Hospital healthy cafeteria initiatives: a systematic review of consumer behaviour related studies

Unhealthy food has become a concerning issue in the context of hospital cafeterias. Cafeterias in hospitals have been value - added with healthy menu offerings and safe eating practices to promote healthy eating to customers. Research in food service operation precisely in hospital healthy cafeterias is abundant. However, the scope of the study, study design and study setting varied. This paper aimed to present a systematic review of hospital healthy cafeteria initiatives - based research within the context of consumer behaviour. Published research between the year 2000 through 2020, conducted within the hospital cafeteria was assessed. The main criteria of selection were the existence of a comprehensive description of the instrument, questionnaire, or interviews used to gather the relevant data. Initial 53 articles underwent careful evaluation independently to ensure that they were empirical research carried out within the context of healthy cafeterias at hospitals. As a result, 29 articles were reviewed. The scope of the study includes consumer behaviour, the effect of promotion on sales in healthy cafeterias, management and service personnel, the audit of the compliance of healthy cafeterias and the effects of food category or nutrition labelling intervention. Quantitative research design has been used in most research compared to qualitative and mixed methods and the subject of studies involved hospital staff, visitors and food service managers. The findings of this review implied that healthy cafeteria initiatives have been introduced in the hospital food service. But more consumer behaviour research, continuous food service staff training, strategic promotion and intervention programs for healthy menus, eating habits and lifestyles could be implemented to make hospital healthy cafeterias successful.


Introduction
Globally, the prevalence of obesity as well as noncommunicable diseases (NCD) including diabetes and hypertension continue to rise over the years due to unhealthy eating practices (Burgoine et al., 2014;Bennett et al., 2018;Jilcott Pitts et al., 2018). It is commonly argued that individuals who would often consume meals away are more suspectable of NCDs (Sahud et al., 2006;Wu and Sturm 2014;Jackson et al., 2016;Onufrak et al., 2019). As of 2016, it was reported that an estimated 40.5 million (71%) of the 56.9 million worldwide deaths were caused by NCDs of which 15.2 million (38%) were amongst individuals aged between 30 years and 70 years, and 23.6 million (58%) above the age of 70 years (Bennett et al., 2018).
A more alarming fact is that the availability of unhealthy food within the setting of hospital cafeterias has also become an issue of concern (Lesser et al., 2012;Lederer et al., 2014;Harel et al., 2015). In several western countries, hospital cafeterias were heavily criticized by scholars for not upholding a positive image of promoting healthy eating practices by selling customers less healthy food as well as permitting fastfood franchises to operate inside the hospital premises (Auer, 2006;Freedhoff and Stevenson, 2008;Lesser et al., 2012;Bell et al., 2013;Harel et al., 2015).
In an attempt to encourage healthy eating which would potentially lower the prevalence of NCD, the healthy cafeteria initiative was launched and implemented beginning with public hospitals before being expanded into other institutions (Abdul Rais et al., 2019;Abdul Rahman et al., 2020). The general aim of this initiative is to foster healthy eating practices among customers by limiting the options for comfort food whilst offering wider varieties of healthier alternatives especially fresh fruits and low-calorie snacks (Dawson et al., 2006;Bell et al., 2013;Donohoe Mather and McGurk, 2014;LaCaille et al., 2016;Moran et al., 2016). Various strategies including food labelling colour coding, stealth marketing, pricing incentives, loyalty programs and food shelf manipulations were imposed to maximize and encourage better healthy food selections at the cafeteria (van Kleef et al., 2012;Thorndike et al., 2014;Lindeman et al., 2016;Patsch et al., 2016;Thorndike et al., 2016;Chan et al., 2017).
To better understand the sustainability of the healthy cafeteria initiative, scholars have done research evaluating various elements. From the perspective of the customers, studies would often focus on customers' acceptance, satisfaction, and food purchase patterns. Meanwhile, from the perspective of the cafeteria operator, studies would highlight staff commitment, the effectiveness of food labelling, and sales figures for healthy food. A holistic understanding is indeed important to ensure that healthy cafeteria initiatives fulfil the long-term aim of producing a healthy society. Therefore, this paper aimed to compile and review the literature relating to the implementation of hospital healthy cafeteria initiatives within the context of consumer behaviour.

