An Evaluation of Local Health Systems Michael L Taylor MD FACP CMO, Truven Health Analytics ©Truven Health Analytics Inc. All Rights Reserved. 1 Introduction and Overview All measures, except for patients satisfaction, are for the period October 2010 through September 2013 The “current time period” results are for the October 2012 through September 2013 Timeframes Patient satisfaction results are for the calendar years 2011, 2012 and 2013 The Medicare Provider Analysis and Review (MEDPAR) File Contains data from claims for services provided to all beneficiaries admitted to Medicare certified acute care hospitals Data sources ©Truven Health Analytics Inc. All Rights Reserved. Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS) A national survey of patients' perspectives of hospital care (i.e. satisfaction) 2 Introduction and Overview The strength of a hospital’s performance can be evaluated as a function of both its results during the most recent time period and the trend over time For broad measures (e.g. overall complication rate) both the trend over 12 quarters and the most current year’s results are shown All results are presented in terms of an observed to expected ratio (OE ratio) The expected amounts are based upon a national acute care hospital average, adjusted for the case mix of the hospital or hospital system being evaluated ©Truven Health Analytics Inc. All Rights Reserved. 3 Introduction and Overview Results are reported at the hospital system level for 5 systems Health System A Health System B Health System C Health System D Health System E In addition, the results are broken out by 15 hospitals within these systems that were designated by the client. ©Truven Health Analytics Inc. All Rights Reserved. 4 Introduction and Overview, Measure Sources Measure Measure Source Cost Per Case (Overall and for Truven Expected Resource Demand Medicine, General Surgery and model (ERD), the ERD is a set of regression models Orthopedics) that predict expected cost, based upon demographic factors, diagnosis and procedure factors. * Patient Safety Events Overall Rate Patient Safety Indicators (PSIs), developed by the Pressure Ulcer Rate Agency for Healthcare Research and Quality (AHRQ) Post-Op Pulmonary to capture potentially preventable adverse events that Embolism or Deep Vein compromise inpatient patient safety. Thrombosis Death among Surgical Pts With Treatable Complications *Cost is based upon the “Cost to Charge Ratio” and reflects underlying facility cost ©Truven Health Analytics Inc. All Rights Reserved. 5 Introduction and Overview, Measure Sources Measure Complication Rate (Overall and for Medicine, General Surgery and Orthopedics) Measure Source Truven Expected Complications Rate Index (ECRI) model is used to determine observed and expected complication rates using patient demographic and clinical information. Hospital Consumer Assessment of Healthcare Providers and Systems (HCAHPS), a national, Patient Satisfaction standardized, publicly reported survey of patients' perspectives of hospital care. Data are adjusted for the effects of patient-mix and mode of survey administration. ©Truven Health Analytics Inc. All Rights Reserved. 6 Cost ©Truven Health Analytics Inc. All Rights Reserved. 7 Overall Cost per Admission Average Cost per Admission Actual to Expected Ratio Health System A + Health System B Health System C Health System D Health System E Trend for Actual to Expected Results Over 12 Quarters (a lower slope is favorable) -0.09 -0.07 Hospital systems were close in performance and showed almost no change in performance over the 12 quarter time period. -0.05 -0.03 -0.01 0.01 0.03 System C showed the best overall cost performance and Systems E and B the weakest. 0.05 0.07 0.09 1.30 All 5 systems had costs which exceed the expected adjusted nation average amount. 1.25 1.20 ©Truven Health Analytics Inc. All Rights Reserved. 1.15 1.10 Current Actual to Expected Ratio (a lower ratio is favorable) 1.05 1.00 8 0.95 0.90 Overall Cost per Admission – By Hospital Cost per Admission On an individual hospital Actual to Expected Ratio basis, performance within -0.09 Trend for Actual to Expected Results Over 12 Quarters (a lower slope is favorable) system was widely -0.07 dispersed for both System A -0.05 and System C—with System -0.03 B showing somewhat -0.01 greater dispersion having 0.01 both the best and worst 0.03 performing hospitals among 0.05 the 15 individual hospitals 0.07 0.09 1.40 being compared. 1.35 1.30 1.25 1.20 1.15 1.10 1.05 1.00 Current Actual to Expected Ratio (a lower ratio is favorable) Blue axis show the experience for the average of the five Health Systems ©Truven Health Analytics Inc. All Rights Reserved. 9 0.95 0.90 Cost, Observations - Conclusions Hospital system performance did not vary widely and did not change substantially over time System C and System A were the best performing hospital systems E and B the weakest Service line and overall cost performance were closely related Hospital systems that were good cost performers overall tended to be good performers in 2 of the 3 individual service lines examined System C showed strong cost performance overall and in all 3 lines of service ©Truven Health Analytics Inc. All Rights Reserved. 10 Quality ©Truven Health Analytics Inc. All Rights Reserved. 