Defining a Bundled Payment Rate for Bone Marrow Transplants

Summary

Insurers and providers are increasingly pushing for bundled payments and other forms of value-based reimbursement in their contracts with health systems.  The ideal candidate for a bundled payment can be described as follows:

• Well-defined care protocol – the typical steps taken to provide care are well defined
• High frequency of occurrence – the provider has ample volume across which to spread risk
• Low variability in costs from patient to patient
• Clear indication of when the care begins and is complete – i.e., the scope of the bundled payment

Hip and knee replacements, for example, are well-suited for bundled payment consideration.  However, we are seeing more and more bundled payment arrangements emerging for care that doesn’t necessarily fit these parameters.  Among these are bone marrow transplants (BMT).  Such transplants occur with relatively low frequency, and despite the standardization of care, patients vary widely in terms of the care and services needed during the transplant process.  Costs vary widely by patient.  These issues present certain challenges from the standpoint of defining a standard case for a bundled rate.

The purpose of this material is to identify some of the critical actuarial issues in defining a bundled payment for autologous BMT (i.e., where the patient is their own donor, versus allogenic, where the donor is a different person).  I discuss issues regarding each of these characteristics in the remainder of this writing, along with how actuaries might address such issues in analyzing the contract and/or developing a bundled payment rate.  Along the way, I will provide examples of results from recent work performed for a health system in the western U.S., along with results from a normative claims database.  Throughout this writing, “BMT” is assumed to refer to autologous bone marrow transplants.

It is important to note that this is written only to reflect the actuary’s viewpoint – when considering and negotiating the bundled rate and contract, it is absolutely critical to have guidance and approval from clinical leadership of the health system.  I focus first on evaluating the contract proposed by the insurer, and then go into greater detail regarding analysis of costs and rate development.  It is assumed that the reader has some familiarity with the typical BMT course of care.

Contract Evaluation

Typically, the insurer will present a draft contract to the health system for the arrangement.  The actuary will want to review all contract language from the perspective of managing the financial risk to the system.

Scope of the Bundle: Defining The Transplant Period

The typical BMT bundled payment is intended to cover the transplant period (TP), which is usually defined by beginning and ending dates.  Below is a description of commonly proposed dates, along with a brief discussion of potential issues to be considered.

• TP Beginning Date – often defined as the start of the mobilization regimen.
  • Mobilization should be clearly defined in the contract. A common interpretation is the start of the administration of specific drugs.
  • The contract should also clarify the beginning date if there are multiple mobilizations. Ideally, only the most recent mobilization should be used to determine the TP beginning date; the earlier mobilizations would be considered part of the pre-TP, and paid accordingly.
  • Health systems may wish to exclude the mobilization step from the transplant period, because of the variability in cost and patient needs. However, this variability is precisely why insurers are seeking to secure bundled payment contracts, and thus it is unlikely that insurers will be willing to accept more restrictive definitions of the transplant period.
• TP Ending Date – often defined as a certain number of days (e.g., 120) following discharge from the hospital following the transplant.
  • BMT is increasingly being performed on an outpatient basis – the contract should clarify what the TP ending date is for both hospitalized and non-hospitalized patients.
  • The contract should be clear as to payment in the rare instance where multiple infusions are required (i.e., there is a failure to engraft). Ideally, if there is a failure to engraft, the TP should end; for payment purposes, either a new TP begins, or subsequent infusion attempts are paid as outliers or part of the post-TP care.

We have also identified certain circumstances where the contract may be unclear:

• Premature closing of the case – in some circumstances, mobilization may begin, signaling the start of the TP, but the infusion does not proceed. The contract should clearly state how payment will be made for the care delivered.
• Tandem transplants – Payment and TP determination for tandem transplants, if performed by the health system, should be clearly defined.
• Covered services – The contract should be clear that the only services covered by the bundled payment are those related to the transplant. Any other services would continue to be paid as determined under the broader provider contract.
• Storage of stem cells – The contract should also be clear as to payment for storage of stem cells. This is typically included under the bundled payment.
• Emerging drugs and technologies – Ideally, new drugs and technologies will be paid at cost, or at a specific percentage of billed charges.
• Timing of payment – Finally, the contract may specify how the bundled payment will be made – commonly in installments, subject to certain reports being submitted by the health system to the insurer. Health system leadership should confirm that such requirements are reasonable and that the timing and amount of the installments are acceptable.

