The impact of face-to-face educational outreach on diarrhoea treatment...

HEALTH POLICY AND PLANNING; 11(3): 308-318 © Oxford University Press 1996

The impact of face-to-face educational outreach ondiarrhoea treatment in pharmaciesDENNIS ROSS-DEGNAN,1 STEPHEN B SOUMERAI,1 PRADEEP K GOEL,1 JAMES BATES,2 JOSEPHMAKHULO,3 NICHOLAS DONDI,3 SUTOTO,4 DARYONO ADI,4

LUCIA FERRAZ-TABOR,5 AND ROBERT HOGAN6

'Department of Ambulatory Care and Prevention, Harvard Medical School and Harvard PilgrimHealth Care, Boston, USA, 'Drug Management Program, Management Sciences for Health, Boston,USA, 3Control of Diarrhoea/ Diseases Program, Ministry of Health, Kenya, 4Control of Diarrhoea/Diseases Program, Ministry of Health, Indonesia, SPRITECH, Indonesia, 6 World Health OrganizationProgramme for the Control of Diarrhoea! and Respiratory Diseases, Geneva, Switzerland

Private pharmacies are an important source of health care in developing countries. A number of studieshave documented deficiencies in treatment, but little has been done to improve practices. We con-ducted two controlled trials to determine the efficacy of face-to-face educational outreach in improv-ing communication and product sales for cases of diarrhoea in children in 194 private pharmacies intwo developing countries. A training guide was developed to enable a national diarrhoea control pro-gramme to identify problems and their causes in pharmacies, using quantitative and qualitative researchmethods. The guide also facilitates the design, implementation, and evaluation of an educational in-tervention, which includes brief one-on-one meetings between diarrhoea programme educators andpharmacists/owners, followed by one small group training session with all counter attendants work-ing in the pharmacies.

We evaluated the short-term impact of this intervention using a before-and-after comparison groupdesign in Kenya, and a randomized controlled design in Indonesia, with the pharmacy as unit of analysisin both countries (n = 107 pharmacies in Kenya; n-87 in Indonesia). Using trained surrogate patientsposing as mothers of a child under five with diarrhoea, we measured sales of oral rehydration salts(ORS); sales of antidiarrhoeal agents; and history-taking and advice to continue fluids and food. Wealso measured knowledge about dehydration and drugs to treat diarrhoea among Kenyan pharmacyemployees after training.

Major discrepancies were found at baseline between reported and observed behaviour. For example,66% of pharmacy attendants in Kenya, and 53% in Indonesia, reported selling ORS for the previouscase of child diarrhoea, but in only 33% and 5% of surrogate patient visits was ORS actually soldfor such cases. After training, there was a significant increase in knowledge about diarrhoea and itstreatment among counter attendants in Kenya, where these changes were measured. Sales of ORSin intervention pharmacies increased by an average of 30% in Kenya (almost a two-fold increase) and21 % in Indonesia compared to controls (p<0.05); antidiarrhoeal sales declined by an average of 15%in Kenya and 20% in Indonesia compared to controls (p<0.05). There was a trend toward increasedcommunication in both countries, and in Kenya we observed significant increases in discussion ofdehydration during pharmacy visits (p<0.05).

We conclude that face-to-face training of pharmacy attendants which targets deficits in knowledgeand specific problem behaviours can result in significant short-term improvements in product salesand communication with customers. The positive effects and cost-effectiveness of such programmesneed to be tested over a longer period for other health problems and in other countries.

at World H

ealth Organization on O

ctober 5, 2014http://heapol.oxfordjournals.org/

Dow

nloaded from

http://heapol.oxfordjournals.org/

Diarrhoea treatment in pharmacies 309

IntroductionDehydration due to diarrhoea remains a major causeof preventable morbidity and death among infants andchildren in developing countries.u Since its incep-tion, the World Health Organization Programme forthe Control of Diarrhoeal Diseases (WHO-CDD) hasfocused on improving diarrhoea case management bygovernment health workers and parents. While theseefforts have increased use of oral rehydration salts(ORS) and reduced mortality,1 inappropriate treat-ment in private pharmacies has limited further gainsin health status. These two studies evaluate a WHO-CDD demonstration project designed to enable na-tional CDD programmes to include in their remitprivate pharmacists and licensed sellers of drugs.

Drug retailers are an important source of health carein many parts of the developing world. In additionto selling drugs prescribed by physicians, pharmaciesoften serve as primary sources of information aboutillness and drug therapy. Consumers in many culturesprefer pharmacies and drug shops to other health careproviders because they are conveniently located,waiting times are shorter, staff are less socially distantthan other health workers, and there is no separatecharge for advice.3"6 Increasing privatization ofhealth care in many countries may further increasethe role of pharmacies and drug shops in primaryhealth care.

