Table 2.
Characteristics and outcomes of included studies (classified by chronological order).
| First Author, Year of Publication, Country |
Study Design | Sample Size | Patient Type | Setting | Intervention Type |
Time of Observation |
Endpoints | Methods of Assessment | Results |
|---|---|---|---|---|---|---|---|---|---|
| Hartwell, 2003 UK [19] |
Observational | n = 62 | Women’s Health (n= 42) and orthopaedic (n = 20) inpatients | Hospital wards (n = 2) | Bulk trolley system = patients choosing food and amount from the trolley vs. Plated meal system = meals are ordered in advance (24 h before consumption) |
3 consecutive days before and after 6 months. |
Nutrient intake Food waste |
Food weighed pre- and post-consumption Microdiet computer software for nutritional content |
No differences between the nutrient content of the food intakes (both lower than recommended dietary values) ↓ plate waste with the bulk trolley service (5.9% vs. 11.6%) but high waste left on the trolley (20.5%) with bulk trolley system. |
| Edwards, 2006 UK [20] |
Observational | n = 52 | Patients presenting a mixture of clinical conditions | NHS teaching Hospital | Steamplicity vs. Cook-chill food service |
3 consecutive days in 2-weeks periods for each arm | Food intake Food waste |
Food weighed before and after the meal using digital weighing scales (individual food components separated when possible) | ↑ food intake in Steamplicity than cook-chill system (daily mean of 282 g vs. 202 g at lunch; 310 g vs. 226 g at dinner) ↓ food waste in Steamplicity system then in cook-chill service (33% vs. 49%) |
| Hickson, 2007 UK [21] |
Observational | n = 57 | Patients presenting a mixture of clinical conditions; not at nutritional risk, without eating problems; and able to choose a menu |
Hospital wards (n=7) |
Steamplicity vs. Traditional bulk cook-chill system |
1 lunch meal/patient between March and April 2006 | Energy and protein requirements met with Steamplicity Energy and protein consumption between the two systems |
Estimates of served food portion sizes Food waste weighed to calculate food intake and energy and protein intake (Nutritional analysis program) Comparison consumption and patient requirements |
Steamplicity does not meet the patients’ energy requirements (36% deficit) ↓ energy intake in Steamplicity than in bulk cook-chill system (p = 0.04) No difference in protein intake between the two systems No difference in food wasted; more protein wasted in the Steamplicity system |
| Rufenacht, 2010 Germany [22] |
RCT | n = 36 | Hospitalised patients with NRS-2002>3 |
Internal Medicine hospital ward | NT: Nutritional counselling with a dietitian + ONS vs. ONS (without nutritional counselling) |
10–15 days | Energy intake Protein intake |
Weighing of all meals before and after consumption Energy and protein intake calculated with nutritional software |
NT group met the energy requirements by 107% and protein requirements by 94% ONS group met the energy requirements before discharge by 90% and protein requirements by 88% |
| Hickson, 2011 UK [23] |
Observational | n = 253 | Hospitalised adult patients at high risk of malnutrition | Two large teaching hospitals | “Protected mealtimes” (PM) vs. Standard food service |
June/July 2008: standard food service October/November 2009: PM |
Nutritional (energy and protein) intake Food waste |
Direct observation of meal consumption Weighing food consumed and food waste Evaluation of the intake by nutritional software |
No impact of PM on energy intake (p = 0.25) ↓ protein intake (p = 0.04) in intervention group |
| Manning, 2012 Australia [24] |
Monocentric observational | n = 23 | Elderly inpatients (almost all at risk of malnutrition) | Hospital 2 wards |
Volunteer feeding assistance program vs. No volunteers (feeding provided by nurses) |
2 days for each arm | Energy and protein intake and % of energy and protein requirements met Food waste |
Weighing of remaining food after meal consumption % of each item consumed Estimated energy and protein intake according to requirements |
↑ energy and protein intake at lunch (p = 0.005; p = 0.009) No difference in daily total energy intake (p = 0.113) ↑ total daily protein intake (p = 0.004) ↑ % of energy requirements met with volunteers (64% vs. 58%, with an additional 448 kJ) ↑ % of estimated protein requirements met with volunteers (71% vs. 59%, p = 0.003). |
| Young, 2012 Australia [25] |
Prospective pre-post |
n = 254 |
Inpatients aged >65 years | Internal Medicine wards (n = 3) of a large metropilitan Hospital |
3 mealtime assistance interventions: PM AIN: A nutritional focused staff-member assisting patients with meals PM + AIN: combined intervention |
