Health sector reforms and human resources for health in Uganda and Bangladesh: mechanisms of effect

BioMed CentralHuman Resources for Health

ss
Open AcceResearchHealth sector reforms and human resources for health in Uganda and Bangladesh: mechanisms of effectFreddie Ssengooba†1, Syed Azizur Rahman†2, Charles Hongoro†3, Elizeus Rutebemberwa†1, Ahmed Mustafa†4, Tara Kielmann*†5 and Barbara McPake†5
Address: 1Health Policy, Planning & Management, Makerere University, Institute of Public Health, Republic of Uganda, 2Department of Public Health and Policy, Health Policy Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, United Kingdom of Great Britain and Northern Ireland, 3Health Systems Trust, 1st Floor Riverside Centre, Belmont & Main Road, Rondebosch, 7700, Republic of South Africa, 4Ministry of Health and Family Welfare, Dhaka, People's Republic of Bangladesh and 5Institute for International Health and Development, Queen Margaret University College, Corstorphine, EH12 8TS, United Kingdom of Great Britain and Northern Ireland

Email: Freddie Ssengooba - [email protected]; Syed Azizur Rahman - [email protected]; Charles Hongoro - [email protected]; Elizeus Rutebemberwa - [email protected]; Ahmed Mustafa - [email protected]; Tara Kielmann* - [email protected]; Barbara McPake - [email protected]

* Corresponding author †Equal contributors

AbstractBackground: Despite the expanding literature on how reforms may affect health workers and which reactions they mayprovoke, little research has been conducted on the mechanisms of effect through which health sector reforms eitherpromote or discourage health worker performance. This paper seeks to trace these mechanisms and examines thecontextual framework of reform objectives in Uganda and Bangladesh, and health workers' responses to the changes intheir working environments by taking a 'realistic evaluation' approach.

Methods: The study findings were generated by triangulating both qualitative and quantitative methods of data collectionand analysis among policy technocrats, health managers and groups of health providers. Quantitative surveys wereconducted with over 700 individual health workers in both Bangladesh and Uganda and supplemented with qualitativedata obtained from focus group discussions and key interviews with professional cadres, health managers and keyinstitutions involved in the design, implementation and evaluation of the reforms of interest.

Results: The reforms in both countries affected the workforce through various mechanisms. In Bangladesh, the effectsof the unification efforts resulted in a power struggle and general mistrust between the two former workforce tracts,family planning and health. However positive effects of the reforms were felt regarding the changes in payment schemes.Ugandan findings show how the workforce responded to a strong and rapidly implemented system of decentralisationwhere the power of new local authorities was influenced by resource constraints and nepotism in recruitment. On theother hand, closer ties to local authorities provided the opportunity to gain insight into the operational constraintsoriginating from higher levels that health staff were dealing with.

Conclusion: Findings from the study suggest that a) reform planners should use the proposed dynamic responses modelto help design reform objectives that encourage positive responses among health workers b) the role of context hasbeen underestimated and it is necessary to address broader systemic problems before initiating reform processes, c)reform programs need to incorporate active implementation research systems to learn the contextual dynamics andresponses as well as have inbuilt program capacity for corrective measures d) health workers are key stakeholders inany reform process and should participate at all stages and e) some effects of reforms on the health workforce operateindirectly through levels of satisfaction voiced by communities utilising the services.

Published: 1 February 2007

Human Resources for Health 2007, 5:3 doi:10.1186/1478-4491-5-3

Received: 27 October 2006Accepted: 1 February 2007

This article is available from: http://www.human-resources-health.com/content/5/1/3

© 2007 Ssengooba et al; licensee BioMed Central Ltd. This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/2.0), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.

Page 1 of 13(page number not for citation purposes)

http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Abstract&list_uids=17270042

http://www.human-resources-health.com/content/5/1/3

http://creativecommons.org/licenses/by/2.0

http://www.biomedcentral.com/

http://www.biomedcentral.com/info/about/charter/

Human Resources for Health 2007, 5:3 http://www.human-resources-health.com/content/5/1/3

BackgroundIn the last two decades, developing country governmentshave implemented a variety of reforms in the health sectoron the understanding that these reforms would create theright individual and organisational incentives for improv-ing health systems performance. However, reform initia-tives have not always considered human resource issuesthat are relevant to their success and have often failed toinclude the participation or perspectives of the healthworkforce in reform planning processes and decision-making.

A number of studies have considered the effects of reformson the health workforce [[1-5] and [6]] and highlight theimportance of human resources to the success of reformobjectives [7] as well as the complexity of human resourcemanagement in the context of reforms [8,9].

These studies have pointed out that human resourceissues need to be a primary consideration in reformdesign, suggesting that reforms can only be implementedsuccessfully where there is consensual participation on thepart of the workforce. Ngufor describes how health staff inCameroon perceived reforms as a punishment inflictedon the nation by the International Monetary Fund (IMF)and the World Bank and as a result developed a laissezfaire attitude to their work resulting in reduction of con-sultation times and absenteeism [4]. Similarly, workers inZimbabwe were reported to perceive reforms as threaten-ing their job security, salaries and training and expressedtheir demotivation in the form of unethical behaviourwith their patients and neglect of work responsibilities[6]. In other parts of the world, health workers haveresisted change on the grounds of conflicting values. InLatin America, for example, reforms were perceived asaiming to undermine the fundamental values that hadinspired the design of the system, and sparked off a waveof resistance and strikes in El Salvador and Mexico. Thisled to the stalling and delay of the reform process [10,11].

Other studies have noted higher motivation levels amongthe health workforce through reforms. In Kazakhstan,reforms that aimed at changing the old Soviet systemthrough the introduction of more market-oriented financ-ing and service delivery, were argued to have resulted inincreased interest in primary care among physicians,increased attention to quality and patient satisfaction,more rational and creative use of resources, and strongercommitment of physicians' personal time and resourcesto improve services for patients [12]

Despite the expanding literature on how reforms mayaffect health workers and which reactions they may pro-voke, little research has been conducted on the mecha-nisms through which health sector reforms either

promote or discourage health worker performance. Howdo reforms affect the health workforce and createresponses that are likely to encourage the success or failureof reform objectives? How does context influence theroutes through which reforms affect provider incentiveenvironments and eventually motivation and perform-ance? To address these questions, the following paperseeks to trace these mechanisms of effect and examineshealth workers' responses to the changes in their workingenvironments by taking a 'realistic evaluation' approach[13].

