Doctors as street-level bureaucrats in a rural hospital in South Africa

Introduction: In the perspectives of implementation of policy, the top-down and bottom-up perspectives of policy-making dominate the discourse. However, service delivery and therefore the experience of the policy by the citizen ultimately depend on the civil servant at the front line to implement the policy. Lipsky named this street-level bureaucracy, which has been used to understand professionals working in the public sector throughout the world. The public sector in South Africa has undergone a number of changes in the transition to a democratic state, post 1994. This needs to be understood in public administration developments throughout the world. At the time of the study, the public sector was characterized by considerable inefficiencies and system failures as well as inequitable distribution of resources. The context of the study was a rural hospital serving a population of approximately 150 000. Methods: An insider-ethnography over a period of 13 months explored the challenges of being a professional within the public sector in a rural hospital in South Africa. Data collection included participant observation, field notes of events and meetings, and documentation review supplemented with in-depth interviews of doctors working at a rural hospital. Street-level bureaucracy was used as a framework to understand the challenges of being a professional and civil servant in the public sector. Results: The context of a resource-constrained setting was seen as a major limitation to delivering a quality service. Yet considerable evidence pointed to doctors (both individually and collectively) being active in managing the services in the context and aiming to achieve optimal health service coverage for the population. In the daily routine of the work, doctors often advocated for patients and went beyond the narrow definitions of the guidelines. They compensated for failing systems, beyond a local interpretation of policy. However, doctors also at times used their discretion negatively, to avoid work or to contribute to the inefficiencies of healthcare delivery.


Introduction:
In the perspectives of implementation of policy, the top-down and bottom-up perspectives of policy-making dominate the discourse.However, service delivery and therefore the experience of the policy by the citizen ultimately depend on the civil servant at the front line to implement the policy.Lipsky named this street-level bureaucracy, which has been used to understand professionals working in the public sector throughout the world.The public sector in South Africa has undergone a number of changes in the transition to a democratic state, post 1994.This needs to be understood in public administration developments throughout the world.At the time of the study, the public sector was characterized by considerable inefficiencies and system failures as well as inequitable distribution of resources.The context of the study was a rural hospital serving a population of approximately 150 000.Methods: An insider-ethnography over a period of 13 months explored the challenges of being a professional within the public sector in a rural hospital in South Africa.Data collection included participant observation, field notes of events and meetings, and documentation review supplemented with in-depth interviews of doctors working at a rural hospital.Street-level bureaucracy was used as a framework to understand the challenges of being a professional and civil servant in the public sector.Results: The context of a resource-constrained setting was seen as a major limitation to delivering a quality service.Yet considerable evidence pointed to doctors (both individually and collectively) being active in managing the services in the context and aiming to achieve optimal health service coverage for the population.In the daily routine of the work, doctors often advocated for patients and went beyond the narrow definitions of the guidelines.They compensated for failing systems, beyond a local interpretation of policy.However, doctors also at times used their discretion negatively, to avoid work or to contribute to the inefficiencies of healthcare delivery.

