introducing quality managementinto primary health care services in

Introducing quality management into primaryhealth care services in UgandaF. Omaswa,' G. Burnham,2 G. Baingana,3 H. Mwebesa,3 & R. Morrow4

In 1994, a national quality assurance programme was established in Uganda to strengthen district-levelmanagement ofprimary health care services. Within 18 months both objective and subjective improvementsin the quality of services had been observed. In the examples documented here, there was a major reductionin maternal mortality among pregnant women referred to Jinja District Hospital, a reduction in waiting timesand increased patient satisfaction at Masaka District Hospital, and a marked reduction in reported cases ofmeasles in Arua District. Beyond these quantitative improvements, increased morale of district health teammembers, improved satisfaction among patients, and greater involvement of local govemment in thedecisions of district health committees have been observed. At the central level, the increased coordinationof activities has led to new guidelines for financial management and the procurement of supplies. Districtquality management workshops followed up by regular support visits from the Ministry of Health headquar-ters have led to a greater understanding by central staff of the issues faced at the district level. The qualityassurance programme has also fostered improved coordination among national disease-control pro-grammes. Difficulties encountered at the central level have included delays in carrying out district supportvisits and the failure to provide appropriate support. At the district level, some health teams tackled problemsover which they had little control or which were overly complex; others lacked the management capacity forproblem solving.

IntroductionThe health status of people in sub-Saharan Africacontinues to lag behind that of those in other regions(1, 2). Morbidity and mortality among children andadults remain unacceptably high. The health sectormust meet increasing demands with resources thatare often declining in real terms. As this is likely tocontinue, better management of existing resourcesoffers the best and perhaps the only hope of improv-ing the quality of health services and increasing thehealth status of the people (3).

In developing countries, a management-by-results approach is a common strategy for improvinghealth care services, i.e. setting quantitative cover-age targets for specific interventions combined withinspection-oriented supervision (4, 5). Yet sustain-able quality improvements are rarely achieved

I Head, Quality Assurance Unit, Uganda Ministry of Health,Entebbe, Uganda.2 Assistant Professor, Department of International Health, TheJohns Hopkins School of Hygiene and Public Health, 615 N. WolfeStreet, Baltimore, MD, 21205 USA. Requests for reprints shouldbe sent to this author.3 Medical Offlcer, Quality Assurance Unit, Uganda Ministry ofHealth, Entebbe, Uganda (deceased).4 Professor, Department of International Health, The JohnsHopkins School of Hygiene and Public Health, Baltimore, MD,USA.Reprint No. 5765

because underlying managerial and logistical weak-nesses are not addressed. These deficiencies oftendiscourage health care workers from applying theirskills and make it difficult for the health system touse effectively the external resources provided.

A fundamentally different management ap-proach, variously termed total quality management(TQM) or continuous quality improvement (CQI),has been widely embraced by health services andindustry in Japan, Europe, and North America (6).Broad-based, process-oriented management, the useof teams, and decentralized decision-making havegenerally replaced the traditional "top down" man-agement approach in efforts to achieve quality inproducts and services.

In the Ugandan public sector, wide-ranging de-centralization has devolved both the authority andfunding of district-level health services to the districtgovernment, opening new possibilities for healthmanagement initiatives. We report here the intro-duction on a national basis of TQM methodsfor health services in Uganda, difficulties encoun-tered during their introduction, and the resultsachieved in the first 18 months following theirintroduction.

Health Services in UgandaAfter two decades of internal conflict that have seri-ously eroded health services, Uganda is undergoing

Bulletin of the World Health Organization, 1997, 75 (2): 155-161 © World Health Organization 1997 155

F. Omaswa et al.

major sociopolitical reform. The three-year nationalhealth plan, drawn up in 1993, reoriented the focusof the Ministry of Health towards primary healthcare (7). Government services, including health serv-ices, have been decentralized to the district level.Accordingly, each district health team prepares anannual work plan and budget which are submitted toits district council for funding from the treasuryblock grants that the council receives. A cost-sharingplan was introduced, allowing collected funds tobe utilized by the health facility concerned. Themanagement of Ministry of Health headquartershas been restructured to be more responsive to theneeds of decentralized districts. The health sector'sshare of the national budget has increased from 1%(1986) to 5.8% (1994), equivalent to a per capitaincrease from US$1.20 to US$5.00. However, cen-tral government health expenditures still remainconsiderably less than those in most sub-SaharanAfrican countries (8).