This approach takes account of the explanatory mecha-nisms and the context of health systems reforms. Thisstands in contrast to the more common evaluationapproach, which tends to focus on a reform programme'smeasures and its intended effects and then seeks to meas-ure the differences between reforming and non-reformingentities, within the dimensions of those intended effects.The research described above suggests that reform pro-grammes cause a multitude of workforce responses whichact as the lynchpin between formal arrangements at theoutset of the reforms and the resulting changes in the sys-tem as experienced by people who use it. Figure 1, origi-nally created by McPake [14], captures the essence of thescenario described above and provides a framework forhealth systems research.

The focus of enquiry lies within the three componentsoutlined above. While the de-jure system sets the contextand defines the organisational structures, the intendedincentives and management procedures, the dynamicresponses reflect how those implementing the de-jure sys-

A conceptual model for health systems research – the Dynamic Responses Model [14]Figure 1A conceptual model for health systems research – the Dynamic Responses Model [14].

Dynamic responses: Informal structurese.g. community groupsnetworks of friendsInformal behaviourse.g. (offering or demanding)under-the-counter paymentprofessionalismInformal relationshipspe.g. clientelism political alliance

De-facto system: Services as experienced bypeoplee.g. access; quality

De-jure system:Organisational structures Intended incentivesManagement proceduresTraining courses



tem respond and create 'informal' arrangements or behav-iours that may or may not encourage the success of'formal' reform objectives initiated by the de-jure system.The de-facto system then encompasses the lived experi-ence of the results of the relationship between the de-juresystem and the dynamic responses. Similar to realisticevaluation, the model emphasises the mechanisms ofeffect between a social programme (component of the de-jure system), and its impact (in the de-facto system), andrecognises these mechanisms as being social: involvingthe interaction of human beings in a particular contextthat determines the nature and form of dynamicresponses [14].

In this paper, we seek to focus on the mechanismsthrough which unintended human resource effects havearisen, using the case studies of Bangladesh and Uganda,two countries that have faced many reforms attempting toimprove health systems performance and delivery of care.Both countries have extremely low per capita expenditureon health (Bangladesh has a public health expenditure of14 international dollars, one of the lowest in the regionand Uganda's public health expenditure lies at 22 interna-tional dollars) and face challenges in the attainment of theMillennium Development Goals and national develop-ment. After 20 years of military regime, Bangladeshbecame a parliamentary democracy in 1991. With rigidcentral government structures and disagreement betweenthe main political parties inhibiting response to localhealth needs, the country began a wide programme ofreforms to address issues of responsiveness. In Uganda, aseries of reforms were introduced from the early 1990s aspart of a larger initiative to restore the health system fol-lowing its collapse during the political crises of the 1970s.

Focusing on selected reforms in each country (see Table 1)the first section of the paper outlines the scope of thereforms implemented and provides observations onworkforce responses made by earlier evaluations. The sub-sequent section then presents the results of this research,exploring the impact of the mechanisms on social interac-tions.

MethodsThe paper is based on a study conducted in 2004 examin-ing mechanisms of effect through which selected healthsector reforms have impacted on human resources in thenational contexts of Bangladesh and Uganda. The studyhad the following main objectives:

1. To critically analyse health sector reforms in relation tothe macro-level environment (policy level analysis:financing, regulation, organisation and management): orto understand the contextual de-jure system.

2. To examine the pathways through which selectedreforms have impacted on the incentive environment forhealth workers (authority and accountability structures,career structures, staff recruitment, payment) and theresponses they have produced: dynamic responses

Study setting and designThe study was undertaken in six district zones in eachcountry. The study districts in both countries were identi-fied using a socio-economic stratification methodemploying the Human Development Index (HDI) rank-ing. The HDI takes three main indicators into account: lifeexpectancy, educational attainment as a proxy measure ofliteracy and GDP per capita as a measure of standard ofliving. Together, these indicators show the level of needand the capacity of districts to benefit from health inter-vention [15]. The method was chosen to capture a widerange of reform experiences across varying socio-eco-nomic development levels. It was expected that recruit-ment patterns, survival activities and communityresponses might operate differently in districts of varyingprosperity levels, and that sampling across the develop-ment level range would increase the breadth of experiencelikely to be captured. Study districts were randomlyselected from three equal strata ranked by HDI, that is,two districts were selected from each stratum. Tables 2 and3 show the selected districts and their HDI ranking inUganda and Bangladesh respectively.

Table 1: Selected reform initiatives in Uganda and Bangladesh

Uganda Bangladesh

• Decentralisation of governance to district councils, 1993 • Unification of health and family planning services at sub-district level, 1998 – 2003• Civil service reforms, 1990 onwards � Procurement

� Pay reforms • Training of health and family planning personnel at sub-district level, 1998 – 2003• User fees implementation in 1992 and retraction in 2001 • Introduction of clinics at the community level, 1998 – 2000• Health sub-district policy, 1999



Methods of data collection and analysisThe information for the study was generated initially byreviewing published and government documents andthen by triangulating both qualitative and quantitativemethods of data collection and analysis among policytechnocrats, health managers and groups of health pro-viders. A comparison was made between providers in thefamily planning and health tracks in Bangladesh andbetween public and non-governmental organisation(NGO) sectors in Uganda. The rationale for the compara-tive groups was based on the nature of the major reformsand the human resource groups that they targeted.

Documents – published and government reports fromboth countries – were analysed in the formative phase ofthe study which was undertaken with the purpose ofunderstanding the reform processes and focusing on theirobjectives, design, implementation and structural impactson health workers at national, district and health facilitylevels.

District level surveys of health workers were conductedbetween February and June 2004. In Bangladesh, 703 pro-viders were interviewed individually, while, in Uganda,the total number of providers interviewed amounted to800 (see Table 4 for breakdown of interviews per district).A questionnaire that had been pilot-tested and translated,where necessary, elicited health workers' perceptions ofthe effects of the selected reforms on their organisationand on themselves as workers. Scaled responses were usedto gauge perceptions of the following characteristics of

their job: management authority, accountability, careerdevelopment, recruitment pattern of staff and theirdeployment, payment schedule, promotion opportunityand survival strategy. Health workers could respond to theclose-ended questions with 'Strongly agree', 'Partiallyagree', 'Do not agree' and 'Do not know', which allowed awider scale of responses and captured the middle ground.In Uganda, all the staff working in the health centres at thestudy sites were interviewed and an attempt was made tointerview at least 50% of doctors and clinical assistantsand 20% of nurses at the selected hospitals. Due to erraticstaff availability and attendance on survey days, theseguidelines were only partially followed. In Bangladesh, asample of health workers was taken from each of the fourdifferent types of public health facilities located at district,sub district and community levels. Clinical workers suchas medical assistants and family welfare visitors wereselected from the community level.

