Peru - Primary Health Project

rs ~~~~~ e~ The World Bank FOR OMCIL USE ONLY RepoIt NoL 12494 PROJECT COMPLETION REPORT PER'U PRIMARY HEALTH PROJECT (LOAN 2,11-PE) NOVEMBER 17, 1993 MICROGRAPHICS Report No: 12494 Type: PCR Human Resources Division Country Department I Latin America and the Caribbean Regional Office This document has a restricted distribution and may be used by recipients only in the performace of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  .,~~~~~~~~~~~~~~~~~~~~ CURRENCY EOUIVALENTS Currency Unit US$ Intis (1982 - 1991) US$ 0.75 Intis (1982) US$ -- 1.50 Intis (1983) US$ 4.40 Intis (1984) US$ 9.38 Intis (1985, January to June) US$ 14.0 Intis (1985, July to December) US$ 14.0 Intis (1986) US$ 20.0 Intis (1987) WEIGHTS AND MEASURES Metric System FISCAL YEAR January 1 to December 31 ABBREVIATIONS AND ACRONYMS ANSSA = Andlisis del Sector Salud en el Peril (Peru Health Sector Analysis) COOPOP = Cooperaci6n Popular (Popular Cooperation) ENNSA = Encuesta Nacional de Nutrici6n y Salud (National Nutrition and Health Survey) IPSS = Insttuto Peruano de Segundad Social Peruvian Social Security Institute MAPS = Modulos de Atenci6n Primaria de Salud (Primary Health Care Modules) MEB - Muebles de Equipamiento Bdsico (Basic Equipment Stands) MEF = Ministry of Economy and Finance MOH - Ministry of Health MOSBAIS - Modulos de Saneamiento Bdsico y Atencion Integral a la Salud (Basic Sanitation and Integrated Health Attention Modules) PEAPS - Proyefto Especial de Atencion Primaria a la Salud (Special Project for Primary Health Care) PHC = Primary Health Care  FOR OFCIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.SA loM of Oiudw4esmI oQmns EvatuaIgn Nommber 17, 1993 MEMORANDUM TO THE EXEC=IDnE DIRECrORS AND THE PRESIMEN. SUBJECF: Project Completion Report on Peru Pdmay Health Project (Loan 2211-PEE Attached is the Project Completion Report on Peru - Prim Health Project (Loan 2211 -PE) prepared by the Latin Amenca and the Caribbean Regional Office. Part II was prepared by the Borrower. The project sought to introduce a comprehensive model for primary health care services While it expanded health care infrastructure substantialy, the project failed to establish this health care model for lack of adequate institutional and managerial capacity, and lack of counterpart financing. During implementation, the Bank's acceptance of a major modification which futher complicated execution, contributed to this unsatisfactory outcome. Even prier to the cessation of ,U disbusements to Peru in 1987, disbursements for this project were very low. Tire project is rated as unlikely to be sustained and as having negligible impact on institutional develojrekt. 'the PCR provides a fhank and full assessment of these events. A new basic health project is being developed which takes a somewhat different approach. An audit is not planned. Attachment Ttd dcuethas a ressed disMbution and may be used by recpipents onty in the peow of Io dtSr dumci d1 conbnX may not oMenwb be dsdosed w, Wodd Bank . - Imd  PER:U FOR OMCIAL USE ONLY EMIMARYHM SPRrJAN 2211-E EYALUAMLON SUNnARX ................................. ii PART-1 P!rWect Review fr2mB4&J3 D=c1a 1. ProjectIdentitit ......................... ... I 2. Background .............................. I 3. Project Objectives and Description ...................... 2 4. Project Desip %nd Organization ....................... 3 S. Ptqet Implementation ..................... *..... 8 6. Project Results ............................ 11 7. ProjectSustnaility ............................ 14 8. Ban Performance ........................... 14 9. Eborower Perfonnanc . .................... 16 10. Project RMeltonship ....................... 17 11. ConsultingServics ....................... 17 12. Proiect Documsentation and Data ............................ 17 13. LeisonsLearned ................................. 18 PART I: JPrgjec Review from Bonrower's PXc 1. Baclgrund .................................. 22 2. Experience in thie Execution of the Inital Project ................. - , 23 3. ReformuMton of thie Initid Project ......................... . 24 4. Project Objectives . ............................... 26 5. Project Componets and Targets ....., ....................... 26 6. Goals of thie Reformulad Project ........................... 27 7. Project RaWts an/o Achievements ........................... 29 8. Negative Aspoct of Project Development ......................... 30 9. ProjectSusWaiihty ... .................... 31 10. Bank-Borrower Relafionshitp ...................... 32 11. ConclLtsions and Lessons Leanmed ........................... 32 12. Summar ............................... ........ ....34 PART m:Statistical Infornnaton 1. Esimated andActualCumuaive Dibursement .... .... 36 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents rnay not otherwise be disclosd without Wottd liank authofization.  2. Planned and Actual Allocation of Bank Financing .......... 37 3. Staffinputs ... . ...... .... .......... *..... 38 4. Use of Bank Misions .......................... 39 1Tables 1. Related Bai Las ............. ... .............. 40 2. Project Timetable ................... 41 3. Loan Disbursements ....................... 42 4. Project Implementation ...... ..... .....a... na 5. Project Costs and Financing A. ProjectCosts ........................... . 43 B. ProjectFinancing ........................... 43 6. Project Results A. DirectBenefits ................... 4 B. Economic Impact .................... .. na C. FinancImpact ..................... a D. Studies ......... .. ......... . 45 7. Status of Key Covenmts . ....................... 46 &. Use of Bank Resources A. Staff Inputs ............... . 47 B. Missons ................ .... . ..... 48 C. Costs .............. .. . n/a  ,PERU S)JCTCOMPLOMN EpR BPMARY IEALTH PROE= aDLAN 221 -PE PREFACEi This is the Project Completion Report (PCR) for the Primary Health Project in Peru, fo' which loan 2211-iPE in the amount of $33.5 million was signed on March 28, 1983. Tbe loan became effective in May 1983 and was closed on August 14, 1991, retoactive to May 1, 1991 (two years behind the June 30, 1989 scheduled closing date), when $5.8 million equivalent had been disbursed and the undisbursed balance of $27.8 milion was cancelled. The PCR was prepared by the Human Resources Division of the Latin America and the Caribbean Country Department I (Peface, Evaluation Summary and Parts I and IIM). The Bank sent the Bonrower Parts I and m on January 5, 1993. Because of the substntial time which had elapsed since the ssnson and subsequent closing of the project, we expected the Government to have difficulty in complying with ftis request. in January, the Bank and the Government held a Joint workshop with key Ministy of Health personnel and former project diectors to review Parts I and m. The Ministry used the workshop proceedings as an input for Part II, which they forwarded to the Bak on March 31, M993. Prepastion of this PCR was strted during the Bank's supervon misonS of 1988 and 1990, and is based, inter alla, on the Staff Appraisal Reort; the Loan Agreement; supeion reports; annual audit reports; coeSpondence between the Bank and the Borrower; and internal Bank memoranda. Bank staff who prepared, appraised and suprised the project reviewed the draft rport and their suggestions have been incporated in this version. i  PERU gR-OIQ MPLREPQRI PRIMARy HEALTH PROJECT (LOAN 221 l-PO EYALUMlMO SUMARX 1. The main objectives of the project, the expansion of primar3 health care services to 3.5 million people and significant reductions in mortality and morbidity from preventable causes, were not achieved. The part of the project which was given prionty in implementation was the construction of additional pimary health care infrastructure. imp] lfltadion Excde 2. The project was signed in March 1983 and became effective in May 1983. Implementation started quite slowly, and from the second through th fourth year of execuon, the project consistently disbursed about 20 - 25% of planned amountL This was due largely to Ministry of Health difficulties in uring adequate countepart contributions and delays in establishing efficient oprting procedures for implementation. n 1985, a new national govemnt frs planned to cancel the project and then decided instead to broaden it to include the trnpedo Andio, the poor mountainous Andean corridor which spans the length of Peru. The mmed project emphaszed the exansion of heat service infIrastructure through the constrin of small health posts rather than the development of a comprehensive health service model. The Bank was initaly spporive of the new government's plans for reprogrmming. Work to prepue and ratify an amendment of the legal agreement was stopped when disbursements to Peru were halted in May 1987. Only 17% of the Bank loan of $33.5 million was disbursed. The loan was closed effective May 1991. 3. The loan was used to support civil works, purchase of vphicles, medicines and materials and technical assistance and training. The i&ative shares of disbursements varied considerably from planned allocations. The loan agreement projected about equal support for civil works and procurement of vehicles, medicines and matras (about 40% each) with the balance going largey to sWport technical assistance and training(7%). In practice, two-thirds of Bank financing supported civil works, 15% supported purchase of vehicles, medicines and materials and 4% financed tec assistance and training. (See  Graph 2.) The loan asssted the Government of Peru to exnd its primary health care infrstructure substantially and to met emergency requirements for replenishing medicine stocks. It failed to establish the omprehensive model for health care it intended to introduce in order to bring about major reductions in mortality and morbidity. It also failed to strengthen the Ministry of Health's capacity to manage health care. 4. The project built 11 health centers, 173 health posts and 4 staff health center housing additions. In theory, this additional health service infastuchtre could provide services to between 800,000 and 1.6 million people (depending on the uban/rural mix). The project also provided 70 rura communities (or 260,000 people) with basic health equipment and medicine. T'hus the project potentally expanded coverage to between 1 and 2 million beneficiaries. 5. As planned, the project also introduced usel innovations which could serve as models for other health activities. It developed six health education manuals and methodologies for training- health workers. It also developed and aired at least 40 radio spots on maternal health, nutrition, breastfeeding, child diarrhea and immunization. The proect trained at least 177 fomal health workers and provided motivational ls on health to at least 8400 community members. -It introduced modular consuction designs for health posts which allowed for their later expansion into health centers. It supported human resource development in prmary health care by developing clear job descriptions for all members of the health care team. Sustnb 6. The project did not strengthen the Ministry of Health's capacity to provide or oversee primary health car servies. While it was desnd for xecution by the exsting branches of the Ministry, their admnsrative and; manageial capacity was very weak. In an effort to make project execution more efficient, the Mmnistry reassigned responsibility for execution to an autonomous agency separate from the normal Ministry stuctue in 1986. This resulted in maginaliation from the normal workings of the Frmistry and neigible impact on Ministry capacity. Findings adle ns Leamned 7. This experience provides valuable lessons for other projects ins design, ozation and participation. It demonstraes the merits of keeping m  project design simple, stardng with small changes and building on successful implementation to introduce additional changes. It also demonstrates hie importance ei appraising only'after the prject is fuly conceptualzed. 8. Major innovations should be introduced on the basis of suOcess on a small scale. For example, the design called for the recritment, taining and placemert of paid health care workers, health auxiliaries, on a large scale (4500) based on earlier smal scale experiences in the country. Before introducing a new level of salaried health worker on a large scale, the project could have exerimented on a pilot basis and expanded after careful assessment of potential problems and the development of responses. Health projects which build infrastructure and provide medical equipment should monitor counterpart budget provisions to ensure their adequacy and should plan for upkeep and mantenance requirements to ensure sustnability. 