Destress Your Sampling Process
Most medical review teams do not have an active role in sampling. In fact, if you were to ask them, many probably don’t want any part of it! It can be stressful, even on investigators! But understanding that in order to thoroughly review an allegation, your sample must be designed so the investigator or SIU staff member requests the records appropriately. This can make the difference between a strong case and one that is indefensible.
But how do you destress the sampling process? Well, aside from having a clear process in place, you should really include the medical review perspective before you sample.
Many of us have experienced the pain of sampling for an investigation only to find out that we failed to remember one important aspect of our data or scenario that renders our sample useless or at least ineffective in evaluating our allegation. In our SIU Survival Guide, we describe a scenario where a medical reviewer could not refute nor substantiate the allegation because the sample design was conducted at the claim level, rather than the member level, which rendered the medical review ineffective.
This type of error can be avoided by keeping 3 key questions in mind:
What is the allegation?
〰️
What is the specialty?
〰️
What is the size of the dataset?
〰️
What is the allegation? 〰️ What is the specialty? 〰️ What is the size of the dataset? 〰️
It is essential to know the focus of your investigation and tailor the sampling design in a way that will allow for an appropriate assessment of the allegation while considering the provider specialty and the size of the dataset.
Will the sample unit be claims? Claim lines? Members? Other?
Should it be narrowed to a particular place of service? Code set?
What is the appropriate date range?
Part of determining how you select the universe for your sample is going to depend on how episodes of care are defined based on the subject of the audit. This determination is best made with input from your medical review team, the ones who will have to actually review the records.
Typically, your sample unit will be a member, claim, claim line or our favorite, a combination of member and date of service. At the beginning of a lead or investigation, this may be unclear, but in the early steps of the investigation, you need to consider this before pulling your sample. Often, we have multiple allegations to contend with. In these instances, it’s important to formulate your sample in a way that allows you to see the full picture.
Here are some examples
Durable Medical Equipment (DME). DME suppliers will typically supply either 1 item or a specific supply kit. The claims associated with a DME supplier are often not interconnected over time and as such, have little to no impact on each other from a record review perspective. What we mean is that even though supplies may be recurring and have limits, the records from one date of service, don’t typically impact another in the way other reviews may. For this reason, it may be appropriate to audit a set of service dates or claim lines attributed to a certain supply.
Primary Care Provider (PCP) / Family Practice. For a PCP audit, the services the provider renders for an encounter are often related to one another - and there will generally be multiple services (claim lines) on a claim. For a medical record review, the reviewer may want to consider requesting records by claim ID (with all lines included) or requesting by the member ID / date of service combination (especially if you suspect the subject is claim splitting). By sampling on a member ID / date of service, your reviewer can get a full picture of the services billed on any given day. Claim splitting is quite common, especially if the subject is trying to circumvent edits.
Home Health. For a home health case, it may be necessary to look at a longer period of time to see the care that members received over several months/years. For this reason, a member-level audit which pulls all claims for all dates of service for the member may be appropriate. This way we can assess whether the patient is getting better and whether the plan of care is being modified as progress is made. Note: if you are doing a statistically valid random sample (SVRS) this can get quite large as your sample unit is a member.
The examples listed above are by no means an exhaustive list of options for record requests. Their purpose is to demonstrate an example of the thought process necessary by the investigator to accurately determine the best way to sample for an audit. Whether you handle all parts of your case internally or have an outsourced team that supports you, when in doubt, phone a friend. If you don’t conduct your own reviews, we recommend reaching out to a medical reviewer to capture their insight… and maybe formalizing that process to make it more efficient. We refer to this process or meeting as collaboration meetings (more on those in the near future)!
The process of sampling to support an investigation is not a new concept in the healthcare fraud, waste and abuse industry. But in order for it to be done properly, it must be a collaborative step in the overall investigative process. Bringing together the investigative and medical review components to ensure that you are sampling appropriately is extremely important. Don’t go it alone - ask for help!
Integrity Advantage is the way healthcare payers reimagine the value of their fraud, waste and abuse program.
We provide FWA services to payers around the country. If you need a program assessment, program growth strategy, investigations, medical reviews or training support -- reach out today.
We are a certified Women’s Business Enterprise (WBE) and an Economically Disadvantaged Woman Owned Small Business (EDWOSB).
For more information click below, call us at 866-644-7799 or email info@integrityadvantage.com.