During development of the three-year plan, arecurrent finding was poor management of hospitalsand health services. In selecting activities to befinanced under the World Bank's Second Health Sec-tor Loan to Uganda, improving the quality of carewas made a priority. A national quality assuranceprogramme was launched in March 1994, and its firsttask was to assist health teams in managing theirdecentralized districts. The methods used built onthe experience of a quality assurance pilot projectcarried out in Kabarole and Bundibugyo districtsin Uganda by the German Agency for TechnicalCooperation (GTZ) and UNICEF.

Evolution of quality improvement methods

Quality improvement methods were introduced inmanufacturing industries in the 1930s when it be-came evident that reliance on inspection was lesseffective than strengthening the production proc-esses (9). These early concepts evolved to includethe quality of design, marketing, distribution, andother organizational activities. Other developmentsemphasized the collection of data about the produc-tion process, use of this information by teams ofworkers, and a renewed focus on determining andaddressing the needs of customers. Some key publi-cations made these concepts an integral part of man-agement science worldwide (10-12). In the 1980squality improvement methods, embodied in theTQM and CQI approaches, were adopted by serviceindustries, including the health sector (13-15). Therecognition that developing country health serviceswere not reaching their potential despite increasedtraining, financial investment, and supervisory ef-forts sparked interest on the part of the United

States Agency for International Development andother groups in the application of quality improve-ment methods that had been successful elsewhere (5,16, 17).

MethodsThe national quality assurance programmeThe national quality assurance programme beganwith establishment of a quality assurance unit com-posed of three medical officers. An initial quality-awareness workshop was conducted for seniorpersonnel from the Ministry of Health, MakerereUniversity Medical School, and Mulago Hospital.During this workshop a 25-member national qualityassurance committee was established, composed ofsenior Ministry of Health personnel, managers of allnational disease-control programmes, and repre-sentatives from the medical school and the centralhospitals. This committee assists the quality assur-ance unit in implementation of the quality assuranceprogramme and integration of activities within theMinistry of Health.

The principal focus of the programme is on im-proving the quality of district-level health services byusing quality management methods to identify andsolve common service-related problems. This ap-proach includes the development and disseminationof standards or guidelines, determining the needs ofpatients and their families, strengthening communi-cation between health care providers and users, andusing data to identify gaps in quality.

Quality management methods were introducedthrough quality-awareness workshops for district-level health teams and the district's administrativeand political leadership. During these workshops,participants selected clinical or administrative prob-lems from their districts to be addressed over thefollowing 6 months by means of quality improve-ment methods. They also developed work plans tocollect the necessary data, develop and apply solu-tions, and measure the resulting changes. After 6months, the district teams met again to compare theresults of their problem-solving activities, and toidentify a further round of problems to be tackled.At the end of the first year, a general meeting washeld for district health teams to share the lessonsthey had learnt.

A key part of the programme is support visits todistricts conducted by teams of two or three personsfrom the national quality assurance committee. Dur-ing these visits, the national team works with thedistrict health team to solve problems related tothe district's work plan, and other administrative or

WHO Bulletin OMS. Vol 75 1997156

Quality management in primary health care services

clinical problems district managers are experiencing.At monthly meetings of the national committee,teams discuss their findings and arrange for furthertechnical assistance from specific national pro-grammes if necessary.

ResultsMinistry of HealthAn important achievement at the central level hasbeen the strengthening of interaction among disease-control programmes, facilitated by monthly qualityassurance committee meetings. District visits bycommittee members have helped staff at the Minis-try of Health appreciate the need for integrateddistrict-level services.

Examples of problems detected through thesevisits were inconsistent drug procurement proce-dures and difficulties in obtaining health funds chan-neled through local government. In response, proce-dures and guidelines were quickly established bythe Ministry of Health and the Ministry of LocalGovernment, which eliminated the problems.

At the Ministry of Health, difficulties were ex-perienced in carrying out district-level support visitsas often as originally envisioned, and cooperationbetween some disease-control programmes remainsuneven.

