CMS’ Star Ratings Miscalculated

Modern Healthcare is reporting that the Centers for Medicare and Medicaid Services (CMS) has miscalculated hospitals star ratings since they were first released in 2016 (1). Officials at Rush University Medical Center in Chicago exclusively disclosed their analysis and correspondence to Modern Healthcare. The investigators found that instead of evenly weighting the eight measures in the safety of care group, the CMS' star ratings formula relied heavily on one measure— The Patient Safety and Adverse Events Composite, known as PSI 90 —for the first four releases of the ratings and then complication rates from hip and knee replacements for the latest release. The single measure accounted for about 98% of a hospital's performance in the safety group, according to Rush's analysis. The safety group can also greatly influence a hospital's overall star rating, the analysis concluded. Rush's findings likely prompted the CMS to announce this week that it would postpone the July release of its star ratings (1).

The statistical model the CMS uses likely caused the miscalculation. The model, called latent variable modeling, uses scores for seven groups of measures to calculate the star ratings:

1. Mortality
2. Safety of Care
3. Readmission
4. Patient Experience
5. Effectiveness of Care
6. Timeliness of Care
7. Efficient Use of Medical Imaging

The three outcome groups—mortality, safety and readmissions—are each weighted the most at 22% each. Measures within each group are supposed to be evenly weighted to calculate the hospital's performance in that area. Rush's analysis found that the weight given to the PSI-90 measure was much greater than the seven other measures in the safety group. Specifically, PSI-90 was weighted 1,010 times stronger than the catheter-associated urinary tract infections measure, 81 times stronger than the C. difficile infection rates measure, 51 times stronger than the central line-associated bloodstream infection rates measure and 20 times stronger than either the surgical site infection rate measure.

Latent variable modeling changes the weighting and is inappropriate for measuring clinical outcomes, said David Levine, senior vice president of advanced analytics and informatics at Vizient (1). "Given the disproportionate weighting of the safety scores over time, they did not represent a composite measure," said Dr. Omar Lateef, an author of the analysis and Rush's senior vice president and chief medical officer (1). Lateef said he and his colleagues at Rush were alarmed by a rating drop from 5 to 3 stars because they have improved performance on five of the eight safety measures since the December release. " Lateef added that although CMS was initially dismissive of Rush’s concerns that CMS has come around since presented with Rush’s analysis.

CMS announced earlier this week that it was delaying release of the star ratings "to address stakeholders concerns." No date has been set for when the new ratings will be released.

Richard A. Robbins, MD

Editor, SWJPCC

Reference

1. Maria Castellucci M. CMS star rating system has been wrong for two years, health system finds. Modern Healthcare. June 15, 2018. Available at: http://www.modernhealthcare.com/article/20180615/TRANSFORMATION01/180619933?utm_source=modernhealthcare&utm_medium=email&utm_content=20180615-TRANSFORMATION01-180619933&utm_campaign=am (accessed 6/15/18).

Cite as: Robbins RA. CMS' star ratings miscalculated. Southwest J Pulm Crit Care. 2018;16(6):338-9. doi: https://doi.org/10.13175/swjpcc078-18 PDF 

Arizona Averages Over 25 Opioid Overdoses Per Day

An Arizona Republic article and the Arizona Department of Health Services Director's blog, Dr. Cara Christ, brings the opioid crisis home (1,2). Christ states that over 3200 opioid overdoses with over 400 deaths occurred between June 15 and October 17 in Arizona. This averages to over 25 overdoses and 3 deaths per day.

Some of the data from Christ’s blog are below:

• Males 25-29 have the highest rates of suspected opioid overdoses.
• 37% of people experiencing a suspected opioid overdose had an opioid prescription in the two months prior to their overdose.
• The majority of overdoses occur at home.
• The most commonly cited pre-existing health conditions of those with suspected overdoses was chronic pain. Depression and other behavioral health conditions were also common health conditions noted.
• Meth and heroin were the most frequently cited drugs involved in reported neonatal abstinence syndrome.
• About 40% of people experiencing suspected opioid overdoses who had a prescription in the Controlled Substances Prescription Monitoring Program had been prescribed both benzodiazepines and opioids in 2017. When these medications are combined, it is so dangerous that the FDA gives it a “black box” warning.
• About 40% of people experiencing a suspected overdose that had prescription history in the Controlled Substances Prescription Monitoring Program (CSPMP) had received opioid prescriptions from 10 or more providers.
• Only about 25% of clinicians prescribing controlled substances checked the CSPMP prior to prescribing.

