Informal Payments and Health System Governance in Serbia: A Pilot Study

Few studies address Serbian providers’ perceptions of informal payments and the association between these perceptions and health system governance. The aim was to investigate civil servants’ perceptions on informal payments of the health care system in Serbia and to link these results with available evidence on informal payments in other Western Balkan countries. A literature review on informal payments in Western Balkan and in-depth interviews of civil servants working with Serbian health care were conducted. Informal payments were mostly taking place in inpatient care and were based on low salaries of doctors, poor resources, a desire to receive better or faster service, fear of being denied treatment, and an expression of gratitude through gifts. Policy measures had limited effect and vulnerable groups were more likely to pay informally. There is a need for further research to highlight how health system governance and prevailing policies affect informal payments in Western Balkan.


Introduction
The European Union (EU) is enlarging to the Western Balkans in the coming years. To become an EU member, the candidate countries must agree to certain values and principles regarding health systems, for example, the main focus should be put on primary health care (Council of the EU, 2007). The countries must also commit themselves to the European policies and strategies for enhancing population health, for instance, the "Europe 2020" strategy (European Commission, 2015). In 2009, Serbia applied officially for EU membership, and in 2012, the country was granted a candidate status. Negotiations for membership began in January 2014 (European Commission, 2014).
To commit to the context of EU integration and its common values and principles, Serbia reformed its health care sector after the overthrow of the Yugoslavian government in 2000 (Simic, Marinkovic, & Boulton, 2012). Some of the challenges Serbia faces after the reform are (a) high referral rates from general practitioners to specialists, (b) a perception that specialists provide high quality care (World Health Organization [WHO], 2010), (c) weak contributions to the Health Insurance Fund by public enterprises experiencing economic difficulties, and (d) a deficit in funding for the health system (Simic et al., 2012). The poor funding has lowered the quality of services, creating long waiting lists and lowering the quality of health care equipment and facilities (Karajicic & Mužik, 2014). In addition, the primary care system suffers from inadequate skills and competencies and the low economic and social status of health professionals (Simic et al., 2012).
The services of general practitioners are often seen as obsolete, and when they provide easy access to specialists, it is perceived as high quality health care (Simic et al., 2012). Because general practitioners are the gatekeepers to specialists or secondary care, informal payments may exist to facilitate access to these services (Simic et al., 2012). Today, the health care system in Serbia is characterized by a universal health coverage system that is both publicly financed and publicly provided. Private insurance can also be obtained to receive quicker access and enhanced consumer choice. These, however, are not included in the public insurance scheme (Karajicic & Mužik, 2014).
In 2015, the Euro Health Consumer Index (Health Consumer Powerhouse, 2015) compared the performance of health care systems in 35 European countries based on 48 indicators. The Index ranked the Serbian health system as the 30th among these countries. One explanation of such performance could be drawn from Transparency International's (TI; 2015) Corruption Perception Index, which ranked Serbia as 78 out of 177 countries worldwide in terms of its level of corruption. This is an overall perception of corruption and includes not only health care but all public sectors in Serbia. According to the United Nations Office of Drug and Crime (UNODC;2011), there are notable variations in bribes between the Western Balkan countries. In this article, bribes are understood as giving, receiving, or offering goods or services to influence the actions of an official. Informal payments are, therefore, understood as a form of bribery. In Serbia, almost 14% of citizens at age 18 to 64 reported that they had directly or indirectly experienced bribery during the study period. Although there were variations between the regions of Serbia, 9% of the interviewees had paid at least one bribe during that period. In 56% of the cases, the citizen offers to pay, and in 14% of the cases, they are asked to make bribes to the public administration staff (UNODC, 2011).
According to several studies (International Federation for Human Rights, 2005;TNS Medium Gallup, 2011), the citizens of Serbia believed that health was the most corrupt sector. UNODC (2011) found that one in three Serbian citizens who participate in any type of bribery do so to receive better service and one in five to speed up the service. Medical doctors are one of the groups of public officials that receive the most bribes. In the present study, these illegitimate bribes are referred to as informal payments. The concept of informal payments also includes gratitude gifts. In this study, the term "informal payments" will, therefore, include both gratitude gifts and bribery.
Several studies (Avdyli, 2010;Colombini, Rechel, & Mayhew, 2012;Gupta, Davoodi, & Tiongson, 2000;Nekoeimoghadam, Esfandiari, Ramezani, & Amiresmaili, 2013) show that informal payments have negative effects on health care access, equity, and population health. WHO (2000) claimed that any demand for out-of-pocket payment (formal and informal payment for health services) will lead consumers to reduce their use of health services. This may hurt particularly the poor by limiting their access to health services. Lewis (2007) claimed that informal payments sometimes exist alongside formal payments, and a distinction between these is difficult, which undermines the reporting of informal payments. Therefore, informal payments may cause inequity in the sense that they can impoverish some groups of society, whereas groups with greater resources receive better quality of care. Consequently, the financing of the health system determines the burden of health care payments in society and the level of protecting poorer groups from further poverty (financial risk protection; WHO, 2000).
Informal payments indicate poor health system governance as they are linked to a lack of control and accountability in health care and weak rule of law (Gaal & McKee, 2005). In addition, information about the existence of informal payments in Serbia is scattered, and there is a poor overview of informal payments and their consequences on the Serbian health care system. To identify future research focus, this exploratory study aimed to investigate Serbian civil servants' (government officials, that is, doctors working in the health care system) perceptions on informal payments of the health care system (Gaal & McKee, 2004;Greer, Wismar, & Figueras, 2016;Vian, 2008). As this is a pilot study, only one country (Serbia) was selected for the empirical data collection. Only little evidence about the topic regarding Serbia was available. Knowledge based on existing literature from other Western Balkan countries completed the results (excluding Macedonia because no such study fulfilled the inclusion criteria of the literature review).

