Findings

The case study hospitals did not have an official organogram. However, observations and discussions with hospital managers and staff identified the existence of a management structure which was highly hierarchical (Figure 2). At the lowest level were frontline healthcare workers (such as pharmacists, medical doctors, and nurses) and non-health staff (such as accountants and maintenance personnel), all of whom were answerable to the heads of their respective departments. These heads of departments were middle level managers for clinical departments (e.g. paediatrics, obstetrics and gynaecology), wards (e.g. adult male, adult female and paediatrics), non-clinical departments (e.g. pharmacy and laboratory) and support departments (e.g. accounts and maintenance) who were themselves answerable to the three senior hospital managers namely the medical superintendent, the hospital administrator and the hospital nursing officer in-charge. The medical superintendent was the chief executive of the hospital and was responsible for the overall running of the hospital. The hospital nursing officer in-charge was in charge of the nursing department and hence all nursing wards in charges. The hospital administrative officer was in charge of all the hospital non-clinical departments. The case study hospitals had 3 management and decision-making committees. First, there was a hospital management team (HMT), comprised of all hospital departmental managers (middle level managers) and senior managers. Second, there was an executive expenditure committee (EEC), comprised of only the senior managers, and third, there was the hospital management committee (HMC) which was an oversight committee that drew its membership from the local resident community. The hospital was represented in the HMC by the medical superintendent, who was also its secretary, and the hospital administrative officer.

Hospital Organogram

The budgeting and planning process was comprised of two distinct activities; quarterly budgeting and the annual work planning (AWP) process. The development of the hospital budget and the AWP were designed to be linked and aligned. At the beginning of each government fiscal year (July 1), hospitals were required to develop and submit AWPs to the central Ministry of Health (MOH) for approval. Hospitals were then required to develop quarterly budgets that outlined the allocation of available resources to the priorities indicated in the AWPs. Hospital AWPs were developed by the HMT and submitted to the regional office for onward transmission to the central Ministry of Health (MOH) for approval. While the range of services provided by hospitals was guided by KEPH, hospital managers had autonomy to allocate available resources across service areas (i.e. prioritize across these services). The budgeting process should begin at the hospital department level, where departmental managers develop a list of departmental needs and present these to the HMT. The HMT then deliberates on the departmental needs and develop budgets that allocate available cash budgets across hospital departments. These budgets should then be deliberated upon and finalized by the EEC and subsequently presented to the HMC for review and approval. Budgets approved by the HMC should then be submitted to the regional level and from there submitted to the MOH for approval.

While the budgeting and planning process was expected to be linked and aligned, in practice, this was not the finding in both case study hospitals. The AWP was developed almost one quarter in the planning year, while the budgets were developed on time at the beginning of every quarter. This meant that the first budget of the year was often developed without the existence and hence any reference to the AWP. Subsequent budgets were also developed without reference to the AWP. The result was that activities budgeted for in the quarterly budgets were dissimilar to activities planned and budgeted for in the AWP. As a result of this non-alignment, hospital managers placed little importance to the AWP process. Very few managers knew what was contained in the AWP, very few participated in the process, and hardly any cared about implementing the AWP.

People just fill the [AWP] template very fast but they don’t even know what they are putting in the plans. If you ask people ‘okay you did the AWP some three months ago do you remember what you did?’ Most of the people don’t have an idea. They’ll tell you ‘we did it and it has already been sent to the province. We finished that business. Middle level manager, Hospital A

Formal and informal criteria were used to allocate budgets. Formal criteria are objective criteria that were used explicitly by hospital decision makers to determine how the hospital budget was allocated across departments and/or services. Informal criteria refer to subjective considerations, which were often implicitly employed, that influenced budget allocation decisions in hospitals. To get an idea of the prominence of criteria used in the case study hospitals, we developed a word cloud by identifying decision-making criteria mentioned in interview transcripts and the number of times they were mentioned (Figure 3). The criteria identified will be discussed next.

Word cloud of priority-setting criteria in the case study hospitals

In both case study hospitals, the dominant criterion used to allocate budgets to hospital departments and services was the revenue generating potential of the departments. Departments or services that generated more revenue from user fee collections were prioritized over departments that generated less revenue and subsequently received a larger share of the hospital budget. The reason given for using the revenue generating potential of departments is that the hospitals experienced a severe scarcity of resources and relied on user fee collection to finance their daily operations (Barasa et al. 2016). To make sure that the hospital continued to run, resources had to be allocated in a manner that assured further generation of revenues:

The hospital generates very little money which means priorities have to change…So first we want to make money, we allocate where we can make money… Middle level manager, Hospital B

Historical budgeting also featured prominently among the criteria used by managers to allocate budgets across departments in both hospitals. Departments often received the same budgetary allocation or increments to previous year’s budgets. The lack of technical competence in budgeting and planning, and lack of priority-setting guidelines, together with resource scarcity also contributed to the use of historical budgeting (Barasa et al. in press). Managers also considered the extent of necessity of a service in making budgetary allocation decisions. Services were considered essential if the hospital could not run without them. The perceived medical need in the hospital’s catchment area was also a determinant of hospital allocations. The need was however based on the volume of patients seeking different services at the hospital rather than any formally assessed need in the community. Other formal criteria used included international and national priorities such as the Millennium Development Goals, the feasibility of implementing the service, and affordability of proposed services.

