Highlights:
- Will AI and Dogs Replace Doctors? Cancer sniffing dogs and AI for early detection.
- While investment on sniffer dogs for medical diagnosis, Junior Doctors on strike in AP
- Not only doctors welfare is being neglected but sanitation gaps exist in Healthcare.
By Dr. Amitav Banerjee, Chairperson of the Universal Health Organisation (UHO)
Will AI and Dogs Replace Doctors? Cancer sniffing dogs and AI for early detection.
Sniffer dogs have been used by police squads, and detective agencies to trace criminals and explosives. Canine sense of smell is far superior compared to humans. A diagnostic startup “Dognosis” is using this capability of dogs to scale up early diagnosis of cancers in humans. Patients breath into a mask which connected to the company’s laboratory where dogs sniff the exhaled breath. The response in the dogs are in turn recorded by Artificial Intelligence (AI) sensors which in turn study the dogs’ respiration, movements, brain activity and body language to determine whether the breath samples carry signs of cancer. It the technology succeed, it will do away with imaging scans, blood tests and biopsies.
This development has been reported in the ET Healthworld. The results of the Phase 2 trials have been published in the Journal of Clinical Oncology. On critical appraisal of the study, UHO offers the following observations.
Dognosis has reported results from a Phase II study involving 3,275 participants, including people with several types of biopsy-confirmed cancer. The study reported approximately 90.8% sensitivity and 91.3% specificity, with an area under the curve (AUC) of 0.962. Sensitivity of about 90.6% was reported for early-stage cancers. These are impressive results for a relatively novel approach and provide evidence that trained dogs can detect patterns in volatile organic compounds (VOCs) associated with cancer in human breath.
However, the reported results need to be interpreted carefully. The Phase II study was essentially a controlled diagnostic-accuracy study rather than a demonstration of how the technology would perform in the general population. The researchers included known cancer patients and controls, which are useful for determining whether the system can distinguish the two groups but does not fully reproduce population screening.
This distinction is extremely important because positive predictive value depends heavily on cancer prevalence. Even a test with approximately 91% sensitivity and specificity can produce many more false positives than true positives when used among largely healthy people, where cancer prevalence is low. Consequently, the technology should not currently be described as a test that can tell an individual whether or not they have cancer. A positive result would need confirmation using established diagnostic methods.
Another limitation is that the Phase II population and controlled testing environment may not represent the complexity of real-world screening. Factors such as smoking, infections, inflammation, medications, diet, other illnesses and environmental exposures could potentially affect breath chemistry.
The role of AI also deserves some caution. Although AI can help interpret the complex behavioral patterns produced by multiple dogs, it does not automatically solve the fundamental biological and statistical challenges. There is a possibility that algorithms could inadvertently learn characteristics specific to particular hospitals, dogs, sample-handling procedures or patient populations. Independent validation at different hospitals and in different populations is therefore essential.
There are also potential conflicts of interest. The Phase II research was funded by Dognosis, and several researchers have connections with the company. Such relationships do not invalidate the findings, particularly when the study has undergone peer review, but they make independent replication particularly important. Ideally, future trials should involve researchers and institutions without financial interests in the technology.
The regulatory position also deserves clarification. The company has indicated that its planned prescreening product may not require conventional regulatory approval because it is a prescreening rather than a diagnostic test. That claim should be treated cautiously until the relevant Indian regulatory authorities and the precise product classification are clear. The distinction between screening, risk assessment and diagnosis can have important regulatory implications.
The evidence does not yet justify replacing established cancer-screening programmes or diagnostic procedures. The key unanswered question is not simply whether dogs can detect cancer, but whether this technology can reliably identify cancer in ordinary, mostly healthy populations and ultimately improve clinical outcomes.
We have concerns that in healthy populations the accuracy of such gimmicky tests will be affected adversely throwing up a large number of false positives who in turn will face a battery of other tests and sometimes unnecessary biopsies and surgeries. Players with various conflicts of interests will enter into the fray and reap huge profits without improving people’s health.