1 day in the first week of hospitalisation | Daily energy and protein intake | Visual estimation of plate waste (none, 1/8, 1/4, 1/2, 3/4, all) Intake evaluation by nutrient analysis software. |
↑ energy intake, no differences between intervention groups (p = 0.16) ↑ protein intake (p = 0.07), no differences between the three interventions (p = 0.20). ↑ adequate EER (p < 0.01), no difference between interventions (p = 0.29). ↑ adequate protein intake (intake > EPR) (p = 0.03); no difference between interventions (p = 0.57). |
| Maunder, 2015 Australia [26] |
Prospective | n = 119 | Hospitalised adult patients | Private hospital |
Bedside electronic meal ordering system (BMOS) vs. Paper menu (PM) group with default meals |
48 h period × 2 | Energy intake Protein intake |
Use of photography and five-point visual wastage scale (0%, 25%, 50%, 75% and 100% wasted). Estimation of dietary intake by total meal eaten weight and calculated by nutritional analysis software analysis |
In BMOS vs. PM group: ↑ energy intake: 8273 vs. 6273 kJ/day (p < 0.05) ↑ protein intake: 83 vs. 66 g/day (p < 0.05) |
| Navarro, 2016 Israel [27] |
Prospective | n = 206 | Adult hospitalised patients | Hospital Internal medicine ward |
Improved meal presentation vs. standard lunch |
Mean 4.7 days (intervention group) Mean 5.25 days (control group) |
Food intake Food waste |
Digital Imaging Method and visual estimation of plate waste (6-point scale: 0%, 25%, 50%, 75%, 90%, 100%) Estimation of food intake by nutritionDay questionnaire |
+19% of food intake in the intervention group compared with control group (p < 0.05) ↓ starch and main course waste in the intervention group compared with control group (p < 0.05) |
| Collins, 2016 Australia [28] |
Parallel controlled pilot study |
n = 124 |
Elderly subacute patients (38% malnourished at admission) | Hospital, subacute geriatric ward | Modified hospital menu with higher energy foods including ONS (and a visual menu) vs. Control group: standard cook-chill meals (no visual menu) |
14 days/group | Nutritional (energy and protein) intake Food waste |
Visual estimation of plate waste before and after meal consumption Calibrated seated scales or self-reported or medical notes (if unable to be measured) Daily energy (kJ) and protein (g) intake estimated from plate waste data by nutritional software |
In intervention vs. control group: ↑ mean energy intake (132 vs. 105 kj/kg/day; p = 0.003) ↑ mean protein intake (1.4 vs. 1.1 g protein/kg/day; p = 0.035) |
| Farrer, 2016 Australia [29] |
Prospective |
n = 65 |
Acute care inpatients prescribed smooth pureed meals |
Acute care hospital | Smooth pureed meals in a moulded format (intervention group) vs. Smooth pureed meals in the standard format (control) |
2 weeks | Food intake Plate waste |
Weighing meal wastage with calibrated electronic scales | ↑ food intake from <1/4 to >3/4 of the meal in the moulded form (p = 0.03) compared with control ↓ 120 g of plate waste in the intervention group compared with control group even if not significant (p = 0.09) |
| Porter, 2017 Australia [30] |
RCT | n = 149 | Admitted to the subacute setting | 3 hospitals 3 wards n.2 geriatric evaluation and management wards and n.1 rehabilitation wards |
PM (Intervention period) vs. Usual care (Control period) |
4 weeks | Daily energy intake Daily protein intake Daily energy deficit |
One quarter portion method per day; per patient per meal period and per interruption Use of nutritional software to estimate energy and protein intake |
No significant differences between the intervention and control conditions for unadjusted analysis. ↓ energy deficit in intervention periods vs control periods if adjusted for age, nutritional status and type of subacute ward. |
| Strotmann, 2017 Germany [31] |
Case study |
n = 367 |
Hospitalised patients |
Hospital surgery | A package of measures including: - Sensitisation of employees to food waste - Order assistance training - Analysis of the flow of communication along the supply chain - Configuration of a food catalogue with detailed description of meals - Change of order and delivery process - Change of portion sizes according to target group-specific standards and their needs vs. Usual care |
2 weeks | Daily food waste rate (per person) Total food waste rate |
Weighing food before and after consumption using electronic scales | ↓ 20% in the average quantity of food served daily per person in hospital (p < 0.0001) No difference in hospital total waste; rate remained the same after implementing measure |
| Barrington, 2018 Australia [32] |
Observational prospective |
n = 96 (control) n = 105 (intervention) |
Oncologic hospitalised patients | Hospital | BMOS vs. PM group with default meals |
2 weeks | Total food intake Energy intake Protein intake Food waste |
