We've implemented HMS at hospitals ranging from 50 beds to 400 beds. The failure mode is almost always the same: vendors build what's impressive in a demo, not what works in a Bangladeshi ward at 11pm with intermittent internet.
We've implemented hospital management systems at eleven facilities in Bangladesh — ranging from a 45-bed private clinic in Sylhet to a 380-bed hospital group in Dhaka. The failures we've seen, and the ones we've been called in to rescue, follow patterns that are consistent and predictable.
The common thread is not technical. The systems that fail aren't failing because the database schema is wrong or the UI is ugly. They fail because they were designed for how a hospital administrator imagines a hospital works, rather than how a Bangladeshi hospital actually operates at 11pm on a Tuesday.
Most HMS systems have a patient registration flow designed for orderly intake: one patient, one clerk, complete the form, proceed. Bangladeshi hospitals — especially public ones and popular private facilities — have chaotic admission patterns. Multiple patients arrive simultaneously. Family members are completing forms on behalf of patients. Emergency admissions bypass normal intake.
Systems that require sequential, complete data entry at registration create bottlenecks that staff work around by not using the system. We've seen hospitals where 60% of patients are registered after the fact — the data entry happens at billing time, not at admission.
Hospital billing in Bangladesh involves: advance deposits, partial payments, family-split payments (where different family members pay different portions), discount authorisations at multiple levels, insurance pre-authorisation (for the minority with insurance), and the informal "arrangement" that exists at many hospitals for established patients.
Western HMS billing modules handle insurance claims and card payments well. They handle cash-first, advance-deposit, multi-payer, discount-heavy billing poorly. This mismatch sends hospitals back to manual billing systems or parallel spreadsheets.
Prescription management, e-consultation notes, diagnostic result review — all excellent on paper, all poorly adopted in practice. The reason: Bangladeshi hospital doctors are extraordinarily busy. A senior consultant at a private hospital may see 60–80 outpatients a day in addition to ward rounds and emergencies.
Any digital workflow that adds friction to a consultation is abandoned. The doctor who hand-writes a prescription in 45 seconds will not switch to a system that takes 2 minutes to navigate. The only HMS features doctors adopt reliably are ones that make something they already do faster or that give them information they currently have to ask for.
Most HMS systems generate reports that look impressive in demos and are useless in daily operations. A 50-column revenue report is not what the charge nurse needs. She needs to know which beds are occupied, which discharges are expected today, and which patients have outstanding balances that need to be resolved before discharge.
Operational dashboards — simple, real-time, role-specific — are consistently the most valuable part of an HMS, and consistently underinvested in vendor systems.
Power cuts and internet outages are a reality in Bangladeshi hospitals outside Dhaka. An HMS that requires continuous connectivity is unusable during load shedding. We've designed all our HMS implementations with an offline-first core: critical functions (patient registration, bed management, basic billing) work on local network even when internet is down.
Start with billing and bed management — the two workflows that affect every patient every day. Get them working correctly before touching clinical features. The hospital's trust in the system is built on the reliability of these core functions. Once staff see that the system is dependable for the basics, adoption of clinical features follows.
Work With Us
From ERP to HealthTech to custom SaaS — we partner with businesses that want software built properly.