Quantifying and describing production waste in two urban healthcare centres with differing foodservice models

Abstract

Aim

To quantify, describe, and compare production waste and food packaging waste in two healthcare centres with different foodservice models.

Methods

In this observational study, all food wasted during production and all food packaging was measured by weight and cost over a 7‐day period for two hospitals: Hospital A (800 beds, cook‐on demand fresh, a la carte menu room service model) and Hospital B (60 beds, traditional cook chill retherm model with weekly cyclical menu and set meal times). Comparisons between sites were made per patient overnight bed day, as an indicator of hospital activity. The average daily waste was calculated, and costed by multiplying cost (AU$) food item/ kilogram by measured weight; to derive the cost of food waste per kilogram. Food waste by weight and cost per overnight bed day was used as a measure to compare the hospitals.

Results

Over 7 days, Hospital A wasted 916.4 kg (AU$6937.22) and Hospital B wasted 69.7 kg (AU$417.10) of food. Most food wasted at both hospitals was edible (78.6% Hospital A and 93.5% Hospital B) and trayline contributed more food waste than bulk production. At both hospitals, plant‐based foods comprised the greatest proportion of edible food waste by weight, but not by cost. When accounting for hospital activity, food waste was lower at Hospital A than at Hospital B (0.15 kg of food per overnight bed day and 0.30 kg, respectively). Hospital A generated 199.5 kg (0.03 kg per overnight bed day) of packaging waste compared to 32.2 kg (0.14 kg per overnight bed day) at Hospital B.

Conclusion

Findings suggest that food waste is costly, that food waste differs between hospitals with different foodservice models, and that overnight bed days is a useful metric for making comparisons. A cook‐fresh, on‐demand room service foodservice model resulted in less food and packaging waste.

1. INTRODUCTION

In Australia, food waste is a significant problem that has notable environmental and economic impacts. ¹ , ² Consumer institutions in Australia, including hospitals, prisons, and schools, are estimated to produce over 248 000 tonnes of food waste per year. ³ , ⁴

Hospital foodservice departments have long acknowledged their responsibility in reducing and managing food waste. ⁵ These settings have the ability to create significant impacts on reducing food waste due to their sizeable and repetitive operations. They are also a key target area to assist in halving Australia's food waste by 2030, as outlined in the National Food Waste Strategy. ¹ , ⁶ , ⁷ That said, comprehensive food waste data is lacking in healthcare settings. Research has traditionally reported on plate waste audits ³ , ⁸ to determine how plate waste is related to nutrition outcomes of patients. ⁸ Data quantifying food wasted during the production of meals (such as in bulk production or food produced but unserved from the trayline) does not appear to be routinely measured or recorded. ⁵

In recognition of this, there has been a shift in recent years to broaden the scope of measuring food waste in hospitals to include kitchen production waste—that is, food and packaging waste as a result of the production and service of patient meals. ⁹ , ¹⁰ , ¹¹ , ¹² One study provided a snapshot of the quantity and type of food waste and packaging waste in three Australian hospitals using two different foodservice models, ⁹ though comparison between hospitals was not possible without data on the overall number of trays or patient bed days. A second study comparing production waste in a single hospital, pre‐ and post‐implementation of a room service foodservice model, found that the implementation of a room service foodservice model appeared to significantly reduce production waste compared to a thaw‐retherm foodservice model. ¹¹ Two additional studies measured kitchen waste from hospitals in the United States and New Zealand, but the focus of the research was primarily on exploring reasons for waste ¹⁰ and the environmental impact of hospital food waste. ¹²

The lack of methodological consistency in measuring production waste, as well as irregular measurement and recording of production waste, poses additional barriers to making meaningful cross‐study comparisons. Some guidelines ¹³ recommend that production waste should be less than 10% of total production, but there is limited evidence to support this target and no standardised method for measurement. Available research has measured waste for 1 day, ¹² 10 days, ¹¹ or 14 days ⁹ and for entire days or by meal. Cook et al ¹⁴ recently developed a hospital food waste consensus tool to assist in standardising the measurement of food and food‐related waste, with the preferred methodology to measure plate and production waste over 14 days using electronic scales within the foodservice kitchen.