This is by no means an all-inclusive list of considerations; the contract should be reviewed in its entirety for any other financial risks to the health system.

The Pre- and Post-Transplant Periods

Pre- and post-TP payments are typically described in the bundled payment contract.  Services performed before or after the TP are typically paid at a percentage of billed charges for hospital services (the percentage may differ between inpatient and outpatient services), and a fee schedule for professional services (commonly presented as a percentage of the Medicare fee schedule).

Defining Outliers

Outlier terms are critical in BMT bundled rate contracts, because of the high variability in costs among patients.  Currently, BMT contracts most often use inpatient days as an indicator of outlier experience.  For example, a contract may specify that any inpatient days beyond 30 will be paid on a per diem basis.

However, we have found that inpatient days are a relatively insensitive outlier term that is rarely invoked, and for which per diems are generally insufficient to cover additional costs.  Therefore, we suggest that health systems seek to define outliers in other ways in the bundled payment contract.  Based on the cost analysis described later in this writing, we suggest seeking the following outlier terms:

• Use of the mobilization drug Plerixafor. Plerixafor was present for about half of the autologous BMT patients we recently reviewed.  We found that the use of Plerixafor was linked with significantly higher overall costs.  In some cases, Plerixafor added as much as $100,000 to the overall case expense. Therefore, it may be desirable to set up the contract such that Plerixafor is paid at cost or on a percent of billed charges basis.
• Time for radiation and/or chemotherapy. As might be expected, we have seen that case expenses also varies with the time spent in radiation and/or chemotherapy.  Rather than using inpatient days as an outlier indicator, the health system may want to target days spent in radiation/chemotherapy as a potential indicator of outlier experience.
• Diagnosis. We found distinct differences in cost between diagnoses of lymphoma and multiple myeloma. Lymphoma cases were on average about $30,000 higher in overall case expense than multiple myeloma cases.  This may not be enough to merit a separate outlier term in the contract – rather, the health system may want to examine the prevalence of each diagnosis in their own patients, and consider the financial impact if the distribution of diagnoses shifts.

Typically, contract review proceeds concurrently with cost analysis (the next section of this writing).  It is critical to incorporate any contract term changes into the cost analysis and determination of an acceptable bundled payment amount.

Cost Analysis

The health system’s experience with autologous BMT should be considered when determining an acceptable bundled payment rate. Such experience may not be available, or more commonly, may not be available for both hospital and professional care delivered.  The analysis can be supplemented with a review of normative claims data.  Both approaches (patient-level analysis and normative data) are described below.  First, we begin with the output from the patient-level analysis.

Deciding on a Rate

The typical cost analysis will produce a table of estimated payments at different target reimbursement levels.  The health system decides what percentage of billed charges is acceptable, and the table illustrates the corresponding equivalent bundled base payment rate.  For example, Table 1 below shows base payments as a percent of billed charges (i.e., the equivalent allowed charges) from a recent assessment:

Table 1

In the example above, the insurer has proposed a case rate of $106,950 with $2,173 per non-ICU outlier day, which equates to about 37.9% of total billed charges on average.  The health system is currently being paid at 85% of charges, which is equivalent to a case rate of $239,638, more than double what the insurer has proposed.

The ultimate negotiated rate lies somewhere between those two values.  The health system must determine what level of reimbursement is acceptable, given the contract terms and the system’s own experience with applicable cases.  More information about this determination is provided later in this writing.

Patient-Level Analysis

In order to construct Table 1, we relied on data submitted by the health system for both hospital and professional charges for approximately 30 autologous BMT patients.  This list was ultimately narrowed down to 27 patients, as three of the patients appeared to have care that either originated or ended outside the health system.