The quantitative data from both countries were analysedusing SPSS and presented in frequency tables.

Key Informant Interviews were conducted with districtlevel health managers and officials in the central govern-ment ministries responsible for health, public service andlocal government. The interviews focused on the context,the content and the process of reforms and elicited per-ceptions of how reforms affected ground-level realitieswithin health facilities and the workforce. In Uganda theinterviews were conducted by social science graduateswho had prior experience in qualitative data collection.

Table 4: Number of provider interviews by districts

Uganda Bangladesh

District Number of respondents District Number of respondents

Kampala 199 Jamalpur 121Jinja 122 Cox's Bazar 115Bushenyi 120 Moulvibazar 119Mubende 139 Barguna 113Moyo 102 Chuadanga 117Arua 118 Gaibandha 118

Table 2: Uganda: Human Development Index of selected study district

Study district HDI Group rank

Kampala 0.593 HighJinja 0.534 HighBushenyi 0.456 AverageMubende 0.458 AverageMoyo 0.361 LowArua 0.383 LowLira 0.405 Average

Table 3: Bangladesh: Human Development Index of selected study district

Study district HDI Group rank

Jamalpur ≥0.501 HighThakurgaon ≥0.501 HighCox's Bazar 0.401-0501 AverageMoulvibazar 0.401-050 AverageBarguna <0.401 LowChuadanga <0.401 LowSirajgonj 0.451–0.501 Average



After two days of orientation regarding the study objec-tives and methods, a semi-structured interview guidedeveloped by the core research team was piloted and thenutilised to conduct the interviews. The guide aimed toexplore the main domains of the reforms and their possi-ble effects on human resources. The findings from theseinterviews were used to identify the themes and questionsfor the individual staff survey. In Bangladesh, questionsfor the key interviews were drafted and discussed with theHealth Systems Development Programme office and con-ducted by the principal and co-investigators at thenational level and by a trained research associate at thedistrict level. National level key informants were inter-viewed about the reform processes and implementationissues, while district level informants were asked torespond to service delivery issues, practical experiencesand constraints at ground-level, as well as issues concern-ing the coordination and cooperation of the two cadres(health and family planning) in Bangladesh, their incen-tives, workload and job environment.

Focus Group Discussions were conducted with differentcadres of health workers (nurses, midwifes and clinicalofficers and associated field cadres) to explore how theroles and support of the organisations concerning theirhuman resource management functions changed underthe reforms. In Uganda six focus group discussions wereheld in Lira. The participants were grouped according totheir cadres i.e. midwives (three groups), nurses (threegroups) and clinical officers and allied health workers(two groups). Members from each group were invitedfrom 2–5 health facilities to Lira, where the discussionswere held. The individual participants were selected froma list of district personnel provided by the District RecordsOfficer. In Bangladesh, eight focus group discussions wereheld in Sirajgonj. Here again the participants weregrouped according to their cadres (Upazila Family Plan-ning Officers (UFPO) and members of the Ministry ofMother and Child Health (MOMCH); Upazila Health andFamily Planning Officers (UHFPO); medical doctors fromthe district hospitals; senior staff nurses from the districthospitals; senior staff nurses from the Upazila HealthComplex; paramedical staff; Sub Assistant CommunityMedical Officers (SACMO) and family welfare visitors(FWV); family welfare assistants (FWA) and family plan-ning inspectors (FPI)) and discussions took place at thedistrict hospital and Upazila health complexes. Upazilarefers to the Health Sub-district level in the Bangladeshihealth system. Guidelines for the topics were preparedand discussions were conducted by a co-ordinator of thestudy together with three trained research associates.

The qualitative information was compiled and analysedfollowing standard qualitative data analysis techniques

with master sheets enabling the identification of emergingthemes and quotes that illustrated the thematic issues.

ResultsThe context of reforms and their effects on intended objectives: de-jure systemBangladeshAfter partition from India in 1947, Bangladesh achievedfull independence in 1971 and became a parliamentarydemocracy in 1991 after 20 years of military regime. Withrigid central government structures and disagreementbetween main parties largely inhibiting response to localhealth needs [16], Bangladesh began a wide programmeof reforms to address issues of responsiveness. The mainreforms in Bangladesh aimed at integrating the two sepa-rate divisions of health services and family planning thusunifying the two programmes with the intention ofimproving their efficiency and responsiveness to the userpopulation.

UnificationThe process of unifying service structures started in 1998at the sub-district (Upazila) level and below with theintention of gradually, and eventually, unifying overallmanagement. Unification entailed reorganisation andrestructuring of health and family planning delivery sys-tems to ensure a common management structure. Allhealth and family planning activities at the Upazila levelwere placed under the unified authority of the UHFPO,who is a medical doctor by profession and formerlybelonged to the health directorate as a cadre. Family plan-ning workers were moved to a formal civil service (recur-rent budget) payroll. This shift had implications forfinancial management, performance appraisal and work-loads.

Historically, health and family planning programmes hadseparate budgets and the drawing and disbursement offunds were handled separately for each track. After unifi-cation, the financial authority to draw and disburse fundsat the Upazila level was delegated to three officers: theUpazila Health and Family Planning Officer in overallcharge, the Upazila Family Planning Officer and theMaternal and Child Health Family Planning Officer.

The unification process reassigned the role of perform-ance appraisal for both the management team at the Upa-zila level and for the operational staff at lower levels. Themajor change in the appraisal system was that individualsfrom a different professional or programme backgroundwere put in charge of appraising personnel [17].

The integration of family planning and health servicesbrought a new set of tasks to be carried out by all frontlinepersonnel. Thirty-seven tasks were identified as compo-



nents of the health worker's job description at the com-munity clinics [18]. Most of these new tasks were addedon to the work of family planning workers while a com-paratively small set of tasks was assigned to workers of theHealth Directorate.

ProcurementThe reforms in procurement and logistics managementcreated a centralised procurement with guidelines necessi-tating a multistage procedure. The separate logistical man-agement systems of health and family planning servicesprior to the Health and Population Service Program(HPSP) were replaced by an integrated logistical system,including unified arrangements for procurement, storage,distribution and transportation. The intention was toimprove service quality, motivate employees to performbetter and reduce cost. An evaluation of the procurementsystem reported that quality and motivation problemspersisted. Continued drug shortages were attributed toextra bureaucracy in the procurement guidelines, lack ofskills in the actual process of procurement, and financiallimitations [19].

