So,. Set . Med. Vol . 2 2 . No . 6, pp . 653-672 . 1986

02277-9536,86 S, 00+0.00Printed in Great Britain . All rights reserved

Copyright C 1986 Pergamon Press Ltd

UTILIZATION OF PHARMACIES ANDPHARMACEUTICAL DRUGS IN ADDIS ABABA, ETHIOPIA

HELMUT KLOos', TSEGAYE CHAMA', DAWIT ABEMO',KEFALO GEBRE TSADIK' and SOLOMON BELAY -

'Department of Geography, Addis Ababa University, Addis Ababa and 'Pharmacist, City CouncilPharmacies, Addis Ababa, Ethiopia

Abstract-This paper examines recent developments in the pharmaceutical retail trade in socialist Ethiopiaand presents the results of two surveys on pharmacy utilization in Addis Ababa . Surveys were carriedout in 6 private and 5 government pharmacies . Objectives are : (1) to examine drug retailer utilization inrelation to locational, transportation and retailer-related factors, (2) evaluate the role of socioeconomicfactors in pharmacy and drug utilization ; and (3) determine distance decay associated with clients' triporigins and the location of their residences as indicators of service areas . Although most clients originatedin Addis Ababa, large numbers came from rural areas, especially in the pharmacies near large marketsand other shopping areas in the inner city . Centrally located retailers also served more Addis Ababaresidents and larger sections of the city than peripherally located retailers, largely due to a combinationof urban structure, distribution of health care facilities, prevailing drug shopping behavior and populationdistribution . Government pharmacies had larger service areas and served larger numbers of clients thanprivate pharmacies, primarily due to lower prices and greater availability of pharmaceuticals . Meandistance from pharmacies to places of origin of trips was smaller than mean distance from pharmaciesto residences of the same clients . Similarly. distance decay gradients were steeper for the former than thelatter in the 4 pharmacies studied in the second part of the survey, indicating the greater suitability oforigin of trip as a parameter of service area . Type and price of drugs purchased were associated withsocioeconomic factors, particularly level of education and housino'environmental health conditions in twodistricts, but there was little variation in the small number of drugs purchased per client . Several formsof drug-purchasing behavior of pharmacy clients and selling practices of private retailers are describedas adaptive responses to prevailing economic and sociopolitical conditions . The study concludes thatpopulation-based studies of disease occurrence and health behavior are needed to better evaluate thehealth needs of the population for the planning of additional drug retailers in Addis Ababa's suburbandistricts .

INTRODUCTION

Health planners must consider a number of locationand allocation parameters as well as travel time andcost, intervening opportunities for medical treatment,and the culture of the population at risk if spatiallyand functionally optimum delivery of care is to beprovided. Although several models have been devel-oped by geographers with the objective of upgradingthe location and allocation of hospital and physicianservices [I], research in one area of health care-thedistribution and utilization of pharmaceutical drugretailers-has been neglected. Provision of essentialdrugs [2] in developing countries has been identifiedas an essential element of primary health care withinthe context of the Global Strategy of Health for Allby the Year 2000 [3] . Pharmaceuticals have con-tributed to the improvement of living conditions andlife expectancy in many developing countries, buttheir supply (primarily through multinational cor-porations) and cost fall far short of health needs andavailable resources in poor countries [4] . Althoughexpenditures for pharmaceuticals in different devel-oping countries accounts for 25-60% of the totalnational expenditure for health care, which represents2 to 4 times the proportion spent in developed

Send all correspondence and reprint requests to Dr H .Kloos, P.O. Box 31609, Addis Ababa, Ethiopia .

ssa _n_o 653

countries, pharmaceutical products are either notavailable in appropriate quantity or they areunaffordable for most of the population [5]. In someof the least-developed countries, including Zaire andEthiopia, the annual public expenditure on health isonly about $1, compared to $450 in the United States[6] . The scarcity of pharmaceuticals in developingcountries is even more clearly brought into focuswhen one considers the widespread distribution ofdisabling and killing diseases, the concentration ofbiomedical services in large cities, and lack of controlover the distribution and utilization of drugs . Tra-ditional indigenous medicines are also widely used inmany developing countries [7] . However, they areunlikely to reduce the growing demand for modernpharmaceuticals. Increasing exposure of indigenouspopulations to commercial drugs in also believed tobe causing far-reaching changes in health behavior,particularly in urban areas [8 .9] .

Problems of achieving equitable distribution andsafe use of pharmaceuticals have also been associatedwith the various phases of their production . market-ing and distribution . Concerns include the pro-motional and sales practices of multinational com-panies and the utilization of drugs by retailers . healthworkers and patients [10] . Partly due to sales pro-motion and the patent and sophisticated pricingsystems of multinationals, many governments havefailed to implement World Health Organization

6 5 4

HELMUT KLOOS et a!.

guidelines [111 for developing rational drug policiesaimed at the procurement of essential (or appropri-ate) and low priced products. In the absence ofadequate drug policies physicians tend to over-prescribe and at the retail level pharmacy workersand unqualified drug peddlers often sell pharma-ceuticals without doctors' prescriptions to an unin-formed and often illiterate public, with frequentharmful effects [12] .

Partly due to strong cultural, social and economicbarriers between people and physicians in developingcountries, pharmacists . druggists and less qualifiedretailers play a greater role in prescribing and dis-pensing pharmaceuticals than in developed countries .Self-diagnosis and self-treatment involving pharma-ceuticals are widespread and drug retailers freelyadvise clients on which products to use, many ofwhich can be sold in developed countries only withprescriptions . The sale of many pharmaceuticals overthe counter that would require prescriptions in theircountries of origin adds to the importance of drugretailers in meeting health needs. Thus, the healthstatus of a population encompasses not only easy andinexpensive access to safe and effective remedies, butalso better understanding of the proper use of phar-maceuticals, including associated risks and benefits .Prescribing and dispensing practices influence howpharmaceuticals are used and who uses them [13] .

Most critics of the production and marketing ofpharmaceuticals in free market competition . and theoperation of multinationals in developing countries,favor their regulation to facilitate more-equitabledistribution of health services (14] . The main argu-ment of proponents of more moderate policies is thatelimination of the profit motive and competition inthe production and marketing of drugs would alsodestroy drug research and result in inefficiency andcorruption [15] . Van der Geest [16] noted that thehighly political character of the debate between rad-icals and moderates may result in strongly biasedresearch and suggested that case studies be carriedout of drug distribution at the community level . Thepresent study examines the utilization of selectedgovernment and private pharmacies in Addis Ababain relation to locational, distance, facility-related,socioeconomic and political factors .

PHARMACEUTICAL DRUGS ANDRETAILERS IN ETHIOPIA

The Ethiopian revolution aims to establish thenecessary economic and social infrastructure for theimprovement of health services in a country withsome of the highest infant mortality rates and lowestlife expectancy world-wide . The following statementby the Head of the General Planning Sector [17]reflects the accomplishments of the revolution andthe change in philosophy and orientation of thehealth services toward primary health care involvingcommunity participation . away from a free marketsystem [18] .

Social ownership of the major means of production anddistribution . the establishment of a strong central planningorganization (National Revolutionary Development Cam-paign and Central Planning Supreme Council) with sevenregional planning offices, the formation of peasant . work-

ers' . youth . women's and urban associations and the pro-motion of producers' and service cooperatives provide idealconditions for development and dissemination of health-related activities as well as for research efforts focused onhealth promotion .

Specific measures for achieving the economic, so-cial and health development goals are set forth in the1984-1994 10-year plan [19] . However, although theMinistry of Health is in the process of reorganizingand restructuring itself to better coordinate, imple-ment and support the new plan . one can hardlyconsider this a national plan comparable to thoseexisting in older, more established socialist countries .Progress toward achieving more decentralized andequitable health services has been slow- partly due toorganizational and conceptual difficulties at variouslevels of the health services and at training centers[20] . Health services remain highly centralized inAddis Ababa, with health facilities in the capital cityusing 33% of the recurrent health budget . Some 28%of the hospital beds, 51% of all physicians . 60°6 ofall pharmacies and 74% of all pharmacists in Ethi-opia are found in Addis Ababa [21] . Moreover, thereferral system between primary care facilities andregional and national hospitals remains under-developed [22] . Although no pharmacy regulationshave been issued by the socialist government ofEthiopia, due to the absence of a comprehensivenational health plan, the Ministry of Health hastaken a series of measures which have had a far-reaching effect on the operation of drug retailers andthe distribution of pharmaceuticals throughout thecountry [23] .