We need better trained doctors than better trained dogs to make a proper diagnosis combining patient’s presenting complaints, history, background, clinical examination followed by appropriate diagnostic tests.
Regrettably, a doctor training is getting diluted with increase in medical colleges without matching faculty and resources, producing doctors with poor clinical and diagnostic skills. At this rate, we may see a future when dogs may replace the doctor’s medical assistant, if not the doctor!
While investment on sniffer dogs for medical diagnosis, Junior Doctors on strike in AP
While dogs are being medically trained, real junior doctors in Andhra are not happy with their life and pay.
The Junior Doctors Association (JUDA), representing postgraduate (PG) medical students, interns and senior residents, continued its protest at Government General Hospital (GGH) in Vijayawada on Tuesday, with the agitation entering its eighth day over a 15% stipend hike, recruitment norms and the proposed increase in the retirement age of professors.
JUDA executive member and PG forensic medicine student Dr Naveen said “We have been asking for a 30 per cent stipend hike, but they told us they will only give 3-5 per cent, so we agreed for 15 per cent, which is according to the government’s GO. The GO says every two years there will be a 15% hike in the stipend. So our ask is to at least meet that. But the government is not agreeing and we have been in negotiations.”
“The government are recruiting non-clinical people in large numbers to teach first years, which is not right, because people who are doing MBBS and MD face a job crunch,” said Dr Pawan Kumar, a PG pulmonology student.
On the stipend demand, Pawan said the doctors were seeking parity with Telangana. “All we are asking is to be on par with the Telangana government because they are able to implement the GO, so now they are getting Rs 70,000-75,000. We don’t want to settle for the 3-5 per cent proposal because then it will become a habit for them. We are protesting peacefully,” he said. The doctors, on the eighth day, withdrew emergency services completely.
Without going in much details, UHO recommends that more should be invested in human resources like doctors to address our public health problems instead of training dogs for the purpose.
Not only doctors welfare is being neglected but sanitation gaps exist in Healthcare.
A new WHO and UNICEF report has highlighted a significant gap in basic sanitation across healthcare facilities worldwide. In 2025, only 40% of facilities met the criteria for basic sanitation services, while 11% had no sanitation service at all, leaving an estimated 936 million people without access to adequate sanitation during healthcare encounters.
The report, which provides the first global estimates for sanitation services in healthcare settings, assessed whether facilities had usable toilets meeting key requirements for patients and staff. These included dedicated staff facilities, sex-separated toilets with menstrual hygiene provisions, and accessibility for people with limited mobility.
The findings varied substantially by region. Basic sanitation services were available in only 23% of healthcare facilities in sub-Saharan Africa, compared with 55% in central and southern Asia.
The sanitation findings were part of a broader assessment of essential environmental conditions in healthcare facilities. Globally, 85% of facilities had basic water services, while 72% met basic hygiene standards and 71% met basic healthcare waste management standards.
Environmental cleaning emerged as another major gap. Only 59% of facilities met basic environmental cleaning standards, which require both appropriate cleaning protocols and trained cleaning personnel. The report estimated that 18% of facilities, serving approximately 1.5 billion people, met neither requirement.
The disparity was even more pronounced in the world’s least-developed countries, where only 24% of healthcare facilities met the basic environmental cleaning standard.
The figures point to an infrastructure problem that extends beyond patient comfort. Gaps in sanitation, cleaning, hygiene and waste management can complicate infection prevention and control efforts and add another layer of operational challenges to already demanding clinical environments.
UHO recommends that to ensure quality health care to the population, we should not be distracted by the glitz and glamour of rapidly emerging technologies without first addressing the fundamentals, i.e. human health resources, medical equipments and good quality health infrastructures and hospitals.
Also Read:
Universal Health Organisation (UHO) Weekly Newsletter – 14 AUGUST 2026