By Christopher Guadagnino, Ph.D.
In the years since its creation in 1986, the Pennsylvania Health Care Cost Containment Council (PHC4) has received an uneven reception by health care providers, purchasers and consumers, at times evoking distrust, apathy, resentment and resignation, and at other times producing enthusiasm and collaboration.
Ushered in chiefly by efforts of the business and labor community to address spiraling health care costs in the mid-1980s by stimulating competition and holding health care providers publicly accountable for their quality and costs, it sparked a wave of skepticism and concern among physicians and hospitals that the publicized cost and quality data would be flawed and subject to misinterpretation.
Ironically, market forces appear to have reversed the polarity of enthusiasm. Businesses and consumers – despite perennial accolades and praise for PHC4’s mission with each published report – have scarcely made use of the data in their purchasing and provider selection decisions over the years, while physicians and hospitals have become the chief users of the Council’s data for quality improvement efforts, with most of their data validity concerns having been satisfied by improvements made to the Council’s severity adjustment methodology. In fact, a consortium of physicians and hospitals in southwestern Pa. has just received national recognition for its clinical quality improvement model, which is built upon its use of PHC4 data and involves the payor and purchaser communities in its model.
This summer, the Council has come up for reauthorization by the Pa. General Assembly and has survived, albeit with key modifications reflecting the changing needs of the commonwealth. Because of a souring economy and increasing financial pressures, Pa. hospitals have successfully argued that they are being harmed by the multi-million-dollar cost burden of providing data to the PHC4. In the wake of its reauthorization, the PHC4 will now be restricted to collecting data comprising no more than 50 percent of hospitals’ discharges and must find a way to reduce hospitals’ data collection and reporting costs by 40 percent within one year.
The Council also emerged from reauthorization with a significantly curtailed ability to report more meaningful health care cost data beyond the hospital charge data it has been reporting. Although the Council had never reported data on actual payments by individual health plans to individual providers, it always had the authority to do so. No longer.
The Council will now have a new role to play, reflecting an increased national and statewide focus on patient safety: it will now be required to report back to hospitals data on their adverse medical events, complications and infection rates. In doing so, it will depart from its public report card function and report the data confidentially to each institution. The Council will also contain a new member with expertise specifically on issues of patient safety.
Recent inflation of health care costs may be reanimating a fresh interest by health care payors and purchasers in cost containment initiatives involving PHC4 data. Some initiatives have already been launched, with mixed success, and others may follow.
Economic Urgency and Data Problems
The PHC4 was formed as an independent state agency in 1986 under Act 89 in order to address rapidly growing health care costs following a deep national recession. The key drivers of health care cost containment initiatives at the time were large companies such as Pittsburgh’s Alcoa, which was a member of Buy Right, a national consortium of major companies that hoped to use their collective purchasing clout to obtain volume discounts in health insurance and to steer patients to lower cost providers who offered high quality, says Floyd Warner, president of the PA Chamber of Business and Industry, and the PHC4’s first chairperson.
The PHC4 represented a free market strategy to contain costs, rather than a regulatory one, and was intended to stimulate competition in the health care market by giving comparative information about the most efficient and effective health care providers to individual consumers and group purchasers of health services, and offering information to health care providers that they can use to identify opportunities to contain costs and improve the quality of care they deliver. Act 89 specifically assigned the Council three primary responsibilities:
• To collect, analyze and publicize data about the cost and quality of health care in Pa.
• To review and make recommendations about proposed or existing mandated health insurance benefits upon request of the legislative or executive branches of the Commonwealth.
• To study, upon request, the issue of access to care for those Pennsylvanians who are uninsured.
The Council’s primary output has been its published reports on health care quality and charge data, while it also reviews proposed mandated health benefits when requested by the Secretary of Health or committee chairs in the Pa. General Assembly. The third activity appears to have atrophied.
The Council’s early years were mired in controversy, as critics charged that it had mishandled funds, issued misleading information and ignored the needs of the medically indigent. A bill to disband the Council had been introduced in 1990 by then-Pa. Senator Richard Tilghman. The Council’s first hospital effectiveness report, released in June 1989, was heavily criticized by the Hospital Association of Pennsylvania (HAP) as using software that inappropriately flagged certain cases without offering sufficient analysis to determine if quality of care was deficient in those cases. Although HAP continued to give the Council qualified support, it continued to criticize the validity of the Council’s data analysis software and complained about the administrative cost of providing outcomes data to an external agency, diverting money away from patient care services.
