Preference and willingness to pay for nutritional counseling services in urban Hanoi

Background: Despite substantial achievement in reducing malnutrition rates in Vietnam, there has been an increasing rate of overweight individuals in urban areas, which may result in a high burden of non-communicable diseases. Nutritional counseling clinics have been introduced in several settings; however, little is known about the preference for this service among urban clients. This study aimed to assess the preference and willingness to pay (WTP) for nutritional counseling services among urban clients. Methods: We interviewed 429 clients who attended Hanoi Medical University Nutritional Counseling Clinic (Hanoi, Vietnam). WTP was determined using double-bounded dichotomous-choice questions and open-ended questions. Results: In total, 78.6% respondents were willing to use nutritional counseling services. The mean amount of WTP for one-time service and one-year package was 96,100VND (~$4.3) and 946,400VND (~$41.9), respectively. Clients’ willingness to use the service was higher among females, those seeking counseling for elderly people and those who preferred face-to-face counseling services (p<0.05). WTP was higher among those who were over 35 years old, those seeking services for the elderly people, those having poor nutritional status, and those having under-6 year old children (p<0.05). Conclusions: The preference and WTP for nutritional counseling services in urban Hanoi were relatively high. Scaling up this service is necessary to actively prevent and control the spread of non-communicable diseases.


Introduction
In recent years, Vietnam has achieved a significant improvement in people's health and nutritional status 1 . This is indicated by an improvement in people's knowledge, attitude and practice on nutrition, and a significant decrease in malnutrition rates among children. According to the National Institute of Nutrition, the rate of marasmus and stunting has been reduced from 19.9% (2013) to 14. 1% (2015) and 32.6% (2013) to 24.6% (2015), respectively 1 . However, in urban areas, significant increases in overweight and obesity rates may result in high burden of non-communicable diseases (NCDs) 2 . A survey among 17,213 people in Vietnam showed that the rate of overweight and obesity was 16.3%. This high rate was fueled by unhealthy diet habits, alcohol abuse and sedentary lifestyles 3 .
In developed countries, nutritional counselling has been recognized as an effective measure to improve awareness and encourage a healthy lifestyle, and has been shown to reduce the risk of obesity and NCDs 4 . Nutritional counseling clinics can be organized in co-location with other general health care services or as standalone sites. However, in resource-scarce settings, this model has not yet to be implemented widely, due to the low responsiveness of health systems, as well as the poor practice of prevention against nutrition-related problems among the population 5 . This condition can be seen in several countries around the world, such as Denmark or Western Australia [5][6][7] .
In Vietnam, nutritional counseling clinics have been recently introduced in metropolitan areas, including Hanoi and Ho Chi Minh City. However, little is known about the profile and preference of the clients that attend these clinics. To inform policy development and support the expansion of this service, the present study was conducted to assess the preference and willingness of clients to pay for nutritional counseling services in an urban site in Hanoi.

Methods
Study setting and sampling method A cross-sectional study was conducted from March to April 2016 in an urban clinic in Hanoi Medical University, Hanoi, Vietnam. Eligibility criteria included 1) clients attending services in the Center of Preventive Medicine at Hanoi Medical University; and clients' parents or guardians (for those who were under 18 years old); 2) aged 18 years and above; 3) agreed to participate in this study and gave written informed consent; 4) able to answer a questionnaire (Supplementary File 1 and Supplementary File 2) for 15-20 minutes.
All eligible respondents from March to April 2016 were invited to participate in the study, resulting in a sample size of 429.