Search parameters
A systematic literature search with electronic databases was conducted for articles published from the year 2000 through 2020. The main criteria of selection were the existence of a comprehensive description of the instrument, questionnaire, or interviews used to gather the relevant data. Articles were searched inside Scopus and manual search through Google Scholar using the keywords of 'hospital healthy cafeteria' and 'hospital cafeteria satisfaction'. A total of 53 search results were obtained. Each title, abstracts, and full text were evaluated independently to ensure that they were within the context of healthy cafeterias at hospitals. Articles that are not within the context of being non-empirical were excluded. Research involving questionnaires or electronic assessment tools was given particular care to assess the validity of the tools described in the articles. Exploratory factor analysis, confirmatory factor analysis, per cent agreement, kappa statistics, and intraclass correlation coefficient or Cronbach's a were particularly looked for.

Sample size and population
Two types of samples were identified which are human respondents and hospital cafeterias. Crosssectional quantitative studies would have a sample size ranging from 321-602 human respondents. Longitudinal intervention studies would have larger samples ranging from 1003 to 5695 respondents. As for qualitative studies, a smaller sample size was observed between 8-41. For studies relating to cafeteria scoring in terms of sales and compliance with regulations, the number of samples ranges from 14-39.

Data collection
Data relating to variables of interest in the context of the hospital's healthy cafeteria were assessed using either quantitative, qualitative and mixed methods. The quantitative studies are administered either using questionnaires or electronic assessment tools. Crosssectional studies would last between 2-7 months. Meanwhile, the longitudinal studies would take up to 24 months to complete. Numbers and types of items differ in each research depending on the variables measured ranging from 9 to 57. Nine studies assessed the validity and reliability of research instruments either through Cronbach's alpha, exploratory factor analysis, confirmatory factor analysis, per cent agreement, kappa statistics, and intraclass correlation coefficient. The qualitative studies were accomplished through interviews or focus groups conducted by at least 2 people which usually last between 45 mins to 2 hrs. To ensure the accuracy of the information, some interviews were even recorded.

Results
A total of fifty-three search results were obtained through Scopus and another three were obtained through Google Scholar using the mentioned keywords. Any similar articles that appeared with the Google Scholar search or irrelevant articles were ignored. Upon reviewing the content of each article, a total of 24 were excluded due to them being conceptual papers, case studies, and not being related to the healthy cafeteria at hospitals. The remaining 29 papers are compiled in Table 1.

Aims of the studies reviewed
A total of twenty-nine studies published from 2000 to 2020 are reviewed to determine the current research regarding hospital healthy cafeterias which look into the perspectives of consumer behaviour, the effect of promotion on sales in the healthy cafeteria, the attitude of service personnel, auditing the compliance of healthy cafeterias and effects of food category or nutrition                Based on the reviews, eight (8) studies are related to consumer behaviour, and aimed to assess customers' perception, acceptance, satisfaction and purchase intention. Six (6) of the studies applied quantitative data collection methods, while another two (2) studies applied qualitative and mixed methods respectively. The subjects of studies involved staff and visitors (for 4 studies), one (1) study carried out research among staff and café operators, two (2) studies were conducted among restaurants' staff, and one (1) study involved parents and food service managers.
A total of nine studies related to sales performance and management of hospital cafeterias were included in the review. A total of five studies evaluated the effect of promotion materials and food labelling on sales figures at healthy cafeterias. Whilst four studies focused on management practices mainly assessing how staff perceived healthy cafeteria implementation and barriers that need to be overcome.
Research focusing on the dietary quality and health in hospital cafeterias was also included in the review. These include two studies aimed to evaluate dietary quality and health of healthcare staff eating at the cafeteria, using multiple tools to gather data for assessment. On the other hand, four (4) studies that audit the compliance of healthy cafeterias, used traditional pen and paper as well as electronic scoring instruments. The majority of studies (6 studies) evaluate the effects of food category or nutrition labelling intervention on healthcare workers' purchase patterns as well as changes in calorie intake.