Definition of Concepts and Theoretical Framework
A variety of professions have acknowledged the existence of informal payments in health care. However, there are substantial differences in the definition of the concept and no definition covering the various aspects of informal payments is available (Cherecheş, Ungureanu, Sandu, & Rus, 2013. Vian (2008 argued that one of the fundamental issues that illustrate this conceptual confusion is the uncertain differentiation between gifts and bribes in health care. One perspective features cultural aspects and sees informal payments as an expression of gratitude, whereas another includes feeling compelled to make unofficial payments to receive higher quality care. These payments vary from flowers, sweets, small tips, or large sums of cash (Gaal, Belli, Mckee, & Szócska, 2006). As concluded in the systematic review of Cherecheş and colleagues (2013, p. 110), the definition of Gaal et al. (2006) is the best yet available. Despite its limitations toward solely considering informal payments as an additional payment on top of any formal out-of-pocket payment (Cherecheş et al., 2013), the present study will take inspiration from the definition described as a direct contribution, which is made in addition to any contribution determined by the terms of entitlement, in cash or in-kind, by patients or others, acting on their behalf, to health care providers for services that the patients are entitled to. Polese (2014) argued that there is a blurred line between gifts and bribes in that they depend largely on the context and long-term relationships that maybe generated from both gifts and bribes. Balabanova and McKee (2002) also doubted whether gratitude gifts exist in health care, as any payment favors the payer in obtaining preferential access to care. Liaropoulos, Siskou, Kaitelidou, Theodorou, and Katostaras (2008) showed that only one fifth of the respondents paid out of gratitude, whereas nearly half of the respondents did so out of fear of otherwise receiving lower quality care. One of those proposing a conceptualization of informal payments as an unregulated fee is WHO (2000), suggesting to define informal payments as an act of corruption (i.e., bribery). Gaal and McKee (2004) attempted to explain informal payments through a theory that takes its starting point from Hirschman's (1970) work on "Exit, Voice, and Loyalty." The theory attempts to analyze peoples' reactions to a decline in organizations and states, which leave consumers (patients) and employees (doctors) unsatisfied. In this case, consumers and employees would either leave the organization (Exit) or complain to available channels (Voice). The latter option assumes "Loyalty" to the organization, as voice is usually less effective and more risky (Gaal & McKee, 2004). Regarding informal payments, other characteristics of Exit and Voice are present insofar as dissatisfied people do not complain (Voice) or leave the organization (Exit) but seek change through informal methods such as payments or connections. The latter resemble an internal/informal Exit, hence the term INXIT (Gaal & McKee, 2004). The theory explains some possible reactions from patients and doctors when unsatisfied with the organization of health care. Lewis (2007) and Gaal and McKee (2005) emphasized that informal payments are symptoms of poor management, underfunding, poor control in health care, lack of accountability, and deficits in rule of law, that is, poor governance. Informal payments consequently appear to be linked to governance in health care, leading public and private spending in health care to have no impact (Lewis, 2006). Vian (2008) illustrated the reasons and opportunities for informal payments and suggested that civil servants abusing public position and power for private gain do so because they feel pressured to abuse, because they are able to rationalize their behavior (e.g., individuals' beliefs, attitudes, and social norms influence informal payments, and to some individuals, it may therefore be morally acceptable to give civil servants money or gifts), or because they have the opportunity to abuse. Civil servants often act in a manner that maximizes their self-interest, weighing the costs and benefits of acting corruptly against acting with integrity (Vian, 2008). The opportunities to act corruptly are influenced by six factors: monopoly, which creates opportunities for informal payments by limiting the ability of citizens to choose other service providers; discretion, referring to the autonomous power of civil servants to make decisions; accountability, which is governments' obligation to demonstrate effectiveness in reducing informal payments; citizen voice, representing the channels and means for active participation by stakeholders; transparency, referring to the active disclosure of information relating to decision making and performance measures; and enforcement, which is the collection of evidence of informal payments and the punishment of those engaging in such acts (Vian, 2008). These six factors are important in capturing informal payments and should be considered in strategies to tackle these.