Introduction
Internationally, in the implementation of public policy, the top-down and bottom-up perspectives dominate how implementation is conceptualized 1 .In the top-down perspective, policy is made at 'higher' levels (such as provincial or national levels) and has to be implemented by the public service employees on the ground.The alternative bottom-up perspective focuses on the needs in the local context as the point of departure for the content of the policy and foregrounds participatory approaches.Yet Michael Lipsky 2 argued that in either approach, the actual implementation of policy in the local public institutions depends on the civil servants at the front line aligning themselves with the guidelines and policy briefs.This phenomenon Lipsky called 'street-level bureaucracy'.
Lipsky 1 further argued that civil servants at the front line face a dilemma in the process of implementing policy in the local context.He described a number of key drivers behind this dilemma, which include that policy usually does not adequately address the complexity of the situation that the officials face or the complexity of the decisions to be made.Furthermore, the policy itself may be too complex to be implementable and the context in which the officials work is in comparative or absolute terms resource poor (particularly as many policies are aspirational).Indeed, a more recent iteration of this dilemma has been to focus on the public service gap as driving discretion 3 .Many local and personal dynamics such as personal relationships and values of the individual civil servant, the local context and cultural setting or organizational culture of the public sector all influence how the dilemma is experienced.
Lipsky proposed that a key strategy that street-level bureaucrats employ to manage the dilemma is to use their discretion regarding how the policy is actually implemented in individual situations.And by using discretion regarding how the policy is being implemented, the street-level bureaucrats make policy -in the sense that they determine how the citizens experience the public service 2 .
The concept of street level bureaucracy has attracted considerable interest throughout the world in the discourses around public administration and more notably in some disciplines, such as social work and policing 4 .It also has problematized street-level bureaucracy in the public sector of different countries (see, for instance, the discussion regarding the UK or Netherlands 5,6 ) or the changes in the international policy environment of public administration (such as the New Public Management 7 ).The role of the civil servant at the front line has remained the focus of the growing body of work on street-level bureaucracy [8][9][10] .
Examples of exploring street-level bureaucracy in the healthcare setting include maternal services in the UK 11 and mental health services in the Netherlands 12 .While most of the published work is from the developed countries, examples from further afield include a description of nurses as street-level bureaucrats in HIV care 13 and the role of community care-givers in Kenya 14 .Despite the variations of local context, the attention to how policies and guidelines are implemented locally is a much more universal concern and seems to have some profoundly universal characteristics 9 .
The idea of discretion as part of street-level bureaucracy 2 offers a particularly interesting angle on how we can explore the work of doctors at the frontline of the public healthcare sector and how frontline work determines what kind of service citizens experience 8 .In many senses, public healthcare systems throughout the world share the kind of local context described above.The healthcare sector is awash with guidelines and algorithms to assist with decision-making covering both clinical and organizational arrangements.Yet in the daily routine of work, many decisions are characterized by high degrees of discretion, such as admission to hospital (often balancing the patient's condition with the availability of hospital beds) or when a patient is deemed to be well enough to be discharged.Yet despite the public sector being the context for the work of many doctors worldwide (and, indeed, it plays an important role in healthcare delivery in many countries of the world), the doctor functioning as a civil servant has remained largely unexamined.This has particular relevance for practitioners in peripheral settings where the local context -and particularly access to higher level services -plays a central role in determining how policies are interpreted and implemented for healthcare delivery for rural communities.
The focus of this article is how doctors functioned as bureaucrats, ie working within the rule-bound public service as civil servants.

Location and context
The public sector in South Africa has undergone major changes following the transition to a democratic state in 1994.The developments have been part of a deliberate process aiming to transform service delivery to meet the needs of the majority of the population 15 .This required an internal realignment of policies while at the same time a massive expansion of service delivery, to address the needs of the previously underserviced majority of the population.The public sector has also been seen as a tool to address aspects of societal transformation -and particularly in the higher echelons a number of overtly political appointments were commonplace 16 .
The right to health care is constitutionally mandated in South Africa and has a well-developed legislative and policy environment.Numerous guidelines produced by the National Department of Health, such as standard treatment guidelines for different levels of care 17,18 , determine not only clinical care but also to a large degree which medications are available at the different levels of care.
Yet at the time of the study, the public healthcare sector has been characterized as a system under stress and struggling to meet its constitutional imperative 19 .The quadruple burden of disease of the country 20 further stretches the already limited resources of the public healthcare sector.Poor management, poor accountability and poor service delivery were key findings of a critical review of the healthcare system 21 .Rural healthcare services are particularly under-resourced; while an estimated 43% of the population of South Africa lives in rural areas, only 12% of doctors and 19% of nurses work in rural areas 22 .Such inequities in terms of resource allocation are not unique to South Africa.
The study was located in a public sector hospital in a rural area in KwaZulu-Natal province in South Africa.It is a 160bed hospital providing district-level care to a population of approximately 150 000 people.At the time of the study, 12 doctors were working at the hospital who were responsible for both the primary care service in clinics within a 70 km radius and the district-hospital care.The population is generally poor (the district consistently has been in the most deprived quintiles in the country) 23 , characterized by high levels of poverty and unemployment rates.The majority of the population is isiZulu-speaking, and live in two small villages, extensive areas of subsistence farming homesteads or as farm workers on commercial farming enterprises.