Numerous studies have shown that pharmacies sell-ing drugs without prescription often do so inap-propriately and with little scientifically substantiatedadvice about medicines or illness.7"9 Since manypharmacists are not allowed to dispense most drugswithout prescription, their training rarely providesadequate information about therapeutics. In addition,although many countries require that pharmacies beowned or managed by a trained pharmacist,customers are often seen by untrained atten-dants.10"13 The lack of scientific knowledge is furthercomplicated in this setting by economic incentives,strong consumer drug preferences, inappropriatecommercial drug information, and intensive pressureby drug company representatives to sell specificproducts.1413

Few studies have focused on methods to improvetreatment in pharmacies.16 A few countries haveestablished social marketing programmes to promotesales of safe contraceptives17"19 and commercialORS20-21 in pharmacies and drug shops, but the

extent of their impact is unknown. One innovativeprogramme in Kenya, jointly sponsored by a drugcompany and the Ministry of Health, reported an in-crease in sales of flavoured ORS marketed in ruralretail kiosks, but the programme was discontinuedbecause it was not commercially viable.22 A study inBangkok showed no impact of mailing educationalmaterial about diarrhoea treatment to pharmacies andretail drug stores.23

CDD programmes have traditionally dealt with thepublic sector. Targeting private sector pharmaciesrepresents a substantial shift in perspective, skills,and training modes. We have previously presenteda behavioural framework for approaching interven-tions in this sector.24 Here, we report short-termresults of a persuasive strategy in two developingcountries.

MethodsStructure of the interventionThis paper describes the results of a field test usingthe WHO-CDD Guide for Improving DiarrhoeaTreatment Practices of Pharmacists and LicensedDrug Sellers (the Guide)23 in two countries, Kenyaand Indonesia. The programme's specific objectiveswere: to increase the capacity of a national CDD pro-gramme to plan and undertake a persuasive trainingintervention; to improve the knowledge of targetpharmacists and counter attendants about diarrhoea,specific drugs to treat it, and proper case manage-ment; and to improve actual treatment practices inpharmacies, specifically to increase sales of ORS, todecrease sales of antidiarrhoeals and antibiotics, toincrease appropriate questioning about diarrhoeahistory and etiology, and to improve advice aboutreferral and prevention.

All programme activities were carried out by im-plementation teams led by the Directors of theMinistry of Health CDD programmes, and includedCDD staff and representatives of university faculty,the National Pharmaceutical Society, UNICEF andother non-governmental organizations, or localprivate sector consulting firms.

The Guide describes a four-stage process to:1) Assess knowledge and actual treatment practices;2) Identify underlying motivations and constraints tochanging practices;3) Design a persuasive educational intervention

at World H



Dow

nloaded from


310 Dennis Ross-Degrvan et al.

through brief face-to-face encounters; and4) Carry out and evaluate the intervention.

Baseline assessment of knowledge, practices andmotivationsAs a first step in understanding the behaviour ofprivate sector drug sellers, the Guide recommendsan interview survey with a sample of pharmacyowners, pharmacists, and counter attendants to assessknowledge about diarrhoea and the drugs used to treatit. Following these interviews, actual practices areobserved through surrogate patient visits by con-federates posing as the mothers of children with diar-rhoea and asking for advice about treatment. Thesevisits have been shown to be reliable methods formeasuring diarrhoea treatment in pharmacies in otherdescriptive studies.7I5-23-26 Confederates are trainedto respond in a standard way in the local vernacularto questions raised by the counter attendant. In thisstudy, confederates described a simple case of waterydiarrhoea of short duration that should require notreatment other than ORS. Confederates purchasedwhatever drugs were recommended during thesevisits, and, after leaving the shop, recorded on a stan-dard data collection form: 1) questions asked by theattendant about signs and symptoms; 2) advice givenabout drugs sold; and 3) advice given about homecase management, prevention, or referral.

In Kenya, the market survey consisted of 70 phar-macies in the towns of Nairobi, Nakuru, Thika,Ruiru, and Kisumu. In Indonesia, the survey tookplace in Jakarta, Bogor, Bekasi, and Tangerang, andthe sample included 19 pharmacies, 22 over-the-counter drug stores, and 8 small scale drug sellers;only the pharmacy results are presented here.

In order to learn more about factors underlyingobserved behaviour, the Kenya team conducted 8focus group discussions (FGDs), four with trainedpharmacists, three with untrained counter attendants,and one with pharmacy owners. The Indonesia teamconducted 6 FGDs with heterogeneous groups in-cluding pharmacy owners, assistant pharmacists, andcounter attendants.

Educational messagesBased on the results of both the market assessmentand the FGDs, the teams developed and pre-testedprinted materials containing the main interventionmessages (see examples in Figure 1). In Kenya, theseincluded three materials promoting ORS, food, andfluids as the recommended treatment for diarrhoea

(a brochure for pharmacy personnel, a poster aimedat mothers to be displayed in the shop, and leafletsto distribute to mothers), and a brochure aimed atpharmacy staff discouraging use of antidiarrhoeals.