Empirical Part
Sampling. A purposive nonprobability sampling was performed (Bryman, 2008) meaning that individuals who were appropriate for answering the research questions were approached. The participants were expected to be educated within the health care field (i.e., doctor, nurse, nursing assistant) and fluent in English (due to the feasibility of the study it was not possible to include translators or other international or local researchers). Online websites were searched to find participants at national and regional public health institutions and primary care centers in Belgrade and Vojvodina. The participants were sent an email, asking them to participate. The online selection of participants was necessary due to lacking network in the area and was a less time-consuming approach. Primary care centers were chosen, as they are public institutions where informal payments have been claimed to take place (Simic et al., 2012). The Public Health Institutes in Belgrade and Vojvodina were selected, as they do research on population health and on performance and functioning of the health care system. Other institutions such as the National Health Insurance Fund and the Ministry of Health were not included due to difficulties in locating contact information on specific officials through the Internet. Networking would have been an appropriate approach for locating these officials, though time constraints made this unrealistic.
Data collection and study participants. The data collection was conducted in Belgrade and Vojvodina in the first week of August 2014. The data consisted of two semistructured group interviews on-site and one follow-up interview through Skype (n = 4). Participants of the first interview were two public health officials from Novi Sad (Vojvodina) and those in the second interview were two health care providers from Belgrade. The Skype interview was with one of the public health official from Novi Sad. The participants were female, medical doctors, or specialists in social medicine and aged between 30 to 60 years. They worked as department heads at an institute of public health or as employees in a primary care center. The respondents of the on-site interviews were paired (two respondents in each interview) so that they could also assist each other with language difficulties, despite speaking fluent English, and to create better flow.
Interview guide and conducting the interviews. The interview guide included questions on the respondents' beliefs about informal payments and their views on the health care system and its governance, the functioning of the center, coordination between secondary and tertiary care, and service delivery. The questions were based on Vian's (2008) framework, theory on trust derived from the study of Aasland, Grødeland, and Pleines (2012). The themes of the guide were "perceptions on health care," "corruption in the health sector," "informal networks," "trust," and "equity." The interview guide was pilot tested and revised twice before the actual interviews. The interviews, conducted in English and recorded on a dictaphone, lasted approximately 1 hr. The interviews were transcribed verbatim into Microsoft Word. Transcripts of the interviews were returned to the respondents to confirm their statements and to allow them to add additional comments.
Data analysis method. The phenomenological hermeneutical approach adopted from Lindseth and Norberg (2004) was used to inductively develop emerging themes. To understand the interviewees' approach to informal payments, it was necessary to first apply this method to investigate their common perceptions of the term. To grasp the meaning as a whole, the text was read through several times, and an overall understanding of the data was formulated. In Lindseth and Norberg's (2004) methodological description, this is called a naïve understanding. This understanding guides the next steps in the analysis and must be validated through reading literature. The validated understanding then further guides the subsequent thematic structural analysis, where themes were identified in the text and separated into units of meaning. These units consisted of the respondents' quotations and were marked with the respondent's participant number. The units of meaning were printed out on paper to better manage the material and to identify sections and quotations that were similar and thus formed a pattern. Subsequently, these units of meaning were reflected against the validated naïve understanding, condensed into neutral summaries, and abstracted into main themes and subthemes. After this categorization, the themes were once again reflected on, reorganized, and sorted into final themes and subthemes. Finally, the findings from the data analysis were sent to the participants for feedback.
Ethical considerations. An application for ethical approval was sent to the Ethical Committee at the Institute of Public Health of Serbia including the informed consent form, the study protocol, and an explanation of the interview process and the interview guide. The Institute of Public Health of Serbia approved the ethical clearance on August 26, 2014. Despite applying for the Ethical clearance in a timely manner, it was not possible to receive the final approval before conducting the interviews. Nevertheless, the researchers corresponded with a member of the committee during July 2014, confirming the approval of the initial qualitative study, with minor revisions. Before conducting the interviews, an informed consent form was signed by the participants.

Literature Review
A literature review was completed to complement the qualitative data. An initial literature review was also conducted before the qualitative study, among other things, to narrow the study of informal payments to only one country (Serbia). The following literature review was focused on health system governance and informal payments in all Western Balkan countries to get a broader view on informal payments in the area. The aim was to answer to the following question: What existing evidence is available on informal payments and how dispersed are informal payments in Western Balkan?
Inclusion and exclusion of studies and search strategy. Inclusion and exclusion of studies was completed in an iterative manner, where search terms were redefined when becoming familiar with the literature. In assuring the comprehensiveness of evidence, the steps of phrasing the research question and searching and selecting relevant published and gray literature were repeated. The three overall search terms "Health Systems Governance," "Western Balkan," and "Informal Payments" were constructed, including 89 sub-keywords connected with the Boolean operator OR. The three overall search terms were initially searched separately in the eight databases of PubMed, CINAHL, Social Sciences Citation Index, Science Citation Index, Applied Social Sciences Index and Abstracts, Worldwide Political Science Abstracts, Conference Proceedings Citation Index Science, and Conference Proceedings Citation Index Social Science & Humanities. Afterward, a combined search of the separate searches was then completed with the Boolean operator AND.
Inclusion criteria were based on the number of Western Balkan countries included; whether concerning governance and informal payments in health care; sufficiency of quality and whether including primary evidence; studies from the time frame 1991-2015; or studies in English language. Excluded studies were unable to fulfill inclusion criteria or were duplicates. Both quantitative and qualitative studies were included, making this a mixed studies review.
A manual search was also conducted by looking through reference lists of relevant literature. Many articles were not in English, did not regard Western Balkan countries, did not meet all inclusion criteria, or were already identified. Therefore, no additional articles were included from this search.

Study selection.
A total of 387 studies were identified from the search after 40 duplicates had been removed. In total, 277 studies were excluded after screening titles and abstracts. Many of these had focused on the business sector or political or police corruption. 70 studies were examined more thoroughly, from which 64 did not meet the inclusion criteria. Finally, six studies were included in the review (Figure 1). Data from the six studies were extracted to Microsoft Excel to organize the data. Information extracted from the studies was completed with inspiration from Centre for Reviews and Dissemination (2008) and included 18 indicators, that is, year, author, study design, aim/objectives, age, gender, and ethnicity of patients.
Appraisal of included studies. The six included studies were assessed by using the Mixed Method Appraisal Tool (MMAT; Pluye et al., 2011), developed by McGill University in Canada. The methodological quality was assessed using the corresponding criteria for qualitative, quantitative, or mixed method studies. A quality scoring system was performed giving quality scores of 25%, 50%, 75%, or 100% for each study. One included study scored 50% (moderate) whereas five of the studies scored 75% (good), due to mainly lack of information in the reporting (Table 1).

