This article was downloaded by: [University of Nebraska, Lincoln] On: 01 February 2015, At: 20:15 Publisher: Routledge Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Global Public Health: An International Journal for Research, Policy and Practice Publication details, including instructions for authors and subscription information: http://www.tandfonline.com/loi/rgph20

Validity and reliability of the Vietnamese Physician Professional Values Scale a

b

c

Nguyen Minh Sang , Alix Hall , Tran Thi Thanh Huong , Le Minh ad

Giang

c

& Nguyen Duc Hinh

a

Center for Research and Training on HIV/AIDS, Hanoi Medical University, Hanoi, Vietnam b

Click for updates

Priority Research Centre for Health Behaviour, Faculty of Health, The University of Newcastle & Hunter Medical Research Institute, Newcastle upon Tyne, Australia c

Department of Epidemiology, Hanoi Medical University, Hanoi, Vietnam d

Department of Ethics and Social Medicine, Hanoi Medical University, Hanoi, Vietnam Published online: 03 Dec 2014.

To cite this article: Nguyen Minh Sang, Alix Hall, Tran Thi Thanh Huong, Le Minh Giang & Nguyen Duc Hinh (2015) Validity and reliability of the Vietnamese Physician Professional Values Scale, Global Public Health: An International Journal for Research, Policy and Practice, 10:sup1, S131S148, DOI: 10.1080/17441692.2014.981830 To link to this article: http://dx.doi.org/10.1080/17441692.2014.981830

PLEASE SCROLL DOWN FOR ARTICLE Taylor & Francis makes every effort to ensure the accuracy of all the information (the “Content”) contained in the publications on our platform. However, Taylor & Francis, our agents, and our licensors make no representations or warranties whatsoever as to the accuracy, completeness, or suitability for any purpose of the Content. Any opinions and views expressed in this publication are the opinions and views of the authors, and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content should not be relied upon and should be independently verified with primary sources of information. Taylor and Francis shall not be liable for any losses, actions, claims,

proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or howsoever caused arising directly or indirectly in connection with, in relation to or arising out of the use of the Content.

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

This article may be used for research, teaching, and private study purposes. Any substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing, systematic supply, or distribution in any form to anyone is expressly forbidden. Terms & Conditions of access and use can be found at http://www.tandfonline.com/page/termsand-conditions

Global Public Health, 2015 Vol. 10, No. S1, S131–S148, http://dx.doi.org/10.1080/17441692.2014.981830

Validity and reliability of the Vietnamese Physician Professional Values Scale Nguyen Minh Sanga*, Alix Hallb, Tran Thi Thanh Huongc, Le Minh Gianga,d and Nguyen Duc Hinhc

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

a

Center for Research and Training on HIV/AIDS, Hanoi Medical University, Hanoi, Vietnam; Priority Research Centre for Health Behaviour, Faculty of Health, The University of Newcastle & Hunter Medical Research Institute, Newcastle upon Tyne, Australia; cDepartment of Epidemiology, Hanoi Medical University, Hanoi, Vietnam; dDepartment of Ethics and Social Medicine, Hanoi Medical University, Hanoi, Vietnam b

(Received 16 August 2013; accepted 26 September 2014) Physician values influence a physician’s clinical practice and level of medical professionalism. Currently, there is no psychometrically valid scale to assess physician values in Vietnam. This study assessed the initial validity and reliability of the Vietnamese Physician Professional Values Scale (VPPVS). Hartung’s original Physician Values in Practice Scale (PVIPS) was translated from English into Vietnamese and adapted to reflect the cultural values of Vietnamese physicians. A sample of clinical experts reviewed the VPPVS to ensure face and content validity of the scale, resulting in a draft 37-item measure. A cross-sectional survey of 1086 physicians from Hanoi, Hue and Ho Chi Minh City completed a self-report survey, which included the draft of the VPPVS. Exploratory Factor Analysis was used to assess construct validity, resulting in 35 items assessing physician’s professional values across five main factors: lifestyle, professionalism, prestige, management and finance. The final five-factor scale illustrated acceptable internal consistency, with Cronbach’s alpha coefficients ranging from 0.73 to 0.86 and all item-total correlations >0.2. Limited floor or ceiling effects were found. This study supports the application of the VPPVS to measure medical professional values of Vietnamese physicians. Future studies should further assess the psychometric properties of the VPPVS using large samples. Keywords: physician values; medical professionalism; validity; reliability; Vietnam

