MANAGED CARE August 1998. ©1998 Stezzi Communications
Health plans can position themselves even before talks begin by getting all the information they can on the population they wish to target.
The saying that the three most important things to consider when buying a house are location, location and location can be tweaked a bit and applied to the process of purchasing a disease management program. The three most important tools health plan representatives need when sitting down at the bargaining table with a vendor are information, information and information — allowing you to shape your expectations and bargaining position. Without such information, shopping for a disease management program is akin to walking into a car showroom and saying aloud that you don't quite know what you want. Gathering data should begin even before the initial contact.
"All of the vendors are more than willing to put your data through their buffing system but they charge you a considerable amount of money for that," says Thomas Morrow, M.D., vice president and medical director for One Health Plan of Georgia. "The more you know about your population and your expectations, the better prepared you are going into the talks."
Once you have the data, there are things you should insist upon, says Al Lewis, president of the Disease Management Purchasing Consortium, a company that has helped 22 health plans negotiate contracts. "I hugely prefer that the program use savings as a performance measurement," says Lewis. "You can't guarantee savings without guaranteeing outcomes because you can't get savings without keeping people healthy."
Morrow says health plans should also seek:
- A deadline for promised improvements;
- A schedule for progress reports;
- A plan for continuing improvements beyond what's promised; and
- A guarantee that the contract can be modified later.
Morrow also likes to see patient satisfaction measures of some kind. "They're talking to our members and I don't know if they're being polite or rude." You'll want assurances that the vendor will be invisible, seamlessly integrating the program with other services you offer.
"There is no vendor that has so much credibility in disease management that patients will feel more comfortable talking to him than they are talking to their own doctor," says Louise Kier Zirretta, M.B.A., group vice president for Isis Research and author of the Disease Management Strategic Research Study & Resource Guide, which she prepared when she was at Migliara/Kaplan Associates, a market research company.
Insist on significant, measurable outcomes, says Lewis. "For instance, with diabetes, there should be far fewer amputations in the out years of a contract and lower hemoglobin A1Cs in the early years. With asthma, there should be a guarantee of X percent fewer nighttime awakenings or missed school or work days. A surprising number of health plans don't insist on guarantees of quality and costs. The program has to cover the entire population of people with the disease. Who gets counted? I'm quite adamant that everybody should get counted."
Peter Smith, CEO of Ralin Medical, a DM vendor in Buffalo Grove, Ill., has "no problem" with Lewis's assertion, but says the term "population-based disease management" can be interpreted in different ways. He points out that many plans, in the beginning, identify the acutely ill "who are clearly costing them money." That might be a way to go, at first. And even well into the program, he says, different services should be allotted to different subgroups that are defined by the severity of their symptoms. "I think you need to separate acute and nonacute patients. You don't want active management of your nonacute patients because you'll wind up paying more than it's costing you today."
You'll want a contract in which the vendor shares the risk for savings when the particular disease lends itself to such a system, says Harry L. Leider, M.D., vice president of health services and corporate medical director at HealthNet, with 450,000 members in Kansas and Missouri.
"Nobody assumes risk management for things like hypertension and high cholesterol because the payback is too long," says Leider. Asthma, end-stage renal disease and congestive heart failure make excellent candidates for risk contracting. "The big issue is, 'Is this a fee-for-service arrangement or a risk arrangement?" I'd want a risk arrangement. It shows our financial people that we're serious about saving money. Risk agreements separate the wanna-be vendors from the ones that really have competence."
Despite the advantages of risk-sharing arrangements, statistical data indicate that many health plans do not insist on such contracts. Kier Zirretta's guide reports that 48 percent of over 100 managed care organizations surveyed said that contracts "currently in place with vendors are on a fee-for-service basis, although [managed care organizations] expressed a clear preference for risk-sharing arrangements that provide for evaluating the program at the end of a defined period and sharing the savings or loss from disease management services with the vendor."
Says Kier Zirretta: "Vendors, taking their cues from managed care organizations, would like to see the programs capitated."
Lewis says the point where much of this often gets ironed out is not in the contract, but in the letter of intent. (See "Reading between the lines") "Cost and quality guarantees and who gets counted should be 80 percent of the value of the contract," says Lewis. "Twenty percent of the bargaining time goes to 80 percent of the value. All the issues of who counts in the program, what gets counted and when — if you get those key points in the letter of intent, you can literally start implementation."
Gathering data
Negotiations involve more than demands. There's a "give" as well as a "take." Says Leider, "What are they going to expect you to do? Do they want you to do patient identification?"
The question brings us back to data-gathering concerns. Vendors, for the most part, want to sell you something that fits your need, says Ralin's Smith. "The more information the plan can provide, the better," he says, pointing out that most disease management contracts are based on historical baseline costs. How much has the plan had to pay to provide services for the population being targeted for disease management? Answering that question is crucial, says Smith. "We have found that once we've been able to agree on the historical data, the process tends to speed up. The biggest problem is just getting the data."
