Artificial Intelligence (AI) is transforming the workplace. It offers immense potential for employers to enhance efficiency, decision-making, and overall performance. However, as AI is increasingly integrated into hiring decisions and other HR functions, employers must take note of the ethical and regulatory considerations that surround its use.
In this blog we review some of the potential benefits of the use of AI for employers as well as regulations considerations in using AI for hiring decisions.
Potential Benefits of AI for Employers
Enhanced Efficiency: AI can automate routine tasks, freeing up time for the HR team to focus on strategic activities. Ai can support activities such as sorting through resumes and scheduling interviews, both activities that typically consume significant resources.
Improve Decision-Making: AI systems can quickly and accurately analyze large volumes of data, providing insights that help employers make informed decisions. This includes identifying trends in employee performance, predicting future hiring needs, and optimizing workforce planning.
Bias Reduction: When properly designed and implemented, AI can help reduce unconscious bias in hiring decisions. AI can objectively evaluate candidates based on data rather than subjective impressions, leading to more fair and more inclusive hiring practices.
Cost Savings: Automating aspects of the hiring process can significantly reduce costs. AI can handle initial screenings, reducing the need for extensive human involvement until the later stages of the hiring process. The efficiency can translate into considerable savings.
Better Candidate Matching: AI-powered tools can analyze job descriptions and candidate profiles to identify the best matches, improving the quality of hires. By considering a broader range of factors, including skills, experience, and cultural fit, AI can help employers find candidates who are more likely to succeed in their roles.
Rules and Regulations Surrounding the Use of AI
While AI offers many benefits, its use in hiring is subject to various restrictions and regulations designed to ensure fairness and protect candidate rights. Employers must be aware of these to avoid legal pitfalls and ethical dilemmas.
Data Privacy Laws: Regulations like the General Data Protection Regulation (GDPR) in Europe and the California Consumer Privacy Act (CCPA) in the U.S. impose strict requirements on how personal data is collected, stored, and used. Employers must ensure that AI systems comply with these laws, particularly when handling sensitive candidate information.
Anti-Discrimination Laws: In the U.S., the Equal Employment Opportunity Commission (EEOC) enforces laws that prohibit discrimination based on race, color, religion, sex, national origin, disability, and genetic information. AI systems must be designed to comply with these laws, avoiding biases that could lead to discriminatory practices.
Transparency and Accountability: Employers must be transparent about their use of AI in hiring decisions. Candidates should be information when AI is used to evaluate their applications and employers should be prepared to explain how AI decisions are made. This transparency is crucial for building trust and ensuring fairness.
Algorithmic Fairness: AI systems must be regularly audited to ensure that they are fair and unbiased. Employers should work with AI vendors who prioritize ethical AI development and are willing to provide insights into their algorithms’ workings.
Bias Mitigation: Employers must actively work to mitigate any biases in their AI systems. This involves continuous monitoring, testing, and updating of AI algorithms to ensure they do not perpetuate or exacerbate existing biases.
Beware of Evolving Regulations
To harness the benefits of AI while navigating its regulatory landscape, employers need to keep up to date with the latest regulatory landscape and best practices related to AI in hiring. The AI regulations are evolving – and vary by jurisdictions. Employers must constantly monitor to ensure that they remain aware of the most recent regulations. Additionally, employers must regularly audit AI systems to ensure that they are fair, unbiased, and compliant with relevant laws. Finally, success will require collaboration of all stakeholders – HR, legal, and IT teams – to ensure that the AI systems that are implemented reflect a holistic approach.
Where Can You Find Out More?
Using AI to remove bias from hiring decision-making:
People with terminal illness experience pain at the end of life and for many, this pain goes untreated. One of the key elements of hospice care is effectively managing the patient’s pain. Untreated or undertreated pain results in needless suffering – due to physical pain and mental distress. However, family caregivers often have a difficult time assessing their loved one’s pain. Further, they are often concerned with side effects of pain medications, including concerns of addiction to or tolerance of pain medications. Further, both patients and family caregivers often have trouble communication with the hospice team the degree and nature of pain that the patient is experiencing. This often leads to ineffective pain management and needless suffering in end-of-life.
What are some considerations when giving pain medication?
Respect
Respect the patient’s wishes regarding pain management. Understanding a patient’s goals and values guides the care team in providing personalized and compassionate care.
Consult
Pain medication decisions are made in consultation with the patient, considering their preferences, values, and goals for care.
Collaborate
The hospice team includes the primary physician, medical director, nurse, social worker, chaplain, hospice aide, caregivers, and patient. Everyone works together to create the right plan.
How should pain medication be administered and monitored?
Individualized and Regular Assessment
Pain medicine is administered based on individualized assessments of the patient’s pain levels. Regular assessments of pain are important for managing pain and ensuring the plan remains effective.
