Abstract
Background
PSIs use inpatient administrative data to flag cases with potentially preventable adverse events (AEs) attributable to hospital care. We explored how many AEs the PSIs identified in the 30 days post-discharge.
Methods
We ran the PSI software (version 3.1a) on VA 2003–2007 administrative data for ten recently validated PSIs. Among PSI-eligible index hospitalizations not flagged with an AE, we evaluated how many AEs occurred within 1–14 and 15–30 days post-discharge using inpatient and outpatient administrative data.
Results
Considering all PSI-eligible index hospitalizations, we identified 11,141 post-discharge AEs, compared to 40,578 inpatient-flagged AEs. More than 60% of post-discharge AEs were detected within 14 days of discharge. The majority of post-discharge AEs were decubitus ulcers and postoperative pulmonary embolisms or deep vein thromboses.
Conclusions
Extending PSI algorithms to the post-discharge period may provide a more complete picture of hospital quality. Future work should use chart review to validate post-discharge PSI events.
Background
Numerous studies have shown that the negative effects of poor hospital quality of care can extend beyond discharge. The consequences of hospital adverse events (AEs) include unplanned outpatient encounters, emergency room visits or hospital readmissions.1–4 Additionally, hospital performance measures limited to AE detection during the index hospitalization may misidentify high and low performing hospitals.5 These studies highlight the need to examine post-discharge data in order to provide a more complete picture of hospital quality and safety performance.
Existing hospital quality measures include 30-day mortality and readmissions,6,7 but there are no endorsed measures to our knowledge that focus explicitly on post-discharge morbidity. The Agency for Healthcare Research and Quality (AHRQ) Patient Safety Indicators (PSIs) are standardized algorithms that use inpatient administrative data to flag cases with potentially preventable inpatient AEs attributable to hospital care.8 Although the PSIs were developed to detect AEs based on administrative data collected during the hospitalization, three previous studies adapted three of the PSIs to detect AEs in the 30 days post-discharge. These studies used post-discharge inpatient data from the state of New York and found that the number of AEs detected increased by 9–20% beyond the PSIs identified during the index hospitalization.9–11
Our study expands upon this previous work by evaluating post-discharge AE detection among a wider set of ten PSIs that was recently validated in the Veterans Health Administration (VA).12 Additionally, we explore the number of AEs detected in outpatient encounters, an area of patient safety research that has received little attention. The objective of this paper is to examine the feasibility of applying PSI criteria to detect AEs in the 30-day post-discharge period using both VA inpatient and outpatient administrative data. The VA is an ideal setting for conducting this study because of the comprehensive administrative data available for VA patients that includes inpatient and outpatient encounters, allowing linkage of episodes of care.
Methods
Data Source
We obtained Fiscal Year (FY) 2003–2007 VA inpatient data from the Patient Treatment File (PTF) (n=2,343,088 hospitalizations) and outpatient data from the Outpatient Care File (OPC) (n=3,513,455 visits).13 We linked the two datasets across years by patient identifiers (i.e., scrambled social security numbers).
AHRQ Patient Safety Indicators
AHRQ spearheaded the development of the PSIs shortly after the Institute of Medicine (IOM) report charged the country to address AEs in healthcare.14 The PSIs were specifically designed to detect hospital AEs using International Classification of Diseases, Ninth Revision, Clinical Modification (ICD-9-CM) diagnosis and procedure codes.15 Since the PSI algorithms favor specificity over sensitivity (i.e., they are set up to maximize identification of true events), they search for diagnoses and procedures coded in secondary inpatient data fields to avoid labeling an AE that was present-on-admission (POA) as a hospital safety event. (Although with increasing implementation of POA codes, the most recent software does include principal diagnoses that are not POA16).
Of the 18 hospital-level AHRQ PSIs, we analyzed ten that were both endorsed by the National Quality Forum (NQF)17 and previously validated in the VA:18–24 Decubitus Ulcer (“Ulcer”), Iatrogenic Pneumothorax (“Pneumothorax”), Central Venous Catheter-Related Bloodstream Infections (“Infections”), Postoperative Hemorrhage or Hematoma (“Hemorrhage/Hematoma”), Postoperative Physiologic and Metabolic Derangement (“Derangement”), Postoperative Respiratory Failure (“Respiratory Failure”), Postoperative Pulmonary Embolism/Deep Vein Thrombosis (“PE/DVT”), Postoperative Sepsis (“Sepsis”), Postoperative Wound Dehiscence (“Dehiscence”), and Accidental Puncture or Laceration (“Puncture/Laceration”).