For each patient, we completed the following steps:

1. Tabulate all hospital and professional experience in chronological order
2. Identify the transplant period (TP) start date based on the proposed contract definition
3. Identify the TP end date based on the proposed contract definition
4. Calculate the billed charges before, during, and after the TP
5. Determine the payments that would be received under the proposed contract

The result was a table showing total TP billed charges for each patient, along with what would be expected under the insurer’s proposed contract.  An excerpt from this table is shown below.

Table 2

	Transplant Period Billed Charges			Proposed Contract Payment
Patient	Hospital	Professional	Total	Base	Outlier	Total	% of Billed
1	$201,296	$4,852	$206,148	$106,950	$0	$106,950	51.9%
2	$211,156	$7,797	$218,953	$106,950	$0	$106,950	48.8%
3	$335,028	$8,654	$343,682	$106,950	$0	$106,950	31.1%
4	$136,539	$5,018	$141,548	$106,950	$0	$106,950	75.6%

Based on the total billed charges across all 27 patients, we then were able to produce the base payment table shown in Table 1.

Normative Data

We also reviewed our normative claims database to provide supplemental cost information.  Our database had about 400 recent autologous BMT cases, of which 283 were deemed complete and appropriate for use in analysis.  Table 3 below provides a summary of our analysis of allowed charges.

Table 3

	All LOBs	Medicare	Commercial	Individual
Number of Cases	283	39	217	27
Net Allowed	$186,500	$104,800	$199,600	$198,800

Table 3 indicates that the average allowed charges across all cases is about $186,500, which appears consistent with the experience obtained from the health system if payments average around 65% to 70% of billed charges (a reasonable assumption).

Patient-Level Analysis versus the Normative Data

Because cost varies so much from patient to patient, a critical consideration is whether the health system’s historical patient experience is indicative of typical experience.  For one health system, we found that their experience included more lower-cost patients than the typical distribution, as shown in Table 4 below.

Table 4

Quartile	Begins:	Database Average:	Normative Distribution:	Health System Distribution:
1	$0	$215,742	25%	51.9%
2	$278,230	$351,865	25%	40.7%
3	$419,662	$502,611	25%	7.4%
4	$624,607	$908,320	25%	0.0%

Therefore, in developing the case rate, the health system must also consider the financial implications if a more typical patient distribution is experienced in the future.

Setting the Case Rate

Given all of this information, the health system may then set the target case rate by taking into account three key components:

1. The health system’s cost for providing the care. The bulk of the target case rate reflects the funds needed to cover the system’s actual cost for providing the care.  A simplistic approach is to review current cost-to-charge ratios.  A cost-to-charge ratio of X% suggests that X% might be the minimum acceptable target reimbursement rate – however, it is important to consider which underlying expenses are fixed versus variable (from the BMT perspective).This approach is complicated by the special nature of BMT services, and in particular, the expensive drugs that make up a large portion of the charges.  The cost-to-charge ratio may need to be adjusted to reflect the system’s effective costs for these drugs.
2. The health system’s administrative burden. The system may want to add a small adjustment to the target reimbursement to reflect any administrative burden associated with operating under the case rate.  This may include, for example, special documentation or reporting required by the payer under the case rate contract, or internal accounting activities required to allocate funds to hospital and physician providers.
3. Risk premium.Similar to insurer practice when taking on risk, the system may wish to charge a risk premium for taking on the financial risk under the case rate.  A risk premium of 5% to 7% is typical, although a higher premium may be desired due to the wide variability of expenses for BMT patients. Statistical analysis to develop an exact risk premium is problematic due to the small number of cases, but could be conducted across a larger dataset to develop a more precise BMT-centered risk premium.

Ultimately, the health system must decide what is an acceptable level of reimbursement.

In Closing

Despite certain challenges, insurers are increasingly looking to providers to take on risk via bundled payments for autologous BMT and other care that isn’t necessarily optimal for such an arrangement. Careful attention to contractual language can help to mitigate the risk.  Regardless, the system should recognize and plan for the variability in cost not addressed in outlier payments or other contract provisions.

Guest Post – Elaine Corrough, FSA, FCA, MAAA