Determinants of Patient Satisfaction in The United Arab Emirates Health Care Industry

Background: The United Arab Emirates is aiming to provide 'world-class healthcare' to its citizens. This study aims to determine the association between overall satisfaction and various aspects of health care services in the United Arab Emirates on a country level. Methods: The data was collected from 5,855 respondents on the affordability, quality, accessibility, and responsiveness dimensions of health care. Besides frequency tables and descriptive statistics, statistical methods, such as Principal Component Analysis and Multiple Regression, were used to reduce and model the association between dependent (overall satisfaction) and independent (affordability, quality, accessibility, and responsiveness) variables. Results: The ndings indicated higher satisfaction levels with quality and accessibility and lower satisfaction levels with affordability and responsiveness. Recommendations are provided to address the latter two factors. Conclusion: The study identies issues in health care service provision in the United Arab Emirates and offers recommendations about enhancing affordability and responsiveness.

patients' perceptions of various dimensions such as affordability, quality of delivery, accessibility, and responsiveness are evaluated. This paper adds to the increasing body of literature available on health care service delivery around the world. It focuses on an oil-exporting country striving to decrease its dependence on oil and diversify its economy. Moreover, the country's health care system is developing and is evolving rapidly to serve the population better despite the current health crisis, the COVID-19 pandemic worldwide.
This study aims to answer the following research question: To what extent do quality, affordability, accessibility, and responsiveness determine overall satisfaction in the UAE health care sector?
Health Care in the United Arab Emirates The UAE, health care sector, is categorized by public and private service providers' corresponding service utilization. Despite the prominent presence of private health care providers in the country, the general public expenditure as a percentage of the overall health care expenditure is relatively high. Between 2000 and 2015, the proportion of public spending on health care has been approximately 67 percent on average of the total expenditure [23].
The regulatory system in the health care industry in the UAE is multi-tiered and includes federal and emirate level regulatory bodies [1]. The Federal Insurance Authority also plays a crucial role in the UAE health care sector, as the UAE government is transitioning towards universal health coverage for all residents in the UAE [24]. The UAE health care industry is dominated by a handful of private health care providers. During the last few years, the sector's mergers and acquisitions have further consolidated the industry [21].
International accreditation is prevalent and encouraged by the UAE regulatory authorities across the health care sector. According to reports by the Ministry of Health and Prevention, over 70 percent of health care institutions and facilities have been accredited by international bodies. In contrast, the other facilities are accredited by 2021, according to reports by the Ministry of Health and Prevention [25,48]. Despite international accreditation of health care facilities, the standardization measures of quality lack throughout local private and public health care providers.
The importance of the insurance industry is paramount universally. The UAE's insurance industry is fragmented, with 61 insurance companies [26] serving a population of approximately nine million people (Federal Competitiveness and Statistics Authority, 2018). The top ve insurance companies have a collective market share of 56.2 percent [27].
There has been signi cant development across the UAE's healthcare system over the last two decades, including improvements in accessibility, affordability, and service delivery of clinical quality. With such developments, the UAE government continuously nanced the health industry over the years, ensuring health care delivery to the population, even though falling oil revenues. One of the dilemmas is to stabilize the healthcare-related costs while extending accessibility and sustaining the quality of health care and affordability.
The National Agenda 2021 for UAE involves several indicators: health care, education, economy, justice, police and security, society, housing, infrastructure, and government services [28]. Under the realm of health care, the UAE government plans to cooperate with all strategic stakeholders in the health care sector and other sectors to provide health care according to clear national and international quality standards and seeks to focus on preventive medicine and to reduce the prevalence of non-communicable diseases in the country [28]. In line with these objectives, AED 4.84 billion (6.89 percent of the annual budget 2020) has been allocated to health care and social services programs [29]. The actions aim to improve both clinical and service quality in the UAE health care sector. This paper will primarily focus on evaluating the perceptions regarding various dimensions of service quality. The results will facilitate an understanding of challenges in health care services in the UAE.