In Indonesia, printed materials included a poster anda counter display aimed at customers; the back of thecounter display contained information useful for phar-macy staff to teach mothers about ORS preparation.These materials contained messages promoting ORS,breastfeeding, and feeding in diarrhoea treatment.In-depth interviews with pharmacy staff revealedreluctance to display materials that would overtlydiscourage use of antidiarrhoeals. However, the train-ing in both countries covered the dangers of antidiar-rhoeals and the reasons for not recommendingthem.27

Face-to-face interventionPharmacists and counter attendants were trained inshort interactive sessions by outreach educatorsfamiliar with techniques of effective communication,using training formats adapted to fit local needs andresources. This training model, when used to pro-mote appropriate scientific practice, has been des-cribed as 'academic detailing'.28 To enhance itsimpact, the interventions were supported by sponsorscredible to pharmacists such as the WHO, the nationalPharmacists' Association, a university school of phar-macy, or the national CDD programme. The effec-tiveness of this approach for changing drugprescribing among physicians in developed countrieshas been consistently proven in previous studies.28"32

There have been no published studies of its use inthe developing world, either among trained healthworkers or in the private retail sector.

Outreach educators in both study countries weretrained about diarrhoea and its appropriate manage-ment in pharmacies, data on current practices, andtechniques of interactive communication for adultlearning.28 Four Ministry of Health personnel servedas the outreach educators in Indonesia, while 6 stafffrom the Faculty of Pharmacy and the Medical Train-ing Centre fulfilled this role in Kenya. In both coun-tries, the intervention began with brief one-on-onemeetings with pharmacists/owners discussing keytraining messages and ways to deal with perceivedbarriers to practice recommendations, followed bytraining of all counter attendants in group sessionsof 5-10 attendees organized close to their shops.These sessions (2-3 hours on a single day in Kenya;two days in Indonesia) covered the aetiology of

at World H



Dow

nloaded from



Manage •>DIARRHOwith THREE

ORS + FLUIDS + FOOD

rORALIT CAIRAN TERBAIK UNTUK DIAPF

K

The WHOrrcommcndi'd

trcHtmvnl

Figure 1. Examples of persuasive printed materials used in Kenya and Indonesia

diarrhoea, its effects on the body, and its propermanagement. Trainees received posters and patienteducation materials for display in their shops.Trainees and pharmacy owners were not informedthat the sales practices in their pharmacies would beevaluated.

Research design and sampleIn Kenya a randomized, controlled trial was imprac-tical, because pharmacies are clustered in cities andtowns, and experimental-control contamination waslikely. Therefore, the study utilized a quasi-experimental design,33 with measurement of out-comes before and after training in a sample of studypharmacies and in a comparison group. The sampl-ing frame included all pharmacies located in Nairobi,Nakuru, Kisumu, and Mombasa, identified fromretail pharmacy lists from the Ministry of Health andthe Pharmaceutical Society of Kenya. In Nairobi, 62pharmacies were randomly selected, while all 50pharmacies located in the other towns were included;analyses are limited to the 58 pharmacies in Nairobiand 49 in other towns in which outcomes weremeasured both before and after the intervention.

Training took place in two phases. Nairobi phar--macies (Group 1, n = 58) received the interventionin the first phase, while pharmacies from Nakuru andKisumu (Group 2, n=24) received training in thesecond phase (and were controls in the first phase);pharmacies from Mombasa (Group 3, n=25)received no training and served as controls throughoutthe study period. The training programme includeda total of 90 pharmacists and 162 counter attendantsin study pharmacies. Outcomes were measured in allgroups before and after the first wave of training, andagain after the second wave in Group 2 and Group3 pharmacies only. Thus, we were able to measurethe short-term impacts of training in Nairobi phar-macies compared to concurrent changes in all othertowns, and at a later point in time, short-term im-pacts of training in Group 2 pharmacies comparedto concurrent changes in Group 3.

The Indonesia team selected a purposive sample of87 pharmacies from Jakarta, and the neighbouringcommunities of Bogor, Tangerang, and Bekasi. Phar-macies were first stratified by geographic location andtheir baseline practices, and then randomly assigned

at World H



Dow

nloaded from


312 Dennis Ross-Degnan et al.

to intervention (n=43) and control (n=44) groupsfrom within these strata. Analyses are limited to the42 pharmacies in the intervention group and the 41in the control group in which outcomes weremeasured both before and after the intervention.

In Kenya, to measure changes in knowledge,educators administered a ten-question instrument toall pharmacy assistants before and after training.Items were presented in a different order in the pre-and post-tests. Of the 128 assistants who attendedtraining, 109 completed both questionnaires. A pairedt-test was used to test the significance of im-provements in knowledge as a result of training.

To measure changes in treatment practice, all phar-macies included in the study in both countries werevisited by surrogate patients as described above.These surrogate patients were blind to the purposeof the study, and to the study or control status of thepharmacies. In Indonesia, a single visit was made toeach pharmacy one month before and after training.In Kenya, surrogate patients made two visits per phar-macy per period (2-4 weeks before the first train-ing, and then 2-4 weeks following each wave oftraining). In all cases, a single surrogate patient nevervisited the same pharmacy more than once.

The significance of observed changes in targetpractices were determined differently in the twocountries. In Indonesia, with one visit per pharmacyper period, we computed group averages and usedsimple t-tests to test the significance of pre-post dif-ferences. In Kenya, with multiple visits per phar-macy, we first calculated for each pharmacy theproportion of visits pre and post during which a givenoutcome was observed (e.g. sale of ORS), then wecalculated the change in this proportion between thetwo waves of data collection. Finally, we used t-teststo determine the significance of differences betweenstudy and control groups.