Recent developments in the Ethiopian drug whole-sale and retail trade are similar to those in the othersections of the health services . particularly in regardto increased government guidelines and control ac-cording to the country's health needs . the purchasingcapacity of its people, and the philosophy of social-ism [24] . The single Ethiopian manufacturer of phar-maceuticals and several importers wholesalers werenationalized and recast in a state corporation re-sponsible for the manufacture, import, quality con-trol and distribution of drugs . Total drug sales inEthiopia increased from 42 .9 million Birr [25] in 1972to 45 .8 million Birr in 1980 . The domestic share ofdrug production increased from 185 to 31.0% duringthe same period [26] . This increased share of thenational drug manufactures was partly due to anincrease in import taxes (from 27 to 40%) designedto encourage indigenous production [27] .

Beginning in 1977 the Ministry of Health nation-alized private pharmacies and druggist shopsthroughout the country and by 1934 it owned 5public pharmacies and all but one of the I I publichospitals in Addis Ababa . Four years later the AddisAbaba Municipality began its own pharmacy pro-gram. By the end of 1984 it operated 5 publicpharmacies, all of them placed for rapid expansionand maximum utilization at major traffic inter-sections and bus stops . As part of their effort tominimize cost of health care and to gradually takeover the private sector, both the Ministry of Healthand the municipality reduced the retail profit marginof drugs sold in retail shops from 40 to 30% andsupplied their pharmacies with certain drugs that

private retailers were no longer permitted to sell,including T .B. drugs, some higher antibiotics . insulin,psychotrophic drugs and several vitamin mineralpreparations . Several private retailers responded bylowering the prices of some drugs . especially for wellestablished clients, in an effort to remain competitive .In November 1984 municipality pharmacies reducedthe profit margin by another 5% . to 25%, a movefollowed by some Ministry of Health pharmacies .with further price reductions planned for the future .The lowest priced drugs were offered by the govern-ment hospital pharmacies since their recent exemp-tion from customs tax but their services were re-stricted largely to inpatients and qualifying poorpeople. Equally important, those hospital pharmaciesthat do serve the public require prescriptions fornearly all pharmaceuticals, thus eliminating fromtheir clients the great majority of the population,which continues to depend heavily on self diagnosisand self-medication . The price reductions, togetherwith their well selected locations in the inner city andthe wide-spread sale of over-the-counter drugs havecontributed to making the municipality-owned phar-macies the most popular and most widely patronizeddrug retailers in Addis Ababa . The municipality is

Utilization of pharmacies and pharmaceutical drugs

planning to establish pharmacies and druggist shopsalong with clinics and health centers in all 9 keftegnas(higher cooperative societies of urban dwellers) [28] atthe urban periphery that lack drug retailers (Fig . l,Table 1) . It is not known, however, if the municipalityor the Ministry of Health will eventually control anddevelop the drug retail trade in Addis Ababa.

Owners of private pharmacies and druggist shopshave developed a pessimistic attitude toward thefuture of their business in view of the increasingthreat of business closures by the government . Inter-views with pharmacy owners indicate that the Minis-try of Health has been closing pharmacies on theground that owners evaded sates taxes . failed to payrent on time, and committed serious errors in dis-pensing drugs . One of the various regulations issuedby the Ministry of Health as part of revised healthpolicy is the prohibition to treat patients in pharma-cies, druggist shops and rural medicine vendor shops[29] . Before the 19'4 revolution most drug retailers inAddis Ababa (and in rural areas) commonly treatedclients for numerous ailments, giving physical exam-inations and injections and performing surgery andother clinical procedures, as observed by the seniorauthor on several occasions [30] . The prohibition to

•

Private pharmacy or druggistshop

o Private pharmacystudied

Government pharmacyA Government pharmacy

studied+ Hospital -

* Clinic or healthcenter

LIM11J, Nterkato

p'IPiazza

R Railroadstation

- Kefregnaboundary

- Kebele boundary

---C;ty limits

Build-up area

Only the 2 hospitalpharmacies studied are shown

0 z 3 km

Fig . I . Drug retailers, hospitals, clinics and health centers in Addis Ababa in December 1984 .

653

656

Table I . Population size and density . land use, housing and environmental health conditions, and numbers of pharmacies, druggist shopsand medical care facilities, by keftegna and study district

'Based on the 1984 Addis Ababa census [5911tAddis Ababa Master Plan, unpublished data .;G-government offices ; 1-international organizations ; M-Merkato and surrounding slums : Ma-manufacturing industry; R-

residential : S-retail shopping area ; U-university .§I-Below average : 2-average ; 3-above average .' No . of pharmacies (No. of druggist shops) No . of hospital pharmacies.

treat clients in drug retail shops is perceived byprivate owners as another means by the governmentto squeeze them out of business .There were 57 pharmacies, 18 druggist shops (Fig .

1) and 15 rural medicine vendor shops in AddisAbaba at the end of 1984 . They included 23 govern-ment pharmacies and 34 private pharmacies . Thirteenof the 23 government pharmacies were in AddisAbaba's I I public hospitals, with 10 public retailerswidely distributed in the more central sections of thecity. Druggist shops differ from pharmacies in thatthey carry fewer pharmaceuticals . are owned or man-aged by druggists rather than pharmacists and arenot permitted to compound drugs themselves [31] .The oldest pharmacies and druggist shops still inexistence in Addis Ababa date back to the 1930s andare in some of the oldest sections of the city . thePiazza and Merkato (Fig. 1) [32] . Rural medicinevendors, originally licenced by the Ministry of Healthto serve rural areas, were permitted in 1964 to opentheir shops in the peripheral urban areas of AddisAbaba with the objective of alleviating the chronicshortage of drug retailers there . However, due to thesmall number of pharmaceuticals rural drug vendorswere permitted to sell-mostly analgesics, antacids,plasters, tonics and other nonprescription items-theresidential areas in the suburbs have remained with-out adequate coverage until today. Nearly half of alldrug retailers in Addis Ababa are still in the commer-cial and crowded residential areas of the Piazza,

HELMLT Kloos et a!.

Housing andenvironmental

healthconditions ;

I to 22

1

I to 2I to 2I to 2I to 2I to 2

2 to 32 to 32 to 3

2 to 32 1o 32 to 32 to 32 to 3

o

2 to 32 to 3

Merkato and railroad station (Figs I and 2),reflecting few changes since 1973, when Kloos [33]first mapped all pharmacies, druggist shops and ruralmedicine vendor shops .

METHODOLOGY

All pharmacies and druggist shops as well as allpublic hospitals, clinics and health stations in AddisAbaba were identified from Ministry of Healthrecords and field checked . Only pharmacies wereselected for study, for comparability of private andgovernment retailers . After initial visits to all phar-macies in the city, 10 pharmacies were seelected bypurposive sampling for preliminary study in May1984 on the basis of location and type of ownership .They included 4 shops studied by Kloos [34] in 1973 .Thus 5 retailers in the central keftegna. Nos 2 and 5and one or two each in the peripheral, lower densitykeftegna (Nos 13, 14, 15 and 18) were selected [FigsI and 4(B)] . Two of the pharmacies were operated bythe municipality, 2 by the Ministry of Health and 6privately . Eight of them were public pharmacies, theremaining 2 being located in a hospital and operatedby the Ministry of Health . One of the municipalitypharmacies is located in the Piazza and the other inthe Merkato . Both government and private retailersare considered to be representative in regard tolocation and size (number of employees and drugs) .Apparently due to concern over possible repercus-

KeJregna

Study district(weredal in

December 1981survey Population

Area(ha)

Population Predominantdensity'

land use ;

1 45,137 153 295 R35 .771 201 178 R. S, G

3 -70,850 225 315 R, M . S

Teklehaimanot 65,41 1 207 316 R, Sf . 55 75 .331 170 433 R, M . S6 65,797 185 356 R, 11 . S

7 74,157 202 367 R8 60,8-1 1 1331 46 R9 45.626 248 184 R10 62.019 920 67 R11 64.766 1370 47 R12 42,565 1198 36 R

13 49.825 361 138 R . U, G . 114 Mekagelegna 53,156 161 204 R.U,G, 1IS 56 .918 363 157 R. L, G . I

16 52,649 1344 39 R17 57,151 2663 21 RIS 47 .109 417 113 R19 53 .78 22 1411 38 R. Ma20 56,291 657 86 R, Ma21 75 .579 349 217 R . G71 33 .08 22 194 171 R23 59 .2;4 1496 40 R24 63-007 1495 42 R25 46,291 1443 32 R

1,41257.5 18 .864 75

No . ofpharmacies anddruggist shops•

No. ofhospitals .

health centersand clinics

1(2)010(2)0

07

5(1)20(0)18(4)0I(I)I

3I4-

3(0(00(1))10(0)10(0)00(1)00(0)0

03I

0(1)35(0)11(2)0

Of l)00(0)02(0)0010)011110612)11(0)00(0)10(0)10(0)0

43(18)13

341

00I0060I2

1

44

Population densityper hectare [591

120-99

ti 100-199

i:VJ 200-299

w »..k.300-399

443

•

Privatepharmacies :Nos 5-10

2

3 km14

Utilization of pharmacies and pharmaceutical drugs

Fig . 2. Population density . b y kefregna, and pharmacies studied .

sions from cooperation with government in-vestigators, including tax collection aspects, com-petitive advantages and strategies and securing ofpharmacy licenses, 4 private retailers declined tocooperate with our study . Random sampling of phar-macies was therefore not possible .