Measurements and instruments
Socio-demographic variables included age, gender, ethnicity, religion, educational attainment, marital status, current occupation, self-assessment of nutritional status and monthly household income (see Table 5 for detail).
Preference for nutritional counseling services included who would receive nutritional counseling, frequency of counseling services and communication methods for counseling.
Willingness to pay for nutritional counseling services were elicited using the bidding game technique, which consists of double-bounded dichotomous-choice questions combined with an open-ended question regarding two service packages: 1) feefor-service; and 2) one-year nutritional management package.
We selected 200,000 VND (~ US$ 9; 2017 exchange rate) and 3,000,000 VND (~ US$ 135; 2017 exchange rate) to be the initial prices for fee-for-service and one-year nutritional management package, respectively, based on the actual price of nutritional counseling services in this clinic. Each patient was asked a series of questions about their WTP at specific prices (see Figure 1 and Figure 2 for the bidding process). Firstly, the clients were asked if they were willing to pay the initial prices. Depending on the choice of either Yes or No, interviewers presented two other bids: the higher bid for respondents answering "Yes"; and the lower bid for respondents saying "No". The question was repeated until the last bid was equal to four times or one eighth of the initial prices. Finally, the respondents were asked an open-ended question "What is the maximum price you would be willing to pay for nutritional counseling services?"

Statistical analysis
Data was analyzed using STATA software version 12.0 (Stata Corp. LP, College Station, TX, USA). A p-value <0.05 was considered statistically significance. A stepwise logistic model with the threshold of p-value < 0.2 was used to identify associated factors with the WTP. Interval regression was used to measure the amount of WTP and identify associated factors.

Ethical approval
Proposal of this study was approved by the Ethical Committee of Hanoi Medical University. Subjects were introduced to the purpose of this study, and asked to give written informed consent if

Amendments from Version 1
We have added another co-author for this paper. Dr. Quynh Le has been involved in the conceptualisation and design of the study, supervising data collection and commenting on the manuscript. After our previous submission, Dr. Quynh Le raised comments and conducted further statistical analyses that confirm the validity of the models we presented in the study. We had some discussions and reached consensus on the results. All authors acknowledged her contributions and would suggest that Dr. Quynh Le should be included as a co-author of the manuscript.  they agreed to participate in the study. Respondents could withdraw anytime they want. Their information was ensured to be confidential.

Results
Demographic and socio-economic statuses of respondents are summarized in Table 1. Most of the clients were Kinh (97.7%), having above high school education (63.2%), single with no children (50.6-60.0%), and were in white collar employment (43.3%). Table 2 shows the willingness to use for nutritional counseling services of clients. Overall, 79.6% clients wanted to use counseling services. The major desire was that respondents' children would receive nutritional counseling (74.8%) monthly or more frequently (39.8%) via meeting physicians face-to-face (64.9%).
The WTP for one-time service is described in Table 3. Overall, a high amount of the respondents were willing to pay for nutritional counseling services (87.2%). The mean amount they were willing to pay was 96,100 VND per utilization (95% CI 81,000-111,000 VND), equivalent to US $4.3 in 2017, which varied across groups.
There was a significant difference in the WTP of the three age groups (p<0.05). Table 4 describes the WTP for the one-year nutrition management package. On average, respondents were willing to pay 946,400VND (95% CI 860,200 -1,032,700 VND) (~$41.9 -2017) for this package, which varied among groups (p<0.05).
Associated factors of the willingness to use and WTP for nutritional counseling services are shown in Table 5. The likelihood of using nutritional counseling services was higher among females, those seeking counseling for elderly people and those that preferred face-to-face counseling services. WTP for one-time service was 95,000 VND higher among clients aged over 35. Meanwhile, WTP for one-year nutritional management services was higher among those seeking services for the elderly people, those with a poor nutritional status and those that have under-6 year old children.