9. The project did not follow a clear plan for p;asing the series of interdependent activities which needed to be well coordinated and carefully timed (building, equipping, medicine procurement, recrutment, traing, community outreach). In implementation, this resulted in problems such as medicine being procured and delivered to health centers and posts still under construction, new health workers being recniited and trained for infrastu that was not completed, etc. 10. The project was planned for implementation by the nisty of Health without identifying and seelkng solutions to the aspects of the Ministry which functioned poorly. Insufficient attention was paid to its adminive and legal structures, and its negotiating abilities. First, in the context of tight budgetary constraints, the Ministry of Health was unable to defend its requests for counterpart funding to the Ministry of Economy and Finance. Second, the Ministry's intrnal administrve structures made the mechanics of project execution cumbersome. Third, the Ministry's legalisc approach to decision maing promoted inacin over problem solving. 11. This project yields useful lessons on o ion. Project implementation arrangements should be based on a reaisdc assessment of Ministry of Health capacity. Key constraints should be addressed explicitly during preparation. In this project, for example, reater attention to the procedures of the Ministry's General Administra Office and Legal Advisor's Office might have enabled it to avoid delays in accounting and budgeting, and might have ensured more rapid understanding of and compliance with Bank procedures. An early assessment of Ministry capacity can lead to a conscious choice to either. (a) strengthen Ministry maagement and administative capacity; or (b) ensure that all project activities can be carrWied out without reliance on the extng Mnisy srucres. The complexities of manang iv  technical assistace activities suggest the usefulness of developing an umbrella contact with an outside agency to manage the idencaion, recruitment and supervision of consultants. 12. The project also points out the importance of involving ky players early on. ZPor example, greater Ministry of Finance support for the project at an early stage might have facilitate - necssary counterpart fimding. 13. Finally, this project suggests two additional lessons on stategy. First, after two years of extremely weak performance, a new Govemment proposed a major modification and expansion of tne project which the Bank estimated would set implementation back considerably. Although Bank staff working on the project recognized that this reproganming woud not address the constraints which had led to weak perfonnance to date, the Bank expressed willingness to consider this reprogramming. The Bank should be fim in rejecting reprogramming of an already weak project which would further complicate execution. n sinmilar cases, the Bank should recommend either: (a) cancelation followed by preparation of a new project which explicitly addrs the factors that contrbuted to weak perIormance of the earlier project; or (b) scaling down of the project with the possibility of later modification only based on successful implemeritation of the scaled down version. 14. Second, when loan disbursements are suspended, the Bank should arrange to review on at least an annual basis the merits of keeping the project open, taking into consideration: (a) the commitment fees which will accrue while the project is suspended in relation to the cost of preparing a new project if the suspended one is cancelled; (b) the liklihood that the suspension will continue; and (c) the project's prior implementation record. In this case, disbursements were suspended on all loans to Peru in May 1987 becaue the national government's declared policy of limiting debt payments to a fixed proportn of export earnings placed it in arrean with the Bank. The natonal government which would be in office until July 1990 was firm in this policy and gave it a high political profile, suggesting it was unlikely to reverse it. It would have made sense to cancel this loan earlier because: (a) the commitment fes which would accrue over the suspension would be equal to the cost to the government of preparing a new project; (b) this goverment would be in place for three years following suspension; and (c) the implementation experience prior to suspension bad not been uccessful. v  PERU I"ARY HEALTH LPROlEa (LOAN 2211-PE) 2ABIJ: Pject evew from Bank's Perspect 1 Proect Identi Name: Primary Health Prject Loan Number. 2211-PE RVP Unit LA1HR Country: Peru Sector: Health Subsector: Pimary Hth Care 2 Bnckgroimd 2.1 Badevye1gmz gbj.gdv: When this project was prepared, the Government of Peru had just embarked on an initiative to introduce a new model for primary health care with the objective of ensuring unversal access to primary health care by the year 2000. Tius model was based on the Primary Health Care Module-a defined package of resources to provide prmary health care sevices to poputon groups according to their size, health rsks and resources avae. This model was in sharp contrast to Peru's publi health care system thogh the mid 1970s, which emphasized curative care and concentratedinfructure and human rouc in u centes, lagely in hospits. In 1980, Peru adopted a Natona Plan for 1982-1985 which sought to reduce mortaity and morbidity (the rate of death and illness) through increased emphasis on primary health care and began to undertake major modifications in the organizaton of the Ministry of Hieath to acbieve these iovements. 2.2 FPlicy contet The national government in place during prepaation of this project sought to achieve eonomic efflciency, remove restrictions on and promote the private sector and create a more equiable sharing of delopment benefits. It placed high priority on eatng loyment, lieraizing trade, promtng eports and fighting inflation.  Serious poverty problems and fiscal constrints which adversely affected social sector spending contributed to the appeal of pmary healtfi care as an efficient, low-cost model for health care delivery. The national government in place during project implementation from 1985 on placed msjor emphasis on redressing neglect of the trapecio Andino, Peru's Andean corridor. 2.3 Linkages a aroject objecdves, sector objctive and mac plicy: The project objective of expanding priury health care services to reach 3.5 million people underserved by existing health care services supported the government's objective of promoting improved health care access and improved health status for the entire population. It closely paralleled macro policy objectives of sharing development benefits more equitably and using scarce fiscal resources efficiently to deliver social services. 3 Lject ObWectives and Dcription 3.1 Pmiect gbjeives: The principal objective of this project was to iese the availability of and access to prmary health care services for about 3.5 million people in four areas of Peru in order to reduce mortality and morbidity. It aimed to reduce mortality from: (a) diarrheal diseases by 20-40%; (b) respiratory diseases by 10-25 %; (c) vaccine preventable diseases by 50-70%; and (d) complications of pregnancy by 20-50%. It also sought to reduce the frequency and swverity of morbidity (illness) from: (a) diarrheal diseases by 20- 50%; (b) respiratory infections by 30-50%; (c) vaccine preventable disae by 30-70%; and (d) sepsis and toxemia by 20-40%. It aimed to reduce the severity Of malnutriion by 20-40% and to increase the number of family planning users by 3-20%. The ranges in the targets reflected diffences in eistng lvels by region and age group. As baseline information for these indicators was not provided in the SAR and project documents do not indicate that they were monitored over the life of the project, it is impossible to compare levels at the beginning and end of the project. Baseline information was to have been collected by the project, yet the ten different outcome objectives with four different regional targets for each one were more complex than necessary to measure project impact. 3.2 Compe: The project had seven components: (a) organizaion and delivery of primary health care sevices; (b) health manpower taining; (c) community health education and picipation; (d) monitorng, supervion and evaluation; (e) maintenance of buildings, fiunitwe, equipment and vehicles; (f) institutional development; and (g) project adminion support. 3.3 Am: The project was designed for implementation in four 2  Health Regions including Lima. In the Health Region of La libertad, it was to include Pacasmayo and Trujillo; in the Health Region of Centro Medio, Junin, Tar, Jauja, Huancayo and Huancavelica; in Lima, Rimac and the Cercado de Lima; and in Huanuco, Pucallpa. These were selected to represent the four principal types of areas-the selva, sierra, coast and the capital city. This was seen as crucial to testing the viabilty and adapting primary health care services for different areas. From 1986 on, implementation was expanded to five other Departments, Puno, Cusco, Apurimac, Cajamarca and Ancash, because of increased government priority on the dpecio Andino, a mountainous coridor of poverty which spans the length of the country. 4 Wjaa Deign and Organizatin 4.1 eQntbaL foundation: The National Primary Healti Care Policy, which emerged from project preparation and was adopted by the Ministry of Health in January 1982, pxovided the foundation for this project The project was based on the development of moduks, packages of resources (human, technological, physical and financial) needed to provide primary health care to about 2,400 families (15,000 population) in scatted rurl setfings or about 6,400 families (40,000 population) in clustered urban areas. The health center was to be the base of each module, complemented by four to six health posts, each staffed by a registered nurse or other health professional. Below the health post, each module was also to include 120 to 150 rijehaes, or volunteer health motivators, and 24 to 32 salaried health auxiliaries. The health care approach of these modules emphasized: (a) community partipation in decision making and resource allocation for services; (b) health professionals to perform functons requiring clinical judgement and delegation to non-professionals of safe and s functions; (c) systematic home visiting by field nursing auxiliaries; (d) use of symptoms and risk factors to screen individuals and determine the appropriate health worker to provide care; and (e) coordination with the development activities of other Ministies. Health auxilaries were to make household visits every two months to every family in their jursdiction. The basic features of the module are set forth in the table below. The elements of this approach had been previously tested in pilot projects in Peru. This project was to.be the first phase toward nationwide adoption. 3  Basic Feahres of Primay Health Care Module Lev:1 Health Act .: Urbaw Rural: I *~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ r~~~~~~~~~~~~~~~~~ _ _._ _ __saff.- Ppi.: Stff: : HeaO Cuat - mdia corsulation 27 6,400 22 2,400 - clinical lab (5 Mds, 1 families; (3 Mds, fhmilies; - x rays dentist) 40,000 112 time 15,000 - ambulatory popi. dentist) popin. professional care - assessmnt of popultion heal and nutrition stats - resourc allocation Hel'bNst - short ten cae of 1 1,200 1 600 sick registered famiies; regitered families; - groups for health nure 7,500 nurse 3,750 education popin. popln. - training of icharies - oontinuing education of auxiiaries - quarterly review of health atu and Community Healt - immunizations 1 200 1 100 Gmp - nutrtion monitong Auiy famili; amliay fmis; - diarhea control (now paid 1,250 (new paid 625 - deparasiltation worker) popln. worker) popln. - ratme of comnon child Mness - vion sreening for shool oidren - pro-nata screening for risk - low risk deliveries - family Planning infomuation - hygie Community HealIh 1 Rijhary 40 1 rnjohasy 20 Unit (volunter families; (volmteer families; motivator) 250 moivator 125 ___________ .________ _I pop2n. ) popLn. 