Identification of Inpatient PSI Events
We applied the AHRQ PSI software version 3.1a to FY03-07 inpatient data to identify hospitalizations that met the criteria for each PSI numerator (i.e., PSI events that occur in the hospitalization) and denominator (i.e., hospitalizations eligible for the PSI) criteria. We calculated PSI observed rates for the index hospitalization according to the PSI guidelines.15
Identification of Post-discharge PSI Events
To estimate the number of PSI events that occurred in the post-discharge period, we limited our PSI denominator to hospitalizations eligible for the PSI that were not flagged for an inpatient PSI event (i.e., the PSI denominator minus the PSI numerator from the index hospitalization). We used scrambled social security numbers to link those hospitalizations eligible for post-discharge PSIs to inpatient and outpatient encounters that occurred within the 30-day post-discharge window. While 30 days is a commonly accepted timeframe for examining post-discharge effects,5,7 we also examined how many PSI events occurred within 1–14 days of discharge to determine whether the majority of AEs occurred shortly after discharge.
We developed the post-discharge AE numerator criteria based on modifications of the PSI criteria used in the index hospitalization. Specifically, our revised numerators looked for principal, not secondary, diagnosis codes in order to identify the primary reason for the inpatient or outpatient encounter. If the PSI numerator was based only on ICD-9-CM diagnoses codes, our revised PSI numerator criteria searched for the presence of these codes only in the principal diagnosis field in the post-discharge inpatient and outpatient administrative data. This rule applied to the following six PSIs: Ulcer, Pneumothorax, Infections, PE/DVT, Sepsis, and Puncture/Laceration.
If the PSI numerator also included ICD-9-CM procedure codes, our modifications were more complex. ICD-9-CM procedure codes are limited to inpatient data, whereas procedures performed in outpatient encounters are coded using the Current Procedural Terminology (CPT).25 We relied on the previous literature and clinical input to crosswalk ICD-9-CM procedure codes to appropriate CPT codes. For Dehiscence and Derangement, which include a limited number of procedure codes in the numerator criteria, we identified matching CPT codes.26 The Derangement PSI includes both diabetes and acute renal failure components; we focused on detecting evidence of post-discharge acute renal failure only. (From our previous work, we found that the diabetes component was rare and had a very low positive predictive value at 13%.)18 For Hemorrhage/Hematoma, the PSI algorithm includes 27 different procedure codes; in prior work we found that several of these codes were relatively non-specific for a postoperative bleeding event, e.g., 8604, “other incision with drainage of skin and subcutaneous tissue.”27 Thus, we were concerned that many of the cross-walked CPT codes would similarly not be suitably specific, so we did not account for procedure codes when looking for outpatient events. For Respiratory Failure, the algorithm specifies either an acute respiratory failure secondary diagnosis code or an appropriately timed intubation/ventilation procedure in reference to an index procedure. We felt we would be more likely to pick up true post-discharge events based on a principal diagnosis of acute respiratory failure; we were concerned that there would be too many clinically unrelated reasons for intubation/ventilation codes. As a result, we did not incorporate these procedure codes into our post-discharge PSI numerator (see Table 1 for the final algorithms).
Table 1.