The research conducted to date offers insight into the value that measuring production waste can offer to individual institutions and suggests that some foodservice models may produce less food waste during production. However, more work is needed in this area to explore the reason for waste, the source of waste, and differentiate between edible and inedible food components. Likewise, identifying a metric that would allow for benchmarking across sites would be useful. The aim of this study is to quantify and describe the types of production waste in two healthcare centres with different foodservice models and to compare production waste by patient bed day across the two sites. This study will also quantify, describe, and compare food packaging waste for these two sites.

2. METHOD

A cross‐sectional analysis of bulk production waste was conducted concurrently over 7 days at Mater South Brisbane (Hospital A) and Mater Private Hospital Redland (Hospital B), both located in Queensland, Australia, in October 2023. Production waste included all waste generated from procured food destined for patient consumption through the bulk production of food, traylines, and associated packaging waste, but not including plate waste. The audit methodology was developed utilising the hospital food waste audit consensus tool. ¹⁴ Over the 7‐day audit, all food and packaging related to production waste set for disposal was individually described and weighed. Kitchen scales were used to measure food and packaging waste, and large bulky items like cardboard boxes were measured using platform scales. Different scales were used at each site due to the concurrent nature of data collection. All scales were calibrated prior to data collection; brand names for scales are not available. Permission and support were obtained from the Dietetics and Foodservices Department responsible for both hospitals. Consistent with the NHMRC National Statement in Ethical Conduct in Human Research, ¹⁵ approval by a human research ethics committee was not required for this research.

Hospital A is a private hospital but includes contracted public health services, with a total of 800 beds and an average length of stay of 2.5 days. Hospital A has two Emergency Departments and provides acute care services to a range of case mixes designated by subgroups in general medical, surgical, and oncology wards, as well as providing maternity services and a subacute rehabilitation unit. This site utilises a cook‐fresh room service model with on‐demand electronic menu ordering. Patients can order meals from an à la carte restaurant‐style menu available anytime between 6:30 am and 7 pm. Meals are trayed, when ordered, from a trayline that holds hot meals in small batches or prepares meals on demand. All patient meals in Hospital A are made to order and distributed via trolleys within 45 min of ordering by in‐house food services staff. All meal components are produced on‐site at Hospital A.

Hospital B is a private hospital and has 60 beds with an average length of stay of 1.8 days. Hospital B provides acute care services to a range of case mixes comprising general medical, surgical, and oncology, as well as a small cohort of rehabilitation patients. This site utilises a traditional retherm foodservice model. The meals are prepared off‐site at a central production kitchen and delivered frozen in single‐use aluminium trays that have been sealed in plastic and packaged in cardboard boxes. Patients order meals using a paper menu of set meal options for consumption the next day. Meals are trayed at set mealtimes and plated hot from the trayline. The menu is on a 7‐day rotation, and the meals are delivered at set mealtimes. A reduced range of foods is prepared onsite, including sandwiches and salads.

Data collection was completed concurrently and consistently at each site by two teams of eight final‐year postgraduate dietetic students from Bond University over 7 days (Monday–Sunday). Data collection was scheduled as two shifts of two students per day between 9 am and 8 pm. A 7‐day audit period was chosen over a 14‐day audit to minimise ongoing disruption for the kitchen staff and to cover at least one full menu cycle. Student researchers were provided with a 1‐h training session on the research objectives and data collection methodology before the waste audit, with additional support provided by project leads during data collection. Local foodservice staff were aware of the audit and instructed not to dispose of any waste into large bins until observed by the data collection team. Data collection at Hospital A was continuous across the day given the nature of the room service model, while data collection at Hospital B occurred after each set meal service (breakfast, lunch, dinner). Each item related to production waste was weighed individually to the nearest 0.001 kg using commercially available electronic scales and recorded on paper forms; no packaging waste was excluded. The recorded waste was disposed of in a separate large waste bin to prevent duplicate recordings. Data collected from food waste included: the date measured, the name of the item, weight in grams, food category, edible or inedible, reason for waste, and food group classification according to plant, animal, or other grouping (Table 1). Data collected from packaging waste included: date measured, name of the item, weight in grams, and type of material. The paper forms were transcribed onto an Excel (Microsoft Corporation, 2018) spreadsheet at the end of each shift. Data entered was cross‐checked by two authors against the paper sheets.