In contrast to the formal criteria identified above, managers in Hospital A felt that allocative decisions were influenced by informal criteria such as the lobbying and bargaining ability of departmental managers.

You see you can have a head of department who is not very vocal and does not articulate your needs as well as they should…some departments…they seem to always get more than others…it all depends on how eloquent and convincing the head of department presents his proposals. Middle level manager, Hospital A

Resource allocation was also dependent on interpersonal relationships and mutual benefit between the middle-level managers and the senior managers.

Allocations depend on your relationship with the hospital administrators…we mean in life sometimes things work because of relationships right? You are a friend of mine and we get along well so we will allocate something to you. Middle level manager, Hospital A

Middle level managers at Hospital A also felt that allocations favoured the senior managers who were part of the EEC. The use of these informal criteria was made possible in Hospital A because there was little deliberative space in the budgeting process. Given that actual allocation decisions were made by a small group of senior managers (EEC), this provided an opportunity for the EEC managers to leverage on their unique position to favour their departments and the departments of those with whom they enjoyed good relationships.

The situation was different in Hospital B where the middle level managers, through the HMT, were empowered to make allocation decisions. While managers in this hospital also felt that the bargaining and lobbying ability of managers had an influence, the general feeling was that favouritism did not influence decisions. The result was that while in Hospital A managers generally felt that the allocation decisions were unfair, in Hospital B the feeling was that allocations were relatively fair.

We don’t get all that we need but we can say that the budgeting is fair. The medical superintendent ensures there is equity. At least each department gets something small. Middle level manager, Hospital B

In this section, we use the framework that we previously developed (Barasa et al. 2015a) to evaluate the budgeting and planning process in the case study hospitals. We first present our findings on the use of consequentialist principles followed by the adherence to proceduralist conditions.

Hospital managers were unfamiliar with mechanisms such as cost-effectiveness analysis (CEA) and program budgeting and marginal analysis (PBMA). When the basics and rationales of these methods were explained to them, they responded that although the methods were potentially useful in decision-making, they lacked the technical skills and data required. However, in both hospitals, budgeting and planning decisions considered the affordability of competing alternatives. This could be argued to be an attempt to incorporate efficiency, given the capacity and data constraints that the hospitals faced. By taking into account the costs and affordability of competing priorities, managers were recognizing budget limitations and the need to make decisions such that the hospital could get the most out of available resources.

In both case study hospitals, the dominance of revenue maximization as a priority-setting criterion meant that departments (and hence patient groups such as children under 5 years) that did not generate user fee revenues were systematically underfunded compared to departments that generated user fee revenues. This practice meant that budget allocations were inequitable. Further, the reported favouritism in resource allocation given to departments headed by senior managers and those whose managers enjoyed good relationships with senior management could also be considered as sources of inequity.

The level of satisfaction with the budgeting and planning process varied between hospitals. In Hospital A, stakeholders (senior and middle level managers, and frontline practitioners) were not satisfied with the budgeting and planning process because the process was generally not inclusive, leaving most stakeholders disgruntled. Further, the scarcity of resources meant that hospital managers were not satisfied with the resources that were allocated to them. The use of revenue generation criterion also left the managers whose departments generated little revenue disgruntled. In Hospital B, the stakeholders reported having some level of satisfaction with the budgeting and planning process. While they were unhappy with the limited availability of resources, they seemed to understand the scarcity situation. It appeared that this general satisfaction with the process was due to the fact that they were included in the budgeting and planning process. However, managers of departments with low revenue generating potential, like in Hospital A, were unhappy with the process.

The level of understanding varied across stakeholders and was related to their level of engagement. For example, while in Hospital A the middle level managers had a low level of understanding of the budgeting process given that they were excluded from it, in Hospital B, the middle level managers reported adequate understanding of the process because they were involved in it.

In both case study hospitals, budgeting and planning processes did not result in shifted resources. This was because budgeting and planning in these hospitals was significantly guided by historical allocations. The budgeting and planning process was therefore not responsive to the changing dynamics of resource needs.