Community clinicsIn order to expand access and coverage to the essentialservice package and to replace labour-intensive and costlyoutreach family planning services with a cost-effectivepackage at one location, community clinics were con-structed for every 6,000 population [20]. However, thepersistent unavailability of drugs and supplies reduced theclinics to the provision of family planning and in somecases child immunisation only [21].

UgandaIn the years after independence in 1962, the public serviceof Uganda, including the health service, was regarded asone of the most effective in sub-Saharan Africa. However,health services for the most part collapsed during the1970s and 1980s [22]. Since then Uganda has imple-mented a series of reforms, specifically the decentralisa-tion of service delivery and accompanying reforms in civilservice which were both aimed at achieving the goal ofmaking the central and local authorities function in anefficient and democratic manner.

DecentralisationDecentralisation of health care delivery in Uganda had adual context. The first entailed decentralisation of all gov-ernment services that were delivered by the central gov-ernment with the aim of devolving power to districtauthorities. The second context was the transformation ofroles in the health sector in response to the decentralisa-tion policy.

Matters relating to personnel were part of the decentral-ised functions. A new structure, the District Service Com-mission, was formed at the district level to perform thefunctions of personnel management, under direct super-vision and guidance from the national Public ServiceCommission and Ministry of Public Service. The mainhuman resource management roles at the district levelwere to identify staff requirements and their trainingneeds and to ensure that health facilities had the mini-mum staffing requirements. In addition, the powers torecruit, exercise disciplinary control, promote and toremove persons from district service were delegated to theDistrict Service Commissions (DSCs).

A critical change in the employment system under decen-tralisation was the demand-driven recruitment into thedistrict service. Each district would advertise the availablejobs within its department before any recruitment wasundertaken. This was a major shift from the previous cen-tralised system where medical and paramedical profes-sions were recruited by the Ministry of Health (MOH)immediately upon graduation and posted directly toselected workstations.

However, the revenue base of the local authorities waspoor and depended on grants from the central govern-ment to pay the district personnel. Grants from centralgovernment were enshrined in hard budget constraints(conditional grants) that did not provide for local flexibil-ity in resource allocation [23].

Civil Service ReformsThe civil service reform (CSR) programme was to addressfour key areas: personnel management; organisationalstructure; performance accountability; and service condi-tions [24].

The excessive size of the Ugandan civil service was under-stood to explain inefficiency, poor performance, andinadequate pay and benefits. Between 1990 and 1997, thepublic service was reduced by 54 per cent without a clearagreement on the target-size of staff between the Ministryof Finance and that of Public Services [25]. Further com-ponents of the CSR included decentralising power overpersonnel management to district service commissions,introduction of results-oriented management (ROM) andcapacity building [22].

Pay reformsBy 1995, after salary increments totalling 85 per cent over1986, wages and salaries in the civil service were still per-ceived as being insufficient to maintain an adequatestandard of living [25,26]. After 1995, the pay reformswere abandoned in light of deficient revenue collection



and alternative demands concerning the initiation of auniversal primary education programme.

The first phase of the salary enhancement plans addressedthe judiciary and justice sub-sector. This boosted earningsof the entry-level lawyers to a point that was three timesthat of a consultant medical doctor. The perceived ineq-uity was responsible for several strikes between 1990 and1995 among medical staff. In response, a lunch allowancefor medical workers was introduced in 1994 [27,28]. Fur-thermore, the government consolidated all allowances,such as housing, medical and transport, into salaries butat the same time imposed a pay-as-you-earn tax (30%) onthe consolidated salary; a decision that reduced the over-all impact of the strategy to increase earnings. However,the regularity of salary payments improved due to the factthat earmarked finances were being sent to the districts asconditional grants [27].

Due to budget constraints, the civil service recruitmentinto the district service has been largely frozen since 1993.Although local governments had the authority to recruit,salary delays and arrears in the district payrolls resulted inhealth workers salaries being reassigned to the centralgovernment payroll in 2000 [29].

User feesDuring the 2001 presidential election campaign, user feesin the public health delivery system were outlawed. Anincrease in the number of clients seeking services occurred[29] as did stock-outs of drugs and essential supplies. TheMinistry of Health increased operational budgets to thehealth facilities in order to cater for the increased demandfor drugs [30]. This, however, did not compensate for thestaff's loss of their previous supplementary incomethrough user fee revenues [31,32].

Health sub-districtLow accessibility to basic health services had persisted andgaps in staffing especially in the rural health units wereattributed to weak management of health services belowthe district [33,34]. A policy decision was made to furtherdecentralise health service delivery to the county level,corresponding with the political constituency or the"Health Sub-district" (HSD). In line with this process,Health unit management committees (HUMC) were for-malised and strengthened to oversee the work of thehealth facilities and also act as the link with the commu-nities.

Several strategies were adopted to achieve the HSD objec-tives. These included the upgrading of some health facili-ties to provide a comprehensive set of essential healthservices as well as harnessing available hospital capacityfor disease prevention and health promotion activities. Inaddition, the health sub-district personnel were requiredto provide managerial support to 7–15 lower-level healthfacilities within their defined geographical area [34].These additional support roles added a substantial work-load related to administration i.e. supervision, financialmanagement, coordination and reporting.

The human reactions to reforms: dynamic responsesSurvey findingsHealth personnel in both Bangladesh and Ugandaresponded to a questionnaire covering questions on theeffects of reforms regarding issues such as management,supervision, training, promotion, salaries, the availabilityof drugs and supplies and incentives for remote deploy-ment (see Table 5).

The survey data suggest that in Bangladesh the reformshave affected the health staff and former family planning

Table 5: Percentage of health staff agreeing with statement on reforms

Bangladesh unification reforms Uganda decentralisation reforms

Health Family planning Public NGO

% Strongly or partially agreeing

The reforms have increased your chances of being promoted 8 7 50 51The reforms have increased the chance for you to keep your job 31 64 55 46The reforms have increased the objectivity of the appraised/performance reports

62 74 74 73

Your salary is always paid promptly 93 59 73 65Your salary increments in the past couple of years have been satisfactory 47 38 19 19The reforms have made your job description clear 67 83 60 55The equipment essential to perform assigned tasks is available in sufficient quantities

75 56 63 58

The drugs and supplies required to accomplish your tasks are always available 70 32 70 65The reforms have made your workload more manageable 64 83 52 49



staff in different ways. While family planning staff con-sider themselves to have least benefited from the reformsconcerning their salaries, promptness of pay and the avail-ability of drugs, supplies and equipment, health staffseem less satisfied with their job security and the manage-ability of their workload. In Uganda, there appears to beless satisfaction with workload and pay, but compara-tively more satisfaction with promotion prospects. Gener-ally, both in Uganda and Bangladesh, health staff seemmore satisfied with the objectivity of performance reportsafter reforms.