Discussion
Nutrition has been a pressing topic of many researchers 8 . There are several studies about nutritional counseling services for patients 9-11 or concerning a particular nutritional component 12,13 , but studies about general and preventive nutritional counseling are still limited 14 . Evidence provided by this study not only imparts information for future research, but also gives nutritional counseling providers a better perception to enhance their services.
In this urban setting, we found a high preference for nutritional counseling services for various target client groups, including elderly people and children. Clients also reported a high WTP for this service, which could be very helpful for expansion of the services. However, a combination of communication methods is needed; we found a higher preference for face-to-face counseling among respondents, knowing that many of them may also seek other health care services.
Overall, the preference for nutritional counseling in this study was quite high (79.6%). Most of the clients who did not have the need for this service were single with no children and self-evaluated their nutritional status as 'average'. The mean amount of WTP for one-time and one-years services was $4.3 and $41.9, accounting for 0.20% and 1.98% GDP per capita in Vietnam in 2015 ($2,111, enumerated by World Bank) 15 , which is an acceptable amount for clients to pay.
Associated factors of the preference and WTP for nutritional counseling services in our study were not in line with some predictions provided by a study in South Korea 16 . Our study showed   that older clients are more willing to pay for nutritional counseling than younger ones. Another noteworthy finding of this study is that clients with a higher educational level were not as willing to pay for the one-year management package as clients who only finished high school. This can be explained by the two occupations of respondents: those whose educational level were above high school were mainly white-collar workers, while almost everyone with lower educational levels were still high-school students or college students (83.8%). This may suggest that the recent nutritional education programs in Vietnam have caused a positive effect on students' attitude toward nutritionally related programs (http://dinhduonghocduong.net/) 17 .
Those who have under-6 year old children and assess their children's nutrition status poorly had a higher WTP for nutritional counseling services. These findings are well expected, thus enhance our study data's validity. We suspected that clients' income was associated with their WTP, as richer clients are more likely to pay a higher amount for nutritional counseling services. However, there was no significant relationship between clients' household income and the WTP for nutrition counseling services.
To elicit a clients' preference and WTP, we used the bidding game technique, as it was proved to be more reliable than open-ended questions or dichotomous-choice questions only 18,19 . However, one of the biggest drawbacks of this technique is that the risk of starting-point bias -the initial bid can have influence on clients' WTP 20 . The initial bids in this study were based on the actual prices for nutritional counseling services in this setting in order to minimize the occurrence of this bias. Additionally, our study may possibly be affected by other biases, such as observation bias, which occurs when the roles of respondents in their families can affect the amount of their WTP 21 . For example, we assumed that those who were the bread-winners in their families tended to have higher WTP for health-related services. Another example is that if information about nutritional counseling may not be sufficiently provided, this may result in lower preference and WTP for nutritional counseling services. To mitigate this bias, we selected highly-experience interviewers and trained them carefully with a standardized protocol for data collection.

Conclusions
The preference and willingness to pay for nutritional counseling services in urban Hanoi is relatively high. These findings may partly contribute to the implementation of maintaining nutritional counseling services Vietnam, thus actively preventing and controlling the spread of non-communicable diseases.

Data availability
Dataset 1: Raw data for Table 1-Table 5. doi, 10.5256/ f1000research.10974.d153260 22 Author contributions NBT, HTL, QNHL and BXT designed the study. All authors undertook data collection and analysis. HVN wrote the first draft. All authors approved the final version to be published.

Competing interests
No competing interests were disclosed.

Grant information
The author(s) declared that no grants were involved in supporting this work. 1.
in a range of settings. Knowing that clients are willing to pay for the service will contribute significantly to sustainability. In addition, some discussion or speculation on whether this would negate the contribution of universal medical insurance would contribute to the discussion.
While I know that the review is meant to focus on the scientific content there are some content issues that need to be addressed Abstract: I would be more general with the Dong described 96100 is too specific for the analysis performed and would be better cited as 96000 add in USD to the $ amount quoted.
page 3 Western Australia is not a country and so this sentence needs to be adjusted.