ResultsBaseline assessmentsIn the Kenya pharmacies, respondents to the baselinesurvey reported an average of 27 customers per weekseeking care for diarrhoea, over half (53 %) for a childunder five. In Indonesia, respondents reported 15customers per week seeking diarrhoea treatment, withabout one-third of these (37%) for children.

In Kenya, 87% of respondents understood about fluidloss during diarrhoea, and 67% considered fluidreplacement to be the most important aspect oftreatment. In Indonesia, however, only 21% ofrespondents adequately understood fluid loss, and16% considered fluid replacement of paramountimportance.

ORS was widely available at the time of the surveysin all pharmacies, with up to 7 different productsfound in Indonesian pharmacies, and up to 6 inKenya. However, antidiarrhoeal preparations wereavailable in much wider variety, with 71 differentbrands reported in Indonesia and 22 brands in Kenya.

Figure 2 contrasts the products that surveyrespondents reported they sold to the previouscustomer having a child with diarrhoea versus theproducts actually sold during baseline surrogatepatient visits. It is clear that reported and observedpractices differ widely. The majority of pharmacystaff, 66% in Kenya and 53% in Indonesia, reportedthat they sold ORS to the last customer, but duringsurrogate patient visits, only 33% of confederates inKenya and 5 % in Indonesia were actually sold ORS.The reverse was true for antidiarrhoeals; in Kenya,reported use was 33 % while observed use was 48 %,and in Indonesia reported vs. observed use were 58%and 74%, respectively.

Motivations and constraintsFocus groups (pre-intervention) examined the reasonsfor the divergence between reported and observedpractices. Awareness of ORS among pharmacy staffwas high in both countries: however, many focusgroup participants expressed a feeling that ORS wasmerely 'good first aid and not the key to treatment.Respondents reported that both they and theircustomers wanted something 'stronger' than ORS tostop diarrhoea and treat its cause. Other reporteddeterminants of practice included the prescribingpractices of local doctors, product advertising, anddrug company sales representatives and information.In Kenya, pharmacy staff mentioned that they did nothave leaflets promoting ORS like other drugs. InIndonesia, respondents exhibited a strong attachmentto particular antidiarrhoeals, and mentioned personalexperience with these drugs as a factor in determin-ing their preferred treatment.

Pharmacy staff in both countries acknowledged theimportance of profit as a motive, preferring drugsthat yielded higher profit. However, the difference

at World H



Dow

nloaded from



Kanya Phaimadaa

OOw Draga

20% 40% 80%% OT CUStOfTMTS

n%

• brtarvtawa(n-70) OStmutatad VWta(r^l18)

PtwrnACwt

OttwrDniga

20% 40% 60%7* of customers

80%

| • lnt»rvWw»(n-H) DSbmdatad VTatta(n-19) |

* Product maritrtod primarty tor treating dtanhoea" EKhar akm* or as an tig red lent in «n antktanhoaal

Figure 2. Diarrhoea treatments reported in interviews with staffin private pharmacies compared to drugs sold during surrogatepatient visits

in relative profitability of ORS compared to specificalternative antidiarrhoeals is not great, especially inthe quantities these products are typically sold. InKenya, respondents also emphasized their desire tobe perceived as competent, 'scientific' practitioners,who could contribute to national health programmes.These motivations later became key themes of thetraining intervention (see Figure 1).

Impact of training on knowledgeKnowledge of counter assistants in Kenya about fluidloss and the role of ORS in dehydration was high bothbefore and after training (Table 1). However,knowledge about signs indicating bacterial etiologyand about the use of antibiotics was much lower.After training, pharmacy assistants recognized moresigns of bacterial diarrhoea (+0.18,95% confidenceinterval = [0.02,0.34]), and knowledge that anti-biotics are useful in only these cases increased(+0.20, 95% CI = [0.08,0.32]). Training was also

associated with an increase in the number of situa-tions (e.g. child has fever) recognized by trainees asreasons for referral (+0.44, 95% CI = [0.27,0.63]).

Impact of training on drug salesIn Kenya, sales of ORS differed at baseline amongGroup 1 intervention pharmacies in Nairobi (46% ofvisits), Group 2 intervention pharmacies in smallertowns (19%), and control pharmacies in Mombasa(26%) (Figure 3). Despite having the highest salesof ORS at baseline, use of ORS increased by 33%in Nairobi following the first wave of training, com-pared to a slight increase of 4% in Group 2 phar-macies and no change in Mombasa pharmacies(intervention vs all other pharmacies = +32%,95% CI = [15%,48%]). After the next round oftraining, use of ORS increased by 23% in Wave 2pharmacies, compared to a decrease of 3% in controls(+26%, 95% Q = [-1%,53%]). In Indonesia,where baseline use of ORS in the study groupswas equal at about 40%, the intervention group in-creased ORS sales by 34% after training, comparedto an increase of 13% among controls (+21 %, 95%CI = [3%, 39%]).

