Dive into Rivet's Revenue Diagnostics, your financial control center, designed to enhance your organization's fiscal health. This tool aids in various net revenue tasks including month-end procedures, revenue projections, cash reserves, physician payroll, and identifying causes of revenue leakage. This article outlines the core components of Revenue Diagnostics, explaining its forecasting mechanism and providing insights into its various features.
1. Date of Service Bill Date or Date of Payment
Explore Three perspectives in healthcare financials: date of service, bill date, or date of payment. Each offers unique insights suitable for different use cases - operational and financial/accounting.
Date of service:
Focuses on the service dates of claims.
Ideal for accrual accounting or operational analysis, such as assessing the impact of internal process changes on denial rates.
In the given example it displays the total charges for claims made between July 1, 2023, and August 1, 2023. The net revenue projections and actual figures in this view correspond to these claims. Given the varied adjudication times of claims, this means that forecasts indicate that we predict that you will eventually collect that much net revenue over the next ~12 months that comprise a normal adjudication timeline. Similarly, the actual net revenue is based on the remittances and Explanation of Benefits (EOBs) for this group of claims, irrespective of the receipt dates of these remittances/EOBs.
Bill date:
The difference between bill date and the date of service comes from a billing lag. A week or two can elapse between the service rendered date of service and the claim sent to the payer.
This means that selecting the March date of service and March Bill date results in a different set of claims, because the claims rendered at the end of Feb might not be billed until March, and claims rendered at the end of March might not be billed until April.
Date of payment:
Centers on the date of payment, or more specifically, the date of the remittances/EOBs.
Best for financial and accounting purposes from a cash perspective, such as monthly net revenue projections.
In the example above, you would see the gross charges of claims with remittances/EOBs dated between July 1, 2023, and August 1, 2023. The net revenue forecasts in this view represent the net revenue we expect you to collect on the selected dates, regardless of when the claim was billed. The same is true for the actuals. For example, if you want to see how much net revenue you can expect to collect in July, you would select the date of payment tab with the date selection above.
2. Understanding Adjudication Over Time
The Adjudication Over Time chart is a feature offering a visual representation of the adjudication process for a selected set of claims. This chart plots the adjudication curve of a set of claims with the selected dates of service.
Rivet initially projects the total net revenue expected to be collected from these claims. Following this, we estimate the pace at which these claims will undergo adjudication. This results in a comprehensive visual that traces the journey from when the service is rendered to the receipt of insurance payments.
For more in-depth information, click on 'Show data' located beneath the chart. This action reveals a table detailing the numbers represented in the chart. If you wish to view the data, with weekly granularity simply click the toggle adjacent to a particular month. For a day-to-day breakdown, select the three dots in the upper right corner and export the report as a CSV file.
3. Net revenue lookback
The Net Revenue Lookback chart provides insights into the service dates of claims that were paid within your selected payment dates. Essentially, it attributes a segment of the actual net revenue to the services that generated this revenue.
This chart is particularly helpful for investigating unexpected outcomes. For instance, if the actual net revenue for July was lower than anticipated due to a spike in denials, using the date picker to focus on July in this chart will reveal which service months were most impacted by this revenue leakage. Following this, you can switch to the date of service or bill date tab and select the identified month to further analyze and understand the underlying root cause.
4. Forecasting in Revenue Diagnostics
Important Notes on Forecasting:
Estimates, Not Certainties: While we believe in the accuracy of our forecasts, it's crucial to understand that they are estimates, not certainties. Rivet advises against treating our forecasts as absolute truths, particularly in fiduciary duty situations, such as mergers and acquisitions.
Directional Guidance: In Revenue Diagnostics, consider these figures as providing directional guidance. They are insightful indicators, but not infallible truths.
Accuracy Considerations:
Considerations for Accuracy: Our forecasting accuracy is largely contingent on the amount of past data available. The more historical data you have, the more accurate the forecasts.
Optimal Data Duration: For the most insightful forecasts, having two or more years of historical claims data is ideal. This duration helps to accurately gauge the seasonality trends specific to your business.
Challenges with Limited Data: With only a year's worth of data, distinguishing between regular patterns and anomalies can be difficult.
Confidence Levels and Data Requirements:
Confidence with Limited Data: Forecasts can be generated with less than a year of data, but the confidence in these forecasts will be proportionately lower.
Role of Settled Claims: The number of 'settled' claims in your account (those fully adjudicated and not subject to further review) acts as a basis for our forecasts.
Guideline for Forecast Duration: As a rough guideline, we can project forecasts for half the duration of the period covered by your settled claims.
Seasonality Analysis: To accurately forecast seasonal variations, a minimum of two years of data is necessary.
5. How we forecast
Revenue Diagnostics uses an advanced statistical technique called linear regression to provide forecasts. Each “line” with its corresponding slope is considered a “forecast”.
Understanding Our Forecasting Approach: When creating a forecast, we analyze your data by plotting it as points on a graph and drawing a line through those points. The slope of that line represents "the forecast," allowing us to predict future values by following the line forward.