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Referee Expertise: nutrition and dietetics I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard. The first, the topic is relevant and interesting. However, the authors need to clarify some points for future replication by other.
Please confirm code of the dependent variable and independent variables in this study.I am not sure whether the dependent variable is Preference or WTP1 or WTP2.I try to replicate the result by using STATA program, but the result of the multiple logistic regressions is not the same those presenting in Table 5. 1.
using STATA program, but the result of the multiple logistic regressions is not the same those presenting in Table 5.
My analysis are: Table 1  Table 2  Table 3 Remarks on Table 5 mentions about the Vietnamese economic status using the monthly income quintile. It would help us gaining better understanding if the authors discuss how the amount of 2,000,000 VND for single service and 3,000,000 VHD for yearly services related to their income status. Furthermore, if it is calculated as % of total expense, we will have better understanding on whether it is considered as high cost, or catastrophic cost or not. However, for sure the expense for this counseling service is not too high. How could we guarantee that this could reflect the real perception of the value of counselling as priority and their willingness to pay when compare to other services such as treatment that have better short term output or outcomes.
The methodology of biding to get the WTP amount is very interesting that the participants choose the amount themselves and use their mean as WTP. Please discuss concerning the amount that finally identify as WTP is appropriate in the real context of their living standard.
Participants of this research are clients who are seeking services at the Center of Preventive Medicine. Researcher should explain whether the sample could represent to population, since they might have some conditions that have higher demand for the services, then the results might be over estimate.
Since this study is interesting: Is the nutrition counseling fee included in their service package (with all service) that they must pay. Or The nutrition counseling is separate service that they could select or not select?

If applicable, is the statistical analysis and its interpretation appropriate? Yes
Are all the source data underlying the results available to ensure full reproducibility? Yes

Are the conclusions drawn adequately supported by the results? Yes
No competing interests were disclosed.

Competing Interests:
Referee Expertise: Health policy, heath service system, epidemiology, spatial epidemiology I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
Author Response 01 Oct 2017 , Institute for Global Health Innovations, Duy Tan University, Vietnam Cuong Tat Nguyen Dear Prof. Wongsa Laohasiriwong, Thank you very much for your review. I will provide explanations for your questions.
1. Both Preference, WTP1, and WTP2 are dependent variables, as they represented clients' willingness to use the nutritional counseling service, willingness to pay for one-time service and for a one-year package.
2. We selected 200,000 VND and 3,000,000 VND to be the initial prices for the one-time service and one-year package, respectively, as they were the actual prices of nutritional counseling services in this clinic at that time. These prices were accounted for 0.43% and 6.4% of GDP per capita in Vietnam in 2015, yet still acceptable prices for Vietnamese clients to pay. As for the comparison between this service and treatment, it is quite a hard question, since comparing the outcomes of preventive medicine and curative medicine is not an easy task. This could be a good idea for further studies about nutritional counseling service's outcome in the future.
3. The mean amounts of WTP for both one-time service and one year package are suitable for the living standard of Hanoi people, as this research was carried out in an urban area of Hanoi, with the majority of participants were students and white collar workers. These prices were also relevant to other medical services' prices such as out-patient examination (~150,000 VND) or chest X-ray (~80,000 VND). 4. We also have taken into account the risk of overestimating the preference of clients for nutritional counseling service, but this was not significant, as the participants were clients who sought other medical services in this clinic, e.g vaccination. The nutritional counseling service was a separate service and was not included in their medical package This research study has been well documented and technically sound, with regard to statistical consideration. The analysis employs standard techniques and sample size has been reasonable (N=429). The report of results should be welcome as the need for healthcare and related data has become increasingly imperative.
The authors of the work have presented interesting findings, especially the significance of a small amount of spending, roughly $4 for one-time counseling service, and $42 for one-year package of services, both significant at conventional statistical levels. Apparently, there are rooms for the authors to report more insights since I can see the opportunity for controlling for some of the most influential socio-economic demographic factors, such as income, or physical state, such as BMI, where conditional probabilities computed from appropriate modeling efforts can potentially be insightful.
I was also a bit surprised with the fact that the rate of approval for use of ICT apps and devices has been quite low, and this reality will be in and of itself an issue worth exploring as they are critically important in today's age of information, and help reduce counseling costs.
All in all, this is a good and useful study, and I am happy to approve it for indexing by F1000Research.
No competing interests were disclosed.

Competing Interests:
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
The benefits of publishing with F1000Research: Your article is published within days, with no editorial bias You can publish traditional articles, null/negative results, case reports, data notes and more The peer review process is transparent and collaborative Your article is indexed in PubMed after passing peer review Dedicated customer support at every stage For pre-submission enquiries, contact research@f1000.com