On October 16, a new mandate went into effect that requires clinicians to check the CSPMP prior to prescribing an opioid or benzodiazepine. Other states implementing such mandates have experienced reductions in people with 4 or more prescribers or pharmacies, reductions in opioid prescribing, and reductions in Morphine Milligram Equivalent daily doses.

The CSPMP requires registration and login but is relatively easy to use (3). You can search not only in Arizona but other states as well. Personally, as a pulmonary consultant I infrequently prescribe opioids or benzodiazepines. However, I have used the website once to check a benzodiazepam prescription for a patient I suspect might be addicted. No other prescriptions were found. It at least gave me some assurance that he was not obtaining prescriptions from multiple practioners while we attempt to wean him off this medication.

Richard A. Robbins, MD

Editor, SWJPCC

References

1. McCrory C. More than 400 opioid-overdose deaths reported in Arizona since June 15. Arizona Republic. October 27, 2017. Available at: http://www.azcentral.com/story/news/local/arizona/2017/10/27/more-than-400-opioid-overdose-deaths-reported-arizona-since-june-15/809157001/ (accessed 10/28/17).
2. Christ CM. Opioid update: latest data and emergency rules update. October 17, 2017. Available at: http://directorsblog.health.azdhs.gov/opioid-update-latest-data-and-emergency-rules-update/ (accessed 10/28/17).
3. Arizona Board of Pharmacy. Arizona PMP Aware. Available at: https://pharmacypmp.az.gov/ (accessed 10/28/17).

Cite as: Robbins RA. Arizona averages over 25 opioid overdoses per day. Southwest J Pulm Crit Care. 2017;15(4):179-80. doi: https://doi.org/10.13175/swjpcc133-17 PDF 

Substitution of Assistants for Nurses Increases Mortality, Decreases Quality

Substituting nursing assistants for professional nurses is associated with poorer quality of care and increased mortality according to a study published in BMJ Quality & Safety (1). Linda H. Aiken PhD and colleagues analyzed the effect of increasing the proportion of less extensively trained nurses at 243 acute care hospitals in Belgium, England, Finland, Ireland, Spain, and Switzerland. They surveyed 13,077 nurses and 18,828 patients who had been in 182 hospitals between 2009 and 2010. They also consulted mortality records for 275,519 patients who had had surgery in 188 of the hospitals between 2007 and 2009.

Overall, 47% of the professional nurses in the study had bachelor's degrees, although they were unevenly distributed, with some hospitals having none. In a hospital that has average nurse staffing levels and skill mix, the researchers estimated that replacing one professional nurse with a lower-skilled worker increased the odds of a patient dying by 21%. Conversely, each 10% increase in the proportion of nurses with high-level skills was associated with an 11% decrease in the odds of a patient dying postoperatively and a 10% decrease in the odds of a patient giving the hospital a low rating.

Overall, the findings paralleled those from the United States and are consistent with the concept that a higher level of education leads to improved care. "We find a nursing skill mix in hospitals with a higher proportion of professional nurses is associated with significantly lower mortality, higher patient ratings of their care and fewer adverse care outcomes," the researchers write. They conclude "that caution should be taken in implementing policies to reduce hospital nursing skill mix because the consequences can be life-threatening for patients."

Richard A. Robbins, MD

Editor, SWJPCC

Reference

1. Aiken LH, Sloane D, Griffiths P, et al. Nursing skill mix in European hospitals: cross-sectional study of the association with mortality, patient ratings, and quality of care. BMJ Qual Saf. 2016. Published on-line 11/15/16. [CrossRef] 

Cite as: Robbins RA. Substitution of assistants for nurses increases mortality, decreases quality. Southwest J Pulm Crit Care. 2016;13(5):252. doi: https://doi.org/10.13175/swjpcc121-16 PDF