A literature review of the disruptive effects of user fee exemption policies on health systems

Background Several low- and middle-income countries have exempted patients from user fees in certain categories of population or of services. These exemptions are very effective in lifting part of the financial barrier to access to services, but they have been organized within unstable health systems where there are sometimes numerous dysfunctions. The objective of this article is to bring to light the disruptions triggered by exemption policies in health systems of low- and middle-income countries. Methods Scoping review of 23 scientific articles. The data were synthesized according to the six essential functions of health systems. Results The disruptions included specifically: 1) immediate and significant increases in service utilization; 2) perceived heavier workloads for health workers, feelings of being exploited and overworked, and decline in morale; 3) lack of information about free services provided and their reimbursement; 4) unavailability of drugs and delays in the distribution of consumables; 5) unpredictable and insufficient funding, revenue losses for health centres, reimbursement delays; 6) the multiplicity of actors and the difficulty of identifying who is responsible (‘no blame’ game), and deficiencies in planning and communication. Conclusions These disruptive elements give us an idea of what is to be expected if exemption policies do not put in place all the required conditions in terms of preparation, planning and complementary measures. There is a lack of knowledge on the effects of exemptions on all the functions of health systems because so few studies have been carried out from this perspective.


Background
If illness constitutes a major shock for households, the cost of healthcare is a second shock that strikes harshly and can drag them into poverty [1]. In Burkina Faso, 80% of poor households go into debt or sell assets to pay healthcare costs [2]. User fees in public services contribute to these catastrophic expenses [3], reduce health services utilization [4,5] and are a major impediment to universal healthcare coverage [6,7]. One solution being considered to improve access to care is to make services free at the point of service. Many countries, such as South Africa, Mali, Niger and Ghana, have therefore implemented exemption policies, sometimes targeted to population groups (e.g. children under the age of five years) or to specific services (e.g. caesareans).
Among the many articles published on this topic, two knowledge syntheses have been carried out. One presented the available scientific knowledge on the emergence, formulation, implementation and effects of exemption policies in five African countries and identified research needs [8]. The other was a systematic review aimed at evaluating the effectiveness of funding mechanisms to promote healthcare access for the poorest, including user fee exemptions [9]. Despite the poor quality of their research design, the authors concluded that the studies had shown that exemptions had a positive effect on health services utilization, but that they were not sufficiently effective to protect the poorest.
Even though these policies have produced an increase in the demand for services and therefore would appear very technically relevant, they raised a number of pragmatic concerns. In fact, these reforms have often been organized within unstable health systems that are not much used by the population and whose organizational deficiencies and dysfunctionalities are well known [10,11]. While still subscribing to the principle of point-of-service fee exemption, decision-makers from 15 African countries reiterated these concerns at a meeting held at the end of 2010 [12].
A health system is a collection of services, persons and actions whose interactions are complex and interrelated. As has been affirmed by the World Health Organization (WHO) and others before [13], for a health system to be effective and equitable, its different functions must work in unison, and any change will therefore have an effect on each function [14]. Thus, exemption can modify a health system's equilibrium and disrupt its essential functions. The objective of this article is to identify the disruptions in health systems that are caused by these new public policies.

Methods
Our analysis was based on a survey of the scientific literature [15]. That survey took the form of a scoping study, an approach which summarizes all scientific knowledge on a specific topic, regardless of the type or quality of the studies' designs. This summary is useful to better understand the nature of the scientific studies currently being done and to identify possible shortcomings. It differs from, and is often seen as a precursor to, systematic reviews that synthesize the most robust scientific data by using methods that limit biases and random errors [16]. We applied the systematization of the scoping study process proposed by Arskey and O'Malley [17]. Details of the documentary research strategy have been described elsewhere [15].
An article was retained if it: 1) dealt with a healthcare user fee exemption experience on a national scale (public policies); 2) reported original empirical data; 3) involved African countries; 4) mentioned pressures or disruptions in the health system; 5) was published in a peer-reviewed journal or monograph; 6) was published between 1988 and 2009 inclusively; and 7) was in French or English. The studies' designs were reported using the classification system of the Mixed Method Appraisal Tool [18], an instrument used to describe the quality of studies, whatever their nature.
We examined the content of the articles to identify the elements that produced bottlenecks or pressures on the health system. WHO has defined a health system based on six essential functions (Table 1). We used this framework to analyze the content of the articles retained.

Description of the studies identified
Our research approach gathered 23 articles (Table 2) from seven different African countries: Ghana, Kenya, Madagascar, Senegal, South Africa, Tanzania and Uganda. The topic most often studied was service utilization or demand (n = 7), not surprising given that improving utilization is usually the primary objective of exemption policies. Few articles looked at policy implementation (n = 3) or changes in service quality (n = 1). The types of study designs included quantitative nonrandomized (n = 12), quantitative descriptive (n = 2), qualitative (n = 3) and mixed (n = 6).