The next section explores the multiple interactionsbetween the key players affected by the reforms and pro-vides an insight into the positive and negative responsestriggered by the changes that took place in Bangladeshand Uganda in the context of the reforms.

Interview and focus group findingsBangladeshIn Bangladesh, the changes that had affected the authoritystructures displayed a multitude of reactions amongsthealth staff, who experienced the impact of reforms ontheir management arrangements, performance appraisalsystems, promotional opportunities, payment schemesand ultimately on their performance and provision ofservices. The combination of unification efforts and theadjustments to familiar arrangements resulted in a mix-ture of responses and interactions between health andformer family health staff, their senior level managementstaff and community members who had been involved inthe formation of the community clinics.

ManagementRespondents reported that the unification at the lower lev-els (Upazila and below) of the health and family planningprogrammes and the resulting assignment of administra-tive authority to personnel in the health track had causeda conflict. Family planning directors and their subordi-nates at the national and district level lost their authorityover staff at lower levels.

The informants at the district management levels in Bang-ladesh were divided on the effects of the unificationreforms at the Upazila level and below. Civil surgeons(managers on the health side of the divide) held optimis-tic views about the new administrative and authoritystructures at the Upazila. In contrast, the deputy directorsof family planning described how the changes had causedturmoil in the functioning of the programmes. It wasreported that the new structures were dysfunctional andhad demoralised workers that belonged to the familyplanning programme. The same split in perceptionsregarding the newly unified structures was observedamong the managers at the sub-district (Upazila) level.

The sub-district managers expressed difficulties in exercis-ing their authority due to the continued, and sometimesperceived as deliberate, interference by higher offices inthe functioning and allegiance of sub-district personnel.Informants confirmed that the unification reform erodedthe authority of managers:

"We are supposed to supervise the Health Assistants, HealthInspectors and Family Planning Inspectors but these people donot follow our advice and orders and the boss (UHFPO) doesnot instruct the persons to cooperate with us" (SACMO andFWV group).

Complaints concerning the bureaucratic obstacles inaccessing administrative services at the management levelwere also voiced in the focus group discussions. Theseincluded delays in allowance payments and bribes thatworkers felt obliged to pay to their managers for theirallowances to be approved and paid.

"We have faced problems with the three Drawing and Dis-bursement Officers and more managers. To get the supplies,transport and other allowances and bills paid we have to moveto several offices, which hampers our work and causes dissatis-faction" (Senior staff nurses group, Bangladesh)

Performance appraisal systemsThere had been a dual system of performance appraisal inthe unified structure. Historically, the family planningpersonnel used a service book, which monitored the dailyactivities undertaken by each worker. Given the prospectsfor transferring the Family Planning personnel to the rev-enue payroll, the service book was used as the basis forsuccessful transfer and in addition determined the level ofsalary to be received upon transfer into the civil service.The administrative arrangements assigned some of thesupervisory roles for the health service track personnel tothe family planning track officers and vice versa. Discus-sants expressed suspicions about performance appraisalreports by superiors who belonged to the rival track,claiming Annual Confidential Reports were being used aspunishment and as a way to settle scores between person-nel and supervisors.

Career structure and promotionAlthough promotional and career structure problems pre-ceded the reforms, the common perception in Bangladeshsuggested that the problems were aggravated or remainedunsolved. Unification reforms assigned a critical elementtowards promotion, that is, annual confidential reports(ACR) from the trusted vertical programme hierarchy to anew set of officers. The competence of the new assessorsof performance was contested partly due to organisationalmistrust and possibly due to the dismantling of the well-knit network of patronage in the previous structures.



Salaries and promptness of payPredominant views emerging from the FGDs (focus groupdiscussions) suggest that salaries had not increased appre-ciably under the unification reforms. Many health work-ers expressed concerns over their income and mentioneda number of strategies to cope with the insufficiencies.Most strategies related to activities outside their officialwork including cultivation, drug shops and work at pri-vate clinics. Low and unpaid salaries were among thedominant concerns arising from focus group discussions.

"We can hardly meet expenses for 2 weeks from our salaries.We have to live at a low standard compared to the status wehold. This is degrading" (FPO and MA at sub-district level).

Among family planning staff, delays in salary paymentwere commonplace, prompting enthusiasm for promisedtransfer to payment from the revenue budget, the samepayment scheme that health staff were benefiting from.

"For six months I have not been paid any salary. My wife askedme if I am working at my job or doing other things" (FWAs &FWI group).

"We are told that 40 per cent of our colleagues have been trans-ferred to the revenue budget and the rest of us will be trans-ferred soon. This will give us an opportunity to earn salary ontime" (Sub Assistant Community Medical Officer & Fam-ily Welfare Visitor group, Bangladesh).

Personnel on the established government civil servicereceived their salaries regularly on the first week of theproceeding month while those that remained on the irreg-ular payroll still suffered from delays in their payments.

Respondents in Bangladesh suggested that the changes inallowances had resulted from the scaling down of field-work and the introduction of new administrative struc-tures that threatened the established systems to accessallowances and challenged the long tradition of networksthat the personnel had built over time with their superi-ors. As mentioned earlier, the distribution of financialcontrol across three different persons after unificationintroduced tighter micro-bureaucracy and gate-keepingthat made access to allowance and claims difficult.

"To process the allowance bills, you go to the Upazila FamilyPlanning Officer and he tells you to go to the Upazila Healthand Family Planning Officer (UFHPO). The UFHPO tells youto go back to Upazila Family Planning Officer and so on. Thebill is not paid for one month ... you have to travel to see whathas passed. When it is passed one has to pay commission to thebosses and to the clerks. It is frustrating" (Family PlanningInspector and Family Welfare Assistant group, Bangladesh)

Community expectationsIn Bangladesh, community participation was demon-strated by donation of a plot of land, construction ofbuildings to house the clinics and the formation of com-munity clinic groups (CCG) for clinic management. Ini-tially most health workers applauded communityinvolvement, especially because it encouraged learningabout the constraints under which they were working.During the interviews and focus group discussions, itbecame clear that the health workers were aware of howthe changes in their performance and the effects on serviceprovision were perceived by the communities.

i. Availability of drugs and supplies

One of the stated objectives of the health sector reformwas to ensure that the procurement and supply of drugs tothe community clinics was brought under unified man-agement and made more efficient in terms of cost reduc-tion and saving time, thus ensuring steady and betterservice delivery. Discussants stated that this objective hadnot been reached and claimed that they were thus unableto adequately perform their tasks due to the irregularity inthe provision of supplies. They further described how pro-curement problems caused anger and incomprehensionamong communities that initially had welcomed thecommunity clinics:

"There was enthusiasm among the community people and theirviews about the Community Clinics were positive ... becausedistances for minor illnesses were reduced. People were compet-ing to provide land for the clinics but when the drug suppliesstopped, the people could not see the benefits ... their views arenow unfriendly" (Upazila Health and Family Planning &Medical Officer group, Bangladesh).

