Not everyone is happy, but Kentucky doing better than the rest!
Yesterday, the Centers for Medicare & Medicaid Services (CMS) released its first iteration of a new and simplified method of presenting the results of the awkwardly named HCAHPS survey that is administered to patients after their discharge. The survey is intended to capture the patients’ own perceptions of their hospital experience and it is now part of the larger CMS Hospital-Compare initiative. As with previously released quality and safety initiatives, the reaction of individual hospitals is mixed.
The survey of 32 questions is administered to randomly selected adult patients who are between 48 hours to 6 weeks after being discharged alive from acute-care Medicare hospitals. Unlike most other quality and safety programs managed by CMS, individual patient participants are not limited to the Medicare or Medicaid programs. Hospitals attempt to reach a target of 300 surveys per year by telephone or mail, but some smaller hospitals struggle to do this. Additionally, hospitals that are exempt from having to submit the more objective Medicare process and outcome measurements (including Medicare’s Critical Access Hospitals) can participate voluntarily in this patient experience program.
Hospitals have been submitting this information voluntarily since 2006, it has been mandatory since 2007, and the results have been reported publically since 2008. A similar program for nursing homes and a few other healthcare providers already exists. Future plans are to include these patient experience scores with the other more blood-and-guts process and outcome measures to provide a single global star-score for all Medicare hospitals. Proprietary quality and safety organizations are already doing this.
What’s it all about?
Twenty-one of the questions in the survey relate to “substantive” issues in the following 11 domains:
- HCAHPS Composites Measures
- Communication with Nurses (Q1, Q2, Q3)
- Communication with Doctors (Q5, Q6, Q7)
- Responsiveness of Hospital Staff (Q4, Q11)
- Pain Management (Q13, Q14)
- Communication about Medicines (Q16, Q17)
- Discharge Information (Q19, Q20)
- Care Transition (Q23, Q24, Q25)
- HCAHPS Individual Items
- Cleanliness of Hospital Environment (Q8)
- Quietness of Hospital Environment (Q9)
- HCAHPS Global Items
- Overall Hospital Rating (Q21)
- Recommend the Hospital (Q22)
The raw data is sent to CMS for processing. An adjustment is made for patient-mix in an effort to make the wide variety of participating hospitals more comparable. [Any such process can, of course, introduce artifacts of its own.] From 1 to 5 stars are awarded for each of the 11 separate domains above. Somehow the 11 separate stars are merged into a single global 1- to 5-star rating which is what most reports will focus on, including this initial one! CMS offers a “Frequently Asked Questionr” document that goes into more detail about the methods.
As is our custom in these pages, I extracted the ratings for all of Kentucky’s hospitals from the public report. Below are the aggregate global rating scores along with the percent of Kentucky hospitals receiving that score. The results of all U.S. hospitals are included for comparison. Links to Excel files for both Kentucky and the U.S. are provided at the bottom so you can manipulate the data yourself.
|HCAHPS Grades for U.S. Hospitals, April 2015|
|Kentucky||All U.S. Hospitals|
|Star-Rating||Number Hospitals||% Hospitals||Number Hospitals||% Hospitals|
The median national and Kentucky score was 3-stars, with a skew towards the higher scores in both. It is obvious that scores were not assigned to evaluable hospitals by quintiles as a score of 5-Stars was received by only 5.4% of all hospitals. Very few hospitals nationally, and none in Kentucky received the lowest score of 1-Star.
Note that almost one quarter of US hospitals did not receive a score at all. Some of those eligible chose not to participate voluntarily. Some of them could not collect the minimum 100 surveys considered necessary for statistical validity. I will look into the reasons for absent scores in further detail at another time. It is fair to ask why a given hospital is not represented!
No hospital or hospital system in Louisville received a score of 5-Stars, but neither did we see a 1-Star. Some hospital systems have chosen to organize themselves for Medicare purposes under a single provider number, often to capture the advantages of enhanced payments attached to being a teaching hospital. Thus, although we have 8 acute-care hospitals in the city, only four scores are available to inform us. I maintain that not all hospitals within a chain should be assumed to be equivalent. The scores are: Baptist Healthcare 4-Stars; Norton Healthcare 3-Stars; Jewish & St Mary’s Healthcare 3-Stars; and University of Louisville Hospital 2-Stars.
Not unexpectedly, hospitals that got less than they thought they deserved are complaining that the things measured are too subjective, that the methodology is flawed, or in essence that these are non-professional individuals offering judgments about the quality of medical practice! I must admit, that originally I might easily have lumped myself in with some of these critics. After all, what do tasty meals, rugs on the floor, quiet hallways at night, or effective advertising have to do with the quality of science-based medical care? I do however believe the multitude of anonymous on-line patient evaluations of doctors are of little or no value and serve mostly to attract viewers to the associated advertisements that fill both electronic and paper pages.
I have written extensively about both the value and the shortcomings of the numerous programs out there that accredit and attempt to evaluate the quality and safety of hospitals and medical professionals. It is already a major proprietary industry! Nonetheless, and despite more than average interest, I myself have trouble knowing what to do with the numbers and scores offered. I am annoyed at the selective ways some hospitals promote the good scores that always seem to be available and ignore or say little about the bad ones. I do think it is possible to over-simplify one of the most complex of human undertakings into a single number or icon. I assume that the more granular patient evaluation will remain available on the Hospital Compare website. I strongly believe that as much information as possible must be made to the public and its advocates to allow addressing specific questions. How to achieve a useful balance between too much and too little information remains a major challenge.
In the end, I do think such structured evaluations by patients have value. The elements asked about are intrinsically important to the provision of even basic professional medical care. Nothing bothers me more than hearing a person tell me that they thought they had a good doctor but one with poor bedside manner or who could not communicate with them effectively. In my view as a lifelong medical educator, there is no such thing as a good doctor who is not responsive to their patient. Patients sent home with insufficient guidance of what they should do when they get there are at high risk for unnecessary readmissions. What is the matter with having patients perceive that their rooms and bathrooms are clean? Would you want to go to a hospital that almost nobody would recommend? Read the questionnaire yourself and tell us why these things are not relevant.
Supporters of the program and even some critics will admit that at the very least, this and other programs stimulate hospitals and medical professionals to do a better job. It is certainly beyond dispute that we need to do so. My bottom line is that if nothing else; how a hospital, its employees and its medical professionals attend to the issues addressed in the HCAHPS survey says a great deal about their overall respect for and commitment to the patients they serve. If any hospital claims that these perceptions of how patients felt they were treated are not as important as the high level of technology that might be offered, or glowing testimonials from selected individuals, or awards received from proprietary commercial sources— that’s a hospital not to go to!
More to come.
The database released by CMS has over 200,000 lines of data, each with many elements. It is too large to handle with the usual spread-sheet programs like Excel and requires a database-type program. I intend to wrestle it into more manageable forms and will post files for the use of my readers. That will take a little more time, but I wanted to get this preliminary summary up quickly. Come back later for more. Tell me what you are interested in.
Peter Hasselbacher, MD
Emeritus Professor of Medicine, UofL
April 17, 2015
Note: The two Excel files above were trimmed to reduce all but the names. locations, and the global scores of the hospitals. If you are interested in more extensive data such as the scores for individual elements or those for an individual state, please contact me. If you use these data for reporting elsewhere, please give KHPI the credit.