Literature Review
The Iron Triangle of Health Care developed by Kissick in 1994 identi ed three critical elements for patient satisfaction: Quality, Cost, and Accessibility. The model is based on the principle of opportunity cost and trade-off, where all three elements are interdependent, and the achievement of two factors will only occur at the expense of the third [19].
Extensive literature is available on quality in health care, and several de nitions of quality have been proposed. According to Mosadeghrad [20] quality in health care can be excellence, value, conform to standards and guidelines, and meet customer needs and expectations. Compliance with these dimensions of quality would result in satis ed end-users or ultimate bene ciaries of the health care service. A patient-centered approach in health care provision dictates "... [quality is that which] exceeds patient expectations and achieves the highest possible clinical outcomes with the resources available" [2]. This encompasses the health care service's clinical and service quality dimensions, encircling care, and service aspects. Donabedian in 1980 [30] referred to these aspects as technical quality and interpersonal quality. Donabedian highlighted that a distinction is drawn between care in health care provision, which is related to the direct technical intervention (i.e., treatment, medication, check-ups, etc.) and service, which is concerned with the patients' experience of their interaction with the health care provider [30]. However, both care and service are interconnected and inseparable and are essential to the overall quality of the patients' experience. While patients might not necessarily be able to assess the technical elements of care, they can evaluate their service quality experience. Cohen highlighted that patients tend to give more weight to the interpersonal aspects of their experience with the health care provider [31].
Several factors contribute to the importance of patient satisfaction as a tool to help health care providers understand patients' perspectives and enhance overall quality. First, patient satisfaction and patient loyalty were found to be correlated. For example, in Yemen, a study found that patients' satisfaction with reliability, empathy, and assurance had a signi cant in uence on patient loyalty [32]. Similarly, Mortazavi et al. [33] concluded that there is a signi cant correlation between patient satisfaction and patient loyalty in nursing care, operating rooms, admission, and administration.
Second, patients can be viewed as customers from a consumerist perspective, especially in private healthcare providers' competitive market. Therefore, when health care delivery is commodi ed, patient (or consumer), satisfaction is crucial to organizational pro tability. Patients are empowered to leave the consumption loop due to dissatisfaction and nd other options [34]. While there is a direct link between organizational pro tability and patient satisfaction [35], the latter can reduce costs associated with resolving customer complaints [36].
Lastly, various aspects of patient satisfaction contribute to policy and organizational reform [37]. Although patient satisfaction surveys are not widely utilized in health care service assessment [38], they can identify performance gaps and indicate improvement areas.
Some health care professionals might dismiss patient satisfaction as being too subjective to determine health care quality [38]; however, patient-centered approaches to health care emphasize the importance of satisfying patients' expectations. In this sense, patient-satisfaction is an evaluative process in which a patient cognitively and emotionally reacts to the health care they receive [39]. Therefore, it is integral to include recipients' perceptions of service quality in the evaluation of overall satisfaction.
Another important aspect of health care is cost; synonymously used with the term affordability. As a generic term, affordability can be best de ned as a measure of someone's purchasing power towards a good or service [40]. According to Glickman [41], "affordability is not a synonym for low prices. It describes a qualitative ability to pay -an interaction of price, disposable income, and judgments about the necessity of a particular good." Accordingly, affordability of health care re ects the end user's purchasing power, derived from his disposable income, to pay for health care services.
To personalize the actual cost and affordability of health care, Emanuel et al. [42] proposed developing an "Affordability Index," [2] which relates the average cost of health care to the average household income. The authors themselves indicated that the index is not perfect in determining the affordability of health care services. Glickman [41] highlighted several limitations of the index, such as regional and national variability in health care spending, the cost of insurance subsidized insurance for low-income families, and double counting employer contributions to premiums. The index also has a limitation concerning its universal applicability as many countries do not have universal or sponsored insurance.
Accessibility to health care is a complex and multidimensional construct. Levesque et al. highlighted that the literature on health care de nes and operationalizes accessibility concerning nancial physical and geographical accessibility, predisposing and enabling factors, availability, accommodation, affordability, acceptability, adequacy, and cultural acceptability [40]. Gulliford et al. broke-down access into "having" su cient health care providers or, in other terms, availability and gaining access to such providers, which depends on actual utilization and affordability [43].
One aspect of accessibility is the availability of health care services when required, which can be operationalized, for example, as the number of patient-physician contacts [44]. Besides, availability can be determined by the number of hospitals and clinics, the number of beds in each institution, and physicians and nurses [45]. Mosadeghrad [20] argued that accessibility (physical, nancial, and conceptual) is crucial, while availability is essential. The author highlighted that accessibility is especially critical when the service costs are high, and insurance is unavailable. Gulliford et al. argued that from a policy perspective, facilitating access to health care refers to helping people to utilize appropriate health resources to preserve or improve their health [43].
Besides the factors discussed above, Mosadeghrad [20] found that responsiveness in health care service provision is also relevant, mainly because information asymmetry exists between patients and service providers. This quality aspect may include active listening, trust, respect, confidentiality, courtesy, and effective communication, referred to as interpersonal aspects [31]. Lack of perceived responsiveness on the part of the service provider may lead to lower levels of satisfaction among healthcare services [46].
Besides the factors discussed above, the literature discusses many other factors that determine perceived service quality in health care provision [20]. It is evident that various aspects of health care delivery are highly interdependent, and their operationalization varies signi cantly in the literature [40]. This paper primarily focuses on three factors (cost, quality, accessibility), as suggested by Kissick [19], and responsiveness in the UAE health care sector. [2] Affordability Index is a ratio that relates health insurance costs to household incomes over time. It is calculated by dividing the mean cost of an employer-sponsored family health insurance policy by median household income.