"It is not serious to go to the clinic daily and sit there doing nowork. People come for family planning or cough and you tellthem there are no medicines. It is risky... some time they abuseus and complain why we are being paid salary" (Sub AssistantCommunity Medical Officer, Family Welfare Visitor group,Bangladesh).

ii. Adequate service provision

In Bangladesh the service integration between familyplanning and health services brought a new set of tasks tobe carried out by all frontline personnel. In effect, the inte-gration added several services, for example, child immuni-sation, basic treatment of ailments and directly observedtherapy (DOTS) to the family planning personnel. Viewsemerged during the discussions with health staff suggest-ing that family planning activities had been displaced bytheir new tasks under integration, at Upazila level. Many



participants indicated that the increase in workload leftless time for patients with family planning needs:

"I am required to provide services at the clinics and visit house-holds for Family Planning. It is impossible to do all this work.Family Planning suffers because there are many Expanded Pro-gramme of Immunisation (EPI) clients to see at the clinic" (FPIand FWA group, Bangladesh).

"At the inauguration of the CC we were supplied drugs once ortwice and the community response was good. But there arefewer persons attending at our units [union level].... the clinicshave not received any supplies now six months after their [offi-cial] opening" (group discussant SACMO and FW, Bangla-desh).

UgandaIn the Ugandan context, the decentralisation effortsbrought about a rapprochement of local authorities, dis-trict boards and health workers who previously hadenjoyed relative autonomy from local management struc-tures. Here the closer contact and growing relationshipswith local authorities together with the changes inaccountability structures brought about a multitude ofresponses amongst health workers, specifically concern-ing their recruitment patterns, job security, supervisionand relationship with community members.

RecruitmentThe new authority structures that came with the decentral-isation policy were the district councils, the district servicecommissions and the health unit management commit-tees. In the discussions, health workers had differing opin-ions regarding the influence and newly acquired decision-making powers of these structures regarding specificallythe issues of recruitment and supervision.

"Recruitment by the district is now faster and makes it possibleto get quickly on the payroll ... this was taking years before"(District Director of health services, Uganda)

"When you go for the interview, these people (district servicecouncillors) interview you in the local language and use com-plex parables and proverbs to fail you. If you do not understandthe questions you are technically out" (Group Discussant, Clin-ical officers, Uganda)

"...some few, especially sons and daughters of those in positionsof influence in the district don't bother with the recruitmentprocess. They are just appointed." (Focus Group Participant,Uganda))

Personnel working away from home viewed the process ofrecruitment as inward looking and biased against workersfrom districts that were not home to the new authorities.

Local informants explained that the expression "sons anddaughters of the soil" is widely understood to describe thephenomenon of preference for workers who originatefrom the district. While territorial patronage systems hadalready been observed before reforms, respondentsclaimed nepotism was present to a higher degree with therelocation of decision-making authority to local bodiesunder decentralisation.

Job SecurityIn Uganda, health workers with management responsibil-ities expressed fears about their precarious relationshipswith the local authorities. They claimed that their con-cerns mainly arose from local authorities' decision-mak-ing power in the dismissal of health workers. The viewsbelow also point towards a patronage network that isclaimed to have strengthened under decentralisation.

"...For some of us who work in rural places if you disagree withyour sub-county bosses on any matter, that boss will automati-cally make sure that he at least punishes you and you lose yourjob" (FGD Midwife, Uganda)

"I think the relationship is good. If you give them [district coun-sellors] what they want, they also make our work easy. .. Some-times they take the only vehicle for burials or campaigns and wehold back the activities. At the end of the day one has to be ingood books or you are thrown out" (Assistant District HealthDirector, Uganda).

SupervisionIn addition to the structural changes that influencedrecruitment patterns and job security, the type and focusof supervising the health workers changed with the newstructures in Uganda. It was reported that HUMC, CAO(Chief Administration Officer), DSC and local politicianswere increasingly undertaking the supervision of person-nel and health facilities. Health staff described how beforedecentralisation, supervision was geared more towardsthe technical quality of workers encouraging them toundertake new service roles or to improve their services.According to their accounts after the reforms, supervisiontook on a more human angle with the direct accountabil-ity of health staff to selected community members. In gen-eral, health staff seemed to appreciate the new changes intheir supervision since relationships to key communitymembers improved through tighter working relation-ships.

"Decentralisation has fostered better supervision of facilities bythe district because they are now answerable to the people at thedistrict or within their areas of jurisdiction" (Secretary DistrictService Commission, Uganda)



"... The health unit management committee has come to realizethat the major problem is not selling of drugs but the smallquantities always supplied. Now the relationship is not so badas it was in the past" (Focus group discussant, clinical Officers,Uganda)

Views also emerged from the focus group discussions inUganda that suggested the health staffs' relationships spe-cifically with community leaders had changed with decen-tralisation. There seemed to be an overall improvement.

"Leaders now appreciate the work we do and are keen to supportus. ... they have been articulating our needs for more staff at thedistrict council" (Key Informant Interview, Facility In-charge)

Despite several attitudes describing the positive relation-ships and increased understanding between local author-ities and health staff through the changes in supervision,several statements displayed the workforce's concernabout relationships with those community members thatwere less involved in overseeing the provision of healthservices.

Community expectationsCommunities expected the quality of services to improvemarkedly when user fees were removed as had been prom-ised during the presidential election campaigns. Theincrease in the utilisation of health services after theremoval of fees put pressure on the availability of drugsand supplies leading to marked shortages. Health workersreported a sense of feeling caught in the middle of thefinancial shortcomings of government and the highexpectations created by the publicity about the removal ofuser fees.

"How do you expect us to handle that problem (too manypatients and too few drugs)? We used to buy drugs from the feesbut now we are just spectators. We tell them OS (out of stock)and they say we are not kind to them" (Group Discussant, Clin-ical Officers, Uganda).