District levelOne of the programme's principal accomplishmentshas been to bring together district health teams withlocal administrators and political leaders to shareresponsibility for strengthening health services.Within the health system, the integration of curativeand preventive activities has been improved, andhospital managers are now in many cases a part ofthe district health teams. A more smoothly function-ing referral system between health units and districthospitals was one of the first benefits of improvedcooperation.

Developing district-level problem identificationand problem-solving capacity is a central objective ofthe programme. Table 1 shows the nature of theproblems that district health managers identified andundertook to resolve. Of the problems tackled, thoserelated to cost-recovery schemes were the most com-mon. In resolving them, weak points in the cost-recovery process were located and strengthened, andstandard procedures established for health centres toallocate the funds thus realized. Following thesechanges, a number of districts noted reduced absen-teeism by health care workers, improved morale,

Table 1: Types of problems selected by districts forresolution using quality improvement methods

No. ofdistricts

Type of problem selectingproblema

Management and administration, including 21financial management

Quality of clinical and preventive services 20Collection and use of information, especially 13

for planning purposesPatient or employee satisfaction 6

a Some districts selected two problems.

and increased patient satisfaction with the servicesprovided.

The following examples of problems addressedin districts illustrate the approaches taken and theresults obtained.

Maternal mortality in Jinja District. In 1993, 17 of 126pregnant women referred from the district's approxi-mately 30 rural health units died after arrival at JinjaDistrict Hospital. The principal causes of deathwere found to be haemorrhage (antepartum andpostpartum), ruptured uterus, and postpartum sep-sis. Many primary health care workers did not rou-tinely identify high-risk pregnancies and some couldnot identify patients in need of emergency referral.Hospital staff and the district health team began aneducational campaign for outlying health units thatstressed the early recognition and prompt referral ofwomen with high-risk pregnancies and obstetri-cal complications. Hospital record keeping wasstrengthened, and a monthly outcome record forpatients referred for obstetrical complications wasestablished. The district medical offlcer arranged fordissemination of this record and the follow-up of theproblems identifled.

In the subsequent 12 months maternal deathswere reduced from 17 (13.5%) to 8 (2.9%), and thenumber of women referred for obstetrical complica-tions increased from 126 to 274. Maternity-ward stafffeel that the educational campaign has resulted inthe earlier referral of many complications. Both hos-pital and district personnel nevertheless recognizethat this level of mortality is still unacceptably high,and a programme of continuing education and sup-port through supervisory visits has been put in placeto continue improvements.

Measles in Arua District. Despite an active im-munization programme, measles cases reported in1993-94 rose relative to previous years. Using anIshikawa diagram, multiple potential causes were

WHO Bulletin OMS. Vol 75 1997 157

F. Omaswa et al.

identified. From this list of potential causes, thedistrict health team identified three areas theybelieved to be major contributors to the problem,and which they could address. These includedweakness in the district cold chain for measlesvaccine, problems with diagnostic accuracy, and apoorly functioning information system. Of the 54health centres, 9 did not have refrigerators and reliedon neighbouring clinics to store vaccine. The neigh-bouring clinics did not always have adequate spacefor the ice packs needed to ensure safe vaccinetransport. Of the units with refrigerators, 9 did nothave reserve gas cylinders. Moreover, when the coldchain was broken, health units did not have anestablished process for dealing with vaccine.

A medical officer following up reported cases ofmeasles found that a variety of skin conditions wereincorrectly classified as measles. In other instances,the diagnosis was deemed correct although somechildren had a record of measles immunization.

Clinic records frequently showed discrepanciesbetween daily outpatient tally sheets and monthlysummaries. Sometimes monthy summaries had to beestimated because of missing daily tally sheets. Avariety of corrective measures were put into placeduring July 1994, including reallocation of refrigera-tors, acquisition of reserve gas cylinders, strengthen-ing of cold-chain monitoring, and provision of anadequate supply of forms. Additional training wasalso provided to strengthen diagnostic skills. Thedecline in reported measles cases is shown in Fig. 1.

Outpatient waiting times at Masaka HospitaLOutpatients often reported early to Masaka Hospi-tal, but many did not leave until very late, and thenfrequently without treatment. Both patients and staff

Fig. 1. Number of measles cases reported from AruaDistrict, before and after introduction of the qualityimprovement programme.