4.2 Gosensus on Met ==ip framework The concept was _nderst by the leadership of the Ministry of Health (MOH). Health personnel from the four project regions particpated in the prpation process and the cental Ministry team made swveral visits to these regions during 4  pIrparation. Preparation was coordinat with other major donors, yet consensus did not eist on the use of salaried health axiliaies nor on the approa of periodic family health visits Bank staff felt that intensive outreach was vital to achlevig quick and significant reductions in morbidity and mortalit and tha only salaried workers realistically could be depended upon to dedicate the time necesary to undertake this work. Bank staff also considered that the project provided the Ministry an opportunity to test this model in four areas to veify its effectiveness for wider expansion. Bank staff did not consider disagrement among donors over approaches to be problematic because of the experimental nature of the project. 4.3 Innovative elements of Lrject design: This project was the first Bank project in Peru exclusively in the health sector; it was innovative in its concepion. It sought to bring about a major transfonration in health sevices through the introduction of a model to embody the national policy on primary health care. It developed models for infrastructure for three different types of geographic areas (coast, highlands and jungle), established a family-oriented outreach approach to service delivery and developed job descriptions for profesional and community health workers to clearly identify their respective roies. nas designs for health posts were modular to allow easy upgrading to health centers as needs changed. 4.4 Project scope and scale a Oateness: The project sought to introduce innovations smulously on a number of different fronts- infrastruce human rources, community outreach-and for a variety of circumstances-the rural sierra, jungle, coast and the margnal areas of Lima. Ihis varit and complexity were too great for the administ and manuagent capacity in the Ministry, particularly because this was the Ministr's first Bank financed project and the first effort to develop a compreheive pdmary health care system. This meant that the Ministry was leaning and adopting Bank procedu for ement, disbursement and consultant contracting at the same time that it was supposed to be achieving major substantive targets in project implementaton. In retospect, tie scope and scale of the project were too large for extng circumstances. The Ministry of Healdth had neither successful experience in the delivery of pnmary health care services nor an established model. 4.5 The complexity of the model may be challenged in two respects: (a) the model alW for home viits to each family every two montis which required hiring this large number of new personnel; and (b) the model did not limit atention to priority target groups among the population (for example, pregnant women and children under 3) in recogidton of the scarcity of human, matefial and financial reources. The project also called for the creadon of a new cadre of health workers-4,500 field nursing auiliaries-rather than the S  reorientation and retraining of existing health workers. While this increased the complexity of the project, the assessment of Bank staff at thfe time was that retAining of existing health workers was not a viable option since the absolute number of health workers was inadequate for the expansion of health service coverage. The project design overestimated the management capacity within the Ministry of Health. The SAR included mortality and morbidity indicators which were not clearly defined and for which baseline data were not readily available. 4.6 Ouality of project peparation: The Bank provided substantial mission support for project preparation (177 staff weeks through appraisal including 106 staff days on 10 missions with experse in economics, health, architeclure and education). Two aspects of the project were very well developed at appraisal. The model for care specified clearly the infrastructure and human resources at each level from the community to the health center. Job descriptions were developed to define the rewspective responsibilities of each health care worker. A number of other cntical aspects, some of which were discused and agreed upon in principle with Ministry of Health counterparts, were not defined by appraisa or explained in the staff appraisal report. These included: (a) how communities would participate itl services; (b) how health services would be organized and supervised within the modules; (c) how new field health auxiliaies would be compensated; (d) how current MOH sing- disproptionately urban based-would be shfifted to fit the new model of services; and (e) how the MOH and the IPSS would coordinate to attain a more rational pattern of service provision. 4.7 Following appraisal, further preparation work was undertaken in one area, the development of health education manuals for different levels of health care workers. This included adaptation of materials to the different situatons of the coast, jungle and highlands and field testing. Liftle operational planning was done in other areas between appraisal and loan effectiveness. The PPF was never used to support operational planning because the local counterpart contribution was not provided. Managerial problems within the MOH also impeded progress in developing detailed operational plans. 4.8 Aopriateness f project timing: In principle, the project tming was very appropriate because it immediately followed and supported the Government's formulation of a National Plan of Coordinated Health Actions (PNACS) for 1982-85 with a strong emphasis on PHC. In practice, fiscal constraints which arose after loan approval placed severe restictions on the countepart funding. A change in the national government in 1985 led to a major shift in the Ministry of Health's strategy for expanding primary health care services. These events came after project preparation and could not realitcally have been anticipated. 6  4.9 Definition of roles and responsibilities: The responsibility of the MOH for project execution was clearly defined and understood, however the Ministry's identified role taxed its capacity, which the Bank team acknowledged to be weak. The project design called for the establishment of a small technical unit to coordinate project implementation activities by line departments of the Ministry. The relationship of this technical unit to existing Ministry divisions was carefully defined. Final responsibility for project oversight was placed with the Vice Minister to ensure the high level attention necessary for the project's success. The responsibilities of the health system below the central Ministry level were less clearly defined. Regional health authorities did not understand their roles in the project initially. 4.10 Impact of defined responsibilities on implementation: The institutional arrangements for project implementation were carefully planned to ensure the project's integration into the existing organizational structure of the Ministry and the high level attention its success required. Yet existing procedures and information systems made accountability for project implementation weak. In 1986, without Bank participation or approval, the Ministry of Health altered the implementation arrangements completely by creating an autonomous body to execute the project. This not only created a tremendous bureaucracy (with a staff of 363 and annual personnel expenditures of $1 million in 1986), it also isolated and marginalized its execution from all other Ministry of Health work. This jeopardized the sustainability of the change the project was intended to bring about in the delivery of health services. It also meant that any mechanisms which the project introduced to improve efficiency had no impact on the Ministry's normal procedures. 4.11 Asc of project design and orgaization which most affected PrQoAec outcome: Despite the Ministry of Health's strong commitment to primary health care and to the project, the project had limited success in introducing a com?rehensive new health care model and improving health conditions for a large number of Peruvians. 4.12 Design: The design called for the recruitnment, training and placement of a new level of paid health care workers, health auxiliaries, on a large scale (4500), and required that a series of interdependent activities be well coordinated and carefully timed (building, equipping, medicine procurement, recruitment, taining, community outreach). In practice, the Ministry's weak management capacity resulted in problems such as medicine being procured and delivered to health centers and posts still under construction, new health workers being recruited and trained for infrastructure that was not completed, etc. 4.13 Organization: The project was planned for implementation by the Ministry of Health without identifying and seeldng solutions to the aspects of the 7  Ministry which functioned -poorly. Insufficient attention was paid to its administrative and legal structures, and to its ability to prewnt the project to the Ministry of Economy and Finance. First, in the context of tight budgetary constraints, the Ministry of Health proved incapable of defending and protecting necessary counterpart funding. Second, the Ministry's internal administrative structures made the mechanics of project execution cumbersome. Third, the Ministry's legalistic approach to decision maldng promoted inaction over problem solving. 5 Project Implementation 5.1 Critical variances from plans: Critical variances occurred between planned and actual project implementation in virtuly all components. Some modifications were innovative and contributed positively to the project. For example, packages of medicines were developed according to the level of spialization of health establishments. Special -muebles de equipamiento basico' (MEB-a combination examining table and medicine cabinet stocked with basic health care equipment and materials) were designed for communities where the construction of health posts was not feasible to provide community health workers with the necessary requirements to carry out community health activities. Modified versions of these MEB, called bodquines comnunes (community pharmacies), are in widespread use in rural and marginal commumnties now. 5.2 Other variances were detrimental to the project's success. Project implementation can be seen in three phases: Phase I, from 1983 to 1985; Phase IT, from the new government's proposal to reprogram the project in 1985 to suspension of disbursements in 1987; and Phase m, from suspension of disbursements in 1987 to cancellation of the loan in 1991. In Phase I, because of the Ministry of Health's failure to secure adequate counterpart funds, project execution was so slow in starting that by October 1984 the Bank was already weighing the merits of suspending disbursements or scaling down the project These options were bypassed in favor of waiting until April 1985 to see if progress improved. This problem was resolved to a certain extent by improvements in counterpart contributions for 1984 and the introduction of a Special Program through December 1985 which increased the Bank's share for civil works from 70% agreed to at appraisal to 90%. By 1984, one year after effectiveness, the Special Account had not been opened. External auditors for 1983 and 1984 found the project recordkeeping system so inadequate that they were unable to state an opinion as to the reasonableness of financial statements and results due to the lack of journals, annexes, monitoring of cost of works, and outstanding advances. The project did not maintain separate accounts. Expenditures for other projects were charged against this project. The project 8  lacked a separate system of controls and kept neither project files nor inventory control. 5.3 In Phase II, project execution was stopped completely in mid- 1985 amid the change in national government. After initially announcing its desire to cancel the loan, the new govemment decided instead to reprogram the project in late 1985 and presented an initial proposal to the Bank in December. The reprogrammed project which the new government began to implement in 1986 shifted emphasis to the rapecio Andino and focused on the expansion of health post infrastructure, rather than on the development of an integrated health care model. Construction designs were modified to reduce costs. While these modifications might have been positive if they made greater use of less expensive labor and less use of more expensive materials, in practce the high labor requirements of constructing with qiwncha and adobe combined with the high overhead for supervizion by a new autonomous project agency led to higher construction costs than originally anticipated. 