Method
The literature a rms that overall satisfaction with health care services is associated with various dimensions, such as quality, cost, accessibility, approachability, and responsiveness, to name a few. This study intends to investigate whether or not these associations are relevant in the context of the UAE. Based on the research question, the study hypothesizes that there is no association between quality, affordability, accessibility, responsiveness, and overall satisfaction in the UAE health care sector.
Given the study's nature, quantitative analysis is most appropriate for data collection and analysis. Quantitative research methods allow broad data collection and analysis across many respondents [47,48].
The target population for the research study consisted of citizens and residents of the UAE. [1] The study data was collected by YouGov and the Mohammed Bin Rashid School of Government in Dubai. A convenience sampling method was used to collect data. While the use of a convenience sample may impact how representative a sample is, a larger sample size may mitigate this limitation to some extent.
The appropriate sample size for a population of 9,121,170 was calculated through the Raosoft Sample Size calculator as an estimate [49,50]. A sample size of n=3,382 was found to be suitable; however, the actual sample size exceeded the appropriate sample size by approximately 65 percent, and n was equal to 5,855.
The respondents were asked to provide socio-demographic information, which is summarized in Table 1. They were also asked to indicate whether or not they have a regular general physician, whether they have visited a doctor in the last three years, and lastly, whether they intend to go back to their home countries for medical consultation.
The survey items for the health services factors were adopted from the survey developed by Marshall and Hays [51] with the authors' permission. The use of SERVQUAL [52] survey is standard in the health care literature. However, this study opted for the above instrument as it was deemed appropriate for exploratory research across the UAE's health care sector and was not targeted to a particular service provider. While eighteen questions were included in the survey, nine a rmative questions were used for data analysis. The survey was produced in English and was then translated into Arabic. These questions were classi ed under overall satisfaction, quality, affordability, accessibility, and responsiveness. The responses to all the questions were recorded on a 5-point Likert scale from 1 (strongly disagree) to 5 (strongly agree).
The English and Arabic surveys were disseminated online. The survey instrument fully complied with the ethics guidelines. All appropriate procedures related to informed consent, con dentiality, anonymity, participants' rights, and participation conditions, including the right to refuse or withdraw without penalty, were observed for the study.
A multi-step data analysis was employed to draw a conclusion about the sample and infer a conclusion about the population. In the rst step, preliminary studies such as frequency tables, descriptive statistics, and correlation estimates were produced. The correlations were used to determine the need for principal component analysis (PCA) with varimax rotation. In the last step of the analysis, the components extracted through PCA were used in a multiple regression model to estimate the association between overall satisfaction and various dimensions of service quality in the UAE health care industry.
Principle Component Analysis PCA is used for dimension reduction in the presence of a large number of variables and multicollinearity. However, another statistical method, such as regression or structural equation, needs to be subsequently used to estimate dependence between the derived and dependent variables.
To ensure the parsimony of the multiple regression model, PCA was employed for dimension reduction. The need for PCA was determined by estimating correlations between the variables included in the data collection instrument (refer to Table 4).
Field [53] indicated that the sample size is of primary consideration when using any form of factor analysis. Since the sample size in this study is substantially large relative to the number of variables involved, the use of PCA is deemed appropriate. The appropriateness of the sample size is determined by the Kaiser-Meyer-Olkin Measure of Sampling Adequacy (KMO). Kaiser [54] indicated that KMO values ranging between 0.7 and 0.9 are considered to be good. In addition to the sample size, Bartlett's test determined the strength of the relationship between variables. A p-value less than the signi cance level (α = 0.05) associated with Bartlett's test statistic is indicative of the appropriateness of using factor analysis.
The appropriate number of factors extracted is determined by 'eigenvalues' and 'scree plots.' The factors having eigenvalues of one or greater than one generally is retained. All such elements explain the most variance and are suitable for statistical methods determining the association between the dependent variable and independent factors.