Address pain early
Addressing pain before it becomes too severe can contribute to more effective pain control and improved quality of life.
Communication
Encourage patients to communicate openly about their pain levels. This information is crucial for healthcare providers to make informed decisions about medication adjustments. Regular communication between caregivers and healthcare providers ensures an accurate understanding about the patient’s pain.
Low dose pain medicine
It is preferable to initiate low-dose medication to maintain alertness and minimize potential side effects.
Titration
Titrate pain medication up as needed, to achieve optimal pain relief. Regular assessments guide the titration process ensuring the right balance between pain control and functionality. Often a long-acting pain medication is given coupled with a breakthrough pain medication, if needed, to keep pain at or below goal level.
Timely administration
Administer pain medicine in a timely manner, adhering to the prescribed schedule. Consistent dosing helps maintain a baseline level of comfort.
Pain log
Use a pain log to track pain levels, related factors, what medicine was given, dosage of medication, and time medication was given.
Education
The hospice team should educate the patient and caregivers on the use of pain medication, including dosage, timing, and potential side effects.
Monitor
Monitor for potential side effects of pain medication and collaborate with the healthcare team to address any concerns. This includes a careful assessment of the patient’s overall well being.
Consider a holistic approach to pain management
Many hospice agencies advocate for a holistic approach to pain management, including physical, emotional, and spiritual care. Alongside pain medications, this involves exploring and integrating non-pharmacological interventions such as massage, music therapy, aromatherapy, and relaxation techniques for a more comprehensive approach to pain relief.
The hospice Special Focus Program (SFP) is conducted by the Center for Medicare and Medicaid Services (CMS). The objective of this program is to identify poor performing hospice agencies, based upon quality indicators, that place hospice beneficiaries at risk. These hospice agencies will then be subject to additional scrutiny and oversight to ensure that they meet Medicare requirements. The SFP is designed to either bring these programs into compliance or force them out of the Medicare program by terminating their Medicare status.
What is the origin of the Special Focus Program?
The hospice Special Focus Program was mandated in the Consolidated Appropriations Act of 2021. That is also when it was clarified that hospices would be surveyed every three years. All hospices now have had a survey since 2021. Some of that data is being used for the hospice Special Focus Program, which is designed to identify the worst performing hospices and either bring them into compliance or force them out of the program by terminating their Medicare status.
How is a hospice agency selected for inclusion in the Special Focus Program?
CMS uses an algorithm to identify the poor performing hospice agencies to include in the SFP. The algorithm combines data from a few data sources to score each of the hospice agencies. The score is based on data from: condition-level deficiencies in standard surveys, substantiated complaints, Hospice Care Index (HCI), and the CAHPS survey. The algorithm does not stratify hospice agency based upon size or location; all hospice agencies are held to the same standard regardless of their size or location. The bottom 10% ranked hospice agencies (which are the hospice agencies with the highest algorithm score) are selected to be included in to the SFP.
What is the impact of a hospice agency being included in the SFP?
Hospice agencies that are included in the SFP will be publicly reported on the SFP website. SFP is a framework for increased oversight. The hospice agencies that are included in the SFP program will be surveyed more frequently — at least every six months. CMS will determine what actions must be taken based upon the survey results.
How will a hospice agency exit the SFP?
A hospice will complete the SFP if in an 18-month time frame the hospice agency has no Quality of Care condition level deficiencies or immediate jeopardies for any two six month SFP surveys and has no pending complaints or have returned to substantial compliance with all requirements. The hospice will receive a letter from CMS that will indicate official completion of the program. If a hospice is unable to meet the completion criteria – due to inability to successfully pass surveys or continued complaints while on the SFP – it will be placed on the Medicare termination track.
Even as hospices work to improve their levels of quality and compliance, there will always be hospice agencies that fall in the lowest 10% of performance relative to their peers. Only by continually monitoring their quality performance and comparing these quality scores to peer performance can a hospice agency stay out of the lower 10% and off of the SFP list.
Financial statements record an agency’s financial activities and convey the agency’s financial health. Financial statements are often audited by accountants to ensure their accuracy and are prepared in accordance with a specific set of accounting rules called Generally Accepted Accounting Principles (GAAP). GAAP ensures that financial statements are consistent and comparable across agencies.
In some instances, agencies are required to follow GAAP in preparing their financial statements. For example, the financial statements that are submitted with hospice cost reports must be prepared in accordance with GAAP.
An agency’s financial statements describe the results of the agency’s operations and provide information about the agency’s assets and liabilities, revenues and expenses as well as cash flow activities. Together, this information provides a picture of the agency’s financial health and activities and can be used to monitor agency performance over time. An agency’s financial statements can also be used to support decision-making.
There are three key financial statements
The three key financial statements are the balance sheet, the income statement, and the cash flow statement.