On a Saturday in May 1984, 20 teams of 2 univer-sity students each [35] interviewed clients in all selec-ted pharmacies between 8 a .m. and 12 noon and 2p.m. and 6 p.m ., working on a rotation basis. Theyused a questionnaire and conducted the interviews inthe Amharic language, asking questions about placeof residence, whether clients obtained all drugs theywanted and reasons for not obtaining all drugs,previous trips to other pharmacies during the sameillness episode, and place of origin of trip . Effortswere made to interview all persons coming to re-tailers, including those who did not purchase anydrugs . Clients were briefed about the purpose of theinterviews and asked for their cooperation . A total of1781 clients were interviewed about place of residenceand of these 1040 were asked the other questions aswell . An estimated 85-90% of all clients of 8 of the10 pharmacies were interviewed . Only 70% of clientsof the 2 municipality retailers were surveyed due tothe large volume of trade .

657

In December 1984, a second, more intensive surveywas carried out in 4 pharmacies, including 3 studiedin May (Nos 1, 5 and 9) and including a municipalitypharmacy newly opened in July 1984 . These wereselected on the basis of ownership (2 were private and2 government owned) and location . Two (Nos I and5) are located in close proximity to each in a relativelypoor, densely populated area with substandard hous-ing and environmental sanitation . The other two(No. 9 and the new retailer) are also adjacent in a lesscongested, more suburban section where large num-bers of government employees, academic personneland students live (Fig . 2) . The study units selected forevaluation of socioeconomic influences on pharmacyand drug utilization are Teklemaimanot Wereda(District), consisting of keftegna 3-6, and the moreaffluent Mekagelegna Wereda (keftegna 13-15)(Table 1) [36] . An expanded questionnaire was usedto interview 815 clients during 3 weekdays . Newquestions included the type, number and price ofdrugs purchased ; level of education ; use of pre-scriptions and traditional medicines : and name of theillness for which the drug(s) was bought . Indigenousillness names and descriptions rather than scientificmedical terms were used to better understand healthbehavior. The question on origin of trip w as modified

to obtain more precise locational information, Inter-views were carried out during the same hours daily asin May. Information on total number of clientsserved in the 2 government pharmacies was obtainedfrom sales records and in the private pharmacies fromcensuses . The allegedly unreliable data on drug inven-tories of private retailers were inaccessible becausethey are considered confidential and proprietary .Thus inventories of private shops were estimated onthe basis of types of pharmaceuticals purchased bythe sample and those the clients could not obtain .Although only one interviewer was used in eachpharmacy, resulting in smaller proportions of totalclient populations being interviewed than in May, thesample is thought to be fairly representative . This wasalso indicated by a comparison of clients servedduring quiet hours and business rush hours . Thenames of pharmaceuticals purchased and their priceswere obtained from visual inspection of containersand sales receipts . The interest clients showed in thequestions relating to their efforts to obtain the desiredpharmaceuticals, the relatively low refusal rate (5%)among the 2596 persons interviewed and the gener-ally more accurate information generated by healthsurveys than sociological surveys in Ethiopia (37] allindicate that client information was fairly accurate .

Analysis of data included mapping of client infor-mation at the kebele level (cooperative society ofurban dwellers), the smallest administrative unit inAddis Ababa, for the analysis of locational anddistance decay relationships . Distance of client travelwas measured by the straight line method and by

Table 3 . Reasons for 284 clients not obtaining drugs in government and private pharmacies

hypothetical location of client residences in the centerof kebele . The size of kebele in Addis Ababa (andthe rest of Ethiopia) is based primarily on popula-tion size, most kebele containing between 3000 and6000 people, with extreme values in kebele in keftegna4 (12,247) and 15 (1739) . The area size of kebeleranged from 4 ha (in keftegna 21) to 1102 ha (inkeftegna 20) [38] .

SURVEY RESULTS

Service areas

The number of kebeles where clients of individualpharmacies lived may serve as an indicator of the sizeof their service areas . During all 3 surveys undertakenon the location of drug retailers and their clients inAddis Ababa in 1973 [39] and in May and December1984, associations could be made between urbanstructure and population density (Tables I and 2) .Centrally located retailers consistently served, as agroup, larger areas of Addis Ababa and more rurallocalities than peripherally located shops . In thepresent study, the role of additional utilization par-ameters, including pharmacy ownership and size,socioeconomic level, shopping behavior and travelpatterns of clients were also evaluated . However, thepresence of numerous drug retailers in the inner city .the small sample and lack of population based dataon health behavior do not permit the delineation ofmutually exclusive service areas . Moreover, apparentdifferences in health needs and health demands,largely due to differences in socioeconomic level,

653 HELMCT KLO(DS el al.

Table 2. Reasons given by 815 clients of 2 government and 2 private pharmacies for choice of drug rcu er

Reason given

GovernmentpharmacyNo . I

The newgovernmentpharmacy (°o)

Private pharmacyNo . 5

Private pharmacyNo.9 19n)

Low cost of drugs 125 82 (39.4) 0 (0)Close to residence 60 68 (24.0) 79 (66.1)

or place of workShopping nearby 9 0 (1 .7) 67 3 (24 .2)Pu blic ownership 14 15.5) 0 0 -Drugs are always 14 (5.I) 0 0 -

availableCredit is available 5 (1 .51 0Fresh medicines 6 (1 .31 0Good service I (0.8) 0Others 6 9 (6.7) Ii (6 .6))Multiple reasons :Close to residence or place 30 20 (9 .5) 0 0

of work and low cost of drugsLow cost and drugs are 4 4 (1 .5) 0 0

always availableLow cost and fresh medicines 3- 1 (0 .8)Others 4 6 (1 .9) (3.1)

Total 296 230 100 163 126 (100)

No. of clientswho obtained some drugs

No. of clientswho obtained no drugs

Notavailable

Tooexpensiveor lack

of moneyNo

prescriptionNot

available

Tooexpensiveor lack

of moneyNo

prescription Total

Government pharmacies 45 10 5 29 5 a 99Private pharmacies 43 14 6 99 23 0 185

Total 88 24 II 128 28 284

Table 4 . Origin of client travel to : peripherally and i certralls located pha :madesPlace of

Place o1employment employment Clinic or

Home

< I :an

> I km

hospitalCentrally

161

63

42

35locatedpharmacies

Peripherally

145

38

43

I'slocatedpharmacies

Total

406

101

85

163

'Mostly smial visits .(Does not consider teinporary stay in Addis Ababa,

require examination of parameters other than servicearea for more comprehensive evaluation of drugretailer utilization . As a means of facilitating theanalysis of the spatial distribution of clients and in aneffort to contribute to developing a conceptual basisfor the study of drug retailer utilization, distancedecay gradients were determined for origin of tripand place of residence of clients in relation to AddisAbaba's population .

Distance decaySteep distance decay gradients of patient travel

have been reported in developing countries . The fewstudies carried out to date associated this with tra-ditional healing alternatives, cost of travel and treat-ment and sociocultural harriers between ill personsand modern medical facilities in rural areas [40] . Inurban Khartoum,'Omdurman Hebert and Hijazi [41]identified socioeconomic level and availability of

10090

80

70

ae 60

50

40

30

20

10

- (A)

100

(a)

--- Origin of trips- - Reidences of clients-Addis Ababa population

-1

11

so-

- 1

U

0 1 1 2 3 4 5̀ 6 7 8 9

0

Utilization of pharmacies and pharmaceutical drugs

659

7 3 4 5 6 7 8 9

Distance Ikml

Dl stance Ikml

medical facilities with utilization patterns in differentsections of the city .

Mapping residence locations of clients of 8 largerpharmacies at the kebele level revealed fairlyprogressive and consistent distance decay . However,the fact that only 39% of the 1040 clients studied inthe 10 pharmacies in May had made the trip directlyfrom home (Table 4) prompted us to determine originof trips and compare the distance decay gradients ofthese two parameters in relation to population distri-bution in order to evaluate per capita utilization . Theresults, based on data for 815 clients of 2 governmentpharmacies (No. 1 and the new shop of the munici-pality) and 2 private pharmacies (Nos 5 and 9) areshown in Figs 3-7. Overall, 69% of the 8I5 clientsmade the trip from places within l km of the pharma-cies and 14% came from the 1-2 km distance band .The remaining 17% of the clients showed aprogressive distance decay pattern right to the city

Fig . 3 . Relationship between distance decay gradients of origin of trips . residences of clients and AddisAbaba's population . (A) Government pharmacy No . I . (B) Private pharmacy No . 5 . (C) The new

government pharmacy . (D) Private pharmacy No . 9 .