Original Patient Safety Indicator (PSI) Numerator Specifications (Version 3.1a) and Revised PSI Numerators for Select Post-Discharge Events
| Indicator | Original PSI Numerator for Index Admission | Revised PSI Numerator for Post-Discharge Events |
|---|---|---|
| Decubitus Ulcer (DU) | Discharges with ICD-9-CM code of decubitus ulcer in any secondary diagnosis field. | Subsequent admissions OR outpatient visits with ICD-9-CM code of decubitus ulcer in any principal diagnosis field. |
| Iatrogenic Pneumothorax (IP) | Discharges with ICD-9-CM code of 512.1 in any secondary diagnosis field. | Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM code of 512.1 in any principal diagnosis field. |
| Selected Medical Infection due to Medical Care (SMI) | Discharges with selected infections defined by specific ICD-9-CM codes any secondary diagnosis field. | Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM codes for selected medical infections in any principal diagnosis field. |
| Postoperative Hemorrhage or Hematoma (PHH) | Discharges with either of the following:
|
Subsequent admissions with ICD-9-CM codes for postoperative hemorrhage or hematoma in the principal diagnosis field, and a corresponding postoperative drainage of hematoma or control of hemorrhage in any procedure code field. OR, subsequent outpatient visits with a principal diagnosis of postoperative hemorrhage or hematoma. |
| Postoperative Physiologic and Metabolic Derangement (PPMD) | Discharges with either of the following:
|
Subsequent admissions OR outpatient visits in the post-discharge period with an ICD-9-CM procedure code or CPT for dialysis among cases with acute renal failure in the index admission. |
| Postoperative Respiratory Failure (PRF) | Discharges with either of the following:
|
Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM codes for acute respiratory failure in any principal diagnosis field. |
| Postoperative Pulmonary Embolism or Deep Vein Thrombosis (PPE/DVT) | Discharges with ICD-9-CM codes for PE or DVT in any secondary diagnosis field. | Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM codes for PE or DVT in any principal diagnosis field. |
| Postoperative Sepsis (PS) | Discharges with ICD-9-CM code for sepsis in any secondary diagnosis field. | Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM codes for sepsis in any principal diagnosis field. |
| Postoperative Wound Dehiscence (PWD) | Discharges with ICD-9-CM code for reclosure of postoperative disruption of abdominal wall (54.61) in any procedure field. | Either:
|
| Accidental Puncture or Laceration (APL) | Discharges with ICD-9-CM code denoting technical difficulty (e.g., accidental cut, puncture, perforation, or laceration) in any secondary diagnosis field. | Subsequent admissions OR outpatient visits in the post-discharge period with ICD-9-CM codes for accidental puncture or laceration in any principal diagnosis field. |
NOTES:
Detailed information about each PSI, including denominator criteria, can be found in the version 3.1a PSI Guide on the AHRQ website: http://www.qualityindicators.ahrq.gov/downloads/psi/psi_guide_v31.pdf
Results
Our merged dataset included 1,196,014 unique patients with at least one hospitalization between FY03-07. There were 40,578 PSI events detected during the index hospitalization and 11,141 (27%) additional PSI events that occurred post-discharge over the five-year period. The observed rates for post-discharge PSI events ranged from 0.07 for Puncture/Laceration to 8.48 for PE/DVT per 1,000. Compared to the PSI events identified in the index hospitalization, using administrative data from post-discharge inpatient and outpatient encounters enabled us to capture from 2% (Puncture/Laceration) to 77% (Hemorrhage/Hematoma) additional events (see Table 2).
Table 2.
Rates of Inpatient PSI Events and Post-Discharge PSI Events Identified in the VA during Fiscal Year 2003–2007
| AEs Identified During the Index Admission | AEs Identified During the 30-day Post- Discharge Period |
% of Additional PSIs Detected in Post- Discharge Data |
|||||
|---|---|---|---|---|---|---|---|
| PSI | Eligible Hospitalizations (Denominator) |
PSI Events (Numerator) |
Observed Rate of Inpatient PSI Events per 1,000 |
Eligible Hospitalizations (Denominator)* |
PSI Events (Numerator) |
Observed Rate of Post- Discharge PSI Events per 1,000 |
|
| Ulcer | 999,594 | 16,399 | 16.41 | 983,195 | 3,216 | 3.27 | 20% |
| Pneumothorax | 2,203,365 | 1,402 | 0.64 | 2,201,963 | 290 | 0.13 | 21% |
| Infections | 1,695,634 | 3,349 | 1.98 | 1,692,285 | 1,324 | 0.78 | 40% |
| Hemorrhage/ Hematoma | 513,841 | 1,998 | 3.89 | 511,843 | 1,545 | 3.02 | 77% |
| Derangement | 269,130 | 578 | 2.15 | 5,406 | 28 | 5.18 | 5% |
| Respiratory Failure | 201,192 | 2,850 | 14.17 | 198,342 | 55 | 0.28 | 2% |
| PE/DVT | 512,527 | 6,080 | 11.86 | 506,447 | 4,348 | 8.59 | 72% |
| Sepsis | 101,564 | 677 | 6.67 | 100,887 | 93 | 0.92 | 14% |
| Dehiscence | 100,285 | 620 | 6.18 | 99,665 | 84 | 0.84 | 14% |
| Puncture/ Laceration | 2,341,976 | 7,203 | 3.08 | 2,334,773 | 158 | 0.07 | 2% |
NOTES:
Ulcer: PSI #3 Decubitus Ulcer.
Pneumothorax: PSI #6 Iatrogenic Pneumothorax.