Use of photography and five-point visual wastage scale (0%, 25%, 50%, 75% and 100% wasted). Estimation of dietary intake by total meal eaten weight and calculated by nutritional analysis software analysis |
↑ average energy intake (p < 0.001) in BMOS ↑ average protein intake (p < 0.001) in BMOS ↑ in receiving the food ordered (p < 0.001) in BMOS ↑ in choosing food that patients liked (p = 0.006) in BMOS No significant differences in average plate waste between the groups (34.3% in the BMOS vs. 35.3% in PM, p = 0.75) |
| McCray, 2018° Australia [33] |
Retrospective analysis of data pre- and post-intervention |
n = 148 |
Case mix of patients (general medical, surgical, and oncology wards) | 2 adult care hospitals | Room service (RS) = meals ordered by patients from a “a la carte menu” and delivered within 45 min vs. Traditional foodservice model = meals ordered completing a paper menu (cook fresh, 14-day cycle) up to 24 h before meals |
A 24-h consecutive period for 4 days | Nutritional intake Energy and protein intake as % of requirements Food waste |
Meal intake observation tool using a five-point visual scale (0, 1/4, 1/2, 3/4, all) Nutrition analysis by nutritional software |
In room service intervention vs. traditional foodservice model ↑ mean energy intake (1588 kcal/d vs. 1306 kcal/d; p < 0.005) ↑ mean protein intake (65.9 g/d vs. 52.3 g/d; p < 0.003) ↑ % of requirements of energy (75.1 vs. 63; p < 0.024) and protein (84.7 vs. 65; p < 0.011) intake ↓ total mean plate waste (12% vs. 29%; p < 0.001) |
| Mc Cray, 2018b Australia [34] |
Prospective | n = 187 | Adult hospitalised patients | Acute care hospital |
Food and Nutrition Solutions (FNS) and Room Service ChoiceTM vs. Traditional model (TM) with paper menu |
4 days | Energy intake Protein intake Plate waste |
Meal intake observation tool using a five-point visual scale (0, 1/4, 1/2, 3/4, all) Calculation of the nutritional intake using the FNS software |
Compared with TM group, in FNS group: ↑ energy intake: 6379 vs. 5513 kJ/day (p = 0.020) ↑ protein intake: 74 vs. 53 g/day (p < 0.001) ↑ % of energy requirements met: 78% vs. 64% (p = 0.002) ↑ % of protein requirements met: 99% vs. 70% (p < 0.001) ↓ total average plate waste 17% vs. 30% (p < 0.001) |
| Neaves, 2021 Australia [35] |
Retrospective analysis | n = 210 | Adult hospitalised patients | Large tertiary hospital 3 wards: surgical, thoracic, cystic fibrosis |
RS vs. Thaw-retherm service control group |
5 weekdays | Nutritional (energy and protein) intake % of energy and protein met Food waste |
Visual tool for nutritional intake and plate waste five-point visual scale (0%, 25%, 50%, 75%, 100%) and weight estimation of % wasted food |
In RS compared to control group ↑ average energy and protein intake (p < 0.001). ↓ plate waste (15% vs. 40%) and production waste (5.6% vs. 15%, p < 0.001) ↓ food waste (p < 0.01) ↓ total average production waste (p < 0.001) |
| Razalli, 2021 Malaysia [36] |
Cross-sectional | n = 95 | Adult patients prescribed with texture-modified diet | Hospital | Texture modified diets 3 types: -Blended diet -Mixed porridge -Minced diet |
from 1 to over 7 days | % plate waste % protein plate waste |
Visual estimation of plate waste through Visual Comstock Scale (6-point scale: 0%, 25%, 50%, 75%, 90%, 100%) Digital food weighing scale |
↑ plate waste (65%) in blended diet (65%) than minced diet (56%) and mixed porridge (35%) (based on weighing method) ↑ protein waste (61.1%) in minced diet compared with other diets (based on weighing method) |
| Berardy, 2022 USA [37] |
Prospective | n = 447 | Adult hospitalised patients | Hospital | Type of protein source vegetarian meals (peanut butter, tofu, black beans, brown lentils and hummus) vs. meat-containing meals |
7 days | Total food waste Food waste of categories of food |
Weighing of containers removing container weight Use of recipes for composite foods to determine proportional weights for individual categories of food |
↑ 34.05 g of food waste (p = 0.05) in patients with meat-containing meals compared with vegetarian meals Largest category of food waste in meat-containing meals: vegetables. Largest category of food waste in vegetarian meals: grains and vegetables |
Abbreviations: AIN, Additional assistant-in-nursing; BMOS, bedside electronic meal ordering system; EER, estimated energy requirements; EPR, estimated protein requirements; FNS, Food and Nutrition Solutions; NRS-2002, nutritional risk score—2002; NT, nutritional therapy; ONS, oral nutritional supplement; PM, protected mealtimes; REE, resting energy expenditure; RCT, randomised controlled trial; RS, room service; TM, Traditional Model; UK, United Kingdom; vs., versus; ↑ increase; ↓ decrease.