Figure 4 summarizes the impact of training on salesof antidiarrhoeals. As with ORS, sales of anti-diarrhoeals in Kenya differed somewhat at baselineamong Group 1 intervention pharmacies (44%),Group 2 intervention pharmacies (36%), and controlpharmacies (62%). Sales of antidiarrhoeals decreasedby 17% in Nairobi following the first wave oftraining, compared to an increase of 2% in bothGroup 2 and control pharmacies (Group 1 vs. allother pharmacies = - 1 9 % , 95% CI = [-36%,-3%]) . After the next round of training, sales ofantidiarrhoeals decreased by 5% in Group 2 phar-macies, while they increased by 2% among controls( - 7 % , 95% CI = [-27%,13%]). In Indonesia,where antidiarrhoeals were used by about two-thirdsof pharmacies in both groups at baseline, interven-tion pharmacies decreased antidiarrhoeal sales by29% after training, compared to a decrease of 9%among controls ( -20%, 95% CI = [-39%,-3%]) .

The use of antimicrobial products was also discour-aged except in cases of diarrhoea with possiblebacterial origin. However, overall rates of antibioticuse were already low at baseline in both countries(13% in Kenya and 2% in Indonesia). Modestdeclines in use following training in Kenya were notanalyzed statistically.

at World H



Dow

nloaded from



Table 1 . Impact of training on key knowledge indicators among pharmacy assistants in Kenya

Indicator Maximum score Pre (S.E.) Post (S.E.) Difference*(95% C.I.)

Total knowledge score

Fluid loss is the reason diarrhoea is dangerous

ORS replaces lost water and minerals

Signs of bacterial etiology (blood in stool,frequency, fever)

Antibiotics useful only to treat diarrhoea ofbacterial etiology

Antidiarrhoeals never useful and areunnecessary risk and expense

When to refer (child has fever, diarrhoeacontinues, bloody stool)

10

1

1

3

1

2

3

4.60(0.15)

0.80(0.04)

0.80(0.04)

1.54(0.08)

0.34(0.05)

0.58(0.04)

1.76(0.08)

5.49(0.17)

0.81(0.04)

0.86(0.03)

1.72(0.06)

0.54(0.05)

0.95(0.06)

2.21(0.10)

+0.88***(0.62,1.16)

+0.01(-0.09,0.11)

+0.06(-0.02,0.14)

+ 0 . 1 8 "(0.02,0.34)

+0.19***(0.08,0.32)

+0.37***(0.27,0.48)

0.44***(0.27,0.63)

Notes: * average difference of paired pre-post scores.•* pre-post increase significant, p<0.05.•*• pre-post increase significant, p<0.01 .

Kenya Pharmacies

% Surrogate Patient Visits Receiving ORS

100%

80% -

60%

Group 1 (n*M) Group 2 (n"24| Control (n°25)

I Pre • Pott Wive 1 D Port Wava 2 |

Kenya Pharmacies

% Surrogate Patient Visits Receiving Antldlarrhoeals

100%

Group 1 (n»6S) Group 2 (n>>24) Control (n-25)

• Pre 0 Post Wsv* 1 • Post Wav* 2

Indonesia Pharmacies

% Surrogate Patient Visits Receiving ORS

100%

Intervention (n-43) Control (n=44)

• PreDPost

Figure 3. Effect of persuasive training on sales of ORS to treatdiarrhoea as measured during surrogate patient visits to privatepharmacies

Indonesia Pharmacies

% Surrogate Patient Visits Receiving Antidlarrhoeals

100%

Intervention (n-43) Control (n-44)

I Pre D Post

Figure 4. Effect of persuasive training on sales of antidiarrhoealsto treat diarrhoea as measured during surrogate patient visits toprivate pharmacies

at World H



Dow

nloaded from



Communication

The number of questions about symptoms of diar-rhoea asked by counter attendants before making aproduct recommendation was low, averaging 1.2 atbaseline in Indonesia and 1.5 in Kenya. In all in-tervention groups in both countries, the number ofquestions increased following training compared toslight reductions observed in control pharmacies,although these increases are not statistically signifi-cant. The frequency of individual types of com-munication recorded in Indonesia was too low tojustify further analysis.

Three key symptoms to determine severity and typeof diarrhoea are blood in the stool, vomiting, andfever.34 In Kenya, questioning about both blood inthe stools and vomiting increased in both interven-tion groups following training compared to controls(Table 2). Questioning about fever showed littlechange in either group. None of these results werestatistically significant. Compared to control phar-macies, dehydration was discussed in 13% morefollow-up visits in both Group 1 (95%CI = [0.02,0.25]) and Group 2 pharmacies (95%CI = [0.04,0.21]).

DiscussionUsing the process outlined in the Guide, teams ledby the CDD Program Directors in Indonesia andKenya were able to assess baseline diarrhoea treat-ment practices in private pharmacies and their causes,and to design and implement a training interventionbased on a persuasive educational model that led tomeasurable improvements in treatment practicesamong pharmacy staff in both countries.

Baseline awareness of ORS and its function was highin both countries, although focus groups revealed thatthe staff felt that ORS was not as 'powerful' as anti-diarrhoeal agents. Knowledge about other products,about non-pharmacologic treatments, and about thedanger signs which would indicate the need formedical treatment was limited. Consistent withstudies in industrialized countries,35 self-reportedpractices (e.g. use of ORS) were substantially over-stated in comparison to observed behaviours.