TABLE 1.

Definitions of production waste categories.

Food waste	Definition of production waste subcategories
Food category
Bulk production waste
Sandwich	Food waste from the bulk production and sandwiches
Juice	Food waste from the bulk production of fresh juice
Fruit salad	Food waste from the bulk production of fruit salad
Recipe	Food waste from the bulk production of recipes
Trayline waste
Breakfast	Foods wasted from the breakfast plating line
Lunch	Foods wasted from the lunch plating line
Dinner	Foods wasted from the dinner plating line
Reason for waste
Loss of quality	Food (held hot on line) wasted due to reduced appeal to customers (colour, taste, texture) despite still being edible. Example: Food from the breakfast and lunch lines that have been out for an extended period.
Over produced	Hospital A: Food wasted at the end of the day due to overproduction. Hospital B: Food wasted after each meal service period due to overproduction. Example: Food left in Bain‐Maries that cannot be safely stored (i.e., gravy, soup, mixed meals, cooked vegetables or cooked meat) dessert items (i.e., jelly, custard or pastries) and beverages (i.e., cordial, high protein milo or milk).
Spoiled ingredient	Individual food ingredients of inadequate quality or no longer food‐safe before serving to customers. Example: soggy tomatoes, mouldy bread or, a prepared ingredient or recipe, such as roast beef, past its expiration date before it hits the line.
Recipe waste	Edible parts of food wasted due to that part of the food not being needed for the given recipe. Example: bread crusts (for easy chew and children's meals), bread ends, bread corners (trimmed to fit packaging)
Inedible	Inedible parts of food that are removed during processing
Food group
Animal‐based foods
Meat	Meat (beef, lamb, pork), poultry (chicken, turkey, duck), fish, shellfish, and eggs
Dairy	Animal‐based; milk, yoghurt, cheese, custard
Meat‐based produced foods	Foods that are pre‐purchased or produced in‐house and are comprised of multiple ingredients and include meat, such as lamb casserole or chicken curry
Plant‐based foods
Vegetable	Fresh and minimally processed (frozen, tinned) vegetables and herbs
Fruit	Fresh and minimally processed (frozen, tinned) fruits
Cereal grains	Cereal grains and products predominately made from cereal grains, including oats, flour, pasta, bread, breakfast cereal, rice, quinoa
Legume	Dried and tinned beans and lentils, tofu
Vegetarian produced foods	Foods that are pre purchased or produced in‐house and are comprised of multiple ingredients and do not include meat, such as pumpkin lasagna, lentil curry
Other foods
Pantry and sauces	Items typically used in the production of other foods and condiments, including baking powder, tomato paste, vegetable stock, salad dressings, tahini, mustard, vinegar, tomato sauce
Beverages	Non‐dairy nutritious liquids including; Juices, cordial, Milo
Discretionary	Foods high in fat, sugar, and/or salt, including processed meats, muesli bars, meat pies, cakes, and chips
Packaging waste
Type of material
Hard Plastic	Rigid plastic packaging that cannot easily be dented, crushed or pierced. Example: 1–2 L bottles of; milk, custard, juice
Soft Plastic	Flexible plastic packaging that can easily be scrunched into a ball. Example: Bread bags, cling wrap, plastic bags containing meat or vegetables
Cardboard	Stiff multilayered and corrugated paper usually in the form of cartons and boxes that hold large quantities of food products during transport and storage.
Glass	Inorganic solid material that is usually transparent or translucent as well as hard, brittle and impervious to the natural elements.
Metal	Rigid cans and tins, primarily made of metals such as aluminium, tinplate or steel. Example: tins of fruit, cans of beans.
Foam	Expanded polystyrene (EPS) packaging in the form of boxes or thin trays
Other	Packaging waste made of materials not found above or a combination of the materials above and cannot be separated. Example: Tetra Pak's (Paper, Polyethylene and Aluminium), Baking Paper, meat twine and netting