Methods
The work presented is part of a larger doctoral study 24 that used qualitative methods to explore the experiences of a professional working in the public sector in a setting of a rural district hospital in KwaZulu-Natal.As part of an insiderethnography 25 , a range of data collection methods were used, including an extended period of participant-observation, field notes covering group discussions and meetings as well as key events.The focus of the participant-observation was the everyday routine and occurrences, rather than the exceptional or isolated events.Semistructured, in-depth interviews were conducted with all participating doctors working at the hospital.The data collection was further supported with the review of documentation, included policy documents, guidelines, minutes of meetings and both scientific and grey literature 26 .
An inductive process was used to analyse the data, reflecting on the professional/civil servant intersection through the lens of Lipsky's framework of street-level bureaucracy.Through an iterative reflective process moving between the data and texts and focusing on key thematic areas, the narrative of the research emerged 26 .

Ethics approval
The proposal for the research was reviewed and approved by the Research Ethics Review board of the University of Pretoria (#7/2008).Participation in the study was voluntary and out of the 14 doctors working at the hospital over the period of data collection, two doctors declined to participate.All participating doctors signed an informed consent form.In the reporting, as also in this article, identifying characteristics of the doctors were removed as far as possible and the participating doctors were randomly assigned initials to anonymize them.

Results
The healthcare professionals, including doctors, functioned between the intentions of the policies and the reality that they face in daily work.In this space, they confronted the disjuncture between the many positive aspirations articulated in the policy and the harsh limitations of local services that did not seem to be designed to address the policy imperatives.'We always need more equipment, more time, more staff' (TC interview).The feeling of resource limitation also added to the perception that, 'there will always be a conflict' (FW interview) between the role of being a doctor and a civil servant.Working in a rural hospital, rather than in better resourced settings, ie 'being out there' (FT interview), meant that doctors had to deal with situations where investigations such as CT scans or blood tests were not available and 'one had to make do with less' (WT interview).And, therefore, 'When you come to work for the government service, you forget about many things.You say, let me put this aside' (FE interview).
Yet, despite the resource limitation and the difficulty, there was abundant evidence in the data of how doctors as a group responded to the imperative of service delivery within the available resources.This included the allocation of staff to different services such as visiting the primary healthcare clinics or working in the operating theatre.Critically, the group of doctors themselves took initiative in most cases to arrange services in a way to optimize coverage of services to the population.
The doctors met every Friday morning to plan the services for the following week and had a handover meeting every morning, where any changes to the plans for the day were discussed.A range of documentation and tools was developed (such as a template of services that were being offered to allocate doctors each week) to structure the planning processes and to serve as documentation regarding the allocations.As services changed, the planning templates were adjusted to accommodate the changes.Checklists and treatment guidelines regulate individual activities and in many instances (such as when a certain drug was not in stock), through a discussion among the doctors, an alternative would be agreed upon.The local interpretation of the policies offered opportunities for finding innovative solutions to often complex problems.
At the hospital, doctors exerted considerable control over the practicalities of their work, both collectively and individually.The implementation of policy relied to a significant degree on the discretion of the doctors themselves to interpret the policy in the context of resource limitation as well as local complexity of the context.It required discretion on the part of the street-level doctor to do the best for the patient, which lies at the heart of how services could be improved.
In the detail of the work of the doctors, considerable variability remained in how individual doctors worked.Beyond the service planning, doctors would also make their own arrangements regarding how ward rounds would be conducted, the pace of their work and how they managed the interactions with colleagues or patients.
During the ward round in the female ward, Dr WT presented a patient he had admitted the previous day.The patient was post-menopausal and had previously been admitted to the ward with chronic vaginal bleeding.During the prior admission at the district hospital she had had an ultrasound that found a mass in the lower uterus.It was not clear, however, whether it was a benign fibroid or a malignancy.After discharge from hospital she had been referred for a pap smear at the primary healthcare clinic.At the time of doing the smear she had been asked to return to clinic after 4 weeks to review the results with the visiting doctor.At the scheduled follow-up visit she was seen by the doctor.Her test results were not back, however, and the doctor wrote: 'Pap smear result not back -TCB [to come back] in 2 weeks,' in the notes.She returned two weeks later, and the subsequent doctor wrote the same note.