Infections: PSI #7 Central Venous Catheter-related Blood Stream Infections.
Hemorrhage/Hematoma: PSI #9 Postoperative Hemorrhage or Hematoma.
Derangement: PSI #10 Postoperative Physiologic and Metabolic Derangement.
Respiratory Failure: PSI #11 Postoperative Respiratory Failure.
PE/DVT: PSI #12 Postoperative Pulmonary Embolism/Deep Vein Thrombosis.
Sepsis: PSI #13 Postoperative Sepsis.
Dehiscence: PSI #14 Postoperative Wound Dehiscence.
Puncture/Laceration: PSI #15 Accidental Puncture or Laceration.
The denominator for events detected in the post-discharge period is equal to the number of hospitalizations eligible for the original PSI less the number of PSI events detected during the index hospitalization. We have one exception to this rule: for Derangement we focused on the dialysis component of the PSI numerator and we restricted the sample of PSI-eligible hospitalizations in the discharge period to those that had acute renal failure during the index hospitalization.
We found that for all PSIs, post-discharge events were more likely to be detected within 1–14 days of discharge than within 15–30 days, e.g., 2.32 versus 0.70 per 1,000 for Hemorrhage/Hematoma, respectively, and 4.62 versus 0.55 per 1,000 for Derangement (see Table 3). There was also variation in where the PSI event was detected post-discharge. For example, PSI events such as Ulcer, Hemorrhage/Hematoma, PE/DVT, and Sepsis, were more frequently detected in outpatient encounters (3.17 vs. 0.13, 2.57 vs. 0.55, 7.68 vs. 1.81, and 0.74 vs. 0.22 per 1,000, respectively), while other events, such as Pneumothorax, Infection, Derangement, and Dehiscence, were more often found in subsequent hospitalizations (0.09 vs. 0.05, 0.51 vs. 0.29, 3.88 vs. 1.85, and 0.75 vs. 0.16 per 1,000, respectively).
Table 3.
Post-Discharge Patient Safety Indicator (PSI) Events by Timing (1–14 Days versus 15–30 Days of Discharge) and Setting (Inpatient versus Outpatient)
| PSI | Hospitalizations Eligible for Post-Discharge PSIs (n) | PSI Events Detected from 1–14 Days of Discharge n (rate per 1,000) |
PSI Events Detected from 15–30 Days of Discharge n (rate per 1,000) |
Total Post-Discharge PSI Events Detected within 30 Days of Discharge n (rate per 1,000) |
||||||
|---|---|---|---|---|---|---|---|---|---|---|
| Inpatient | Outpatient | Either | Inpatient | Outpatient | Either | Inpatient | Outpatient | Either | ||
| Ulcer | 983,195 | 62 (0.06) | 1,681 (1.71) | 1,730 (1.76) | 70 (0.07) | 1,432 (1.46) | 1,486 (1.51) | 132 (0.13) | 3,113 (3.17) | 3,216 (3.27) |
| Pneumothorax | 2,201,963 | 114 (0.05) | 66 (0.03) | 168 (0.08) | 87 (0.04) | 47 (0.02) | 122 (0.06) | 201 (0.09) | 113 (0.05) | 290 (0.13) |
| Infections | 1,692,285 | 509 (0.30) | 312 (0.18) | 812 (0.48) | 351 (0.21) | 171 (0.10) | 512 (0.30) | 860 (0.51) | 483 (0.29) | 1,324 (0.78) |
| Hemorrhage/ Hematoma | 511,843 | 207 (0.40) | 1,018 (1.99) | 1,189 (2.32) | 75 (0.15) | 295 (0.58) | 356 (0.70) | 282 (0.55) | 1,313 (2.57) | 1,545 (3.02) |
| Derangement | 5,406 | 19 (3.51) | 8 (1.48) | 25 (4.62) | 2 (0.37) | 2 (0.37) | 3 (0.55) | 21 (3.88) | 10 (1.85) | 28 (5.18) |
| Respiratory Failure | 198,342 | 23 (0.12) | 20 (0.10) | 39 (0.20) | 12 (0.06) | 5 (0.03) | 16 (0.08) | 35 (0.18) | 25 (0.13) | 55 (0.28) |
| PE/DVT | 506,447 | 569 (1.12) | 2,445 (4.83) | 2,791 (5.51) | 348 (0.69) | 1,447 (2.86) | 1,557 (3.07) | 917 (1.81) | 3,892 (7.68) | 4,348 (8.59) |
| Sepsis | 100,887 | 14 (0.14) | 49 (0.49) | 61 (0.60) | 8 (0.08) | 26 (0.26) | 32 (0.32) | 22 (0.22) | 75 (0.74) | 93 (0.92) |
| Dehiscence | 99,665 | 70 (0.70) | 12 (0.12) | 76 (0.76) | 5 (0.05) | 4 (0.04) | 8 (0.08) | 75 (0.75) | 16 (0.16) | 84 (0.84) |
| Puncture/ Laceration | 2,334,773 | 48 (0.02) | 51 (0.02) | 99 (0.04) | 23 (0.01) | 36 (0.02) | 59 (0.03) | 71 (0.03) | 87 (0.04) | 158 (0.07) |
NOTES:
Ulcer: PSI #3 Decubitus Ulcer.