5.4 The Ministry of Health established the autonomous agency (PEAPS), which was formally instituted by Presidential Decree in May 1986, to execute the project. PEAPS could sign legal agreements with public and private agencies and had technical, administrative, economic and financial autonomy from the MOH, but this structure was never fully ratified. The Ministry of Health rever approved its organization chart; the Ministry of Labor never provided necessary approval of its organizational manual; and PEAPS never named its Board of Directors. The Bank was preparing to submit an amendment to the loan agreement for Board approval of the reprogramming when all Bank disbursements to Peru were suspended on May 5, 1987. 5.5 In phase m, project execution was severely hampered by the suspension of Bank disbursements. Nonetheless, the Government demonsatd its commitment to the project by continuing and even increasing its contribution to the project Ifrom 1987 through 1990. 5.6 Possibility of avoiding these variances: The initial slowness of project implementation might have been possible to avoid had the Bank sought assurances of budgetary allocations to the project for the first year of implementation as a condition of Board presentation or effectiveness. The major vanances between the project as planned and as implemented which occurred after the new government reprogrammed the project in 1986 could have been avoided only had the Bank conveyed a clear message that it would not consider a reprogramming proposal which would expand the scope of the project and shift implementation into even more difficult areas. This would likely have led to the cancellaton of the loan. Cancellation could have been justified by the 9  poor implementation record of the project in its two years of implementation. Bank staff noted in September 1985 that the proposed reprogramming could imply furnher delays in project completion when the project was already 2.5 years behind schedule. Bank staff also noted, however, that the new government was very enthusiastic about making a major push to expand primary health care, the areas where the government wished to work were underserved, the government had high level support for the project and assurances of counterpart funds which had not existed before, and would not have been interested to continue a scaled down version of the existing project. 5.7 Identification of project risks: Project risks were correctly identified as: (a) management wealmess of the Ministry of Health; and (b) the need to successfully introduce a new health strategy. The seriousness of these risks was underestimated. At appraisal, staff noted some meassures which would protect against these risks: (a) the project's limitation to a relatively small portion of the country; and (b) the conceptual simplicity of basic elements of the scheme. In retrospect, management weaknesses of the Ministry of Health were an important factor in poor performance. The introduction of a new health strategy adapted to four radically different settings (coast, highlands, jungle, vmarginal urban area) was more complex than anticipated and inhibited implementation. The difficulty of supervising a complex project in four substantially different types of areas widely dispersed throughout the country could have been foreseen and could have led to a phased approach, beginning with areas where implementation would be easier and expanding to more challenging areas on the basis of experience from the first phase. 5.8 Impa of un-foreen- factors on project implementation: A significant deterioration of the security situation in several project areas affected project supervision, protection of completed and equipped infrastructure and staffing. One project engineer was murdered, two new health posts were taken over by terrorists, one project vehicle burned, and construction of 11 incomplete health posts abandoned due to terrorist activities. This could not reasonably have been foreseen in 1982. Resistance of physicians and nurses to the participation of non-formal health auxiliaries in the health system could have been foreseen. A concerted strategy could have been developed to gain their approval and support by demonstrating the advantages the project potentially held for them (such as improved infrastructure to work in, improved supplies, and additional opportunities for training and scholarships). 5.9 Actions and decisions which affected project implementation: Many actions and decisions affected the project negatively. In phase I of implementation, the decision by the Ministry of Economy and Finance to curtail the public treasury contribution led to slow start up and limited achievements. In phase II, the government's decision to reprogram led to a shift in emphasis 10  away from a comprehensive model for health care. The move from the selected four project areas to the trapecio Andino led to more difficult implementation and supervision and greater security risks. The 1986 decision to set up an autonomous body (PEAPS), while intended to improve efficiency in project procedures, jeopardized the sustainability of project impact. PEAPS grew to have over 300 staff; its salary payments (at levels above those of the Ministry) came to absorb 90% of the project budget after Bank suspension. The mushrooming of this autonomous agency drew competent Ministry of Health staff away from other Ministry functions, led to resentment on the part of Ministry staff who were not as well paid, and led to reluctance on the part of the Ministry of Finance to continue supporting the project after Bank suspension. The project decision for COOPOP, a government agency which promoted community participation in projects nationwide, to participate in construction of health posts was intended to promote greater community participation. In practice, COOPOP's involvement resulted in delays and high construction costs. When the project agreement with COOPOP for construction was rescinded, the autonomous agency's decision to undertake construction direcdy also resulted in high construction costs because of the agency's high overhead spread over a low volume of construction. On the other hand, a project decision to have UNICEF provide supplies directly to the health center was positive because it avoided logistical problems of distributing supplies to dispersed centers. 5.10 Actions not taken which affected implementaffon: The actions not taken by the MOH which most affected implementaton were: (a) not assigning adequate counterpart funds for the PPF, which impeded detailed planning and preparation prior to effectiveness; (b) not recruiting the technical assistance which was provided for in the project; and (c) not developing a clear and stable project management structure with adequate authority and technical expertise. 6 Poect Results 6.1 Achievement of Objectives: The project did not achieve its objectives. Variances between planned and actual results can be explained largely by the ambitious scope and scale in light of previously mentioned constraints; the inability of the MOH to establish an effective project management structure with continuing support from the ministry level; and inadequate GOP financial resources and uncertainties and delays in their provision. 6.2 Changes in Objectives: Although project objectives did not change during implementation, the strategy was modified considerably in implementation to place primary attenior. on expanding infrastructure. The 11  govremment elected in 1985 proposed to the Bank in 1986 that: (a) the project area be expanded to include the trpecio Andino; (b) the number of modules be increased from 131 to 250; (c) priority be given to new construction; and (d) construction designs be modified to use low cost local materials. The MOH created an autonomous management unit to execute the project. The reogrammed project sought to provide primaty care services to 6.7 million people rather than 3.5 million as estimated at appraisal. Although Bank missions were supportive of these proposed changes, no formal amendment was made to the project loan agreement and disbursements were suspended before the amendment was prepared and ratified. After the Bank's agreement in principle in April 1986 and its expression of willingness to amend the legal agreement in November 1986, the new strategy was adopted in practice by the Peruvian government. The expansion led to more health posts being constructed than orginaly planned and in more widely dispersed areas. 6.3 Explanation for variance between planned and actual results: The project was planned to result in the systematic establishment of primary health care modules which would serve 3.5 million beneficiaries. Instead of leading to this systematic introduction of primary health care, it made some isolated advances. It constructed some primary health care infrastrucure, introduced new materials for health education, disseminated some health messages and led to the recruitment and training of additional auxiliary health workers. The variance is due to the early inability of the Ministry of Health to secure adtquate national resources to complement the Bank loan, the later shift in Government priorities toward expanding physicil infrastructure, and the persistent inability to manage a series of activities which were conceptually simple but complex to implement systematically. 6.4 Eect ImpUac: The project was not implemented as it was designed and did not achieve its intended health impact. Nonetheless the investments it supported did expand available primary health care infrastrt, although in different areas and in different combinations of centers and posts than originally planned. It built 11 new health centers, 173 new health posts and 4 new health center staff housing additions. Using initial project estimates of coverage and assuming adequate staffing and other inputs, these health centers and posts could provide services to between 800,000 and 1.6 million people (depending on the urban/rural mix). The project also provided 70 rral communities (or about 260,000 people) with basic equipment and medicines. Thus the project potentially expanded coverage by between 1 and 2 million beneficiaries. The project developed the designs and technical plans for two health equipment maintenance centers, although these were not built. 6.5 The project also provided medicine to 184 centers, 773 posts and 70 communities. Again using the original project coverage estimates, the 12  project supplied medicine to facilities capable of serving between 6 and 13 million beneficiaries. The project also provided 33 vehicles; their potential impact on services was limited since most were used for supervision of construction, one was destroyed by terrorists and one was stolen. 6.6 The project led to the development of manuals and methodology for training health workers as planned. The six manuals developed were "Education for Health: Ideas for its Development in Primary Health Care and Basic Health Services" and training manuals for: (a) Volunteer Health Promotors; (b) Field Auxiliary Nurses; (c) Health Post Nurses; and (d) Health Module Physician Directors. The project also developed and aired at least 40 radio spot messages on maternal health, nutrition, breastfeeding, child diarrhea and immunization. The project trained at least 177 formal health workers and auxiliaries and provided motivational talks on health to at least 8400 community members. 6.7 The project participated in the formulation of plans for the ENNSA, the national survey on health and nutrition which was conducted in 1984. Terms of reference developed by the project for a health finance study were incorporated into the 1985/86 health sector analysis supported by the U.S. Agency for International Development (AID). 6.8 The project influenced health sector policies in several ways. It supported and deepened the Government's nascent commitment to promote primary health care. By monitoring and promoting Government compliance with loan agreement covenants, it: (a) led to the introduction of salried community health workers (the health auxiliary); (b) encouraged the integration of the Ministry of Health and the Peruvian Social Security Institute health services; (c) promoted incentives for personnel assigned to depressed zones; and (d) encouraged decentralization of management and decision-maling in the health sector. Sustained impact on policy is less clear since the Ministry has backtracked on some of these policy initiatives since. 