Discussion
Consumer behaviour related studies involving a consumer who directly dine at hospital cafeterias have reported from average (Abdul Rahman et al., 2020) to good level of acceptance and satisfaction (Dawson et al., 2006;Kimathi et al., 2009;Rafidzah et al., 2020) and positive perceptions (Azizan et al., 2016). A similar positive perception was also observed amongst parents and foodservice managers despite the low availability and labelling of healthy food options (Bell et al., 2013). Focus group interviews amongst customers highlighted that affordability as well as commitment from cafeteria operators as major factors in effectively fostering healthy eating at the hospital cafeteria (Tamrakar et al., 2020). The affordability of healthy food alternatives has also been seen to be an issue concern, especially in developing nations (Cox et al., 1998;Muhihi et al., 2012). Strong administrative commitment and policies to offer subsidies for healthier food options may be implemented to encourage purchases and consumption (Kahn-Marshall and Gallant, 2012;An, 2013). However, it should be cautioned that such strategies would be challenging to implement especially for cafeterias that are profit-oriented in both public and private hospitals (Pridgeon and Whitehead 2013). These cafeteria operators usually operate in urban hospitals are able to offer choices of local and international menu items compared to those operating in rural areas that tend to focus on local dishes based on ingredient availability. As contended by Abdul Rais et al. (2019), re-patronization at a healthy cafeteria can only be achieved through the moderation of perceived value as well as the mediation of eating behaviour and satisfaction. Meanwhile, Kimathi et al. (2009) argued that a high satisfaction or acceptance rate does not always guarantee repatronization (Adebanjo, 2001) as their studies found that sales figures of healthy food were much lower compared to sales of comfort food. Chu et al. (2014) contended that menu labelling does not significantly hamper sales and revenue thus relieving concerns about the adverse effect of menu labelling on food sales at the hospital cafeteria. However, other studies relating to the sales performance of hospital healthy cafeterias reported that the intervention of colour labelling, nutrient labelling, health promotion and price intervention can indeed increase sales purchase of healthy food (Sato et al., 2013;Sonnenberg et al., 2013;Thorndike et al., 2014;Patsch et al., 2016). Sales figures of selected healthy food can increase as much as 8% through the appropriate intervention means. The findings are in line with the advocation of Chan et al. (2017) that price discounts can contribute about 5% to a sales increase of healthy food. However, to maximize sales, the authors suggested adapting reward points that can boost sales as high as 28%.
A number of studies have also taken the route of evaluating the hospital's healthy cafeteria initiatives purely from the operator's point of view. As highlighted by Lederer et al. (2014), some cafeteria managers perceived that lack of demand for healthy foods was the main limitation in making healthy changes to hospital cafeterias. On-going training is essential for improvements to the food environment at the hospital cafeteria, especially for those staff who may lack the knowledge and experience in nutrition. Managers also tend to express inadequate selections from vendors, customer complaints negative responses by staff, difficulties with marketing changes internally and externally as well as additional expertise required for menu labelling as potential barriers that need to be eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER overcome (Dauner et al., 2011;Jilcott Pitts et al., 2016). In order to financially sustain a hospital healthy cafeteria, Jilcott Pitts et al. (2018) posed that the best practices for boosting sales of healthy food include evaluating well-received new dishes, shortlisting popular healthy food recipes through point of sale (POS) and nutrition analysis software, launch tastier healthy food options, higher food costs for the new healthier menu can be offset by adjusting the price point of other offerings, realign employee and customer discounts to favour healthier food choices.
Individual eating behaviours are increasingly acknowledged to be influenced by major food environment characteristics.
Essentially, health institutions need to be consistent in creating healthy food environments that fulfil healthy dietary guidelines. Several studies assessing the effectiveness of hospitals' consumer food environment have been conducted in the United States (Winston et al., 2013a;Derrick et al., 2015) and Australian healthcare settings (Tsai et al., 2018). Psychometrically tested tools used were the Hospital Nutrition Environment Scan for Cafeterias, Vending Machines, and Gift Shops (HNES) (Winston et al., 2013b) and paper-based audit tools designed to evaluate the key food environment guidelines (Tsai et al., 2018). Findings from these studies highlighted that the average consumer nutrition environment of the participating hospitals was found to be less supportive of healthful