Pneumothorax: PSI #6 Iatrogenic Pneumothorax.
Infections: PSI #7 Central Venous Catheter-related Blood Stream Infections.
Hemorrhage/Hematoma: PSI #9 Postoperative Hemorrhage or Hematoma.
Derangement: PSI #10 Postoperative Physiologic and Metabolic Derangement.
Respiratory Failure: PSI #11 Postoperative Respiratory Failure.
PE/DVT: PSI #12 Postoperative Pulmonary Embolism/Deep Vein Thrombosis.
Sepsis: PSI #13 Postoperative Sepsis.
Dehiscence: PSI #14 Postoperative Wound Dehiscence.
Puncture/Laceration: PSI #15 Accidental Puncture or Laceration.
Discussion
We found it was feasible to modify PSI algorithms to detect AEs in the post-discharge period. In fact, applying PSI algorithms to post-discharge data enabled us to identify nearly 30% more safety events that may be attributable to care received during the index hospitalization. These events were more likely to occur within the first 14 days of discharge. With respect to the setting where post-discharge PSIs were detected, this varied across the PSIs. With some exceptions, AEs detected in inpatient encounters likely reflect the greater severity and acuity of the complication.
Our research builds on recent evidence that some inpatient safety events may be the result of care received in the previous hospitalization.28 Earlier studies of the PSIs found 9%, 20% and 25% more cases of Hemorrhage/Hematoma, PE/DVT, and Infection, respectively, using data from post-discharge inpatient encounters.9–11 When we limited our post-discharge analysis to inpatient data, we had similar findings (14%, 15% and 25% respectively). However, by including both inpatient and outpatient encounters in our study, we found 77%, 40% and 72%, more AEs respectively for these PSIs. This highlights the importance of incorporating outpatient data into post-discharge AE detection.
Our study also expands on the literature by examining two post-discharge time periods. Although a 30-day period is commonly used as the cut-point in post-discharge hospital quality measures (e.g., readmissions), it is likely that the ideal window in which to capture post-discharge AEs varies according to the nature and severity of the complication. We found that most PSIs occurred shortly after discharge with many fewer events detected more than two weeks later. However, Ulcer and Puncture/Laceration had nearly as many events detected in the second half of the 30-day window as in the first. There is evidence that decubitus ulcers may manifest over a longer period of time,29 such that a post-discharge window beyond 30 days may be appropriate. In the case of Puncture/Laceration, one would expect this complication to manifest during or shortly after the index procedure. Thus, it is possible that the higher frequency of these events in the 15–30 day timeframe is the result of intervening episodes of care, indicating that a shorter post-discharge window may be more appropriate. Future studies may want to explore additional time periods to detect post-discharge AEs attributable to the index hospitalization.
As expected, many of the AEs that would be considered as more severe complications were more likely to be detected in post-discharge inpatient encounters (e.g., Dehiscence and Derangement); however, we found at least one exception to this rule. We were surprised to find more cases of Sepsis in the outpatient setting. (On the whole, there were relatively few cases in either setting.) It is possible that these outpatient cases were misdiagnosed, miscoded, or more likely to be diagnosed in a VA setting without inpatient acute care – in this scenario, the patient would have been sent to closest acute inpatient care facility for treatment, which in many areas is a non-VA hospital. Additionally, although many post-operative hemorrhages and hematomas may be treated in an outpatient setting, it is possible that many of the Hemorrhage/Hematoma outpatient cases may not reflect true AEs. As a result of poor CPT to ICD-9-CM procedure code matching, the algorithm we applied to outpatient data was not as specific as the original PSI algorithm. In the case of Sepsis, merging post-discharge VA with non-VA data may provide a more comprehensive and clearer picture of VA care. In the case of Hemorrhage/Hematoma, a nationally-validated ICD-9-CM and CPT crosswalk would greatly facilitate the application of PSI algorithms to outpatient data. For all PSIs, chart review would also confirm the reliability of outpatient coding and determine whether the PSI algorithms may be further adapted to detect events in outpatient data.