If the whole data-gathering process leaves you frustrated, you are not alone. Here, again, is the Disease Management Strategic Research Study & Resource Guide, saying that managed care organizations "cited the lack of adequate information technology as the most challenging obstacle to overcome in the implementation of a disease management program. Information technology is expensive and time-consuming to develop, and thus a ripe market exists for vendors who can satisfy this need."
Morrow reports that "the biggest thing we've come across is system limitations. If you can't get information that you can give to the vendor that he can use and feed back to you, then you're dead in the water." For instance, one vendor Morrow negotiated with needed the phone numbers for patients with asthma. "We don't have the numbers for many patients," he says. "We had to pay another vendor to find those numbers. Also, you can call any household in Atlanta during the workday and there's nobody there. The problem is identifying your population. If Jane Doe needs to belong to a diabetes program, you've got to identify exactly who Jane Doe is, including her medical history. Most insurance companies can't aggregate all the information needed."
Claims information
Drew Palin, M.D., would agree with that. Palin is CEO of ThinkMed, a Milwaukee company that sells software that's intended to help HMOs gather and stratify claims information. "If you have a population of 100,000 members, 800 to 900 of them are going to be diabetics. But all diabetics aren't the same."
Sound like a pitch? Companies like ThinkMed, while not DM vendors, could help plans during the prenegotiation stage. In fact, says Palin, ThinkMed provides that service for some of the plans that use its software.
But "claims data don't necessarily give you clinical data," Smith notes. "Claims data may not be that difficult to pull, but trying to track back to the clinical data isn't always easy." Do vendor and plan sometimes read different things from the same data? "It's not so much that we would disagree," says Smith. "It's just that the whole process is quite complicated."
Palin admits that claims data aren't always translatable into clinical data, but points out that "the advantage of using claims information is that it's universal. It is a very good starting point and, with appropriate interpretation, it can serve as a proxy for clinical information."
Lewis says the steps leading to negotiations work like this: "Typically, there's a proposal from several vendors and then presentations are made and then, usually that same day, health plan officials vote on who they want to do this with. Then draft letters of intent go back and forth five or six times before the letter works for both sides."
Of course, he says, "There are presentations that wind up going nowhere. Often, the health plan was just being polite. Often the plan didn't have a process in place for working up a letter of intent. After the presentation, the plan has every intention of getting a letter to the vendor but it often doesn't happen."
Morrow agrees that the letter of intent is vital. "It's sort of like the engagement," he says. "It's a serious component for the whole process." After the letter comes another trouble spot, says Morrow. "There's often sticker shock because then the vendor is going to come back with a proposal and it's going to have a dollar sign attached to it."
But Leider says sticker shock can be avoided if you prepare for negotiations. "You need to know what the competitive deals are out there," says Leider. For instance, say a vendor tells you he can save you ten percent. "How do I know if ten percent is good or bad for most health plans?"
NCQA-style reports?
Morrow points out that knowing what you're trying to accomplish helps. Some disease management programs are no more than public relations efforts so that "the plan can tell employers it has something in place." More intensive programs delve into how to utilize services to reduce costs. Morrow says, "A third way is to set up a National Committee for Quality Assurance-type report to help you with quality improvements."
Can a disease management program help a plan meet NCQA standards? It depends on whom you ask but the answer seems to be, "Yes, in a roundabout way."
Morrow warns: If a vendor begins talking about how he can help you with NCQA data, listen carefully, because much of what NCQA requires can't be done by a vendor. "The adaptation of guidelines has to be a plan function," says Morrow. "They can tell you what NCQA requires and provide you with some of the outcomes material, but you still have to do it. You need to ask them to clearly define what they mean when they say that they can help with NCQA. There are a whole lot of things NCQA would like you to do that are usually not part of any program you can buy."
However, Leider of HealthNet asserts that a well-run DM program will impress NCQA officials. "NCQA wants to see you improving the care of populations of people. A good program will help you attain NCQA accreditation."
This sounds all too familiar to Jody Barish, director of clinical program development at Highmark Blue Cross Blue Shield. She recently completed negotiations with three different vendors. "The vendors will say 'We can help you get NCQA accreditation,' but it's really the health plan's responsibility to measure its own improvement activities," says Barish.
Kier Zirretta points out that "The NCQA measurements, even with the new HEDIS, have not really been disease-specific. They've really been process-specific. Having a well-run disease management program in place will definitely add to the quality of a plan. But nothing in the NCQA accreditation process says you have to have a disease management program."
All of which comes down to the practical wisdom that you should understand what you're buying, says Ralin's Smith. Most plans dealing with prospective vendors go onsite to see how programs work where they're already functioning. "We let the health plan officials come in and actually look over the shoulders of our nurses doing their jobs," says Smith.
Says Leider: "You need to find out if the vendor has ever achieved real outcomes. If the outcomes data the vendor offers during his presentation come from one site, you've got to ask yourself, 'Would the same system work here?'"
It's a question worth answering because, as Smith points out, a good disease management program will produce "enough savings so that everyone will come out ahead."