6.9 The project supported human resource development by developing clear job descriptions for all the members of the primary health care team (physicians, nurses and auxiliaries) and by developing, field testing and utilizing training materials for them. It also promoted incentives to attract human resources to the rural areas and primary health care. The project's capacity to foster human resource development further through scholarships and technical assistance was not fully exploited. Project managers may have seen scholarships and technical assistance as costly items which would be easy to do without. 6.10 The project's impact on the physical environment is limited to the apropriate placement of health care infrastructure through careful pre- 13  construction topographical studies. The project contributed to the technological environment by developing the MEB, an appropriate technology for providing community health care where health post construction was not a viable option. It also demonstrated the feasibility of utilizing local construction materials and techniques (quincha, adobe) and more labor intensive methods for health infrastructure. 6.11 It is unlikely that the project had a significant impact on the social environment by affecting poverty levels or income distribution. The project's potential for strengthening the central Ministry or regional health institutions was diluted by the creation of a separate autonomous executing agency during implementation. 7 ectA $ustainability 7.1 Factors affecting outcome: Despite the project's success at expanding infrastructure and providing medicine to primary health care centers and posts, a functioning PHC system does not exist on a national level as an outcome of the project. Two key issues call into question sustained impact, namely the capability of the MOH to lead and direct the development of PHC services; and the ability of the Government to provide adequate financial resources for primary health care. 7.2 Ways to address risk factors: Project sustainability could have been strengthened by emphasizing institutional development in the Ministry of Health, executing the project through the normal Ministry channels as originally designed, involving regional authoritics at an early stage, and introducing innovations in separate phases and on a smaller scale. 8 Ba&nk hf=ance 8.1 Strengths and weaknesses of Bank performance: The main strengths of the Bank's performance were: (a) substantial technical and planning assistance to MOH officials responsible for planning and implementing the project and (b) continuity of an experienced, professional and understanding staff. 8.2 The main weaknesses of the Bank's performance were: (a) the decision to appraise before crtical planning and decision making were completed; (b) inadequate assessment of the complexity of the model and its relationship to other donor-supported PHC work; (c) inadequate assessment of the management weakmesses within the Ministry of Health and how they would 14  hamper implementation; (d) unclear messages to the Govemment about the Bank's likely response to the reprogramming proposed in 1985/86; and (e) excessive reluctance to cancel the project loan. 8.3 In general, the Bank performed well during the early stages of the project cycle by providing substantial assistance to those preparing the project. 8.4 The Bank's performance during supervision was good through 1985, based on the number and composition of missions. Bank staff provided frequent attention to the project, conveyed a supportive attitude and maintained good reporting. Scoring of project progress and problewas was too lenient. 8.5 In 1986, the Bank's communication with the Government about reprogramming was not as immediate and clear as it could have been. In an April 1986 mission, Bank staff expressed a willingness to consider a reprogramming proposal. This willingness was probably mistaken because project implementation had been unsuccessful largely for structural problems which the reprogramming would not address. The scale of the project was already noted as too ambitious and the reprogramming would expand it fiuther. Project records do not indicate that Bank staff explained to the Government immediately that no expenditures incurred on the basis of the new model would be recognized by the Bank until after Board approval. Project documents also do not indicate that they explained immediately that the Bank would not accept the inclusion of areas that Bank staff could not supervise directly. In November 1986, Bank staff advised Government counterps that Board approval of the amendment would probably talce three months (i.e. February 1987), yet when disbursements were suspended six months later in May 1987 no amendment had been presented to the Board. After the 1987 suspension of disbursements, the Bank's four year delay in cancelling the undisbursed balance of the loan was axcessive and the Government incurred over $800,000 more in commitment fees without ever regaining use of the loan. 8.6 Main lessons: The primary lessons are: (a) pay greater attention to the appropriate project scale in relation to project, sector and country conditions; (b) do not assume that the provision of TA is a panacea or that it can substitute for informed decision maldng in critical areas by responsible officials; (c) ensure that all essential components of a project are carefully planned and reasonably assured of success before appraising; (d) involve and draw on the expertise of parties whose approval and support is crucial to project performance in the planning process at an early stage ( for example, legal advisors, administrative officers, Ministry of Finance officials); and (e) even when the design is simple conceptually, consider carefully how complex the project will be to implement at all levels. 15  9 agrwe Piomance 9.1 Strenuths and weakses: The main strengths of the Govemment's performance were the strong interest at various levels of the ministry in implementing a new health policy emphasizing primary health car and the participation of many competent, committed persons in the project. The main weaknesses were: (a) the inadequate planning and management capability and inflexible and burdensome internal admninistrative controls; (b) weak coordination among MOH units; (c) insufficient negotiating skills to ensure countepart funding from the MEF; and (d) mid-course changes in approach. 9.2 Borrower staff worked conscientiously at the preparation stage but lacked staff time and experience to do detailed planning for project implementation. Subsequent to project preparation but before implementation, the lack of local financing prevented MOH staff from carrying out tasks which were to be accomplished under the PPF. The lack of a strong project management group added to the difficulties. During project implementation, the borrower had s.gnificant managerial, financial and legal problems, all of which combined to seriously impede effective implementation. 9.3 Use of staff and consultant resouc.es: Except in the area of training, the MOH did not recruit consultants to assist with the project, particularly in management and budgeting. In the absence of a strong MOH commitnent to improve in these areas, their recruitment would have had litde impact on MOH policies and procedures. 9.4 Mail lgssons: Three main lessons can be drawn from this experience for the borrower: (a) first, the MOH should do some detailed preliminary planning and budget projections, and identify critical policy and other decisions which must be made for tie project to be successful (for example, funding for a substantial number of new positions, a realistc plan for transferring urban based personnel to rural areas; the implications for the central ministry and the feasibility of giving increased authority to regional and local levels); (b) second, the Ministry should test new health service strategies in limited areas in circumstances ensuring availability of all necessary resources and draw heavily upon expenences of pnor pilot projects before introducing them nationwide; (c) third, the MOH should assign its legal and administrative staff to learn Bank procedures anld iron out any incompatibilities foreseen as early as possible. This process should be supported by the central government agencies responsible for defining government procurement and accounting regulations. 16  10 ProWeA Relation-ship 10.1 Bank-BorrowerWeTlationship: Bank-MOH relationships generally were good. Communications were frequent and collegial. Problems in project implementation arose more from the situation of the health sector in general and from the severe financial constraints facing the GOP than from direct project actions. If the Bank had encouraged a closer relationship between the Ministry of Health and the MU, local financing might have been forthcoming in a more tiroely fashion. 10.2 Bank relaffonships with other parties: Bank staff and consultants could have sought closer worldng relationships with other donor agencies providing assistance in PHC. A consensus message from the donors might have contributed more to strengthening the Ministry's managerial capacity and organization. 10.3 Borrower's relationships with other parties: MOH relatbinships with the MEF and the IPSS were not adequate to secure required local financing and to reach early agreement on rationtion of health services between MOH and IPSS. At the reprogramming exercise undertaken by the subsequent government, the MOH initially was able to secure a very substantal increase in financial resources for the health sector and to male progress on an agreement with the IPSS. 11 Consuting Services 11.1 The Bank used consultants from project przparation on. The MOH used consultant services in much lower volume than anticirated at apraisal, and only in training. The substantial delays the Ministry experienced in contracting consultants for management and budgeting suggest that either the project directors were not convinced of the value of consultant serces, or the arrangements for hiring consultants required administrative mechanisms not easily avaiable within the MOH. To avoid the former, the Bank should address the project implementing agency's concerns with consultants explicitly. To resolve the latter, the Bank could encourage the Ministry to set up umbrella arrangements for administering consultant services with experienced agencies (for example, UNDP, PAHO, large consulting firms). 12 Project Documentation and Data 12.1 L&galagreement: The legal agreement was well prepared and included covenants to address key issues for the project's success. It might have 17  been helpful to include a provision that the GOP commit a specified level of financing for the first year of project implementation as a condition of effectiveness. This would have supported the Ministry of Health's budget request to the MEF. 12.2 Staff Appraisal Report: The SAR provided a usefil framework for the Bank during project implementation. It is not clear that it served the Borrower to the same extent. The SAR's presentation of health outcome indicators was not as useful as it could have been for monitoring project achievement. It included numerous health outcome indicators disaggregated by age group, population subgroup (e.g. pregnant women) and region, each expressed as a percent change from existing levels. Yet iio information was included on baseline levels. Some of the targeted reductions would not be possible to detect even with a good information system and universal udlization of services. Other indicators would have been difficult to monitor because they were not clearly defined (for example, what does a percentage reduction in the severity of malnutrition mean?). In retrospect, targets could have been fewer, more sensitive to change and capable of measurement in the timeframe of project implementation. 