The Pennsylvania Medical Society (PMS) took a position of non-support of PHC4’s activities, believing that it fell short of providing valid, meaningful and useful data. Joseph Danyo, M.D., president of the PMS in 1990, warned that the medical community, put on the defensive by publicized report cards of questionable validity, might practice defensive medicine, which would raise costs. Gordon MacLeod, M.D., PMS president in 1991, further warned that decision-making by the business community, which he believed emphasized cost containment over quality of care, could adversely affect care.
The Council’s executive director at the time, Ernest Sessa, defended the Council by noting that it had been wrongfully perceived as a threat to the medical profession, and stressed that publication of outcomes and charge data was not intended to put medical providers on the defensive, but to make consumers more educated about purchasing health care.
Economic Urgency Fades While Data Improves
PHC4 data fell into relative disuse by purchasers during the mid 1990’s as incentives to scrutinize health care purchasing choices would fade with a booming national economy. According to Warner, improvements in the economy were so good during this time that the business community shifted its attention away from cost containment initiatives and focused instead on their core business activities.
The rate of health care cost inflation relative to general inflation was flat. “When that happens,” says Warner, “businesses virtually forget about health care” and absorb medical cost as a price of doing business. Few in the business and labor community had much incentive to manage or use PHC4 data, Warner adds. Buy Right disbanded.
The provider community continued to criticize the validity of the Council’s published outcome data and injected their objections into the Council’s legislative reauthorization hearings. Changes made to the Council upon its reauthorization in 1993 – including the addition of a Technical Advisory Committee to improve the validity of the Council’s risk-adjusted hospital mortality data – began to ease providers’ concerns, according to Bernie Lynch, PMS’ senior director of practice economics. If providers were to be held publicly accountable for their outcomes data, the Council was now to be held accountable to make those data more clinically accurate, and comparisons among providers more fair.
Further easing provider defensiveness, perhaps, was the benign neglect on the part of the public and business community, of PHC4 data.
Urgency Returns But Wherewithal Fades
A deteriorating economy and escalating health care costs since the late 1990s have brought cost containment concerns to the forefront once again, but may also, ironically, have curtailed the use of the very data that is designed to bring health care costs down. According to Warner, it is now too costly for many Pa. businesses to analyze and implement PHC4 data in their health care purchasing decisions.
With an economic downturn and the rate of health care premium increases having hovered in the upper teens in recent years – reaching as much as 29 percent this year, Pa. businesses have been faced with a dilemma, says Warner: they have a database in the PHC4 that he says is unparalleled in the country, but have largely ignored the data for at least two reasons. Major corporations, historically, have eliminated their health care experts when downsizing, leaving them little expertise to manage and respond to health care cost and quality data. Additionally, steering employees to selected providers, based on cost-to-quality ratios, carries a legal risk and a potentially costly management of employee resistance.
Between 1999 and 2000, the PA Chamber of Business and Industry ran a series of conferences about the how health care premiums were increasing at two- to three-times the rate of inflation, and strategies to deal with that trend. “Nobody came,” says Warner, illustrating the dilemma. To date, he maintains, a report card mechanism featuring medical cost and quality has not driven health care purchasing decisions in Pa.
While systematic use of PHC4 data by purchasers has not materialized in the manner originally intended, businesses do continue to support the Council’s reports. Warner co-authored a preface to the PHC4’s 2002 Annual Report with the president of the PA AFL-CIO, indicating that the business and organized labor communities “are strong supporters of PHC4 and the many beneficial endeavors that the agency has undertaken,” and that they “look forward to working with the Council to continue releasing groundbreaking and internationally renowned reports that allow our members and all citizens to make informed choices about health care.” But the preface itself acknowledges that “we believe that PHC4 represents one of the Commonwealth’s most underutilized tools for economic development.”