Costs of food items purchased were provided by the procurement team through an export from the financial operations system. This data set included a list of food items purchased during the 6 months prior to data collection, the purchase unit, and the price. Additionally, Hospital A retrieved the cost of recipes from the electronic menu management software. The number of overnight bed days during the data collection week was provided by the Finance and Operations Manager for each site. Overnight bed day is a common and somewhat useful metric for foodservices as it indicates the level of patient activity and is related to overall patient meal production requirements, excluding nil by mouth or fasting patients and those receiving meals who do not stay overnight. It broadly allows comparison to be made between sites in regard to patient volume and subsequent food production requirements within a hospital.

Production waste data cleaning and analysis were completed in Excel. Data cleaning involved creating uniform naming conventions, amending incorrectly entered classifications (food category, reason for waste, food group and type of material) and excluding items with insufficient information to be categorised. A descriptive analysis of food waste was performed separately by food category, edible and non‐edible, reason for waste, and food group as an aggregate of the 7 days. Analysis of food groups excluded inedible food. Packaging waste was analysed separately by type of material.

Each analysis calculated the total, percentage, and daily average waste by weight to 0.1 kg and relative cost in Australian dollars for each subcategory. The average daily waste was calculated by dividing the total waste for the week by the 7 days of the audit. Costing of food waste was calculated by multiplying the procurement cost of each food item in Australian dollars per kilo by its measured weight. The cost of food waste per kilogramme was calculated by dividing the total cost by the total weight. Food waste by weight and cost per overnight bed day were calculated by dividing the total weight and cost of food waste by the number of overnight bed days at each site.

3. RESULTS

The final dataset included a total of 203 and 167 unique food waste items weighing 916.4 kg and 69.7 kg with an estimated cost of $6937.22 and $417.10 for Hospitals A and B, respectively (Tables 2 and 3). As a proportion of the total food procurement budget, approximately 7% of foods purchased in each site were wasted (7.3% for Hospital A; 7.4% for Hospital B). Hospitals A and B excluded 13 kg and 29 kg of food waste, respectively, from analyses due to being unclassifiable or not production waste. The large proportion of excluded data at Hospital B was largely due to poor description at the time of data collection. As such, accurate descriptions and cost estimates could not be applied.

TABLE 2.

Description of production waste by weight over 7 days.

