Patient safety remains a global healthcare challenge: the World Health Organization (WHO) estimates that around 1 in 10 patients is harmed during care, while more than 3 million deaths occur each year because of unsafe care. The available figures span global estimates, OECD comparisons, U.S. reporting systems, and historical improvement programs, so measurement period and geography matter when interpreting them.
Contents
- Global scale of patient harm
- Preventability and economic impact
- Healthcare-associated infections
- Medication safety and reported events
- Diagnostic, surgical, and imaging safety
- What hospital trend data show
Global scale of patient harm
The WHO Patient safety fact sheet places the overall burden at around 1 in every 10 patients harmed in healthcare. In low- and middle-income countries, as many as 4 in 100 people die from unsafe care. The WHO Global Patient Safety Observatory also estimates 134 million adverse events in hospitals in low- and middle-income countries.
These figures describe different measures. “Harmed” is broader than “died,” and the 134 million estimate concerns hospital adverse events in low- and middle-income countries rather than all healthcare encounters worldwide. The Observatory’s key figures repeat the global estimate that about 1 in 10 patients is harmed. By May 2023, the Observatory had received official patient-safety survey responses from 108 WHO Member States; across that survey, 27% of criteria were fully met.
The OECD Patient safety topic provides a complementary view from member and partner health systems. In 2020, the share of patients reporting a medical mistake during the previous two years ranged from about 6% to above 10%, depending on the country. Among hospitalized adults, patient-safety incidents during the last hospitalization ranged from 4% to 17%.
Staff experience also varies. Only about 6 in 10 health workers said that handoffs and information exchange were adequate for maintaining a safe care environment, and perceptions of safe handoffs and transitions differed by more than 20 percentage points across countries. These measures capture patient and workforce experience rather than a single global incidence rate.
Preventability and economic impact
More than 50% of patient harm is preventable according to the WHO Patient safety fact sheet—approximately 1 in 20 patients when expressed as the equivalent proportion in the fact sheet. Half of preventable harm is attributed to medications. In primary and ambulatory settings, some estimates suggest that as many as 4 in 10 patients are harmed, and up to 80% of that harm may be avoidable.
Patient harm also has a measurable economic dimension. The WHO estimates that harm potentially reduces global economic growth by 0.7% each year, while the indirect cost of patient harm amounts to trillions of U.S. dollars annually. Unsafe medication practices and medication errors alone are estimated to cost US$42 billion each year, almost 1% of global health expenditure.
OECD data estimate that up to 13% of healthcare spending across OECD countries goes toward treating patients harmed during care. The OECD also attributes 15% of hospital activity and expenditure directly to patient harm. The same 15% figure appears on the WHO Patient safety facts page for total hospital activity and expenditure in OECD countries.
Improvement can produce savings. Focused safety improvements in U.S. Medicare hospitals produced an estimated US$28 billion in savings between 2010 and 2015, according to the WHO Patient safety facts page. AHRQ separately estimated about $28.2 billion in cumulative savings from reductions in hospital-acquired conditions from 2011 through 2015.
Healthcare-associated infections
Healthcare-associated infections (HAIs) are a major patient-safety measure because they can affect patients during hospitalization or other healthcare exposure. At any given time, 7 out of every 100 hospitalized patients in high-income countries acquire an HAI, compared with 10 out of 100 in low- and middle-income countries, according to the WHO Patient safety facts page.
In the European Union, around 3.2 million patients are infected with HAIs each year, and 37,000 die as a direct consequence. WHO reports that simple infection-control measures, including hand hygiene, could reduce HAIs by more than 50%.
The U.S. CDC estimates that about 1 in 31 U.S. hospital patients has at least one HAI on any given day. In nursing homes, about 1 in 43 residents contracts at least one infection associated with healthcare each day. More than 38,000 active hospitals, long-term care facilities, and other healthcare facilities provide data to the CDC’s National Healthcare Safety Network (NHSN).
The 2024 CDC HAI Progress Report covers acute care hospitals, critical access hospitals, inpatient rehabilitation facilities, and long-term acute care hospitals. Its national comparison with 2023 showed declines in most reported categories:
| 2024 measure compared with 2023 | National change |
|---|---|
| Central-line-associated bloodstream infections (CLABSI) | 9% decrease |
| Catheter-associated urinary tract infections (CAUTI) | 10% decrease |
| Ventilator-associated events (VAE) | 2% decrease |
| Colon-surgery surgical-site infections (SSI) | 4% decrease |
| Hospital-onset MRSA bacteremia | 7% decrease |
| Hospital-onset C. difficile infections | 11% decrease |
| Abdominal-hysterectomy SSI | 8% increase |
The CDC report also found that 17 states performed better on at least two infection types in 2024 than in 2023, 6 states performed better on three, and 4 states performed better on four or more. Compared with the 2015 baseline, 50 states performed better on at least two infection types, 23 performed better on five or more, and 3 states performed worse on two or more.