100 I I

C, OU'

Co

co'O 6000l.m 40

20

May June July Aug Sept Oct Nov Dec Jan Feb Mar AprMonth

had become increasingly dissatisfied. By means of anIshikawa diagram, multiple potential causes forthese delays were identified. Subsequently, throughan examination of outpatient records, observationsof patient flow, and discussions with patients and keystaff, the following problems were identified whichthe hospital staff felt could be addressed immedi-ately: low morale among health workers, shortage ofsupplies, inadequate supervision by hospital man-agement, poor coordination of patient flow, andinefficient dispensing of drugs. The hospital manage-ment and the outpatient staff together developed awork plan in mid-1994 to address each problem area.Uniforms were provided to staff, and a work sched-ule for each staff member was established. In addi-tion, to supplement low salaries, some of the fundsraised through outpatient fees were distributedamong outpatient staff as an incentive (an acceptedpractice in Uganda). The ordering system for sup-plies functioned poorly and a new system was estab-lished. Where supplies were not available throughgovernment distribution channels, arrangementswere made for their local purchase. The hospitalmanagers developed a regular supervision pro-gramme to support outpatient staff. Duplicate andunnecessary steps were found in the flow of patientsthrough the outpatient department, and a new pa-tient flow pattern was devised. Pre-packaging ofcommon prescriptions reduced delays in dispensingmedications to patients.

By the end of 1994 long delays had been elimi-nated, and patients arriving in the morning weretreated and released by noon. Reorganization of pa-tient flow eliminated overcrowding at certain steps inoutpatient processing. Both patients and staff feltgreater satisfaction with the new system. The utiliza-tion of outpatient services in the second half of 1994increased by 46.7% over the first half of the sameyear, as shown in Table 2.

Difficulties at the district levelAt the district level the capacity to identify gaps inquality and to address them varied widely. Somedistricts identified appropriate problems which theysolved in a methodical and efficient manner, withlittle outside assistance; some districts selected prob-lems that were too complex, or over which they hadlittle control; other districts identified suitable prob-lems, but lacked the capacity for their resolution. Insome cases, visits from members of the nationalquality assurance committee did not provide districtteams with the assistance needed for solving theproblems they had chosen. None the less, most dis-tricts were able to make substantial improvements inthe problem areas they selected.


. onI

158


Table 2: Number of outpatients seen at Masaka Hospi-tal before and after quality improvements in 1994

Month No. of outpatients

Before improvementJan 3513Feb 3249Mar 3519Apr 3868May 5203June 3535 (22887)a

After improvementJuly 5650Aug 5891Sept 7166Oct 6830Nov 4534Dec 3498 (33569)

a Figures in parentheses are 6-month totals.

DiscussionIn its first 18 months, the quality assurance pro-gramme raised awareness of the importance of thequality of health services, both at the central and thedistrict level. At the central level, the integration ofnational programmes has been promoted, and thereis increased awareness of the needs and capacities ofdistrict health teams. The use of quality managementmethods by district health teams has led to bothsubjective and objective improvements in the qualityof services. One important achievement has been tobring local political leaders and district health teamstogether to improve health services. These resultsindicate that the principles and many of the methodsof TOM are applicable in Uganda and probablyin other countries with similar problems, and thathealth workers can master them.

Key factors contributing to the successfullaunching of the quality assurance programme, andto the improvements occurring during the past 18months, include the decentralization of the healthand other sectors of government, the restructuring ofhealth services, a loyal and generally well-educatedworkforce, and a stable political climate. Recogni-tion by the Ministry of Health of the importance ofimproving the quality of services led to the develop-ment of the quality assurance programme, and theircontinuing support has been critical. Another impor-tant factor has been the introduction of cost sharingin a way that directly benefits health workers andstrengthens local services. The decentralization offinancial responsibility to district councils and dis-trict health teams has greatly facilitated the optimaluse of funds. The success of the programme to date

clearly demonstrates that the principles and tools ofquality management, when suitably adapted to meetthe needs of the Ugandan health system, are bothreadily understood and appreciated. Furthermore,implementation of this programme has requiredrelatively little external technical assistance. Itshould be emphasized that the improvements wehave reported here have largely been achieved withexisting resources. However, further improvementswill require additional resources that can increaseaccess to primary health care services (18). Obtain-ing more of these needed resources from withinUganda is a major challenge.

