THE smell hits first. In a district hospital just outside Dhaka, a patient’s family member searches for a working toilet. The door hangs half-broken; inside, there is no running water, no soap, and no record of when it was last cleaned. A nurse shrugs, maintenance is ‘someone else’s job.’ Down the corridor, a cleaner from a historically marginalised Harijan community finishes a shift with no formal training, no protective equipment, and no pathway to advancement.
Walk into almost any public hospital — Dhaka, Rajshahi, Chattogram and Sylhet — and before you encounter a doctor, before you find a ward, before anyone takes your blood pressure, you will find a toilet. Or what remains of one. Broken latches. Overflowing bins. Floors that have not been disinfected since the previous administration. Lights that have not worked in months.
Now imagine a different scene: the same cleaner scans a small QR code pasted outside the toilet door with a low-cost smartphone. A simple interface prompts a checklist — cleaned surfaces, refilled supplies, functional taps. The scan logs a timestamp, uploads to a central dashboard, and flags missing items. A supervisor across the building sees a red alert: ‘Water supply interrupted, Block B.’ Within minutes, maintenance is notified. The act is small, almost mundane. Yet it is also transformative.
On paper, Bangladesh is living through a technological moment. Smartphones are everywhere. Digital payments are routine. Telemedicine platforms are multiplying. And yet, walk into too many government hospitals and you will encounter a scene that belongs to a different century: a toilet that is locked, broken, or unusable; no soap near the sink; overflowing bins; and families improvising care in corridors.
This is not merely a failure of maintenance. It is a failure of systems. They are the most visible symbols of a deeper crisis: a health system where accountability is weak, basic operations are undervalued, and dignity is treated as optional. The connection between broken hygiene and institutional failure is neither abstract nor incidental. Public hospitals in Bangladesh frequently struggle with non-functional sanitation facilities, irregular cleaning schedules, and weak infection control practices. These are not merely aesthetic shortcomings; they directly affect patient outcomes and public trust, pushing patients toward out-of-pocket private care and informal providers.
The future of healthcare is undoubtedly digital. But in Bangladesh, before artificial intelligence transforms diagnostics or telemedicine expands access, it must first fix the basics — through traceability, accountability and behavioral change.
Start with the toilet.
This may sound reductive, even provocative. But sanitation is not peripheral to healthcare; it is its frontline. Hospital-acquired infections are often the downstream consequence of upstream neglect. When toilets are broken, handwashing falters. When cleaning schedules are inconsistent, pathogens spread. In ICUs, where antibiotic use is high, such lapses accelerate antimicrobial resistance, a growing public health threat that Bangladesh cannot afford to ignore.
A 2022 report of the health ministry noted gaps in infection prevention protocols and inconsistent monitoring across facilities. When patients encounter visibly neglected environments, they infer, often correctly, that unseen processes may be equally fragile. Trust erodes not through policy failures alone, but through everyday experiences of neglect.
Developed health systems learned this lesson decades ago. In the United Kingdom, the National Health Service introduced strict, auditable cleaning protocols after high-profile hospital infection outbreaks in the 2000s. Cleanliness audits became routine, with ward-level scorecards publicly displayed and linked to funding and management accountability. Similarly, hospitals in Japan have long embedded hygiene into operational culture: cleaning logs, standardised checklists and visible accountability are part of daily routines, not afterthoughts. In the United States, infection control programs tied to agencies like the Centres for Disease Control and Prevention require hospitals to track and report hospital-acquired infections, with digital systems flagging anomalies in near real time. The lesson across these contexts is consistent: hygiene improves when it is measured, monitored and made visible.
Closer to home, South Asia offers both cautionary tales and promising innovations. India’s Swachh Bharat Mission demonstrated how combining infrastructure investment with behavioral monitoring can shift public norms. In several states, municipal bodies experimented with QR codes in public toilets, allowing citizens to rate cleanliness and report issues instantly — creating a feedback loop between users and administrators. Meanwhile, in urban hospitals in Sri Lanka, structured infection control units have begun digitizing cleaning schedules and supply tracking, contributing to relatively stronger hygiene outcomes compared to regional peers. Even in resource-constrained settings, the principle holds: small digital interventions can reinforce accountability where manual systems fail.
