Trial Magazine

You are litigating a nursing home abuse or neglect case and suspect the defendants understaffed their facility to increase revenue. However, the defendants drag their feet in producing relevant documents; staff are hesitant to speak about the conditions at the nursing home in depositions; and the parent and consulting companies assert that they were not involved in the operation, management, or control of the facility.

But you know that labor costs are the largest component of the nursing home budget and that short-staffing is a key strategy to increase the facility’s profitability. How can you prove these allegations when the defendants refuse to produce a single document? The answer lies with the Centers for Medicare and Medicaid Services’ (CMS) staffing data for the facility.

CMS is the federal entity that regulates the operation of most skilled nursing facilities in the United States. The agency identifies staffing as one of the vital components of a nursing home’s ability to provide quality care.¹ Indeed, ample evidence exists that links resident outcomes to staffing levels in a facility—multiple studies show a strong positive relationship between the number of nursing home staff who provide direct care to residents on a daily basis and the quality of care and quality of life of those residents.²

Higher registered nurse (RN) staffing levels are associated with better resident care quality in terms of fewer pressure ulcers; lower restraint use; decreased infections; lower pain; improved activities of daily living (ADL) independence;³ less weight loss, dehydration, and insufficient morning care; less improper and overuse of antipsychotics; and lower mortality rates.⁴

Publicly Available Data

The good news is that CMS posts staffing information on the Nursing Home Compare website and uses it in the Nursing Home Five-Star Quality Rating System to help consumers understand the level of staffing in different nursing homes.⁵ On a quarterly basis, each facility must electronically submit direct care staffing information based on payroll and other auditable data through the Payroll Based Journal (PBJ).⁶

CMS then publishes public data sets (also known as public use files or PUFs) that display daily data on nursing home staffing levels based on the information submitted through the PBJ and Minimum Data Set (MDS) records.⁷ The staffing data combined with census information (the number of residents in a facility at any given time) provide valuable information on whether a nursing home was staffed appropriately.

CMS also calculates the number of direct care nursing staff it “expects” each skilled nursing facility to provide based on the residents’ Resource Utilization Group (RUG-IV) categories.⁸ Every resident in a skilled nursing facility is assigned a RUG score. The higher the RUG score, the more help and nursing time the resident needs.

For example, a resident who requires extensive services will either be RUG ES3, ES2, or ES1. A resident who has behavioral symptoms or cognitive performance symptoms, such as difficulty with recall or repeating words, will be assigned BB2, BB1, BA2, or BA1.⁹ Each category corresponds to a recommended amount and type of care for the resident.

The CMS Staff Time Resource Intensity Verification (STRIVE) Study, which began in 2005, measured the average number of RN, licensed practical nurse (LPN), and nurse aide minutes associated with each RUG score. In other words, CMS determined how much time it takes to care for a resident who fits within a particular RUG category. These are referred to as “case-mix hours.”¹⁰

The case-mix hours for each nursing home are based on the daily distribution of residents by RUG score group in the quarter covered by the PBJ reported staffing data and the estimates of necessary daily RN, LPN, and nurse aide hours from the STRIVE Study.¹¹ CMS provides case-mix nursing minutes by RUG-IV Group and nursing staff type in the “Technical Users’ Guide.”¹²

Let’s imagine a nursing home has 100 residents. Fifty of those residents fall under RUG-IV Code RHB, meaning they need 203.78 total nursing minutes. The other 50 residents fall under RUG-IV Code LC2 and need only 191.36 total nursing minutes. To find the case-mix hours for this facility, CMS averages the total nursing minutes needed by the residents in each RUG-IV category. These residents require an average of 197.57 minutes of total nursing time, and CMS would calculate the facility’s case-mix hours as 3.29. Therefore, the facility needs to provide 3.29 total nursing hours (RN + LPN + aide) per patient day.¹³

There are criticisms of the STRIVE study as it only measured time provided by the facilities participating in the study and did not measure the nurse staffing levels necessary to achieve the quality of care for each resident to attain or maintain the “highest” practicable level of well-being under federal regulations.¹⁴

STRIVE nursing times in certain categories were lower than the lowest staffing minimums recommended by experts.¹⁵ Despite these problems, CMS uses the STRIVE times to adjust nursing home data for comparison purposes. However, remember that the CMS case-mix hours may not represent the hours required for high-quality resident care. Just because a facility is meeting the case-mix hours for each nursing category does not mean the facility is meeting the residents’ needs.