Why Linear Regression: We choose linear regression for its optimal blend of accuracy and explainability. Linear regression is deterministic, meaning the forecast remains the same with the same data. Although it may smooth out variance to some extent, providing accuracy on average, it can be less effective at predicting outliers like seasonality.
Consideration of Advanced Methods: While we considered advanced methods in Machine Learning, the potential accuracy gains were outweighed by the lack of explainability. Advanced Machine Learning techniques, while powerful, aren't deterministic and pose challenges in explaining their inner workings.
Rivet's Stance on Explainability: Given the critical role of net revenue forecasting, Rivet prioritizes explainability over incremental accuracy improvements when possible. We use historical data to create a forecast line that extends into the future.
Forecasting Accuracy Considerations: The longer the forecast period (e.g., a full year) or further into the future (e.g., a quarter three quarters from now), the more likely inaccuracies may occur, especially towards later dates. It's akin to the importance of precision when walking across a room versus flying across the country. Being one degree off on a short walk has an infinitely smaller impact on your journey than being one degree off on a cross-country trek.
Dynamic vs. Static Forecasts: Keeping forecasts static allows for observing variances between forecasted metrics and actual results, aiding in identifying revenue leakage sources. However, we recommend re-forecasting every six months for continued accuracy.
6. Dynamic vs. Static Forecasts
Defining 'As of' Date: The 'As of' date feature in Revenue Diagnostics allows users to specify the cutoff date for historical data included in forecasts. This feature enables viewing both dynamic (constantly updating) and static (fixed) forecasts.
Dynamic Forecasts:
Automatically Updating: Dynamic forecasts in Revenue Diagnostics update in real-time, incorporating new data as it's ingested by Rivet. While the dynamic nature of these forecasts means that they will continuously get more accurate, it also means that they are less useful for use cases where you want to compare actual results to forecasted results.
Use Cases: Dynamic forecasts are primarily useful for use cases where accuracy is the primary concern, such as determining physician compensation and cash on hand. To use dynamic forecasts, do not assign a value to the “As of date” filter.
Static Forecasts:
Fixed Forecast Values: In static forecasts, the forecasted values do not change once they’ve been saved. This means that the variance between a forecasted value and an actual result is not affected.
Use Cases: Useful for budget adherence or identifying revenue leakages. To set a static forecast, select the “As of date” filter from the filter menu and select the last day of historical data you want to consider for this report. For example, if you want the forecasts you’re currently viewing not to change, set the 'As of' date to today. Once you’ve selected your as of date, click on the “Save report” button on the left-hand side. The “As of” date filter also allows for more advanced use cases, such as seeing a forecast from a perspective in the past. For example, if you meant to save a report for Q1 before the quarter started, you can set an “As of” date on the last day of Q4 and get the forecast results you would have seen then.
Caution with 'As of' Date:
Impact on Accuracy: Setting the 'As of' date too far back can reduce the accuracy of forecasts due to less historical data being used.
Minimum Data Requirement: Avoid setting the 'As of' date earlier than 9 months from your first service date in Rivet, as this is the minimum required data for accurate forecasting.
Remember, the 'As of' date is a versatile tool in Revenue Diagnostics but should be used judiciously to ensure the reliability of your forecasts.
7. Lock forecasts function
When you save a report, you will notice an option labeled 'Lock Forecast Values.' By selecting this checkbox, you can guarantee that the forecast values in this report remain constant and will not be subject to future changes. This effectively sets the 'As of' date to today, ensuring that the data reflects the forecast as of the current date.
8. Patient Responsibility limitations
As of now, Rivet cannot determine how much actual patient responsibility has been collected to date. The figure in the overview table represents the patient responsibility that is owed, not how much has been paid. The actual payer responsibility accurately reflects the cash collected from payers. To enhance the accuracy of Rivet's forecasts, we recommend exporting the report from Rivet and applying your standard patient collections calculations to it. This may include multiplying by a collections rate and incorporating a time lag.
9. How we use contractual rates
Customers subscribed to Rivet’s Payer Performance product will find that Revenue Diagnostics employs the contractual rates that have been uploaded into their accounts, as seen in the 'Contract Evaluation' table. The forecasts presented in this table are based on the contractual rates you have provided. In contrast, the actual amounts shown reflect the allowed amounts derived from actual claim and remittance data. Therefore, the discrepancy between the allowed amounts and the net revenue indicates revenue leakage, which is primarily due to denied and underpaid claims. Generally, allowed amounts should be higher than net revenue because they represent an “ideal world” without underpayments or denials. If your allowed amounts are lower than net revenue, likely, you haven’t loaded all of your rates into Rivet. When Rivet does not have rates for a specific claim, it cannot produce an allowed amount for it. Please reach out to support if you have this issue.
If you have any additional questions about Revenue Diagnostics feel free to chat in or send an email to support@rivethealth.com.