The PSIs are a set of quality measures that use hospital inpatient discharge data to provide a perspective on patient safety. Specifically, the PSIs identify problems that patients experience through contact with the health care system and that are likely amenable to prevention by implementing system level changes. The problems identified are referred to as complications or adverse events. There are currently 27 PSIs that are defined on two levels: the provider level and the area level. They are risk adjusted using a model that incorporates DRGs (with and without complications aggregated); a modified comorbidity index based on a list developed by Elixhauser and colleagues;14 and age, sex, and age-sex interactions.
At the provider level, the PSIs present a picture of patient safety within a hospital and provide information about the potentially preventable complication for patients who received their initial care and experienced the complication of care within the same hospitalization. The PSIs use secondary diagnosis ICD-9-CM codes to detect complications and adverse events. The measure set covers a variety of areas such as selected postoperative complications, selected technical adverse events, technical difficulty with procedures, and obstetric trauma and birth trauma.
The 20 provider-level PSIs include:
1. Complications of Anesthesia
2. Death in Low-Mortality DRGs
3. Decubitus Ulcer
4. Failure to Rescue
5. Foreign Body Left During Procedure
6. Iatrogenic Pneumothorax
7. Selected Infections Due to Medical Care
8. Postoperative Hip Fracture
9. Postoperative Hemorrhage or Hematoma
10. Postoperative Physiologic and Metabolic Derangements
11. Postoperative Respiratory Failure
12. Postoperative Pulmonary Embolism or Deep Vein Thrombosis
13. Postoperative Sepsis
14. Postoperative Wound Dehiscence
15. Accidental Puncture or Laceration
16. Transfusion Reaction
17. Birth Trauma – Injury to Neonate
18. Obstetric Trauma – Vaginal with Instrument
19. Obstetric Trauma – Vaginal without Instrument
20. Obstetric Trauma – Cesarean Delivery
The ability to assess all patients at risk for a particular patient safety problem, along with the relative low cost of collecting the data, are particular strengths of the datasets that use administrative data. However, many important areas of interest, such as adverse drug events, cannot currently be monitored well using administrative data and using this data source to identify patient safety events tends to favor specific types of indicators. For example, the PSIs cited in this chapter contain a large proportion of surgical indicators, rather than medical or psychiatric measures, because medical or psychiatric complications are often difficult to distinguish from comorbidities that are present on admission. In addition, medical populations tend to be more heterogeneous than surgical populations, especially elective surgical populations, making it difficult to account for case mix.
While PSIs may be more applicable to patient safety when limited to elective surgical admissions, the careful use of administrative data holds promise to identify problems for further analysis and study. The limitations of this measure set include those inherent with the use of administrative data, clinical accuracy of the discharged-based diagnosis coding, and indicator discriminatory power.
Administrative data are unlikely to capture all cases of a complication, regardless of the preventability, without false positives and false negatives (sensitivity and specificity).
When the codes are accurate in defining an event, the clinical vagueness inherent in the description of the code itself (e.g., hypotension) may lead to a highly heterogeneous pool of clinical states represented by that code.
Incomplete reporting is an issue in the accuracy of any data source used for identifying patient safety problems, as medical providers might fear adverse consequences as a result of full disclosure in potentially public records such as discharge abstracts.
The heterogeneity of clinical conditions included in some codes, lack of information about event timing available in these datasets, and limited clinical detail for risk adjustment all contribute to the difficulty in identifying complications that represent medical error or that may be at least in some part preventable. These factors may exist for other sources of patient safety data as well. For example, they have been raised in the context of the Joint Commission’s implementation of a sentinel event program geared to identifying serious adverse events that may be related to underlying safety problems.
Yet, despite these issues, the PSIs are a useful tool to identify areas in patient safety that need monitoring and/or intervention for improved patient care.