When Dr WT saw her at the PHC clinic at her subsequent visit some 2 months after the pap smear had been taken, the results had still not arrived.This time, instead of writing 'Pap-smear results not back.TCB (to come back) in 2 weeks', from the peripheral clinic he called the cytology laboratory in the regional centre to trace the pap smear number and in the process learned that that there was a backlog of about 5 months for reading routine pap smears.Owing to the high index of suspicion for a malignancy, he asked the laboratory staff to prioritize reading the smear and to give him feedback that same day, which they agreed to.He also traced the results of other blood investigations that were taken previously (that also had not been recorded in the notes) and decided to admit the patient so that 'things can be sorted out' and 'she will not fall through the cracks' (Field notes).
The data revealed that Dr WT understood the clinical implications of the investigations.Yet he was also able to translate the clinical information into required steps within the healthcare system to assist the patient.He had an intimate knowledge of the healthcare system, including the referral pathways and requirements, and the functioning of the different levels of care.He had understood where the bottlenecks were and how to negotiate a way around these in order to facilitate the patient's journey through the pathway to care.He had a sound understanding of the system in which he was working.These competencies are distinctly different from having the capacity of making a diagnosis, staging or deciding on treatment for cancer of the cervix from a clinical perspective.
Dr WT had clearly come across the, at times, very specific requirements that limit a referral (such as types of investigations required in order to facilitate the appointment at the oncologist) and had worked out an approach to how best to fast-track such arrangements.
Discretion was not only evident in acting for patients in the bureaucracy (like Dr WT) or choosing to increase efficiency and cooperation.It also had negative aspects, such as using discretion to leave work early, avoid work or taking responsibility.
Out-patient Department was full -all the chairs in the waiting area were occupied.When I arrived, I saw Dr FW and Dr TW standing in conversation at the entrance of one of the consulting rooms in front of the full waiting room.Dr FW and Dr TW were talking fairly loudly about where they were the previous weekend.There seemed to be no urgency to attend to the waiting patients.The patients sat waiting and the two doctors stood there in the corridor, talking.In the meantime, the two other doctors working called patients into their consulting rooms at regular intervals.The queue moved slowly.After observing this for some time, I walked past and asked how OPD was going, to which Dr FW replied 'Eish, it is really full ...', pointing to the waiting room.And then he added, 'I suppose we must get cracking', and called the next patient in the queue to his consulting room (Field notes).Doctors themselves were the bottleneck in the system, being the limiting factor to how quickly patients would be able to complete the visit at the outpatient department.This position affords doctors considerable freedom to pace themselves and to control their work, which in many instances is used to the doctors' benefit.A number of bureaucratic processes such as how the waiting room is managed control the access of patients to doctors.
The examples also illustrates that not all doctors view their responsibility in the same way or treat patients in the same manner.In a number of instances, the very same doctors described above engaged with patients in a humane and caring manner.
At Emmaus it is possible to escape the definitions that are thrown upon you and one does not have to behave in a way that is expected from civil servants.It is possible to be less officious and have a different relationship with the people around you. (CF interview) I am across the courtyard from OPD and on the side Dr FW is standing in a small circle of people.One of them is a young woman, dressed in the hospital-issue gown; the others are dressed neatly and seem to be relatives of the woman.The doctor is explaining something to the relatives and the woman.As I pass I hear an older woman asking the doctor 'And are you going to be here, when we come back with her?' To this the doctor replied, 'Yes, I will have everything ready then, just go to OPD and join the queue there, I will see you in OPD ...' I don't hear anything more, as I have passed.As I continue and turn the corner to theatre, I see first the young woman and then the older woman hugging the doctor and then turn with the rest of the relatives and walk to the ward.(Field notes) The engagement with patients and their family in this manner was not an uncommon sight and in many instances arose out of ongoing clinical care for the individual.It required the engagement (and therefore discretion) of the doctor for such a relationship to form and to be maintained over time.Such ongoing relationships were not a requirement in terms of policy regarding how the service was delivered within the healthcare system, and the bureaucratic processes within the hospital at the time did not facilitate such longer-term relationships.However, within the local context of the rural hospital, personal relationships and how these formed part of the way doctors made meaning of their work at the hospital seemed to have a large impact on how some people were being treated.The relationships referred to above were larger than that of a professional: they point to a more human contact and interaction not regulated by professional or bureaucratic rules of being a doctor in the public sector.

