3.1 Patient Safety Data Sources

3.1 Patient Safety Data Sources

Measuring patient safety is a critical pre-requisite for improving care outcomes. To identify areas of vulnerability and track the effectiveness of interventions, healthcare organizations must collect and analyze both quantitative data and qualitative data. These two data types serve complementary purposes: quantitative data provides numerical counts, rates, and statistical distributions that tell organizations what is occurring and how often, whereas qualitative data captures descriptive narratives, perceptions, and contextual factors that explain why and how safety failures occur.

Quantitative vs. Qualitative Data in Patient Safety

Effective patient safety measurement requires a balanced integration of both data paradigms.

• Quantitative Data: This type of data is structured and numerical. It is easily aggregated, trended over time, and used for benchmarking across facilities. Common examples include incident report counts, clinical infection rates (such as central line-associated bloodstream infection, or CLABSI, rates), 30-day readmission rates, and medication error counts. Quantitative data allows organizations to identify statistical outliers and establish historical baselines.
• Qualitative Data: This data is unstructured and descriptive. It focuses on the human factors, workplace culture, and latent systemic issues that numbers alone cannot reveal. Common sources include Root Cause Analysis (RCA) narratives, safety culture survey comments, leadership safety WalkRounds feedback, and patient or family interviews. Qualitative data is essential for understanding the clinical and organizational context that led to a specific safety failure.

Voluntary Incident Reporting vs. Active Surveillance

A fundamental challenge in patient safety measurement is the underreporting of adverse events. Historically, hospitals have relied on voluntary incident reporting systems, where frontline staff submit electronic or paper forms when they observe a near miss or adverse event. However, research consistently demonstrates that voluntary reporting systems capture only 10% to 15% of actual adverse events. Staff are less likely to report events due to fear of retribution, lack of time, or normalization of deviance.

To overcome the limitations of voluntary reporting, organizations implement active surveillance. The most prominent active surveillance methodology is the IHI Global Trigger Tool (GTT). Developed by the Institute for Healthcare Improvement, the GTT utilizes trained reviewers to conduct retrospective reviews of a random sample of patient charts. The reviewers look for specific clinical "triggers" — such as the sudden administration of naloxone (indicating a potential opioid overdose) or an abrupt drop in hematocrit (indicating unrecognized hemorrhage). Once a trigger is identified, the chart is audited to determine if an adverse event occurred. Active surveillance via trigger tools captures up to 10 times more adverse events than voluntary reporting systems, providing a much more accurate representation of the organization's safety profile.

AHRQ Common Formats

To standardize the way healthcare organizations collect and report patient safety events, the Agency for Healthcare Research and Quality (AHRQ) developed the Common Formats. These formats provide standardized definitions, data elements, and reporting templates for a wide range of patient safety events, enabling the aggregation of data across different electronic systems.

The AHRQ Common Formats are designed for use by Patient Safety Organizations (PSOs), which were established under the Patient Safety and Quality Improvement Act of 2005 to allow providers to report and analyze safety events in a federally protected, non-punitive environment. The Common Formats cover 10 distinct event categories:

1. Medication (e.g., dosing errors, wrong-patient administration)
2. Fall (e.g., assisted or unassisted patient falls)
3. Device or Medical Technology (e.g., malfunction or misuse of a pump)
4. Surgery or Anesthesia/Procedure (e.g., wrong-site surgery)
5. Blood or Blood Product (e.g., transfusion reactions)
6. Pressure Injury (e.g., hospital-acquired pressure ulcers)
7. Infection (e.g., healthcare-associated infections)
8. Venous Thromboembolism (VTE) (e.g., deep vein thrombosis)
9. Perinatal (e.g., maternal or neonatal injury)
10. Other (e.g., patient identification errors or elopement)

By using these standardized formats, PSOs can aggregate data from thousands of hospitals to identify rare patterns, systemic risks, and national safety trends.

Patient Safety Indicators (PSI) and PSI-90

Another critical quantitative data source is administrative billing data, which is analyzed using Patient Safety Indicators (PSIs) developed by AHRQ. Unlike voluntary reports or trigger tools, PSIs are derived from discharge data using ICD-10 diagnostic codes. They identify potential in-hospital complications and adverse events that occur during a patient's stay.

The Patient Safety and Adverse Events Composite (PSI-90) is a weighted composite measure used by the Centers for Medicare & Medicaid Services (CMS) for national benchmarking, public reporting, and financial penalty programs (such as the Hospital-Acquired Condition Reduction Program). PSI-90 aggregates 10 individual indicators:

• PSI 03: Pressure Ulcer Rate
• PSI 06: Iatrogenic Pneumothorax Rate
• PSI 07: Central Venous Catheter-Related Bloodstream Infection Rate
• PSI 08: In-Hospital Fall with Hip Fracture Rate
• PSI 09: Perioperative Hemorrhage or Hematoma Rate
• PSI 10: Postoperative Acute Kidney Injury Requiring Renal Replacement Therapy Rate
• PSI 11: Postoperative Respiratory Failure Rate
• PSI 12: Perioperative Pulmonary Embolism (PE) or Deep Vein Thromboembolism (DVT) Rate
• PSI 13: Postoperative Sepsis Rate
• PSI 15: Accidental Puncture or Laceration Rate

While PSI-90 is a powerful tool for high-level monitoring, it has limitations. Because it relies on billing codes, its accuracy is highly dependent on the quality and completeness of clinical documentation. It can also penalize hospitals that are highly vigilant in documenting complications, creating a "detection bias" where safer, more thorough hospitals appear to have higher complication rates. Quality leaders must validate PSI-90 data through clinical chart audits before using it to drive quality improvement projects.