Synthesis methods.
A thematic synthesis method was used to interpret the data, to maintain consistency between the qualitative study and the literature review. The analysis was completed with inspiration from Popay and colleagues (2006), who also argue that thematic syntheses maybe used when synthesizing qualitative, quantitative, and mixed method studies. The quantitative variable labels that were included in the individual studies were extracted as themes in the same way as conceptual themes were extracted from the qualitative studies (Popay et al., 2006). These were then summarized narratively and are reported in the literature findings section.

Findings of the Empirical Part
According to the main themes (italics) from the structural thematic analysis, the respondents emphasized primary care as the main function of the health system. They also argued that primary health care is poorly governed by the state, which creates low incentives for doctors to perform. This poor motivation caused the doctors to not live up to people's expectations for service provision. Furthermore, only the public provision of health care services is available in Serbia, but some larger cities have private beneficiaries to support their health budget, causing an uneven balance between cities. In general terms, some respondents believed that primary care functions well. However, respondents also stated that informal payments exist in health care and that poor communication caused informal rules to take over formal rules. Finally, transitioning into the EU was seen as a positive step, although the process of transitioning is perceived to move slowly. The main themes will be further elaborated in a narrative hermeneutic way.

Perceptions on Health Care, Service Provision, and Governance in Transition
The respondents believed that primary care was important because of its closer patient-provider relationship compared with secondary and tertiary care. They believed that health care benefited economically from a focus on primary care. Concerning governance, the data indicated that health care was poorly managed, which indicates poorly governed primary care. This was drawn from statements, among others, that there was poor communication between the three levels of health care. General practitioners were believed to be obsolete and incapable of providing sufficient care to patients. This was related to poor communication between the three levels of health care, as the respondents believed that specialists undermined general practitioners, which caused inefficiency and minimized trust between health professionals. Therefore, discrepancies between people's expectations of services and actual service provision were reported. The interviewees argued that trust in general practitioners had decreased primarily due to their low social status. However, the respondents felt that the poor funding of primary care caused more damage than this lack of trust. The respondents believed there were a lack of funding and an emphasis on specialist treatment within secondary and tertiary care.
They do not have enough tools to work with. So it is very difficult to work in primary health care because in general, we said that the most important part of the health care is the primary health care but when the Insurance Fund decided where to give the money, they give the money for hospitals and other centers: 25% [of] financial resources are going to the primary health care and they said they must solve 80% of the whole health problems and needs.
Due to underfunding and poor resources in primary care, the respondents believed that citizens perceived primary care doctors to be incompetent. Specialist care was perceived as higher quality. Doctors, in general, were also believed to be unfamiliar with general practice when graduating from school because schools emphasize specialist medicine. All of the above is believed to create an expensive health care system that is based on treatment rather than prevention. These factors were perceived to create low incentives for doctors' performance and poor motivation among primary care doctors, causing more inefficient service provision. The data indicated that poor resources in primary care were seen to be the reason why general practitioners referred patients to specialists. Consequently, general practitioners were perceived to have an obsolete function.
The main problem is, if the doctor wants to solve some problem (treat a patient in the primary care clinic), he doesn't have enough tools and resources. So he cannot do that. So it is the main problem: it is easier to send him to the specialist.
Another example of poor incentives to perform, according to the informants, was the low salaries of doctors.
. . . one, they are not enough paid. Salary is disgracefully low. They are not paid enough to do their work. Today, also, the salaries are disgracefully low for that kind of quality professionals.
Low salaries also indicate a low position in the society of general practitioners. As a consequence of these factors, the respondents believed that some doctors have left the country to find more favorable jobs elsewhere.
The informants believed that there was a difference between rural and urban primary care centers. For instance, Novi Sad, an urban city, reported having several private beneficiaries assisting with health care; these are lacking in rural areas.
So NGOs together with health institutions are providing some health care programs for the citizens of Novi Sad. All other, as to my knowledge, other municipalities, towns or cities in Vojvodina do not have that kind of budget.
Private practice was not included in service provision. Instead, patients who could afford to do so had the option to supplement public services with private insurance services.
Persons can also add private insurance. So, everyone is free to pay contributions to private fund and then make some additional health services, etcetera, or better conditions in hospitals, etcetera. So that is how it works in our country. Now.
The respondents believed that only a low percentage of citizens could afford additional private insurance and that private providers should not be included with public service provision. They also stated that they believed that a future plan of the government is to include private practice in public service provision.
The informants felt that the transitioning into the EU would benefit Serbia. However, one respondent believed that some laws from the previous regime, Yugoslavia, were more favorable than those from the EU. For example, the EU immunization laws are believed to favor collective rights over individual rights. The huge way of solutions in the European Union is good, and we are happy to go there because, I think, maybe they will change the system of managing, and the transition become the real thing, not only the thing on paper. But in some way, they can look in local legislative and see maybe it is better.
Corresponding to the previous results on governance, two respondents believed that Serbia still had long way to go before becoming a democratic European country but that it was gradually transitioning into the EU.
Well, we are learning. We are learning the hard way, I think. It's not something that is natural for us, still. We are far from that, you know. We do have it on paper, you know. We have a multiparty system or something, but deep down, here and here [points to the heart and head] I don't feel that very much. I don't feel that.
There appeared to be a perception that traces of the previous Yugoslavian regime remained in Serbia. The informants believed, for example, that politicians did not fulfill their duty to transition into democracy. It was also alluded that general citizens did not regard solidarity and social capital (including informal and formal social networks) as common principles of society. Half of the informants believed that there were different societal aspects to living in rural and urban areas, and therefore trust and democratic participation were different between these areas.