Service delivery
Many articles mentioned increased service utilization after the implementation of exemption policies, particularly in Madagascar [19], in Uganda for primary care visits [20], and in Ghana and Senegal for institutional deliveries [21][22][23]. In Uganda, it appeared that the free services could not satisfy the increased service demand, prompting patients who were better-off to use services that were not free [24][25][26]. This increase in demand therefore put pressure on the health system. This pressure was somewhat alleviated by additional reforms implemented on the supply side (increased budget for drugs, improvements in the stocking system, salary increases) [20]. However, the increase in demand had a negative effect on service quality (non-availability of drugs, longer wait times, decreased motivation among health workers) and, in the end, on malaria control and on users' confidence in the health system [27]. The decline in service quality could also be attributed to an abrupt change in policy and to underfunding.

Health workforce
In Uganda, the 2001/2002 administrative data showed that additional salaries were eliminated for auxiliary and technical staff [28]. The absence of this staff who were originally paid by user fees contributed to the human resources crisis [29]. In Uganda, this crisis was seen in the decline in health workers' morale and in their attitude toward their work [27,28]. Elsewhere, this crisis translated into a heavier workload experienced by health workers [30][31][32][33], a feeling of inadequacy among medical personnel [32], and a sense of being overworked and exploited in the face of increasing work demands [30]. Negative effects on practices were reported in South Africa [30]. The loss of income for personnel caused by eliminating user fees could lead doctors to devote themselves more to private practice, especially if the problem of discontinuity in drug supplies were to persist [28].

Health information system
Only one study, in Ghana, reported that little information was available at the central level on the number and type of deliveries carried out in health centres and on the amount of reimbursements [31].

Medical products, vaccines and technologies
The experiences indicated an overall shortage of drugs, highlighted in South Africa [30], Kenya [33], and Madagascar [19]. In Uganda, the stock forms showed an increase in the quantity of drugs received after fees were abolished, whereas the actors spoke of problems of availability, particularly for antimalarials, and delivery delays caused by administrative red tape at the district leve [27,29,34]. Patients responded to this shortage by turning to private services to purchase these drugs [27,29]. Some authors reported that health workers perceived the drastic decline in service utilization after seven months of exemption to be a consequence of the shortages in additional drug stocks [28].

Health systems financing
The major problem is the underfunding of the exemption policies. In Ghana, the funds paid to the districts at the start of the fiscal year were not enough to cover the year's expenses [31,35]. The policy was underfunded by 34% in 2004 and 73% in 2005. In Uganda, despite an increase in the budgets allocated, the health centres lost income [20,25] and were unable to cover recurring expenses [27]. In Ghana and Kenya, some health centres had resumed charging for services and drugs in order to deal with funding shortfalls [31,33]. Governments sometimes had to gradually reintroduce user fees after they had been abolished [36,37]. In Senegal, health centres increased the fees for certain acts that could still be charged in order to compensate for lost income from those acts that had become free [23].

Leadership and governance
Most of the countries appeared to have provided clear orientations regarding which services or which populations were the targets of exemption. However, these orientations did not always appear to have been expressed in clear directives, as was seen in South Africa [30]. Ghana [31,32], Kenya [33], and Senegal [23], where it seemed that various actors in the system had been inadequately informed. As well, the exemption policies appeared sometimes to interfere with other health policies and programs, such as community-based health insurance [31]. In Ghana, the difficulty of obtaining accounting reports from managers was an important element for actors in that system [35]. However, the complexity of the reimbursement process and the multiplicity of actors were also factors that impeded a real transparency in the process and made it difficult to assign clear responsibility to actors at different levels in the system ('no blame' game).
The various pressures mentioned in the literature are described in Table 3 by country, health system functions, and authors.

Discussion
Understanding health systems as complex collections of components that interact in an effort to achieve their objectives offers an opportunity to see the effects of health policies through a new lens [42]. As has been explained by The Alliance for Health Policy and Systems Research (AHPSR) [43], every system-wide intervention, such as user fee exemption policies, has an effect on the health system's different components. This new literature review showed that, in responding to population needs that had up to then not been addressed, particularly because of an unfair funding model, the exemption policies generally created an increase in health services utilization. However, at the same time, this increase produced bottlenecks at different points in the system (Table 4) as was confirmed in a recent supplement of Health Policy and Planning on user fee removal in the health sector in low-income countries [44]. That being said, increased utilization was not the only source of pressure on the system. There was a constellation of variables that acted, at the same time, as both causes and consequences of the policies, particularly Table 1 The six building blocks of a health system Health system functions Definition (1) Service provision Good health services are those which deliver effective, safe, quality personal and non-personal health interventions to those who need them, when and where needed, with minimum waste of resources.
(2) Health personnel A well-performing health workforce is one which works in ways that are responsive, fair and efficient to achieve the best health outcomes possible, given available resources and circumstances.
(3) Health information A well-functioning health information system is one that ensures the production, analysis, dissemination and use of reliable and timely information on health determinants, health systems performance and health status.
(4) Drugs and vaccines A well-functioning health system ensures equitable access to essential medical products, vaccines and technologies of assured quality, safety, efficacy and cost-effectiveness, and their scientifically sound and cost-effective use.
(5) Funding A good health financing system raises adequate funds for health, in ways that ensure people can use needed services, and are protected from financial catastrophe or impoverishment associated with having to pay for them.

