KM: The software itself has required fields for data entry. Unless the data that you’re entering is accurate and comports to the field, it won’t allow you to move on. Additionally, dose is automatically calculated within the software, which also provides the physician, nursing staff and clerical staff with dropdown screens with pre-populated information rather than having to manually put in information. From a treatment standpoint, it’s intended to provide the physician with a real-time reference of what general clinical practice would say would be an acceptable range of treatment. But it’s not a requirement that they practice within that because every patient is different. We’re looking at it as a more general reference point than as a requirement for practice.

KM: The vendor, Innovative Solutions, has a stellar Scientific Advisory Board that meets twice a year to review the clinical practice guidelines and enhance them as the state of practice changes. Fundamentally, the guidelines are a product of this scientific advisory board. Now that this program is getting promulgated and deployed in different areas, the local quality advisory committees are beginning to develop liaison relationships with the scientific advisory boards to provide their input, as far as refinement of these guidelines.

KM: Innovative Solutions had a very successful pilot program in the Detroit market with employees and dependents for Chrysler and General Motors. They used about seven facilities in that area. Their outcomes were very positive after their 18-month trial. I spoke with every physician that was involved with it; all were very positive as far as the efficacy of the program and the integrity of the guidelines. What Highmark was interested in was not coming out with another, "We say, you do" type of approach. We wanted a collaborative approach where we give physicians the tools that would help them more efficiently and more effectively practice medicine that they were trained to practice and let us step out of the way and provide informatics support and some administrative capabilities to tie together coalitions. For example, I’m trying to engage Pennsylvania’s malpractice carriers to potentially provide our physicians a discount in their premiums if they utilize this software and then to study how the use of this might affect claims and jury awards over time, and potentially pass those savings back on to the physician.

KM: Yes. Every time the physician enters in the description of the clinical information, the guideline pops up for their reference. They can see the bibliography that went into the development of the guideline and actually take a look at the abstract that was used. The Scientific Advisory Board has collated the guidelines promulgated by the American College of Surgeons, Radiation Therapy Oncology Group, National Cancer Care Network and the National Cancer Institute. There are a variety of different guidelines out there that are reasonably similar but duplicative. They’ve conducted a meta-analysis of the research and the existing guidelines, collated that information and summarized it in the form of these ranges of treatments so that there can be a quick reference source of information for the practicing physician.

KM: The guidelines at the time of treatment really are meant to be a general benchmark as far as their treatment approach, but also serve as a benchmark for some of those rare diseases that a community practice physician might see very infrequently and have some concern about the treatment approach. When we take a look at treatment patterns versus guidelines, and even more importantly, outcomes of patient care versus best practice outcomes, the guidelines provide a useful benchmark to get a general sense of a physician’s overall practice. The program is not intended to encourage the physician to treat one way or another. We want to provide general references and encourage the physician to think about their overall treatment approach—not the treatment approach on a specific patient. For example, a patient may be able to get into treatment only once a week. That means the physician cannot give the total dosage in as many fractions as they might like, which means that the potential complications might be greater. That might fall outside of guidelines. However, for that specific patient, that’s a treatment consideration that could only be made in that one-to-one interaction between physician and patient.

KM: We contract with virtually every credentialed radiation oncologist in the 29-county western Pennsylvania area, as well as their facility. We’ve asked for one administrative representative and one physician representative from each facility to serve on the Quality Advisory Committee. We typically get at our meeting greater than 50 percent of our network showing up, which is absolutely phenomenal for any kind of physician-driven committee. I think the physicians are interested in this process and, now that we’re far enough along that they see we are truly interested in outcomes, they’re starting to take an ownership. The leadership structure that has developed has been based on physician and administrative volunteers being co-chairs of committees and having additional responsibilities. The committee has met quarterly. The group tells up what type of reports will be relevant, what the interpretation of data might mean or might not mean, what things might require additional follow-up. We’re very careful to keep the information confidential so we’re not naming individual physicians or facilities. We want it to be a pure quality approach and we’re hoping that the kind of collegial approach that the physicians have experienced in medical school and residency can be replicated here in a manner that the physicians feel is worthwhile.

KM: There aren’t any absolutely clear paths on treatment improvement with this project. We are embracing the principles of continuous quality improvement fully and we are very cognizant, with the combination of statistical analysis and individual treatment patterns, that one might not necessarily reflect what the other is showing. We are erring on the side of further research before we feed back to the group and say, "In general the physicians are practicing thus and such and we would recommend that you would change to something else." Just the sharing of the data, looking at variations, and the physicians knowing where they stand on the curve, I think, is helping in potentially reducing inappropriate variations. The other thing we want to be able to show is that there may be some very good practice patterns out in community settings that do not get highlighted in any literature because they’re not involved in NCI trials. That will show up as a variation, but we still want to be able to highlight that as a best practice. The only way we know that is to wait until we see what the outcomes are.

KM: We provide the blinded reports to the facilities and physicians and we will speak with each of the physicians about what the information seems to show and receive their input. We don’t want to put physicians on the committee in a position where they’re seen as adversarial or any way other than being supportive. We have a strong feeling that, with this group, even though there are different facility affiliations—many competitive with each other—when they get into this room, all the competitive issues are off the table and we’re all there just to focus on patient care. We’re really taking the road of collegial communication and letting the information speak for itself.

KM: No, there’s no penalty. And that really was a big concern at the start of this program. Some thought this was a precursor to narrowing our network, changing fee schedules, etc., and we said, "No, there’s none of that. That is off the table. That is not what this is about." We wouldn’t be investing all of this time and energy, analytic support and software licensing if all we were looking at was reducing our cost. What we want to be able to say is, "We’re getting, in general, the right treatment for the right patients at the right time." We think there will be a natural evolution in this process where, if physicians are not practicing within acceptable standards of care, then that will change over time, but more importantly they’ll be able to benchmark their own patterns with their peers and with national results regardless of whether they’re in a small freestanding facility near Warren, Pa. or in an academic medical center.

KM: I don’t think we even understand the variation well enough at this point. We don’t have any quantitative goals.

KM: We have around 65 physicians and 40 facilities within the 29-county area, enough to put together a good body of knowledge, but not so much as to be overwhelming. The science involved in radiation oncology has been out there for a long period of time and is fairly well-developed. The physicians are very well-trained scientifically and tend to be more computer-literate because the equipment is computer-based. And the data we looked at showed tremendous variation in radiation therapy that we just couldn’t understand. The positive results of the pilot project in Detroit using the ISH software, we felt, is something that would allow Highmark to play a role within the medical community that was consistent with our role as an insurer but did not stand in the way of the physician practicing medicine or adding a lot of bureaucracy to their day-to-day practice.

KM: It’s been offered to all. The only major facilities that aren’t using it are Altoona Hospital, which has chosen to provide data in an independent format to us, and I’m still at the final stages of my negotiations with UPMC. They stated an intent to participate, but we haven’t ironed out all the issues.

KM: Yes. We’re looking at expanding it within cancer at this point. We have some preliminary candidates that we think are interested in piloting medical oncology software. We hope to get that off the ground in April. If that goes well, then we’d like to offer the program to the medical oncology network by January 1, 2003.