EXTENT OF HUMAN AND ENVIRONMENTAL ANTIBIOTIC USE IN

EXTENT OF HUMAN AND ENVIRONMENTAL ANTIBIOTIC USE IN RURAL BANGLADESH Results of qualitative research Roess AA 1, Winch PJ 1, Afsana A 2, Afroz D 2, Ali NA 2, Shah R 2, Seraji H 2, Baqui AH 1, Darmstadt GL 1, Arifeen SE 2 Second International Conference on Improving Use of Medicines, Chiang Mai, Thailand March 30 -April 2, 2004 Funding provided by the Center for a Livable Future, JHSPH, USA Doctoral fellowship for AA Roess provided by National Institute of Child Health and Development, NIH, USA

Abstract Problem Statement: Studies from the United States and the European Union have demonstrated a direct link between livestock antibiotic usage and human infection caused by antibiotic-resistant pathogens. No information is available from resource-poor settings on the extent of environmental antibiotic usage or the prevalence of antibiotic-resistant bacteria in animals. Objective: To assess the extent of environmental and human antibiotic usage in rural Bangladesh, surrounding practices and beliefs, and the impact on antimicrobial-resistant bacteria in human populations. Setting and Study Population: Between July 2002 and May 2003, interviews were conducted with 20 village doctors, local “pharmacists, ” homeopaths, clinic workers, and local animal care providers to learn about their antibiotic prescription practices. Interviews were also conducted with 40 randomly selected households in a remote region of Bangladesh to learn about human and animal antibiotic usage practices and surrounding beliefs at the household level. A survey instrument was developed to quantify these results in 700 households and is currently in use. Outcome Measures: Average number of drugs/household, number of drugs/child case, number of drugs/childhood disease type, percentage of drugs prescribed/provider type, average cost of drugs/household, average cost of drugs/child, number of animal antibiotics/household, cost of animal antibiotics/household. Results: Current human antibiotic use, defined as use within the last 1 month, was reported in every household sampled. Children were the recipients of about 50% of antibiotics in use. Animal antibiotic use was found in 60% of sampled households. Human and animal antibiotic use at the household level was at suboptimal doses. Veterinarians are increasingly concerned about the unregulated use of animal antibiotics in rural areas, and several described treating “resistant” cases with second-line antibiotics. An interesting parallel emerged between care-seeking behavior, antibiotic use, and prescriber practices for both humans and animals. Conclusions: The extent of antibiotic use in rural Bangladesh is much greater than previously thought. Most surprising was the widespread use of animal antibiotics and the anecdotal reports of “resistant” animal infections. We are currently collecting specimens from people and animals to determine the prevalence of antibiotic-resistant bacteria in the populations and the relationship between environmental antibiotic use and the emergence of antibiotic-resistant human pathogens. We are working with the Bangladesh Ministry of Fisheries and Livestock to assess the problem and make recommendations for future surveillance.

Increasing prevalence of antimicrobial resistant (AMR) microbes in S Asia Community-acquired infections – – – – Multidrug resistant pneumococci Drug-resistant H. influenzae FQ- and ESC-resistant Salmonella Multidrug resistant Shigella FQ-resistant gonococci Multidrug-resistant M. tuberculosis Drug-resistant malaria CHRP 2002

Risk factors for AMR Antibiotic use Recent hospital stay/visit Crowding Travel Animal antibiotic use (food animal & pet)

Study Conceptual framework Drug Use Policy Human ab use Provider practices SES Human Carriage AMR HH Crowding Town Clinic Animal ab use Animal Carriage AMR Human disease AMR

AMR conceptual framework with pic 23 Feb 2004. doc

Objectives To understand antibiotic use practices for people and animals at the household level in Sylhet To design an instrument to quantify antibiotic use

Methods Field Observations In-depth interviews – – Ministers of Fisheries and Livestock Government veterinarians and scientists Pharmaceutical representatives Village doctors, drug-sellers Semi-structured interviews – 14 human use – 24 animal use – 10 village doctors Village doctor network

Crowding as a risk factor Crowding – Especially in the capital and in the town centers throughout Bangladesh is the 8 th most populated countries with a population of about 140 million—for some perspective that’s equivalent to about half of the US’s population in Wisconsin.

Exposure to animals -- Dhaka

Town to village delivery Pharmaceu tical presence: Such carts from pharmaceu ticals filled with human and animal antibiotics and medicines make deliveries to bazaars. Smaller carts driven by rickshaws or pulled by delivery boys delivered to more remote areas.

Flooding/Infrastructure This aerial shot was taken 3 months before the Monsoons. Statistics on flooding included that more than 50% of the country is underwater more than 50% of the year. Untreated human waste, animal waste, and pollutants enter this water. People are exposed to this water either by contact or through using the water for household work of for drinking.

Human medicines in stock at home Antibiotic Paracetamol s Homeopath (Vitamins) <1 year old 7 1 3 1 -10 years 8 5 2 Adults Total packages 9 7 3 14/14 61 9/14 5(8)/14

Care-providers/drug sellers roles “Village doctors” Treat Sell human Treat Sell animal infants med animals med 43 39 27 8 11 37 women were first asked who the 3 “most popular village doctors” were. Each woman was then asked to name an animal doctor if she hadn’t already. 43 different “doctors” were identified. 2 of the village doctors originally identified were thought to only treat animals.

Village doctor network VD multiple roles Treatment patterns Examination/Symptoms Dosage

Animal ownership Animal Cows Any 1 -5 >5 17 13 4 5 12 5 8 4 Swans/pigeons 24 4 19 4 Elephant 1 1 Goats Ducks Chickens 5 24 households that owned at least 1 chicken were included to learn about animal medicine use and animal husbandry.

Expenditure for animal purchase Animal Cows Goats Ducks Chickens Swans/pigeons Elephant 0 30 -200 Tk US$1 -4 2 1 201 -500 US$4 -10 501 -9 K 10 -200 10 -25 K 200 -500 15 1 2 8 8 10 2 3 2 4 2 2 1 Animals are a large investment because they are sources of food (meat milk and eggs), and are a good source of income.

Animal use Eggs % 100 Meat 83 Milk 75 Agricultur e Raise and sell 54 54

Animal medicine use Heard of Gave animal antibiotics/ med? % 29 Yes 92 Yes Animal ever ill? 100 Yes Care seeking? 96 Yes

Animal diseases Cow—diarrhea, back pain, blisters Goat—diarrhea, stomach swelling Duck/Chicken—white diarrhea, fever from river, “doze” Evil eye, bad spirits

Animal “medicine” Oxytetracycline Metronidazole Growth promoter (2/24) Paracetomol Homeopathic Injections (5/24)

Animal medicine cost 20 -60 tk (US$1) Total 1 61 -120 (US$2) 3 121 -181 (US$3) 181 -220 (US$4) 2 220 -2 k (US$12) 12 (50%) Human medicine cost 10 -60 tk (US$1) Total 61 -120 (US$2) 121 -181 (US$3) 181 -220 (US$4) >220 (US$5) 1 4 4 2 3 (22%) • About 50% of households spent on average 12 USD for animal medicine while more than 50% spent less than 3 USD on human medicine. • Since animals are such a big investment this is understandable and expected by many of those in the animal medicine business in Bangladesh.

Animal care expenditure Animals food + income Buy animals – 5, 000 -18, 000 Tk per cow – 0, 100 -120 Tk per chicken Treatment – 200 Tk injection – 16 -60 Tk/tablet (oxytetracycline)

Human care expenditure 10 Tk-320 Tk, m=40 Tk Vitamin Homeopathic Paracetomol Antibiotics

Animal ab use

Drug sellers A drug seller gave us some insight into the human and animal medicine business. He described starting out as a human pharmacist and then making a gradual transition to human and animal medicine. Eventually he switched to the exclusive sale of animal medicines because it is much more profitable and associated with less risk. His story was repeated several times in our other interviews.

Animal husbandry Animal Shed “under bed” House Cows Goats Ducks 15 3 4 1 2 6 2 Chickens 8 13 3 Swans/pigeons 4 Elephant 1 In rural areas proximity to livestock and household members is greater and so is the potential risk. For example we found that while the majority of cows had their own shed or special house chickens and goats not only lived in the house but lived in special baskets “under the bed”.

Animal care-taker Animal Male Cows Goats Chicken 13 4 3 2 24 0 Female household members take care of both chickens and children, thus they introduce to their children another source of antibiotic resistant pathogens.

Animals, Fish, Food, Water Animals can transmit resistant bacteria and resistance agents through contamination of water. This type of setting is common throughout the world, fish are caught in this water that is contaminated with both human and animal waste, fish are consumed, water is used for many purposes, and resistance spreads.

Future data collection Currently we are using a survey instrument to quantify human and animal level antibiotic use, and related household and SES variables. Data will be obtained on the prevalence of human and animal carriage of AMR bacteria and the relationship of the two. This information will be shared with the Government of Bangladesh and other stakeholders with the goal of informing programs to reduce the spread of resistant microbes.

Survey design AMR risk – Animal exposure (crowding, waste, ab use) – Current antibiotic use (look at available medicine) – Human waste removal – Human crowding Antibiotic use – Human

Acknowledgements Many thanks go to the Centre for a Livable Future at Johns Hopkins University School of Public Health and especially to Robert Lawrence and Polly Walker for their support and encouragement. JHSPH Ellen K. Silbergeld Timothy Baker Diane Lynn Francisco Mathuram Santosham Paul A. Law ICDDR, B, DSH Tamanna Sharmin Ashraful Alam Neeloy Samir K. Saha