Patient Safety 2026: The world has named the harm. Now hospitals must build the system.

01At a glance

What the world’s leading patient safety conferences, regulators and journals said in 2025 and 2026, and the seven moves hospital leaders should make next.

02A movement at its midpoint

Patient safety in 2026 sits at an unusual moment. It is more than 25 years since To Err Is Human and five years into WHO’s Global Patient Safety Action Plan 2021–2030.

In August 2025 the movement lost Lucian Leape, whose work helped start it, at 94. A 2024 JAMA Internal Medicine viewpoint by Singh, Sittig and Classen judged that only marginal improvements have been realised since the Institute of Medicine report. That verdict is better read as a challenge than as a failure.

Across São Paulo, Anaheim, Manila, Chennai, New Delhi, Cape Town, Cairo and Dublin, practitioners spent 2025 and 2026 saying the same thing in different accents. Hospitals now know what harms patients. The next decade belongs to organisations that can see harm, learn from it and act faster than it recurs.

This paper distils that conversation for the people who do the work every day: quality heads, nursing leaders, NABH, JCI and GAHAR coordinators, risk managers and hospital chief executives.

03Harm: the quiet epidemic

The headline numbers have barely moved, and that is the finding. WHO’s 2026 campaign restates that 1 in 10 patients is harmed in care and about half of that harm is preventable. Its first Global Patient Safety Report (2024) adds that 1 in 20 patients suffers preventable medication harm, with 53 per cent of it occurring at prescribing.

The burden is heaviest where systems are thinnest. Opening the 7th Global Ministerial Summit in Manila in April 2025, WHO’s Western Pacific Regional Director cited 134 million adverse events a year in low- and middle-income countries, linked to 2.6 million preventable deaths. The OECD (2022) estimates more than 3 million deaths a year from unsafe care worldwide.

Well-resourced systems suggest that 1 in 10 is a floor. The SafeCare study of 2,800 charts across 11 Massachusetts hospitals found an adverse event in 23.6 per cent of admissions; a quarter were preventable and a third serious or worse. Lead author David Bates said the results show a need for better tools and increased measurement of patient harm.

The economic case is plain. WHO puts the cost of unsafe care at up to 12.6 per cent of health spending in high-income countries, and estimates that prevention could avert about USD 1.17 trillion a year. For a finance director, safety spending protects revenue.

Progress is fragile but recoverable. US central-line infections ran 94 per cent worse than expected in 2021, yet by 2024 CDC data showed CLABSI down 9 per cent and CAUTI down 10 per cent on the prior year. Disciplined prevention shows up in the data.

04The policy agenda

The global framework is the WHO Global Patient Safety Action Plan 2021–2030. Its first objective makes zero avoidable harm a state of mind and a rule of engagement; six more cover high-reliability systems, safe clinical processes, patient engagement, workforce, information and partnerships. The first Global Patient Safety Report (2024) shows how far there is to go.

Manila and the Mandaluyong Declaration

Sixty-four countries met in Manila on 3 and 4 April 2025 for the 7th Global Ministerial Summit and adopted the Mandaluyong Declaration. It frames patient safety as a pillar of resilient, people-centred health systems, and is the first in the series to bring safety into disaster, emergency and climate planning.

Dr Tedros Adhanom Ghebreyesus, Director-General, WHO · Manila, April 2025

At the same summit, WHO Western Pacific Regional Director Dr Saia Ma’u Piukala warned that blame culture remains deeply entrenched. The Philippine Health Secretary, Teodoro Herbosa, reminded delegates that behind every statistic is a person.

From first breath to lifelong care

World Patient Safety Day 2025 focused on safe care for every newborn and child, under the slogan “Patient safety from the start!”. WHO South-East Asia’s Dr Catharina Boehme said no child should be harmed by care meant to heal them. The 2026 theme, safe care for noncommunicable diseases, carries safety beyond the hospital into long-term management, homes and communities. Noncommunicable diseases cause 74 per cent of deaths worldwide.

05The 2025 to 2026 conference circuit

From São Paulo to Cape Town, conference programmes converged. Every 2026 programme reviewed here carried an AI or digital safety track, and learning over blame, workforce wellbeing and patient voice recurred throughout.

ISQua: technology, humanity, compassion. São Paulo 2025 put patient safety first among eight tracks. Among its most remembered moments, Francisco Lima and his wife spoke about losing their daughter Julia to medical error. One ISQua scholar summed up the mood: safety is not a technical metric but a moral commitment. Dublin 2026 explicitly named Safety-II, human factors, algorithmic bias, virtual care and wearables in its safety track.