Perceptions on Informal Payments
The respondents expressed belief that informal payments were present in the health system. Some of the respondents believed that specialists and the biggest branches of medicine were the most burdened, and therefore additional payments for services occurred. Half of the respondents said that the public perceived primary care to be transparent and open. They did not believe that informal payments occurred in primary care, especially given the increased licensing and accreditation in the area. This was, however, not the case for hospitals, as they were difficult to monitor because they provide a wider range of services.
They are trying to do that (monitor hospitals), but it is very hard. It's hard to monitor their services because there is so wide [a] range of services. And I think it will be some time before they have to do like us.
Respondents further claimed that primary care should be the core of any health system, due to the positive patientprovider relationship and the economic benefits. Although in Serbia, general practitioners were believed to be obsolete and incapable of providing sufficient care to patients. Also, the poor funding in primary care and an emphasis on specialist treatments increased the discrepancies of people's expectations of services and the actual service provided.
Due to underfunding and poor resources in primary care, the respondents believed that citizens perceived primary care doctors to be incompetent, perceiving specialist care as higher quality. The data indicated that poor resources in primary care were seen to be the reason why general practitioners referred patients to specialists, indicating that informal payments may take place in this process.
The main problem is, if the doctor wants to solve some problem (treat a patient in the primary care clinic), he doesn't have enough tools and resources. So he cannot do that. So it is the main problem: it is easier to send him to the specialist.
Low salaries and poor resources in the health system were also perceived as a reason for these informal payments for services, because low salaries indicate a low position in the society of doctors. Although, the respondents believed that the payments were based on patients' willingness to access higher quality services and not on doctors asking for money.
You go to the operation and you want the best. Patients feel they have to pay, for that. For the best service. Some respondents said that it was difficult to separate informal payments from gratitude payments and gifts. Giving gratitude gifts, such as flowers and cakes to doctors, appears to be a common cultural phenomenon. Such gifts are not perceived to be an informal payment in Serbia, according to the respondents. Both paying additional cash informally and giving informal gifts were mostly to express gratitude for satisfactory care, and the respondents felt that doctors would find it rude to deny such gratitude. It was also argued that these gifts create a positive patient-provider relationship. All respondents believed that any payments for extra care through informal payments would be of such a small amount that they were not significant. Therefore, they did not perceive gratitude gifts as being unethical or informal payments as being a critical issue in health care. One of the respondents further believed that corruption was more dispersed in politics and that this was the area where action would be needed.
Bigger corruption is within politicians, not with the doctors. Believe me.
The media was also believed to exaggerate corruption, and the respondents perceived reality to be different.
They talk all the time, but they don't say that it is one percent, ten percent or zero point one percent, of the health professional who were given some money or something. I don't think it is fair to talk about that, because we are living in twenty-five years in very bad condition[s], and they never tell that. That, for example, a clinical doctor, they work for twenty-four hours for ten Euro. Do you know that, for example?
Respondents also perceived that some informal rules took precedence over formal rules in the health care system. They described poor communication in health care as arising from the low status of general practitioners. Poor communication and mistrust in the competency of general practitioners created illegitimate informal practices. Consequently, informal procedures took the place of formal rules, as general practitioners used their friendly relationships with other doctors to navigate the health care system rather than their professional networks. It was evident that some general practitioners felt disregarded and engaged in informal behavior as a coping strategy. Therefore, when needing assistance or advice regarding treatments, general practitioners tended to contact friends rather than a more appropriate specialist for advice.

Description of Included Studies
Five of the six included studies were peer-reviewed articles and one a working paper, mostly concerning Albania. Two studies considered informal payments in Serbia and the rest concerned Bosnia Herzegovina, Montenegro, Kosovo, or Croatia. No Macedonian studies were in the included data. Designs were typically quantitative, with only one qualitative and one mixed method study. Sample sizes also varied from 131 to 49,848 people (Table 1).
When examining the included studies, three overall themes appeared: (a) effects of poor governance and frequency of informal payments in Western Balkan, (b) causes of informal payments in Western Balkan, and (c) informal networks in health care systems of Western Balkan. These themes will be elaborated below.