A central lesson of the programme is the impor-tance of involving local political and governmentleaders in the quality-improvement process fromthe beginning. Decentralization means that districthealth teams and district political leaders must worktogether in setting priorities and allocatingresources.

Although results from the first 18 months areencouraging, much remains to be done before qual-ity improvement methods become a part of everydayhealth management in Uganda. The commitment ofthe Ministry of Health and district health managersto improve quality must be sustained if long-termimprovements are to be realized (19). Many districtteams have recognized the need for better informa-tion for planning, and through the quality assuranceprogramme have set about to improve data col-lection and use. However, more work is neededbefore health teams will have ready access to de-pendable data and possess the skills to use these datafor strengthening health services. A further objectiveof the programme is to increase rapidly the numberof health workers in Uganda with the ability to usequality assurance management methods. Such con-cepts must be built into health training programmesas a standard management approach. To date, 434health care personnel from all 39 districts haveattended at least one round of quality assurancemanagement workshops and have had follow-upsupport visits. Quality assurance managementtraining has now been introduced into the jointMinistry of Health/Makerere University Master ofPublic Health curriculum. A major gap in theintroduction of quality improvement methods isin the private health sector and among the manynongovernmental organizations providing healthcare in Uganda. Tertiary hospitals have yet toparticipate fully in the quality improvementprogramme.

To create a culture of quality- one which doesnot accept mediocrity - requires commitment fromhealth workers, patients, and communities. Barriersto establishing such a culture include an entrenched


F. Omaswa et al.

attitude among health professionals that they alonecan best determine community health needs. Lowhealth worker morale, low pay, and the fact thatmany health facilities are still awaiting rehabilitationalso contribute to a continuing tolerance of lowstandards. Meeting the basic needs of health work-ers, the "internal customers" of the health system,is essential to improving their performance. Com-munities, on the other hand, often have becomepassive recipients of the services provided. The re-cent establishment of community-based manage-ment committees for all hospitals and healthunits constitutes a major step towards forging thepartnership between communities and health-service providers that is essential for the continuousimprovement of services.

Ultimately, quality stems from an attitude ormindset fostering continuous service improvement.Achieving quality often requires a major shift in ex-isting thinking about health care, a shift whichUganda has begun to make. This change from theidea of health as a commodity made up of treat-ments, procedures, and training applied in definedand enforceable ways represents the birth of the con-cept of providing a service that meets the needs pa-tients and communities perceive, and that isdelivered in conformity with established standards.Not making this shift means remaining locked in thetrap of not having enough resources, yet wastingmuch of what there is through inefficient servicesprovided by unmotivated personnel.

AcknowledgementsWe express our appreciation to the hospital and districthealth staff in Jinja, Arua, and Masaka Districts for theirassistance, and to the Uganda Ministry of Health for per-mission to publish these results. The Uganda Quality As-surance project is funded through the World Bank SecondHealth Sector Loan. Methodology development was sup-ported by the USAID ARCCOS Cooperative AgreementDPE 5992-A-00-0050-00.

ResumeIntroduction de la gestion de la qualitedans les services de soins de sant6primaires en OugandaUn programme national d'assurance de la qualit6 aete institu6 en Ouganda en 1994. II a ete cree auMinistere de la Sante une unit6 d'assurance de laqualit6 qui devait servir de secretariat et diriger lesactivites d'une commission nationale d'assurance

de la qualite compos6e de responsables et decadres des programmes nationaux de lutte contreles maladies. Des programmes de formation auxm6thodes d'amelioration de la qualite ont ete mis entrain a l'intention des personnels d'encadrement auniveau central et a celui des districts.

On a constat6 des am6liorations a la fois objec-tives et subjectives de la qualite des services desante de district. Dans le district de Jinja, le person-nel de sante s'6tait alarme de la forte mortaliteobservee parmi les femmes enceintes envoyeespar les unites rurales de sant6 et a elabore unprogramme qui a permis de reduire fortementcette mortalite. A l'H6pital de district de Masaka,les patients ambulatoires devaient endurer desattentes particulierement longues; le personnel del'hopital s'est servi des methodes d'ameliorationde la qualite pour determiner quels 6taient lesproblemes, reduire les temps d'attente et accrottrela satisfaction a la fois des patients et des agents desant6. L'6quipe de sante du district d'Arua a utiliseles demarches enseignees dans le cadre des pro-grammes de formation a I'assurance de la qualitepour resoudre un probleme d'augmentation descas de rougeole signal6s, malgre 1'existenced'un programme de vaccination apparemmentbien gere.