In addition to the PBJ reported staffing data and the case-mix hours, CMS assigns each nursing home facility a rating of one to five stars for both RN staffing and total staffing. And then for each nursing home, an overall staffing rating is assigned based on the combination of the RN and total staffing ratings. The rating categories are set using a percentile-based method determined by factoring in clinical evidence on the relationship between staffing and quality.

For instance, if the total nurse staffing hours per patient day are between 3.108 and 3.579, but the RN staffing hours are below .317, the staffing rating for the facility will be one star. If the total nurse staffing hours are greater than or equal to 4.408, but the RN staffing hours are still below .317, the staffing rating for the facility will be three stars. The most recent staffing and rating values can be found in the “Technical Users’ Guide.”¹⁶

---

If you know where to look, with just a few keystrokes you can download the defendant facility’s daily staffing data for a quarterly period.

---

The most important aspect of this data is that it is at our fingertips. If you know where to look, with just a few keystrokes you can download the defendant facility’s daily staffing data for a quarterly period.

Download the PBJ Data

CMS provides downloadable files of the PBJ data.¹⁷ The full datasets contain more records than most spreadsheet programs can accept, so you’ll have to filter the data for a more manageable dataset.

Step 1. Choose the year and quarter you want to search. For example, if your client was injured in July 2021, search for “Q3 2021” in the top left corner of the page to evaluate the staffing data at the facility during that time.

Step 2. It will be helpful to manage the columns included in your dataset. Not all columns are relevant to the exploration of the staffing data. I recommend checking only the following under “Manage Columns”:

• PROVNUM: this is the provider number and can help when searching data discussed later
• PROVNAME: this is the name of the facility
• CY_Qtr: this is the year and quarter
• WorkDate: this is the year/month/day of the data in that row
• MDScensus: this is the census at the facility on the date indicated
• Hrs_RN: this is the total number of RN hours for the date indicated
• Hrs_LPN: this is the total number of LPN hours for the date indicated
• Hrs_CNA: this is the total number of certified nursing assistant (CNA) hours for the date indicated.¹⁸

Step 3. Enter the provider name or provider number in the search bar. This creates a dataset for the specific facility for that quarter and year, which can be exported into an Excel spreadsheet. I recommend organizing the data by WorkDate before exporting so you will have the data in chronological order.

Now you have the reported nursing hours for each category of nurses, aides, or both at the facility on any given date in the quarter, as well as the facility’s census. For example, if your client’s injury occurred on July 30, 2021, you may see a total of 11.5 RN hours for a resident census of 52, meaning there were just .22 RN hours per resident on that date.

There are also categories for Hrs_RNDON (total hours for the director of nursing (DON)); Hrs_RNadmin (total hours for RNs who serve in an administration role, such as the MDS coordinator); and Hrs_LPNadmin (total hours for LPNs who serve in an administration role rather than as direct care staff). Arguably, these categories of hours do not accurately reflect hands-on nursing time provided to patients because administrative nurses and the DON rarely provide hands-on care to residents. The hours in these categories are not included in the other categories and do not need to be deducted from those totals. You can simply exclude them from your filtered spreadsheet.

Since these datasets are updated quarterly and available up to three years back, it is a good practice for someone in your firm to export the data for each facility you are in litigation with at the end of each quarter.

Download the Case-Mix Hours

A downloadable file that contains the “case-mix” and “reported” hours is included in the nursing home provider information data table available in the Provider Data Catalog on cms.gov.¹⁹ The reported hours are a monthly average of the nursing hours the facility submitted through the PBJ system. The case-mix hours are calculated by CMS based on the STRIVE study and represent the nursing hours recommended for the particular patient population.

Step 1. Download the .zip file for the month and year you are interested in. The file will be titled “nh_archive_month_year.zip” or “nursing_homes_including_rehab_services_archive_month_year.zip.”

Step 2. Once the file is unzipped, download the file titled “NH_ProviderInfo_MonthYear.” The Excel spreadsheets contain columns with specific information, including:

• PROVNUM
• PROVNAME
• Columns Y, AA, AC, AE, AG, AI, and AK with various CMS ratings on quality and staffing
• Column AO: reported nurse aide staffing hours per resident per day
• Column AP: reported LPN staffing hours per resident per day
• Column AQ: reported RN staffing hours per resident per day
• Column AU: case-mix nurse aide staffing hours per resident per day
• Column AV: case-mix LPN staffing hours per resident per day
• Column AW: case-mix RN staffing hours per resident per day.

Step 3. It is easiest to manipulate this spreadsheet if you freeze the top row so you can always view the categories of data. Then, use CTRL-F to find the name of the facility you are interested in. Columns also can be “hidden” in Excel so you only view relevant columns and data. It can take some time and practice to perfect manipulating this spreadsheet, but once mastered, you can screenshot the data you need and create documents that you can later use for exhibits or demonstrative aids in your case.