12.3 Availability of data relevant to PCR: The SAR included a long list of targets for each of the project's seven components (number of centers and posts built, equipped and supplied, months of technical assistance, number of community volunteers trained, etc.). Data on the achievement of these targets were not all available. From project reports and supervision reports, it appears that these indicators were not interpreted or monitored consistently during project implementation. Project reports at different times appear to measure a variety of different indicators and also to apply different intexpretations to the same indicators. This makes it difficult, for example, to consolidate the annual numbers for training and reach meaningful conclusions. This suggests ta it would have been useful for the Bank and the Ministry to have developed and consistently applied a format for monitoring project implementation complete with definitions of key terms (such as criteia for considering someone "trained", for example). This format could have been used to monitor progress over time and could have formed the basis for semiannual Borrower project reports to the Bank and for Bank supervision reports. This would have provided a good picture of project implementation over time. 13 Lessons Lened 13.1 This project provides valuable lessons for other projects in design, organization and participation. It demonstrates the merits of keeping project design simple, starting with small changes and building on successful learning 18  to introduce additional changes, and appraising only when the project is fully conceptualized. 13.2 Major innovations should be introduced on the basis of success on a small scale. For example, the design called for the recruitment, training and placement of a new level of paid health care workers, health auxiliaries, on a large scale (4500). Before introducing a new level of salaried health worker on a large scale, the project could have experimented on a pilot basis and expanded after careful assessment of potential probh- ms and the development of responses. 13.3 Health projects which build infrastructure and provide medical equipment should include counterpart budget provisions and plans for their upkeep and maintenance to ensure sustainability. While this project included them as part of the recurrent expenditures needed for operation, the government did not provide them. 13.4 Project design should inclth%ie plans for phasing interdependent activities whose introduction needed to be well coordinated and carefully timed (building, equipping, medicine procurement, recruitment, training, community outreach). This project's failure to phase activities adequately resulted in problems such as medicine being procured and delivered to health centers and posts still under construction, new health workers being recruited and trained for infrastructure that was not completed, etc. 13.5 Institutional strengthening is vital for project success. This project was planned for implementation by the Ministry of Health without identifying and seeldng solutions to the roles of the Ministry which functioned poorly. Insufficient attention was paid to its administrative and legal structures, and its negotiating abilities. First, in the context of tight budgetary constraints, the Ministry of Health proved incapable of defending and protecting necessary counterpart funding. Second, the Ministry's internal administrative structures made the mechanics of project execution cumbersome. Third, the Ministry's legalistic approach to decision-making promoted inaction over problem solving. 13.6 Project implementation arrangements should be based on a realistic assessment of Ministry of Health capacity. Key constraints should be addressed explicitly during preparation. In this project, for example, greater attention to the procedures of the Ministry's General Administration Office and Legal Advisor's Office might have enabled the project to avoid delays in accounting and budgeting, and might have ensured more rapid understanding of and compliance with Bank procedures. An early assessment of Ministry capacity can lead to a conscious choice to either: (a) strengthen Ministry management and administrative capacity; or (b) ensure that all project activities 19  can be carried out without reliance on the existing Ministry structures. The complexities of managing technical assistance activities suggest the usefulness of developing an umbrella contract with an outside agency to manage the identification, recruitment and supervision of consultants. 13.7 This project highlights the importance of the Bank orienting the counterpart agency about Bank procedures for disbursement, procurement and auditing. Through project supervision, the Bank should also monitor project coordination units carefully to ensure that they operate as small, technical units and collaborate closely with line departments. 13.8 The experience of this project also supports the approach of involving key players early on. For example, greater Ministry of Finance support for the project at an early stage might have facilitated necessary counterpart funding. Regional and local authorities need to be involved at an early stage to build commitment to the project and to broaden the project's interest group. 13.9 Finally, this project suggests two additional lessons on strategy. After two years of extremely weak performance, a new Government proposed a major modification and expansion of the project which the Bank estimated would set implementation back considerably. Although the Bank recognized that this reprogramming would not address the constraints which had led to weak Performance to date, the Bank expressed willingness to consider this reprogramming. This willingness was due in part to the Bank's desire to use this project as a means to maintain a positive country dialogue under difficult circumstances. This experience suggests that project decisions which are not supported by project considerations will do little to achieve broader country relations aims. The Bank should be firm in rejecting reprogramming of an already weak project which would further complicate execution. In similar cases, the Bank should recommend either (a) cancellation followed by preparation of a new project which explicitly addresses the factors that contributed to weak performance of the earlier project, or (b) scaling down of the project with the possibility of later modification only based on successful implementation of the scaled down version. 13.10 When loan disbursements are suspended, the Bank should arrange to review on at least an annual basis the merits of keeping the loan open, takdng into consideration: (a) the commitment fees which the Government will incur during suspension versus the costs to the Govemment of preparing a new project; (b) the likelihood that the suspension will continue; and (c) the project's prior implementation record. In this case, disbursements were suspended on all loans to Peru in May 1987 because the national governmnent'q derlared poition of limiting dcbt payments to a fixed proportion of export earnings placed it in 20  arears with the Bank. The national govemment had made this position a central point in its international discourse, was therefore unlikely to back down and would be in office until July 1990. Earlier anellation of this loan would have made sense because: (a) the Government would spend more in commitment fees to keep the project loan during suspension than it would cost to prepare a new project; (b) this government would be in place for three years following suspension; and (c) the implementation experience prior to suspension had not been successful. 21  PERU PROJECT COMPLETION REPORT PRIMARY HEALTH PROJECT (LOAN 2211-PE) PART I: Project Review from Government's Perspective 1 Bakgrond 1.1 The Ministry of Health adopted the social goal of the Pan American Health Organization-Health for All by the Year 2000-(Alma Ata- USSR 1978 meeting) to tackle flaws in the health care system that left 25% of the population totally deprived of health services. Health For All encompasses: a) primary health care, the general integration of community-planned, programmed, implemented and administered health, economic and social development actions organized with the support and advice of the state and other national and international organizations; b) holistic health care, the result of the unified and informed implementation of outreach, protection and rehabilitation actions targeted to benefit both individual family members and the community; and c) universal access to health care, to permit better health status that would enable most of the population to be socially and economically productive for fte benefit of the individual and the society. 1.2 At the request of the Ministry of Health, a special World Bank mission visited Peru on July 1980. The mission agreed to assist in the preparation of a Primary Health Project that would be submitted to the World Bank for financing. To this effect, a worldng group was established at the Ministry (RM No. 0157-8WSA/MS, dated November 1, 1980). World Bank staff visited Peru in January and June 1981. Together with members of the National Commission for Primary Health Care, they worked to define the scope, goals and size of the project. The draft project proposal was sent to the Bank in September 1982. The Banlc in turn sent a three-week pre-appraisal mission (October, November 1982) that defined the general guidelines for project appraisal. The latter was prepared by a working group established on December 31, 1981 for that purpose. In February 1983, the project proposal was approved by the leadership of the Ministry of Health and presented to the World Bank. The Bank sent a six-member appraisal mission that prepared a Staff Appraisal Report indicating that with slight modifications the project proposal would be technically, administratively and financially feasible. On this basis, the mission decided to recommend financial support from the Banl.  1.3 The Loan Agreement between the Republic of Peru and the IBRD for the US$33.5 million operation was approved by the Government for signing (DS No. 046-83-EFC) on February 25, 1983. The loan proceeds would finance the implementation of the Primary Health Project and the Borrower would provide counterpart funds of no less than US$22.5 million in accordance with the provisions of article 3 of the Loan Agreement. The Loan Agreement was signed on March 28, 1983, and became effective on May 20, 1983. 2 Experience in the Execution of the Initial Project 2.1 The main objective of tme project that was initiated in 1983 was to satisfy the priority health needs of the rural and marginal urban populations in selected areas' by developing essentially preventive outreach activities and community-based health services. The strategy was based upon the Primary Health Care Module, a package of human resources and physical facilities designed to provide primary health care, medical and community-based health services to the aforementioned populations. This was the first level of contact between the individual, the family and the community, and the national health system, and it became the first link in a permanent chain of health care. 2.2 Implementation was quite slow during the first three years which demonstrated clearly the gap between the initial project plan and the actual execution. This shortfall was mainly due to: (a) lack of administrative capacity; (b) independent administration of the project components without overall project management; (c) untimely remittance of funds due to the weak iquidity position of the government; and (d) lack of trust in the effectiveness of the project and its goals. For these reasons, the achievement of the programmed goals was behind schedule and only 6% of the approved amount was financially implemented. 2.3 The change in the financial costs as a result of the inflationary process that produced the negative liquidity position of the Treasury caused an opportunity cost loss of unused financial resources and unfulfilled cost-benefit that had an adverse impact on society. 2.4 The construction of the health facilities utilized conventional techniques that did not take into consideration Peru's economic, social and cultural realities and increased the cost of the project, at the expense of greater coverage. X Centro Medio, La Libertad, Lima and Ucayali. 23  2.5 With regard to equipment, only the imported clinical/medical components were purchased (UNICEF) in accordance with the stipulations of the corresponding contractual clauses. The locally-manufactured complementary inputs (furniture, implements, linens) were not purchased. 2.6 With regard to human resources development, out of a total of 27,589 persons, only 1,416 were trained, including physicians, nurses, auxiliaries, technicians, and outreach workers. 2.7 The medicines and materials component and the overall project administration performed far below the proposed goal due to the lack of coordination between the different components during the three-year implementation phase. 