IMPROVING THE USE OF ARTEMISININ-BASED

COMBINATION THERAPY IN RURAL ZAMBIA

Third International Conference for Improving Use of MedicinesAntalya, Turkey

November 15, 2011

Kojo Yeboah-AntwiCentre for Global Health and Development

Boston University

Background • Many sick children in rural Zambia are seen by

community health workers (CHW) because public health facility-based services are not readily accessible

• Zambia has changed first line drug for uncomplicated malaria to more expensive and effective artemisinin-based combination therapies (ACTs)

• Relatively little known about how to optimally deploy ACTs to the community level

• Concerns about potential overuse of ACTs and development of resistance if use by CHWs is not guided by rapid diagnostic tests (RDTs)

Overall Study Goal

Demonstrate the effectiveness and feasibility of using CHWs to manage malaria (with artemether-lumefantrine) guided by RDTs

Objectives Will RDT use lead to reduction of overuse of

ACTs?

How well will CHWs follow a treatment algorithm and adhere to results of the RDTs?

Will children and CHWs experience any side effects with the pricking?

What happens to children with RDT negative results who do not receive ACTs?

Will the community accept CHWs performing RDTs and prescribing ACTs?

Study Location Southern Province of

Zambia Mazabuka and

Siavonga districts Chikankata Mission

Hospital area Population: 70,000 1 Mission Hospital

and 5 Rural Health Centers

Study Sites

Chaanga RHC

Mwanamunzya CHP

Hamukombwe CHP

Study Design Cluster randomized, controlled trial

around community health posts manned by CHWs

Community health posts matched according to distance from rural health center

Intervention Both intervention and control CHWs trained in

classification and treatment of febrile illness

Both CHWs supplied with ACTs and antipyretics

Intervention CHWs received additional training on RDTs and infection control

Intervention CHWs received RDTs and supplies for waste disposal and infection control

Enrolment Children aged 6 mo to 5 yrs with fever were

enrolled

Children with severe illness excluded and referred

History, examination, RDT (in intervention arm), classification and treatment

Baseline form completed (findings, results, treatment, address)

Follow-up Patients seen at day 5-7 to collect information on

visit, outcome of treatment, additional care received

Determine current condition and advise as necessary

Monthly data from CHPs on patient seen, supplies and referrals

Post intervention FGDs and IDIs of caregivers, CHWs, health workers and community leaders

Baseline Characteristics of CHWs

Intervention (n=18) Control (n=19)

Male 83.3% 89.5%Age in years: Mean (range)

40.3 (26-53) 40.0 (27-55)

Education: Secondary 72.2% 64.4%Considered as full time 5.6% 26.3%Years of practice: Mean (range)

10.2 (1-26) 7.3 (1-22)

Trained by Chikankata 55.6% 63.2%Last Refresher Course: less than a year

55.6% 52.6%

Supervision By RHC in last 3 months

44.4% 42.1%

Distance of CHP from RHC: Mean (range) km

9.2 (1-15) 9.3 (3-15)

Baseline Participant CharacteristicsCharacteristics Intervention

(n=1017)Control

(n=2108)Sex (female) (%) 47.6% 48.8%Age (mean) (SD) months 22.6 (14.0) 23.6 (14.7)Children underweight (WAZ score <-2.00)

28.1% 30.3%

Mother’s Education:No formal education 45.4% 37.7%

Primary 45.5% 54.2%Mother’s occupation:

Farmer 58.1% 48.9%Housewife 36.5% 46.4%

Households with 6 or fewer persons

64.2% 62.6%

Immunizations up to date 59.5% 67.5%Slept under ITN last night 71.3% 69.5%

Results

Intervention Control RR (95% CI)

Correct Classification1

99.7% 98.9% 1.01 (1.00 – 1.01)

Appropriate treatment2

98.2% 99.3% 0.99 (0.97 – 1.00)

Febrile children receiving ACTs

27.5% 99.1% 0.23 (0.14 – 0.38)

1Intervention: classify as malaria if RDT (+), and not malaria if RDT (-)Control: classify as malaria if fever (+), and no malaria if fever (-)

2Prescribe ACT if classified as malaria; ACT not prescribed if classified as not malaria

Results 975 RDTs done of which 271 (27.8%) were positive and 704

were negative

3 of 704 RDT negatives received ACTs from CHWs

4 of 271 RDT positives did not receive ACT from the CHWs

5 of the RDT negatives who did not receive ACTs from CHWs were not satisfied and managed to get ACTs from other sources

91.2% of the RDT negatives got well with only antipyretics and needed no additional treatment

Adverse Effects with RDT

Of 975 RDTs done: 3 children with minor bruises

2 children with skin infection

14 children with minor bleeding

1 incident of self prick

Community Acceptance

Caregivers were comfortable with CHWs pricking children to test for malaria

Caregivers trusted results of the RDTs

CHWs felt confident in performing RDTs and treating with ACTs

Health workers endorse CHWs use of RDTs and ACTs Community leaders happy with intervention and want

expansion to other areas and adults

Conclusions CHWs are capable of performing RDTs and appropriately

dispensing ACTs

Use of RDTs and ACTs by CHW is safe and effective

Use of RDTs at the community level by CHWs has potential to reduce the overuse of ACTs and improve ACT use

Health workers see CHWs involvement in malaria treatment with RDTs and ACTs as positive and has the potential to reduce workload at health facilities

Community acceptance of malaria treatment by CHWs is overwhelmingly encouraging

Limitations More CHP attendance in Control arm

resulted in relative imbalance between the two study arms

“full time” CHWs Varied cluster size No evidence of “contamination”

Recall bias

Designed not to repeat parasitological test at follow up

Study Team Kojo Yeboah-Antwi Portipher Pilingana Kazungu Siazelee William B. MacLeod Katherine Semrau Penelope Kalesha Busiku Hamainza Donald Thea

Davidson H. Hamer Lora L. Sabin Karen Kamholz Euphasia Mtonga Pascalina Chanda Arthur Mazimba Phil Seidenberg

Funding USAID/Washington through CFAR

Cooperative Agreement GHSA-00-00020 with BU

President’s Malaria Initiative (Washington DC)