Labour equity sits at the center of this system. The sanitation workforce, often drawn from Horijon communities, has long been excluded from formal training, fair wages, and professional mobility. Their work is essential, yet stigmatised. This is not only unjust; it is inefficient. Cleanliness in healthcare settings is a skilled practice, requiring knowledge of disinfectants, infection pathways, and safety protocols. Developed countries treat it as such. In Singapore, hospital cleaning staff undergo standardised certification and are integrated into infection control teams, with clear performance metrics and career ladders.
Bangladesh’s future healthcare system will certainly be digital. Artificial intelligence will eventually help read X-rays, interpret scans, and optimise clinical decisions. But if we want a health revolution that actually reaches ordinary people, AI must first do something less glamorous and far more transformative: fix the basics through traceability, accountability, and behavioural change.
In clinical care, we talk about `patient experience’ as if it is a soft metric. It is not. Experience is evidence. Evidence of whether a system can reliably deliver what it promises. A functioning, clean toilet is the most basic test of a hospital’s respect for human beings. If a facility cannot keep toilets open, clean and stocked, how confident should we be that it can maintain sterile procedures, segregate medical waste, enforce hand hygiene, or manage antibiotic stewardship?
This is where Bangladesh’s digital ambition has often gone wrong. We have built pockets of computerisation, what some quietly call ‘MIS theatre,’ that generate reports upward but change little on the ground. A database is not a health system. An app is not accountability. Digitisation that merely reproduces old habits in electronic form can even worsen cynicism: people see ‘modernisation’ on banners, while living the same dysfunction in wards.
Developed systems are already extending such approaches. In Germany, hospitals increasingly use sensor-based monitoring to track hand hygiene compliance and restroom usage, feeding into centralised dashboards that optimise cleaning schedules. In South Korea, smart public sanitation systems in cities like Seoul integrate QR access logs with maintenance alerts, ensuring rapid response to faults. These are higher-cost versions of the same idea Bangladesh can implement at a fraction of the price using QR codes and smartphones.
The most actionable use of artificial intelligence in Bangladesh today is not futuristic diagnosis. It is operational reliability.
Here is a proposal modest enough to be dismissed and radical enough to matter: put a QR code on every hospital toilet across the country. Not a decorative one. A live one. Linked to a real-time dashboard accessible to hospital administration, district health officers and eventually the health ministry.
A practical intervention that can primarily be piloted: installing QR-code every toilet door in public hospitals (and later, every ward washroom, handwashing station and waste disposal area).
Every time a cleaner completes a round, they scan the code. Time-stamped. Geolocated. Logged. Patients, attendants, and staff encounters a problem, a broken fixture, a missing soap dispenser, a locked door, they scan and report. They can scan and submit a 10-second feedback rating (clean/unclean, soap available, water available, door functional). For those without smartphones, an SMS short-code can work. Given that smartphone penetration in Bangladesh crossed 45 million users and continues to climb steeply, this is not a technological leap. It is a management decision.
This is not about shaming workers. It is about shifting from a culture of vague responsibility to a culture of measurable service. What gets measured gets managed. What gets tracked gets fixed. Crucially, the data must be visible to those who can act: the ward supervisor, the hospital director, and a small oversight cell at the ministry or the Directorate General of Health Services. A dashboard that sits in Dhaka and never reaches the matron or the facility manager will become another performative screen.
Technology alone will fail if we ignore the human system beneath it, especially the cleaning workforce. This is the workforce we pretend not to see. Sanitation work in Bangladesh has historically been assigned to marginalised Dalit communities, often referred to locally as ‘Harijan,’ a term many now consider outdated. Many workers have done this labour for generations with limited training, limited protective equipment, low pay, and social stigma. We cannot build `smart hospitals’ on an employment structure that remains feudal in practice.