Following training, product sales in both countriesinvariably changed in the direction of key recom-mended practices. Overall, sales of ORS nearlydoubled in comparison to baseline values. In Kenya,ORS sales increased significantly in Nairobi com-pared to control pharmacies, while increases just

failed to achieve statistical significance in othertowns, due to smaller sample size. In Indonesia,despite the fact that ORS sales in control pharmaciesalso increased by a third, intervention pharmaciesnevertheless experienced significant increases in ORSuse. In contrast, pre-intervention sales of antidiar-rhoeal products were reduced by about a third in phar-macies which received training. In Kenya, mirroringthe findings for ORS, declines in antidiarrhoeal saleswere significant in Nairobi but not in other towns;in Indonesia, antidiarrhoeal sales declined somewhatin the control group, but sales in intervention phar-macies were nevertheless significantly lower.

Were the observed changes due simply to increasedknowledge about diarrhoea and drugs? In the caseof antidiarrhoeals where knowledge was poor, the in-tervention both increased knowledge and decreasedproduct sales. In the case of ORS, baseline knowledgewas already quite high. Nonetheless, ORS sales in-creased dramatically, indicating that such a persuasiveintervention can also enable behaviour change byaddressing barriers to change.

While we observed a trend toward increased com-munication, particularly in questioning customers forsigns of bacterial diarrhoea and dehydration in Kenya,communication between counter attendants andcustomers remained poor in both countries.

Several threats to the validity of this controlled studydeserve comment. In Kenya, selection of study andcontrol groups was not random, but by geographicarea. However, results of the intervention were con-sistent following both waves of training, thusreplicating the findings in different environments. InIndonesia, where randomly allocated control and in-tervention pharmacies were mixed in the samegeographic areas, changes in product sales indicatepossible contamination of the control group; however,this would have reduced the size of observed effectsreported above.

Surrogate patient visits did not measure behaviour inresponse to customer requests for specific products.However, studies of similar interventions in in-dustrialized countries have succeeded in equippinghealth providers with skills to overcome suchpressures,28 and strategies to overcome patient de-mand were also included in training in this study. Itis possible that some pharmacy staff were aware ofthe role of the surrogate patients and changed theirbehaviour accordingly. However, different assessors

at World H



Dow

nloaded from



Table 2. Impact of training on communication during sales encounters in Kenya pharmacies

IndicatorTraining Group 1

Beforetraining

Change aftertraining (S.E.)

Training Group 2*

Beforetraining


Group 3 Controls*

Beforetraining


Average H history 1.69 +0.54 1.34 +0.19 0.92 -0.42questions (0.30) (0.36) (0.23)

Ask about blood 10% +14%** 20% +3% 9% - 4 %in stool (0.05) (0.08) (0.04)

Ask about presence 32% - 3 % 17% +4% 8% - 2 %of fever (0.06) (0.07) (0.06)

Ask about presence 29% +9% 12% +16% 14% - 2 %of vomiting (0.06) (0.08) (0.05)

Discuss signs of 9% +13%** 0% +8%** 4% - 5 %dehydration (0.05) (0.04) (0.06)

Notes: * Scores before training are values during Wave 2 of the simulated patient visits.** Pre-post difference significant, p<0.05.

visited each pharmacy in the baseline and follow-upassessments to minimize this possibility.

Not informing pharmacy staff that their practiceswere being evaluated also raises an ethical issue.However, taking consent from each pharmacy wouldhave substantially increased the likelihood of bias.Surrogate patient visits have been used frequently indrug use research, but the circumstances in whichthey can be used ethically need to be clarified.

ConclusionsThis pilot study of the WHO Guide represents thefirst controlled trial documenting the impact of per-suasive training implemented by national CDD pro-grammes in changing private sector retail practices.Although the results support the short-term efficacyof this approach, several questions remain aboutthe 'real-world' effectiveness of such interventionmodels. In particular, more data are needed on thesustainability of the positive effects on drug sales.Given that surrogate patient visits proved relativelyquick and inexpensive to implement, it is crucial thatfuture interventions use this technique to measureboth short and longer-term impacts. If the short-termimprovements observed in this study do not persist,strategies will need to be developed to reinforce andsustain them.

If changes in behaviour appear to be long-lived, wealso need more information about the resources and

incentives necessary to continue such an activity overtime and on a larger scale. One promising strategyto reinforce improvements and to extend die poten-tial scope of these activities might be to involve ORSmanufacturers and distributors as active partners indesigning and implementing this intervention. Fin-ally, future research should examine whether meseresults are generalizable to other country settings andto other important health problems commonly treatedin pharmacies (e.g. acute respiratory infections ormalaria) for which improvement in sales practicescould substantially lessen me burden of illness ininfants and children.

References1 Grant J. State of the World's Children 1993. New York:

Oxford University Press for UN1CEF; 1993.2 World Bank. World Development Report 1993; Investing in

Health. New York: Oxford University Press; 1993.3 Igun UA. Why we seek treatment here: retail pharmacy and

clinical practice in Maiduguri, Nigeria. Soc. Sci. Med. 1987;24:689-95.