(6) Governance and leadership
Leadership and governance involves ensuring strategic policy frameworks exist and are combined with effective oversight, coalition-building, the provision of appropriate regulations and incentives, attention to system-design, and accountability.
planning and attendant measures that often proved to be limited [8,44]. There is a risk that, by disrupting the system's functioning, these interventions will produce the opposite effect from what was intended, that is, a return to low utilization of services. However, given the recent nature of these policies, the timeframe adopted in this review, and the possibility of publication bias, which exposes problems (although often implicitly) more often than successes, we do not yet have enough distance or any evidence that these reverse effects-such as a return to low utilization-actually exist.
The AHPSR's observation in 2009 is still relevant with respect to user fee exemption policies: ". . .systemic factors and their impacts have hardly been studied" [43]. The articles included in the present literature review were not aimed at understanding the systemic effect of these policies, but generally focused on a single function of the health system. Two-thirds of the articles addressed only one or two system functions. This perspective adopted by the authors also limits the lessons we can draw from national experiences to support decisionmakers involved in developing and implementing such policies. Only a few of the authors attempted to put the results of their studies into perspective in the context of health systems [20,27]. The major lesson to be drawn from this scoping study is thus the need for empirical studies aimed at understanding more clearly, on a nation-wide or district scale, how these policies affect, at one and the same time, all the functions of the health system. Now that the positive effects of user fee exemptions on utilization have largely been demonstrated, researchers need to direct their attention toward more systemic questions. This approach fits within the movement that promotes developing and implementing reforms while taking into account the different potential effects on health systems [45].
Addressing these more systemic questions, by looking particularly at implementation, will likely call for new forms of collaboration among researchers, political decisionmakers and practitioners [46]. In doing this survey, it was not our intent to report all the available knowledge on these interventions or to discuss the Table 3 Pressures mentioned in the literature by country, health system functions, and authors (1988 to 2009)

Medical products, vaccines and technologies
• No information on the number of acts and the amount of reimbursements [31] Financing • Funding unpredictable, insufficient and discontinuous; problems with reimbursement in cases of referrals [31] • Funding unpredictable, insufficient and discontinuous; informal payments; health centres going into deeper debt [35] Leadership and governance • Lack of information and complexity of funding procedures; poor supervision; problems in assigning responsibilities [31] • Lack of information and communication (funding); competition with other interventions; poor supervision; 'no blame' game [35] Kenya Service delivery • Increase in demand for service [36,37] Health workforce • Increase in workload [33] Health information

• No information
Medical products • Problems of availability and insufficiency of drugs and kits [33] Financing • Informal payments [22] • Informal payments; loss of income for health centres [33] • Insufficient funding; informal payments [36] Leadership • Poor understanding of the policy; problems in assigning responsibilities [33] Madagascar Service delivery • Increase in utilization [19] Health workforce • No information

Health information • No information
Medical products • Problems of availability of drugs [19] Financing

• No information
Medical products • Problems of availability of drugs [29] Financing

• No information
Medical products • Problems of availability of drugs [24,26,27,29,34] Financing • Insufficient funding [25,34] • Difficulties in meeting recurrent expenses; informal payments [27] Leadership • Interference with other types of interventions [27] methods used, as this has been done by others already. Our primary aim was to bring to light the elements related to exemption policies that were disruptive to the system and to highlight certain areas where there was a lack of knowledge. The main limitation of our study, aside from its being more narrative than explicative, is therefore that we have not highlighted the successes of these policies, except for the increase in service utilization. Moreover, the elements reported here are only potential disruptions and do not necessarily reflect what is actually happening across all systems as a whole. For this reason, we repeat our call for further empirical studies on this matter. Nevertheless, these disruptive elements may also be seen as demonstrations of what might occur if decision-makers, in trying to implement exemption policies, do not ensure the necessary conditions of preparation, planning and attendant measures from a systemic perspective, as is shown in the new empirical studies published in the recent supplement of Health Policy and Planning [44].

Health information
Lack of information on the number and type of services carried out in the health centres and on the amount of reimbursements.

Drugs and vaccines
Problems of availability of drugs