The balance sheet provides a snapshot of the agency’s financial health at a given point in time. It lists the agency’s assets and liabilities as of the last day of a specified period, also known as its capital structure. The balance sheet will usually be produced on a quarterly basis.
The income statement, also known as the profit and loss statement, provides information on all the agency’s revenue and expenses during a specified period. The income statement is usually prepared quarterly, often as an aggregation of internal monthly ledgers, and is used to review revenue and expenses over the period. Subtracting expenses from revenues results in the agency’s profit, also called the net income.
The cash flow statement details cash inflow and outflow over a specified time period. Cash flow statements are generally prepared quarterly and are separated into operating activities, investment activities, and financing activities. The cash flow statement, which tracks how the agency uses its cash over a period of time to fund operating expenses and investments and pay off its debt obligations, provides a view of the agency’s liquidity and cash sources and uses, which can provide insight into the ability of the agency to generate sufficient cash to support its operations over the near- and long-term horizon.
How can an agency director use financial statements?
Financial statements are critical to assess the financial viability of an agency, its capital structure, income generation and cash flow needs. Financial statements can also be used to measure the impact of initiatives on the agency’s bottom line. Compare direct expenses to revenue, as reported on the income statement. Did revenue increase? Were expenses reduced? How did the net income compare to expectations?
Financial statements also provide information on agency spend and cash flows and can provide insight into how cash uses are allocated. Periodically reviewing expenses, when necessary on a line by line basis from the relevant internal ledgers, can provide visibility into potentially unnecessary costs and areas where expenses can be reduced and spend can be reallocated to more productive areas.
Analysis of financial statements can provide an overall agency view of revenue and expenses, identifying how different areas in the agency are contributing to overall financial performance. This can support discussions with different area leads on goals and budgeting and helps management reassess overall agency performance. Financial statements can also be used to motivate and engage the agency’s teams. The income statement can show employees how their activities impact revenue, providing them with an understanding of how they can directly impact agency financial health. Remember, only a financially strong agency can deliver improved patient services and higher quality of care.
Patient length of stay is monitored to detect instances of inappropriate use of the hospice benefit or other deficiencies in the services delivered by the hospice provider. Length of stay is monitored for both very short length of stay as well as for length of stay that is longer than the norm.
What may unusual length of stay tell a hospice provider?
When patients are discharged alive with a short length of stay it may signal that the patient did not understand the hospice benefit when the patient was admitted to hospice. Or, patients may discharge live from hospice after just a few days because they were not satisfied with the services delivered by the hospice provider. Patients with length of stay longer than 180 days could be indicative of a patient who is no longer hospice eligible. Patients who are no longer eligible for service should be discharged from hospice and any payments that were received from Medicare while the patient was no longer eligible for services should be returned to Medicare. Failure to discharge the patient or failure to return the funds are examples of fraud and abuse.
How is length of stay calculated?
Length of stay is calculated based on the number of days that a patient receives hospice care. Specifically, for a patient who is discharged from hospice (whether or not the patient is discharged alive), the patient length of stay is calculated as follows:
Patient length of stay = [patient discharge date]-[patient admission date]+1
Which patients are included in length of stay calculation?
The length of stay calculation assumes that only discharged patients are considered in the calculation – since the formula expressly refers to the patient discharge date. When only discharged patients are considered (whether live discharges or discharges due to death), the hospice provider only has a backward-looking view on performance relating to length of stay. For example, if a hospice provider has been providing service to a patient for 12 months and the patient is still on service, the patient will not be included in the traditional average length of stay metric – since the patient has not yet been discharged. On the other hand, once the patient is discharged the patient’s length of stay will be at least 365 days since the patient – while still currently active – has already been on service for 365 days. If active patients are considered in a length of stay calculation, it gives a hospice provider a metric that can be used to highlight patients whose clinical charts and documentation of care may benefit from additional review.
What length of stay metrics should be calculated?
In addition to computing average and median length of stay based on discharged patients only, average and median length of stay can be computed for active patients. Patient length of stay for an active patient is calculated as follows:
Active patient length of stay = [end of evaluation period date]-[patient admission date]+1
For example, suppose the current date March 15, 2023 and a hospice wishes to calculate the active patient length of stay as of the end of 4Q 2022 for a patient who was admitted on December 1, 2022. The calculation is as follows:
End of evaluation period date: 12/31/22
Patient admission date: 12/1/22
Active patient length of stay = (12/31/22) – (12/1/22) + 1 = 31 days
The active patient length of stay as of the end of 4Q 2022 is 31 days.
If the hospice wishes to calculate the active patient length of stay as of current date, the calculation is as follows:
End of evaluation period date: 3/15/23
Patient admission date: 12/1/22
Active patient length of stay = (3/15/23) – (12/1/22) + 1 = 105 days
Average and median length of stay would be computed as usual. If any concerning value — such as long length of stay – is identified based upon the active patient length of stay, a hospice provider can immediately investigate and determine if any remediation action is required, rather than waiting until patients are discharged. Delay can lead to additional fines or further action from Medicare.