Our study has several strengths. We used a large sample of linked inpatient and outpatient administrative data, enabling us to track patients over time. We also developed and applied numerator criteria to test seven PSI algorithms for post-discharge safety events that had not previously been explored in the literature. Finally, we measured the extent to which problems resulting from inpatient care could potentially be detected in post-discharge outpatient encounters, an important area for future research.
This study also has a few limitations. Some of the post-discharge PSI events we detected may be false positives (previous work examining the validity of selected PSIs found that after accounting for POA, the positive predictive values ranged from 46–90%).12 Thus, we may have overestimated the yield of true events that can be detected by applying modified PSI criteria to post-discharge data. We suggest that future studies use chart review to validate the post-discharge PSIs. On the other hand, we also limited our study to VA data only despite the fact that 30% of Veterans also have Medicare coverage.30 As a result, our study may have underestimated AEs if post-discharge care occurred in Medicare-reimbursed hospital settings. Finally, we could not always crosswalk exact ICD-9-CM procedure codes with CPT codes, resulting in somewhat imperfect matches between the inpatient and post-discharge PSI algorithms; however, this would have only affected the specificity of one of the PSIs we evaluated (Hemorrhage/Hematoma).
Since the PSIs are a reasonably valid and cost-efficient means of identifying AEs,12 policymakers may consider using the PSI algorithms to measure post-discharge morbidity associated with hospital quality. Unlike 30-day readmissions, the PSIs identify specific complications associated with hospital care. A better understanding of safety events and where they are detected can help hospitals target quality improvement efforts. Additionally, the PSIs are increasingly being used for hospital profiling with respect to public reporting and pay-for-performance initiatives.17,31 If the modified post-discharge PSIs detect true patient safety events, they may increase the yield of hospital-associated AEs and provide a more accurate picture of hospital performance. Future research should evaluate the predictive validity of the post-discharge PSIs, as well as whether the inclusion of post-discharge events alters hospital profiles.
Conclusion
Although the PSIs were designed to detect AEs that specifically occurred in the inpatient setting, by modifying the PSI numerator criteria, we identified many more events that occurred post-discharge. These post-discharge PSIs may represent a unique opportunity to widen AE detection beyond the index hospitalization period, particularly because they use standardized algorithms that can be applied across diverse settings of care. We demonstrated the feasibility of using the PSIs to detect post-discharge AEs; evaluating all AEs associated with inpatient care may provide a more complete picture of hospital quality. Future work should refine the post-discharge PSI algorithms and use chart review to validate the PSI events detected in the post-discharge period.
Acknowledgments
Funding
This study was supported by Department of Veterans Affairs Health Services Research and Development Service grant number SDR 07-002.
Special thanks to: Amresh Hanchate, PhD for his advice on methodology; Susan Loveland, MAT for programming expertise and assistance with identifying the hospital sample and flagged cases.
Footnotes
Conflict of Interest Declaration
All authors do not have any current or foreseeable financial and personal conflicts of interest. Separate conflict of interest and financial disclosure form are provided to managing editor Deborah Meiris by fax.
Contributor Information
Hillary J. Mull, Email: hillary.mull@va.gov.
Ann M. Borzecki, Email: amb@bu.edu.
Qi Chen, Email: qc2112@bu.edu.
Marlena H. Shin, Email: marlena.shin@va.gov.
Amy K. Rosen, Email: akrosen@bu.edu.