Multiple Regression Analysis
Regression analysis enables the study to nd an association between one dependent variable and one or more independent variables. The regression equation requires the variables to be numeric. This study uses one dependent variable: the respondents' overall satisfaction and four independent variables that resulted from the PCA. The independent variables are classi ed under cost, quality, accessibility, and responsiveness.

The resulting regression equation can be written as
Overall Satisfaction = β 0 + β 1 (Affordability) + β 2 (Quality) + β 3 (Accessibility) + β 4 (Responsiveness) + ε Where β 0 is the constant term β 1 to β 4 are the coe cients of the factor's affordability, quality, accessibility, and responsiveness, and ε is the error term The coe cients' magnitude will ascertain the importance of each factor in determining the respondents' overall satisfaction concerning various dimensions of health care services in the UAE. Table 1 presents the pro le of the respondents. The male respondents comprised 59 percent of the sample. The sample appears to represent the population concerning age, population distribution across the seven emirates, and population composition (Emiratis and non-Emiratis) [55,56].

Results
For education and income, 53 percent of the participants indicated having a Bachelor's degree, approximately 34 percent reported having some form of postgraduate education, and 78 percent of the respondents speci ed having a monthly income of AED 25,000 or less. Nearly 50 percent of the respondents indicated having a sponsored insurance plan. Approximately 66 percent of the sample does not have a family physician, and 20 percent reported not having a medical check-up in the last three years. Moreover, 51 percent of the respondents intend to go back to their home countries for medical treatment if the need arises.
The frequencies of responses to various questions are presented in Table 2. Two factors stand out clearly -responsiveness and affordability. While all the other factors recorded a higher number of responses on the right side of the spectrum, that is, agreed and strongly agreed, the responsiveness and cost factors have a high percentage of neutral responses, 42 percent, and 36 percent, respectively. This neutrality, while ambiguous, may be interpreted as dissatisfaction with health care services on these dimensions. This interpretation also strengthens due to low means and medians associated with responsiveness (mean 3.19; median 3) and affordability (mean 3.31; median 3). All other factors recorded a mean and median of 4, indicating relatively higher satisfaction levels with them.
The correlations between various factors were statistically signi cant (α = 0.01). The correlation estimates were relatively high (refer to Table 4), suggesting interdependence between various variables. Accordingly, PCA was employed for dimension reduction.
The KMO of 0.88 indicated the suitability of employed PCA for dimension reduction purposes. Accordingly, three factors were identi ed a priori, and these three factors (quality, accessibility, and responsiveness) explained nearly 71 percent of cumulative variance. The affordability factor was excluded from the PCA as it had only one question, and it was then included as a separate independent variable in the multiple regression equation.
As presented in Table 5, the quality of health care service delivery in the UAE has the highest Eigenvalue (>3) and explains nearly 50 percent of the total variance. The other two factors are also important; however, the Eigenvalues are less than one. The percentage of variance explained by them is approximately one-fth of the quality aspect and ranges between 10-11%.
In the last step of data analysis, four aspects (quality, accessibility, responsiveness, and affordability) were included in multiple regression analysis. The estimated model was statistically signi cant (F-statistic 1764.265; p-value 0.000), with an R 2 of 0.547 and an adjusted R 2 of 0.546. The R 2 and adjusted R 2 values indicate that nearly 55 percent of the perceived overall satisfaction variability can be explained by the four factors included in the model as independent variables. There is a negligible difference between R 2 and adjusted R 2 values suggesting that the speci cation of the model is appropriate, and all the factors are relevant in estimating the model. All the independent variables are statistically signi cant, with a p-value of 0.00 (α = 0.05). Quality has the highest coe cient, followed by accessibility, responsiveness, and, lastly, affordability. The positive coe cients indicate that better perceptions related to all the four factors would lead to higher overall satisfaction.
Based on the results of multiple regression, the model can be presented as: Overall Satisfaction from Healthcare Delivery in the UAE = 3.32 + 0.51 (Quality) + 0.28 (Accessibility) + 0.18 (Responsiveness) + 0.11 (Affordability) The multiple regression model results indicate statistically signi cant relationships between the dependent and independent variables under consideration. Therefore, there is enough evidence to reject the null hypothesis.