Warner says that today’s financial pressures on companies mirror conditions that led to the PHC4’s creation in the mid-1980s. Companies facing thin margins, sapped so greatly by health care cost increases, are reaching the point at which they can no longer pass premium cost increases on to employees and will have to look elsewhere to trim costs. Warner believes that the cost-benefit equation is changing enough for companies once more to justify organizing their health care purchasing decisions around data such as PHC4’s cost and quality data, which he maintains is sufficient to make sound purchasing decisions. With a large proportion of Pa. workers choosing to reside in Pa. after retirement, says Warner, companies may find it worthwhile to relook at those decisions just from the standpoint of retirement plan benefits.
One employer coalition did spearhead an attempt to use PHC4 data to steer patients to the best providers in its region. The Lehigh Valley Business Conference on Healthcare (LVBCH), representing 65 employers in eastern Pa., had begun in 2000 to study data on open heart surgery volume and quality at the five facilities offering the procedure in its region, with the intention of offering Aetna enrollees full coverage for non-emergency open heart surgery at preferred facilities, and charging deductibles and copays to Aetna enrollees choosing to go to other facilities, says LVBCH President Catherine A. Gallagher.
The project, which would have been expanded to other procedures if successful, never came to fruition, as Lehigh Valley Hospital had dropped its Aetna contract and forced two large employer members of LVBCH, including Bethlehem Steel, to drop out of the initiative. Further, point-of-service plans became less popular and gave way to preferred provider organization plans, which Gallagher says complicates the financial incentive mechanism and makes it more difficult to steer patients.
In addition to having limited practical impact on purchaser decision-making, PHC4 data never did a great job of infiltrating the consumer marketplace. Although the PHC4 website gets tens of thousands of hits, and consumers do look at the outliers on the provider reports, there is no evidence on whether or not they follow up on that information, according to David Nash, M.D., chair of PHC4’s Technical Advisory Group and associate dean and director of Thomas Jefferson University’s Office of Health Policy and Clinical Outcomes.
Patients most likely choose their heart surgeons and hospitals primarily on the basis of referrals by their cardiologist, their history with specialists or by recommendations by their friends and neighbors, believes Cliff Shannon, a member of the PHC4 and president of SMC Business Councils, a coalition representing 5,000 small companies in southwestern Pa.
When employers paid the premiums, only they had an interest in costs, says Warner. However, with copays and deductibles increasing, the patient becomes a consumer again, and has a greater incentive to look at the data, says Warner.
The deteriorating economy, escalating health care costs and reimbursement cuts have intensified financial pressure on Pa.’s hospitals, making data collection activities of the PHC4 too costly, in their view. A spotlight on the cost burden to hospitals of collecting and reporting mandated data to the PHC4 split the medical community’s position on how much data the Council should continue to collect, and made the Council vulnerable to political lobbying efforts by the hospital community to reduce the mandated quantity of data reporting. At the same time, recently renewed efforts by the Council to collect actual payor information beyond mere hospital charge data – which has proxied in the Council’s reports for the amounts insurers actually pay for medical services – energized lobbying efforts by the health insurance community to shield that information from public scrutiny.
The Council almost didn’t survive the political battle, says Warner, and the business community agreed to accept the resulting political compromise, although it would like to have seen the Council’s abilities remain unencumbered. Lobbying efforts instead succeeded in passing amendments to restrict what data the Council could collect and report.
HAP supported reauthorization of the PHC4, acknowledging that academic medical centers use PHC4 data extensively in their research and other hospitals use the data to benchmark internal performance improvement activities. But for cost reasons, HAP wanted to eliminate the requirement to submit to PHC4 a separate clinical data stream in addition to patient chart data, arguing that the majority of states that publicly report hospital data rely on the patient chart data with no significant loss in predictive power of severity adjustment.
Since the Council’s inception, Pa. hospitals have been required to submit the two data streams to the Council: UB-92 patient medical record data, plus clinical data specially abstracted for the Council’s MediQual risk adjustment software, according to Paula Bussard, HAP’s senior vice president, policy and regulatory services. HAP complained at the Council’s reauthorization hearings that the average cost per Pa. hospital of data reporting for MediQual averages $152,000 or $20 per discharge for an estimated statewide total cost of $26 million – money that HAP says represents scarce resources being diverted from patient care. At a time when 41 percent of Pa. hospitals and health systems are experiencing negative total margins, says Bussard, all efforts at cost containment are prudent.
