Publication/Presentation Date
7-25-2014
Abstract
Abstract:
Lehigh Valley Health Network’s orthopedic surgery rating on Healthgrades, the online hospital rating site, is only one star for both hip and total knee replacements. This does not accurately reflect the healthcare services provided. Healthgrades calculates their ratings with clinical outcomes data obtained from Medicare, focusing on mortality and complication rates. An in depth analysis of complications obtained as a result of hip and total knee replacements indicated higher than average rates of acute renal failure, 584.9, and complications affecting other body systems, 997.**, most notably other respiratory problems, 997.39. The instances of acute renal failure are decreasing from 2010 to 2014 with a standardized change in the diagnosis related group. While the trend of all 997.** complications show a high spike in 2012 and 2013. These complications were miscoded by outside contractors and a screening and educating initiative has been put in place to stop the incorrect coding. Each year, orthopedic surgeries’ Healthgrades will improve as the instances of complications continue to decrease.
Background:
With many competing hospital facilities in the same area, patients now have options of where to obtain their healthcare. The customers turn to online ratings to aid in their choice of facility. Healthgrades is an independent online healthcare rating system that assesses patient safety, clinical quality, patient experience, readmission rates, and timely and effective care for hospitals, physicians, dentists, and nursing homes based exclusively on clinical outcomes. Potential customers can compare these rankings to determine which healthcare facility will provide the best care for their medical needs. Healthgrades uses a star based rating system with one star, indicating that a hospital or physician performed worse than expected in an area, to five stars, meaning they performed better than expected.1 In general, Lehigh Valley Health Network shows average to high rankings for its services, except for orthopedic surgeries, which only have a one star rating for both hip and total knee replacements.2 By improving the Networks orthopedic Healthgrade rating, it will attract more patients and increase the Networks competitive edge against the several other orthopedic specialty hospitals in the Lehigh Valley area.
In order to improve the orthopedic rating, it is necessary to gain an understanding of how these ratings are determined. To define accurate ratings across diverse hospitals and physicians, Healthgrades obtains public data that is representative of the entire population from Medicare claims records. Healthgrades then looks at in-hospital complications and risk-adjusted mortality rates over a three year period using information from approximately 40 million patients discharges.1 Healthgrades estimates a predicted value for mortality and complications based on risk factors such as age, gender, current health condition, etc.3 Since Medicare waits two years before releasing this data for use, the 2014 rankings are determined by clinical outcomes from 2010 to 2012.1
Complications postoperatively, either a direct result of surgery or hospital stay, are common in joint replacement patients. The average complication rate following hip and total knee replacement surgeries was 3.4% nationwide in 2013, calculated from Medicare claims data.4 For all total joint replacements, serious complications occur in less than two percent of patients.5 An individual hospital’s complication rates can be skewed by many factors outside of the operating room. A study conducted at Northwestern University proved that a hospital with a respected reputation could have high complications rates that are not indicative of poor quality, but rather inconsistent record keeping.6 An in depth examination of hip and total knee replacement procedures will help to uncover the reasons behind orthopedic surgeries Healthgrades rating.
Methodology:
Healthgrades does not publicize their algorithm for determining their rankings, but the Medicare data they use is available to the public. To accurately determine the weakest area in orthopedics that greatly affects the rating, first, information was obtained from the Lehigh Valley Health Network databases on all cases from 2010 to 2014 that involved either a total knee replacement or a hip replacement. The patients’ age, gender, diagnosis, procedure, diagnosis related group (DRG), and complications, either present on admission or obtained during their hospital stay, were included for each patient analyzed. There were 3011 total knee replacement patients and 1883 hip replacement patients. Only patients who underwent elective surgeries and did not have a preexisting fracture were included in the analysis, since Healthgrades also discards these patients from its algorithm as high risk for complications and mortality.1
Excel was used to analyze and uncover trends in this data based on date, DRG, and complications. Patients with the most common complications were further analyzed by looking at their medical records, including history and physical, lab test results, and physician’s notes. After initial analysis, more research into coding procedures was required to fully grasp areas for improvement. Hip and total knee replacement patients’ data from 2013 and 2014 were then analyzed to confirm trends in the found in earlier years and get a picture of what the Healthgrades rating will be in the furture.
Results:
After Microsoft Excel analysis of hip and total knee replacement data from Lehigh Valley Health Network patients from 2010 to 2014, the following results were obtained.