4 Kloos H, Chama T, Abemo D, Tsadik KG, Belay S. Utiliza-tion of pharmacies and pharmaceutical drugs in Addis Ababa,Ethiopia. Soc. Sci. Med. 1986; 22(6): 653-72.

3 Ferguson AE. Commercial pharmaceutical medicine andmedical ization: a case study from El Salvador. Cult. Med.Psychiatry 1981; 5(2): 105-34.

6 Mitchell MF. Improving pharmacy practice in the ThirdWorld: The Jamaican experience. Mobius 1985; 5: 129-34.

7 Berih AA, Mclntyre L, Lynk AD. Pharmacy dispensing prac-tices for Sudanese children with diarrhoea. Public Health1989; 103: 455-8.

at World H



Dow

nloaded from



8 Calva JJ, Ceron E, Bojalil R, Holbrook A. Antibiotic con-sumption in a community of Mexico City. n . Survey ofpurchases at pharmacies. Bol. Med Hasp. Infant. Mex. 1993;SO: 145-50.

9 Schorling JB, de Souza MA, Guerrant RL. Patterns of anti-biotic use among children in an urban Brazilian slum. Int.J. Epidemiol. 1991; 20: 293-9 .

10 Unda R. The abuse of antidiarroeal agents and antimicrobialdrugs in the treatment of infant diarrhoea in Mexico.PRTTECH Assign No. DC229. 1986.

1 ' Blanco RA. Concepts of pharmacy employees concerning themanagement of diarrhoea in children. Bol. Med Hasp. Infant.Mex. 1989; 46: 463-9 .

12 Wolffers I. Drug information and sale practices in some phar-macies of Colombo, Sri Lanka. Soc. Sci. Med. 1987; 25 :319-21 .

13 van der Geest S. The illegal distribution of western medicinesin developing countries: Pharmacists, drug peddlers, injec-tion doctors and others. A bibliographic exploration. Med.Anthropol. 1982; Fall: 197-219.

14 Weisberg DH. "Drugs-retail and wholesale": The druggist aspractitioner and medical middleman in a northern Thai town.Paper presented at the Annual Meeting of the AmericanAnthropological Association. Washington, D.C. , December,1982.

15 Amidi S, Ajamee G, Sadeghi HR, Yourshalmi P, GharebjehM. Dispensing drugs without prescription and treating patientsby pharmacy attendants in Shiraz, Iran. Am. J. Public Health1978; 68 : 495-7 .

16 Kafle KK, Gartoulla RP, Pradhan YMS, Shrestha AD, KarkeeSB, Quick JD. Drug retailer training: Experiences from Nepal.Soc. Sci. Med. 1992; 35(8): 1015-25.

17 Bailey PE, Janowitz B, Solis M, Machuca M, Suazo M. Con-sumers of oral contraceptives in a social marketing programin Honduras. Stud. Fam. Plann. 1989; 20 : 5 3 - 6 1 .

18 Janowitz B, Suazo M, Fried DB, Bratt JH, Bailey PE. Impactof social marketing on contraceptive prevalence and cost inHonduras. Stud Fam. Plann. 1992; 2 3 : 110-7.

19 Rizo A. Drugstore employees and family planning in LatinAmerica. Pathpapers 1979; 5: 1-18.

2 0 Bates J. Interviews with staff of Bangladesh Contraceptive andORS Social Marketing Project. Boston: Management Sciencesfor Health; 1990.

21 Ross-Degnan D. Interviews with staff of the Ghana SocialMarketing Programme. Boston: Management Sciences forHealth; November, 1989.

2 2 Kenya PR, Gatiti S, Muthami LN et al. Oral rehydrationtherapy and social marketing in rural Kenya. Soc. Sci. Med.1990; 31(9): 979-87.

2 3 Podhipak A, Varavkhya W, Punyasatabandhu P, VathanophasK, Sangchai R. Impact of an educational piogiaui on the treat-ment practices of diarrhoeal diseases among pharmacists anddrug sellers. Southeast Asian J. Trop. Med Public Health1993; 24(1): 3 2 - 3 9 .

24 Goel P , Ross-Degnan D , Soumerai S, Reich M , Herman P .Childhood diarrhoea treatment at retail pharmacies: Keydeterminants and persuasion as a policy intervention,(submitted)

23 World Health Organization Diarrhoeal Disease Control Prog-ramme. Guide for Improving Diarrhoea Treatment Practicesof Pharmacists and Licensed Drug Sellers. World HealthOrganization; Geneva: 1993.

26 Tcanson GB, Sterky G. Self-prescribing by way of pharmaciesin three Asian developing countries. Lancet 1986; 6 2 0 - 2 .

2 7 World Health Organization Diarrhoeal Disease Control Pro-gramme. The Rational Use of Drugs in the Management ofAcute Diarrhoea in Children. Geneva: World HealthOrganization; 1990.

2 8 Soumerai SB, Avorn J. Principles of educational outreach('academic detailing') to improve clinical decision makingJAMA 1990; 263(4): 549-56.

