Increasing AHRQ PSI-90 Rates Requires Facility Attention

CDI people, processes, and technologies must adapt to changes if a facility’s outcomes are not to be unintentionally misrepresented.

The Centers for Medicare & Medicaid Services (CMS) assesses inpatient care quality in part through its analysis of inpatient claims employing the Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicators (PSIs).

Pertinent calculations include the following:

• PSI 04: Death Rate among Surgical Inpatients with Serious Treatable Complications – affecting a facility’s reputation as reported on the CMS Hospital Compare website
• PSI 90: Patient Safety and Adverse Events Composite – affecting CMS’s Inpatient Quality Reporting score contributing to a potential 1-percent traditional Medicare inpatient revenue reduction, and comprising the following PSIs, with varying weighted impacts:
  • PSI 03 Pressure Ulcer Rate
  • PSI 06 Iatrogenic Pneumothorax Rate
  • PSI 08 In-Hospital Fall with Hip Fracture Rate
  • PSI 09 Postoperative Hemorrhage or Hematoma Rate
  • PSI 10 Postoperative Acute Kidney Injury Requiring Dialysis Rate
  • PSI 11 Postoperative Respiratory Failure Rate
  • PSI 12 Perioperative Pulmonary Embolism (PE) or Deep Vein Thrombosis (DVT) Rate
  • PSI 13 Postoperative Sepsis Rate
  • PSI 14 Postoperative Wound Dehiscence Rate
  • PSI 15 Abdominopelvic Accidental Puncture or Laceration Rate

PSIs also affect many reputational websites, such as:

• CMS Hospital Star Ratings (https://data.cms.gov/provider-data/topics/hospitals/overall-hospital-quality-star-rating/)
• Leapfrog (https://www.hospitalsafetygrade.org/media/file/Fall2021HospitalSafetyGrade_MethodologyChanges_FINAL.pdf)
• Maryland’s Quality-Based Reimbursement (QBR) Program (https://hscrc.maryland.gov/Documents/Quality_Documents/QBR/RY2023/QBR%20RY23%20FINAL%202020-12-02%20FINAL%20Final_%20For%20Web.pdf)
• Veterans Hospital Quality (https://www.data.va.gov/dataset/SAIL-FY2021-Hospital-Performance-All-Facilities/y9x8-349i)
• Healthgrades (https://www.healthgrades.com/quality/mortality-and-complications-outcomes-2021-methodology)
• Many others

Consequently, inpatient hospitals must diligently understand how AHRQ calculates each PSI and partners with their medical and coding staffs to ensure that provider documentation correctly reflects all PSI-sensitive conditions, inclusions, exclusions, and risk adjustments. Such facilities also must ensure that all inpatient coding departments compliantly apply official ICD-10-CM conventions, guidelines, and advice to this documentation, in order to report PSI-sensitive codes within the first 25 slots, as allowed by version 5010 of the X12 standards for HIPAA transactions. 

AHRQ published its updated criteria for 2021 fiscal-year (FY) PSI calculations in July 2021 – which, for certain PSIs, significantly increases their frequency, impact, or risk adjustment, summarized at https://www.qualityindicators.ahrq.gov/Modules/Log_Coding_Updates_PSI_v2021.aspx. Those of immediate interest include:

• COVID-19 – AHRQ amended its software so that COVID-19 cases (e.g., ICD-10-CM code U071) are excluded by default, and if one does include COVID-19 cases, there will be no risk adjustment, smoothed rates, or composites for hospital indicators. They stated that “the recommended and default choice is to exclude COVID-19 discharges,” which should be welcome news for hospitals that experience COVID-19-related hypercoagulopathies resulting in pulmonary emboli affecting PSI 12 (and surgical inpatients with pneumonia as a secondary diagnosis who die during their hospital stay, affecting PSI 04).
• PSI 04: Death Rate among Surgical Inpatients with Serious Treatable Complications – AHRQ revised the order of the PSI 04 strata hierarchy to shock, sepsis, pneumonia, GI hemorrhage, and DVT/PE, swapping the positions of GI hemorrhage and DVT/PE (whereby DVT/PE is last), and added ICD-10-CM code Z66, Do Not Resuscitate, as a significant risk adjuster, weighted even higher than metastatic cancer. AHRQ worded it as “lifesaving intervention goes against patient’s documented preference;” thus, while not explicitly stated, it appears that Z66 does not have to be present on admission to qualify.
• PSI 08: In-Hospital Fall with Hip Fracture Rate – Removed the requirement that a patient who suffered a hip fracture not present on admission undergo a hip procedure to qualify for the PSI. They also moved most of the exclusions for seizure disorders, syncope, stroke, coma, cardiac arrest, poisonings, trauma, delirium, anoxic brain injury, metastatic cancer, lymphoma, and other bone diseases, leaving only peri-prosthetic hip fracture as an exclusion, significantly increasing PSI 08’s frequency in facilities where post-admission hip fractures occur. The only way to inhibit this increase in PSI 08 is to prevent these in-hospital fractures from occurring in the first place.
• PSI 11: Postoperative Respiratory Failure Rate – Removed the exclusion for Major Diagnostic Category 5, Diseases & Disorders of the Circulatory System, which will likely increase the frequency of this PSI, since many physicians document “postoperative respiratory failure” in intubated or ventilated patients after cardiothoracic surgery. In addition, AHRQ increased PSI 11’s weight in the PSI-90 composite from 16.8 to 24.0 percent, increasing its impact. Careful attention to ICD-10-CM coding conventions, guidelines, and advice and documentation requirements for ICD-10-CM codes J95.82- (a PSI 11 trigger), J96-, J95.1, and J95.2 (all three of which are not PSI 11 triggers) by coding and CDI teams in all surgical inpatients thus increases in importance.
• PSI 14: Postoperative Wound Dehiscence Rate – Removed the exclusion for AHRQ’s definition of the immunocompromised state, such as with end-stage renal disease, severe malnutrition, and HIV disease.
• PSI 15: Abdominopelvic Accidental Puncture or Laceration Rate – While not directly mentioned by AHRQ, Coding Clinic’s advice in its edition published for the second quarter of 2021, requiring the reporting of an accidental laceration complication code after a serosal tear resulting in an unplanned bowel excision, will likely increase this PSI rate.
• Totally Revised 2021 AHRQ Elixhauser Risk Model Implementation – AHRQ uses the Elixhauser risk model, available at https://www.hcup-us.ahrq.gov/toolssoftware/comorbidityicd10/comorbidity_icd10.jsp, to risk-adjust the likelihood that a PSI will or will not occur. In 2021, AHRQ amended this model by removing the highly weighted electrolyte disturbance cohort and requiring certain “semi-chronic” conditions to be present on admission to qualify, such as heart failure, drug-induced bleeding disorders, strokes, stroke consequences, pulmonary hypertension, peptic ulcer disease with bleeding, and malnutrition.

Unlike HCCs or DRGs, the Elixhauser risk model ignores the principal diagnosis in its risk calculation. Rigorously developed preoperative assessment infrastructure and workflows implemented by properly trained assessors are essential to capturing Elixhauser comorbidities and other factors affecting AHRQ PSI (and other methodology) risk adjustment.

Facilities interested in how they are currently performing with PSIs may view their and their competitors’ data at https://data.cms.gov/provider-data/dataset/ynj2-r877. Analytics groups, such as CDIMDtracker.com, IBM Watson, internal data analytics teams, and others can regroup current data sets using the FY 2020 and FY 2021 methodologies to determine what impact AHRQ’s new changes will have on a hospital’s performance.

CDI people, processes, technologies, and efforts must adapt to these significant changes if a facility’s outcomes are not to be unintentionally misrepresented.

Programming Note: Listen to Dr. Kennedy report this story live today during Talk Ten Tuesdays at 10 Eastern.