References
- 1.Encinosa WE, Hellinger FJ. The impact of medical errors on ninety-day costs and outcomes: an examination of surgical patients. Health Serv Res Dec. 2008;43(6):2067–2085. doi: 10.1111/j.1475-6773.2008.00882.x. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Friedman B, Encinosa W, Jiang HJ, Mutter R. Do patient safety events increase readmissions? Med Care May. 2009;47(5):583–590. doi: 10.1097/MLR.0b013e31819434da. [DOI] [PubMed] [Google Scholar]
- 3.Forster AJ, Clark HD, Menard A, et al. Adverse events among medical patients after discharge from hospital. CMAJ. 2004 Feb 3;170(3):345–349. [PMC free article] [PubMed] [Google Scholar]
- 4.Rosen AK, Loveland S, Shin M, et al. Examining the Impact of the AHRQ Patient Safety Indicators (PSIs) on the Veterans Health Administration: The Case of Readmissions. Med Care. 2012 Sep 29; doi: 10.1097/MLR.0b013e318270c0f7. [DOI] [PubMed] [Google Scholar]
- 5.Bilimoria KY, Cohen ME, Ingraham AM, et al. Effect of postdischarge morbidity and mortality on comparisons of hospital surgical quality. Ann Surg Jul. 2010;252(1):183–190. doi: 10.1097/SLA.0b013e3181e4846e. [DOI] [PubMed] [Google Scholar]
- 6.Khuri SF, Daley J, Henderson W, et al. The Department of Veterans Affairs’ NSQIP: the first national, validated, outcome-based, risk-adjusted, and peer-controlled program for the measurement and enhancement of the quality of surgical care. National VA Surgical Quality Improvement Program. Annals of Surgery Oct. 1998;228(4):491–507. doi: 10.1097/00000658-199810000-00006. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Centers for Medicare and Medicaid Services. [Accessed October 18, 2011];Hospital Quality Initiatives Outcome Measures. https://www.cms.gov/HospitalQualityInits/20_OutcomeMeasures.asp#TopOfPage.
- 8.Agency for Healthcare Research and Quality. AHRQ Patient Safety Indicators Overview. [Accessed May 3, 2011];AHRQ Homepage. 2011 http://www.qualityindicators.ahrq.gov/Modules/psi_overview.aspx.
- 9.Gallagher B, Cen L, Hannan EL. Readmission for selected infections due to medical care: expanding the definition of a patient safety indicator. In: Henriksen K, Battles JB, Marks E, Lewin DI, editors. Advances in patient safety: from research to implementation. Rockville, MD: Agency for Healthcare Research and Quality (AHRQ) and Department of Defense (DoD); 2005. pp. 39–50. [PubMed] [Google Scholar]
- 10.Weller WE, Gallagher BK, Cen L, Hannan EL. Readmissions for venous thromboembolism: expanding the definition of patient safety indicators. Jt Comm J Qual Saf Sep. 2004;30(9):497–504. doi: 10.1016/s1549-3741(04)30058-4. [DOI] [PubMed] [Google Scholar]
- 11.Shufelt JL, Hannan EL, Gallagher BK. The postoperative hemorrhage and hematoma patient safety indicator and its risk factors. Am J Med Qual Jul-Aug. 2005;20(4):210–218. doi: 10.1177/1062860605276941. [DOI] [PubMed] [Google Scholar]
- 12.Rosen AK, Itani KM, Cevasco M, et al. Validating the patient safety indicators in the Veterans Health Administration: do they accurately identify true safety events? Med Care Jan. 2012;50(1):74–85. doi: 10.1097/MLR.0b013e3182293edf. [DOI] [PubMed] [Google Scholar]
- 13.Veterans Health Administration. [Accessed May 1, 2009];VIReC Research User Guides. 2009 http://www.virec.research.va.gov/References/RUG/RUG.htm.
- 14.Institute of Medicine. To Err is Human: Building A Safer Health System. Washington, DC: National Academy Press; 1999. [Google Scholar]
- 15.Agency for Healthcare Research and Quality. AHRQ Guide to Patient Safety Indicators Version 3.1a. Rockville, MD: AHRQ; 2007. [Google Scholar]
- 16.Geppert J, Bott J. [Accessed February 11, 2013];AHRQ Quality Indicators Webinar: AHRQ Present on Admission (POA) – User Overview. 2010 http://www.qualityindicators.ahrq.gov/Downloads/Resources/Webinars/2010/May%2012.pdf.
- 17.National Quality Forum. [Accessed July 28, 2010];NQF Endorsed Standards. 2010 < http://www.qualityforum.org/Measures_List.aspx>.