Effects of Poor Governance and Frequency of Informal Payments in Western Balkan
In their mixed method study from Serbia, Arsenijevic, Pavlova, and Groot (A) found that women frequently report paying both quasi-formal payments and informal payments for maternity services. Consequently, quasi-formal payments were a result of discrepancy between official and hospital guidelines, highlighting the poor governance in the system. The reported existence of informal payments varied a great deal from 2% in what they identify as the "pro-government" publications (from 2009), 10% to 14% in the "non-government" publications (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009)(2010), and 22.1% in the online questionnaires (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008). In this study, the highest paid bribe amounted to 500 Euro. Despite these payments, women still reported problems with the quality of care they received.
Informal payments were also reported in a cross-sequential study from Kosovo, Serbia, Bosnia Herzegovina, Montenegro, and Albania, conducted by Bredenkamp, Mendola, and Gragnolati in 2011 (B). They found that informal payments represent a relatively large share of total health expenditure, and often among the poor population. Accordingly, Albania had the highest mean share of informal payments of all five countries. Also Albanian household in the lowest and poorest quintile paid on average 8% of their total health expenditure in the form of informal payments, compared with 4% in the richest quintile. In Kosovo, the share is more evenly distributed; however, the poorest still paid 2% of their total health expenditure in informal payments, whereas the richest paid 1%. In Serbia, however, the richest quintile paid a slightly greater share of informal payments (3%), compared with the poorest (1%).
In addition to this study, one working paper (D) and one article (E), both using data from the 2002, 2005, and 2008 Albanian Living Standard Measurement Survey, found similar results. The former study (D) found that the incidence of informal payments were fairly high in Albania; however, these were less frequent in outpatient care than in inpatient care. A general decline was also found in informal payments in 2005 and 2008, whereas the other study (E) found an increase in all health expenditure, including informal payments, from 2002 to 2008. (In all categories of expenditures, inpatient and outpatient public and private services related to medical fees, laboratory tests, drugs purchased, gifts paid to medical staff, and transport).
Both studies (D, E) revealed that poorest groups in Albania were those where a decline in informal payments was least pronounced. In fact, the amount that the poorest group paid informally increased almost 5 times over the years (121 ALL in 2002-777 ALL in 2008, whereas in the richest groups, the increase were more moderate (E). This increase in informal payments accordingly showed that the poorest group had an increased expenditure of informal payments from 7% in 2002 to 10% in 2005 and to 30% in 2008 (of their total expenditure). Although, the increase was not accompanied by an increase in the relative per capita total expenditure and, further, shows that policy measures through 2002 and 2008 have had a negative impact on the poorest groups in this study. Finally, in 2008, the poorest households had a relatively less budget than the rich if compared with 2002, and the poorest faced higher out-of-pocket payments.
Older people in general paid higher informal payments for inpatient and outpatient care, maybe as they require more care than others (D). Generally, working-age adults were the ones that paid larger amounts of informal payments and patients in higher consumption quintiles spend more on informal payments for both inpatient and outpatient services. However, they also found that health insurance lowered the probability of paying informally for both health care services, but this factor did not affect the amount paid. Correspondingly, the results showed that generally vulnerable people having a smaller budget to pay for health care were more likely to pay informally (D). These results demonstrate an imbalance of informal payments and indicate that vulnerable groups are more prone to informal payments, and that people with more resources spend more on informal payments.
In a Croatian study, Radin (C) found that in 2007, those who experienced corruption and those that consider corruption in health care to be the biggest problem (salience) also did not prefer using public health care facilities. Although, a preference toward choosing these facilities increased in both indicators with age and settlement size. Also, an increase in age increased the odds of trusting public health facilities; however, in overall terms, corruption still lowered the odds of preferring public health facilities.
From the opinion polls of 2009, the results were different. Those with experience of and salience toward corruption in public health care were more likely to prefer public health care. Although, this preference decreased when having a greater degree of education, improvement in economic situation, and increase in settlement size. These results from 2009 did not support the fact that health care corruption lowers trust in health care of the public sector. In fact, the effect of experiencing health care corruption had a positive effect on trust in public sector health care.

Causes of Informal Payments in the Health Care System in Western Balkan
In maternity care in Serbia (A), researchers discovered that women paying informal payments expressed dissatisfaction with having to pay, and reported that they only did so for the safety of their child. Among the women who did not pay and who did not have connections, 14 reported that they regretted this and would pay next time. These payments were in addition to the quasi-informal payments. Others paid informally due to the low salaries of providers. Patients in Albania correspondingly recognized low salaries and considered their informal payments as contributing salaries. Providers further rationalized their behavior and claimed a higher standard of living (F).
Respondents of the just mentioned Albanian study recognized that some informal payments were given as gifts to express gratitude. Although, other members of the public were, as in the study of Arsenijevic et al. (A), also upset to pay informal payments. Respondents reported that patients willingly paid informally to receive better attention, faster care or for a fear of substandard care, and a feeling of security. Many respondents, both public and providers, also reported situations were no treatment was given if a payment was not made (F). This corresponds to findings in the study of Arsenijevic and colleagues (A), where respondents reported, that is, obstetricians to most frequently be paid, to secure his or her presence during childbirth, out of fear for substandard care and to secure better quality of care.

Informal Networks in Health System of Western Balkan
Several studies recognized that people in informal networks (i.e., friends with or related to a health care provider) are less likely to pay informal payments. One study (A) found in their questionnaires that 26 of the women who did not report bribes indicated that they had connections, such as friends, relatives, or colleagues, who helped them avoid informal payments by not accepting payments from them and ensured special treatment and adequate care. The sample indicated that during 2005-2008, the number of women reporting having special connections increased, whereas reporting informal payments decreased in the same time frame. Another study (F) similarly found that providers described lack of social connection as an indicator for increased willingness of providers to ask for or accept informal payments. Some providers even indicated that an informal payment was a way to "warm-up" or create closer provider-patient relationship. Tomini and Groot (D) described that rural residents paid less, but in general, they paid higher amounts of informal payments in outpatient care compared with residents of urban areas. This indicate that these rural residents do not pay in their own area due to social connections, but pay higher amounts in urban areas where such network connections are lacking and higher opportunity costs are involved. This is, for example, when needing care at larger hospitals in urban areas. The fact that vulnerable people were more likely to pay informally is attributed to the possibility that people with more social connections and information (social capital; for example, health care providers) are less likely to pay informal payments due to their network.