Anotherdrug

retailerShop?mgnearby Other'

Ruralarea+

Noanswer

Total No .of clients

16 19 36 616

1 10 41-4

78 16 114 46 1040

6 6 0

HELNLT KLoos ei a+.

limits . By comparison, only 40% of the residences ofthese clients were within I km of the pharmacies,22% between I and 2 km and 34% at greater dis-tance. Considerable variation was noted betweenpharmacies, reflecting the influence of urban struc-ture, intervening opportunities and pharmacy-relatedfactors. The number of clients making the trip fromlocations within the I km radius varied between 60°x°[Fig . 4(A)] to 83% [Fig . 7(A)] . Residences of all clientpopulations of the 4 pharmacies were more dispersedthan origin of travel, only 28-55% of them fallingwithin the I km radius, indicating efforts by thepopulation to minimize transportation cost and time .The greatest differentials between location of resi-dences and origin of trip were noted for the 2 retailersin the Merkato . where large numbers of people workand do their shopping . According to the 1984 popu-lation census, only 14% of the population livedwithin I km of the 2 retailers in the Merkato and aneven smaller proportion (4.2%) near the other 2retailers, with 74.9% at 2-6 km distance (Fig. 3) .These sharp gradients in per capita utilization ofpharmacies indicate that Addis Ababa residents arehighly distance conscious and tend to use pharmaciesin their neighborhoods and places of work andshopping whenever possible .

(A)

Number of clients.

1

• 2-5 .

• 6-10

•

11-20

Governmentpharmacy No. 1

0

1

2

3 km

Urili_ation of government and private pharmacies

Government pharmacies served, on the average,both larger number of clients and kehele (P <0 .65)than private pharmacies . These differences were evengreater than the interview results indicate, due tounderreporting of clients in the former . This wasconfirmed by analysis of daily sales records and clientcensus data during the December survey . Whereas2971 and 1453 clients purchased drugs in governmentpharmacy No. I and the new municipality retailer .respectively . only 329 persons were recorded in pri-vate pharmacy No . 5 and 374 in private pharmacyNo. 9. The large client load of the new pharmacy .which was opened at a new location only 5 monthsprior to the survey, illustrates the rapid growth of thisand the other municipality-owned retailers . The smallnumber of clients in the 2 hospital pharmacies maybe explained by a combination of restricted access tothe public and the location of this two-pharmacyhospital in a low-density residential area withoutcommercial centers. Government pharmacies alsoserved more people from rural areas, half of the totalbeing recorded in the 2 pharmacies in the Piazza andMerkato areas (Table 1). The decline of private drugretailers is also shown when results are compared

Fig . 4(A) Origin of trip to government pharmacy No . I .Figs 4-7 . Origin of clients trips and location of their residences, by kehele.

161

with those of the 1973 study, when the 3 sampledretailers in the Merkato and Piazza all reportedincreasing sales. In 1984 two of these retailers com-plained of decreasing sales and the third said saleswere stagnant . All 3 retailers attributed the downtrend in their business to the presence of governmentpharmacies nearby .

The rapid growth and wide use of the municipalityoperated pharmacies reflects the public's growingawareness and appreciation of the lower prices,greater availability of pharmaceuticals, and certainqualitative advantages they offer . Of the 1040 clientsinterviewed in May, 284 (27 .3%) did not obtain allthe drugs they wanted . Of these . 156 clients wereunable to get drugs due to their unavailability or highprices-significantly more of them in private thangovernment shops (Table 3) . These findings weresubstantiated by the results of the December survey .When asked why they had come to the pharmacieswhere they were interviewed, about one-third of allclients of government pharmacies mentioned lowdrug prices. Another third gave various other reasonsassociated with government retailers, includinggreater availability and freshness of drugs (due totheir rapid turnover), good service, public ownershipand availability of credit . None of the clients inter-

Utilization of pharmacies and pharmaceutical drugs

20 I

3

24

25

B10

9 1

17

16

Fig . 4(B) Residence of clients of government pharmacy No . E

6 6 1

viewed in the private shops mentioned any of thesefactors, stating that they had come primarily due tocloseness of retailers to their homes or places of work,for personal reasons, or due to mere chance . "Ialways come here" was a common answer by clientsof private pharmacies, reflecting their tong associ-ations with these older drug retailers . Casual obser-vations in various pharmacies and druggist shopsshowed that many people patronize private shops asexpressions of social relations of long standing . Ex-tended visits by clients with pharmacy owners andpersonnel due to friendship, kinship or ethnic ties,favors in the form of price reductions, free samples,the sale of prescription drugs without prescriptionsand the sale of more than the maximum number ofdrugs allowed per person (5) all serve to develop andmaintain personal client/pharmacy relationships .They are of lesser importance in the newer govern-ment pharmacies, where personnel have no profitmotive since they are paid salaries only and experi-ence high turnover rates . Equally important. govern-ment pharmacies must account for all sales trans-actions daily and in monthly internal audits .minimizing the danger of illegal sales [42). Whereasall government retailers give receipts for purchases toclients, only 2 private retailers were observed to do

66 2

1

Number of clients

• 2-5• 6-10ell-20

0 Private pharmacy No . 5

HEL%t T KL005 et a! .

so. Failure of the private sector to use detailed andaccurate accounting methods has prompted the gov-ernment to levy sales taxes on the basis of estimatedsales, resulting in frequent tax evasion charges .

Location of pharmacies and medical care facilities andurban structure

The greater utilization of centrally as opposed toperipherally located pharmacies (Table 4) is due to acombination of factors including population density,location of places of employment and medical treat-ment centers, transportation networks and prevailingshopping behavior. Pharmaceuticals are commonlypurchased as part of multi-purpose trips to theMerkato, the hub of economic activity in Ethiopia[43] . and the Piazza, another popular shopping area .People from rural areas also converge here for buyingand selling of goods and social activities . The highpopulation densities, the numerous places of employ-ment and convergence of the major transportationroutes in the areas of the Merkato and Piazza allcontribute to concentrating the resident and dailycommuting populations in the city center . Multi-purpose shopping trips and stop-overs in pharmacies

Fig. 5(A) Origin of trip to private pharmacy No . 5 .

on the way to or from work are common practices ina population concerned with minimizing travel costand to some extent also time . Specific aspects ofshopping behavior are considered below .

Thirty-four of the 44 hospitals, clinics and healthcenters in Addis Ababa are in the 14 centrally locatedkeftegna, most of them clustered in the same neigh-borhoods as the drug retailers . Prescriptions areissued by physicians in all 3 types of medical facilities .In our sample 23.2% of all clients purchased drugswith prescriptions . Some patients obtain their drugsimmediately after they see the physician while otherswait days or weeks until they can afford them or afterthey have shopped around town for the properpharmaceuticals at the lowest price . One reasonpeople resort to over-the-counter drugs and self-medication is the fear of high-cost prescription drugsand overprescribing. Our data show that whereas themean cost of all drug purchases without prescriptionswas 3.33 Birr, that with prescriptions was 8 .91 Birrand this pattern was found in all 4 pharmaciesstudied in December. Solomon Ayele [44] identifiedthe long waiting lists of the clinics and high physicianfees as an additional factor influencing the acces-sibility of modern health services .

In 1984, general practitioners in Addis Ababacommonly charged 3-5 Birr per visit and specialists10-15 Birr . The concentration of medical services inthe city center is more pronounced than shown inFig. I since most of the illegal clinics that are run byhealth assistants, laboratory technicians and even laypersons are also concentrated in the inner city-alongwith injectionists . The total number and precise dis-tribution of all services of this transitional medicalsystem [45] is not known . According to local healthofficials, illegal practitioners constitute a major healthproblem in Addis Ababa due to the wide range ofclinical procedures they perform in hygienically un-acceptable settings and their use of crude instrumentsand practices in performing uvulectornies, clitori-dectomies, circumcisions, tooth extractions and cau-terizations, among others . Moreover, they commonlyillegally prescribe and use pharmaceuticals that theyobtain from official drug retailers . These and otherpractices were described by Freij er at. [46] in one partof Addis Ababa . In a first step to control illegalpractitioners, health officials removed the signs oftheir clinics without however closing them . Our sur-veys and interviews with pharmacists indicate thatthis underground but socially accepted medical sys-tem obtains pharmaceuticals primarily from private

Utilization of pharmacies and pharmaceutical drugs

Fig . 5(B) Residence of clients of private pharmacy No . 5 .