If the country wants clean hospitals, it must professionalise cleaning as essential healthcare work: Modern training for modern facilities (chemicals, mops, waste segregation, infection-control routines, restroom maintenance); certification and career ladders so sanitation work is not a dead-end role but a recognised skill; protective equipment and occupational safety as a non-negotiable standard; and respectful supervision that treats cleaning as part of patient safety, not as an afterthought.
One pragmatic model already exists in Bangladesh: the disciplined housekeeping systems of five-star hotels. Government hospitals can partner with hospitality training teams to design short, practical modules — adapted to clinical realities. It is a scandal that in a place where infections can kill, hygiene protocols are often weaker than those used to protect hotel ratings.
Some will ask: why spend political attention on toilets when Bangladesh needs ICU upgrades, more doctors, more medicines?
Because toilets are not separate from clinical outcomes. Poor sanitation and weak infection control increase hospital-acquired infections. More infections mean more antibiotics. More antibiotics — used under weak oversight — mean more resistance. And antimicrobial resistance is already one of the gravest, most expensive threats facing modern medicine.
In ICUs, antimicrobial resistance is not only a matter of which drug a physician selects. It is the sum of dozens of small operational behaviours: hand hygiene, surface disinfection, isolation practices, waste management, catheter care and the ability to track pathogens and prescribing patterns over time.
That is exactly where simple AI-enabled systems can lay the groundwork: Stock and supply analytics to prevent shortages of gloves, disinfectants, and soap (because ‘no supplies’ becomes ‘no hygiene’); digital antibiotic stewardship logs that flag unusually broad-spectrum prescribing, prolonged courses without review, or unit-level spikes; outbreak detection through basic clustering signals from lab results and ward locations — starting small, improving as data quality improves.
Bangladesh does not need to leap immediately into genomics-driven artificial intelligence. It needs the discipline of data capture and compliance, and then the advanced analytics can follow. Telemedicine, e-Health cards, and AI-assisted diagnostics will only succeed if people trust the system. Trust is not built by speeches; it is built by daily proof. We must keep in mind that small tech builds trust for big digitisation. When patients see cleaner wards, functioning toilets, predictable queues and fewer stock-outs, they become more willing to share data, adopt digital records and use referral pathways. Clinicians become more willing to document care digitally when they see reduced chaos and less administrative burden.
The government should not outsource its health operating system to expensive, closed vendor solutions that create dependency and lock-in. Instead, it can catalyze local innovation, quickly and credibly, through targeted public funding. First, it needs creating a Health Operations Innovation Fund that commissions Bangladeshi universities to develop open, modular tools such as QR hygiene tracking, supply-chain dashboards, SMS feedback systems, anomaly detection, and simple reporting interfaces; pilot in one large government hospital per division, publish results, iterate rapidly, and scale what works.
The government can fund institutions like the Bangladesh University of Engineering and Technology to design AI-driven hygiene tracking systems built for local conditions: low bandwidth, multilingual interfaces, interoperable with existing hospital management infrastructure, affordable for upazila-level health facilities. BRAC University’s public health faculty, ICDDR,B’s epidemiological expertise, the TMMS experiences in primary healthcare delivery in rural area, and Dhaka University’s social science researchers could be convened around a national digital health architecture commission, not another committee that produces a report, but a funded, time-bound design sprint with implementation targets.
This is not an abstract dream. It is an achievable public-interest engineering project, one that brings together clinicians, engineers, behavioural scientists, hospital managers, and sanitation workers. The country’s healthcare future will indeed include AI reading scans and supporting diagnosis. But if we skip the basics, hygiene, logistics, supervision, respect for frontline labour, high-tech medicine will sit on a low-trust foundation.
Bangladesh has extraordinary technical talent, remarkable institutional resilience, and a public health tradition, from the cholera research of ICDDR,B to the community health innovation of BRAC, that the world has legitimately looked to for inspiration. The ingredients for transformation are present.
What is needed now is the political courage to start where the need is most visible, most basic, and most human.
Mohammad Aminul Islam is a senior lecturer in media studies and journalism at the University of Liberal Arts Bangladesh.