Introduction Over the last century health care systems worldwide have transformed from a servicefocused profession to a commercialised industry (Hafferty, 2006), and Vietnam is no exception. In 1986 the Vietnamese Government introduced the economic reform known as Doi Moi, which transformed the Vietnamese economy from a socialist, predominantly subsidised economy to a market economy (Beresford, 2008; Ladinsky, Nguyen, & Volk, 2000; Segall et al., 2002; Witter, 1996). Consequently, the Vietnamese health care system has gone from a government-financed service to being predominately funded by private and for-profit health care services (London, 2008). In addition, medical professional

*Corresponding author. Email: [email protected] © 2014 Taylor & Francis

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

S132

S.M. Nguyen et al.

salaries are low with many physicians having to rely on a fee-for-service income rather than a salary income (Sepehri, Chernomas, & Akram-Lodhi, 2005). The commercialisation of the Vietnamese health care system has also impacted heavily on patients. For instance, heath care service outcomes have been shown to reflect more favourably for financially wealthy patients compared to the financially disadvantaged (Sepehri et al., 2005). Medical professionalism is seen as a highly influential trait of health care providers. It is described as ‘a set of values, attitudes and behaviours that results in serving the interest of patients and society before one’s own’ (Diaz & Stamp, 2004; Reynolds, 1994). However, the commercialisation of health care systems threatens the underlying core physician values that define medical professionalism. For instance, there is now a greater emphasis on the professional and personal development of physicians (Hatem, 2003), and an increased importance on financial reward and monetary incentives (Reynolds, 1994; Swick, Bryan, & Longo, 2006). Most of these changes have been seen as undermining the very essence of medical professional values (Hafferty, 2006; Relman, 1998; Swick et al., 2006), creating a fear that physicians will no longer place patient and society needs above their own (Hilton & Slotnick, 2005). For example, physician prescribing habits and professional behaviour have been found to be influenced by physician’s interactions with pharmaceutical companies (Chew et al., 2000; Lexchin, 1993; Orlowski & Wateska, 1992; Wazana, 2000). To combat this threat a greater emphasis is being placed on teaching and evaluating medical professionalism (Hatem, 2003; West & Shanafelt, 2007). For instance, medical professionalism is now considered by some medical systems as a core component of medical competence (Hatem, 2003). Although medical professionalism and physician professional value are relatively new concepts in Vietnam, they are also increasing in importance. In 2007 medical professionalism and ethics were introduced into the Vietnamese medical curriculum. If we are to effectively influence medical professionalism in this expanding commercial world, it is necessary that we accurately understand the current state of physician professional values. Consequently, there has been a call for valid and reliable tools that assess medical professionalism (Epstein & Hundert, 2002). To assess medical professional values, Hartung, Taber, and Richard (2005) developed the Physicians Values in Practice Scale (PVIPS). The 38-item PVIPS is relatively succinct (Hartung et al., 2005) and illustrates strong psychometric properties (Hartung et al., 2005). However, there are no psychometrically rigorous assessment tools to assess the professional values of Vietnamese medical physicians. Vietnamese medical practices have been heavily influenced by both Western and Eastern cultures (Chu, 2008; Ladinsky, Volk, & Robinson, 1987; Thompson, 2003; Wahlberg, 2006). Consequently, the professional values of Vietnamese physicians may differ to those of Western medical professionals. To understand and promote medical professionalism in Vietnam it is necessary that we develop psychometrically valid measurement tools that accurately assess aspects of medical professionalism, such as physician’s professional values. Data from such measures can be used to inform medical education and curriculum, and to identify areas in need of improvement in relation to medical professionalism in Vietnam (Lynch, Surdyk, & Eiser, 2004). The aim of this study was to develop and assess the face, content and construct validity; internal reliability; and floor and ceiling effects of the Vietnamese Physician Professional Values Scale (VPPVS).

Global Public Health

S133

Methods Study setting This was a cross-sectional study undertaken in three Vietnamese cities: Hanoi, Hue and Ho Chi Minh (HCM) City.