663

retailers in the Merkato and other districts in theinner city . Failure of private retailers to preparedetailed sales records allows them to sell unnoticedlarge quantities of drugs, especially antibiotics, anal-gesics and antacids, often in clinic packs of 1000tablets or capsules . Drug peddlers sell their ware notonly to illegal clinics and individual practitioners intheir homes but also directly to an uninformedpublic. In Addis Ababa per capita self-medicationinvolving pharmaceuticals was found to be 2-4 timeshigher than in rural areas, which was attributedto the greater availability of and demand forpharmaceuticals in urban areas [47] .

As part of the general retail trade and broadercommercial scene in Addis Ababa more than two-thirds of all pharmacies are located on major roads,nearly all of which constitute the outward radiatingboundaries of keftegna (Fig. I) and are major busand small vehicle routes . The advantages of locatingpharmacies on major roads and intersections and atbus stops have been known by owners for a longtime, contributing to the persistence of the pharmacydistribution patterns of the 1960s and 1970s with itsdowntown clusters [48] . The Addis Ababa Munici-pality also recognized the importance of accessibilityof pharmacies and placed pharmacy No . I in the

664

I

Number of clients. 1• 2-5• 6-10•

11-20

a The new government pharmacy

HELNUT KLaUS et p/ .

Fig . 6(A) Origin of trip to the new government pharmacy,

center of the Merkato, next to the major shoppinghalls and the central bus stop . Government pharmacyNo. 2 and the new pharmacy of the municipality wereopened at major bus stops, and the 2 remainingmunicipality shops were located near the major inter-national hotels and the large stadium in Keftegna 14 .As a group, however, government pharmacies weremore accessible than private ones (Table 5), furtheremphasizing the importance of ownership in theirutilization .

Pharmacy-related factors other than type ofownershipand location

Large differences existed between pharmacies inthe number of drugs in stock, number and type ofemployees, and age of shops . The largest numbers ofbrand names and generic drugs were reported by the3 municipality pharmacies . The 2 hospital pharma-cies and private pharmacies Nos 7. 8 and 10 reportedsmaller numbers of these drugs, while the fewest werefound in private pharmacies Nos 5, 6 and 9 (49] .Pharmacy No. 7. also a drug importer, had greateraccess to pharmaceuticals, and pharmacies Nos 8 and10, located in relatively high-income residential areas,carried a wide range of prescription and high-cost

I--

imported brand name drugs to meet the higher healthdemands of their catchment area populations . Al-though no detailed information could be obtained onsales volume, especially in the private shops, numberof drugs in stock and number of clients served inindividual pharmacies (Table 5) appear to be onlymoderately strong indicators of the relative success ofdrug retailers . Thus although number of drugs andnumber of clients served were highest in governmentpharmacy No . 1, this shop carried fewer brand drugsthan either one of the other 2 municipality pharma-cies studied . These differences reflect the munici-pality's effort to meet the needs of the population byincreasing the sale of cheaper generic drugs in theMerkato and other poorer sections of the city andproviding the higher socioeconomic segments of thepopulation with better known imported brands .Socioeconomic influences on drug utilization arediscussed in greater detail below . The relatively largenumber of drugs kept in stock by government phar-macy No . 4 may be explained by the need for a broadrange of drugs in hospital settings regardless of thenumber of clients served .

Number of staff was fairly closely associated withnumbers of clients served by individual retailers . but

considerable variation was noted in regard to profes-sional qualifications and the extent to which familymembers were employed (Table 5) . All governmentpharmacies were headed by pharmacist managerswith university pharmaceutical training. They weresupported in their work in most shops by druggistsor a second pharmacist, 3-6 dispensers, cashiers .accountants and cleaners, all unrelated persons work-ing on a contract basis. Four of the 6 privatepharmacies (Nos 7, 8 . 9, 10) were staffed by universitytrained pharmacists and the remaining two (Nos 5and 6) by pharmacists trained on the job during the1940s and 1950s (Table 5) . Employment of familymembers and relatives, a widespread practice amongprivate retailers in Addis Ababa . has the advantageof providing for more loyal and flexible staff regard-ing business practices and working hours re-spectively .The older, private pharmacies had a more stable

clientele than the newer, rapidly growing governmentpharmacies. Having met the drug needs of theirrespective neighborhoods for 25 years or more andhaving become well respected pharmacists, these re-tailers are widely known and respected in the commu-

Utilization of pharmacies and pharmaceutical drugs

Fig. 6(B) Residence of clients of the new government pharmacy .

665

nity (Table 5) . Informal interviews with some of theirclients indicate that especially older persons valuetheir associations with these well established pharma-cies and their personnel, many of whom they havesought for medical advice for years . Late in 1984,pharmacy No. 8 was acquired by the municipality,resulting in changes in personnel and some drugs, butthe impact of this change from private to governmentownership on utilization is not known .

Origin and purpose of trip, type of transport and illness

According to the results of the May 1984 study,about 40% of all clients had come directly fromhome, more than half of the total sample havingmade the trip to 7 of the 10 study pharmacies fromplaces of employment, markets . shops and places ofsocial engagements . Eighteen percent of the clientsgave places of employment as origin of their trips todrug retailers and 16% clinics or hospitals . withsmaller proportions mentioning social visits, otherdrug retailers and shopping nearby (Table 4). Thesecond most important places of origin . besideshome, of clients to the 2 pharmacies in the Merkatowas the open market and shops nearby . In the 2

666

'Numbers in parentheses indicate numbers of family members or other relatives .tNumbcrs of clients in parentheses are based on an estimated 10-15% under-reporting of all clients during interviews is 3 pharmacies and

30% under-reporting in government pharmacies Nos I and 2 .'Estimated number of drugs, based on interviews with pharmacists or owners and personal observations .

§This pharmacy was taken over by the municipality from a private owner in 1982.

hospital pharmacies most clients had come withprescriptions from the outpatient clinic in the samehospital .

Due to the majority of the clients travelling topharmacies from places other than their places ofresidence, a follow-up study was undertaken in De-cember to examine relationships between origin ofpharmacy trip, distance, type of transport, purpose oftrip and illness . Seventy-four percent of all clients of

HELMUT KLoos e( al .

Table 5 . Pharmacy-related factors and number of clients served

the 4 pharmacies studied walked, with the largestproportion coming from their places of residence andshopping or social visits . Ninety-six percent of allpedestrians walked up to 2 .9 km and only two (0 .3%)of them walked more than 5 km, although the relativeimportance of the different source areas and meandistances walked varied considerably among individ-ual pharmacies, largely reflecting their accessibilityand urban structure (Table 6) . Whereas most people

Table 6 . Distances travelled by 815 clients of 4 pharmacies, by mode of transport and origin of trip

No . ofdrags

in stockNo . of

personnel'Age of

pharmacy Accessibility

No . ofclientsserved-

Governmentpharmacies

I 893 I Pharmacist ! manager, 3yr Excellent 441 (630)

928

1 Druggist,6 Dispensers .I Accountant,2 Cashiers

I Pharmacist/manager, 2yr§ Excellent 402(571)

New 936

I Druggist,4 Dispensers,I Accountant,2 Cashiers

1 Pharmaeisymanager, 5 months Excellent No data formunicipality 4 Dispensers. May surveypharmacy I Accountant.

3 795

2 Cashiers

1 Pharmacist/manager, 27 yr Good I33(178)

711

2 Druggista,3 Dispensers,I Cashier

I Pharmacist/manager. 27 yr Good 47 (44)

Privatepharmacies

5 150-200(

1 Pharmacist

I Pharmacist/owner, 32 yr Excellent 131 (146)

6 450 ;2-3 Dispensers (2)

1 Pharmacist, 30 yr Excellent 105(117)

7 850'+1-3 Dispensers (2)I Pharmacist, 33 yr Good 151 (170)

8 800$2 Dispensers (l)

I Pharmacist/owner, I3yr Excellent 1%(214)

9 550$3 Dispensers (1)

I Pharmacist/owner . 24 yr Excellent 16 a (116)

10 85012 Dispensers (2)

I Pharmacist/owner IIyr Good .̀4(71)I Dispenser

Walking TaxiMode of transport

Own car or service car Bus

Distanceclass(km) Home Work

Shopping .socialvisitsetc. Home Work

Origin of tripShopping,

socialvisits

Work

Shopping,socialvisitsetc. Home Work

Shopping.socialvisitsetc Total 01.)etc . Home

0.5 62 74 173 8 7 13 6 1 0 0 00 .5-0.9 79 37 43 6 2s 12 5 3 0 0 0

561 (69)

1 .0-2 .9 84 IS 25 9 7 5 0 8 1 6 175 (22)3 .0-4 .9 IS 5 1 3 3 3 10 1 ' 53 (7)5 .0-6 .9 1 0 3 0 1 5 3 1 22 (3)7 .0+ 0 0 1 0 1 0 1 1 0 4 (0.5)

241 127 236 46 -2 38 20 10 6 24 6 9 815 (100)

I

I

Utilization of pharmacies and pharmaceutical drugs

travelling on foot to the 2 retailers in the Merkatohad come from shops and markets, most pedestriansusing the 2 retailers in Mekagelegna District camefrom home, with mean distances travelled 0.5 and1 .0 km in the former and 1 .2 and 1 .3 km in the latter .Taxies were the second most commonly used type oftransport (17%) . Only 36 persons (4%) used theirown cars or service cars and 39 came by bus . Meandistance travelled by pedestrians was 1.1 km, by taxi2.7 km, by private car or service car 39 km and bybus 5.8 km. This mix of transport means and traveldistances reflects cost and time budgeting strategiesand availability of motorized transport .