Des difficultes sont apparues dans la miseen oeuvre du programme au niveau central, carles visites d'appui aux districts ne pouvaientavoir lieu aussi fr6quemment que prevu, et auniveau du district, parce que les problemes sesont reveles plus complexes que ne l'avaientenvisage les equipes de sante ou comportaientdes elements importants qui echappaient a leurcontrole.

Le programme a 6te couronne de succesgrace a la participation des fonctionnaires locaux,a 1'engagement sans relache des hauts fonction-naires du Ministere de la Sante, a 1'existenced'un large programme de decentralisation enOuganda, et a un personnel d'encadrement com-pose d'agents de sante loyaux et gen6ralementbien formes.

Les methodes de gestion de la qualite offrentun moyen d'aborder bon nombre des problemesque rencontrent les services de sante des pays endeveloppement. L'experience de l'Ouganda montreque les agents de sante sur le terrain peuventapprendre a appliquer les m6thodes d'ameliorationde la qualit6. Bien que ce programme ait largementmis I'accent sur les ameliorations de l'efficacite etde la rentabilite des ressources existantes, il nefait aucun doute que la prestation de services desante primaires necessitera dans certains cas desressources additionnelles.



References1. Feachem RG, Jamison DT, Bos ER. Changing pat-

terns of disease and mortality in sub-Saharan Africa.In: Feachem RG, Jamison DT eds. Disease and mor-tality in sub-Saharan Africa. Washington, DC, WorldBank, 1991: 3-27.

2. Cornia GA, Jolly R, Steward F. Adjustment with ahuman face. Oxford, Clarendon Press, 1990: 1-47.

3. Gilson L, Magomi M, Mkangaa E. The structuralquality of Tanzanian primary health facilities. Bulletinof the World Health Organization, 1995, 73: 105-114.

4. Forsberg BC, Barros FC, Victora CG. Developingcountries need more quality assurance: how healthfacility surveys can contribute. Health policy and plan-ning, 1992, 7: 193-196.

5. Bjorck M, Johansson R, Kanji N. Improving thequality of primary care services in Angola. Healthpolicy and planning, 1992, 7: 290-295.

6. Gaucher EJ, Coffey RJ. Total quality in health care.San Francisco, Jossey-Bass, 1993: 3-25.

7. The three year national health plan frame 1993/4-1995/6. Entebbe, Uganda Ministry of Health, 1993.

8. World development report 1993: investing in health.New York, Oxford, 1993: 52.

9. Shewhart W. Economic control of the quality ofmanu-factured product. New York, Van Nostrand, 1931.

10. Deming WE. Out of the crisis. Cambridge, MA, MIT-CAES, 1986.

11. Juran JM. Juran on planning for quality. New York,Free Press, 1988.

12. Crosby P. Quality is free: the art of making qualitycertain. New York, The New American Library, 1979.

13. Donabedian A. The quality of medical care. Science,1978, 200: 856-964.

14. Berwick DM, Godfrey AB, Roesner J. Curing healthcare. San Francisco, Jossey-Bass, 1991: 18-45.

15. Vuori H. Quality assurance of health services: con-cepts and methodology. Copenhagen, WHO RegionalOffice for Europe, 1982 (Public Health in Europe, No.16).

16. Roemer Ml, Montoya-Aguilar C. Quality assess-ment and assurance in primary health care. Geneva,World Health Organization, 1988 (WHO Offset Publi-cation No. 105).

17. Barros FC et al. Quality assurance of a diarrhoeacontrol programme in northeastern Brazil. Scan-dinavian journal of social medicine, 1994, 1: 35-40.

18. Okello 00 et al. The challenge to restoring basichealth care in Uganda. Social science in medicine (inpress).

19. Gaucher EJ, Coffee RJ. Total quality in health care.San Francisco, Jossey-Bass, 1993: 246-280.


introducing quality managementinto primary health care services in

Documents