Deconstruct the Data

Once you’ve downloaded the relevant information, what do you do with it? Given the considerable evidence of a relationship between nursing home staffing levels and resident outcomes, the most impactful comparison of the data is between the reported staffing hours per resident day from the PBJ and the case-mix staffing hours per resident day.²⁰

For example, if the total reported RN hours per resident per day was .22 on the date of your client’s injury, but the case-mix RN hours per resident per day is .34457, there is a deficit of .12 RN hours per resident that day that should have been provided according to CMS. If your client had an infection, but there was a delay in sending her to the hospital because an RN was not available to do an assessment, this data can be very convincing to a jury that the facility’s staffing decisions resulted in harm to your client.

Often, the larger discrepancy is with nurse aide hours per resident per day—meaning the nursing home is reporting fewer nurse aide hours than CMS has calculated are needed for that facility’s resident population. Nurse aides provide most of the hands-on resident care, such as dressing, feeding, bathing, and changing soiled linens. If the nurse aide hours per resident per day are short of what that patient population requires, residents simply aren’t receiving the care they need, which may result in infections, pressure ulcers, dehydration, and malnutrition.

You will be able to evaluate the CMS data for each day your client was a resident or look at trends over the course of your client’s residency. On the NH_ProviderInfo spreadsheets, the reported staffing hours for each category of nurse are an average for the month. So the reported data on the PBJ system provides a more accurate snapshot of the facility’s staffing on a particular day. However, you have to divide the nursing hours provided on the PBJ spreadsheets by the census because the PBJ data is based on a total number of nursing hours rather than the nursing hours per patient day used for the case-mix hours detailed in the Provider_Info spreadsheets.

For instance, if the PBJ spreadsheet shows that on Jan. 1, 2019, the facility reported a total of 43.34 RN hours (in column “Hrs_RN”), to determine the per patient day hours you must divide 43.34 by the census for the facility. If the census reported for that day is 141 (in column “MDScensus”), 43.34/141 is .307—the amount of RN hours per patient day. Compare .307 with the case-mix hours for RNs on the NH_ProviderInfo spreadsheet (in column “CM_RN”).

If your client’s injury was caused by a single incident—such as a fall—understanding the staffing on that specific day is extremely important. However, if your client’s injury is a pressure ulcer that developed over time, it may be helpful to look more closely at the quarterly numbers that give a broader overview of staffing at the facility. Either way, the CMS data is effective to illustrate allegations of short-staffing and how it contributed to your client’s harm.

Demonstrate Why the Data Matters

While the CMS data is helpful to understand whether a facility was understaffed, it also can assist in painting a picture of corporate malfeasance and the corporate entities’ role in the resident’s harm. Often, nursing home suits are filed against multiple defendants, including related corporate entities with significant financial ties to the facility that are making the financial and operational decisions for staffing and care and treatment resources.

In these lawsuits, which intertwine patterns of systemic corporate malfeasance with professional and ordinary negligence, defendants avoid producing relevant materials by re-characterizing the plaintiff’s theories and allegations as nothing more than isolated instances.

In my firm’s practice, we typically get the most pushback in discovery on the production of grievances and complaints, facility-wide occurrence reports,²¹ and budget documentation for the facility. These documents provide insight into the overall conditions at the facility, as well as the allocation of resources for residents’ care and treatment.

When we file a motion to compel, the CMS staffing data demonstrates to a judge that this is not an isolated incident. The staffing data and staffing rating provide evidence of understaffing, which can affect the quality of care provided to residents. If the staffing rating of a facility is a one out of five for a period of four quarters, or if the reported nursing hours per resident day are consistently lower than the case-mix nursing hours per resident day, there is a clear basis for the argument that systemic failures existed in providing adequate care and treatment to residents.

This data also demonstrates that defendants had notice and knowledge of these conditions. When the data is available to the public on the internet and is provided by the nursing home itself, the defense argument that the facility was adequately staffed simply will not pass muster. As a result, a judge is more likely to grant your motion to compel the production of grievances and complaints about these issues across the facility and not only your client’s specific complaints.

This same premise applies to the facility-wide occurrence reports for events such as falls, infections, and pressure ulcers. If you can demonstrate to a judge that your client’s injury is the result of budgetary decisions to understaff, facility-wide occurrence reports become relevant to reveal the facility’s conditions, as well as defendants’ notice and knowledge of the issues related to your client’s injury.