Medication safety and reported events
The 2024 AHRQ National Patient Safety Database (NPSD) Medication Chartbook includes 486,543 medication or other substance events submitted through December 31, 2023. Of these, 164,272 had complete description data for the substance event. Among the complete-description events, 91.9% were classified as “Incorrect action,” and 71.7% of incorrect-action events were incidents that reached the patient.
Among events with known substance type, 75% involved medications, representing 130,434 events. Other reported categories included 2,470 biological products, 1,275 nutritional products, 579 contrast-media events, 204 medical-gas events, and 61 events involving expressed human breast milk.
The administering stage was the origin for 29,537 of 81,818 events with a known stage—about 36%—while 22,133, or about 27%, originated in prescribing. Across medication or other substance events with stage and concern reported, 64.0% were incidents.
Incorrect doses represented 24.0% of known incorrect-action incident types. Among 26,834 incorrect-dose incidents with description known, 33.0% were missed or omitted doses, 32.5% were overdoses, 18.3% were underdoses, and 9.3% were extra doses. Overdoses had the highest harm rate among these subtypes at 8.6%.
Of 61,328 reported incorrect-action incidents with known harm status, 94.5% indicated no harm and 5.5% indicated harm. Harm rates varied by stage: monitoring-stage incidents had the highest rate at 8.3%, administering-stage incidents had a 7.0% rate, and storing-stage incidents had the lowest at 2.1%.
The chartbook reports human factors as a contributing factor in 10.0% of 92,394 incidents with contributing factors reported; communication was reported in 4.7%, and staff qualifications in 2.4%. Of 59,268 incidents with preventability reported, 91.9% were considered likely preventable.
Near misses provide another view of the medication-safety process. The Near Miss section reports 35,302 incorrect-action near misses with stage of origin reported. Of these, 9,902, or 28.0%, originated in prescribing or ordering. Incorrect patient or family action accounted for 9,777 near misses; 3,747 of those, or 38.3%, originated in prescribing or ordering, while 726, or 7.4%, originated in administering.
Diagnostic, surgical, and imaging safety
Diagnostic error is another significant patient-safety concern. The WHO Patient safety facts page reports that at least 5% of adults in the United States experience a diagnostic error each year in outpatient settings. Diagnostic errors contribute to approximately 10% of patient deaths in the United States and account for 6% to 17% of adverse events in hospitals.
Surgical care also carries measurable risks. At least 7 million people each year experience disabling surgical complications, and more than 1 million people die from surgical complications, according to the WHO Patient safety facts page.
The scale of imaging makes radiation safety relevant to healthcare systems. More than 3.6 billion X-ray examinations are performed worldwide each year; around 10% are performed in children. Worldwide, there are also more than 37 million nuclear medicine procedures and 7.5 million radiotherapy procedures each year. These are procedure volumes, not estimates of harm.
The OECD reports wide variation in a specific postoperative safety measure: patient-safety incidents after hip and knee surgery ranged from 57 pulmonary embolism or deep-vein-thrombosis cases per 100,000 surgical discharges in Italy to 1,192 in Australia. The figures are country-specific rates and should not be treated as a global average.
What hospital trend data show
Longer-term U.S. hospital data illustrate how safety measures can change over time. AHRQ estimated 21% fewer hospital-acquired conditions in 2015 than in 2010, with an estimated 115 HACs per 1,000 discharges in 2015. AHRQ also estimated nearly 125,000 fewer in-hospital deaths because of HAC reductions from 2010 to 2015 and more than 3.1 million fewer HACs over 2011–2015 relative to a flat 2010 rate.
Among frequent HACs, the AHRQ chartbook reported that pressure-ulcer rates fell from 40.3 per 1,000 discharges in 2010 to 30.9 per 1,000 discharges in 2014. A separate AHRQ patient-safety chartbook update reported pressure ulcers and pressure-ulcer-related events as 31.6% of HACs, at a rate of 36.3 per 1,000 hospital discharges in the cited update.
Measurement can also shape incentives. Under the CMS Hospital-Acquired Condition Reduction Program, hospitals above the 75th percentile of Total HAC Scores receive a 1% payment reduction. CMS says the program applies to subsection (d) hospitals, which are general acute care hospitals.
Taken together, these statistics show several distinct dimensions of patient safety: the frequency of harm, preventability, infection and medication events, diagnostic and surgical risk, staff communication, and changes measured against a baseline. Comparing any figure requires keeping its source, geography, population, definition, and measurement period attached to the number.