Table 1: Rates of Complication Following Hip and Total Knee Replacement
2010
2011
2012
2013
2014
Hip
8.8%
5.5%
8.2%
7.1%
3.9%
Knee
10.6%
7.6%
8.5%
8.0%
2.8%
Table 1 shows the total complication rates for hip and total knee replacement surgery at Lehigh Valley Health Network from 2010 to 2014.
Table 2: Complication Descriptions
Coding Number
Complication
998.59
Other Postoperative Infection
998.32
Disruption of External Operation
998.11
Hemorrhage
997.5
Urinary
997.49
Other Digestive System
997.4
Digestive System
997.39
Other Respiratory
997.1
Cardiac
997.09
Other Nervous System
997.02
Latogenic Cerebrovascular Infarction or Hemorrhage
996.42
Dislocation of Prosthetic Joint
867.0
Bladder or Urethra w/o Open Wound into Cavity
785.59
Shock w/o Trauma
584.9
Acute Kidney Failure, Unspecified
584.5
Acute Kidney Failure, Tubular Necrosis
518.81
Acute Respiratory Failure
507.0
Aspiration Pneumonia, Food or Vomit
486.0
Pneumonia, Organism Nonspecific
482.9
Bacterial Pneumonia, Unspecified
453.42
Acute DVT or Embolism, Distal Lower Extremity
453.41
Acute DVT or Embolism, Proximal Lower Extremity
453.4
Acute DVT or Embolism, Lower Extremity
428.33
Acute on Chronic Diastolic Heart Failure
428.31
Acute Diastolic Heart Failure
427.5
Cardiac Arrest
415.11
Latrogenic Pulmonary Embolism and Infarction
410.71
Subendocardial Infarction, Initial
410.51
Acute Myocardial Infarction Lateral, Initial
410.01
Acute Myocardial Infarction anterolateral, Initial
293.0
Delirium, Other Conditions
292.81
Drug-Induced Delirium
292.12
Drug-Induced Psychotic Disorder, Hallucinations
038.9
Unspecified Septicemia
038.43
Septicemia, Pseudomonas
038.42
Septicemia, Escherichia Coli
Table 2 describes each complication based on its code. It includes all the complications in the hip and total knee replacement data obtained from 2010 to 2014 that patients developed after admission.7
Figure 1: Distribution of Complications from Total Knee Replacements
Figure 1 shows the distribution of all complications from total knee replacement surgery from 2010 to 2012. Acute renal failure, 584.9, accounts for 39 percent of all complications and other respiratory problems, 997.39, accounts for 13 percent. Together, these two complications make up over half of all coded complications during this time period.
Table 3: 584.9 and 997.** Complication Trends for Total Knee Replacement
Complication
2010
2011
2012
2013
2014
584.9
4.9%
3.1%
2.3%
1.7%
1.3%
997.**
2.7%
0.7%
3.8%
4.7%
0.2%
997.39
0.3%
0.2%
3.0%
4.2%
0.2%
Table 3 shows the percent of total knee replacement procedures that resulted in acute renal failure, complications affecting a specified body system, and other respiratory complications by year from 2010 to 2014. The instances of acute renal failures indicate a declining trend, while the instances of complications affecting a specified body system, particularly other respiratory complications, show a high spike in 2012 and 2013.
Figure 2: Distribution of Complications from Hip Replacements
Figure 2 shows the distribution of all complications from hip replacement surgery from 2010 to 2012. Acute renal failure, 584.9, accounts for 38 percent of all complications and other respiratory problems, 997.39, accounts for 17 percent. Together, these two complications make up over half of all coded complications during this time period.
Table 4: 584.9 and 997.39 Complication Trends for Hip Replacement
Complication
2010
2011
2012
2013
2014
584.9
4.4%
2.7%
1.5%
0.9%
1.0%
997.**
1.8%
1.5%
4.2%
4.9%
0.5%
997.39
0.0%
0.6%
3.3%
4.3%
0.0%
Table 4 shows the percent of hip replacement procedures that resulted in acute renal failure, complications affecting a specified body system, and other respiratory complications by year from 2010 to 2014. The instances of acute renal failures indicate a declining trend, while the instances of complications affecting a specified body system, particularly other respiratory complications, show a high spike in 2012 and 2013.