The physician community was strongly in favor of retaining the MediQual data reporting, according to PHC4 Executive Director Marc P. Volavka, and fought to preserve it.
The compromise embodied in the Act 14, the reauthorization legislation signed by Gov. Rendell on July 17 this year, retained the MediQual clinical data submission but reduced the quantity required to be submitted to 35 diseases, procedures or medical conditions, representing less than half of hospitals’ discharges, and requiring the Council to find a way to reduce hospitals’ data abstraction costs by 40 percent within one year. Bussard says the change should reduce the data abstraction burden on hospitals by an average of 33 percent, while the narrower data set still includes the chief procedures and medical conditions that PHC4 had released reports about.
Volavka notes that a significant number of hospitals continue to abstract the full data set for their own internal quality review and performance activities.
The Council will also include one new member each from the physician and hospital communities, for a total of two each, with one hospital member representing the interests of rural hospitals.
The Council emerged from reauthorization with another significant limitation: on its ability to report what health plans actually pay for medical care. Although it has for years been publishing data on hospital charges, the Council has always had statutory authority to collect and publish data from insurance companies on payments to providers. One of Warner’s frustrations is that the Council never succeeded in securing that data, as early efforts met with obstacles ranging from the absence of a uniform payment reporting system across health plans to outright refusal by health plans to provide data, says Warner.
About three years ago, the Council created a Payor Advisory Group which has been attempting to work through compliance issues in a systematic attempt to collect payment information from health plans, says Volavka. Now, an amendment under Act 14 prohibits the Council from publishing payment data that is specific to individual health plans or individual providers, although statewide aggregate reporting of payment data is permitted.
Beginning next year, the Council will begin to collect data on health plan administrative costs and payments for hospitalization and selected ambulatory surgery procedures, says Volavka, who believes the inability for the Council to disclose payment data specific to individual entities may ease the reluctance of health plans to provide the data.
Reauthorization has also brought a new focus and new duties to the Council. As a national culture of patient safety and medical error reduction continues to take root in Pa., the PHC4 has been newly required to report to patient safety committees of individual hospitals data on adverse medical events, complications and hospital-acquired infection rates, according to Ann Torregrossa, senior policy manager of Pa.’s Office of Health Care Reform (OHCR), which lobbied for the amendment.
The amendment marks a departure from the PHC4’s traditional model of promoting public accountability among providers because the data will not be made public, in order to minimize the threat of lawsuits and prevent a culture of blame, says Torregrossa.
Based on PHC4 data requested by the OHCR, she says, there were 5,294 adverse medical events, 88,000 complications and over 250,000 hospital-acquired infections reported at Pa. hospitals and ambulatory care centers in 2001, which are estimated to have added over $2 billion to medical costs.
Because many hospitals had already been purchasing PHC4 data on these three categories – which the Council abstracts from its existing collected data – the OHCR thought it important for the Council to routinely provide such information to all hospitals without charge, says Torregrossa. In early October, Pa. hospitals will receive from the PHC4 baseline data from 2002 on the three data categories, after which they will receive ongoing reports.
This activity will parallel separate activities of the newly created Patient Safety Authority, which has just contracted with ECRI, a nonprofit health research agency in Montgomery County, to create a Web-based reporting system to collect from Pa. hospitals – and feed back to them on a confidential basis for internal improvement efforts – information on serious events, near misses and infrastructure failure leading to medical error, Torregrossa says, noting that these represent new data categories not collected by the PHC4 or obtainable from its data.
The OHCR has also added a new member to the PHC4, a Governor appointee required to have expertise in the application of continuous quality improvement methods in hospitals – specific expertise which the OHCR did not believe currently exists on the Council, says Torregrossa.
Medical Community Becomes Primary User
A rocky reauthorization notwithstanding, the PHC4’s mission to hold providers accountable to cost and quality has come full circle, although perhaps not by the market mechanism that was intended. While the PHC4 was created to provide medical cost and quality information to purchasers and consumers whose decision-making would spur providers to improve quality, the principal users of PHC4 data have become the providers themselves, looking to link quality improvement to reimbursement enhancement.
Trepidation over what purchasers might do with PHC4 data, whether eased by improvements the Council has made to the data or by lack of its use by the purchasers, has largely given way to provider-driven quality improvement initiatives using that very data.