Discussion
Based on the data analysis results, responsiveness and affordability appear to be lagging on customer satisfaction. Both responsiveness and affordability have a positive association with overall satisfaction. This indicates that when the customers' expectations concerning these two factors are met, the perceived total satisfaction with health care services will increase. While affordability has the lowest coe cient among the four aspects of health care delivery in the UAE, the factor is statistically signi cant in the model, and therefore, requires attention. It also appears that the respondents perceive that doctors do not spend adequate time with them.
The demand and supply gap of health care services has led to increased demand-side costs, resulting in a negative impact on the affordability of health care in the UAE [57]. The country is considered to be one of the most expensive destinations in the Middle East for medical treatment. At the same time, the average price of a doctor's visit is US$ 69, with an approximate 80 percent higher as compared with Singapore and 240 percent higher than in Saudi Arabia [57,58]. Moreover, the UAE's medical in ation is estimated to be 9.9 percent, which is the highest in the GCC [21]. The UAE government has chosen to present price control mechanisms for health services and medicines [59] to ensure that the health care services' eventual recipients are not experiencing extremely higher costs. Besides, the UAE government is transitioning towards universal health coverage for the population [24,[60][61][62][63] to reduce out-of-pocket health care expenses.
As discussed in section 2, the UAE's health care sector appears to be consolidated with a few players in the market. While having a few large players in the industry may promote investments in technology and human capital, it may also entail the risk that more signi cant players may exploit the end-users by offering high cost but low-quality services. Reducing the multi-level regulatory burden may lead to lower entry barriers and promote competition, leading to increased affordability and improved service quality.
Unlike the UAE health care industry, the insurance industry is fragmented, with 61 insurance companies serving a population of approximately nine million. This negatively impacts insurance companies' capacity to achieve economies of scale and scope and invest in providing cost-effective health insurance policies to retail and corporate customers, which would result in increased affordability. The UAE government has made efforts to consolidate the insurance industry by raising capital requirements for the insurance providers. However, concentrated efforts may be required to strengthen the insurance industry to drive the health insurance premiums down. A well-developed and mature insurance sector will enable risk pooling and diversi cation, resulting in lower costs and premiums for the insurers and the insured, respectively.
The ndings related to responsiveness require further investigation. According to the Dubai Health Authority [64] annual statistics report, there are 2.9 doctors for every 1,000 residents (0.5 percentage points decline as compared to 2016) in Dubai. This ratio is high compared to that in many developed countries. For example, the United States of America, Canada, and the United Kingdom is reported by the World Health Organization to have 2.6 (2014), 2.5 (2015), and 2.8 (2016) doctors per 1,000 residents, respectively [65]. Having a comparable density of physicians with developed countries does not indicate human capital's su ciency in the health care industry. The shortage of doctors leads to increased pressure to spend less time with patients, resulting in displeasure on the service quality's responsiveness factor. The most obvious solution to this is to increase the number of physicians. This can be achieved by encouraging youth in the country to opt for medicine as a career of choice and attract foreign talent. The recent announcement regarding changes in the immigration laws may pave the way for quali ed foreign doctors to join the UAE workforce.

Conclusion
This paper aimed to examine the association between overall satisfaction and affordability, quality, accessibility, and responsiveness in the UAE health care sector. Statistical analysis techniques such as PCA and multiple regression were used to model the relationship between the dependent and independent variables. The use of PCA enabled data-driven dimension reduction, resulting in the four factors mentioned above. The results of data analysis highlighted responsiveness and affordability as the most concerning factors for respondents. While the sample was large and appeared to represent the UAE population in some aspects, the use of convenience sampling may limit the generalizability of the ndings. The study can be extended by classifying the respondents according to their socio-demographic pro les and evaluating whether signi cant differences exist between perceptions of different groups of respondents. Furthermore, a study can be conducted to isolate the perceptions of those respondents who intended to have medical treatment in their home countries.