IHI: courage as a safety competency. At the December 2025 IHI Forum in Anaheim, with about 2,400 attendees, new chief executive Sylvia Trent-Adams and co-founder Don Berwick reframed safety as moral courage, including the courage to speak up about harm caused by policy.

Dr Donald Berwick, President Emeritus, IHI · IHI National Forum 2025

HSJ: from talk to delivery. The United Kingdom’s 2026 Congress theme moved beyond honest conversations to delivering safer care: embedding PSIRF, listening to patients, digital and AI, and safety under pressure.

PSMF: the zero-harm flag still flies. The Patient Safety Movement Foundation’s 12th Summit brought together CMS, UnitedHealth Group and Kaiser. Founder Joe Kiani said zero preventable harm is not an abstract goal but a responsibility.

The Global South leads with scale. CAHOCON 2026 in Chennai became the first ISQua regional conference for South Asia. NABH’s conference in New Delhi gave a full track to AI, ABDM and cybersecurity. In Cape Town, the COHSASA-partnered forum argued that safe care can thrive under resource constraints.

06Seven defining themes

1. AI is the top hazard and the top opportunity

No theme moved faster. ECRI ranked risks from AI-enabled health technologies first on its 2025 hazards list and misuse of AI chatbots first in 2026. Chief executive Marcus Schabacker reminded the field that medicine is a fundamentally human endeavour.

The opportunity is just as real. In 2026 the US National Academy of Medicine launched Patient Safety in the Era of AI, a two-year effort towards a national strategy, co-chaired by leaders from Mayo Clinic and CommonSpirit and patient advocate Sue Sheridan.

2. Most harm goes unseen

The most uncomfortable data point came from the US Office of Inspector General in July 2025: hospitals did not capture half of patient harm events in their incident reporting or surveillance systems. Staff often did not recognise events as harm, and few captured events led to improvements.

3. Learning, not blame: PSIRF and Safety-II go mainstream

England’s Patient Safety Incident Response Framework has become the reference model for proportionate, system-focused learning. An October 2025 HSSIB review found broad support, though resourcing was uneven.

Mel Ottewill, Senior Safety Investigator, HSSIB · October 2025

Safety-II, which studies why work usually goes right, is now named in ISQua’s programme and runs through the BMJ Oslo resilience theme.

4. The zero-harm debate

WHO and the Patient Safety Movement Foundation keep zero avoidable harm as their rallying cry. In England, NHS England’s national patient safety director has called zero harm unachievable and warned against false hope. The positions are closer than they look. Zero works as an ethical direction, not a performance indicator: hospitals that target it count incidents down, while hospitals that steer by it learn their way forward.

5. Workforce and speaking up are safety infrastructure

ECRI’s 2026 list includes persistent workforce shortages, and its chief medical officer warned of the risk when clinicians do not feel psychologically safe reporting concerns. Queensland’s nurse-ratio legislation was followed by lower mortality, readmissions and length of stay (Lancet, 2021). Staffing data is safety data.

6. Patients and families are partners

ECRI’s 2025 list put dismissing patient and family concerns at number one, yet only 13 per cent of countries have patient representatives on most hospital boards. Bereaved families at ISQua, the HSJ patient-empowerment theme and the COHSASA community stream brought that voice to the podium.

7. Measurement must go beyond counting incidents

An emerging theme: the Office of Inspector General wants a national harm taxonomy, Egypt has produced its first national patient safety report, and NABH has put data-driven quality improvement on its agenda. The direction is from volume of reports to validity of detection.

07Regional lens

India: accreditation and digital public infrastructure

NABH’s 6th edition hospital standards, effective January 2025, give more weight to digital health. Chief executive Dr Atul Mohan Kochhar framed them as part of elevating the quality and safety of healthcare services. A second edition of NABH’s Digital Health Standards is in draft, and ABDM passed 100 crore ABHA-linked records in May 2026.

The harm evidence is sobering. A 2025 review of 40 Indian studies covering over 300,000 inpatients found a median medication error incidence of 34 per cent, rising to 74 per cent in emergency departments, with most errors at prescribing and administration. Accreditation still reaches a small minority of hospitals, and the National Patient Safety Implementation Framework (2018–2025) has lapsed without a confirmed successor. Until policy catches up, NABH standards and ABDM are the main levers, and hospitals that digitise quality now will set the benchmark.

Singapore: the human in the loop

Ong Ye Kung, Minister for Health, Singapore · Synapxe AI Accelerate, June 2025

Behind that principle sit a regulatory sandbox for AI medical devices with post-market surveillance, national chest X-ray AI by the end of 2026, and AI risk scoring in the next phase of Healthier SG.