11 Quality Overview Quality Domains Patient safety events Complications of care The overall performance for the patient safety event rate and complications was reported by hospital system, individual hospital and service line Performance for the 3 specific patient safety indicators (PSIs) listed below were reported by hospital system and hospital Pressure Ulcers Death among surgical patients with serious treatable complications Postoperative pulmonary emboli or deep vein thromboses (i.e. clots) ©Truven Health Analytics Inc. All Rights Reserved. 12 Quality, Complications ©Truven Health Analytics Inc. All Rights Reserved. 13 Overall Complication Rate Overall Complication Rate Systems B and D were Actual to Expected Ratio Health System A + the strongest hospitals in Health System B Health System C Health System D Health System E this quality domain and Trend for Actual to Expected Results Over 12 Quarters (a lower slope is favorable) -0.09 the only 2 hospital -0.07 systems to have an -0.05 overall complication rate -0.03 lower than expected. -0.01 System E was the 0.01 weakest performer by far 0.03 with a complication rate 0.05 almost 20% higher than 0.07 0.09 1.30 expected. 1.25 1.20 1.15 1.10 1.05 1.00 Current Actual to Expected Ratio (a lower ratio is favorable) ©Truven Health Analytics Inc. All Rights Reserved. 14 0.95 0.90 Complications, General Surgery – By System Complications General Surgery 2013 Results 1.40 1.24 1.20 1.09 1.16 1.03 1.00 0.84 0.80 0.60 0.40 0.20 0.00 Health System A Health System B Health System C Health System D Health System E While there was a strong correlation between service line performance and overall performance, this was particularly pronounced for General Surgery. As was the case across all service lines, B and D systems were the top performers. ©Truven Health Analytics Inc. All Rights Reserved. 15 Complications, Medicine – By System Complications Medicine 2013 Results 1.40 1.20 1.19 1.11 1.00 0.88 0.80 0.76 0.78 Health System C Health System D 0.60 0.40 0.20 0.00 Health System A Health System B Health System E It is noteworthy that this service line was the only service line where System C, an at best middle of the pack performer overall and in other service lines, shows excellent results. ©Truven Health Analytics Inc. All Rights Reserved. 16 Complications, Orthopedics – By System Complications Orthopedics 2013 Results 1.40 1.20 1.20 1.01 1.00 0.80 1.07 0.79 0.68 0.60 0.40 0.20 0.00 Health System A Health System B Health System C Health System D Health System E System B was the leader in this quality domain for this service line. ©Truven Health Analytics Inc. All Rights Reserved. 17 Complications, Further Analyses Subsequent to the initial presentation of results Truven Health examined the relationship between cost and quality, providing some statistical analysis of the results There was a weak negative correlation between cost and quality performance (r=-.33), hospitals that performed somewhat better in the cost domain performed somewhat worse in the quality domain Cost and Quality Performance Comparison by Hospital System 1.40 1.20 1.00 0.80 Complications 0.60 Cost 0.40 0.20 0.00 Health system A ©Truven Health Analytics Inc. All Rights Reserved. Health system B Health system C 18 Health system D Health system E Quality, Patient Safety Events ©Truven Health Analytics Inc. All Rights Reserved. 19 Overall Patient Safety Event Rate Overall Pt. Safety Event Rate Actual to Expected Ratio Health System A + Health System B Health System C Health System D Health System E correlation between the -0.10 Trend for Actual to Expected Results Over 12 Quarters (a lower slope is favorable) There was a strong overall patient safety -0.08 event rate and the -0.06 complication rate -0.04 (previous section). -0.02 Systems B and D are the 0.00 0.02 frontrunners and 0.04 Systems A and E the 0.06 weakest. 0.08 0.10 1.40 1.30 1.20 1.10 1.00 0.90 0.80 Current Actual to Expected Ratio (a lower ratio is favorable) ©Truven Health Analytics Inc. All Rights Reserved. 20 0.70 0.60 Patient Safety Event Rate, General Surgery – By System Pt Safety Events General Surgery 2013 Results 1.40 1.33 1.32 1.20 1.05 1.04 1.00 0.80 0.64 0.60 0.40 0.20 0.00 Health System A Health System B Health System C Health System D Health System E There is a strong correlation between overall performance and performance in this service line as well as with the complication rate. The hospital systems that performed well for patient safety for this service line were the same ones that had low complication rates. In both cases the System D was the top performer with System B and E also showing good results. ©Truven Health Analytics Inc. All Rights Reserved. 21 Patient Safety Event Rate, Medicine – By System Pt Safety Events Medicine 2013 Results 1.20 1.06 1.13 1.08 1.00 0.80 0.73 0.60 0.42 0.40 0.20 0.00 Health System A Health System B Health System C Health System D Health System E The General Medicine and Orthopedics service lines (see next slide) appear to be System C’s opportunity to stand out. As was the case for complications, it was the best performing hospital system for the General Medicine service line whereas elsewhere its performance was not exceptional. ©Truven Health Analytics Inc. All Rights Reserved. 22 Patient Safety Event Rate, Orthopedics – By System Pt Safety Events Orthopedics 2013 Results 1.20 1.12 1.00 0.80 0.77 0.79 0.69 0.60 0.40 0.21 0.20 0.00 Health System A Health System B Health System C Health System D System C was a strong performer across all 3 service lines ©Truven Health Analytics Inc. All Rights Reserved. 23 Health System E Quality, Patient Safety Events - Specific Safety Indicators ©Truven Health Analytics Inc. All Rights Reserved. 