		Weight (kg) Hospital A				Weight (kg) Hospital B
	# unique items	Total weight	% total weight	Daily average (kg)	# unique items	Total weight	% total weight	Daily average (kg)
Food waste
Total	203	916.4	100.0%	130.9	76	69.7	100.0%	10
Edible vs. inedible waste
Edible	181	720.7	78.6%	103.0	61	65.2	93.5%	9.3
Inedible	22	195.7	21.4%	28.0	17	4.5	6.5%	0.6
Food waste category
Bulk production
Sandwich	21	59.2	6.5%	8.5	8	18.9	27.1%	2.7
Juice	6	90.2	9.8%	12.9	0	0	0.0%	0
Fruit salad	14	92.6	10.1%	13.2	4	0.9	1.3%	0.1
Recipe	54	77.3	8.4%	11.0	2	0.6	0.9%	0.1
Trayline waste
Breakfast	28	79.1	8.6%	11.3	14	8.3	11.9%	1.2
Lunch	76	110.8	12.1%	15.8	38	26.2	37.5%	3.7
Dinner	126	407.2	44.4%	58.2	33	14.8	21.3%	2.1
Reason for waste
Bulk production
Loss of quality	19	25.6	2.8%	3.7	0	0	0.0%	0
Over produced	50	51	5.6%	7.3	7	1.3	1.9%	0.2
Spoiled ingredients	11	15.3	1.7%	2.2	2	0.7	1.0%	0.1
Recipe waste	5	35.7	3.9%	5.1	1	17.8	25.5%	2.5
Inedible	18	191.6	20.9%	27.4	4	0.5	0.7%	0.1
Trayline waste
Loss of quality	69	155.5	17.0%	22.2	0	0	0.0%	0
Over produced	120	396.3	43.2%	56.6	48	43.1	61.8%	6.2
Spoiled ingredients	26	30.8	3.4%	4.4	5	2.2	3.2%	0.3
Recipe waste	5	10.5	1.1%	1.5	0	0	0.0%	0
Inedible	5	4.1	0.4%	0.6	14	4	5.7%	0.6
Food groups (edible only) a
Animal based foods
Meat	36	105.7	14.7%	15.1	9	3.8	5.8%	0.5
Dairy	10	17.3	2.4%	2.5	5	0.4	0.6%	0.1
Meat‐based produced foods	9	74.2	10.3%	10.6	11	9.8	15.0%	1.4
Plant‐based foods
Vegetable	49	227.8	31.6%	32.5	12	11.4	17.5%	1.6
Fruit	18	18.7	2.6%	2.7	5	1.2	1.8%	0.2
Cereal grains	18	121.2	16.8%	17.3	4	24.2	37.1%	3.5
Legume	5	12.6	1.7%	1.8	2	0.8	1.2%	0.1
Vegetarian produced foods	12	86.2	12.0%	12.3	7	7	10.7%	1
Other foods
Discretionary	8	18.5	2.6%	2.6	2	0.9	1.4%	0.1
Pantry and sauces	6	7.2	1.0%	1.0	3	5.6	8.6%	0.8
Beverage	10	31.4	4.4%	4.5	1	0.1	0.2%	0

Note: The total measured food waste from the 7‐day audit by unique item, total weight (kg), percentage weight, and daily average weight from Hospital A and B respectively. Each of these calculations was further broken down by the defined food waste classifications and subgroups defined in Table 1.

^a Calculations for ‘food groups’ were only based on the edible food measured.

TABLE 3.

Description of production waste by cost over 7 days.