Reading between the lines
Al Lewis, president of the Disease Management Purchasing Consortium, says the most sensitive stage in negotiations between health plans and vendors negotiating a disease management agreement involves a letter of intent. Here's where the big issues — guarantees for cost, quality and performance — get hammered out.
The letter is usually the first step undertaken by health plans after deciding to work with a specific vendor — a decision usually made after vendor presentations.
Here are highlights from a generic letter of intent that Lewis uses as a template for actual letters leading to disease management agreements. Lewis explains the meaning and relevance of many of the items listed. (The letter is copyrighted and may not be reproduced without his permission.)
| August 20, 1998 John Doe President Any Vendor Inc. Dear Mr. Doe: Generic Health Plan would like to enter into a nonbinding Letter of Intent with Any Vendor Inc. to provide asthma disease management services. The terms of those services are detailed in the May 1998 proposal from Any Vendor to Generic Health Plan, with the following material exceptions: The per case fee is $_____ per member receiving Any Vendor services, payable one half at enrollment and one half 90 days following enrollment. |
|
| This means that the vendor is promising savings in excess of fees, and if it does not achieve those savings, it must reduce its fees in the year ahead. This is like a guaranteed-savings deal, but because the health plan, rather than the vendor, keeps the excess savings, there is no need to reconcile claims to figure out who owes what at the end of each period, unless the health plan feels there is a chance the targets were not met. | The vendor represents that savings in ER and IP claims will exceed the per-case fee by at least _____ percent. If not, the vendor must reduce the subsequent year's fees by the amount of the shortfall. |
| An important point, because you need to have a solution for members who join the plan after the program is implemented. It's fair to assume their claims experience is similar to the experience of current members. | Members may be added to the program at Generic Health Plan's option even if they do not have 12 months' worth of baseline data, using the previously calculated baseline as a proxy for their baseline, and the full-recourse progress payments associated with such members will also be $___ the first year. |
| This is needed because finding the members is such a critical part of implementing a disease management program for the plan's entire population. You don't want to let the vendor off the hook for members who aren't easy to find because they are likely to be less compliant generally, but it's not fair to make the vendor -play detective without compensation. | Members will have valid or forwardable addresses in 95 percent of cases, and reachable phone numbers in 80 percent for group health and Medicare, and 80 percent valid or forwardable addresses for Medicaid; to the extent these targets are not reached, Generic Health Plan will pay Any Vendor $20 per participant below the threshold. |
| This is important because to ensure savings without statistical bias, you must do the whole population, and you can't let vendors financially walk away from people who won't comply — their claims experience is likely to be high and must be included in final savings calculations. | Members will be deemed to consent to participate in the program if they are findable. |
| No inflation adjustment; actual changes in ER and IP rates will be used to adjust the baseline. | |
| This is fair because it is harder to save money if costs are low and easier if costs are high. Since you usually don't have the data when you sign the letter of intent, you don't know whether your numbers are high or low. This protects both sides from unexpected historic claims experience levels. | Also, the full-risk option is based on having an average amount of asthma medical losses. (ER and IP claims coded for asthma account for roughly 0.4 percent to 0.5 percent of all medical losses.) To the extent that the number is less than/greater than 0.5 percent, the guaranteed savings percentage will be adjusted proportionately. |
| It is easier to save money on a stable population. Changes in stability change the terms. | If annual Generic Health Plan turnover varies, up or down, more than two points from current levels, the guaranteed savings percentage will change inversely by one point. |
| Here, several important points are made mostly to ensure that while 80 percent of the potential points of dispute are in the letter of intent, the other 20 percent are at least considered at the time of the contract. | Generic Health Plan understands that there will be other terms Any Vendor may wish to negotiate, involving among other things:
|
| This helps speed up the process because it allows the vendor to rely on the letter of intent to start implementation. But it also prevents the vendor from speeding up implementation to create a large contingent liability should the health plan decide not to consummate. Basically, both sides feel pain if the deal is not ultimately consummated, making a completed contract much more likely than if one party could easily walk away. | However, all price-related terms are included in this letter. While this is not a "binding" letter of intent in the sense that no expectancy damages would be payable in the event that Generic Health Plan ultimately elects not to consummate an agreement with Any Vendor, Generic Health Plan would like Any Vendor to proceed with early stages of implementation effective upon receipt of this letter. To demonstrate Generic Health Plan's belief that it will ultimately consummate an agreement, Generic Health Plan is willing to reimburse Any Vendor one third of all documented out-of-pocket expenses solely related to implementation of the Generic Health Plan asthma program, in the event that Generic Health Plan does not proceed with the program substantially along the terms outlined in this note or the proposal. Please inform us along the way of every major budget increment ($5,000 or above) you are incurring, in order to allow us to monitor this process. If we believe there is a chance we will not consummate this agreement, we will ask you to postpone further spending. |
| A complete listing is much better than getting references because anybody can find a couple of decent references. | As part of our due diligence, we would like from you:
|