Discussion
Findings from the study indicated that the present way of governance of the Serbian health system might create negative effects for the system's functioning. For example, the underfunding of health care and low salaries for doctors seemed to create imbalances in the Serbian health care system, whereby informal payments were seen as a coping strategy. This is in line with the findings of previous studies on the subject (Arsenijevic, Pavlova, & Groot, 2013Janevic, Sripad, Bradley, & Dimitrievska, 2011) and which connect informal payments with poor management, underfunding, or a lack of accountability (Gaal & McKee, 2005;Lewis, 2007). Accountability is especially important in a democracy to hold civil servants responsible for inappropriate actions and to reduce the likelihood of informal payments (Greer et al., 2016;Vian, 2008). UNODC (2011) identified corruption and lack of transparency in the recruitment procedures when hiring for the public sector in Serbia. This may illustrate poor public management of the health care system that lacks integrity and transparency. When investigating Serbian's perception of transparency, United Nations Development Programme (UNDP) and TNS Medium Gallup (2011) further found that citizens regard transparency as being less important in the fight against corruption.According to Vian (2008), transparency is a key factor in the fight against corruption. Comparing this with the INXIT theory of Gaal and Mckee (2004), the declining performance of health care in Serbia may leave patients and doctors' needs unsatisfied. Patients and doctors may choose to leave the system (Exit); however, for patients in Serbia, private practice is expensive and not commonly available and they may have no ability to Exit. This may foster the presence of informal payments, as patients secure themselves higher quality care (INXIT). Through Voice, patients can choose complaint mechanisms. However, studies show that some patients in Serbia perceive reporting of informal payments to be pointless, whereas others benefited from making informal payments (TNS Medium Gallup, 2011;UNODC, 2011). Informal payments in Serbian health care appear to be a strategy to informally "Exit" the system, violating the established rules and diminishing democracy.
Despite only describing others' experiences with paying informally, the interviewed Serbians argued that informal payments were more pronounced in inpatient care services and that such payments in outpatient care services were not possible as the centers were working toward becoming licensed, which required more monitoring. The monitoring of hospitals, however, was perceived difficult, and there were still no intention on the part of policy-makers to start enforcing greater control in these institutions. Even though the respondents may not have experienced paying informally firsthand, this is worrying in contrast to the ideas of poor technical skills in primary care providers, which facilitated high referral rates to specialist providers, and the informal networks between providers. It could mean that providers, being perceived as having poor skills and not being included in an expert network, are passive about recommending policies for health as stakeholders. These issues relate to accountability and citizen voice in Vian's (2008) framework and affect the success or failure of policies (Greer et al., 2016). The study by UNDP and TNS Medium Gallup (2011) in Serbia similarly found public passivity and ignorance concerning the fight against corruption. There was also a perception among our respondents that corruption was given too much attention in the media and that the real problem in health care was not corruption. This may stem from a lack of knowledge about informal payments and gift giving in health care. Providers in Serbia may, therefore, not be held properly accountable for accepting informal payments.
There was a general perception in the findings that patients were aware of doctors' low salaries and poor working environment; hence, they gave them gifts and money out of gratitude or to receive (better) care. Providers further rationalized their behavior and claimed a higher standard of living. Thompson and Witter (2000) found such low salaries to be an important reason why patients give both monetary and nonmonetary gifts to providers. In Vian's framework, this is identified as rationalization and pressure to abuse. Doctors may feel a need to accept informal payments because they have debt and need to make a living. Patients rationalize their behavior by regarding informal payments as being acceptable in the current situation despite free entitlement to health care. Although as several of the Albanian studies in the literature review showed (D, E), such payments create inequity in access to services, as vulnerable groups have a higher likelihood to pay informally. Doctors having the opportunity to accept informal payments (e.g., when a health system is poorly governed) will often act to maximize their self-interest. This may not be the cause of informal payments in some countries, although as Belloni and Strazzari (2014) described it, war can cause illegal informal practices and corruption from informal actors, who exploit their influence in a transitioning country in an effort to maximize their self-interest, that is, due to high discretion or monopoly in health care services. Corruption is, according to these researchers, not a deviation caused by a few public officials but rather a structural matter that becomes standard operating procedure.
It is widely discussed whether informal payments are paid willingly or out of fear of substandard care. Petrovic (2014) found that people in Serbia bribe doctors for better treatment (providers identify an opportunity to abuse) and that corrupt officials may thrive in a system with low salaries (pressure to abuse and rationalization), low health care funding, and high unemployment rates for doctors (providers feel pressure and have an opportunity to abuse). When patients are ill-equipped to assess the quality and adequacy of physicians' decisions and actions, informal payments maybe given to secure fair treatment. However, quality is often not maximized due to the patient's inability to judge medical quality (Thompson & Witter, 2000). The poor knowledge of patients and the dominance of public service provision (monopoly), as is the case in Serbia, limit the ability of patients to choose other providers for services, consequently facilitating informal payments.
Civil servants of the qualitative interviews further described that only a low percentage of doctors engaged in corruption, but that gifts to medical professionals were accepted. Giving gifts to health care providers in gratitude were only identified in one of the reviewed studies from Albania (F). Other studies concerning Serbia (A, B), Kosovo (B), and Albania (B, D, E, F) found patients paying informally for health care services, illustrating that these payments maybe more pronounced than the Serbian respondents initially perceive. In Serbia, gifts are seen as a token of gratitude to doctors (patients' rationalization). Doctors have difficulty refusing these gifts, as patients will feel offended and will accordingly feel a pressure to abuse. Gift giving and favors among family, friends, and local government officials in postcommunist European countries are commonly viewed as social logic (Grødeland, 2013). The reciprocity of "giving and taking" is rooted in sociocultural factors, where patients may express gratitude through gifts (Gaal & McKee, 2004;Polese, 2014). Informal payments may, therefore, be underreported in Serbian health care, as gift giving is undetermined. From the same perspective, the present study illustrated that gift giving created a positive patient-provider relationship, which maybe a rationalization by both patients and providers. However, some patients (A, F) were upset to pay extra for services that they were entitled to. Polese (2014) believed that informal payments may not only create a long-term relationship and social contacts but also produce obligations of reciprocity. Petrovic (2014) argued that many doctors in Serbia are accustomed to receiving small gifts of gratitude but that up to 40% of doctors ask for such gifts, and a number of these also charge informal cash payments. It becomes a moral discussion then whether gifts are seen as gratitude or bribes. Petschnig (1983) denied the validity of gratitude gifts, exclusively believing that they are fees to obtain personal gain. Although WHO (2000) defined any form of informal patient payment (including gifts) as an act of corruption, in practice, it is difficult to determine whether payments or gifts are tips or bribes.
Despite innocent and kind gestures to medical professionals, both WHO (2000) and several researchers (i.e., Colombini et al., 2012;Grødeland, 2013;Gupta et al., 2000) showed that informal payments for health care can have negative effects on health status, health care access, and equity, as people delay seeking care or sell assets to pay for care. WHO (2000) believed that co-payments and informal payments for services will reduce demand in lower socioeconomic groups. Some segments of society will not be able to pay additional health care costs, whereas other groups may receive additional care due to their higher socioeconomic status. It also becomes difficult for decision-makers to draw a line between gratuities and bribes when legitimating gifts in the health care system. In Serbia, further confusion is seen in the charging of co-payments for services. Distinguishing between formal and informal payments becomes difficult, affecting the reporting and perception of informal payments. In Serbia, it maybe uncertain to patients and providers whether they are acting corruptly because there is no clear understanding in the general public of informal payments.