The most arduous users of PHC4 data are hospitals, which use it for internal evaluation and quality improvement, according to Nash.
Physicians are also beginning to find uses for the data. In an endorsement appearing in the Council’s 2002 Annual Report, PMS Executive Vice President Roger Mecum praised the Council for offering one of the best models for collaborating with physicians to advance the delivery of high quality medical care.
“One of the most constructive collaborative efforts the Council has ever been involved in,” says Volavka, is the Pittsburgh Regional Healthcare Initiative (PRHI), a consortium of 44 hospitals; hundreds of physicians; and the business, labor and health insurance communities in southwestern Pa. The PRHI is developing outcome benchmarking studies for five clinical areas and has used special PHC4 reports on cardiac care, obstetrics care, diabetes, orthopedic surgery and depression to drive quality improvement efforts. The PRHI has recently received national recognition for its quality improvement model in a recent issue of Health Affairs.
Guided by the principle that overcoming competitive barriers and sharing outcomes data can form the basis for rewarding higher quality of care, the PRHI is adding process of care details to PHC4 outcome data to develop a sophisticated care tracking database for every patient in the region. The goal is to help clinicians identify clinically relevant issues that lead to reduced variation and best outcomes, and to help them share that information on a quarterly basis to drive improvement across the region, says PRHI Director Ken Segel.
The mandatory data collection activities of PHC4 were a necessary catalyst of the effort. “If we had to ask for voluntary participation to get started on data submission, we would never have gotten started. PRHI would not exist in its present form if we hadn’t had the PHC4’s existing data source to bring people to the table, and doing that in a way that wasn’t immediately threatening to caregivers,” says Segel.
By sharing a refined database of clinical information across physicians at 44 hospitals in the region, the PRHI hopes to engineer reward structures whereby providers ultimately are offered bonus incentives for care that is clinically superior to elsewhere, says Segel.
Collaboration with PRHI has also led to refinements in the PHC4’s data reports, such as the addition of readmission analysis in its heart surgery reports, and the expected release of a new, physician-specific orthopedic report, which Volavka says may be published by mid-2004.
Why no consortium like the PRHI has yet materialized elsewhere in Pa. remains a mystery. The Pittsburgh region had powerful support for the idea in the personage of Alcoa’s Paul O’Neill, says Nash, while others speculate that it was Pittsburgh’s long history of cooperation within its business community that allowed for unified support of the initiative. The Jewish Healthcare Foundation of Pittsburgh had also devoted resources and expertise to spur the initiative’s development.
In the spring, the OHCR plans to co-sponsor with HAP a summit on best practice sharing which will highlight the activities of the PRHI, says Torregrossa, who hopes the Patient Safety Authority and individual hospitals’ patient safety committees will foster such an approach.
A payment-for-performance model may emerge in southwestern Pa. for diabetes management, according to Shannon. At a recent stakeholder meeting, the PRHI presented PHC4 data showing a huge variation in the way that care and disease management is provided to diabetics in western Pa. Initiative participants hope within one year to set up a platform whereby all health insurers, laboratories and physicians agree to standard formats for reporting test results of diabetic patients, and physicians agree to receive telephone or email prompts from health insurers to remind diabetic patients to make scheduled physician office visits, as well as to remind physicians to make timely reviews of testing results, says Shannon. Health insurers would agree to reimburse physicians for additional time interacting with patients outside of office visits, while purchasers would agree to foot some of the bill for at-home diabetic management equipment and supplies for patients, Shannon adds.
Shannon predicts that the Centers for Medicare & Medicaid Services will expand its payment-for-performance model beyond demonstration projects, and that employers, hospitals and health plans will be given new incentives to work on such issues. Outcomes data such as PHC4’s could play a central role in the model.
Independence Blue Cross (IBC) in southeastern Pa. has already begun using PHC4 data as part of its Quality Incentive Payment System for hospitals, under which hospitals receive annual reimbursement increases based in part on their performance against agreed-upon quality parameters, including PHC4’s hospital data on risk-adjusted mortality and readmissions, as well as JCAHO and Leapfrog Group data.
According to IBC, four hospital systems have agreed to participate in the program, while some hospitals have declined.