The implementation of budgeting and planning decisions was fairly similar between the case study hospitals. The planning processes in both hospitals were considered to be mainly an activity on paper that was hardly implemented in practice. A number of reasons, which we have reported elsewhere, led to the lack of implementation of decisions including the lack of resources, reduced motivation due to reduced autonomy of hospital managers over planning decisions, a culture where hospital staff lacked a sense of duty and commitment to their roles and responsibilities, and the lack of strong internal accountability mechanisms (Barasa et al. in press).

The degree of stakeholder engagement varied across the case study hospitals, with the budgeting and planning process being more inclusive in Hospital B, compared to Hospital A. While hospital budgets were discussed by the HMT in Hospital A, final budgeting decisions were made by the EEC. Given that the EEC was a smaller committee that comprised of senior managers only, middle level managers felt excluded from the budgeting process. In Hospital B however, as mentioned above, final budgeting decisions were made by the HMT which was a larger committee that comprised of both senior and middle level managers. The HMT meetings also allowed for greater deliberation and discussion.

We present budgets and people are asked to say why they need the money. At least we get to understand why a department’s budget is like this or like that. People also see why for example they are going to get less than what they asked for….because we also discuss what [resources] is available and how much departments can get. Middle level manager, Hospital B

In both hospitals, however, frontline clinicians rarely participated in budgeting and planning processes. While it was reported that they were not invited in Hospital A, frontline clinicians did not participate in Hospital B despite being invited. As we have discussed elsewhere, it appeared that the main reason for non-participation of clinicians was professional identity (Barasa et al. in press). Clinicians in both hospitals did not seem to think that managerial responsibilities such as budgeting and planning were part of their roles as professionals. They identified themselves more with their clinical roles and considered time spent doing managerial duties as ‘wasted time’ (Barasa, et al. in press). The shortage of clinical staff also contributed to the non-participation of clinicians in budgeting and planning meetings (Barasa et al. in press). As will be discussed below, community members were involved only very peripherally in the budgeting and planning processes in both case study hospitals.

The level of empowerment of different stakeholders varied between the case study hospitals. In Hospital A, middle level managers appeared to have a low level of empowerment to participate in budgeting and planning activities compared to Hospital B.

Decision making is not democratic. I think it’s dictatorial because at the end of the day whatever decisions are made at HMT meetings, we’re still going to hear of another meeting that was held with another committee and basically whatever we had come up with will not even be considered. Middle level manager, Hospital A

Further, actors who were not engaged in the priority-setting process (clinicians and the community) were clearly not empowered to contribute to decision making either.

The extent to which the budgeting and planning process was transparent varied between the case study hospitals. Generally, Hospital B exhibited more transparency. In Hospital A, there was no mechanism in place for disseminating budgeting and planning decisions, and once the final budgets and AWPs had been prepared, they were not shared with the hospital managers. Only selected senior managers had access to these documents, and for both processes, the reasons for decisions were not communicated to the managers. Front line practitioners also reported that they were in the dark as far as budgeting and planning decisions in the hospital were concerned. In Hospital B, a more inclusive budgeting and planning process meant that managers were generally more aware of the budgeting and planning decisions and the rationales behind them. They therefore reported that the process was transparent. Nevertheless, as with Hospital A, they reported that final budgets and work plans were not made available to them unless they individually sought them out.

In both case study hospitals, decisions were rarely made based on information/evidence. Information was gathered using formal channels such as the hospital management information system but ignored. Decision makers often used their gut feeling and hearsay as the basis for decision-making. When information was used, the use was more symbolic rather than functional. Decisions were first made and then information was sought to justify the decisions. One of the reasons given for the low use of information was that the quality of information available was questionable. Managers reported that data captured in clinic registers often had gaps and did not capture all events. They also complained that the data captured in clinic registers were inaccurate.

In both case hospitals the budgeting and planning process did not have a provision for a formal appeals and revision process. Once the quarterly budget or the AWPs had been prepared and approved, they could not be changed or altered over the course of the planning period. This meant that the decision-making process was inflexible and could not be improved with emerging information. It also meant that there was no formal avenue for parties to contest planning and budgeting decisions.

In both case hospitals, community views were obtained through two mechanisms namely the suggestion box and community representatives in the HMC. Both mechanisms were however felt to be ineffective as mechanisms for channelling community views. In both case study hospitals, the suggestion box was hardly ever opened by the hospital administration. The incorporation of community representatives in the HMC was also shown to be an ineffective mechanism for obtaining community values in both hospitals. This mechanism was shown to have two main shortcomings. First, the method of appointing community representatives into the committee was not thought to be transparent and inclusive. Senior hospital managers were perceived to influence the selection process to appoint preferred individuals, who were then thought to simply ‘rubber stamp’ hospital decisions. The community representatives in this committee were therefore not empowered to ask questions and contribute to decision-making.