	Cost ($AUD) Hospital A				Cost ($AUD) Hospital B
	Total cost	% total cost	Cost per kg	Daily average	Total cost	% total cost	Cost per kg	Daily average
Food waste
Total	$6937.22	100.0%	$6937.22	$991.03	$417.13	100.0%	$5.98	$59.59
Edible vs. inedible waste
Edible	$5588.76	80.6%	$7.75	$798.39	$388.17	93.1%	$5.96	$55.45
Inedible	$1348.46	19.4%	$6.89	$192.64	$28.96	6.9%	$6.37	$4.14
Food waste category
Bulk production
Sandwich	$287.63	4.1%	$4.86	$41.09	$78.68	18.9%	$4.17	$11.24
Juice	$532.92	7.7%	$5.91	$76.13	$0.00	0.0%	$0.00	$0.00
Fruit salad	$676.27	9.7%	$7.30	$96.61	$24.96	6.0%	$28.33	$3.57
Recipe	$561.69	8.1%	$7.27	$80.24	$11.27	2.7%	$18.15	$1.61
Trayline waste
Breakfast	$333.36	4.8%	$4.22	$47.62	$32.14	7.7%	$3.87	$4.59
Lunch	$1080.39	15.6%	$9.75	$154.34	$185.47	44.5%	$7.09	$26.50
Dinner	$3464.95	49.9%	$8.51	$494.99	$84.60	20.3%	$5.70	$12.09
Reason for waste
Bulk production
Loss of quality	$189.88	2.7%	$7.42	$27.13	$0.00	0.0%	$0.00	$0.00
Over produced	$242.46	3.5%	$4.75	$34.64	$27.34	6.6%	$21.03	$3.91
Spoiled ingredients	$225.49	3.3%	$14.74	$32.21	$22.83	5.5%	$32.61	$3.26
Recipe waste	$100.94	1.5%	$2.83	$14.42	$61.08	14.6%	$3.43	$8.73
Inedible	$1299.74	18.7%	$6.78	$185.68	$3.66	0.9%	$7.32	$0.52
Trayline waste
Loss of quality	$947.55	13.7%	$6.09	$135.36	$0.00	0.0%	$0.00	$0.00
Over produced	$3111.11	44.8%	$7.85	$444.44	$267.64	64.2%	$6.21	$38.23
Spoiled ingredients	$741.70	10.7%	$24.08	$105.96	$9.28	2.2%	$4.22	$1.33
Recipe waste	$29.62	0.4%	$2.82	$4.23	$0.00	0.0%	$0.00	$0.00
Inedible	$48.73	0.7%	$11.88	$6.96	$25.30	6.1%	$6.33	$3.61
Food groups (edible only) a
Animal‐based foods
Meat	$2172.92	31.3%	$20.55	$310.42	$34.11	8.8%	$8.98	$4.87
Dairy	$89.59	1.3%	$5.19	$12.80	$14.86	3.8%	$37.15	$2.12
Meat‐based produced foods	$420.85	6.1%	$5.68	$60.12	$109.25	28.1%	$11.15	$15.61
Plant‐based foods
Vegetable	$1410.51	20.3%	$6.19	$201.50	$76.49	19.7%	$6.71	$10.93
Fruit	$162.87	2.3%	$8.69	$23.27	$25.43	6.6%	$21.19	$3.63
Cereal grains	$258.06	3.7%	$2.13	$36.87	$81.37	21.0%	$3.36	$11.62
Legume	$54.94	0.8%	$4.37	$7.85	$2.93	0.8%	$3.67	$0.42
Vegetarian produced foods	$776.94	11.2%	$9.01	$110.99	$21.18	5.5%	$3.03	$3.03
Other foods
Discretionary	$128.18	1.8%	$6.94	$18.31	$12.76	3.3%	$14.18	$1.82
Pantry and sauces	$37.25	0.5%	$5.19	$5.32	$9.44	2.4%	$1.68	$1.35
Beverage	$76.66	1.1%	$2.44	$10.95	$0.34	0.1%	$3.44	$0.05

^a Calculations for ‘food groups’ were only based on the edible food measured. All costs are in Australian dollars.

Most of the food wasted at both hospitals was edible (78.6% by weight at Hospital A and 93.5% by weight at Hospital B) and in both hospitals the trayline contributed more food waste than bulk production by weight and cost. At Hospital A, the main reasons for waste by weight and cost were overproduction on the trayline (43.2% by weight) and inedible waste in bulk production (20.9% by weight) At Hospital B, the main reasons for waste by weight were overproduction on the trayline (61.8% by weight) and recipe waste in bulk production (25.5% by weight). At both hospitals, plant‐based foods comprised the greatest proportion of edible food waste by weight, but not by cost. By example, at Hospital A animal‐based food waste contributed 27% of waste by weight, but nearly 48% of waste by cost (Figure 1).

FIGURE 1.

(a) Comparison of weight and cost of edible food waste by food group in Hospital A. (b) Comparison of weight and cost of edible food waste by food group in Hospital B.

When accounting for overnight bed days, food waste appeared to be lower at Hospital A than Hospital B (Figure 2a,b). At Hospital A, 0.15 kg of food per overnight bed day was wasted compared to 0.30 kg at Hospital B. A similar observation was made when comparisons were made by cost.

FIGURE 2.