Staff in Uganda described how communities were sur-prised when medical supplies and drugs would suddenlyrun out in times of need. In fact, FGD discussants reportedhow patients would question the trustworthiness ofhealth staff in the facilities.

"There has been a concept... that medical people steal drugsfrom health units. Drugs kits would be delivered and three dayslater we would tell them (patients) that the drugs are finished.They did not believe ...they would say that we have stolen thedrugs..." (Group Discussant, Clinical Officers, Uganda)

While the quantitative data provide a general picture ofstaff attitudes and their responses to the reforms in the

two countries, the qualitative data explore how reformshave been interpreted and perceived, and how dynamicsbetween key players reveal whether or not the workforceis likely to consist of active and enthusiastic implementersof reform or dissenters, actively or passively resistingimplementation. The findings from the qualitativeresearch suggest less satisfaction with the changes that hadtaken place with the reform initiatives compared to thesurvey results. This could be due to the nature of the meth-ods, if people are less likely to answer negatively whenprompted with a questionnaire than in focus groups orinterviews, using open-ended questions.

DiscussionThis paper has presented findings from a comparativeanalysis of two country case studies investigating theimpact of health sector reforms on human resources inBangladesh and Uganda. Pawson and Tilley's [13] propo-sition for realist explanation provided a useful frameworkfor the exploration of the mechanisms by which healthsector reforms affect health workers' micro environments,thus changing provider incentives and creating a multi-tude of responses. By using the dynamic responses modelfor health system research, this paper provided an insightinto the de-jure system from which the reform objectiveswere planned and initiated and traced the resultingdynamic responses between different levels and types ofhealth staff and the communities.

In both Uganda and Bangladesh, reform plannersneglected the role of context in their planning of reformobjectives and assumed that the workforce would act as apassive element in the reform implementation. In thissense, the study highlighted several issues in both coun-tries that demonstrate the importance of careful analysisof contextual factors in the design and implementation ofreform objectives and the significance of recognising theworkforce as an important and adaptive factor contribut-ing to the success or failure of the reforms.

In Bangladesh, the effects of reforms resulted in a powerstruggle and general mistrust between the family planningand health tracks. A strong sense of inequity in the natureof effects of the reform was illustrated by family planningpersonnel who perceived they had lost out while healthtrack personnel had gained from the reforms both interms of authority and actual reduction of tasks underintegrated provision. There remained considerable dis-trust within the new unified structure, with reluctance totake orders from the new managers. The non-unified topadministration above the integrated Upazila level was per-ceived as a potential source of conflict and ill intent. Inaddition, drug shortages and general procurement failuresemerged as a source of bad public relations between theworkforce and the communities despite originating in



broader systemic problems such as financing and budgetconstraints that had already existed before the initiationof reform objectives. The communities participated in theplanning process and priority setting through communitystructures, however, their decisions were not bindingsince the Ministries of Finance imposed budget cuts andceilings motivated by broader pressures in the economiccontext.

On the other hand, positive responses from the majorityof former family planning staff expressed hopefulnessregarding the changes that would secure salary paymentsand inequities in schemes and demonstrated optimismconcerning prompter salary payments under the revenuebudget. This suggests a rise in motivation levels in balancewith the right incentives.

Ugandan findings show how the workforce responded toa strong and rapidly implemented system of decentralisa-tion that had insufficient competence for human resourcemanagement. Power of local authorities was influencedby resource constraints and nepotism in recruitment. Sub-sequently, these constraints suppressed the incentives ofworkforce related to promotion, job security and profes-sional growth. Health workers were more insecure as aresult of decentralisation due to their dependence on thebenevolence of the new authorities in order to guaranteetheir jobs.

However, closer ties with selected community membersalso had positive effects for health workers. It was sug-gested that community leaders in their new supervisoryfunction were able to witness the precarious position thathealth workers had as middlemen between budget con-straints and procurement failures at higher levels of thesystem and the communities who expected smooth run-ning services and consistent availability of drugs andequipment. In this sense, local leaders with their newlyacquired authority were able to lobby for necessary finan-cial and human resources for the efficient operation of thehealth centres.

While the de-facto system has not specifically beenaddressed in this paper, the outcome of the relationshipbetween the de-jure system and the dynamic responses forthose that ultimately use the services are varying. Whileusers on the one hand may be confronted with demoral-ised and in some cases absent staff, lack of necessary drugsand equipment and dilapidated health facilities, the shiftof authority and supervisory functions towards the ruraldistricts, on the other hand, seems to have positivelyaffected communities who are now actively involved inthe shaping of their health service provision with theirnewly acquired responsibilities.

ConclusionThis paper has emphasised several key points that aresummarised below:

1. By keeping the dynamic responses model in mind,national and international reform planners can designreform objectives that ultimately enhance and improveservices as felt by the communities by encouraging favour-able responses amongst the workforce.

2. Reform planners need to take a closer look at the con-text within which the health system operates in order torecognise potential 'inhospitable elements' which mayhinder reform objectives or 'hospitable' elements whichmay support reform initiatives and provide a basis forimprovements in the operation and management ofhealth systems.

3. Reform programs need to incorporate active implemen-tation research systems to learn the contextual dynamicsand responses, as well as have inbuilt program capacityfor corrective measures.

4. Health workers are key stakeholders in any reform proc-ess and should participate at all stages, that is, conceptual-isation, design and implementation. Reforms tend tocreate losers and winners or can change power structuresbut it is important, at the least, that winners and losersunderstand the purpose of change and have confidence inthe process of consultations on which change has beendetermined.

5. How health workers perceive their relationship with thecommunity will affect their job motivation and perform-ance. This is an important but neglected criterion for eval-uating the impact of human sector reforms.

Competing interestsThe author(s) declare that they have no competing inter-ests.

Authors' contributionsCH, FS, BM were involved in the development of the pro-tocol. FS and ER carried out the fieldwork in Uganda andSAR and AM in Bangladesh. FS, SAR, ER, AM, CH, BM andTK undertook the analysis of the data and TK, BM, FS, SARand ER drafted the manuscript.

AcknowledgementsThis work was undertaken with the financial support of the Alliance for Health Systems Policy Research while the authors were employed through the Department for International Development knowledge programme on Health Systems Development (2001–2006). Both funding sources and WHO, as the institutional home of the Alliance for Health Systems Policy Research, are gratefully acknowledged. We would also like to thank the health staff in Bangladesh and Uganda and all those involved in the setting



Publish with BioMed Central and every scientist can read your work free of charge

"BioMed Central will be the most significant development for disseminating the results of biomedical research in our lifetime."