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

Sample size The study sample was stratified by city and public or private health care sectors, in which the population size (number of doctors) of Hanoi, Hue and HCM City is 4435, 664 and 5167 (in percentage 43%, 7% and 50%), respectively. The WHO sample size calculation was applied to calculate a stratified sampling. As this is also a finite population, sample size determination using the finite population correction factor (with design effect =2) was applied. The estimated sample size was 946. However, the actual sample size was 1086 physicians who completed a survey. The percentage of doctors merely working in private health care sector ranged from 10% to 15% for each city according to Department of Health report in 2010. Sampling To select physicians for the study, we first selected hospitals where physicians were invited to participate in our study. A randomly selected list of hospitals from the level of public and private sectors was developed (based on the data provided by the Department of Health in 2010) in order to accurately reflect the various types of hospitals presenting in the three cities. Hospitals that are a part of other ministries were not included in this study (e.g., Ministry of Agriculture and Rural Development, Ministry of Public Security, etc.). The number of hospitals at each level was calculated so that we obtained enough participants for the study. From the public health care sector, first we selected central hospitals. We randomly selected five in total of 18 (Hanoi), three in total of three (HCM) and only one central hospital (Hue). Second, we selected provincial hospitals. We randomly selected 5 out of 33 (Hanoi), 4 out of 18 (Hue) and 5 out of 31 (HCM). Last, we selected district hospitals/ health centres. Due to a considerable number of district hospitals/health centres, the number of selected district hospitals/health centres was calculated so that we obtained enough physicians working in public health care sector for the study. We estimated to have approximately four district hospitals/health centres in each participating city. From the private health care sector, 4 hospitals out of 25 private hospitals were selected in Hanoi and 4 out of a total of 11 private hospitals in HCM were selected, including both specialty and general hospitals. There were no private hospitals in Hue. The number of private clinics was calculated so that we obtained enough physicians merely working in private health care sector for the study. Private clinics with less than five doctors working were excluded. We estimated to have approximately 6 out of 43 private clinics (Hanoi); 3 out of 12 (Hue) and 5 out of 35 (HCM). Participants Physicians working in health facilities that were listed in the randomly selected list of hospitals (including: tertiary central hospitals, provincial hospitals, district hospitals, community health centres or private health facilities) were invited to take part in this study. Physicians were eligible to take part if they were currently practicing clinical

S134

S.M. Nguyen et al.

Table 1. The number of hospitals and doctors recruited.

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

Public health care sector Number of doctors

City

Number of hospitals

Hanoi

5 Central hospitals

202

5 Provincial hospitals

120

3 District hospitals/ health centre Total 1 Central hospitals 4 Provincial hospitals

27

Hue

HCM City

349 56 83

4 District hospitals/ health centre Total 3 Central hospitals

181 81

5 Provincial hospitals

269

4 District hospitals/ health centre Total

Private health care sector Number of hospitals 2 Specialty hospitals 2 General hospitals 6 Private clinics

Number of doctors 15

10/29 districts

24 23 62

3 Private clinics

Number of districtsa

411 doctors 5/9 districts

19

42

47 397

2 Specialty hospitals 2 General hospitals 5 Private clinics

19 15

200 doctors 10/24 districts

25 38 78

475 doctors

Note: Private clinics which had less than five doctors working and hospitals belonging to other ministries were not included. a Number of districts indicated refers to the location of hospitals including central districts and suburban districts.

medicine. Participation was voluntary. Table 1 shows the number of hospitals and doctors recruited from the three cities. In the public health care sector, we selected 9 central hospitals, 14 provincial and 11 district hospitals/health centres across the three cities. From the private health care sector, we selected 4 specialty private hospitals, 4 general private hospitals and 14 private clinics, each of which had a number greater than or equal to 5 doctors working. The selected hospitals/clinics were located in 10 districts (Hanoi), 5 districts (Hue) and 10 districts (HCM) with the total of 29, 10 and 24 districts of the three cities, respectively.

Question design The VPPVS was developed based on the original 38 scored items from Hartung’s (Hartung et al., 2005) PVIPS. Hartung’s (Hartung et al., 2005) original PVIPS was developed to assess the professional values of medical students with a purpose of assisting medical students in determining a relevant specialty (Hartung et al., 2005). The 38 scored items represent six domains: Prestige (10 items), service (9 items), autonomy (7 items), lifestyle (4 items), management (5 items) and scholarly pursuit (3 items). Each item is a statement relating to one of the six domains of professional values. Respondents indicate the extent to which each item is important to their medical practice, using a 5-point Likert scale (Hartung et al., 2005). Scores range from 1 (strongly disagree) to