(a) Weight of food waste per overnight bed day at Hospital A and B. (b) Cost of food waste per overnight bed day at Hospital A and B.

In total, 199.5 kg (0.03 kg per overnight bed day) and 32.2 kg (0.14 kg per overnight bed day) of packaging waste were measured and categorised into seven types of materials at Hospital A and Hospital B, respectively (Figure 3). The largest source of packaging waste at Hospital A was soft plastics constituting 41.9% (83.6 kg). The largest source of packaging waste at Hospital B was cardboard with 37% (11.9 kg). Soft plastic waste mainly consisted of the original packaging of procured items such as meat, vegetables, and bread, with limited soft plastic from repackaging bulk‐produced foods. No glass packaging waste items were found at either site during the audit.

FIGURE 3.

(a) Weight (kg) of packaging waste at Hospital A. (b) Weight (kg) of packaging waste at Hospital B.

4. DISCUSSION

The aim of the current study was to quantify and describe the types of production waste in two healthcare centres with different foodservice models and to compare production waste across the two sites, using patient bed days as an indicator of hospital activity. A secondary aim of the study was to quantify and describe the packaging waste for these two sites and explore differences that may be attributable to the different foodservice models across the two sites. Findings from the current study suggest that food waste is a costly issue for healthcare, that food waste commonly results from overproduction, and reasons for waste may differ between hospitals with different foodservice models. Hospital overnight bed days, as a common measure of hospital activity levels, may be a useful metric for comparing food waste between hospitals of different sizes and with different foodservice models.

Overall, the current study suggested that the hospital with the cook‐fresh room service model produced less waste (as measured by overnight bed day) than the hospital with the traditional retherm model. This finding is novel as we are unaware of any research to present this metric previously. That said, in a pre–post evaluation of the implementation of a cook‐fresh, room service foodservice model of an urban Australian hospital, Neaves et al. found that production waste was reduced significantly compared to the previous thaw‐retherm foodservice model. ¹¹ Previous research among foodservice workers has highlighted the important role that forecasting plays in production waste. ¹⁰ In traditional cook chill/freeze retherm foodservice models, it is likely that there is a much longer time between forecasting and meal service than in cook‐fresh, made‐to‐order room service foodservice models. Likewise, traditional retherm foodservice models are likely to have more food waste as the meals are rethermed in large trays based on forecast or estimated predictions of number and type of meals required, rather than real time number of portions needed. This differs markedly from a cook‐fresh, room service model where product is prepared, typically in smaller batches, and trayed meals prepared on demand or as actually ordered. While we believe these findings are largely explained by the different foodservice production models, the differences between the sites in meal ordering (paper vs. electronic) may have also played a role.

Findings from this study revealed that over 65%–70% of all food waste measured was wasted from the traylines in both hospitals. It is anticipated that there will always be some trayline waste as hospital procedures require that unserved food from traylines cannot be rechilled to ensure food safety and quality. The majority of trayline waste in the hospital with the traditional retherm foodservice model was due to overproduction, but in the hospital with the cook‐fresh, room service model, the majority of trayline waste was attributable to overproduction and loss of quality. This finding is perhaps not surprising. For hospitals with the traditional retherm foodservice model, challenges with forecasting and retherming bulk large trays of food are likely to lead to excessive food waste. In contrast, the cook‐fresh, on‐demand room service model allows for all food on the set menu to be ordered anytime. ¹⁶ , ¹⁷ While this increases the variety of food offered, increases patient satisfaction and nutritional intake, and reduces food waste, it also requires all menu items to be available at all times. All‐day menu availability challenges food services to accurately forecast the required servings of each food item on any given day to prevent overproduction, but may mean that some food waste due to loss of quality is unavoidable.