Methodological Considerations
Sampling of the empirical part of the study was purposive to identify civil servants working in the health system. A further intention was to include patients and/or the general public to identify a broader perspective on informal payments, but this was not feasible, as a translator or local researchers would have been needed and also given the limited time and resources available for this research. For the same reason, only Englishspeaking participants were included, potentially missing information that could have been valuable to the study. Sampling bias is, therefore, a potential limitation to the study, and the purposive sampling was sufficient to develop an initial understanding, but is not generalizable. The literature review was conducted to strengthen the evidence of these qualitative data, providing a more solid understanding of the phenomenon of interest. Inclusion of studies was, however, narrow and could have provided a stronger evidence base with broader inclusion criteria. The purposive sampling of the public health institutes and primary care centers included inclusion and exclusion criteria for doctors and public health officials within Belgrade and Vojvodina. Other cities could have been included, and more interviews could have been conducted. However, networking and additional calls would have been necessary for additional recruiting, which was unrealistic in terms of time constraints. It was also necessary to obtain approval from management before conducting interviews, which further limited recruiting. Although this is a pilot study, the included data are sufficient to suggest research that is needed to fill the knowledge gaps. Furthermore, a wider variety of professionals would still have represented the same institutions as already included, as other institutions declined to participate. Nonetheless, the literature review was conducted in a comprehensive manner, including data from several Western Balkan countries and enabling a comparison between the empirical findings and what is already known on the subject.
Three potential participants denied participating in the study when realizing the topic to be connected to corruption and in fear of consequences to the individual participants. Budak and Rajh (2012) argued that corruption maybe underreported due to citizen's awareness that they live in a "surveillance society," combined with their memories of an undemocratic regime. This might prevent the reporting of corruption experiences. Some civil servants were also provided an option to fill out a questionnaire rather than to provide a live interview. Questionnaires were, therefore, sent to these respondents and to all potential participants who did not respond, to accommodate any reluctance to conduct a live interview, but none were returned. Sampling bias in the qualitative study is therefore present, as only those subjects who were willing and able to talk about health system governance and informal payments were interviewed.

Conclusion
Primary care was not considered to currently be serving as the core of health care and was perceived to be underfunded and inefficient. Underfunding and the poor motivation of doctors were perceived as incentives for corruption and informal payments according to the literature from other Western Balkan countries and the Serbian respondents of the empirical study. A related issue is the culture of gratitude gifts identified at all levels of health care. Difficulties in terms of defining the difference between informal payments and gratitude gifts were obvious, and if informal payments are perceived as legitimate, it might lead to underreporting, making the prevention of informal payments difficult. Moreover, a poor implementation of strategies tackling informal payments may worsen the situation, creating an increase in patients paying informally. Such consequences may exist particularly for vulnerable groups, as these were found in some studies to be more likely to pay informally. It may also be difficult for patients and providers in Serbia and other Western Balkan countries to separate legitimate and illegitimate co-payments for health care. From the statements of respondents, scientific research (Gaal & McKee, 2005;Lewis, 2006Lewis, , 2007, and theoretical concepts (Gaal & McKee, 2004;Vian, 2008), this study suggested an association between poor governance of the health care system and informal payments in Serbia. One of the most imperative issues identified was the value-laden and culture-bound perception of what is considered an informal patient payment.

Implications for Future Research
A more comprehensive systematic review would be needed to examine research gaps such as the lacking scientific evidence on informal payments by some vulnerable groups, that is, migrants. This includes also governance of the health care systems and implications of paying informally. More scientific research is needed on the nature of gifts as gratitude to doctors and on patients' trust in health care providers (determining, for example, their selection of health care facilities) and the quality of treatment that may lead them to pay doctors informally. Future research should further focus on the implementation and effectiveness of interventions on informal payments such as the effectiveness of policy measures.
A more in-depth study through interviewing the general public and civil servants of the health care system would also be needed. The study maybe used to inform decision-makers on positive and negative aspects of ongoing anticorruption strategies. The aim should, therefore, focus on generating new theories on ways to implement strategies to combat informal payments in the Western Balkan health care settings.