eating, thus warranting nutrition-improvement interventions (Winston et al., 2013a;Derrick et al., 2015;Tsai et al., 2018). Indeed, the influence of healthy hospital initiatives in improving the nutrition environments within hospital cafeterias was evident. The cafeterias in health initiative hospitals that posted calories on the menu and labelling "healthy" foods were reported to have recorded far greater mean nutrition composite scores compared with hospitals that are not part of the health initiative (Derrick et al., 2015). Other empirical evidence also pointed to the importance of creating good consumer nutrition using a point-ofpurchase intervention that could deliver a large impact on consumers' food choices (Webb et al., 2011;Levy et al., 2012;Van Kleef et al., 2012;Vanderlee and Hammond, 2014;Thorndike et al., 2016). Continuous and objective audits assessing the nutrition environments of hospital settings are therefore vital to eliminate selfreport bias and to ensure the success of compliance with government policies relating to healthy food and beverages considering the health facilities as public sector institutions that should serve as a good benchmark in exposing consumers to proper healthy food environments.
Healthy eating interventions at workplaces are reported to have the potential in improving dietary intake, hamper weight gain, and improve employees' health, thus trimming substantial healthcare costs of obesity-related illness the employers. A longitudinal study by  from 11 large urban hospitals in the US showed that the cafeteria healthy eating program that utilizes traffic light labels and choice architecture resulted in a sustained decrease in calories purchased, documenting a 6.2% decrease in calories per transaction over 2 years. Using a dynamic model of weight change, the employees who were frequent purchasers were estimated to lose weight up to 2 kg over 3 years after the implementation of the program, assuming that compensatory changes in diet or physical activity did not take place. In contrast, a 12-month multicomponent obesity prevention program based on a social -ecological framework involving employees from a midsize healthcare system in the US has shown no effects in hampering weight gain (Lacaille et al., 2016). Several other studies that used similar ecological approaches have also reported mixed findings, recording very small to no success in weight gain prevention (Pratt et al., 2007;French et al., 2010;Goetzel et al., 2010;Lemon et al., 2010;Linde et al., 2012;Williams et al., 2014) Although the beneficial effects seen in weight changes were substandard, significant positive impacts on the other health behaviours such as greater increases in walking activity, nutritional knowledge, peer health discussions and positive attitudes were apparent (Lacaille et al., 2016). It was suggested that implementing lowintensity physical activity such as walking and stair climbing may be easily intergraded into the worksite environment compared to utilizing far more complex behaviours of changing dietary habits and involving more vigorous exercise (Lacaille et al., 2016). Of note, the overall diet quality and cardiometabolic risk of an American hospital's employees (n = 602) were improved following a worksite healthy eating intervention using traffic-light food labelling . Given the high proportion of poor dietary quality foods commonly obtained at work (Onufrak et al., 2019), improvements to the diet quality were particularly instrumental for it is known to lower the risk of major chronic disease, regardless of weight change (Sotos-Prieto et al., 2017). These findings showed that incorporating simple behavioural nudges in the worksite food environment such as choice architecture and food labelling could provide more excess towards healthier choices, hence improving employees' overall diet and health.

Conclusion
This review has looked at the 29 studies in the eISSN: 2550-2166 © 2022 The Authors. Published by Rynnye Lyan Resources FULL PAPER hospital healthy cafeterias researched on consumer behaviour, management and health concerns among various subjects using different methods, globally. Consumers perceived a healthy cafeteria positively although the healthy menu is not the reason for repatronage because healthy menus are more costly than normal menus. Despite that, it is evident that supplementing a healthy menu may increase sales in healthy cafeterias provided that price discounts and reward-point systems were made available. Healthy cafeteria management may also need to increase the promotion of healthy menus to increase their demand and at the same time invest more in training the staff for better comprehension of the concept of healthy menus in order to prepare them. Healthy eating intervention programs together with the promotion of healthy lifestyles at work were found to be successful in the healthcare premise. Thus, consumer behaviour must be well understood, staff of foodservice premises must be well trained and the promotion and intervention program of healthy menus, eating habits and lifestyles must be regularly implemented in order to improve the health status of both consumers and healthcare workers.