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

Global Public Health

S135

5 (strongly agree). The PVIPS has evidence of reliability and validity in samples of medical students in the USA and Australia (Hartung et al., 2005; Rogers, Creed, Searle, & Hartung, 2010). The original PVIPS was developed and psychometrically evaluated in medical students in the USA, whereas comparatively for this study the target population was Vietnamese physicians. Due to the differences in the target population, languages, cultures and countries, between the original PVIPS and the VPPVS it is necessary that, for the PVIPS to be a valid and reliable measure of the professional values of Vietnamese physicians, the scale must undergo an extensive translation and cultural adaptation (Guillemin, Bombardier, & Beaton, 1993). The research team translated Hartung’s (Hartung et al., 2005) original PVIPS from English into Vietnamese following Beaton’s Cross-cultural Adaptation Process and Bulligen’s recommendations for transforming psychometric scales (Beaton, Bombardier, Guillemin, & Ferraz, 2000). Following the translation of the scale the researchers then reviewed the content of each item to ensure they were relevant and appropriate to assess Vietnamese physician’s values for the Vietnamese culture. Several minor changes were made to Hartung’s (Hartung et al., 2005) original items with some being reworded and others being removed. A convenience sample of medical specialists from Hanoi was provided with a copy of the first draft of the VPPVS and asked to reflect on the appropriateness of this scale to the Vietnamese context. Five discussion groups were held between the medical experts and project research staff. Between 8 and 10 doctors attended each discussion group, representing the specialist areas of paediatrics, dentistry, traditional medicine and obstetrics and gynaecology. During these sessions the medical experts provided feedback on the appropriateness of each item of the VPPVS and provided recommended changes to the scale. Based on the results of this process several items were reworded, several items were removed (e.g. ‘Involve myself in the lives of my patients’, ‘Speak to local organizations such as educational and religious groups’) and several new items specific to the Vietnamese culture were included to the scale (e.g. ‘Be recognized as a virtuous doctor’, ‘Become an important person that members of family, relatives and friends can count on when they have problems related with their health’). This process resulted in 37 items in the first draft version of the VPPVS. The scores ranged from one to five, with one indicating strongly disagree, three indicating neutral and five indicating strongly agree. The drafted version was then reviewed by a group of sociologists, anthropologists and medical experts who again assessed the cultural relevance of the scale to the Vietnamese context. The draft VPPVS was then refined and pilot tested with 30 physicians, who reviewed and provided final recommendations on the wording and cultural appropriateness of the scale. Following this process slight wording changes were made to finalise the expression of each item and ensure all items were easily understood by Vietnamese physicians. It must be noted that a backwards translation was not performed in the development of this scale, due to limitations in the time, human resources and funding available for this study. In addition to the VPPVS, seven questions were included in the survey to assess participant’s socio-demographic characteristics, including, city, gender, age, years of clinical practice, specialty, level of education and level of income (Appendix). Data collection The head of each randomly selected hospital was contacted and asked to invite clinical physicians working within their hospital to take part in this study. Physicians who were

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

S136

S.M. Nguyen et al.

invited by the head of the hospital to participate in the study attended a meeting conducted by members of the research team. During these meetings physicians were provided with a cover letter, consent form and self-report survey. The information contained in these documents included a description of the goals and procedures of the study. Physicians were informed that the participation in the study was completely voluntary, and they would not receive any direct benefits from taking part in this study. Physicians were also informed that no names or any other identifiable data would be collected in this study and that participants’ answers to the survey would not be disclosed to a third party or would not be used to influence their professional work at the hospital. Consenting doctors completed the self-reported pen and paper survey during the meeting. This study received ethical approval from the Human Research Ethics Committee of Hanoi Medical University. All study procedures were conducted in line with the ethical principles outlined by the Human Research Ethics Committee.

Statistical analysis All analysis was conducted using STATA Version 11.1 statistical software. It is recommended (Beaton et al., 2000) that following the translation and adaptation of a psychometric measures the new version of the measure should undergo a rigorous psychometric evaluation to ensure the new scale retains adequate psychometric properties (Beaton et al., 2000). Following this recommendation we undertook a rigorous psychometric evaluation of the VPPVS. Construct validity The distribution, frequency and percentage of missing values were calculated for each of the 37 items of the VPPVS. Due to the differences in the target population and differences in the items of the original PVIPS and VPPVS, an explanatory factor analysis was used to assess the construct validity of the VPPVS instead of a confirmatory factor analysis. The principal factors method of exploratory factor analysis was employed to assess the construct validity of the VPPVS (Fabrigar, Wegener, MacCallum, & Strahan, 1999). Multiple factor analyses were performed and compared. The number of factors to rotate for each factor analysis was determined by the results of the following tests: the Eigen value more than one rule; break point in the scree plot and parallel analysis (Costello & Jason, 2005; Fabrigar et al., 1999).The most robust factor structure was determined if it: contained at least three items per factor; all items obtained factor loadings above 0.3 and was considered conceptually relevant (Worthington & Whittaker, 2006). As several of the original PVIPS factors were significantly correlated (Hartung et al., 2005) the promax quadratic rotation was used to simplify the structure, to allow for correlations between the final factors (Costello & Jason, 2005; Fabrigar et al., 1999). Any items that failed to obtain a factor loading of 0.3 or more were removed from the scale (Field, 2013). Observations with any missing data were removed from the factor analysis using listwise deletion. The correlation between each of the final factors was calculated. Internal reliability Cronbach’s alpha and corrected item-total correlations were calculated for all final factors. Factors with a Cronbach’s alpha coefficient between 0.7 and 0.95 (Terwee et al., 2007) and item-total correlations above 0.2 (Field, 2013; Pearce, Sanson-Fisher, &