All reported illnesses were categorized arbitrarilyinto light and severe in the assumption that theirseverity and thus the need for life-saving drugs wouldincrease with distance due to scarcity of drug retailersin the peripheral sections of the city . Light illnessesand diseases include acute upper respiratory diseases,most gastro-intestinal diseases, venereal, skin and eyediseases as well as light wounds and arthritis . Severediseases include liver and kidney diseases, malaria,diabetes, heart disease, cancer, tuberculosis andpneumonia . Limitations of this classification wereour inability to determine the severity of individualillness episodes and the large number (77) of ver-

Fig. 7(A) Origin of trip to private pharmacy No . 9.

66 7

nacular categories given by clients . Hod bashua (lit-erally 'stomach disease'), for example, coven a widerange of gastro-intestinal problems, including dys-entery, ill-defined stomach pain (often due to liverdisease), intestinal helminthiasis, and protozoal andbacterial infections. These signs and symptoms mayalso be described as rekmat (dysentery), hod .turret(pain in the stomach) or gubet bashita (liver disease) .No significant differences in types of illness werenoted between clients travelling less than 1 km andthose coming from places further away. This may bedue to a combination of inadequate diseaseclassification, the small proportion of cases withsevere diseases (4% of all clients), the many drugretailers in the city. which represent intervening op-portunities for persons seeking life-saving drugs andthe high proportion (38%) of clients purchasingdrugs not for themselves but for other persons,apparently mostly relatives, friends and neighbors .When considering individual diseases, however,larger proportions of persons with diabetes, venerealdiseases and dysentery travelled more than I km (andup to 9 km), although differences were not statisti-cally significant . This pattern may be due to thegreater availability and lower cost of the relativelyexpensive antidiarrheal, T.B ., and higher antibiotic

668

HELStLT KLOOS et al .

(BI

Number of clients1

• 2-506-10

011-20

O Private pharmacy No . 9

0

1

3 km tI

Fig. 7(B) Residence of clients of private pharmacy No . 9.

drugs in the municipal pharmacies than the periph-erally located private pharmacies, druggist shops andrural medicine vendor shops . It must be kept in mind,however, that the surveys were not population basedand therefore unable to reveal per capita utilizationrates in relation to health needs in different sectionsof the city .Although clients economized on travel time and

cost, often waiting days or weeks to purchase drugsuntil other family or household needs warrantedmulti-purpose trips to the markets, shops and placesof business, many of them shopped around for drugsunavailable or too expensive in the first pharmacy oftheir choice. Forty percent of the 1040 clients studiedin May had visited more than one and up to 4 drugretailers . Of the 284 clients who failed to obtain eithersome or all of the drugs they wanted in the 10pharmacies where they were interviewed, 216 (76%)found that they were unavailable. 52 (18%) con-sidered the available drugs to be too expensive orlacked money and 16 (6%) lacked prescriptions(Table 3). Interviews with pharmacists indicate thatclients themselves, because of preference for specificbrands that have been replaced by generic drugs orother brands and their unwillingness to use substi-tutes, contributed to the perceived unavailability of

drugs. This was substantiated by our surveys, duringwhich many people brought empty medicine contain-ers when asking for drugs, usually rejecting efforts bypharmacists to sell them comparable substitutes . Theincreased emphasis on generic and essential drugs bythe Ministry of Health will thus also increase the needfor education of clients by pharmacy personnel toavoid unnecessary difficulties in drug distribution andin meeting the health needs of the population .

Socioeconomic factors

Comparison of clients living in Teklehaimanot andMekegalegna districts in regard to level of education,mean number of drugs purchased per client, meanprice of each purchase, mean price of individualdrugs, proportion of clients with prescriptions andproportion of clients purchasing antibiotics gavevariable results . Whereas level of education, meanprice of drugs and mean price of individual drugswere significantly higher (P < 0 .05) for persons livingin Mekagelegna than Teklehaimanot, no significantdifferences were found in mean numbers of drugspurchased per client and proportions of clients hav-ing prescriptions or purchasing antiobiotics .Differences were greater between the 2 private andthe 2 government pharmacies, and these differences

remained when clients living outside the 2 studydistricts were included in the analysis (Table 7) . Thesepatterns suggest, considering the generally highersocioeconomic level of the Mekagelegna population,that pharmacy populations are not necessarily repre-sentative of the population in the community . Thesmall number of drugs purchased per client in allpharmacies corresponds with the relatively smallnumber of drugs kept in the homes of Ethiopians(<1), with only 50% of Addis Ababa householdsstoring any pharmaceuticals (50] . The mean numberand price of drugs purchased by rural clients was evenlower than those for Addis Ababa clients in all 4pharmacies, but it is not clear to what extent this isdue to socioeconomic level, differences in distancedistribution or use of traditional medicines . Alsosignificant from the point of drug distribution is thereduction of bulk drug purchase by rural visitorswhen compared to our findings in 1973 [51]. Acombination of greater availability of drugs in ruralareas through the expanding primary care system andgovernment regulations prohibiting the sale of largeamounts of drugs to unlicensed individuals mayexplain this apparent trend .

In individual pharmacies as well, level of educationwas associated with mean cost of drug purchases .This appears to be due not only to the low incomelevels of the general population but also to severaldrug purchasing strategies the population of AddisAbaba has developed in response to scarce resources.The cost consciousness of many people prompts themto obtain the cheap and widely-used anti helminthics,antidiarrheals, antacids and analgesics in partialdoses, and thus at lower prices, from private retailers,who freely sell them upon request . This practice isbeing discouraged by government retailers in view ofincreasing drug resistance of pathogens in AddisAbaba [52] . For more expensive, pre-packaged items,which cannot be sold in partial doses, including

Utilization of pharmacies and pharmaceutical drugs

Table 7 . Number and price of drugs purchased per client, number of prescriptions and traditional medicines and level of education ofclients of pharmacies who live in Teklehaimanot, Mekagelegna and other districts

No . of clientsMean No . of

drugs perclient

Mean price ofeach purchaseper client(Birr)

Clients withprescriptions

Clients usingtraditionalmedicinessimultaneously

Level of educationUp to 8th

gradeAbove 8th

grade'Number of clients, with percentage in parentheses .

5.5.54 22 t E

syrups and other liquid preparations, many people goto government pharmacies instead, which sell themcheaper than private retailers . More affluent (andmore educated) persons tend to purchase all theirdrugs in government shops. That mostly poorerpeople purchase small over-the-counter items in pri-vate pharmacies is indicated by our observations onthe type and cost of drugs they purchased, as well asthe senior author's observation of drug purchasingbehavior in other private shops [53] . These drugpurchasing strategies have been developed in re-sponse to different pricing structures, the increasinglydifficult position of private retailers, financial stress ofa population experiencing a high disease burden, andthe absence of a national drug policy . They providepoorer people with some latitude for action and theymay be expected to persist as long as private andgovernment retailers compete for business in theabsence of an effective national drug policy thatregulates drug distribution and dispensing practices .

Another indicator of financial stress of ill personsin Addis Ababa is the monthly pattern of pharmacyuse. Weekly fluctuations of pharmacy utilization andour interviews with pharmacy personnel indicate thatresidents tend to purchase drugs soon after pay-day around the beginning of each month, while theyhave sufficient money . This excludes governmentemployees, students and qualifying poor people,who can obtain drugs free of charge in governmentpharmacies .