2 9 Avom J, Soumerai SB. Improving drug-therapy decisionsthrough educational outreach. A randomized controlled trialof academically based "detai l ing". NewEngL J. Med. 1983;308: 1457-63.

3 0 Soumerai SB, Avorn J. Economic and policy analysis ofuniversity-based drug "deta i l ing" . Med. Care 1986; 24:313-31 .

31 Soumerai SB, Salem-Schatz S, Avom J, Casteris C, Ross-Degnan D, Popovsky MA. A controlled trial of educationaloutreach to improve blood transfusion practice. JAMA 1993;270(8): 961-6 .

32 Soumerai SB, Mclaugh l in TJ, Avorn J. Improving drugprescribing in primary care: A critical analysis of theexperimental literature. Milbank Quarterly 1989; 67(2):268-317 .

33 Cook T D , Campbell DT. Quasi-experimentation: design andanalysis issues for field settings. Boston: Houghton Mifflin,1979.

34 Ministry of Health, Republic of Kenya. The Policy Statementfor die Kenya Control of Diarrhoeal Diseases Program.Nairobi: 1993.

33 Lomas J, Anderson G M , Domnick-Pierre K, Vayda E , EnkinM W , Hannah WJ. Do practice guidelines guide practice? Theeffect of a consensus statement on the practice of physicians.NewEngL J. Med. 1989; 321: 1306-1311.

AcknowledgementsWe wish to acknowledge the invaluable cooperation and offer ourthanks to: Patricia Whitesell and Susan Toal of ACT Internationalfor assisting in the design and editing of the Guide;USAID/PRITECH for providing examples of educational materialsthat were incorporated as prototypes in the Guide; James Trostleof Harvard Institute for International Development for observingand commenting on die use of qualitative methods; Jennifer Jones,Jeanne Madden, and Laura Goldberg for assistance in preparingmaterials for the Guide; David Alnwick and Sandra Childress ofUNICEF/Kenya for co-funding the Kenya intervention; GkauMwangi and the Pharmaceutical Society of Kenya, faculty of theUniversity of Nairobi Faculty of Pharmacy (James Ombega, J.Chege, Anastasia Guantai, Minal Gudka) and the Medical Train-ing Centre, Nairobi (Edward Odhiambo, Sarah Chuchu) forassisting in the implementation of the Kenya intervention; VivianaMangiaterra of WHO/Indonesia for assisting in die implementa-tion of the Indonesia intervention; Ann Pay son for assistance inpreparing this article; and to all the trainers and pharmacy staffwho participated in the interventions.

This work was supported by a contract from the World HealthOrganization Programme for the Control of Diarrhoeal andRespiratory Diseases to Management Sciences for Health,Washington DC, and Harvard Medical School, Boston.

BiographiesDr Dennis Ross-Degnan, the Principal Investigator for this study,is an Assistant Professor in die Department of Ambulatory Care

at World H



Dow

nloaded from



and Prevention of Harvard Medical School and Harvard PilgrimHealth Care, and the Coordinator for the Harvard Support Groupof the International Network for Rational Use of Drugs.

Dr Stephen B Soumerai is Associate Professor of Ambulatory Careand Prevention and Director, Drug Policy Research Group,Harvard Medical School and Harvard Pilgrim Health Care.

Dr Pradeep K Goel is a Research Associate at Harvard MedicalSchool and Brigham and Women's Hospital, and directed the im-plementation of the WHO Drug Sellers Training Project in Kenya.

James Bates, currently the Director of the USAJD-funded RationalPharmaceutical Management Project for Management Sciences forHealth, Washington, DC, directed the WHO project under whichthis work was completed.

Dr Joseph Makhulo is Manager of the Control of DiarrhoealDiseases Programme, of the Division of Family Health, KenyaMinistry of Health.

Nicholas Dondi, formerly a staff member of the Control ofDiarrhoeal Diseases Programme of the Kenya Ministry of Health,now works as a private health consultant.

Dr Sutoto is the Director of the Infectious Diseases Hospital inJakarta, Indonesia. At the time of this work he was Chief of theSubdirectorate Control of Diarrhoeal Diseases, Directorate Generalof Communicable Diseases Control and Environmental Health,Ministry of Health, Indonesia.

Daryono Adi is Chief of the Evaluation Section for the Subdirec-torate for Control of Diarroheal Diseases of the Ministry of Health,Republic of Indonesia.

Lucia Ferraz-Tabor is the Resident Advisor for the Program forAdvancement of Commercial Technology, Child and Reproduc-tive Health project (PACT-CRH), USAID India and IndustrialCredit and Investment Corporation of India (ICICI).

Robert Hogan is a Programme Management Officer at the Con-trol of Diarrhoeal and Respiratory Diseases Programme at theWorld Health Organization in Geneva.

Correspondence: Dr Dennis Ross-Degnan, Dept of AmbulatoryCare & Prevention, Harvard Medical School, 126 BrooklineAvenue, Suite 203, Boston, MA 02215, USA.

at World H



Dow

nloaded from


The impact of face-to-face educational outreach on diarrhoea treatment...

Documents

Transcript of The impact of face-to-face educational outreach on diarrhoea treatment...