3 Reformulation of the Initial Project 3.1 In 1985, on the basis of a new national health policy that supported actions geared to achieving the highest level of health services for the population, the new Government that had come into office on July 28, 1985, decided to appoint a multidisciplinary team of professionals to undertake an overal evaluation of the original Primary Health Project including the legal, contractual, technical and accounting documents. The group concluded that the original design of the project should be reformulated to make it fully compatible with the national health policy. 3.2 The economic and social development policy of the government in office at that time focused its attention on depressed areas and promoted a more efficient and democratic public action. It was intended to gradually abandon centralism and to promote the development of rural areas. 3.3 The emergency zones (trapecio andino) on which the national development policy focused its attention, are areas with low income levels, poor living conditions, low productivity, underutlized natural resources, a predominantly communal and associative social structure, and high Andean areas that hamper national unity. 3.4 With this in mind, the Peruvian Government decided to urgently implement microregional development programs in the Departnents of Puno, Cuzco, Apurimac, Ayacucho, Huancavelica, the highland provinces of Arequipa, Moquegua and the Andean regions of Tacna in order to promote the economic and social development of depressed areas. Its aim was to coordinate public sector action that would include the Departmental Development Corporations (CORDES). 24  3.5 The basic guidelines of Peru's health policy were set forth within the framework of the national policy. The policy was devised to ensure universal and permanent access to health care for the entire population, to close the large and inequitable gap between health care for a selected minority and the large and neglected segments of the population; and to rechannel the financial and human resources devoted to health care in order to achieve a more equitable distribution based upon the scope and nature of the health problems that afflict the overall population. The main goals were to assign a larger share to the most deprived and to promote the participation of the population as actors and beneficiaries in all health activities so that the management, supervision and control of the health system would be in the hands of the public. 3.6 The following were important elements of the health policy which shaped the policy orientation of the reformulated Primary Health Project: a. Community mobilization and participation at all levels and at different stages of the health system. The community must participate in: the idendfication of health problems; the establishment of priorities according to the scope, importance and urgency of identified problems; the definition of ways to resolve priority problems taldng inteo consideration the nature of the problem, availability of resources, appropriate mix to attain goals bearing in mind regional and local characteristics; the decision- making process in regard to plans and programs liberalizing thie planning system; and the administration of health services, including monitoring, evaluation and social control of atl health activities. b. Effective decentralization of health services delegating responsibility and authority to the most peripheral local facility. This decentralization must encompass programming, implementation and evaluation of health activities and administration of financial, human, physical and technological resources. c. Intersectoral action in health. Health is a complex social problem that surpasses the field of medicine and requires multisectoral action. Therefore, all institutions and sectors must join efforts to find solutions for priority health problems. Intersectoral action must be based upon specific local action and the participation of key sectors. 3.7 In that sense, the Primary Health Project was geared towards the achievement of a better way of life for rural and marginal urban populations in 25  the most depressed areas. Its goals could have been achieved only by treating the project as an important component of the economic and social development process, since health and disease are complex biological, psychological, and social phenomena that have a cause-and-effect relationship with political, economic and cultural events. 4 Projec Qkj 4.1 A system of basic services was established within the model of Universal Health Care in order to execute actions and activities mainly for infrastructure development (construction, equipping and implementation of health facilities) in local com"iunities targeted for universal development. From the institutional standpoint, tle system included political and administrative aspects, peripheral services, support hospitals and former Departmental Health Units (UDES) and from the social point of view, political and social aspects, family units, communities and microregions. 4.2 The main goal of the Primary Health Project was to provide universal health care to the most deprived populations (rural areas of the trapecio andino and marginal urban areas) by means of a system of basic services (health centers and health posts), with the active participation of the community and joint efforts with related sectors and institutions devoted to integrated community development. 5 Project Components and Targets 5.1 In keeping with the health sector strategy, the project targeted the most needy groups, specifically the rural and marginal urban populations. In view of this, the original project was redefined and coverage was extended nationwide to rural communities with more than 500 inhabitants that lacked health services or to communities where expansion of existing infrastructure was warranted by increased community size. 5.2 The project gave priority to the trapecio anino for its strategic political situation (depressed area), to the microregions where services were generally the weakest, and to the marginal urban areas that needed health services due to their limited access to medical attention. 5.3 The five components of the project were: (a) construction of health centers and posts; (b) equipping of health centers and posts; (c) training; (d) medicines and materials; and (e) an administrative component. The main objective of the construction component was to provide physical infrastrucure 26  for health services to the most needy rural and marginal urban communities. This took into account the following principles: (a) to standardize primary health services, consistent with decentralization; (b) to favor the use of labor and resources from the areas where the establishments would be built thus creating temporary job opportunities; (c) to design the prototype on the basis of rationalized building and equipping, while reducing the purchase of equipment with the use of built-in and installed furniture; and (d) to locate health facilities to serve as outreach agents not only for health but also for other productive and social sectors. The target was to build 419 health posts and 67 health centers. Non-conventional building techniques included the use of quincha, adobe and wood. The choice among these materials for each site depended on the topographic and climate conditions, availability of raw material and human resources trained in the use of the particular technique. 5.4 The equipping component aimed to provide medical, clinical, hospital and office equipment for the health centers and health posts to be built; and to equip other project-related facilities that lacked necessary equipment. The targets for this component were the same health centers and posts as for the construction component. For this component the purchase of imported equipment was foreseen (agreement with UNICEF), as well as complementary locally-manufactured equipment. 5.5 The objective of the training component was to ensure effective operation of health services in order to offer timely, effective and efficient health care with active community participation. The training activities included: workshops for community agents; refresher workshops for health auxiliaries; and workshops for health center professionals. The preparation of teaching material for each health facility was also foreseen. 5.6 The medicines and materials component was devised on the basis of packages stocked with basic health care equipment and materals. Each package would take care of 500 persons per year. Twelve packages a year were programmed for the health posts and 24 for the health centers. 5.7 The administrtive component included supervision, evaluation and administrative support activities. 6 Gols of the Reformulated Projct 6.1 The reformulated project had five goals. First, it sought to provide Peru's most depressed rural and marginal urban areas with more primary health services by expanding coverage. Second, it aimed to improve the health status of the country's marginal population through holistic activities 27  in the planned health facilities, communities and families within the framework of the new strategy for primary health care. Third, it aimed to use non- conventional construction techniques and provide equipment capable of satisfying the country's real health needs bearing in mind the ethnic and cultural levels of the population. Fourth, it sought to increase the managerial and implementation capacity of the project and provide for overall administration in order to overcome earlier problems. Finally, it aimed to restructure the Primary Health Care Modules (MAPS) and create the Basic Sanitation and Integrated Health Attention Modules (MOSBAIS) as a project strategy. The MOSBAIS were to be a dynamic force for integral community development and would define duties and establish permanent coordination mechanisms within the community. It was in the interest of the health sector, other sectors and institutions to join and multiply efforts utilizing the community potential and institutional resources to achieve an overall solution for health problems in the localities aiming, at the swme time, to reach a self-sustainable community development. The MOSBAIS were made up of four resources-the family, the community health unit, the community health group and the basic equipment stands (MEB). 6.2 On April 24, 1986, legal provisions were approved at the highest level declaring the public need and preferential social interest of the Primary Health Project. An autonomous executing unit PEAPS was established for the planning, coordination, implementation, supervision and evaluation of the project. It was also decided to grant emergency status to the project until its completion. 6.3 Initially the Bank did not accept the changes proposed by the project reformulation. It considered that: a) it was necessary to keep the pilot plan for the five representative Departments of the different types of areas-the coast, highlands and jungle; b) the trapecio andino was a center of terrorist activities and would be difficult to supervise; and c) amendments would have to be made to the Legal Agreement. Furthermore, the Bank did not agree with the use of non-conventional construction techniques. The reformulated project planned to use stabilized adobe, improved quincha and bricks depending on the geographic location of the construction site. 6.4 At that time, Government project representatives responded to the Bank's concerns with the justifications for the proposed reformulation. They countered that pilot plans should be implemented in those areas in most urgent need of health services. Regional diversity should not be construed as the decisive factor in implementation. When health issues were at stake, the region- -coast, highlands or jungle-was of secondary importance. The main issues were: who was suffenng and who was not; who had access to services and who did not; and who had been the most neglected by the State. 28  6.5 The Government team noted that as far as the trapecio andino was concerned, while it was true that terrorist activities make it a dangerous area, this evil had flourished due to conditions of extreme poverty and neglect. Also, the project team would develop a reprogramming proposal for the Bank to study. They further noted that construction techniques that used local and technologically improved materials would create job opportunities and produce benefits in the long run. 6.6 The Bank decided to support the reformulation based on these considerations although it was still reluctant to support activities in the trapecio andino. The Government reformulation proposal was submitted to the Bank in 1986 and was executed until May 1987 when the World Bank halted disbursements. 