Discussion
Since the work of Lipsky, the public service has changed considerably throughout the world and the public sector context that the study was based in is significantly different from the more classical top-down rule-driven bureaucracy on which Lipsky 2 based his theorizing of the frontline policymaking.Yet discretion, as described by Lipsky, was evident in the role of being a professional as a street-level bureaucrat.As professionals (on individual and collective levels), the doctors used their discretion to ensure equitable service delivery within the resources available.They also championed individual patients and found quick and efficient pathways through the system.There was no inherent requirement of a doctor working in the public sector to do so.In fact many guidelines required adaptations -and at times significant modification -to be implementable in the local context.
Discretion was also evident in the personal engagement, the decision to negotiate with the family regarding the follow-up with the patient.Doctors would regularly advocate for patients, engage with the families, while at the same time abstract individuals into a 'queue' of 'patients' that needed to be 'sorted out'.And yet, doctors also regularly found ways to limit access of patients and avoid taking responsibility.
The merger of the professional requirements of evidencebased approaches and organizational systems of service delivery formed a near-unified bureaucracy.It is within this bureaucracy made up both of the profession and the public sector that the frontline healthcare worker operates.Similar mergers are evident in the professional bureaucracies in countries such as the UK, Australia and elsewhere.
While appearing to be in conflict, the merging of the roles of the health professional and the bureaucrat is required to be able to function effectively within the healthcare system.Being a doctor and being a civil servant are synergistic in daily work, and as a result it is difficult to neatly differentiate professional and civil servant roles in decision-making.The convergence of roles enables considerable agency, as is clearly demonstrated by a doctor's ability to leverage the system for patients, arrange their work and be clinically more efficient.Ignoring one or the other role limits the agency that a doctor in the public sector would have.The agency that derives from such discretion is also a source of considerable professional satisfaction.Doctors are motivated, because they feel that they are able to make a difference and effect change 27 .In the literature, this has also been linked to patient satisfaction 7 .
As part of this agency, in the day-to-day decisions it was evident how doctors acted as street-level bureaucrats, applying values and modifying tasks and rules.This finding challenges Lipsky's interpretation of discretion as a mostly negative by-product of the dilemma of implementing policy 28 .In a context as described above, services are only provided because doctors are able to find local mechanisms to align their practices with policies through discretion.From the descriptions, doctors used their discretion to not only implement policy, but to compensate for inefficiencies and failures (such as an overloaded cytology service) in how the system functioned.Such an engagement is a step beyond policy implementation as here doctors take initiative in a terrain where the policy is no longer functional.With many policies being designed primarily for an urban context, the compensation for system challenges described is commonly seen in rural practices.Responding to the local need (particularly with, at times, minimal supervision that enforces adherence to policy imperatives) increases the space around using discretion in day-to-day decisions.At the same time, as the vignette in the outpatient department illustrates, doctors sometimes used their discretion to maintain and perpetuate inefficiencies within the healthcare system, such as long waiting times in outpatient departments.
This resonates with the role of doctors in the public service throughout the world.An important area to explore further is the role of agency of doctors (and other healthcare professionals) as a critical aspect of public sector transformation.The kind of innovative advocacy that was evident in the example of Dr WT above, offers insights into how services can be improved.It points to the critical value of discretion for system-wide implementation.

Conclusions
Despite the specifics of the particular environment and transitions in the nature of the role of civil servants such as doctors, the concept of street-level bureaucrats remains useful as it correctly sees them as having specific agency.Working at the front line or street level, civil servants are agents who must actively interpret policies in the local context and engage within the public service in order to practise their profession and deliver services to people.In the tension between the top-down policy direction and the bottom-up pressure to be adequately included, street-level bureaucracy acts as an important terrain for improving the implementation of services and therefore advocacy and health system improvement.