Conclusion:
An in depth investigation of the medical records of 28 patients from both the hip and total knee replacement populations who suffered from either 584.9 or 997.39, the most frequent complications shown in Figures 1 and 2, did not yield significant results. There was not a standard way of measuring the change in creatinine or BUN levels to warrant coding acute renal failure versus a different complication. The same unstandardized and confusing method of diagnosing a respiratory complication was also evident.
The decreasing trend of acute renal failure shown in Table 3 and 4, unspecified code is explained by the a change to CMS, Centers for Medicare and Medicaid Services, coding standards in October 2010 from 584.9 being classified as an MCC to a CC. An MCC is a major complication or comorbidity of surgery, while a CC is just a complication or comorbidity of surgery. This means that coding a 584.9 no longer increases the payment hospitals receive from insurance agencies, like Medicare, thus explaining the steady decrease in frequency of this code occurring. The decreasing trend continues in 2013 and 2014 with only around one percent of all hip and total knee replacement patients experiencing renal failure postoperatively.
An exploration of coding procedures and standards over this date range indicated many misdiagnosed complications by outside contractors used in instances of backlogs, including 997.1, 997.39, 997.4, 997.49, and 997.5 codes. 997.** is not normally coded as a complication directly received from surgery, but rather just a conditions that arose post-operatively, unless it is specifically documented by the physician in the medical record. These complications make up 33 percent of total complications for hip replacements from 2010 to 2012 and 27 percent for total knee replacements. When looking at discharges from 2014 so far, only three cases have been assigned any of the 997.** codes. A reeducation process for all coding staff about the use of 997.** codes is underway. A prebilling initiative is being set up to catch any of these undesirable codes for orthopedic surgeries as well as acting as a second reminder for the coding staff not to use these codes. Additionally, no outside contractors have coded any of the cases in 2014 so far this year.
These instances of inconsistent or incorrect coding are common in hospitals as proved earlier from the Northwestern University study. In a few years, once all the poor complication rates are eliminated from Healthgrades’ calculation, Lehigh Valley Health Network will have an orthopedic rating that accurately reflects the medical care provided. The rate of complications for total knee replacement in 2014 is below the national average from 2013 and for hip replacement it is only 0.5% above the national average. In future years, it would be beneficial to conduct the same analysis of complication rates to prevent more penalizing trends to continue for multiple years.
Works Cited:
1. Rating Methodology. (2014). Retrieved June 2014, from healthgrades: https://www.cpmhealthgrades.com/index.cfm/solutions/products-services/quality-achievements-plus/ratings-methodology/
2. Lehigh Valley Hospital - Cedar Crest. (2014). Retrieved June 2014, from healthgrades: http://www.healthgrades.com/hospital-directory/pennsylvania-pa/lehigh-valley-hospital-cedar-crest-hgstd26ae6a6390133?#Ratings
3. Mortality and Complications Outcomes 2014 Methodology. (2014). Retrieved from healthgrades: http://www.healthgrades.com
4. Rau, J. (2013, December 17). Medicare Identifies 97 Best And 95 Worst Hospitals For Hip And Knee Replacements . Retrieved July 2014, from Kaiser Health News: http://www.kaiserhealthnews.org/stories/2013/december/17/medicare-best-and-worst-hospitals-for-hip-and-knee-surgery.aspx
5. OrthoInfo. (2011, December). Retrieved June 2014, from AAOS: Association of American Orhtopedic Surgeons.
6. Avril, T. (2014, January 26). Debate over readmission data after joint replacements. Retrieved July 2014, from philly.com: http://articles.philly.com/2014-01-26/news/46641290_1_readmissions-medicare-eight-complications
7. Appendix C Complications or Comorbidities Exclusion list. (n.d.). Retrieved July 2014, from CMS: Center for Medicare and Medicaid Services: http://www.cms.gov/icd10manual/fullcode_cms /p0031.html
Published In/Presented At
McCrossin, d., (2014, July, 25) Improving Healthgrades Ranking for Orthopedic Suregeries. Poster presented at LVHN Research Scholar Program Poster Session, Lehigh Valley Health Network, Allentown, PA.
Department(s)
Department of Surgery, Research Scholars, Research Scholars - Posters
Document Type
Poster
Comments
Mentor: Brian Leader, Cathyann Feher