Middle East: accreditation as a lever for reform

Egypt is the region’s fastest mover. GAHAR accredited 945 facilities across 26 governorates in FY2025/26 and had its own ISQua accreditation renewed. Accreditation is linked to Universal Health Insurance contracting. In October 2026 Egypt launched a National Quality and Patient Safety Strategy 2026 to 2030 with a confidential national adverse-event reporting system and the country’s first national patient safety report, aimed at institutional learning rather than punishment. Saudi Arabia’s Patient Safety Center brought in a national sentinel event policy from January 2025.

South Africa: when regulators cannot be everywhere

The Office of Health Standards Compliance completed 3,116 inspections in 2024/25 against a target of 3,741, and reported hospitals borrowing equipment to impress inspectors. A preliminary 2026 study of ten public hospitals found only 7 to 15 per cent of incidents formally reported, driven by fear of blame and workload. With National Health Insurance approaching, hospitals need continuous, self-generated evidence of safety.

08Digital quality as the backbone

Across every region and conference reviewed, one conclusion recurs: hospitals cannot improve harm they do not see, and cannot learn from reports that never close.

The Office of Inspector General’s diagnosis goes further than low reporting. Harm is missed because of narrow definitions, culture and weak follow-through. A digital form that only makes submission easier will not close that gap. The evidence points to five capabilities a modern digital quality management system must deliver.

• Multi-source detection: voluntary reports combined with triggers, infection surveillance, audits, patient feedback and indicator signals.
• A shared harm taxonomy, so the bedside and the board classify events the same way.
• Proportionate learning responses: PSIRF-style after-action reviews, thematic analysis and root cause analysis as one tool among several.
• Closed-loop action tracking, where every learning becomes an owned, dated, measured action.
• Accreditation-ready evidence for NABH, JCI, GAHAR and OHSC as a by-product of daily work, not a scramble before a survey.

Medblaze Infini is built around this model: incident reporting and root cause analysis, risk registers, audits, infection surveillance, patient feedback and quality dashboards on one data spine, mapped to the accreditation frameworks of India, the Middle East, Singapore and South Africa. The wider market agrees: MarketsandMarkets sizes healthcare quality management at USD 1.26 billion in 2024, rising to USD 2.51 billion by 2030.

09What leaders should do now

For hospital chief executives, medical directors and quality heads, the 2025 and 2026 conversation comes down to seven moves.

1. Measure what you miss, not just what you log.

Run a trigger-tool review on a chart sample this quarter and compare it with your incident log. The gap is your real baseline.

2. Adopt a learning response framework.

Replace blanket RCAs with proportionate responses modelled on PSIRF and mapped to NABH, JCI or GAHAR. Report learning and actions closed to the board, not just incident counts.

3. Close every loop.

Give every significant event a named owner, a deadline and a measure of effect. Make actions completed and verified a board-level indicator.

4. Govern AI before it enters the workflow.

Set up an AI oversight group, inventory every clinical AI tool, require a human decision-maker for each AI-supported judgment, and rehearse downtime for digital darkness events.

5. Make speaking up safe and visible.

Measure psychological safety, thank reporters publicly, and never blame individuals for system failures. Treat staffing data as safety data.

6. Give patients and families a formal voice.

Create an escalation route for families, seat patient representatives on quality committees, and mine complaints as safety intelligence.

7. Digitise your quality system end to end.

Move off paper, spreadsheets and messaging groups to one integrated system, so incident, risk, audit, infection, feedback and accreditation data inform each other.

10From naming harm to building the system

The world no longer lacks knowledge. It knows the scale of harm, has a ten-year plan, a ministerial declaration and a shared language of learning over blame. What it lacks is execution: seeing the harm that reporting misses, learning in proportion, and proving that actions changed care. That capability is being built hospital by hospital, not in Geneva.

The quality professionals who run safety huddles, close out root cause analyses and prepare for NABH and GAHAR surveys are at the centre of global health reform. Over the next five years their work becomes the evidence that the 2030 plan is being delivered. The hospitals that pair human judgment with digital visibility will lead.

See how Medblaze Infini turns incident reporting into a learning system →

AFrequently asked questions

What are the biggest patient safety issues in 2026?

AI in diagnosis and decision-making, which is ECRI’s top patient safety concern for 2026, along with unrecognised and unreported harm, a persistent culture of blame, workforce shortages, and medication and diagnostic errors. World Patient Safety Day 2026 focuses on safe care for people living with noncommunicable diseases.