Global Public Health

S137

Campbell, 2008; Scientific Advisory Committee of the Medical Outcomes Trust, 2002) were considered internally reliable. Floor and ceiling effects Domain scores were calculated by summing all items in the domain and dividing by the number of non-missing items in the domain. Domains with more than 15% of participants obtaining the lowest or highest score were considered to portray evidence of a floor and/ or ceiling effect (Terwee et al., 2007).

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

Differences in domain scores by physician’s socio-demographic characteristics Differences in the domain scores by physician’s sex, years of practice, medical speciality and level of income were assessed using Mann–Whitney test for categorical variables with two groups or the Kruskal–Wallis test for categorical variable with more than two groups. Non-parametric analyses were used due to the skewed distribution of the domain scores.

Results Participants A total of 1086 physicians completed a survey. Physician demographics are described in Table 2. Over a third of physicians were from HCM City (43.7%), male (56.4%), specialised in internal medicine (41.8%) and had less than 10 years of clinical experience (46.1%). The average age of physicians was 41 years old (Table 2). Construct validity Missing values ranged from 0.3% to 1.6%. All items were skewed; however no item was answered by more than 90% of participants using only one response option. For several items, the full range of response options was not used. Eight factor analyses were performed and compared, including a 10-, 9-, 8-, 7-, 6-, 5-, 4- and 3-factor models. The five-factor model was deemed the most suitable model with all items illustrating a factor loading of ≥0.3 on at least one factor, all factors contained three or more items and items illustrated few cross-loadings. In addition, the five-factor structure was deemed most conceptually appropriate. The final five-factor structure contained 35 items. The factor loadings of the five-factor structure are described in Table 3. Factor 1 contains 11 items and was labelled lifestyle with items loading highest on this factor reflecting values relating to a comfortable, predictable and autonomous work life. Seven items loaded highest on factor 2, which was labelled prestige. Items within factor 2 reflect a clinician’s desire to be recognised as a leading and influential figure within the community. Factor 3 was labelled professionalism with eight items depicting values concerning a clinician’s desire to work hard for the better of others and contribute to their community. Factor 4 contains six items relating to management and administrative values and was thus labelled management. Finally factor 5 was labelled financial, with three items assessing a desire to maximise one’s financial earnings from being a clinician. The correlation coefficients between the final five factors are shown in Table 4. A moderate positive correlation (≥0.4) was found between the following domains: prestige and lifestyle; professionalism and lifestyle; professionalism and prestige; and management and prestige.

S138

S.M. Nguyen et al.

Table 2. Participant socio-demographic characteristics. Characteristics

N = 1086

Downloaded by [University of Nebraska, Lincoln] at 20:15 01 February 2015

Age Mean ± SD Min–Max City Hanoi Hue HCM City Gender Men Women Years of practice Less than 10 years 10–20 years 20–30 years More than 30 years Medical specialty Internal medicine Surgery/obstetric Paediatric Dentistry Oncology Othera Level of income per month Under 5 million VND (under $US250) 5–10 million VND ($US250–$US500) 10–20 million VND ($US500–$US1000) More than 20 million VND ($US1000)

41 ± 11 24–77 %

N

37.9 18.4 43.7

411 200 475

56.4 43.6

612 474

46.1 28.3 20.4 5.2

501 307 222 56

41.8 20.6 12.9 5.6 5.5 13.6

454 224 140 61 60 147

35.7 50.7 9.3 4.3

388 551 101 46

a

Emergency, radiology, psychiatry, ophthalmology, otorhinolaryngology and traditional medicine.

Internal reliability All corrected item-total correlations were above 0.2, ranging from 0.37 to 0.68. All Cronbach’s alpha coefficients were between 0.70 and 0.95 (Table 5).

Floor and ceiling effects As shown in Table 5, none of the final five factors exhibited substantial floor or ceiling effects, with

Validity and reliability of the Vietnamese Physician Professional Values Scale.

Physician values influence a physician's clinical practice and level of medical professionalism. Currently, there is no psychometrically valid scale t...
199KB Sizes 0 Downloads 6 Views