6.7 This stage presents several lessons. First, the Integrated Health Attention Module (MOSBAIS), originally a logical operation, turned into a runaway construction race spurred by political interests and political meddling that did not focus on the overall operation. Second, the lack of institutional srengthening precluded a strong stance for project development. The Ministry of Health had zero participation in health care activities related to the project. Efforts to equip 125 health posts which would have allowed the completion of 16 MOSBAIS for future operation were sometimes frustrated by Ministry of Health officials. There was concern about meeting infrastructure targets without evaluating the effects on the targeted population. There was a lack of ongoing evaluation and project dissemination; no definition or monitoring of key indicators that would have allowed an ex post evaluation of the implementation process and a lack of supervision and adequate monitoring. The scarce economic resources did not permit manning and operating the health facilities that were built. 7 roject Results and/or Achievements 7.1 The original projr-t achieved only 6% of its goals. When comparing the original project anm Lhe reformulated project, progress between 1985 and 1987 was made in the following areas: a. Coverage reached approximately 2 million people. b. The construction of 11 health centers was completed by force account and construction of the Ucayali health center was finished. c. Twelve health centers were totally equipped. 29  d. Construction of 155 new health posts was completed and construction of 22 health posts initiated under the original project was finished, i.e., a total of 177 health posts. e. Ninety new health posts and 22 from the original project were equipped, i.e. a total of 112 equipped health posts. f. A total of 1.850 medicine packages were purchased for the health centers and the health posts. g. Outreach and training activities were held in the Departments of Lima, Junin, Huancavelica, La Libertad, Ancash, Cajamarca, Cuzco, Puno and Apurimac. These activities included surveys, presentations, seminars, educational plans, and prepartion of teaching material. h. Q'iechua and Aymara rural communities that had not interacted before were trained together on maternal health, breastfeeding, nutrition, building techniques, etc. L. The initial modular growth concept was modified. It was decided that the health post should not evolve into a health center. It should remain as a community meeting center for teaching and training activities of different nature. j. An investment project wac implemented consistent with the national health policy. It responded to a national and sectoral goal and, therefore, to the implementation of a government policy that gave priority to health action in the Departments of the trapecio andino, the microregions and marginal urban areas. 8 Negative Aspects of Project Development 8.1 The project did not fulfill its main goals due to economic and bureaucratic difficulties; the coverage was approximately 2 million people; and the latter was achieved by awakening awareness that health is not a handout but rather something obtained through the efforts and cooperation of all people. 8.2 The wrong path was chosen when defining goals; rather than seeklng the tangible goal of measurable results, efforts should have been devoted to create awareness that everybody would have access tc rimary health care if we work for it and educate people to stay healthy. 30  8.3 The first step should have been to train the units that would be in charge of the administration and execution of the project; to create awareness and promote dedication and ;o shift the perective of health s'xvices from medical care and healing to promotion and prevention; in general, to change the health care system and make it more universal. 8.4 The creation of an autonomous executing unit (PEAPS) to enhance efficiency resulted in a bureaucracy that was used as a political instrument. It defeated its own purposes and in 1986 had 371 employees and an annual payroll of approximately US$1 million. 8.5 After the loan was cancelled, the government continued to support the project with the desire of keeping a transient political presence in all areas of execution without a clear picture of goal enhancement. 8.6 The construction of a health post is not feasible unless it is understood that it requires equipment and technical personnel to operate it. 8.7 The project lacked continuity. There were five Ministers of Health during the 1985-1990 Presidential period. The changes in project management plus the interference of partisan politics and the lack of decisions to make execution feasible hampered its sustainability. 8.8 In sum, integrated health care must be viewed as consistent actions for outreach, protection, healing and rehabilitation, that have been planned, programmed, implemented, and managed with the community and that are geared to the overall development of the individual, the family, and the community. 9 Project Sustainability 9.1 Project achievements through implementation have not resulted in a system based upon the primary health care strategy in spite of the coverage attained through the construction, ea 1ipping, and supply of medicines to the new health posts and centers. 9.2 The 1985-1990 Government limited primary health care actions to community participation leaving execution in the hands of nongovernmental organizations, local govemments, universities and churches in some instances. 9.3 As of July 1990, the new Government started to take action in order to implement the primary health care program. It established guidelines in health sector policies and developed strategies to support these policies 31  through institutional strengthening, relocation of human resources and the creation of Integrated Health Development Zones (ZONADIS). In order to f0ifill the chief goal of primary health care, these elements would help prepare .ae dialogue, and coordinate and harmonize the actions of all agents that participate in the delivery of health services, utilizing the existing physical infrastructure built with the project. 9.4 On the other hand, the Ministry of Health and the World Bank have been joining efforts to strengthen primary health care through the preparation of the new Basic Health and Nutrition Project. 10 Bank-Borrower Relationship 10.1 We agree with the World Bank that Bank-Borrower relationship was good and that communication flowed easily. The problems that emerged in implementation were caused by the disorganization in the Ministry of Health during 1985-1990 and by the financial policy of the past government that hampered a larger and timely flow of counterpart capital to achieve the proposed goals. 1I Conclusions and Lessons Learned 11.1 In public health it is not worthwhile to build, equip or take any action in places without pressing needs. The attack must be staged in areas where poverty and hunger are endemic. 11.2 Before acting it is essential to know the real environment. 11.3 Pilot plans work well only in areas with urgent needs and when they can be replicated in other areas on the basis of their pathological patterns and not geographic considerations. 11.4 An innovative project must be executed bearing in mind contributing factors that may support its feasibility and the project's catalytic effect to improve the standard of living of the population which in turn improves health status. 11.5 Contrary to the Bank's assertion, the project did have a plan to introduce and adjust interdependent phases (construction, equipping, medicines and training). 11.6 The problem arises when political interests intervene in a 32  technically-designed project, intended to be executed by professionals. In countries like ours, political interests can tun any technically and well- conceived project into a series of inconsistencies, local interests, and pursuit of personal gains that often result in failure. National interests many times become of secondary importance and the target population endures the consequences of deplorable behavior. What must be achieved and understood is that overall political support is a key element for project development; political soundness must ensure its priorty and continuity; and the effective and guaranteed relation between political and financial aspects legitimize a project. These are considertions that the present Government is implementing. 11.7 Since there was no institutional strengthening, the project's management capacity (through its administrative and legal systems) was weakened; programming did not guarantee efficiency of action. 11.8 It is necessary to establish central mechanisms for the design and initial implementation. These will support the new role of the Ministry of Health as a policy-setting and supervisory body, will strengthen the administration and the operational capacity of health services and will endorse decentralization keeping close and permanent coordination with te central governing body. The Government has been pursuing these actions by means of the rationalization and structuring of the administrative bureaucratic apparatus of the health sector. 11.9 The Ministry of Health was far behind in mechanisms, pr cedures and technological and managerial developments. Therefore, it was practically impossible to manage a project of this size without altering agreed timetables. For example, Bank procedures were understood only after project management staff travelled to Bank Headquarters for a week and learned how they worked. 11.10 Decentralized implementation should have been promoted to gain the necessary flexibility at the local level. Regicnal and local authorities had an odd perception of the project. Therefore, there was no sharing of decision- making and little success in worldng with the community. 11.11 We agree with the Bank that this experience provides valuable lessons for other projects in design, organization and participation. It demonstrates the merits of keeping project design simple, starting with small changes and building on successful learning to introduce additional changes at different stages. 11.12 Projects that include construction of infrastructure and the supply of medical equipment should include budgetary allocations and maintenance plans to ensure continuity. 33  11.13 We agree with the Bank insofar as additional lessons on strategy are concerned and with the proposed recommendations: (a) cancellation followed by preparation of a new project which explicitly addresses the factors that contributed to weak performance of the earlier project; and (b) scaling down of the project with the possibility of later modification only based on successful implementation of the scaled-down version. 12 Summary 12.1 Public health projects must be implemented in extremely poor areas that lack services and where urgent action is required to avoid a deep erosion of the standard of living and outbreaks of disillusionment and violence. 12.2 If pilot projects are devised, they should take into account not only geographic conditions but also ethnic and cultural characteristics that permit replication. 12.3 There must be a way to find and put in place mechanisms to avoid political interference from individual or groups that may alter political decisions oriented to favor national interests. 12.4 It would be advisable for the Bank to include immediate implementation provisions to avoid implementation disruptions; also, the loan should include allotments providing for at least three years of project operation to ensure continuity at the implementation stage. 34  PERU PRQJLTh C 1MpEI REPR PRSMARY lIEALtIHPOIC (LOAN 2211-PE) ParT m: Sttifg Information Please see Graphs 1 through 4 and Tables I through 8 attached. 35  PERU - PRIMARY HEAiTH PROJECT (Loan 2211 -PE) PROJECT COMPLETION REPORT Graph 1. Estimated and Actual Cumulative Disbursements Cumuldve Diaburmnenu ts (US000,00) .36 l"*3222*2***f*i::: F:::th" '': 2:-i::: --:: :*:.:v- " i 3 i 22.3s}r b.i}:u:z::li3~ i.2-2fi} u ;; 2 *33fl.r 1: ..40j}z,j}}}}"}ii} lIti. . : ..::i::::::.srts::::::5..d:::::::23:..:s:::::::: ::2:::::i22 ..:_:: _1 2id U332t33ii.'jHiviti. 222 l ibi7inj. ihiiJi 90 "" vrn.7 rat}... Er i ld !si!riiiiiiNi--f-Aiii 227222 2.2....3222 i33r3:2". 222 22....22.iFii-i'L-! 120 L$mEises 5irlm low low lim 1990