What is the WHO Global Patient Safety Action Plan 2021 to 2030?

WHO’s ten-year framework, adopted at the 74th World Health Assembly, aiming for zero avoidable harm through seven strategic objectives. WHO’s 2024 report found that only about a third of countries have a national plan and 11 per cent report sufficient funding for planned interventions.

What is PSIRF and how does it relate to Safety-II?

The Patient Safety Incident Response Framework is England’s approach to patient safety incidents. It replaces blanket root cause analyses with proportionate, system-focused learning responses. Safety-II complements it by studying how everyday work usually succeeds, rather than only investigating the occasions when it fails.

Why is incident reporting software alone not enough?

Voluntary reporting misses most harm. US federal auditors found in 2025 that hospitals did not capture half of harm events in their own systems. Effective detection combines voluntary reports with trigger tools, surveillance data, a shared harm taxonomy and closed-loop action tracking, which is what a digital quality management system provides.

How should hospitals use AI in patient safety?

Use AI to surface signals, cluster themes and support review, while trained people make the final judgment. Govern it with an oversight group, an inventory of every clinical AI tool, clinician training and rehearsed downtime plans, in line with ECRI and Singapore Ministry of Health guidance.

BReferences

• WHO, World Patient Safety Day 2026
• WHO, World Patient Safety Day 2025
• WHO, Global Patient Safety Report 2024
• PSLHub, summary of WHO Global Patient Safety Report 2024
• WHO news, progress on patient safety (June 2024)
• WHO, 7th Global Ministerial Summit on Patient Safety (April 2025)
• WHO WPRO, keynote by Dr Saia Ma’u Piukala, Manila
• Malaya, Mandaluyong Declaration released
• WHO SEARO, World Patient Safety Day 2025
• OECD, The Economics of Patient Safety (2022)
• Mass General Brigham, SafeCare Study (NEJM 2023)
• Newman-Toker et al., BMJ Quality & Safety 2024
• Singh, Sittig and Classen, JAMA Internal Medicine (2024)
• Boston Globe, Lucian Leape obituary (2025)
• Axios, CMS National Impact Assessment (2024)
• CIDRAP, CDC healthcare-associated infection progress report (2024 data)
• ISQua São Paulo 2025
• ISQua blog, John Byabagambi’s reflections
• ISQua Dublin 2026
• Fierce Healthcare, IHI Forum 2025
• HSJ Patient Safety Congress
• HSJ, “Zero harm is unachievable”
• Patient Safety Movement Foundation, 12th World Patient Safety Summit
• BMJ Healthcare Improvement Forum, Singapore 2025
• BMJ Healthcare Improvement Forum, Oslo 2026
• ISQua, CAHOCON 2026 announcement
• NABH, Patient Safety Conference 2026
• COHSASA, African healthcare leaders reimagine patient safety
• Daily News Egypt, National Quality and Patient Safety Strategy
• Daily News Egypt, GAHAR FY2025/26
• PSL Hub, Saudi Patient Safety Center sentinel event policy
• Sleep Review, ECRI Health Technology Hazards 2025
• ECRI, Health Technology Hazards 2026
• Healthcare Dive, ECRI Patient Safety Concerns 2025
• HIT Consultant, ECRI Patient Safety Concerns 2026
• National Academy of Medicine, Patient Safety in the Era of AI
• HSSIB, PSIRF review (October 2025)
• AHRQ PSNet, nurse-to-patient ratio legislation (Lancet 2021)
• HHS OIG, Hospitals Did Not Capture Half of Patient Harm Events (2025)
• Becker’s, OIG: hospitals report 1 per cent of adverse events (2012)
• AHRQ PSNet, Global Trigger Tool versus voluntary reporting
• Population Medicine, incident reporting in South African public hospitals (2026)
• JAMIA 2026, AI for incident causation analysis
• MOH Singapore, Minister Ong Ye Kung at Synapxe AI Accelerate (2025)
• Digital Health News, NABH 6th edition standards
• NABH, Digital Health Standards implementation programme
• News On AIR, ABDM 100 crore records
• Medical Dialogues, medication errors in Indian hospitals (IJCCM 2025)
• Legacy IAS, “Making health care safe for every Indian” (2025)
• TimesLIVE, OHSC inspections (October 2025)
• MarketsandMarkets, healthcare quality management market

This paper summarises publicly reported information for educational purposes and is not legal or regulatory advice. Figures are reproduced as published by the cited sources; where studies use older data or overlapping populations, the exhibit notes say so.