Only 39 of the 815 clients interviewed in Decembersaid that they used traditional medicines concurrentlyfor the same illness. All but 3 persons allegedly usedtaenicidal plant materials, including kosso (Hageniaabyssinica), enkoko (Embelia schimperi) . metere(Glinus lotoides) and kechemo (Myrsine africana), allof which are widely sold in the Merkato and otherlocal markets, are cultivated in gardens or are gath-ered from trees, bushes and roadside places in the

669

Government Private New Privatepharmacy pharmacy government pharmacyNo . I No . 5 pharmacy No .9

Tekleh .Otherdistrict Tekleh .

Place of residenceOther Other

district Mekag .Otherdtstrictdistrict Mekag.

139 157 SI 82 149 81 941 .25 1 .23 1 .03 1,00 1 .24 1 .03 1 .02 X 1 .09

3 .65 6 .30 0 .86 1 .00 5.97 4 .22 3.99 7 .34

38(28)' 48(31) 4p5) 0 37(25) 31(39) 12(13) 7C^--)

7 6 5 8 3 I

88(63)' 75 (48) 68(84) 63(77) 38 (25) 39(48) 51(54) b(47)

51(37) 82(52) 13(16) 19(23) 111(75) 42(52) 43(46) I'r(53)

6711

vicinity of Addis Ababa [54] . The use of these andother home remedies is probably more widespreadthan the survey results indicate due to the nature andsetting of the interviews in modern pharmacies . Thisis suggested by the general reluctance of many people,especially more educated persons, to discuss thistopic. The utilization of traditional healers and theirinfluence on drug purchasing behavior could not beexamined in this study . No accurate data are avail-able on the distribution of various types of healers inAddis Ababa, including the debtera (church scribe)and sogesha (bone setter and surgeon). A pre-liminary survey of traditional heaters by the De-partment of Traditional Medicine and Healing Prac-tices of the Ministry of Health as part of an effort toregister them and seek their cooperation in primaryhealth care identified 182 healers in Addis Ababa,who constitute 4% of the 4603 healers registerednation-wide. These figures must be considered asunderestimates in view of the general suspicion andlack of cooperation displayed by debtera, wogeshaand other healers toward cooperation with themodem medical system during that survey [55] .

DISCLSSION

The utilization of drug retailers, including clientflow patterns, exhibits some similarities as well asimportant differences with health behavior involvingmedical treatment facilities . Due to the usual fric-tional effect of distance, pharmacy-related factorsand intervening opportunities, fairly distinct andpredictable service areas may be identified in urban(and probably rural) environments. However, thenature of pharmaceuticals as a market commodityavailable to the public at retail shops and thus in anenvironment less controlled by medical institutions[56] gives greater latitude of action to clients thanpatients seeking care in treatment facilities . Similarly,as manufactured products. pharmaceuticals are muchmore uniform in their actual and perceived quality orhealing properties than medical practitioners, con-tributing to the relatively small service areas of drugretailers. In Addis Ababa drug purchasing is wellintegrated into peoples' daily routines, particularlyemployment, shopping and social activities, usuallyas a secondary activity, thus contrasting with themore scheduled and planned trips for medical care intreatment facilities . Such multi-purpose trips reducetransport costs and travel time and also facilitateshopping around for the desired drugs in a city whereself-diagnosis and self-prescription are widespread .Origin of trip is thus a supplemental and possiblymore useful distance parameter than residence ofclients . Concern over high cost of medical care in apopulation experiencing high disease burden andlacking adequate resources has also facilitated theexpansion of the recent government pharmacies, al-though government measures other than price reduc-tions also contributed to the competitive edge of theseretailers over the private sector . By locating drugretailers at the most accessible locations in .AddisAbaba. monopolizing the sale of some drugs andreducing prices, the government has employed strate-gies of the free enterprise system to achieve itsobjective of providing more equitable distribution of

HELMUT KLOOS et al.

health care under socialism . In the absence of a wellformulated national health program, the drug retailtrade in Ethiopia retains, for the time being, elementsof the free enterprise system that facilitate procure-ment and distribution of a wide range of pharma-ceuticals at reasonable prices . General lack ofinefficiency and corruption of government drug re-tailers in Addis Ababa contrasts with the situation inother developing countries [57] and appears to be duein part to a combination of competition, not onlybetween the government and private sectors, but alsobetween the Addis Ababa Municipality and Ministryof Health programs, and the development of govern-ment regulations aimed at improving drug distribu-tion and sales practices . Problems of illicit drug salesand underdosages associated with private pharmaciesare largely due to survival of certain practices frompre-revolutionary times .. lack of enforcement of exist-ing government regulations in the private sector andstrategies developed by the increasingly threatenedprivate retailers .

The observed steep distance decay gradients oftravel to pharmacies, with the great majority of allclients originating from within I km of shops needs tobe confirmed by studies in more peripheral sectionsof the city before recommendations on where tolocate new retailers in the suburbs can be made .Perhaps more crucial is the need for population basedstudies to determine health resources and needs . Suchstudies are now being carried out in several kebeleswith different socioeconomic levels and located indifferent parts of the city and preliminary results havebeen reported [58, 59] .

Acknmdedgements-The authors want to thank Ato EshetuWondimagegnehu, Head of the Department of Pharmacyand Laboratory, Ministry of Health, Ato Abebe En-dedasew, Head of the Department of Health and SocialServices . Addis Ababa Municipality, and the owners, man-agers and other personnel of the pharmacies studied fortheir cooperation and support and the clients who gavefreely of their time . We are also endebted to the Departmentof Geography, Addis Ababa University for material supportand to Dr Oscar Gish for stimulating discussions .

REFERENCES

I . Pyle G. F . Applied Medical Geography . Wiles, NewYork, 1979: Shannon G. W . and Dever G. E. A . HealthCare Delirerr : Spatial Perspectives . McGraw-Hill. NewYork, 1984; Gesler W . M . Health care practitioner usein Calabar. Nigeria . Soc . Sci . ;Med. 13D, 23-30. 1979 ;Stock R . Distance and the utilization of health facilitiesin rural Nigeria . Sac . Sci. Med. 17, 563-570, 1983 .

2 . The terms 'drugs' and 'pharmaceuticals' arc used inter-changeably here to denote medical uses of these prod-ucts although there is increasing preference for using thewords 'medicine' and 'pharmaceuticals' instead of'drugs' . in an effort to distinguish more clearly betweentheir licit and illicit uses. See for example U .S . De-partment of Health. Education and Welfare. PublicHealth Service . Guidelines for AnaAsis of Pharma-ceutical Supply System Planning . DHEW PublicationNo. (PHS) 79-50086 . Rockville, Md .

3 . World Health Organization . Global Strategy for Healthfor All by the Year 2000 . World Health OrganizationDocument EB67.PC.W'P 3 . Geneva . 1980 .

4 . Report of a WHO Expert Committee. The use ofessential drugs . Technical Report Series, No . 685, 1983 ;Report of a WHO Expert Committee. The selection ofessential drugs. Technical Report Series, No . 615,1977 .

5 . Gish O . Guidelinesfor Health Planners . Tri-Med Books,London . 1977; United Nations Industrial DevelopmentOrganization . Notes on UNIDO activities relating topharmaceutical products in the context of primaryhealth programmes . UNIDO Document UNIDO;IOD.207 . 23 August, 1978 ; Peretz S . M . Pharmaceuticaland health problems in the least developed countries .Pharmacy Int . No. 2, 8-12, 1981 .

6 . Provisional Military Government of Socialist Ethiopia,Ministry of Health . Comprehensive Health Services Di-rectory 1974 E.C. (1981/82 G .C .), p . 14 . Addis Ababa .1982 ; Bowers A, Cost-effectiveness : are studies needed?In Pharmaceuticals in Developing Countires : Cost-Effectiveness of Pharmaceuticals. Proceedings of theInternational Federation of Pharmaceutical Manu-fdcturers Associations' Eleventh Assembly, pp . 103-I11 .International Federation of Pharmaceutical Manu-facturers Association, Zurich, 1982 .

7. Bannerman R . H ., Burton J. and Ch'en Wen-Chum .Traditional Medicine and Health Care Coverage . WorldHealth Organization, Geneva. 1983: Good C. M . .Hunter 1 . M ., Katz S . H. and Katz S. S . The interfaceof dual systems of health care in the developing world :toward health policy initiatives in Africa . Soc . Sci . Med.13D, 141-154, 1979 ; Neumann A . K. and Lauro P .Ethnomedicine linking . Soc . Sci. Med. 16, 1817-1824,1982 .

8 . Bledsoe C. and Goubaud M . The reinterpretation ofWestern pharmaceuticals among the Mende of SierraLeone . Soc. Sci . Med. 21, 275-282, 1985 ; FergusonA. E. Commerical pharmaceutical medicine and medi-calization: a case study from El Salvador . Cult . :Med.Psychol . 5, 105-134, 1981 .