When requesting budget information for the facility, the CMS staffing data supports allegations that the entire corporate structure is aimed at increasing revenue at the expense of adequate care and treatment. Without the CMS staffing data showing a facility is understaffed, it can be difficult to convince a judge to order production of budget documents.

While homing in on this data is time-consuming and tedious, it is the building block of your corporate nursing home case. At the end of the day, we are trying to show jurors that these corporate players cut corners and diverted funds from staff and resources. And that these budgetary decisions were intentional and resulted in residents not receiving the care and treatment necessary to prevent harm. By developing evidence of understaffing through the CMS data, you are taking the first step to show that defendants put profits over resident health and safety.

---

Anne K. Moore is an attorney at Connor & Connor in Aiken, S.C., and can be reached at anne@theconnorfirm.com.

---

Notes

1. Ctrs. for Medicare & Medicaid Servs., Staffing Data Submission Payroll Based Journal (PBJ), https://tinyurl.com/3vj66xa7.
2. Charlene Harrington et al., Appropriate Nurse Staffing Levels for U.S. Nursing Homes, 13 Health Servs. Insights 1 (June 29, 2020), https://www.ncbi.nlm.nih.gov/pmc/articles/PMC7328494/.
3. ADL are skills required to manage one’s basic physical needs, including personal hygiene or grooming, dressing, toileting, getting in and out of bed or a chair, walking, and eating.
4. Harrington et al., supra note 2.
5. Ctrs. for Medicare & Medicaid Servs., Medicare.gov, https://tinyurl.com/wbh7tnmc; Ctrs. for Medicare & Medicaid Servs., Five-Star Quality Rating System, https://tinyurl.com/msr3ahpf.
6. 42 C.F.R. §483.70(q); Staffing Data Submission Payroll Based Journal (PBJ), supra note 1.
7. The MDS is a compilation of screening, clinical, and functional status elements that forms the foundation of a comprehensive assessment for all nursing home residents who participate in Medicare and Medicaid. The items in the MDS standardize communication about resident problems and conditions and provide a comprehensive assessment of each resident’s functional capabilities. Ctrs. for Medicare & Medicaid Servs., Long-Term Care Facility Resident Assessment Instrument 3.0 User’s Manual (Oct. 2019), https://downloads.cms.gov/files/mds-3.0-rai-manual-v1.17.1_october_2019.pdf; Ctrs. for Medicare & Medicaid Servs., Minimum Data Set 3.0 Public Records, https://tinyurl.com/3jymeu6x.
8. RUGs flow from the MDS and drive Medicare reimbursement to nursing homes. Skilled nursing facilities are required to classify residents into one of 66 RUGs based on assessment data from the MDS. U.S. Dept. of Health & Human Servs., Nursing Home Resident Assessment Resource Utilization Groups (Jan. 2001), https://www.oig.hhs.gov/oei/reports/oei-02-99-00041.pdf.
9. Health Care Compliance Ass’n, SNF PPS: RUG IV Categories and Characteristics, https://tinyurl.com/2p937zhk.
10. Note that the term “case-mix hours” replaces the term “expected hours” that was used prior to April 2019.
11. Ctrs. for Medicare & Medicaid Servs., Design for Care Compare Nursing Home Five-Star Quality Rating System: Technical Users’ Guide (Apr. 2022), https://tinyurl.com/2p8h72sj.
12. Id. at 25.
13. “Staffing hours per patient day” means the number of full-time non-managerial care staff who ordinarily will be assigned to provide direct patient care divided by the expected average number of patients on which such assignments are based.
14. Harrington et al., supra note 2 at 8. For the regulations, see 42 C.F.R. §483.75.
15. Id.
16. Technical Users’ Guide, supra note 11.
17. Ctrs. for Medicare & Medicaid Servs., Payroll Based Journal Daily Nurse Staffing, https://tinyurl.com/5dwffsy9.
18. Note that there are columns for contract employees—for example, “Hrs_RN_ctr.” You would want to include these columns if it appears that your facility used contract nurses to have a more accurate picture of the staffing levels at the facility.
19. Ctrs. for Medicare & Medicaid Servs., Nursing Homes Including Rehab Services Data Archive, https://tinyurl.com/4jnsj7zk.
20. Harrington et al., supra note 2 at 3, 9.
21. This can include infection control reports, pressure ulcer reports, or CASPER reports, which show the occurrence of events throughout the entire facility. For example, a weekly pressure ulcer report lists every resident in the facility who has a wound with the location of that wound, the date it was acquired, the stage of the wound, the size of the wound, and other information.