the spreadsheet.
Results from the questionnaires were coded and entered into spreadsheet (Excel) and
checked for accuracy. The data were then imported into the statistical program
“Statistical Package for the Social Sciences” (SPSS) and analysed to address the overall
purpose of the study. Means and standard deviations were calculated and t-tests for
unrelated scores used to compare the results: levels of statistical significance used were
p = <0.05. Interviews were analysed by content analysis which allowed for developing
themes to be incorporated and a hierarchical flagging system was established.
Authenticity was ensured by including raw narrative within discussion.
Results and preliminary discussion
A comparison of the factors associated with the two systems is given in Table 1.
Insert table 1 here
As can be seen from these results, ‘Steamplicity’ scored higher for all variables,
although not all of these were significantly different. Results for the two variables
associated with food choice are similar; indicating that despite what at first sight might
appear to be a lack of choice with ‘Steamplicity’, it is sufficient, although care must be
taken with this assumption as patients would not have experienced ‘Steamplicity’ long
enough to produced menu fatigue.
Food ordering is important and results for the ‘Steamplicity’ system are significantly
higher; patients are well aware of what is available. This is advantageous in that with
pre-information, consumers can mentally plan what to eat, not only for the next meal,
but also for subsequent meals. There were no significant differences in the ability of
patients to select their own meals; similarly, there were no significant differences for the
two variables associated with food delivery. All of these factors have the potential to
affect acceptability and enjoyment hence the lack of significance is important.
Food quality scores for ‘Steamplicity’ were higher for all four attributes although only
two of these were significantly different. The most likely reason for this is the high
spread of mean values as indicated by the standard deviation, but only further testing
would confirm this. Again with the overall impression of the meal, the mean values for
‘Steamplicity’ were higher with two being significantly different, perhaps because of
the higher Standard Deviation. One variable that might be of concern is the portion
size, and it is encouraging to note that the mean value was significantly higher for
‘Steamplicity’ indicating that patients were satisfied with the size of the portion as
Food Acceptability Download full-text
Due to the limited number of different foods selected, foods have been grouped by
category, i.e. meat, fish etc and an analysis undertaken. These results are given in Table
Insert table 2
As can be seen, the overall acceptability of the food groups is much higher with the
‘Steamplicity’ system, in most cases, significantly. However, care must be taken with
these results as the number of groups, particularly rice, is quite small, and the
comparison is with food groups rather that dish. Even so, there appears to be a clear
preference for the food items.
Food wastage with ‘Steamplicity’ is considerably lower. The plate waste at the midday
meal was 20% and at the evening meal 13%, giving a mean of 16.5%. Wastage with
the cook-chill system was 27% from the trolley and 22% from the plates, giving a total
of 49%. It is perhaps interesting to note that plate waste was 5.5% higher with the
cook-chill, indicative perhaps of food quality.
Patient Food Intake
The total weight of food consumed in the ward at each meal was divided by the number
of main courses served. Results for the Cook-chill are given in Table 3 and for
‘Steamplicity’ in Table 4.
Insert table 3 and table 4 here