- 18.Borzecki AM, Cevasco M, Chen Q, Shin M, Itani KM, Rosen AK. How valid is the AHRQ Patient Safety Indicator “postoperative physiologic and metabolic derangement”? J Am Coll Surg Jun. 2011;212(6):968–976. e961–962. doi: 10.1016/j.jamcollsurg.2011.01.001. [DOI] [PubMed] [Google Scholar]
- 19.Borzecki AM, Kaafarani HM, Utter GH, et al. How valid is the AHRQ Patient Safety Indicator “postoperative respiratory failure”? J Am Coll Surg Jun. 2011;212(6):935–945. doi: 10.1016/j.jamcollsurg.2010.09.034. [DOI] [PubMed] [Google Scholar]
- 20.Cevasco M, Borzecki A, O’Brien W, et al. Validity of the AHRQ patient safety indicator 7: Central venous catheter related bloodstream infections. J Am Coll Surg. 2011 doi: 10.1016/j.jamcollsurg.2011.02.005. [DOI] [PubMed] [Google Scholar]
- 21.Cevasco M, Borzecki AM, Chen Q, et al. Positive predictive value of the AHRQ Patient Safety Indicator “Postoperative Sepsis”: implications for practice and policy. J Am Coll Surg Jun. 2011;212(6):954–961. doi: 10.1016/j.jamcollsurg.2010.11.013. [DOI] [PubMed] [Google Scholar]
- 22.Cevasco M, Borzecki AM, McClusky DA, 3rd, et al. Positive predictive value of the AHRQ Patient Safety Indicator “postoperative wound dehiscence”. J Am Coll Surg Jun. 2011;212(6):962–967. doi: 10.1016/j.jamcollsurg.2011.01.053. [DOI] [PubMed] [Google Scholar]
- 23.Chen Q, Rosen A, Cevasco M, Shin M, Itani K, Borzecki A. Detecting patient safety indicators: How valid is “foreign body left during procedure in the Veterans Health Administration? J Am Coll Surg. 2011 doi: 10.1016/j.jamcollsurg.2011.02.003. [DOI] [PubMed] [Google Scholar]
- 24.Kaafarani HM, Borzecki AM, Itani KM, et al. Validity of selected Patient Safety Indicators: opportunities and concerns. J Am Coll Surg Jun. 2011;212(6):924–934. doi: 10.1016/j.jamcollsurg.2010.07.007. [DOI] [PubMed] [Google Scholar]
- 25.American Medical Association (AMA) [Accessed Dec. 19, 2012];Current Procedural Terminology Homepage. 2012 http://www.ama-assn.org/ama/pub/physician-resources/solutions-managing-your-practice/coding-billing-insurance/cpt.page.
- 26.Borzecki AM, Lee A, Wang SW, Brenner L, Kazis LE. Survival in end stage renal disease: calcium carbonate vs. sevelamer. J Clin Pharm Ther Dec. 2007;32(6):617–624. doi: 10.1111/j.1365-2710.2007.00871.x. [DOI] [PubMed] [Google Scholar]
- 27.Borzecki AM, Kaafarani H, Cevasco M, et al. How valid is the AHRQ Patient Safety Indicator “postoperative hemorrhage or hematoma”? J Am Coll Surg Jun. 2011;212(6):946–953. e941–942. doi: 10.1016/j.jamcollsurg.2010.09.033. [DOI] [PubMed] [Google Scholar]
- 28.Tsilimingras D, Bates DW. Addressing postdischarge adverse events: a neglected area. Jt Comm J Qual Patient Saf Feb. 2008;34(2):85–97. doi: 10.1016/s1553-7250(08)34011-2. [DOI] [PubMed] [Google Scholar]
- 29.Lyder CH, Ayello EA. Pressure Ulcers: A Patient Safety Issue. In: Hughes RG, editor. Patient Safety and Quality: An Evidence-Based Handbook for Nurses. Agency for Healthcare Research and Quality (AHRQ); 2008. AHRQ Publication No. 08-0043. [PubMed] [Google Scholar]
- 30.Hynes DM, Koelling K, Stroupe K, et al. Veterans’ access to and use of Medicare and Veterans Affairs health care. Med Care Mar. 2007;45(3):214–223. doi: 10.1097/01.mlr.0000244657.90074.b7. [DOI] [PubMed] [Google Scholar]
- 31.Centers for Medicare and Medicaid Services (CMS) [Accessed July 28, 2010];Details for: Quality Measures for Reporting in Fiscal Year 2009 for 2010 Update. 2010 < http://www.cms.hhs.gov/apps/media/press/factsheet.asp?Counter=3044&intNumPerPage=10&checkDate=&checkKey=&srchType=1&numDays=3500&srchOpt=0&srchData=&keywordType=All&chkNewsType=6&intPage=&showAll=&pYear=&year=&desc=false&cboOrder=date>.