The current study presented all food waste findings by both weight and by cost. These two metrics are needed depending on the reason for conducting the waste audit (e.g., estimating budgetary impact or considering options for waste disposal). For most indicators, the findings were similar by weight and by cost. The most notable exception was with the animal‐based food waste. In both hospitals, food waste from animal‐based foods comprised approximately 40% of the foods by cost. This proportion was much smaller when comparing by weight. Apart from the notable implications for healthcare budgets, the environmental costs are extreme. Animal‐based foods are far more resource‐intensive to produce and dispose of. ¹⁸ , ¹⁹ Future opportunities to explore methods for reducing animal‐based foods on the hospital menu, while maintaining protein quality, are warranted.

Food packaging contributes a significant amount of waste, and it appears there is more packaging waste per overnight bed day in the hospital with the traditional, retherm foodservice model than in the hospital with the cook‐fresh, on‐demand room service model. It is notable that the major sources of waste varied by site, and that may reflect the nature of the foods that are available onsite (produced onsite or pre‐purchased) by differing menus and foodservice models. These findings have implications for healthcare organisations, with regards to procurement (working with suppliers to reduce the use of soft plastics) and disposal (through re‐use and recycling).

The strengths of the current study are in the comprehensive collection of production waste and food packaging, and the descriptions of this waste, across 7 days. That said, there are a few limitations to consider when interpreting these findings. First, it is recommended for hospital food waste audits to be conducted over 14 days. ¹⁴ For the current study, we deemed 7 days of auditing to be sufficient as Hospital A has a static, room service menu and Hospital B has a 7‐day menu cycle. While individually measuring production waste items allowed for a more robust data set to inform the scale of production waste, the process was extremely labour‐intensive and may not offer a practical method for future research evaluating interventions to reduce food waste. The process of measuring food waste was not perfect. Errors occurred through waste being disposed of by kitchen staff before weighing or not being able to accurately describe the waste. This limitation is particularly notable for Hospital B, where a higher proportion of data were excluded for this reason. While overnight bed days is a measure of activity level of the hospital, it recognises patients in beds rather than the direct number of meals produced. Patients who are nil by mouth or fasting and those receiving meals who do not stay overnight should ideally be excluded. However, it broadly allows comparison to be made between sites in regard to patient volume and related food production requirements within a hospital. The accuracy of packaging waste may be limited across sites, with residual food remnants within the packaging slightly inflating the measured weight. Additionally, the total weight of cardboard at Hospital A had to be estimated by counting the individual pieces and multiplying by a common weight due to impracticalities and safety concerns of individually measuring cardboard waste. Increasing the capacity and awareness of foodservice staff and researchers performing data collection, as well as addressing inefficiencies in the methodology of future audits, is likely to increase the overall accuracy of the data. Technological applications, such as those that use AI to monitor food waste, may offer practical and efficient strategies for measuring food waste in a more sustainable way.

Findings from the current study provide a comprehensive overview of the quantity and nature of foods wasted during production in two urban acute healthcare settings with different foodservice models. We found that by using patient overnight bed days, we could make comparisons of food waste between the two sites, irrespective of hospital size or foodservice model. Our findings suggest that a cook‐fresh, on‐demand room service foodservice model results in less food waste through production and less packaging waste, but more research in other hospitals is needed to confirm these findings. Moreover, research that evaluates the impact of strategies to reduce food waste in hospitals, investigates and trials methods for repurposing food waste, and explores methods for circular waste disposal is urgently needed.

AUTHOR CONTRIBUTIONS

KB, JU and SM designed the research study and protocol and supervised all aspects of the study. NC and JH oversaw data collection and data analysis. NC, JH and JU drafted the manuscript. All authors critically reviewed the manuscript and approved the final version. The authors acknowledge the Bond University Master of Nutrition and Dietetic Practice students (2024 graduating cohort) for their assistance in data collection.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflicts of interest.

Cook N, Habel J, McCray S, Utter J, Brennan K. Quantifying and describing production waste in two urban healthcare centres with differing foodservice models. Nutrition & Dietetics. 2025;82(4):434‐444. doi: 10.1111/1747-0080.70013

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions.