In the summer of 2020, researchers at the London School of Hygiene & Tropical Medicine and the Medical Detection Dogs charity trained six dogs to detect COVID-19 from odour samples obtained from infected patients. The results, published in peer-reviewed trials, showed the dogs achieving sensitivity rates between 82 and 98 percent and specificity above 90 percent — performance figures that compared favourably with the rapid antigen tests being rushed into deployment worldwide. One dog, working in a hospital setting, screened 153 patients for SARS-CoV-2 with 96 percent diagnostic sensitivity and 100 percent specificity. The researchers noted that a single trained dog could screen up to 250 people per hour.
The medical establishment largely moved on. Vaccines arrived, case counts eventually fell, and the urgency of finding a non-laboratory screening method receded. But five years later, as the global health community builds pandemic preparedness frameworks designed to prevent a repeat of COVID-19's catastrophic spread, the science of medical detection dogs has not only continued — it has strengthened, deepened, and raised a question that matters directly to countries like Bangladesh: why haven't we built this into our health infrastructure?
What the Science Actually Shows
The research literature on canine COVID detection has grown substantially since the initial 2020 studies. A German trial published in BMC Infectious Diseases trained eight detection dogs over just one week to distinguish between SARS-CoV-2 positive and negative saliva samples in 1,012 randomised trials, achieving an overall detection rate of 94 percent with diagnostic sensitivity of 82.63 percent and specificity of 96.35 percent. A subsequent study by US researchers at Assistance Dogs of Hawaii, published in the journal Open Forum Infectious Diseases, reported even stronger results from sweat samples — 98 percent sensitivity and 92 percent specificity in a testing phase, and 96 percent sensitivity with 100 percent specificity when one dog screened 153 hospital patients in a real-world setting.
The JAVMA study published in 2024, assessing dogs trained to detect COVID-19 in live individuals rather than inert samples, found 94 to 96 percent agreement with PCR testing — "a faster and more efficient screening method" in the journal's assessment. Crucially, the study documented that dogs were able to detect the scent of infection two to three days before individuals either showed symptoms or tested positive on conventional tests. For pandemic management, where pre-symptomatic transmission has been identified as a primary driver of spread, that detection window represents a potentially transformative public health tool.
The economic case has been made as well. A large-scale study in Frontiers in Medicine that examined 5,253 samples estimated the daily average cost of scent detection dog screening at $79 USD — roughly equivalent to the cost of 24 rapid antigen tests. Beyond the threshold of approximately 24 samples screened per day, canine detection becomes cost-saving compared to antigen testing. In mass screening scenarios — airports, transit hubs, stadiums — the cost differential becomes substantial. The researchers noted that the cost of mass testing with dogs is relatively constant regardless of the number of people screened, while antigen test costs scale linearly with volume.
A 2025 study published in Frontiers in Public Health, surveying public perception at concert venues in Germany, found that 70 percent of respondents preferred canine testing before events — rising to 72 percent after experiencing it. The non-invasive character of the screening, requiring no nasopharyngeal swab or physical contact with the dog, drove much of the preference. Public acceptance, a factor that complicated the rollout of many COVID-era testing protocols, appears to be a comparative advantage of canine screening rather than a barrier.
The Bangladesh Diagnostic Gap
Bangladesh's experience with COVID-19 exposed the structural vulnerabilities of a health system that, despite notable achievements in primary care delivery and immunisation coverage, was not equipped to absorb a rapid-onset diagnostic emergency. At the pandemic's peak, the country's PCR testing capacity — concentrated in Dhaka's major hospitals and a handful of divisional centres — was overwhelmed by demand from a population of approximately 170 million. Testing turnaround times stretched from hours to days. In rural upazila health complexes, diagnostic options were effectively non-existent beyond clinical judgement.
The inequality in diagnostic access played out in mortality and morbidity patterns that public health analysts have noted were likely underreported. Bangladesh's COVID-19 official case count and death toll have been questioned by researchers, with the structural limitations of the testing infrastructure meaning that asymptomatic and mildly symptomatic cases — the majority of infections — were systematically undercounted outside major urban centres. The country's 2024 Joint External Evaluation of International Health Regulations core capacities, completed in July 2024, acknowledged the need for continued laboratory capacity strengthening and resource mapping to fill technical and financial gaps in the national health security system.
Bangladesh is not starting from zero. The CDC has maintained an office in Dhaka since 2003 and works with the Institute of Epidemiology Disease Control and Research (IEDCR) and icddr,b on nationwide respiratory virus surveillance. The Pandemic Fund allocated $15.5 million to eight South-East Asian countries including Bangladesh to enhance collaborative surveillance and diagnostic readiness. The WHO has invested in a National Influenza Center, now operational since 2007, that tracks COVID-19 and other respiratory pathogens. These institutional foundations exist. The question is what innovative, cost-effective, and rapidly scalable tools could complement them.
Why Detection Dogs Deserve Serious Consideration
The case for investing in canine detection programmes in Bangladesh rests on the specific characteristics of the country's health infrastructure — its strengths, its gaps, and its geography.
Bangladesh has a population density of approximately 1,100 people per square kilometre, the highest of any major non-city-state country in the world. That density means infectious diseases spread rapidly, and that mass screening at chokepoints — ports of entry, railway stations, densely populated markets, garment factory complexes, and Rohingya refugee settlements in Cox's Bazar — could have disproportionate public health impact if conducted efficiently. A single detection dog screening 250 people per hour at Hazrat Shahjalal International Airport, or at the Benapole land port — Bangladesh's busiest border crossing — represents coverage that cannot be replicated by PCR testing at equivalent cost.
Bangladesh already has a deep institutional relationship with working dogs. The Bangladesh Police maintains a K-9 unit with trained dogs used for explosives detection and crime scene investigation. The Army and Border Guard Bangladesh deploy dogs in operational roles. The training infrastructure for canine handlers, while not oriented toward medical detection, represents a foundation that could be adapted. Countries that have piloted COVID detection dogs at airports and public venues — Finland, Chile, Colombia, the UAE, and several others — have generally worked through existing security dog training programmes rather than building medical detection capacity from scratch.
The One Health dimension is significant for Bangladesh specifically. The country sits at the intersection of zoonotic disease pathways that make pandemic risk permanently elevated: Nipah virus outbreaks have occurred periodically since the 1990s; highly pathogenic avian influenza circulates in the country's extensive live bird market network; and the Rohingya refugee crisis in Cox's Bazar — 900,000 people in the world's largest refugee camp — creates a disease surveillance challenge that conventional diagnostic infrastructure has never fully solved. Medical detection dogs trained on SARS-CoV-2 have shown cross-detection potential for other respiratory pathogens; the technology is not single-use, but a platform that could be extended across multiple disease threats.
The Limitations Are Real
The scientific literature is consistent in one important caveat: canine detection is a screening tool, not a confirmatory diagnostic. A dog indicating a positive result requires PCR or antigen test confirmation. False positives generate anxiety and logistical burden. False negatives — while relatively rare given the high sensitivity figures achieved in controlled trials — carry consequence in high-stakes settings. The 2023 Frontiers in Medicine research topic review noted that studies using only negative controls as discriminators produced higher reported sensitivity figures than those comparing against positive samples from other respiratory infections; the distinction matters for accuracy estimates deployed in public health planning.
Dogs are also variable across individuals. A 2023 research synthesis noted that detection sensitivity varied meaningfully between individual dogs and was affected by distractions — unusual noises, the presence of children, other animals — in live screening environments. The performance figures from highly controlled laboratory trials do not fully translate to the field without continued reinforcement training and handler competence. Training a COVID detection dog takes weeks under optimal protocols; maintaining performance requires ongoing sampling and reward cycles that depend on access to confirmed positive samples, which becomes harder to obtain as case prevalence falls.
Cultural considerations matter in Bangladesh. The Islamic tradition classifies dogs as haram or ritually impure, and while the jurisprudential mainstream position permits dogs for working purposes — guard, hunting, police work — the introduction of dogs into medical settings and mass public screening programmes would require careful communication with religious authorities and communities to avoid misunderstanding and resistance.
From Research to Health Policy
Bangladesh's second Joint External Evaluation of IHR core capacities in July 2024 called for the development of a new five-year National Action Plan for Health Security. The evaluation emphasised resource mapping, investment case development, and the identification of technical capacity gaps. Medical detection dogs represent precisely the kind of low-technology, high-throughput screening innovation that resource-constrained health systems need to evaluate systematically — not as a replacement for laboratory infrastructure, but as a complement to it.
icddr,b, one of the world's leading centres for diarrhoeal disease and infectious disease research, has the scientific capacity to conduct rigorous validation trials of canine COVID and respiratory virus detection in Bangladeshi conditions. The institution's research record — it has produced foundational work on cholera, rotavirus, and enteric disease that has shaped global health policy — positions it to evaluate whether the sensitivity and specificity figures produced in European and North American trials replicate in South Asian populations and climatic conditions, and whether dogs can be trained on locally obtained confirmed positive samples across the disease strains circulating in this region.
The global health community's post-COVID reckoning has produced a consensus that the next pandemic will require tools that work at scale, rapidly, cheaply, and without centralised laboratory infrastructure. Bangladesh, having navigated COVID-19 with a health system that performed beyond many expectations but exposed deep diagnostic gaps, has every incentive to be at the front of that innovation curve rather than waiting to receive its benefits second-hand. A dog that can detect infection before the infected person knows they are sick, at an airport or a factory gate, at a fraction of the cost of conventional testing, is not a novelty — it is a public health asset waiting to be properly deployed.
WinTK covers global health innovation and its implications for South Asian health systems. For more analysis on Bangladesh's health preparedness landscape, visit our news and analysis section.