24 Quality, Patient Safety Events - Post Op Clotting* * Postoperative Pulmonary Embolism or Deep Vein Thrombosis Post Operative Pulmonary or Deep Vein Thrombosis There was a good Actual to Expected Ratio Health System A + Health System B Health System C Health System D Health System E correlation between the 3 specific patient safety indicators and the overall patient safety event rate. Consistent with this, System D was the top performer across all 3 measures and the C system the weakest performer. All hospitals exceed expected performance levels for this measure. ©Truven Health Analytics Inc. All Rights Reserved. 25 Quality, Patient Safety Events - Pressure Ulcers Pressure Ulcer Actual to Expected Ratio Health System A + Health System B Health System C Again System D showed Health System D Health System E the best performance in this area as well, far outdistancing the other systems. They also had the greatest rate of improvement. ©Truven Health Analytics Inc. All Rights Reserved. 26 Quality, Patient Safety Events - Mortality* *Death among Surgical Inpatients w/Serious Treatable Complications Death among Surgical Inpatients w/Serious Treatable Complications Actual to Expected Ratio Health System A + Health System B Health System C Health System D Health System E In general, performance was poor in this area compared to the adjusted national average. ©Truven Health Analytics Inc. All Rights Reserved. 27 Quality, Observations - Conclusions System A was a consistently weak performer quality- wise, showing high rates of complications and patient safety events System B also showed strong results System C was very inconsistent Because this system In the General Medicine experience was broken service line it had a lower out by hospital we can than expected see that in addition to complication and patient Of particular note was the poor performance of one of this systems hospitals, it showed weak performance for patient safety events and complications overall as well as across all service lines greatly impacting the system results showing good quality safety event rates Health System A Health System B ©Truven Health Analytics Inc. All Rights Reserved. results, it also showed less inter-hospital variation in performance Overall patient mortality rate was very strong Elsewhere it showed than the much more middle or lower poorly performing performance System A. 28 Health System C Quality, Observations - Conclusions System D consistently showed excellent performance in nearly all quality areas, it was the quality leader Health System D ©Truven Health Analytics Inc. All Rights Reserved. While showing better results in more areas than the System A, System E, generally showed weak quality performance Health System E 29 Patient Satisfaction ©Truven Health Analytics Inc. All Rights Reserved. 30 Patient Satisfaction The responses were to the question, “How do you rate the hospital overall?” Responses are on a 1 to 10 scale, higher is better Overall rate is the average score The system specific, cross hospital rates are a straight, non-weighted average across the hospitals in each system The satisfaction ratings of the individual hospitals within a system were consistent with respect to the rate of improvement If the system as a whole was improving or deteriorating all of the individual hospitals tended to do likewise ©Truven Health Analytics Inc. All Rights Reserved. 31 Patient Satisfaction Health System Average Satisfaction 90 Health System A 85 Health System B 80 Health System C 78.89 78.22 76.67 75 Health System D Health System E 70 Total US 65 2011 2012 2013 The satisfaction with the hospital systems was similar, the highest and lowest scoring hospitals were within 6% of each other and all were above the national average in 2013. This was particularly true of the top performers. System A, B, and D scores were within 1 point of each other as the top performing hospitals. With the exception of System A, all systems showed improvement over the 3 year period. ©Truven Health Analytics Inc. All Rights Reserved. 32 Satisfaction Observations - Conclusions Satisfaction appears to be commodity-like at the system level To the extent that differences do exist, the performance of System D and B were consistent with their strong performance in the quality area The satisfaction ratings of the individual hospitals within a system were consistent with respect to the rate of improvement If the system as a whole was improving or deteriorating all of the individual hospitals in that system tended to do likewise ©Truven Health Analytics Inc. All Rights Reserved. 33 Overall Conclusions Cost and quality are not positively correlated Higher quality performance does not predict lower cost Quality was where hospital systems differentiated themselves Hospitals were not widely separated by cost or satisfaction System D shows consistent strength in all quality areas and was highly rated from a satisfaction stand point While not the least expensive system on a case mix adjusted basis, it is in the middle of the pack cost-wise in a very tight market ©Truven Health Analytics Inc. All Rights Reserved. 34 What’s Next Drawing conclusions from the data Identify opportunities for improvement Identify opportunities for the delivery of cost effective quality care Develop baseline metrics Identify on-going measurement strategy Coordinate of communication Work to align with benefits Remember this is not one and done this is a process ©Truven Health Analytics Inc. All Rights Reserved. 35 More than Data. Answers. ©Truven Health Analytics Inc. All Rights Reserved. 36