Patients are eligible for hospice if they have a terminal diagnosis and a prognosis of six or fewer months to live if their disease runs its natural course. A patient who lives longer than six months can still get hospice care if the medical director or other hospice physician recertifies that the patient is still terminally ill.
What is hospice patient length of stay?
Hospice length of stay is an important metric that is monitored by both CMS and by hospice providers. Hospice length of stay measures the count of days that a patient receives hospice services, from the day that the patient is admitted into hospice until the day the patient is discharged (either alive or deceased). In 2018, 25% of Medicare beneficiaries received hospice care for seven days or less and 54% of Medicare beneficiaries received hospice care for 30 days or less.
Why should a hospice monitor patient length of stay?
Monitoring patient length of stay can aid in detecting cases of possible fraud or abuse – instances where ineligible patients continue to receive the hospice benefit. This metric also helps monitor whether the hospice benefit is being adequately utilized. Although patients are eligible for hospice when they have six months or less to live, most patients receive less than 30 days of hospice care.
Agency patient length of stay is also trended over time and is also compared against the value for patients in the same region, state, or nationwide. The metric may also be analyzed for patients in subpopulations – for example patients with the same disease, race, or ethnicity.
How is patient length of stay calculated?
Patient length of stay is calculated using all patients discharged by the hospice provider during the reporting period. For example, if the hospice would like to compute the length of stay for patients during the 4Q 2022, all patients who were discharged during 1Q 2023 would be included in the calculation. For each patient, the number of days from the date of patient admission until the date of patient discharge is counted; this represents the patient length of stay.
Patient length of stay = [patient discharge date]-[patient admission date]+1
What are common measures of length of stay?
Two common patient hospice length of stay measures are Average Length of Stay (ALOS) and Median Length of Stay (MLOS).
Average length of stay
Average length of stay is the arithmetic mean of the data collected. Specifically, if d is patient length of stay and N is the total number of patients then average length of stay (ALOS) is calculated as follows:
ALOS = ( d1 + d2 + d3 + …. + dn ) /N
Where di = patient length of stay for patient i
Median length of stay
Median length of stay is the middle number in the sequence of numbers. Specifically, compute the length of stay for all N patients. Then, order these N numbers in ascending order. The middle number is the median. If the number of patients is even then there is no middle number. Instead, the median is calculated by taking the average of the two numbers in the middle.
Comparing average and median length of stay
The average is sensitive to outliers in the data. That is, if there are a few patients with a very high length of stay while all other patients have a significantly lower length of stay, the average will be biased by these outliers and will give a misleading assessment of overall patient length of stay. Below, we give an example to provide greater intuition into the impact of outliers on average length of stay and the difference between mean and median length of stay.
Suppose a hospice agency discharged 35 patients during 4Q 2022. The patients’ lengths of stay are as follows:
We compute the average length of stay by summing each of the 35 patient’s length of stay (in the “Length of Stay” column) and dividing that total by 35 (the total count of patients).
Average length of stay (ALOS) = 38.5
We compute the median length of stay by sorting the patient’s length of stay in ascending order and identifying the central number. Since there is an odd number of patients, there will be a single central value. In this case, the central value is 20.
Median length of stay (MLOS) = 20
Average length of stay is almost double the median length of stay. What is leading to these significant differences between ALOS and MLOS? Observe the outliers in the data. There are two patients with length of stay that exceeds 200 days. There are two additional patients with length of stay exceeding 100 days. Since ALOS is sensitive to outliers, ALOS is being pulled to a higher value due to the presence of these outliers.
To provide additional insight, we have plotted a histogram of the length of stay values. A histogram shows the count of observations in the data that fall in each of the specified ranges.
The table on the left shows the count (frequency) of observations of patient length of stay in the data for each of the ranges: 0-10 days, 10-20 days, 20-30 days, 30-40 days, and greater than 40 days. There are 11 patients with length of stay between 0-10 days, 7 patients with length of stay between 10-20 days, 6 patients with length of stay between 20-30 days, 6 patients with length of stay between 30-40 days, and 4 patients with length of stay that exceeds 40 days.
Think about this histogram and now consider the MLOS and ALOS. Median length of stay is 20 days – it falls well in the middle of the data. Average length of stay, however, equals 38.5. It falls, essentially, in the final bar of this histogram and well beyond where the majority of the data lies. The provides a visual demonstration of the impact of outliers on ALOS.
Providers should monitor both ALOS and MLOS. Significant differences between these numbers would indicate the presence of outliers and should be investigated.
Print ‘n take hospice keys
Understanding the difference between the average (mean) and the median