Sir Paul Nurse, Cancer Research UK

Your research papers will be:

available free of charge to the entire biomedical community

peer reviewed and published immediately upon acceptance

cited in PubMed and archived on PubMed Central

yours — you keep the copyright

Submit your manuscript here:http://www.biomedcentral.com/info/publishing_adv.asp

BioMedcentral

up and operational details of the study, without them this project would not have been possible.

References1. Wiskow C: The effects of reforms on the health workforce.

Background paper to the World Health Report 2006; funded by DFID2006.

2. Alwan A, Hornby P: The implications of health sector reformfor human resources development. Bulletin of the World HealthOrganization 2002, 80:56-60.

3. Franco LM, Bennett S, Kanfer R: Health sector reform and publicsector health work motivation: A conceptual framework.Social Science and Medicine 2002, 54:1255-1266.

4. Ngufor GF: Public service reforms and their impact on healthsector personnel in Cameroon. In ILO/WHO Public service reformsand their impact on health sector personnel: Case studies on CameroonGeneva: International Labour Organization/World Health Organiza-tion; 1999.

5. Kohlemainen-Aitken RL: Decentralization and HumanResources: Implications and Impact. Human Resources for healthDevelopment Journal 1998, 2:1.

6. Mutizwa-Mangiza D: The Impact of Health Sector Reform onPublic Sector Health Worker Motivation in Zimbabwe. InMajor Applied Research 5, Working Paper 4 Bethesda, MD: Partnershipsfor Health Reform Project, Abt Associates Inc; 1998.

7. Rigoli F, Dussault G: The interface between health sectorreform and human resources in health. Hum Resour Health2003, 1:9 [http://www.human-resources-health.com/content/1/1/9].

8. Ambegaokar M: Managing people in the health sector: Consid-ering district team performance contracts in Cameroon.HPU, London School of Hygiene and Tropical Medicine. London;2004.

9. Martineau T, Buchan J: Human resources and the success ofhealth sector reform. In American Public Health Association (APHA)Washington DC; 2000.

10. Humedes N, Ugalde A: Human resources: the Cinderella ofhealth sector reform in Latin America. Human Resources forHealth 2005, 3:1.

11. Schlette S: Public service reforms and their impact on healthsector personnel in Colombia. In ILO/WHO Public service reformsand their impact on health sector personnel: Case studies on CameroonGeneva: International Labour Organization/World Health Organiza-tion; 1998.

12. Abzalova R, Wickham C, Chukmaitov A, Rakhipbekov T: Reform ofPrimary Health Care in Kazakhstan and the Effects on Pri-mary Health Care Worker Motivation: The Case of Zhezka-zgan Region. Major Applied Research 5 Working Paper 3, Partnershipsfor Health Reform 1998.

13. Pawson R, Tilley N: Realistic Evaluation London: SAGE Publications;1997.

14. McPake B, Blaauw D, Sheaff R: Recognising patterns: health sys-tems research beyond controlled trials. In HSD Working PaperLondon: LSHTM; 2006.

15. UNDP: Human Development Report. New York: UNDP; 2004. 16. DFID: Bangladesh Health Briefing Paper DFID Health Systems Resource

Centre: London; 1999. 17. Normand C, Iftekhar MHM, Rahman SA: Assessment of the

Effects of Unification on Health and Family Planning Serv-ices. Study Report. Ministry of Health & Family Welfare Government ofBangladesh & London School of Hygiene & Tropical Medicine, University ofLondon 2002.

18. Rahman A: The Study to Assess Implementation of HPSPusing Essential Services Package (ESP). Dhaka, Health Econom-ics Unit, Ministry of Health and Family Welfare 2003.

19. Normand C, Reza Hasan AH, Rahman SA, Iftekhar MHM: Study onEnhancing Health System Performance using Procurementand Supply of Goods under Health and Population SectorProgramme. HSD Programme, Policy & Research Unit, Ministry ofHealth & Family Welfare Government of Bangladesh & London School ofHygiene & Tropical Medicine, University of London 2002.

20. Routh S, Shams-El A, Jahan SA, Begum A, Thwin AA, Baqui AH:Developing Alternative Service Delivery Strategies forMCH-FP Services in Urban Areas: Findings From an Experi-ment. International Center for Diarrhoeal Disease Research, Bangladesh(ICDDR, B). Dhaka 1997.

21. MTR/APR: Mission Team Report 2001:76-78.22. Corkery: Public Service Reforms and their impact on health sector person-

nel in Uganda Geneva, ILO and WHO; 2000. 23. Jeppsson A: Financial priorities under decentralization in

Uganda. Health policy and Planning 2001, 16(2):187-192.24. Okutho G: The way forward. In Post Conflict Uganda: Towards an

effective Civil Service Edited by: Langseth P, Mugaju J. Kampala: FountainPublishers; 1996.

25. Okutho G: Civil Service Reform in Southern and Eastern Africa – UgandaMinistry of Public Service; 2004.

26. Orech M: Civil Service Reform in the context of structural AdjustmentKampala: Ministry of Public Service; 1995.

27. Kanyesigye EK, Ssendyona GM: Payment of Lunch Allowance: ACase Study of the Uganda Health Service. Joint Learning Initia-tive working paper 4-2 Global Health Trust 2004.

28. MOFP&ED: Background to the Budget Ministry of Finance, Planning andEconomic Development. Kampala; 1994.

29. MOH: Health sector Strategic Plan 2000/01 – 2004–05: Midterm ReviewReport Ministry of Health. Kampala; 2003.

30. Konde KJ, Okello D: User Fees in Government Health Units in Uganda:Implementation, Impact and Scope. Small Applied Research Paper No. 2,Partnership for Health Reform Abt Associates Inc: Bethesda; 1998.

31. Acaya G: Increased Utilization Reduced Quality: the Result ofAbolition of Cost-Sharing in Busia. Uganda Health Bulletin 2001,7(4):49-53.

32. Karamagi HC: In Whose Interest Does the State Act? AReview of the User-fee Policy changes in Uganda. UgandaHealth Bulletin 2001, 7(2):34-37.

33. Nzabanita A, Nganwa A: Progress Report on Decentralization of HealthServices below the District Level in Uganda Ministry of Health. Entebbe;1999.

34. MOH: Health Sub-District in Uganda. Concept paper. Kampala 1999.






http://www.human-resources-health.com/content/1/1/9




http://www.biomedcentral.com/info/publishing_adv.asp