9. Low S . M. The urban patient : health-seeking behaviorin the health care system of San Jose, Costa Rica . UrbanAnthrop . 10, 27-51. A comparison of rural and urbanethnomedicine among the Kamba of Kenya. In Tra-ditional Health Care Delivery in Contemporary Africa(Edited by Priscilla R . and Segall M . H .), pp . 13-56 .Syracuse University Press, Syracuse, 1980 .

10. Gish O . and Feller L. L . Pharmaceuticals for PrimaryHealth Care ; The Supply and Utilization of Drugs in theThird World. American Public Health Association,Washington, D.C . . International Health ProgramsMonograph Series, No . 2, 1979 .

II . World Health Organization. National drug policies .Wld HIM Org . Chron . 29, 337-349, 1975 .

12 . For reviews of works on this topic see Van der Geest S .Anthropology and pharmaceuticals in developing coun-tries . Afed. Anthrop . Q . 5, 59-62, 87-90, 1984 : MelroseD. Bitter Pills ; Medicines and the Third World. Oxfam,Oxford, 1982 ; see also Van der Geest S. The illegaldistribution of western medicines in developing coun-tries : pharmacists, drug pedlars, injection doctors andothers . A bibliographic exploration . Med. Anthrop. 6,197-219 . 1982 .

13. U .S . Department of Health . Education and Welfare, op .cit . [2] .

i4 . For a summary of studies favoring regulation see Vander Geest S. op. cit . [12] : for discussions of health andill health under capitalism and socialism see Elling R . H .Cross- .Varional Study' of Health Systems . Transactions .New Brunswick . N . J ., 1980 : Terns M . The three worldsystems of medical care: trends and prospects. Wld HlthForum 1, 78-86, 1980 ; Ehrenreich J . (Ed .) The CulturalCrisis of tfodern Medicine . Monthly Review Press, NewYork, 1978 .

15 . See for example Silverman M ., Lee P . R . and Lydecker

Utilization of pharmacies and pharmaceutical drugs

671

M. Prescriptions for Death : The Drugging a(the ThirdWorld. University of California Press . Berkeley . 1982 .

16 . Van der Geest S . op . cit . [12] .[7 . Mersie E. Introduction : a note on health and develop-

ment . Ethiop . J. filth Den . 1, 7-9, 1984 .18 . For discussions of the pre-revolutionary health care

system in Ethiopia, see Conacher D . G . Medical care inEthiopia. Trans. R . Sac. Trop . tried. Hyg. 70, 141-144 .1976; and Chang W . P. Development of basic healthservices in Ethiopia . Ethiop . Observ . 12, 230-238, 1969 .

19 . For a brief summary of the major health goals of the10-year plan see Gish O . Some observations abouthealth development in three African countries : Ethi-opia. Mozambique and Tanzania . Soc . Sci. .Med. 17,1961-1969, 1983 .

20. Gizaw T. Introduction : the Ministry of Health andHealth Development . Ethiop . J . filth Dee . I, It, 1984 .

21 . Provisional Military Government of Socialist Ethiopia,Ministry of Health, op . cit. [6], pp . 17, 19 .

22. Hailu M . and Mehari W . Reaching the people : someissues in the utilization of selected hospitals in AddisAbaba . Ethiop . J. Hlth Dec . 1, 65-72, 1984.

23 . The pharmacy regulations of 1964, under that title inthe government publication Negurit Ga_era No. 18,101-116, 1964, remain the most recently publishedregulations . For the only study on the location andallocation of drug retailers in Addis Ababa before therevolution see Kloos H . The geography of pharmacies,druggist shops and rural medicine vendors and theorigin of customers of such facilities in Addis Ababa .J. Ethiop. Stud. 12, 77-94, 1974 .

24 . Editorial . Workshop on health opens in Nazareth .Ethiopian Herald. 22 February, p . I, 1985 .

25 . The official rate of exchange of I Bur was S'- .05 in 1984 .26. Solomon A. The drug business and its implications for

the development of health in Ethiopia . Ethiop . J. Dec .Res. 3, 33-50, 1979 .

27. Bedassa A. The Ethiopianization of pharmacies(1935-1983 G .C .) . Paper presented at the 9th AnnualConference of the Ethiopian Pharmaceutical Association,Addis Ababa, 1984.

28 . Keftegnas are higher cooperative societies of urbandwellers, of which there are 25 in Addis Ababa [seeFig . 4(B)].

29 . Rural medicine vendors are small drug retailers ownedand operated by nurses, health assistants and otherauxilliary health workers .

30 . Kloos H . op . cit . [23] .31 . Pharmacy regulations of 1964, op . cit . [23] .32. Bedassa A . op. cit . [27] .33 . Kloos H . op. cit . [23] .34. Kloos H . op. cit . [23] .35. Students from the Geography 431 class (Research

Methods in Spatial Analysis), which the senior authorwas teaching .

36. The only comparable information on socioeconomicconditions in the 2 woreda pertains to housing con-ditions, showing that 24 .5% of the housing units inTeklemaimanot and 35 .0% in Mekagelegna wereowner-occupied and 70 .1% and 61 .3% of the housingunits consisted of one or two rooms, respectively .Provisional Military Government of Socialist Ethiopia .Central Statistics Office. Addis Ababa Manpower andHousing Sample Survey, 1976 . Statist . Bull. No. 15 .1977, see pp . 46, 53 .

37 . Pausewang S . Methods and Concept of Social Researchin a Rural Developing Society : A Critical AppraisalBased on Experience in Ethiopia. Weltforum, Munich,1973 .

38 . Central Statistical Office, unpublished data . 1983 .39 . Kloos H. op . cit . [23] .40 . Stock R. op . cit . [1] .41 . Hebert D . T . and Hijazi N . B . III-health and Health-care

6 '

HELMUT KLOOS et al.

personnel are responsible for the accuracy of theirrespective transactions .

43. DeYoung M . An African emporium : the Addis Mark-ate . J. Ethiop . Stud. 5, 103-122, 1967 .

44 . Solomon A. op . cit. [26] .45 . Buschkens W. F . L . and Slikkerveer L . J . Health Care

in East Africa : Illness Behavior ojthe Eastern Oromo inHararghe (Ethiopia), p . 53. Van Gorcum, Assen, 1981,used this term to describe the intermediate health systemin eastern Ethiopia which incorporates elements of boththe modern and traditional medical systems but whichis beyond the control of either .

46 . Freij L., Yemane K . . Sterky G . and Wall S . Exploringchild health and its ecology : the Kirgos Study in AddisAbaba . J. Der . Res. 2, Suppl ., 1978 .

47. Yayehyirad K . Drug hoarding and utilization in adeveloping country . Paper presented at the 9th AnnualConference ofthe Ethiopian Pharmaceutical Association,Addis .Ababa, [984 .

48. Kloos H . op . cit . [231 .49. This small number of drugs in private pharmacy No . 5

was due to the death of its owner, resulting in discon-tinuation of drug orders. Normally this retailer was saidto keep about 400 drugs in stock .

Yayehyirad K. op . cit . [47] .Kloos H . op . cir. [23 1 .See, for example Eyassu H .-G . Use and misuse ofantimicrobial drugs. J . Ethiop. Pharmaceut . Ass . 4,14- Id, 1979 ; Warndortf J . A . Adverse effects of somefrequently used topical medications (Abstract) . Ethiop .rated . J. 18 . 130, 1980; Afeworki G.-Y . and YetneberschL. Multiple drug resistance within the Shigella sero-groups. Erhiop. med. J. 18, 7-10 . 1980; Messele G .Increasing incidence of resistance of Escherichia coli toantimicrobials . Ethiop . med. J. 21, 61-69, 1983 .

53 . Kloos H. unpublished data, 1984 .54. Kloos H . et al. Preliminary studies of traditional med-

icinal plants in nineteen markets of Ethiopia : usepatterns and public health aspects . Ethiop. med . J . 16,33-43 . 1978 .

55. Dr Yilma Desta, personal communication, 1984 .56 . Ehrenreich 1 . op . cit . [14] .57 . Van der Geest, op . cit . (12] ; Van der Gecst S . The

efficiency of inefficiency : medicine distribution in southCameroon. Soc . Sci. Med. 16, 2145--2153, 1982 .

58. Kloos H . et al . Buying drugs in Addis Ababa : aquantitative analysis . In Pharmaceutical Anthropology(Edited by Van der Geest S . and Welsh R .) . Reidel,London. In press .

59. Office of the Population and Housing Census Com-mission . Ethiopia 1984 . Population and Housing CensusPreliminary Report, Vol. 1, No . I . Addis .Ababa, 1984 .

in Khartoum Omdurman . Soc . Set. Med. 18, 335-343, 50 .1984 . 51 .

42 . Sales account must be balanced daily by pharmacy-employed professional accountants and all pharmacy

52.